Singapore Comparative Law Review 2019 (SCLR 2019) by The UKSLSS

Patron’s Foreword It is again my pleasure to write the foreword to the second issue of the Singapore Comparative Law Review under its revamped title. This issue contains a rich collection of 17 articles, 7 commentaries and 5 case notes. A novel and useful feature of this issue is that some topics are covered by two opposing articles. The Editorial Committee led by Edwin Teong Ying Keat of Bristol University is to be commended for bringing out this issue. In the past few years, the Singapore courts have delivered judgments of various issues of private law that depart from English decisions and if this trend continues, it will provide a regular source of judgments for comparative study with English judgments. The articles, commentaries and case notes on comparative law between English and Singapore law are reviewed very briefly below. I. ARTICLES This year’s issue starts off with “Should common law jurisdictions adopt the doctrine of unconscionability?” by Adelle Yii of University College London. Recently, in BOM v BOK1 (“BOK”), the Singapore Court of Appeal upheld the narrow doctrine of unconscionability and set aside a deed of trust. The Court applied the established test in Fry v Lane2 and Cresswell v Potter3:- if (1) the purchase was made from ‘a poor and ignorant man’, (2) at a considerable undervalue and (3) with no independent legal advice. Where the test above is satisfied, the burden of proof shifts to the Respondent to show that the purchase was ‘fair, just and reasonable’. In BOK, the Court of Appeal added to the first requirement the donor’s “infirmity” of mind or decision making due to physical or mental distress or an emotional reaction due to such distress. By adding this factor, the Court may be said to have broadened the requisites of the doctrine. The Court rejected the supposed broad doctrine adopted by the High Court of Australia decision of The Commercial Bank of Australia Limited v Amadio4 (“Amadio”) on the ground that the test was overbroad and subjective, leading to uncertainty in contractual relations. The test in Amadio is that a dealing is unconscionable where (i) a party to a transaction was under a special disability in dealing with the other party with the consequence that there was an absence of any reasonable degree of equality between them and (ii) that disability was sufficiently evident to the stronger party to make it prima facie unfair or ‘unconscientious’ that he procure, or accept, the weaker party’s assent to the impugned transaction in the circumstances in which he procured or accepted it. Where such circumstances are shown to have existed, an onus is cast upon the stronger party to show that the transaction was fair, just and reasonable. Adelle argues that Australia and other common jurisdictions should consider adopting the narrow doctrine of unconscionability to protect contractual certainty. The difference between the narrow and 1 2 3 4

[2018] SGCA 83 [1888] 40 Ch D 312 (Ch) [1978] 1 WLR 255 [1983] 151 CLR 447

the broad doctrine is that the former specifies the disabilities of the plaintiff whereas the latter refers to “a special disability” to allow the court to expand the categories in appropriate circumstances. The Australian test has the advantage of flexibility. While unequal bargaining power is not, per se, a reason to set aside a commercial transaction, it is clearly a factor the courts must pay special attention to ensure that the weak has not been taken advantage of by the strong, as is human nature when it comes to matters of money. In “You’ve got mail!: Examining the postal rule in the technological age”, Pang Wenn Ler, University of Bristol discusses the established postal rule that a contract is formed when the offeree posts his letter of acceptance. However, in Entores v Miles Far East Co,5 Denning LJ held that the postal rule would not apply to instantaneous communications such as a telex machine. However, the telex machine has gone the way of the dodo. The fax machine is about to suffer the same fate. Today, most communications in commercial transactions are by email, and where email is stipulated in the offer as the means of acceptance, the recipient rule should apply. Acceptance takes place when the email has not been rejected by the server. When it has not been rejected by the server, the email would have been deposited in the offeror’s inbox. Whether he retrieves it or not is irrelevant. One of the few cases on electronic contract formation in Singapore was Chwee Kin Keong v Digilandmall. com Pte Ltd6 that was primarily about unilateral mistake. Nevertheless, Rajah JC’s obiter provided careful analysis about the postal rule in a digital age and its use in our local context. The obiter is by no means a precedent for what rule is to be used. In “Quenchless feuds and ill will – should the tort of malicious prosecution of civil proceedings be recognised in Singapore?”, James Dylan Toh, University of Bristol discusses the decision of the Singapore Court of Appeal in Lee Tat Development7 which contrary to the the U.K. Supreme Court decision in Willers v Joyce,8 declined to extend the tort of malicious prosecution to civil proceedings on the ground of “essential differences” between the “character and consequences” of the criminal and civil processes. It was held that the character of criminal prosecution punishes a public wrong, whereas a civil claim vindicates a private right. There are of course strong policy reasons on both sides. However, James is of the view that on balance, the tort should be recognised on the principle ‘ubi jus ibi remedium (where there is a wrong, there must be a remedy)’. The tort is recognised by the courts in Hong Kong, Australia and New Zealand. Hence, the Singapore decision is an outlier in the major common law jurisdictions. James also points out that the decision in Lee Tat Development was unnecessary because the facts of the case could not sustain such an action even if there were such a tort in Singapore. In “In Light Of The Technical Impracticality Of The Right To Erasure, What Answers Can ActorNetwork Theories Provide?”, Justin Kwik, University of Bristol discusses the right to be forgotten or the right to erasure under English law which directly applies European Union Regulations (GDPR Article 17). Justin argues that Article 17 is ill adapted to robustly protect an individual’s right to be forgotten in an Artificial Intelligence era. This essay is highly technical and is difficult to understand as it discusses subjects such as data removal in machine learning, data removal in blockchain, cyberregulatory theory, network communitarianism, ANT-Foucauldian power lens. Perhaps, a more interesting issue from a moral or philosophical angle is whether any person should have the right, or to what extent, to be forgotten in cases where the person has gone public or become a public figure. 5 6 7 8

[1955] 2 QB 327 [2004] 2 SLR 594 [2018] SGCA 50 [2016] UKSC 43

The larger issue seems to be connected with one’s reputation in this life. In “Abandoning Old Banks and Brooking New Streams: A Case for the Reformation and Statutory Codification of the English Law on A Bank’s Duty of Confidentiality”, Filbert Lam, University of Edinburgh discusses banking confidentiality under the principles laid down in the well-known decision of Tournier9 which held a banker has an implied duty of confidentiality arising from the banker-customer relationship. This duty is subject to four qualifications: (a) Where disclosure is under compulsion by law; (b) where there is a duty to the public to disclose; (c) where the interests of the bank require disclosure; (d) where the disclosure is made by the express or implied consent of the customer. As Tournier is a common law authority, it is always subject to statutory modification for the purpose of combatting financial crime, such as money-laundering via banks. Filbert questions the coherence of Tournier in the light of recent English decisions, which he suggests has revealed important lacunae in the law, and that they support the case for the clarification of this duty in a statute. Singapore did away with Tournier in 1970 when it enacted the Banking Act to create a statutory framework for developing Singapore into a financial centre - s47 provided that “[c]ustomer information shall not … be disclosed … except as expressly provided in this Act”. However, Filbert has misunderstood the Court of Appeal’s decision in Susilawati v American Express Bank Ltd10 that s47 has set out an exclusive regime that prevents parties from entering into express agreements to regulate banking confidentiality on a contractual basis. In “Divergent Paths: The Doctrine on Equivalents in Singapore and the United Kingdom”, Joel Sherard Chow, King’s College London discusses the difference between Singapore law and UK law following Lee Tat Cheng v Maka GPS Technologies Pte Ltd11 where the Singapore Court of Appeal favoured a purposive approach and rejected the decision of the UK Supreme Court in Actavis UK Ltd v Eli Lilly12 which adopted the doctrine of equivalents. In evaluating the delicate balance between ensuring adequate patentee protection and certainty for third parties in constructing claims, Joel appears to favour the equivalents doctrine as it would encourage innovation, research and development, particularly in developing areas of pharmaceutical products and technology and would not conflict Singapore’s statutory regime. We will have to wait and see whether the Singapore Parliament agrees with Joel. In “Have the Rules Governing the Interpretation of Pari Passu Clauses in Sovereign Bonds Developed in Equal Step Under English and New York Law?”, Filbert Lam, University of Edinburgh discusses the different approaches of English courts and New York courts in interpreting pari passu clauses in sovereign bonds. Filbert says that until the 1990s, the pari passu clause provided for equal ranking of holders’ unsecured debt with other external debt of the same nature (“equal ranking simpliciter” interpretation), stemming from the “equal and fair” treatment assurances which governments provided to foreign creditors. However, since the early 1990s, some clauses contained an additional modifier which prima facie contained an equal ranking in the creditors’ right of payment (“equal payments” interpretation). This multiplicity of interpretations has introduced uncertainty in the sovereign bond markets and increased transactional costs and default risks for states. The US Second Circuit Court of Appeals held, in NML v Argentina (“NML”),13 if a pari passu clause contained both the equal ranking and payments elements, they conferred separate but related rights on the bondholders. Filbert concludes from his analysis of how the NY courts and the English courts view 9 10 11 12 13

[1924] 1 KB 461 [2009] 2 SLR(R) 737 [2018] 1 SLR 856 [2017] UKSC 48 699 F3d 246 (2d Cir 2012)

“boilerplate” clauses, that English courts would have arrived at the same decision as NML, albeit for different reasons. Filbert believes that in any event, there are solutions available to sovereign issuers to circumnavigate the legal effects of NML. This is a thoroughly argued and very thoughtful article. However, given the subject matter, the article deserves a different and wider readership, perhaps in some international business law journal. Joel Sherard Chow, King’s College London dives into the murky waters of the 1962 Water Agreement between the State of Johor and the City Council of the State of Singapore to see whether it can be seen as a treaty under international law. As neither Singapore nor Johor was a “State” in international law, the 1962 Agreement would not qualify as a treaty. However, Joel argues that since both Singapore and Malaysia succeeded to the obligations of the original parties, they are likely to be considered “States” under the Vienna Convention. However, the water gets murky as to the second requirement that a treaty must be governed by international law, and the 1962 Agreement is governed by the laws of Malaya. Joel argues that the 1962 Agreement gives rise to international obligations by reason of the conduct of parties in 1965, and also the decision of the PCA in the Chagos Arbitration14 could be relied on to argue that obligations under the 1962 Agreement were likely “elevated” to the international plane, purely by virtue of Singapore’s independence from Malaysia in 1965. This is an interesting issue that awaits for the parties to make an issue of it. In “Paternalism v Autonomy: Examining the moral concerns of criminalizing Suicide and Physician Assisted Suicide”, Carol Yin, King’s College London applauds the recent decriminalisation of attempted suicide in Singapore under the Law Reform Act 2019. She supports the amendment for its more humanitarian approach to suicide (and compares it with the position in UK where it was decriminalised in 1961). Carol however goes further to advocate physician assisted suicide should be allowed because “patients remain in control over their lives in PAS and their intent should justify the conduct of PAS.” However, she does not mention euthanasia in her proposal. There are two essays on the loss of a chance of having a longer life. In “Why Carol Ann Armstrong does not introduce uncertainty: a comparative analysis”, Tan Yin Theng Sarah, University of Bristol, sees some merit in the High Court decision in Carol Ann Armstrong.15 In 2009, THW, a doctor, misdiagnosed an ulcerous skin lesion of PT (the deceased) as non-malignant. Three years later, PT developed a swelling under his armpit which was diagnosed as cancerous and which had metastasised. PT died eighteen months later, and his widow, CAA, brought an action against QL and THW for breach of duty. The Judge found THW had breached his duty, and that there was a loss of an early opportunity for treatment. The issue was whether THW’s breach of duty was an effective cause of PT’s death. CAA’s expert witness relied on US statistics to show that PT had a 10-year survival of at least 68%, and closer to 80% had the melanoma been diagnosed in 2009. THW’s expert witness Prof Chia opined that it was less than 50%. The Judge accepted the assessment of CAA’s expert.16 14 [2015] XXXI UNRIAA 359 15 [2018] SGHC 66 16 Carol Ann Armstrong [2018] SGHC 66 at [19] “…I will accept [CAA’s expert’s] assessment, but with some caution, and hold that [THW’s] negligence had caused [PT] to lose a fighting chance, and also probably caused him to die years earlier than he would have done. But I am not accepting that therefore [PT] had a 68% chance of surviving 10 years. Percentages and data are less helpful when we consider the facts that we do know – that [PT] survived almost four years even though he was misdiagnosed; that though it is possible for his cancer to be dormant, the evidence suggests more likely that it was not. It was growing and developing but only in the lymph nodes under the armpits. And I therefore leap where Lord Nicholls had leapt, and estimate that [PT] might have lived twice that number had he been properly diagnosed. Eight years from 2009 are what I think is as fair an estimate as anyone can hope, given what we know, and not plucking percentages from the myriad unnamed cases spewed from the data. Assertions such as ‘68% chance

The Judge’s decision was problematic. There was obviously an evidentiary gap in causation that the Judge ignored. His approach directly contradicted the decision of the Supreme Court (3:2 majority) in Gregg v Scott17 that rejected any claims for damages for negligence unless the claimant proves on the balance of probabilities of 51% that the negligent act was the cause of the death. The Judge preferred the minority decision in Gregg v Scott. Sarah supports the Judge’s approach in departing from the strict rule in Gregg v Scott, as it establishes a more flexible rule that allows justice to be done in the circumstances of the case. There is always uncertainty in assessments of this nature. Who can really tell with any degree of confidence whether the patient had a less than 50% or 50% chance of surviving for any number of years due to a loss of early treatment arising from a misdiagnosis? In “On the basis of fairness? Taking a chance on recognising lost chances in Singapore: a comparative analysis”, Lau Yi Hang, Singapore Management University argues that the High Court judgment has profoundly changed the law of negligence by recognising loss of chance as an actionable head of damage in the context of physical harm, where there is an evidentiary gap between the negligent act and the cause of death. Lau disapproves of the judgment as policy-driven, that policy is an unruly horse, and that the Judge’s policy reason for doing justice is insufficient to persuade convincingly why loss of chance should be recognised. The decision does not explain why fairness is a compelling reason to circumvent causation and the recognition of lost chance as actionable damage. The introduction of loss of chance claims in physical injury may result in greater uncertainty in the realm of medical practice and negligence law whilst engendering increased medical litigation and defensive medical practice that is antithetical to doctor-patient relationships. Lau proposes that “rather than analysing the recognition of lost chance claims from a perspective of causation, perhaps future courts can address whether lost chance claims can be, and ought to be recognised as a new head of actionable damage. After all, as Baroness Hale in Gregg and Justice of Appeal Andrew Phang in ACB v Thomson Medical18 rightly pointed out, damage is the gist of negligence.” There appears to be two problems of proof in this case. The first is whether THW’s misdiagnosis caused PT’s death on a balance of probabilities of 51%. The second is, if it did, whether it caused PT to lose 8 years of his life, and not 5-10 years. This decision has gone on appeal. We can expect a substantial judgment from the Court of Appeal on the fundamental issues of causation and loss of chance. There are two articles on the subject of Deferred Prosecution Agreements in UK and Singapore. In “The United Kingdom’s deferred prosecution agreement regime five years on: is it an effective tool in addressing economic crime perpetrated by companies?”, Wee Toh Loo, University of Bristol examines the UK DPA regime after 5 years of operation, and concludes that it is unlikely to be an effective tool in addressing economic crime perpetrated by companies for two main reasons. First, the threat of prosecution, is at present, weak. The second is the regime’s ability to change corporate culture through effecting genuine improvements in compliance practice is hamstrung by legitimacy issues arising from its structural focus on companies and other similar organisations. In “Evaluating Deferred Prosecution Agreements in the Context of Singapore”, Tan Yann Xu, National University of Singapore, analyses the Singapore regime in detail and concludes that while the of survival for 10 years’ are intended for doctors to advise their patients. 17 [2005] 2 WLR 268 18 [2017] SLR 918

Singapore DPA regime has many strengths and rests on firm theoretical foundations, its weakness is the level of transparency in the decision making process as to when a DPA will be entered into. There is complete opacity in this area, and this may not only facilitate abuse of prosecutorial discretion, but also the court’s oversight function because of inadequate guidance regarding the factors which the Prosecution must consider. This is early days yet. How the DPA regime will work out depends on why it was established in the first place. There are two articles on the best interests of dying patients/(legally incompetent) children. In “Nothing gold can stay: No better ‘substitute judgment’ against the ‘best interests’ of dying patients”, Edwin Teong Ying Keat, University of Bristol, makes a strong argument from UK experience derived from English court decisions that the courts should be allowed to decide in the best interests of dying patients for three reasons: (a) they have developed the ‘best interests’ test, and have the institutional capacity to apply it; (b) being objective and neutral, they are best placed to carry out the delicate balancing exercise between autonomous wishes and paternalistic concerns, and also the concerns between laymen and hospitals; (c) the alternative substituted judgment test (which empowers family members or doctors to decide what is in the best interests of patients) which is based on the intention of the patient if he were mentally capable of making the judgment. In “Not my decision to make: The courts, best interests, and legally incompetent children”, Soh Kian Peng, Singapore Management University’s School of Law argues that mediation is a better approach than a court decision because (a) the “best interest” test is doctrinally problematic, and (b) the fluidity of the factual matrix adds to the difficulty of applying the “best interest” test. The fact is Singapore’s courts have no experience in making “life and death” decisions for dying patients who are mentally incapacitated (temporarily or otherwise). The reason is that physicianassisted suicide and euthanasia amount to homicide under the Penal Code. Even as regards cases where the life support system is withdrawn or withheld, someone has to make the decision. And if he is not lawfully authorised to make the decision, he would be assisting in the homicide of the patient to the extent of hastening his death. Under current Singapore law, termination of life support may be authorised by any patient under an Advance Medical Directive (AMD). Legislation is required to empower judges to make such kinds of decisions. Soh also discusses the interesting case of Re LP19 (adult patient: medical treatment) where Mount Alvernia Hospital made an application for its doctors to amputate the legs of a patient to save her life. The doctors were unable to obtain the consent of FL as she had fallen into a coma after requesting her doctors to “save her legs at all costs”, and before the doctors could advise her that if amputation was not done she would soon die. Clinically, or intellectually, the decision was not really difficult. It was deciding between whether to allow LP to die or chance to live (she might die during or after the operation), albeit without legs, and even if there was no certainty that she would recover consciousness (of which the doctors could not say). The Judge decided that it was in LP’s best interests that the doctors be allowed to proceed with the amputation. However, the Judge did not articulate the legal basis on which he granted the order. There is no statutory provision which gave the court such power. The court was asked to approve an operation to save LP and not to kill her. It was a case of necessity because no one could give consent on her behalf. In Re F (Mental Patient: Sterilisation)20, the House of Lords extended the legal principle of ‘necessity’ 19 20

[2006] 2 SLR 13 [1990] 2 AC 1

to permit the sterilisation of a 36 year old woman who was suffering from a serious mental disability. The doctors could have proceeded with the operation without a declaration from the court because of Section 92 of the Penal Code.21 The Judge said that there was some concern about litigation, which could only mean that if the doctors botched the operation, and she died. But, that should be an illegitimate concern. If it were otherwise, doctors might as well not treat patients at all. In “Lift not the Painted Veil”, Professor Walter Woon, Dean, Singapore Institute of Legal Education provides a clear guide to practitioners as to the situations when it is necessary or not necessary to invite the court to lift the corporate veil. Corporate personality is intended to shield the incorporators as investors from unlimited personal liability for the debts of the company. Unfortunately, crooked controllers enrich themselves using the resources of the company, and use its separate personality to shield themselves from personal liability. Lifting of the corporate veil is necessary to make such controllers disgorge assets fraudulently taken by them or to impute personal liability for such conduct. Professor Woon’s advice is that lawyers should not attempt to lift the corporate veil unless absolutely necessary. This is good advice, although it has to be said that lawyers normally resort to lifting of the veil as a back-up argument. In “Liability for Road Accidents Caused by Driverless Cars”, Professor Ken Oliphant, Head of Law School, University of Bristol, writes on the civil liability of autonomous vehicles (“AVs”), such as driverless cars, for harm caused to other road users. Who should be liable if the AV has no driver? The possible defendants are the manufacturers, the distributor and the owners. How should the law allocate liability that best serves state interests? The Professor discusses the policy considerations, and scheme set out in the UK Automated and Electric Vehicles Act 2018, which provides for no fault liability as the primary liability. Singapore has also reached a stage of social and economic development where the AV is being experimented in controlled environments. Sooner or later, Parliament has to legislate if the Government considers that the much touted economic benefits of using AVs on public roads are worthwhile and realisable, such as in reducing road accident, improving mobility and increasing leisure time. II. COMMENTARIES In “Ready for repeal: Lessons from the Hart-Devlin Debate”, Alexander Yean, University of Oxford, revisits the Hart-Devlin debate and asserts that time has demolished Devlin’s disintegration of society argument and supports Hart’s argument (see Court of Appeal’s judgment in Lim Meng Suang v Attorney General22). Yean argues that “Hart’s lessons for Singapore are simple: first, the argument that the normalisation (or indeed, the mere decriminalisation) of homosexual behaviour will lead to societal disintegration is not only divisive, but erroneous. Second, the moral environment that perpetuates such a notion should, if anything, change rather than resist change. We are quite ready for repeal.” However, the Government is not convinced in relation to Singapore, as yet. In “#Mentoo: a response to the Criminal Law Reform Bill 2019”, Matthew Choo, King’s College London comments on extensive reforms set out in the Criminal Law Reform Bill 2019. His main criticism is that in one particular area, the reform did not go far enough. The definition of ‘rape’ in s375 of the Penal Code was amended to include a new subsection 1A (which was previously the first 21 Penal Code, S92 provides “Nothing is an offence by reason of any harm which it may cause to a person for whose benefit it is done in good faith, even without that person’s consent, if the circumstances are such that it is impossible for that person to signify consent, or if that person is incapable of giving consent, and has no guardian or other person in lawful charge of him from whom it is possible to obtain consent in time for the thing to be done with benefit”. 22 [2015] 1 SLR 26 at [162]-[164]

part of s376(a)(a) in order to provide an ‘appropriate label’ to such acts. However, Matthew argues that female compelled-penetration should be criminalised as rape and not just sexual assault, owing to the equivalence in harm of both the prevailing conception of rape and with compelled-penetration. For this reason, he argues that the second part of s376(1) should have been transferred to s375 by way of amendment. This would make rape “gender-neutral, providing better protection to victims of non-consensual forced penile-vaginal, anal or oral penetration, regardless of gender…and to reflect changing societal norms and views on the roles of men and women.” The criticism is well meaning but it seems to be a dispute on using the correct label for the offence. A Rafflesia smells as bad by any other name. In “Criminal Law (Temporary Provisions) (Amendment) Bill- An unjustified infringement on the right to a fair trial?” Noel Low, University of Oxford, argues that although the Bill was stated to provide a list of offences for which criminals could be detained without trial, here previously there was no list, and also clarify the scope of the Criminal Law (Temporary Provisions) Act (“CLTPA”), and prevent arbitrary use of the power to detain, its actual effect is likely to increase the scope of the power to detain, in light of the context laid down by Tan Seet Eng v Attorney General.23 Noel is concerned that the Fourth Schedule (“participation in, or facilitating, any organised crime activity as defined in the Organised Crime Act 2015”) might have the effect of reversing the decision in Tan Seet Eng, in that the requirement that the activity must pose a threat to the public order within Singapore might no longer hold. Further, concerns abound that the amendment allows detention for activities of a more minor nature, so long as they fall within a crime specified under the Fourth schedule. This would circumvent the second test laid down in Tan Seet Eng, that the criminal activity must be of a sufficiently serious nature to justify detention without trial. Noel is also concerned about the implication of s30(2)24 which provides that the decision of the Minister to make a detention or supervision order “final”, even though the Minister has clarified that the finality clause does not affect the courts’ power of judicial review, based on the traditional tests of illegality, irrationality and procedural impropriety, and that the clause is intended to mean the Minister can “establish the facts of a case, apply those facts to the relevant rules and considerations, and exercise his discretion.” Noel argues that detention without trial “is an unacceptable interference with the right to liberty and fair trial”, and that the problems with the CLTPA can be resolved by adopting the Closed Material Procedure(“CMP”) in the UK. The CMP provides for a closed court hearing where the identity of witnesses who fear reprisal from testifying is not compromised as the defendant is excluded, but who is represented by a special advocate, who is prohibited from disclosing confidential information to the defendant without permission but whose role is to take instructions and defend his interests. The CMP is thus normatively justifiable on principles of constitutional theory and likely to create a fairer process for alleged criminals facing the CLTPA. Noel might wish to bring his proposal to the attention of the Minister in case he is not aware of the CMP. In “True or False or Misleading? “[A]dequate judicial oversight” over Part 3 Directions under the Protection from Online Falsehoods and Manipulation Act (‘POFMA’)”, Lee Kay Howe, University of Oxford comments on the “fake news” law recently passed by Parliament. Like many, Lee agrees that POFMA is desirable because disinformation or ‘fake news’ is an abuse of freedom of speech and a serious threat to democracy and social order. Like many, Lee is also concerned about its ‘chilling’ effect on legitimate free speech and discussion of matters of public interest, including content critical of the government. He is also concerned with the wide powers given to any Minister to issue 23 24

[2015] SGCA 59 Criminal Law (Temporary Provisions) Act, s30(2)

Correction Directions25 and Stop Communication Directions26, to any person as long as the Minister determines, per s. 10(1)(a), that “a false statement of fact… has been or is being communicated in Singapore”, and, per s. 10(1)(b), “is of the opinion that it is in the public interest to issue the Direction.” Lee points out that what is a statement of fact or opinion is difficult to determine in many cases. Similarly for when a fact is false or misleading. The question of who has the burden of proving that an objectionable fact is true is also not clear. If it is only the defendant, it would be unfair on grounds of information asymmetry between the Government and the individual. Lee raises many other legal issues, but space does not allow for a proper assessment of the merits of his arguments. However, he emphasises that adequate judicial oversight is needed to prevent these Directions from being abused to political ends, or to stifle freedom of speech. At this point of time, one can only say that the proof is in the pudding. In “A call for judicial activism in response to current Anti-Terror legislation”, U. Sudharshanraj Naidu, University of Bristol examines the law in Singapore on detention without trial under the Internal Security Act after it was amended in 1988. He advocates that the courts should declare that the restraints imposed by s8 of the ISA 1988 unconstitutional to allow judicial review to be brought to court detainees. The courts can then extend the grounds of review to apply so as to prevent arbitrary detention orders. Except for referring to Lord Hoffman’s powerful dissent in the Belmarsh27case, Naidu does not explain the legal principles on which the courts may declare s8 unconstitutional since it was enacted pursuant to a constitutional amendment. In “Rage against the machines: AI or Nay?” Annabelle Lee, University of Warwick discusses Artifical Intelligence and its benefits to society and to legal practice. She refers to a host of related issues such as ownership, liability, and data discrimination that remains to be solved, and concludes that lawyers should focus on developing frameworks to resolve these issues “rather than fixate on the anxiety surrounding this transitory disruption brought about to the legal industry by AI technology”. Suang Wijaya, Eugene Thuraisingam LLP provides a companion piece to Commentary piece number 4. It focuses on s61 of POFMA.28 Wijaya points out that the Explanatory Statement does not explain (a) the persons or classes of persons to whom s61 is meant to apply, or (b) the purposes that are intended to be achieved by s61. He argues that s61 is unconstitutional on the principle of the separation of powers in the Constitution (see Prabagaran a/l Srivijayan v PP29) if it is intended to allow the Minister to free any person who has committed an offence under the Act. He argues that s61 would constitute a direct or an indirect interference with the judicial power, citing Steve Ferguson v The Attorney General of Trinidad and Tobago.30 Wijaya accepts that legislation may contain exempt provisions, e.g., s29 of the Enlistment Act.31 But he argues that each exemption clause must be scrutinised in light with the entire legislative scheme and context, to ascertain whether the exemption clause: (a) is a legitimate conferral of powers to 25 Protection from Online Falsehoods and Manipulation Act, s11 26 Protection from Online Falsehoods and Manipulation Act, s12 27 A and others v Secretary of State for the Home Department [2004] UKHL 56 (‘Belmarsh’) at [97], “The real threat to the life of the nation, in the sense of a people living in accordance with its traditional laws and political values, comes not from terrorism but from laws such as these” 28 Protection from Online Falsehoods and Manipulation Act, S61 provides “The Minister may, by order in the Gazette, exempt any person or class of persons from any provision from this Act.” 29 [2017] 1 SLR 173 30 [2016] UKPC 2 31 Enlistment Act, S29 provides “[t]he proper authority may by notice exempt any person from all or any part of the liability of that person under this Act”.

make rules of general application in relation to a person’s or class of person’s rights and duties; or (b) is an illegitimate ad hominem legislation targeted at securing a particular outcome for specific persons (or classes of persons). Wijaya has raised an interesting issue as to the extent of the powers of the Minister under s61. Can he exempt any person or class of persons from (a) any or all of the offence-creating provisions of the Act; or (b) all the provisions of the Act? Will any such classification violate Article 12(1)? III. CASE NOTES In “Noor Azlin bte Abdul Rahman v Changi General Hospital Pte Ltd & Ors32”, Suchetra Sivakumar, Queen Mary University of London, finds the Court of Appeal decision notable in imposing a lower standard of care on doctors working in the A&E Department because their working conditions require them to prioritise their treatment due to the high volume of patients they have to attend to. Making quick decisions are a necessity, which can only be achieved by a ‘targeted approach’, which deals with the ‘treatment of the patient’s presenting symptoms and the elimination of life-threatening conditions’. Incidental findings cannot be disregarded, especially when it is grave. Depending on the incidental finding, seriousness, and the patient’s medical history, an A&E doctor should either make a referral to the appropriate department or conduct relevant tests. However, the execution of these orders is reliant on the system of the hospital and not on the particular doctor. Suchetra considers this a landmark decision for A&E doctors and is especially crucial in light of rising numbers of elderly and overcrowding of the A&E in public hospitals. In “Prabagaran a/l Srivijayan v Public Prosecutor33: Approach to Prosecutorial Discretion”, Tan Yan Shen, University of Oxford notes that the Court of Appeal decided in this case that held that s33B(2)(b) and s33B(4) Misuse of Drugs Act both prescribe a “subjective assessment of an objective condition”, i.e., the offender’s assistance had yielded actual results in disrupting drug trafficking activities. This involves a “multi-faceted inquiry” taking into account a “multitude of extra-legal factors such as the upstream and downstream effects of any information provided, the operational value of any information provided to existing intelligence, and the veracity of any information provided when counterchecked against other intelligence sources”. The court further reasoned that considerable deference should be accorded to the Attorney-General as the operational effectiveness of the CNB will be otherwise hampered. To the last reason, Tan argues that the prosecution lay down guidelines as to what “substantive” means, and even if every of the prosecution’s negative certification is challenged, this approach is consistent with current treatment of death penalty cases. Under s313 of the Penal Code, persons sentenced to the death penalty by the trial judge has a right of appeal to the Court of Appeal. This highlights the emphasis placed on ensuring that there are no grounds of doubt that the person suffering the death penalty is truly deserving of the punishment. In the same vein, there should be no reason why the prosecution should not be made to justify their decision given the wide discretion they currently have and the disproportionate impact the certification decision bears on the accused. Drawing inspiration from R (on the application of Purdy) v Director of Public Prosecutions,34 Tan submits that the prosecution should promulgate a policy that outlines the factors taken for the subjective assessment of the objective condition, and this policy should count towards the “relevant standard”. This would still ensure deference towards the prosecution while still being able to hold them to 32 33 34

[2019] SCCA 13 [2016] SGCA 67 [2009] UKHL 44

account. Above all, this will greatly support the rule of law as recognised in Singapore. In “B2c2 Ltd v Quoine Pte Ltd35”, Evan Ng, University of Birmingham comments on a very interesting judgment of the SICC on the technical aspects of cryptocurrency currency trading. The issue was whether the defendant, an cryptocurrency exchange operator, was liable for breach of contract and breach of trust for reversing a customer’s Bitcoin (“BTC”)/Ethereum (“ETH”) trades. The trades in dispute were executed automatically by the deterministic software programs of the plaintiff and the defendant without any human intervention. The sub-issues included the recognition of cryptocurrencies as property as well as matters of segregation and trust over cryptocurrencies. The underlying principles were all contractual. Potential dangers of modern-day transactions involving minimal or close to zero human interaction—specifically deterministic algorithm trading— were examined to determine its ramifications on the application of the doctrine of mistake. The SICC held that where the actions were carried out by deterministic computer software, the issue of identifying and assessing knowledge would be to determine the state of mind of the programmer of the software, at the time it was written. In “Easing the law onto unchartered terrain: Regency Park Villas Title Ltd and others v Diamond Resorts (Europe) Ltd and others36”, Edwin Teong Ying Keat, University of Bristol discusses the decision of UK Supreme Court in recognising recreational easements in a development of timeshare apartments. Lord Briggs held that recreational use can constitute accommodation of dominant tenements for the purposes of constituting an easement as the law needs to accommodate societal change and new uses of land such as timeshare apartments which he equated, in terms of ‘service, utility and benefit’, to the grant of rights over a communal garden. This decision does not impact Singapore law. The Land Titles Act requires every easement to be registered before it can be enforced. Therefore, whether a recreational easement of the kind decided in this case exists in Singapore depends on whether it is registered as an easement. In “The future of third party (non- spousal) property interest being adjudicated in relation to divorce proceedings: BUE v TZQ37”, Margaret Yeow Tin Tin, HOH Law Corporation discusses competing claims to a share of a HDB flat on the basis of the presumption of advancement in a common intention constructive trust. The registration of the flat was in the joint names of father and sons, where the father and the sons contributed unequal amounts towards purchasing the flat, with the step-mother contributing towards the renovations. The articles, commentaries and case notes discussed in this issue show a continuing trend on the part of our courts to strike out on their own in developing Singapore law for Singaporeans within the common law world. Chan Sek Keong Distinguished Fellow of the NUS Law Faculty (Chief Justice 2006-2012) (Attorney General 1992-2006)

35 36 37

[2019] SGHC(I) 03 [2018] UKSC 57 [2018] SGHC 276

United Kingdom Singapore Law Studentsâ&#x20AC;&#x2122; Society Executive Committe

President: Koh Cheng Jun

Vice-President and Sponsorship Director: Joshua Wang

General Secretary: Valerie Seah

Finance Director: Annabelle Lee

Editor-in-Chief: Edwin Teong

Marketing Director: Tan Si Wen

Editorial Committe

Editor-in-Chief: Edwin Teong

Managing Articles Editor: Filbert Lam

Deputy Articles Editor: James Dylan Toh

Managing Commentary Editor: Lee Kay Howe

Deputy Commentary Editor: Noel Low

Managing Case Note Editor: Sean Loh

Deputy Case Note Editor:

Suchetra Sivakumar

President’s Address Dear Readers, We celebrate the 20th anniversary of the United Kingdom Singapore Law Students’ Society (‘UKSLSS’) this year. The Singapore Comparative Law Review (formerly known as the Lex Loci) has been an indispensable feature of our society ever since its inaugural issue 14 years ago. The publication serves as a platform for Singaporean students in the United Kingdom to keep abreast of, analyse, and reflect upon the latest legal developments in the two common law systems through a comparative lens. The Singapore Comparative Law Review’s publication process is entirely student-run. This year, it is led by our indefatigable Editor-in-Chief, Edwin Teong Ying Keat, and his Editorial Committee. The new written moots segment is a brainchild of Edwin, which facilitates interaction between students studying law in the United Kingdom and those studying in the local universities. The contributions from luminaries of the legal profession in the articles section is similarly a first. It is a testament to the quality and reputation of the Singapore Comparative Law Review, which is now available on both the HeinOnline and LawNet legal databases. Equally noteworthy is the publication of monthly newsletters this year by our Editorial (Newsletter) team, headed by Edwin (Editor-in-Chief), Shermaine Lim (Managing Editor), Ronald Lee (Deputy Editor), and Kimberley Ng (Design Director). The newsletters provide our members with timely updates and succinct analyses of developments such as the new Protection from Online Falsehoods and Manipulation Act, the Criminal Law Reform Act, and the increased application of technology (in its various forms) in the legal profession. I am immensely grateful to all our editors and contributors for their hard work and perseverance throughout the past year. The UKSLSS has had an eventful year. My team and I began our administration by welcoming Professor Walter Woon, Dean of the Singapore Institute of Legal Education (SILE) for a fireside chat hosted by the Singapore High Commission in London last November. The Vacation Scheme Helpdesk, held in the same month at King’s College London, was also enthusiastically attended by our members. The attendees were given the opportunity to interact with Singaporean lawyers who are currently working in London across various international law firms. Most of them are former members of the UKSLSS. The Rajah & Tann-Cambridge Moot and the subsequent Recruitment Tea held in London demonstrate a strong desire by local firms in engaging our members and a sustained demand for Singaporean students graduating with a law degree in the United Kingdom. The quality of our members and success of our events have been instrumental in building a robust partnership with local firms and I would like to welcome TSMP Law Corporation on board as a new sponsor and WongPartnership as a new Gold Tier sponsor this year. On this point, I would like to thank all our sponsor firms for their continued support over the years, without which we would not be able to conduct our events while ensuring their accessibility to all of our members. In March, the UKSLSS welcomed students and lecturers from Temasek Polytechnic’s Diploma in Law and Management Course. I would like to thank our university representatives— Lim Einle (University College London), Carol Yin, (King’s College London), Regina Soh (University of Bristol), Srishti Murthy (University of Bristol) and Gan Wen Xuan (University of Birmingham) who generously gave up their time to host the students and shared their experiences of studying law in the respective universities. All of our university representatives have played an indispensable role this year in helping the Executive Committee to disseminate information and organise the students for our events.

The legal profession is in a state of flux. At the 29th Inter-Pacific Bar Association Annual Meeting and Conference this year, Chief Justice Sundaresh Menon observed that “technology will be the single most potent force to reshape our legal profession in the years to come”. The UKSLSS strives continually to prepare our members to face new challenges and our monthly newsletters have played a part in increasing understanding of these technological incursions. The 1st panel at the UKSLSS’s annual flagship event— the Singapore Legal Forum 2019, is titled AI (Artificial Intelligence) and its Disruption in the Legal Industry and is specifically designed to deepen the understanding of new technological products and the involvement and competition from “alternative legal service providers” (ALSPs). As observed by one of our panelists Paul Neo, COO of the Singapore Academy of Law in an interview for the March’s edition of our monthly newsletter, the main hurdle to developing an ecosystem where legal innovation can thrive is “still (the) mindset change” (or lack thereof). The importance of flexibility and openness to change cannot be understated. Perhaps it is apt here to take a leaf out of the ancient Chinese philosopher, Lao-Tzu’s Tao-Te-Ching (Book of the Way), Verse 76: The living are soft and supple; the dead are rigid and stiff. In life, plants are flexible and tender; in death, they are brittle and dry. Stiffness is thus a companion of death; flexibility a companion of life. An army that cannot yield will be defeated. A tree that cannot bend will crack in the wind. The hard and stiff will be broken. The soft and supple will prevail. I would like to thank all my predecessors, in particular Adelle and her team for their guidance. The UKSLSS has an exceptionally strong alumni network which my administration consistently tapped on for advice and support throughout the past year. Special thanks are also reserved for Brenda Tan, our liaison from the Overseas Singaporean Unit (OSU) in London for her unwavering support this year. If you are a pre-university or current university student reading this, I hope you will join this big family. The UKSLSS (and its offerings) is akin to a buffet spread—individual members take what they need. However, I hope that you will also give a little back in one way or another. Last but not least, I would like to thank the rest of my Executive Committee—Joshua Wang (Vice President and Sponsorship Director), Valerie Seah (General Secretary), Annabelle Lee (Finance Director) and Tan Si Wen (Marketing Director) for their tremendous work over the past year. They are my pillars of strength and I will definitely miss our weekly Skype meetings. Yours Sincerely, Koh Cheng Jun President of the United Kingdom Law Students’ Society

Editorial Foreword Dear Members, The Singapore Comparative Law Review (‘SCLR’) 2019 marks the 14th edition of the annual legal publication produced by the United Kingdom Singapore Law Students’ Society (‘UKSLSS’). Originally conceived as Lex Loci, the name change in 2018 was to ‘more accurately frame the publication’1 and strive for a ‘comparative’2 rather than ‘bilateral’3 comparison of the United Kingdom and Singaporean jurisdictions. In this regard, comparative analysis, which remains ‘fundamental’4, informs the arguments made. Aligned with the UKSLSS’ 20th anniversary, the editorial committee has worked hard to canvass and edit quality submissions. This year, we have a total of 29 contributions - 17 articles, 5 case notes and 7 commentary pieces. In particular, we have the immense honour of publishing a novel piece by Head of Law School (Bristol) and a luminary in Tort Law, Professor Ken Oliphant. Professor Oliphant has written on and explored tortious liability in the age of autonomous vehicles. This is closely aligned with the journal’s theme: ‘Tradition vs. Modernity’. As we enter a new technological age, an evaluation of the law is required. This sometimes involves transcending traditional mindsets and precedents. Also, we are equally honoured to have Professor Walter Woon, a luminary in Company Law and the Dean of the Singapore Institute of Legal Education, sharing his insight on piercing the corporate veil. We are also thankful to Ms Margaret Yeow Tin Tin from HOH Law Corporation who has contributed a case note on BUE v TZQ and Mr Suang Wijaya from Eugene Thuraisingam LLP who has contributed a commentary piece on evaluating the constitutionality of S61, Protection from Online Falsehoods and Manipulation Act 2019. This year, the SCLR 2019 is proud to introduce a new segment: Written Moots. These feature writers arguing opposing stances on the same legal issue. In line with the comparative element of the journal and the need for the UKSLSS to stay grounded, I have invited writers from the National University of Singapore and Singapore Management University to partake in these written moots. This year, the 3 written moots relate to Tort Law, Medical Law and Criminal Law. These areas of law are the bread and butter of any law student. Also, other published contributions span many areas of law including Contract Law, Company Law, Criminal Law, Trusts Law, Tort Law, Constitutional Law, Land Law, Regulatory Law, Privacy Law, Banking Law, Insolvency Law, International Law, Intellectual Property Law and every law student’s ‘favourite’, Jurisprudence. These contributions have been angled and edited to reflect the journal’s theme. Most importantly, we are immensely grateful for the continued support and insight of our esteemed patron, former Chief Justice, Mr Chan Sek Keong. On this note, I wish to thank past Editors-InChief who ignited this humble flame, my editorial team and all contributing writers for burnishing this flame. Sennett posits that industrial life has not eroded craftsmanship, for it is ‘an enduring, basic 1 Singapore Comparative Law Review 2018 at p. 13 2 ibid.

3 ibid. 4 ibid.

human impulse, the desire to do a job well for its own sake.â&#x20AC;&#x2122;5 It is with this belief that readers should appreciate the effort devoted to the SCLR 2019. May this humble flame burn brighter with every passing year and shed some light when the path ahead is tenebrous. Yours Sincerely, Edwin Teong Ying Keat Editor-In-Chief

Richard Sennett, The Craftsman (Penguin Books 2008) at p. 9

Contents Articles Adelle Yii, University College London Area of Law: Trusts Law Title: Should common law jurisdictions adopt the doctrine of unconscionability? 26 Pang Wenn Ler, University of Bristol Area of Law: Contract Law Title: ‘You’ve got mail!’: Examining the postal rule in the technological age 35 James Dylan Toh, University of Bristol Area of Law: Tort Law (Malicious Prosecution) Title: “Quenchless feuds” and ill will – should the tort of malicious prosecution of civil proceedings be recognised in Singapore? 41 Justin Kwik, University of Bristol Area of Law: Privacy Law Title: In Light Of The Technical Impracticality Of The Right To Erasure, What Answers Can ActorNetwork Theories Provide? 48 Filbert Lam, University of Edinburgh Area of Law: Banking Law Title: Abandoning Old Banks and Brooking New Streams: A Case for the Reformation and Statutory Codification of the English Law on A Bank’s Duty of Confidentiality 68 Joel Sherard, King’s College London Area of Law: Intellectual Property Title: Divergent Paths: The Doctrine on Equivalents in Singapore and the United Kingdom 80 Filbert Lam, University of Edinburgh Area of Law: Insolvency Law (Pari Passu clauses) Title: Have the Rules Governing the Interpretation of Pari Passu Clauses in Sovereign Bonds Developed in Equal Step Under English and New York Law? 86

Joel Sherard, King’s College London Area of Law: International Law Title: The murky waters of Treaty Law: the 1962 Malaysia-Singapore Water Agreement 106 Carol Yin, King’s College London Area of Law: Criminal Law Title: Paternalism v Autonomy: Examining the moral concerns of criminalizing Suicide and Physician Assisted Suicide 111 Written Moot 1 (Tort Law) –Does the High Court judgment in Carol Ann Armstrong introduce uncertainty into the doctrine of lost chance in Tort Law? No Tan Yin Theng, Sarah, University of Bristol Title: Why Carol Ann Armstrong does not introduce uncertainty: a comparative analysis 117 Yes Lau Yi Hang, Singapore Management University Title: On the basis of fairness? Taking a chance on recognizing lost chances in Singapore: a comparative analysis 123 Written Moot 2 (Criminal/Company Law) – Evaluate the effectiveness of the Deferred Prosecution Agreements UK – Not Effective Loo Wee Toh, University of Bristol Title: The United Kingdom’s deferred prosecution agreement regime five years on: is it an effective tool in addressing economic crime perpetrated by companies? 137 Singapore – Effective Tan Yann Xu, National University of Singapore Title: Evaluating Deferred Prosecution Agreements in the context of Singapore 151 Written Moot 3 (Medical Law) – Should the courts decide in the best interests of dying patients? Yes Edwin Teong Ying Keat, University of Bristol Title: Nothing gold can stay: No better substitute judgment against the best interests of dying patients 163

No Soh Kian Peng, Singapore Management University Title: Not my decision to make: The courts, best interests, and legally incompetent children 174 Guest Writer 1: Professor Walter Woon, Dean, Singapore Institute of Legal Education Area of Law: Company Law Title: Lift not the Painted Veil 185 Guest Writer 2: Professor Ken Oliphant, Head of Law School, University of Bristol Area of Law: Tort Law Title: Liability for Road Accidents Caused by Driverless Cars 190

Case Notes Suchetra Sivakumar, Queen Mary University of London Area of Law: Tort Law (Medical Negligence) Title: Noor Azlin Bte Abdul Rahman v Changi General Hospital Pte Ltd and others [2019] SGCA 13 198 Tan Yan Shen, University of Oxford Area of Law: Constitutional Law (Prosecutorial discretion) Title: Prabagaran a/l Srivijayan v Public Prosecutor [2016] SGCA 67: Approach to Prosecutorial Discretion 203 Evan Ng, University of Birmingham Area of Law: Contract Law (Cryptocurrency) Title: B2C2 ltd V Quoine Pte Ltd [2019] SGHC(I) 03 207 Edwin Teong Ying Keat, University of Bristol Area of Law: Land Law (Easements) Title: Easing the law onto unchartered terrain: Regency Park Villas Title Ltd and others v Diamond Resorts (Europe) Ltd and others [2018] UKSC 57 210 Guest Writer 1: Margaret Yeow Tin Tin, HOH Law Corporation Area of Law: Family Law/Land Law (Resulting Trust) Title: The future of third party (non- spousal) property interest being adjudicated in relation to divorce proceedings: BUE v TZQ 214

Commentary Alexander Yean, University of Oxford Area of Law: Jurisprudence Title: Ready for repeal: Lessons from the Hart-Devlin Debate 220 Matthew Choo, King’s College London Area of Law: Criminal Law Title: #Mentoo: a response to the Criminal Law Reform Bill 2019 227 Noel Low, University of Oxford Area of Law: Criminal Law Title: Criminal Law (Temporary Provisions) (Amendment) Bill- An unjustified infringement on the right to a fair trial? 234 Lee Kay Howe, University of Oxford Area of Law: Constitutional Law Title: True or False or Misleading? “[A]dequate judicial oversight” over Part 3 Directions under the Protection from Online Falsehoods and Manipulation Act 239

U. Sudharshanraj Naidu, University of Bristol Area of Law: Constitutional Law Title: A call for judicial activism in response to current Anti-Terror legislation 248 Annabelle Lee, University of Warwick Area of Law: Regulatory Law/Artificial Intelligence Title: Rage against the machines: AI or Nay? 253

Guest Writer 1: Suang Wijaya, Eugene Thuraisingam LLP Area of Law: Constitutional Law Title: Evaluating the constitutionality of S61, Protection from Online Falsehoods and Manipulation Act 2019 258

ARTICLES

Should common law jurisdictions adopt the narrow doctrine of unconscionability? Adelle Yii, University College London Introduction Singapore has long grappled with the issue of whether the doctrine of unconscionability should be given limited application or whether a broader doctrine should be adopted. In the recent case of BOM v BOK (“BOK”)1, the Singapore Court of Appeal upheld the narrow doctrine of unconscionability and set aside a deed of trust. While the narrow doctrine adopted in Singapore shares similarities with the position in the United Kingdom, the Singapore Court of Appeal further developed the test to apply to modern situations. Firstly, the Court must first consider whether the purchaser was poor or ignorant and secondly, establish whether the purchase was considerably undervalued. Lastly, the Court must determine whether the purchaser obtained independent legal advice prior to the purchase. The doctrine is narrow as it holds the applicant to a high standard of proof in relation to the law of evidence, such that only exceptional circumstances will qualify as unconscionable conduct. Where an agreement is unconscionable, it may be set aside by the Court such that parties will be free from their contractual obligations. This article will briefly set out the origins of the narrow doctrine and the way in which it was adopted in Singapore. As a point of contrast, the broad doctrine adopted by the High Court of Australia will be discussed, although it is pertinent to note that the broad notion attracted much criticism in BOK. It is posited that although the adoption of a narrow doctrine carries its own set of disadvantages, the Court of Appeal in Singapore rightly restricted its application in order to maintain legal and contractual certainty.It is further argued that other common law jurisdictions 26

[2018] SGCA 83.

should consider moving towards the narrow conception of unconscionability. This article comprises of seven sections, beginning with the narrow doctrine of unconscionability adopted in the United Kingdom. This sets the foundation and allows for developments made by the Singapore Court of Appeal in the adoption of the narrow doctrine to be highlighted in the second section, through the test set out in BOK. The third section will then briefly set out the broad doctrine of unconscionability adopted in Australia, given that the test applied differs greatly from that in the United Kingdom and Singapore. Once the background and seminal cases have been laid out, the fifth and sixth sections of this article will examine the advantages and disadvantages of the narrow doctrine of unconscionability. This provides a foundation for the comparative assessments of whether Australia should consider adopting the narrow doctrine and in turn, whether Singapore should adopt the broad doctrine of unconscionability. Foundations of the Narrow Doctrine The narrow doctrine of unconscionability has been established in two seminal cases the United Kingdom – Fry v Lane (“Fry”)2 and Cresswell v Potter (“Cresswell”).3 In Fry, the Plaintiff assigned shares left under a will to the Respondent upon the advice of an inexperienced solicitor who acted for the Respondent. It later emerged that the transaction was considerably undervalued, and the shares were worth far more than their sale price. It is germane to note that an undervalued transaction in itself was insufficient to set the assignment aside and unfair dealing had to be 2 3

[1888] 40 Ch D 312 (Ch). [1978] 1 WLR 255 (Ch).

ARTICLES

shown, which allowed for the consideration of unconscionability.4 The Court applied a narrow conception of unconscionability and found that the assignment could be set aside.5 The first stage of the test requires the Court to consider if the purchase was made from ‘a poor and ignorant man’, secondly, at a considerable undervalue and thirdly, with no independent legal advice.6 Where the Plaintiff satisfies the test above, the burden of proof then shifts on the Respondent to show that the purchase was ‘fair, just and reasonable’.7 The doctrine is considered to be narrow as it demands a high burden of proof, effectively allowing only exceptional circumstances to qualify, such as where there was a significant imbalance of bargaining power and a claimant in a clear position of vulnerability. The test is affirmed and developed in Cresswell. Following a divorce, Cresswell signed a document of conveyance for property purchased during the marriage which effectively released all her interest in the property in return for an indemnity against liabilities. She did not have independent legal advice and claimed that she signed the conveyance believing it would merely allow the sale of the property without affecting her rights in it. Crucially, the agent sent to obtain the release did not explain the effect of the document and merely mentioned that the document was for ‘the sale of the house’.8 The Court held that the document of conveyance could be set aside and upheld the narrow test for unconscionability. Alongside the test in Fry, Megarry J recognised that ‘there may be circumstances of oppression or abuse of confidence which will invoke the aid of equity’.9 The success of Cresswell’s claim depended heavily on the unusual nature of the agreement,10 given thatthe Defendant himself had initially regarded it as incapable of giving 4 Fry (n 2) 321 (Kay J). 5 ibid 323-24. 6 ibid 322. 7 ibid. 8 Cresswell (n 3) 256 (Megarry J) 9 ibid 257. 10 ibid 259.

him any legal rights and the wording had no clear legal implications.11 Therefore, this limits the doctrine of unconscionability to a narrow interpretation. Adoption of the Narrow Doctrine in Singapore The case of BOK concerned a Deed of Trust drafted by the Wife, a former lawyer, which was signed by her Husband after he was appointed executor of his mother’s will and testamentary trust. The Deed of Trust provided that the Husband and Wife would hold all the Husband’s assets on trust for their son. Upon entering into a deed of arrangement over his mother’s will, the Husband requested that the Deed of Trust be delivered up to be destroyed. He claimed that he did not receive independent legal advice and did not understand it to be legally binding or that it would include the assets under his mother’s will and testamentary trust. The Court of Appeal held that the Deed of Trust could be set aside on grounds of mistake and misrepresentation,12 but also further narrowed down the doctrine of unconscionability applicable in Singapore. Phang JA set the test to be whether the plaintiff was poor and ignorant and whether the situation was one where the Plaintiff was suffering other forms of infirmities, whether mental or physical in nature.13 The infirmity must be of ‘sufficient gravity as to have acutely affected the plaintiff ’s ability to “conserve his own interests”’.14 This test differs from that in Fry and Crewsswell, by placing more emphasis on the presence of infirmities rather than the value of the transaction. However, the test still draws inferences from Fry and Cresswell by looking at whether the plaintiff was ignorant and had a lack of understanding the implications of the transaction. If these elements are satisfied, the Respondent is then responsible for showing that the transaction was ‘fair, just and 11 ibid. 12 BOK (n 1) [55] (Phang JA). 13 ibid [141]. 14 ibid (emphasis added).

ARTICLES

reasonable’.This seems to suggest an evidential test concerned with the burden of proof, which is shifted to the Respondent before the determination of any remedies. Although the requirements of independent legal advice and the transaction being at a considerable undervalue are important factors for consideration, they are not mandatory factors to establish unconscionability.15 This further narrows the doctrine of unconscionability, as the requirement of an infirmity of “sufficient gravity” raises the threshold for finding unconscionability. The term “sufficient gravity” was defined as an infirmity which acutely affected the Plaintiff ’s ability to protect his own interests.16 In BOK, this high threshold was satisfied as the Wife was aware of the Husband’s impaired mental state following his mother’s passing and took advantage of this in coercing him to sign the Deed of Trust.17 The position in Singapore was largely founded on the test set out in Fry and Cresswell but was further limited through the introduction of the infirmity element. This acts as a new addition to the three-stage test which is arguably beneficial to the doctrine of unconscionability. By imposing a stricter test, the doctrine of unconscionability can only be utilised by those who have suffered a grave disadvantaged caused by the other party. The addition of the infirmity element also protects contractual certainty by imposing a higher threshold, such that those who are capable of understanding the implications of the contract cannot easily claim that the contract should be void for unconscionability. The narrow doctrine adopted in Singapore also confers greater respect for privity of contracts, which is the principle that contracts are a private matter between parties and can only be enforced by parties.18 This essentially means that it should be up to parties to determine whether a contract is to be enforced and Courts

15 ibid. 16 ibid. 17 ibid [154]. 18 Brian Bix, Contract Law: Rules, Theory and Context (Cambridge University Press 2012)

should not intervene unnecessarily. Therefore, the narrow doctrine confers greater respect for privity due to the high threshold, demonstrating the Court’s reluctance to intervene in private contracts. Broad Doctrine Adopted in Australia In contrast to the English and Singapore positions, the High Court of Australia adopted the broad doctrine of unconscionability. In Commercial Bank of Australia v Amadio (“Amadio”),19 the Amadios signed a guarantee to the bank in relation to their son’s business which effectively gave the bank a mortgage over a building which the Amadios owned. The bank was aware that the son’s business was failing and did not disclose this to the Amadios. As a result, the Amadios claimed that they were at a special disadvantage and therefore the guarantee was unenforceable. The majority held that the bank could not rely on the guarantee as it would be unconscionable to do so.20 Gibbs CJ reasoned that the guarantee was unenforceable because the bank’s failure to disclose information regarding the son’s failing business to the Amadios constituted unconscionable conduct. A common law duty of disclosure was placed on the bank where it would affect the nature of degree of the surety’s responsibility.21 This means that banks are not guilty of unconscionable behaviour where it does not alter the relationship between the plaintiff and the bank. A Plaintiff seeking to establish unconscionable conduct must show that unconscientious advantage has been taken of his disabling condition or circumstances.22 While unconscionable conduct refers to conduct which is extremely unjust, unconscientious advantage must specifically involve taking advantage of a disabling condition suffered by one party.23 Although this seems to set a 19 [1983] HC 14, [1983] 151 CLR 447 (High Court of Australia). 20 ibid [20] (Gibbs CJ). 21 ibid [13] 22 ibid [7] (Mason J). 23 ibid.

ARTICLES

barrier as high as that in the United Kingdom or Singapore, it is the application by the Courts that makes the doctrine broad. Given that the mere failure to disclose information was sufficient to constitute an unconscientious advantage, there is no real need for any exceptional circumstance in order to rely on unconscionability. This leaves a wide discretion to the judges and gives rise to a broad application of the doctrine. Advantages of the Narrow Doctrine The first and arguably most significant advantage of adopting the narrow doctrine of unconscionability is that it respects fundamental principles of contract law – the sanctity of contracts. Sanctity of contract is the idea that once parties enter into a contract, they must honour their obligations under the contract.24 The importance of the sanctity of contracts was highlighted in Tee Soon Say v AG.25 In obiter, Phang JA notes that an ordered society depends on parties observing promises made under the law to each other.26 This reiterates the idea that contracts should not be set aside easily due to unconscionability, thus it is more appropriate to adopt a narrow conception. In BOK, Phang JA criticised the broad doctrine of unconscionability as a vague and general doctrine which does not ‘furnish sufficient legal criteria.’27 A vague stance on unconscionability introduces incoherence in contract law, as there will be uncertainty on when contracts will be set aside due to a the plaintiff being on the losing end of an agreement. The narrow conception of unconscionability will only be applicable where there a grave disadvantage has been suffered and the contract does not reflect the parties’ intentions. This preserves contract certainty, which is of utmost importance common law jurisdictions given that it is used both in economic circumstances through commercial contracts and on a micro scale through personal 24 Pearlie Koh and Ho Tham Chee and Woan Lee Pey, ‘Contract Law’ (2007) SAL Annual Review 150 25 [2007] 3 SLR 133 26 ibid [109] (Phang JA) 27 BOK (n 1) [121] (Phang JA).

agreements. The narrow doctrine ensures that parties cannot easily renege on agreements due to the high burden of proof required in the first stage of the test. While the test in English law adopts a different test, the same outcome of contract certainty is still achieved. English law places emphasis on the need for ignorance and the lack of independent legal advice instead of a mere undervalued transaction, imposing a higher threshold. This essentially requires a demonstration of vulnerability and a lack of understanding of the contractual implications before a contract can be set aside. Furthermore, the narrow doctrine whereby unconscionability is not easily raised will encourage parties to seek independent legal advice as a form of personal protection. The Court of Appeal in Singapore took a slightly different approach from Fry and Cresswell. The lack of independent advice was not a strict requirement in proving unconscionability, but Phang JA recognised that it weighed heavily in the decision-making process.28 It is argued this should not be a mandatory requirement in Singapore as it creates a loophole whereby plaintiffs will avoid seeking independent legal advice prior to a contract, in an attempt to claim that the contract should be set aside in the future. In BOK, although the Husband’s lack of legal advice tipped the balance in favour of upholding the unconscionability claim, it can be argued that the more important factor was the considerable undervalue of the transaction which ‘will often underscore and highlight the exploitation of an infirmity’.29 However, it is important to note that the infirmity and undervalued transaction must be of sufficient gravity, whereby the plaintiff was incapable of protecting his own interests, to establish exploitation. This factor was applied bearing in mind that the Husband suffered from infirmities arising from the acute grief of his mother’s passing, coupled with the lack of legal advice.30 Thus, in order to avoid such a 28 ibid [155]. 29 ibid. 30 ibid [154].

ARTICLES

predicament, parties to similar transactions will likely be better off having obtained legal advice and being clear of implications arising from agreements they may enter into. Therefore, the knowledge that unconscionability is narrow and hard to establish will encourage parties to exercise caution prior to entering into contracts. This serves a wider public interest by encouraging parties to seek independent legal advice and protects the sanctity of contracts. Disadvantages of the Narrow Doctrine The narrow doctrine places an onerous burden on individuals who are already in difficult predicaments. Unconscionability in its narrow form sets out high barriers to be crossed for a course of conduct constitute unconscionable conduct. This can be seen through the need to establish that a mental or physical infirmity that was of such gravity to find that the Respondent had taken advantage of this vulnerability. Claims for unconscionability are often brought by those who are already in a vulnerable position as a result of the transaction. It can be argued that these requirements place them in a tougher situation instead of protecting them. Backhouse v Backhouse31 concerned divorce proceedings, whereby the Wife executed a transfer of the matrimonial home to the Husband whilst in an emotional state, without seeking legal advice. Although the transfer was not given effect under the Matrimonial Causes Act 1973 (‘MCA 1973’), in obiter, Balcombe J expressed that unconscionability would not have been established in this case.32 Balcombe J applied the three-stage test as established in Fry and found that unconscionability would fail on the poverty and ignorance ground as he deemed the Wife to be an intelligent woman.33 The Wife was considered to be intelligent due to the fact that she managed the finances of the household throughout the marriage.34 It is argued that the Wife was already vulnerable due to the distress

31 32 33 34

[1978] 1 All ER 1158 (F). ibid 252 (Balcombe J). ibid 251. ibid 245.

of divorce proceedings and managing household finances does not equate to an understanding of legal implications. In addition, if not for the provisions of the MCA 1973, the Wife would have lost her home. Moreover, majority of plaintiffs’ act under pressure and without independent legal advice, yet we expect them to be competent enough to satisfy an onerous burden of proof for unconscionability. Alongside evidential burdens, such cases tend to be drawn out due to the need of evidence and to hear both sides of the story. In BOK, the Husband and Wife disputed whether the Deed was signed willingly or under coercion.35 The high thresholds of the narrow doctrine tends to lengthen the process as more evidence must be put before the Court and a lengthy process may compromise parties’ interests. This means that claimants invest time, effort and money into providing evidence only to find that more likely than not, unconscionability will not be found. Moreover, the test of whether the transaction was fair, just and reasonable introduces policy considerations which may work against the plaintiff by subsuming the first two steps of the narrow doctrine in Singapore.36 This introduces uncertainty as policy does not make reference to a clear line of principles, but are fact-dependant.37 A plaintiff may satisfy the first two steps of the test, yet unconscionability may not be established if the Courts find that there are policy reasons not to impose liability on the defendant. This raises questions as the plaintiff would have proven that there was an infirmity of an extremely grave nature giving rise to unconscionable conduct, yet they are left with no recourse as a matter of policy. Therefore, it is argued that the narrow doctrine in Singapore does little to provide a remedy to a plaintiff where there is evidence of 35 ibid [75] (Phang JA). 36 Lord Neuberger, ‘Some Thoughts on Principles Governing the Law of Torts’ (Singapore Conference on Protecting Business and Economic Interests: Contemporary Issues in Tort Law, Singapore, 19 August 2016) <https://www. supremecourt.uk/docs/speech-160819-03.pdf> accessed 5 June 2019. 37 ibid.

ARTICLES

unconscionable behaviour and may introduce an incoherent application of the doctrine. The narrow doctrine adopted in Singapore also seems to overlap with other doctrines such as duress and undue influence. Duress relates to where a person enters an agreement as a result of threats and in such cases, the contract may be set aside.38 In Tjong Very Sumito and others v Chang Sing En and others,39 it was held that there are two elements to constitute duress. The first element is to consider whether there was an exertion of illegitimate pressure and the second element is to determine whether the pressure compelled the victim to act.40 Phillips argues that in reality, duress is ‘no more than the doctrine of unconscionability in disguise’ due to the common principle that one party was coerced into entering into a contract.41 This is especially given the common reliance on the availability of independent advice. Despite these similarities, unconscionability is still distinguishable from duress and very much necessary. Promises made under duress are not freely made but are made in response to improper pressure, thus there was no legitimate intention to enter into the contract.42 However, unconscionability suggests that the starting point is that the contract was freely entered into as there is no illegitimate pressure. Instead, unconscionability serves to show that while parties intended for there to be an agreement, the exact terms of the agreement do not reflect the intentions of parties. Hence, duress and unconscionability serve to deal with different situations and the two doctrines cannot be conflated. On the other hand, undue influence exists where a contract has been entered into under pressure which falls short of duress. In Royal Bank of Scotland v Etridge (No. 2),43 Lord Nicholls noted 38 Brian Bix, Contract Law Rules, Theory and Context, (Cambridge University Press 2012), 47. 39 [2012] SGHC 125 40 ibid [247] (Steven Chong J) 41 John Phillips, “Protecting Those in a Disadvantageous Negotiating Position: Unconscionable Bargains as a Unifying Doctrine” (2010) 45 Wake Forest Law Review 837. 42 Charles Fried, Contract as Promise: A Theory of Contractual Obligation, (Oxford University Press 2015), 92. 43 [2001] UKHL 44; [2002] 2 AC 773 (HL).

that the objective of the doctrine is to ensure that ‘the influence of one person over another is not abused’.44 Undue influence requires the defendant to have exercised possessive influence over the claimant in a way that amounted to an abuse of that influence, which brought about the impugned contract.45 A defendant has possessive influence where he has the ability to engage in misuse without any specific overt acts of persuasion.46 While unconscionability and undue influence both require elements of influence, there are differences between the two doctrines. As with duress, undue influence concerns an individual who has entered into a contract under the influence of another and not out of freewill. This is distinguished from unconscionability, where parties intended to enter into an agreement but due to unconscionable behaviour, the agreement is not an accurate reflection of these intentions. Spark argues that where the defendant has actual or constructive knowledge of the claimant’s weakness which impairs consent, such conduct falls within the scope of unconscionability.47 By placing emphasis on the presence of knowledge in unconscionability, the Courts are concerned with the defendants conduct in attempting to obtain a benefit for himself. On the other hand, undue influence is more concerned with the plaintiff being dependant on the defendant and whether this dependency affected the quality of the plaintiff ’s consent. Therefore, undue influence focuses more on the relationship between parties whilst unconscionability looks to the conduct of the dependant in procuring the agreement or contract. This means that the two doctrines tackle different situations and it would be wrong to say that unconscionability can be subsumed into the doctrine of undue influence. The Court in Singapore focused heavily on the Wife providing a false or wrong interpretation of 44 ibid [6] (Lord Nicholls). 45 Gareth Spark, Vitiation of Contracts, (Cambridge University Press 2013), 251 46 ibid 255. 47 ibid 286.

ARTICLES

the contents in the Deed and the misunderstand of its contents. In BOK, the Deed of Trust was set aside due to the finding of undue influence48 and in obiter, it was noted that this was due to the fact that it did not give effect to the Husband’s true intentions. The requirement of an infirmity that was taken advantage of arguably falls within the doctrine of undue influence. On the other hand, the coercion placed on the Husband arguably fell within duress. In these circumstances, it seems that the doctrine of unconscionability is unnecessary. If the doctrine of unconscionability is so narrowly applied that cases are to be determined on the basis of other legal doctrines, it seems that there is little point in recognising unconscionability as a separate doctrine at all. However, it is argued that the narrow version of the doctrine of unconscionability is still necessary. Duress provides a remedy where the plaintiff had no real intention to enter into a contract while undue influence is applicable to situations where a dependant relationship is present, which affected the quality of the plaintiff ’s consent. Narrow unconscionability accepts that parties intended to enter into a contract but due to the exploitive behaviour of the defendant, the terms of the contract are not a reflection of the plaintiff ’s true intentions. Should Australia Adopt the Narrow Doctrine of Unconscionability? The adoption of the narrow doctrine in jurisdictions such as Australia would mean that many of their existing unconscionability cases would fall foul of the doctrine. For example, Amadio is unlikely to satisfy the test of the narrow doctrine as there was no exploitation of a physical or mental infirmity as it merely involved the non-disclosure of facts. In Amadio, the High Court of Australia classified the lack of knowledge as a special disability, which would not have been the case under the narrow doctrine.49 This does not satisfy the requirement of an infirmity as the lack

48 49

ibid [92] (Phang JA). Amadio (n 15) [18] (Wilson J).

of knowledge can easily be remedied through simple investigations or communications with their son.It is possible that the scope of unconscionability was expanded in Amadio due to statutory developments at the time. The Contracts Review Act 1980, implemented three-years prior to Amadio, provided relief for unjust contracts, whereby ‘unjust’ was defined as unconscionable.50 The House of Representatives has made clear that it intended for unconscionability to be given a broad interpretation, with the intention of protecting interests of vulnerable parties.51 It is extremely plausible that the High Court of Australia adopted a broad doctrine of unconscionability in order for the common law doctrine to reflect Parliamentary intentions. Thus, in considering whether the narrow doctrine should be adopted, consideration must be given to the existing Contracts Review Act 1980. The broad doctrine of unconscionability was criticised by the Court in BOK as being a ‘broad discretionary legal device which permits the court to arrive at any decision which it thinks is subjectively fair in the circumstances’.52 In obiter, Phang JA noted that this was the case particularly because Singapore did not have the benefit of an ‘initial legal platform’53 in the form of a narrow doctrine to provide guidance to the judiciary. This essentially means that the broad doctrine lacks coherency, as there is no fixed threshold or principle as to when unconscionability can be established. Depending on the application of the Courts, may eventually be exploited as a back-door method of reneging on contracts simply because one party is on a losing end. The Singapore Court of Appeal also justified the adoption of a narrow doctrine by stating that the burden of proof is lower than the threshold of duress and undue influence,54 providing two layers of protection for vulnerable individuals. 50 S4(1), Contracts Review Act 1980. 51 Commonwealth, Parliamentary Debates, House of Representatives, 29 November 1985, 4249 (Lionel Bowen). 52 BOK (n 1) [148] (Phang JA). 53 ibid. 54 ibid [152].

ARTICLES

This provides certainty in the law around unconscionability, whilst devising a safety net for individuals who do not meet the threshold of duress or undue influence. Therefore, it can be argued that the narrow doctrine does serve its purpose of protecting the vulnerable whilst respecting contractual certainty and other jurisdictions, such as Australia, should adopt it. However, the narrow doctrine of unconscionability will directly contradict the broad definition in the Contracts Review Act 1980, which is an indication of Parliament’s intentions. Although it is argued that Australia should adopt the narrow doctrine to provide coherence, it is understandable if no change is made as to do so would require a lengthy process of obtaining Parliamentary approval or the judiciary setting out detailed reasons justifying the reform. Should Singapore Adopt the Broad Doctrine of Unconscionability? The narrow doctrine adopted in Singapore bears some resemblance to the broad doctrine adopted in Australia as the “fair, just and reasonable” test introduces extralegal norms and practices in the process.55 Thus, it can be argued that a broader conception of unconscionability will not be radically different from the present position in Singapore. In the broad doctrine, this is evident through the question of whether parties have acted in a morally acceptable manner.56 If the broad doctrine was applied in Singapore and the United Kingdom, it is argued that unconscionability would have easily been established in each of the respective cases. In BOK, the husband clearly suffered from disabling circumstances following the passing of his mother and his lack of legal knowledge to understand the full implications of the Deed.In Fry, the plaintiff ’s disabling circumstance would have been his ignorance 55 Colleen McCullough, ‘Unconscionability as a Coherent Legal Concept’ (2016] 164 University of Pennsylvania Law Review 779, 823. 56 Amadio (n15) [22] (Deane J).

to legal consequences of the agreement which the defendant took unconscientious advantage of by recommending an inexperienced solicitor to the plaintiff. Likewise, in Cresswell, the plaintiff ’s lack of legal understanding would have constituted a disabling condition, which was taken advantage of when the agent sent by the defendant failed to explain the effects of the conveyancing agreement. However, it is argued that Singapore should not adopt the broad doctrine of unconscionability as the current law has struck a good balance between protecting the interests of individuals and ensuring contractual certainty. The narrow doctrine exercises caution in ensuring that parties cannot easily renege on contracts simply because they did not understand the nature of it, as this usurps the sanctity of contracts. Instead, it ensures that the Claimant must demonstrate that he had an infirmity of sufficient gravity which the Respondent took advantage of if he were to argue that the contract should be set aside on grounds of unconscionability because the contract could not have reflected the true intentions of the parties. Given that a strict test was applied in a private agreement, it is a clear that a strict test will also be applied in the commercial context, protecting contractual certainty. While contractual certainty is important in both private and commercial environments, it is argued that this is more so in the commercial context. Private contracts affect the individual parties while a commercial contracts affect the operations of the corporation. Furthermore, commercial parties will often have the resources to obtain independent legal advice prior to the contract, allowing them to ensure that the contract reflects their intentions. Moving towards a broad doctrine of unconscionability will only disrupt the coherence that the Court has preserved in BOK and will put contract law in a state of confusion as plaintiffs’ can easily establish unconscionable behaviour to set contracts aside.Coherency cannot be sacrificed simply to allow for moral grounds to seep into the law of unconscionability. A 33

ARTICLES

broad doctrine will also be detrimental to public interests as it easily undermines the integrity of the bargaining process behind contracts.57 Therefore, Singapore will benefit greatly from the narrow doctrine of unconscionability and should not attempt to expand its scope of application. Conclusion This article has highlighted that both the narrow and broad doctrine of unconscionability brings its own set of advantages and disadvantages and a determination of which is more suitable depends on the weighing of contextual policy aims. Australia has adopted the broadest definition of unconscionability vis-à-vis the narrow doctrine initially founded in the United Kingdom which was later further restricted in Singapore. The narrow doctrine protects the sanctity of contracts by requiring parties to prove that the unconscionable conduct led to the contract being an inaccurate reflection of the parties’ intentions. The narrow doctrine adopted in Singapore also carries its own imperfections, namely through imposing strict burdens on parties who are already in vulnerable positions. Furthermore, the unconscionability has a stark resemblance to the doctrines of duress and undue influence, making it difficult to justify maintaining unconscionability as a separate doctrine. Although there are clear disadvantages in restricting the scope of unconscionability, it is argued that it would be best if Singapore were to keep the law in its current state. The narrow doctrine is well-balanced in protecting individuals whilst respecting the sanctity of contracts by imposing a strict test in the finding of unconscionability. If the broad doctrine were to be adopted in Singapore, BOK would have been less contentious as the plaintiff would have had no difficulty in proving a disabling circumstance which was taken advantage of.It

57 John Spanogle, ‘Analysing Unconscionability Problems’ (1969) 117 University of Pennsylvania Law Review 931, 932.

is further argued that other jurisdictions such as Australia should consider adopting the narrow doctrine of unconscionability in order to protect contractual certainty. However, such a pivotal shift will require the judiciary to reconsider the application of the doctrine in its entirety and the judiciary may defer to Parliament to impose any reforms, due to its dependence on the Contracts Review Act 1980.

ARTICLES

‘You’ve got mail!’: Examining the postal rule in the technological age Pang Wenn Ler, University of Bristol Introduction The postal rule, long established since 1818 has always been an outlier in contract law for its simple conclusion that acceptance would be completed upon posting to the Post Office. It was seen to be a drastic departure from the fundamental principles of contract1 as it places the risk of fault on the offeror in a disproportionate way. With much discussion of whether the rule is still required and if the rule should apply to electronic mail, the postal rule is once again cast in the spotlight. This essay seeks to examine the current state of the postal rule across the jurisdictions of the United Kingdom and Singapore. Due to the many changes in the world, technology has slowly become one of our default communication methods and hence, it is critical to explore which rule would be more appropriate for contracts and transactions made through electronic means in our technological era. Historical Origins of the Postal Acceptance Rule The rule was established in a time where a rule was required and was meant to provide a guideline for certainty of contracts2 especially when post was frequently used and had many issues that were not in the control of either party. It was established in Adams v Lindsell3 where Law J established that when the offeree posts his letter of acceptance, consensus ad idem is completed, which concludes the offer and provides effect to the acceptance. This gave 1 Hill Simone W.B, Email Contracts - When is the Contract Formed? (Journal of Law, Information and Science Issue 46 2001) 2 Macdonald Elizabeth, Dispatching the dispatch rule? The postal rule, e-mail, revocation and implied terms (Web Journal of Current Legal Issues Vol 19 No 2 2013) 3 Adams v Lindsell [1818] 1 B & Ald 681

rise to the postal rule and the reasoning behind could be due to a lesser departure from the idea of consensus that has been established in contract law.4 The true reasoning behind the court’s choice of time of posting to be an indication of consensus is vague and has been debated by many scholars and criticised for its arbitrariness. After Adams, the courts returned to the issue in various cases throughout the 1800s such as in Stocken v Collin5 and Duncan v Topham6 with an acceptance of the rule but without any expressly articulated reasoning. Gardner proposed certain explanations such as the understanding of the post office as an agent for the offeror in the equation and hence acceptance can be concluded once the letter has been committed to the post office.7 Another explanation for the rule refers to the idea of business convenience where there would be no rule that would not be fair to either side, Thesiger LJ recognises that it would be more convenient to fix acceptance on posting than receipt. His judgment in the case of Household Fire and Carriage Accident Insurance Co v Grant8 provides some explanation for supporting this rule, mainly with the control that offeror has, including being able to stipulate the form of acceptance and being able to enquire the offeree on his acceptance. Understanding the postal rule in the context of business convenience has not been the most persuasive but has been cited in the argument for the rule, focusing on how the way acceptance is received can be stipulated. The explanation works well in the modern context where a wide range of communication 4 Gardner Simon, Trashing with Trollope: A Deconstruction of the Postal Rules in Contract (Oxford Journal of Legal Studies Vol 12 No 2 1992) 5 Stocken v Collin [1841] 173 ER 997 6 Duncan v Topham [1849] 137 ER 495 7 Gardner Simon, Trashing with Trollope: A Deconstruction of the Postal Rules in Contract (Oxford Journal of Legal Studies Vol 12 No 2 1992) 8 [1879] 4 Ex D 216

ARTICLES

options exist such as fax machines, emails and other forms of instantaneous communications. With the rise of technology, the postal rule was once again placed in the spotlight for its application on modern day communication. The issue of postal rule was perceived slightly different in Scotland with no wide recognition of the rule even though its outcome has been followed in later cases. Before 1800, contract formation in Scotland did not follow the strict mode of offer and acceptance unless in the situation of house purchases but the case of Smith v Duke of Gordon9 brought that concept to light. Fellow lawyer-jurist Stair commented on the issue specifically on the idea of offer that it is non-obligatory before acceptance with the offeror reserving the rights to revoke it at anytime.10 This understanding matches with English law on the matter. This was a relatively big change in Scottish law as their general Scottish understanding of contract formation was the idea that direct communication in terms of letters or posts was not key to the existence of contractual obligations, but the expressed will of both parties being more important.11 Hence, with that understanding, postal rule in Scotland was less perceived as the requirement of a letter being posted or the exact timing of when the contract was formed was not as important as the intentions and wills of the parties. Bell, who helmed the position of Scots Law Chair was one of the first to provide a direction of the doctrine of offer and acceptance where he provided his support for the doctrine despite the past development and mentioned the need of documentation, that could be inferred to refer to postal rule.12 This development continued in the case of Dunlop v Higgins13 where Lord Fullterton rejected that posting was enough and stated that the offeror has to know the acceptance had already been made before any contract

9 [1701] Mor. 16987 10 MacQueen Hector L, “It’s in the Post”: Distance Contracting in Scotland 1681-1855 (Essays in Conveyancing and Property Law, 2015) 50 11 ibid 52 12 ibid 53 13 Dunlop v Higgins [1848] 1 H.L.C 381

could be made. It seemed to place emphasis on the need of actual communication being made and the postal rule being inadequate. However, much of this changed with time where the Lord Chancellor affirmed the postal rule in Scotland in Dunlop and other cases that followed where he added comments that seemed to referred to Bell’s view of communication of acceptance was different from offer and so, is not necessary.14 The important part of a contract to him was simply the concurring consent of parties that would not be changed with the use of post. This concept continued its influence with Lord President in the case of Thomson that “the act of acceptance was completed by the putting of the letter into the post-office” and found authority in Bell’s commentaries.15 The development of the rule in Scotland reflects a similar issue with English law where the rule has brought about certain questions on its usage but has generally found its place in both areas of law as a way of communication of acceptance in forming contracts. Influence of Telecommunication Technology on the Postal Acceptance Rule The creation of the postal rule was primarily meant to provide for certainty in a time where communication took place mainly by post, as established. This became an exception to contract formation but was justified by the particular mode of communication at that time. Yet, the postal acceptance rule might not apply with our current ways of communications. With electronic and instantaneous communications, it has brought about questions whether the postal rule is still required and how it applies. The rule has generally seemed to be shafted aside with cases such as Entores v Miles Far East Co16 and Brinkibon Ltd v Stahag Stahl GmbH17 due to how instantaneous technology has become such as email, reverting back to ac14 15 16 17

ibid at [59]-[60] ibid at [66] Entores v Miles Far East Co [1955] 2 QB 327 Brinkibon Ltd v Stahag Stahl GmbH [1983] 2 AC 34

ARTICLES

ceptance has to be directly communicated and received. The distinction made in Entores perfectly illustrates the courts’ acknowledgement of how the rule will not apply in such situations. It was first established in the case of Entores where Denning LJ found that postal rule would not apply to instantaneous communications such as a telex machine. He made a clear distinction between instantaneous communications and communicating by post by virtue of the speed of the communication, where he expressed his opinions, “Communications by these means are virtually instantaneous and stand on a different footing.”18 The main reasoning behind his judgment relates to the idea of both parties having a consensus and for such communications, the offeree would have sufficient reason to know if his acceptance has been received as well as the offeror being able to send a request if any delay or error were to have occurred. Entores ushered a different era to the postal rule and set out rules that govern acceptance in instantaneous communications. This was important due to the changing landscape of companies where many started to adopt technology that affected the formation of contracts. Thus, the courts had to consider the exact time and place where the contract is formed when technology is used when the disputes came pouring in and establishing the rule of acceptance upon receipt of the offer made. The rule was further developed in the case of Brinkibon that was about the formation of a contract using telecommunication. The judges in the case followed precedent and applied the principle established in Entores but provided more opinions on the formation of contract via telecommunications. Lord Wilberforce in the leading judgment expressed that he agreed that it is a sound rule in the context of telecommunications but does not see the rule being applied in all circumstances.19 He specifically mentioned about many possible agents that could be involved or errors that could occur due to 18 19

Entores, 333-334 (per Lord Denning) Brinkibon at 42 ( per Lord Wilberforce)

these machines being operated by third persons and hence the inability to have a universal rule to cover all cases but “must be resolved by references to the intentions of the parties, by sound business practice and in some cases by a judgment where the risks should lie.”20 This provided a further development to acceptance in the modern context where Lord Wilberforce provided certain criteria that could be considered when deciding acceptance and a return to the idea of consensus through the intentions of the parties. In a recent case of Greenergy S A v Memphis Biofuels21, the question of contract formation through email was settled with the judgment that the acceptance took place where the email was received, which seems to infer that the postal rule would not apply in these situation. Apart from decisions from the English courts that provide some clarity on the application of the rule, the development of the rule in electronic communications is still important and a matter for future courts. Thomas v BPE Solicitors22 for example, highlighted that while there was no ‘straightforward’ answer as regards when an emailed acceptance is deemed effective, due consideration must be made to the intention of parties, sound business practice or in some cases a judgment of where the risks should lie.23 Modern Understanding of Electronic Communications Acceptance through electronic means was never meant to be a simple area of law to navigate given the lack of a universal rule. However, this uncertain area of law has received greater clarity with time. It has been discussed whether instantaneous communications such as emails are actually instant and looking at the speed of their delivery, which was mentioned by those in support of applying the rule in emails as it was a similar reason being stated in support for the

20 ibid. 21 Greenergy S A v Memphis Biofuels [2008] All ER (D) 15 22 Thomas v BPE Solicitors [2010] EWHC 306 (Ch) 23 ibid. at [90]

ARTICLES

existence of the postal rule.24 With the increased usage of emails, it has been understood that there is a complex system behind electronic communications between emails being sent out until they have reached the inbox of its intended recipient. The general breakdown of the flow in electronic communication is when the email being sent will stay in the sender’s outbox before it goes into the servers of the internet, finally reaching the inbox of the recipient that can be downloaded when the recipient is connected. Hence, some argue that the servers seem to act like a postal service for the email and that there is enough similarity to consider the postal rule to apply in these instances as well, but has not been accepted or addressed directly by the courts.25 It has been generally recognised that traditional definitions of instantaneous communications were not instantaneous as conceived, with the issues of having different servers for emails as well as the engaging of third party communication management.

principles as mentioned in Thomas. However, contract formation through emails has been accepted as common phenomena of the digital age and without a clear idea, it could bring about possible discrepancies between parties during the period of contract formation. An important question left unanswered is when is acceptance effective? The answer can be divided into upon dispatch or upon receipt.27 This question’s importance cannot be emphasised further due to how it affects the formation of the contract and the terms inside. The problem arose when discussions about these issues through conventions or model laws provide no definition for which timing validated the acceptance. Both timings are significantly different with one being the time when the offeree sent the acceptance, the other being when the acceptance appeared in the offeror’s inbox and the need to consider a possible delay between both events.

The issue continues to be problematic, as the English courts have not provided a preference of any rule or definition. The few explanations that the courts have provided is illustrated in Brinkibon where other judges agreed with Lord Wilberforce’s statement that there is no universal rule that will cover all situations but there is a need to look at each situation on its own. With this development, it provides a large grey area for the postal rule in emails as some argue for the rule to apply due to similarities between the server system in emails to postal service26 while some dispute that extending would be too much, which will bring about further uncertainty.

Different countries adopted different interpretations when addressing the questions regarding the validity of the rule and the possible legal obligations that follow. This led to a type of consolidation on an international scale where the United Nations Commission on International Trade Law (UNCITRAL) provided a model law that countries could adopt into their jurisdiction if they wished to.28 The model law seems to mirror the position of the postal rule with the decision in Brinkibon where an email is deemed as received when it appears in the offeror’s inbox and can be retrieved for reading. It differs from the rule, as its effectiveness does not begin when the email has been sent but as long as it has successfully gone through the Internet servers and appeared in the recipient’s inbox that would conclude the contract formation. The model law reflects an acceptance of the intricacies of electronic

Due to the lack of direction, it has been generally accepted that the courts will continue to follow Brinkibon and look at sound business

24 Capps Deveral, Electronic mail and the postal rule (International Company & Commercial Law Review, Issue 15(7), 2004) 25 ibid 26 Cummiskey Siobhan, The Effect of the Electronic Commerce Act 2000 on the Operation of the Postal Rule (Irish Business Law Quarterly Vol 1 Issue 4, 2006)

Postal Rule in International Commerce

27 Mik Eliza, The Effectiveness of Acceptances Communicated by Electronic Means, Or – Does the Postal Acceptance Rule Apply to Email (Journal of Contract Law, 26(1) 68-96, 2009) 28 United Nations Commission on International Trade Law, Model Law on Electronic Commerce 1996, Article 13

ARTICLES

communications. Nonetheless, the model law is a suggestion that does not impose any mandate. That can be seen in the adoption of the law by certain countries such as Singapore, Australia but has not been incorporated into UK. The model law aims to provide clarity on the situation but has avoided mentioning any particular part about the postal rule or when acceptance would be concluded for the sake of avoiding a clash with different countries. Hence, it is useful but to a limited extent. The law was adopted by Singapore when the Electronic Transactions Act29 was first enacted in 1998 and provides clear definitions for when dispatch and receipt is but no clarification about which factor would be used in deciding acceptance. The adoption of the law in Singapore could reflect the preference of interpretation where postal rule and electronic communications could be incorporated together and applied, as shown by a local case. One of the few cases on electronic contract formation in Singapore was Chwee Kin Keong v Digilandmall. com Pte Ltd30 that was primarily about unilateral mistake. Nevertheless, in the judgment, Rajah JC’s obiter provided careful analysis about the postal rule in a digital age and about its use in our local context. The obiter is by no means is a precedent for what rule is to be used or be adopted by the courts, but it critically reflected certain local attitudes on acceptance when using electronic mail. Rajah JC’s obiter began with an affirmation of contract law in the digital age that it would still continue to apply. “There is no real conundrum as to whether contractual principles apply to Internet contracts. Basic principles of contract law continue to prevail in contracts made on the Internet.”31

29 Electronic Transactions Act 1998 30 Chwee Kin Keong and others v Digilandmall.com Pte Ltd [2004] 2 SLR 594 31 ibid at [91]

Rajah JC highlighted the idea that emails could be regarded as non-instantaneous due to the form of delivery, “Furthermore, unlike a fax or a telephone call, it is not instantaneous. Emails are processed through servers, routers and Internet service providers,” 32 and observed a similar loss of control with postal mail, “Once an offer is sent over the Internet, the sender loses control over the route and delivery time of the message. In that sense, it is akin to ordinary posting.”33 As such, in this regard, Rajah JC concluded that there is a possibility for the rule to be applied in emails. This is further supported by academics like Reed, who has given two justifications for the rule to apply34: emails could have significant delays, similar to posting mail due to the need of having emails sent through packets that have to navigate their way through the various servers and that the offeror has already given their agreement that the offeree could accept through a third party as long as they have done whatever required for the acceptance to reach the offeror.35 Currently, these justifications have not been accepted, as the issue is still inconclusive. Nevertheless, these ideas with special reference to the system that electronic mail exists within bear resemblance to the post office in real life that have been supported by Mik36 as a strong reason for the rule to apply in modern concern. But it was noted that Rajah JC appeared to be inclined towards the general rule when he commented, “ There are, however, other sound reasons to argue … in favour of the recipient rule” and “The recipient rule is therefore more convenient and relevant in the context of both instantaneous or near instantaneous communications. Notwithstanding occasional failure, most e-mails arrive sooner rather than 32 ibid at [97] 33 ibid at [98] 34 Tan Harry SK, Electronic Commerce on the Internet: Formation of Online Contracts (Asia Business Law Review, No 17, July 1997) 35 Chris Reed, Computer Law (2nd ed, 1990) 261 36 Mik Eliza, The Effectiveness of Acceptances Communicated by Electronic Means, Or – Does the Postal Acceptance Rule Apply to Email (Journal of Contract Law, 26(1) 68-96, 2009)

ARTICLES

later.” From his words, it can be observed that Rajah JC seems to favour the view that there are reasonable reasons for postal rule to not apply. However, he acknowledged that the arguments for applying and not applying are both valid and that at this current point, there is no default rule in Singapore though was of the opinion that having one could bring about certainty.37 A recent case of 1L30G Pte Ltd v EQ Insurance Company Ltd38 brought about a discussion about the postal rule where a similar stance from the case of Digiland was reiterated with an added emphasis that the judge has to decide whether the parties have ultimately given their authorisation for the postal rule to be used. Lee Sieu Kin J found that both the postal rule could apply, “Postal rule could apply to emails as they are technically not instantaneous and the control is lost”39 and the receipt rule as well, “But same time for instant or near instant ones receipt rule could apply”.40 In the end, Lee J mentioned that the courts would provide flexibility and look at the circumstances to decide which rule would be appropriate, as seen in “For these reasons, the postal acceptance rule only applies in circumstances where both parties intend for it to apply.” 41 It can be observed from the case that the local courts’ stance on the issue remain to be undecided and meant to be permissive,42 which was what the model law by UNCITRAL set out to do. In general, the flexible approach allow for the courts to better analyse the contract and the parties’ intention, which despite might result in more uncertainty would better encompass all the complexities within current technology43 as highlighted by Mik. 37 Phang Andrew, Contract Formation and Mistake in Cyberspace – The Singapore Experience (The Singapore Academy of Law Journal, 17, 2005) 38 1L30G Pte Ltd v EQ Insurance Company Ltd [2017]

SGHC 242

39 ibid. at [29] 40 ibid. 41 ibid. at [30] 42 Chwee Kin Keong and others v DigilandMall.com Pte Ltd [2004] 2 SLR 594 at [92] 43 Mik Eliza, Certainty at last?: A “new” framework for electronic contracting in Singapore (Journal of International Commercial Law and Technology, Vol 8, No 3, 2013)

Conclusion The different attitudes held by both countries can be seen with the different adoption of rules. The UK courts have made strong distinction between using post and using means of instantaneous communications including emails where postal rule would not be applied unlike in Singapore, which seems to be more accepting of applying the postal rule and adopting a more flexible approach in deciding between the rules. Rajah JC classified emails to be noninstantaneous due to the possibilities of error while processing as compared to fax or a call as well as the need for the emails to pass through servers similar to a mail being sent through a postal service. This understanding is supported by the ideas by scholars such as Mik, Tan and Reed about how emails should be subjected to the postal rule as well due to the similarity of having a middleman that is delivering the acceptance. The difference in treatment could be due to Singapore’s unique position with information technology being a significant part of the city’s trade, resulting in a possible need to keep a more open approach. At the same time, the UK has established case law that would have to apply due to precedence. Nonetheless, a similarity can be drawn between Singapore’s approach of flexibility and UK’s recent case of Thomas v BPE that mentions sound business principles, where both countries acknowledged the difficulty in the technicalities of emails, resulting in the courts needed to take a more case-by-case analysis for a more holistic consideration.

ARTICLES

“Quenchless feuds” and ill will – should the tort of malicious prosecution of civil proceedings be recognised in Singapore? James Dylan Toh, University of Bristol Introduction In the recent case of Lee Tat Development 1(“Tat Development”), the Singapore Court of Appeal considered, inter alia, if the tort of malicious prosecution could be extended to maliciously instigated civil proceedings. In declining an extension, it departed from the majority position of the U.K. Supreme Court in Willers v Joyce.2 The decision in Willers v Joyce, not uncontroversial in itself, was decided by a 5:4 majority with compelling arguments from both sides. In light of these decisions, this article aims to examine the policy and doctrinal reasons underpinning the courts’ conclusions in both cases. By reflecting on the historical development of the tort and undertaking a comparative survey of English and Singaporean jurisprudence, this article considers that the tort of malicious prosecution should be extended to civil proceedings. This is chiefly because contemporary civil claims brought maliciously effect significant damage to defendants, and present remedies are either inadequate or nonexistent. Accordingly, it is submitted that simple justice should prevail, since extending the tort will not necessarily unleash the floodgates, nor bring about excessive uncertainty to the law. Historical development and substantive requirements To understand contemporary debate surrounding the scope of the tort of malicious prosecution, a brief historical reflection is 1 Lee Tat Development Pte Ltd v Management Corporation Grange Heights Strata Title Plan No. 301 [2018] SCGA 50 2 Willers v Joyce [2016] UKSC 43, hereafter known as ‘Willers’.

essential. The scope of the tort was historically “equivocal”.3 Although early caselaw appears limited to malicious criminal proceedings, endorsements of potential claims for malicious civil proceedings are found in Waterer v Freeman 4 and Atwood v Monger.5 Further, a passage from Coke’s Institutes 6 clearly states: “Where two or more conspire to harass any person by a false and malicious suit, whether criminally or civilly, it is a crime punishable by indictment, or the parties injured may sue for damages…” Halsbury’s Laws of England declares that a claimant seeking redress under this tort must prove 5 requirements7. 1. The process must have been without reasonable and probable cause. 2. The process must have been instituted maliciously. 3. The process must have terminated in the claimant’s favour. 4. The claimant must suffer one of three kinds of damage. 8 i. damage to the claimant’s fame ii. damage done to the claimant’s person, meaning life, limb or liberty iii. damage to the claimant’s property 5. The claimant must have been prosecuted. Yet the significance of these requirements must be understood in their context. Winfield and Jolowicz note that this ancient tort was shaped 3 Stephen Todd, ‘Liability for the Malicious Institution of Civil Proceedings’ (2017) 4 Journal of International and Comparative Law 123 4 Waterer v Freeman (1618) Hobart 266 5 Atwood v Monger (1653) Style 378 6 Co Inst 561, 562 (emphasis added) 7 Halsbury’s Laws (5th edn, 2015) vol 93, para 725 8 Halsbury’s Laws (5th edn, 2015) vol 93, para 733

ARTICLES

in 17th century England, an era with “no formal state system for investigation and prosecution”.9 Indeed, prosecution for criminal offences was for the most part initiated by private individuals exercising a public function. Since successful prosecution for criminal sentences entailed heavier sanctions (i.e. imprisonment, corporal punishment), the tort was formulated to inhibit maliciously instigated criminal proceedings and to avoid tainting the reputation of the system. However, four centuries of restructuring to the criminal prosecution system have rendered the tort in its ancient form “all but defunct”.10 Presently, criminal prosecution is the nearexclusive preserve of the public prosecutor’s office. Winfield and Jolowicz accordingly caution against the rigid application of “principles established under the old regime to present circumstances”, since doing so might lead to the obsolescence of the tort. Indeed, recent cases where malicious criminal prosecution was pleaded successfully are exceedingly rare – a notable example being the 1995 case of Martin v Watson. In that case, the defendant made repeated (false) complaints to the police with the sole aim of procuring charges of indecent exposure against a neighbour he despised. Though he did not technically conduct the prosecution, he was nonetheless found liable under the tort. The case demonstrates the difficulty with a literal or rigid application of this historical tort – a fact recognised by the House of Lords in allowing the claim. Further, as Lord Wilson appositely observed in Crawford Adjusters (Cayman) Ltd v Sagicor General Insurance (Cayman) Ltd11 (“Crawford Adjusters”), in this day and age there is a “much less chance” of being the victim of malicious criminal proceedings than of a civil action brought maliciously and without reasonable and probable cause.12 Thus, the dissolution of the private-individual/public-functionrole and the consequentially narrower scope of the tort

9 James Goudkamp, Winfield and Jolowicz on Tort (Edwin Peel ed, 19th edn, Sweet & Maxwell 2014) 636 10 Crawford Adjusters (Cayman) Ltd v Sagicor General Insurance (Cayman) Ltd [2013] UKPC 17 [121] (Sumption J) 11 [2013] UKPC 17 12 ibid [68] (Wilson J)

means that a fundamental reassessment of the civil/criminal distinction underpinning the tort is appropriate and welcome. Willers v Joyce Facts In Willers, the Appellant, Mr Willers was a former director of Langstone Leisure Ltd (“Langstone”). He was directed by Mr Gubay, the controller of a corporate group, which included Langstone, to pursue an action in Langstone’s name against a rival company - Aqua Design and Play Ltd (“Aqua”). This action (“the Aqua claim”) was abandoned on Gubay’s instructions shortly before trial. Subsequently, Langstone sued Willers for alleged breach of contractual and fiduciary duties in causing it to incur costs in pursuing the Aqua directors. Before trial, Langstone gave notice of discontinuance and was ordered to pay Willers’ costs. In the present case, Willers alleged that the claim brought against him by Langstone was part of a campaign by Gubay to do him harm. The Supreme Court did not dispute that the present claim included “all the necessary ingredients for a claim of malicious prosecution”; the question for the court was whether such an action existed in English civil law. Ruling In deciding whether or not an action for the tort of malicious prosecution of civil proceedings was sustainable under English law, the first instance judge in Willers v Joyce was faced with two conflicting opinions. In Gregory v Portsmouth City Council,13 the House of Lords held that English law did not recognise the tort of malicious prosecution in relation to civil proceedings. However, in Crawford Adjusters, the Privy Council by a 3:2 majority upheld the claim for malicious prosecution of civil proceedings. The judge allowed a leapfrog appeal to the 13

Gregory v Portsmouth City Council [2000] 1 AC 419 (HL)

ARTICLES

Supreme Court, which, by a 5:4 majority, endorsed the findings of the Privy Council in Crawford Adjusters and upheld Willers’ claim. Lord Toulson, who delivered the majority judgement in Willers v Joyce, laid down the elements of the new tort.†† 1. Civil proceedings must have been instituted against the claimant 2. These proceedings must have ended in the claimant’s favour, whether by settlement, abandonment or judgement 3. The proceedings must have been procured by the defendant, although it is unnecessary that the defendant brought the proceedings him or herself 4. The proceedings must have been without reasonable and probable cause 5. The proceedings must have been malicious, meaning the defendant has “deliberately misused the process of the court”.14 Malice is entwined with, but distinct from, the fourth requirement. 6. Loss must be suffered as a result of the prosecution of the proceedings. Lee Tat Development In Tat Development, the Singapore Court of Appeal was asked to consider whether or not the tort of malicious prosecution could be extended to civil proceedings generally. In this present claim, the claimant, Lee Tat, argued that it was the target of malicious prosecution by MCST in respect of two separate claims, one in 2005 and the other in 2008. These underlying claims concerned a disputed right of way, and were part of a bitter legal dispute spanning several decades, reflecting the longstanding animosity between the parties. In those prior proceedings, a property developer, Lee Tat, sought to deny the purchasers of a neighbouring condominium development a right of way over a narrow strip of land owned by Lee Tat. In †† These requirements are quoted verbatim from Goudkamp’s succinct analysis: James Goudkamp, ‘A tort is born’ (2017) 167 NLJ 7753 14 Willers (n2) 55 Toulson J

the 2005 action, the Court of Appeal held that the residents of the condominium had a right of way. However, in 2008, the Court of Appeal overruled its earlier decision, holding that the right of way had in fact been extinguished. Unlike the appellant in Willers v Joyce, the appellants in Tat Development did not satisfy the requirements for a hypothetical claim in malicious prosecution. This was because Tat Development failed to demonstrate that the underlying actions were without reasonable and probable cause, or that malice was present in those disputed actions.15 The Court nonetheless issued detailed judgement on this previously undecided point of law, ruling that the tort of malicious prosecution should not be extended to civil proceedings generally. Reasoning, implications and recommendations In the U.K. Supreme Court as well as in the Singapore Court of Appeal, strong policy and doctrinal reasons were raised on both sides of the argument. Given the extraordinary time, costs and complexity associated with modern civil/commercial proceedings, the decision to accept or decline an extension of this tort has important legal, policy and economic ramifications. On consideration, this article submits that extending the tort of malicious prosecution to civil proceedings is more favourable for the reasons explored below. i. Distinction between civil and criminal processes In declining an extension of the tort to civil proceedings, the Singapore Court of Appeal cited “essential differences” between the “character and consequences”16 of the criminal and civil processes. It was held in Tat Development that the character of criminal prosecution was such as to punish a public wrong, whereas a civil claim to vindicate a private right. Yet, to draw such a distinction arguably overlooks the improper motive behind the claim. In both 15 16

ibid [129] ibid [86]

ARTICLES

instances, the defendant deliberately advances a claim he knows to be baseless. The crux of genuinely founded malicious prosecution claims is thus not about the vindication of public or private wrongs, but about the mala fide use of the court’s process. Indeed, inasmuch as a critical feature of the tort lies in the defendant’s intentional abuse of state power, this “[applies] as much to civil proceedings”.17 In essence, both are ultimately about maintaining the rule of law18, and so to use civil proceedings for some collateral purpose is just as much an abuse. Thus, for reasons illustrated here and above, the distinction between the nature of civil/ criminal processes in the context of malicious prosecution is ultimately “unconvincing”.19 A related point advanced in Tat Development concerns the graver consequences of successful criminal prosecution compared to civil proceedings20 – an accused person may “have his property seized, liberty deprived”, and in Singapore, his “life taken as a result of his conviction”.21 Thus, the Singapore Court of Appeal held that even if the criminal case were terminated in favour of the defendant, the psychological stress inflicted during the interim coupled with the “public character” and associated stigma of criminal proceedings evinces fundamental differences between the civil and criminal process and justifies restricting the tort to the realm of the former. However, the consequences of civil litigation in contemporary society are arguably no less onerous. In cases of serious allegations of wrongdoing and fraud, defendants may be forced to engage in costly litigation lasting several years. 22 Given the substantial

17 Todd (n3) 124 18 John Murphy, ‘Malice as an Ingredient of Tort Liability’ (2019) CLJ 1, 23 19 Tom Ng, ‘The torts of malicious prosecution and abuse of legal process’ (2014) 130 LQR 43, 46 20 Tat Development (n1) [88] 21 Tat Development (n1) [89] 22 K Garbett and M Karagoz, ‘Shaping the tort of malicious prosecution of civil claims’ (2018) 168/7776 NLJ 11, 12 <https://www.newlawjournal.co.uk/content/shaping-tortmalicious-prosecution-civil-claims> accessed 25 March 2019

resources expended in the process, this would surely constitute significant interference with the defendant’s property, rendering the consequences of malicious civil suits closer to that of successful criminal prosecution than might appear at first glance. Meanwhile, substantial reputational damage may be caused by the “immediate and irreparable” impact of the false allegations.23 Indeed, negative publicity is an issue faced equally by defendants of a civil suit – in Crawford Adjustors, the individual who initiated the malicious civil suit was also “instrumental in alerting a journalist” working for a local newspaper to the false allegations.24 Furthermore, given that psychological distress is subjective, it may well be that a defendant facing a malicious civil suit suffers equal distress to that faced by an accused in a malicious criminal suit. Thus, the traditional explanation in Quartz Hill that ordinary civil actions do not “naturally or necessarily involve…injury to reputation”25 is “no longer plausible”26 A more fundamental flaw with the suggestion that defendants in malicious criminal and civil suits suffer different consequences is that such an assumption is premised on an means of identifying and quantifying valid kinds of damage (pecuniary, reputational, etc, as opposed to more trivial forms of emotional damage such as hurt feelings and social stigma) suffered by the defendants. The legal validity of certain types of damage over others is by no means self-evident and attempting to draw distinctions could open up a policy quagmire and introduce arbitrariness in the law. Equally, from an evidential perspective, the absence of coherent principles to quantify such damage precludes any easy conclusions about the different nature malicious of criminal/civil suits. At any rate, even assuming such problems could be overcome, and that defendant of malicious criminal suits did in fact suffer a unique kind of injustice/damage, their greater/unique suffering 23 24 25 26

Garbett and Karagoz (n22) 11 Crawford Adjusters (n10) [19] Todd (n3) 123 Gregory (n13) 428 (Steyn J)

ARTICLES

per se would not be a principled reason to deny a remedy to the injustice faced by defendants of malicious civil suits.

the United States.34 Equally, since a successful defendant of a groundless civil suit would be awarded costs, this would minimise the financial damage suffered.

ii. Necessity/Duplication of remedies argument A related argument advanced is that defendants of malicious civil suits may rely on other existing “legal mechanisms” or civil procedure rules27 – principally costs awards.28 Accordingly, extending the tort of malicious prosecution to civil proceedings is unnecessary since it would create a “duplication of remedies”.29 However, in reality, these remedies have proven inadequate or completely non-existent. As the Supreme Court in Willers v Joyce noted, the principal reason given by Lord Steyn in Gregory for not extending the tort was that “any manifest injustice arising from groundless and damaging civil proceedings” was already sufficiently protected under “other torts”.30 Yet, the facts in Gregory, Crawford Adjusters and Willers v Joyce plainly disprove this – in all cases the claimant had in fact no other means of legal recourse. Thus, Lord Steyn’s proposition that the existence of “other torts”31 provides an adequate remedy is “hardly true”32 since no alternative remedies exist. The duplication argument should thus be rejected. Alternatively, it was suggested in Tat Development33 and by Lord Neuberger in Crawford Adjusters that costs awards provided a mechanism for addressing the injustice suffered. In his dissenting judgment in Crawford Adjusters, Lord Neuberger concluded that the availability of the tort in the United States was due to the fact that in the U.K. (and indeed in Singapore) the losing party paid the winning party’s costs and this was a sufficient deterrent to baseless claims, whereas no equivalent mechanism existed in

27 28 29 30 31 32 33

Tat Development (n1) [121] Tat Development (n1) [120] Willers (n2) [47] Willers (n2) [47] Gregory (n13) [422] Todd (n3) 124 Tat Development (n1) [120]

However, as Todd notes, this “ignores the reality” that a costs award is “unlikely to be an adequate compensation” for the actual costs incurred.35 In Willers v Joyce, the defendant of the malicious suit incurred £3.9m of legal expenses, yet was only awarded £1.7m. The result is that where innocent defendants are forced to engage in lengthy and costly litigation, he is bound to suffer substantial injustice. As the Singapore Court of Appeal conceded, even an award of costs on an indemnity basis “does not furnish a complete remedy” but only a “partial” one.36 Thus, to echo the words of Lord Toulson, the notion that a costs order “necessarily made good the injury” caused by malicious prosecution was “almost certainly a fiction”.37 iii. Floodgates/finality argument Another policy argument advanced by the Singapore Court of Appeal and the minority in Willers v Joyce is that allowing a tort of malicious prosecution of civil proceedings might encourage substantial “satellite litigation”††† and in so doing undermine the principle of finality in the law.38 These concerns are speculative. In Crawford Adjusters, Lord Kerr observed that evidence of an “avalanche of litigation” was “conspicuously missing”.39 Equally, the Singapore Court of Appeal acknowledged that empirical analysis of Ontarian jurisprudence demonstrated that extending the tort had not generally opened the floodgates of litigation.40 Thus, inasmuch as one could argue that extending the tort might encourage further 34 35 36

Crawford Adjusters (n10) [165] Todd (n3) 124 Tat Development (n1) [120]

37 Willers (n2) [58] ††† meaning a separate lawsuit connected to and concerning a main lawsuit. For example, a trial adjudicating on the admissibility of certain evidence in a criminal case. 38 39 40

Tat Development (n1) [105] Crawford Adjusters (n10) [108] Tat Development (n1) [114]

ARTICLES

litigation, this argument can be easily turned on its head. This is because it might equally be said to deter those attempting to misuse the legal system for “personal vendettas and dishonest claims”.41 Hence, the speculative nature of the floodgates argument precludes it from bearing any real normative force. Additionally, although at first glance the tort may undermine the principle of finality in the law, this is in practice unlikely to be true. Claims for malicious prosecution do not involve a review of facts or merits of the first case, nor do they constitute a “collateral attack”42 on the outcome of the first proceedings. Accordingly, an action for malicious prosecution of civil proceedings cannot accurately be said to undermine the public interest in finality of the law.

have a “chilling”46 or “deterrent”47 effect on honest claimants, who might fear a retaliatory malicious prosecution suit if their original claim fails. Given that access to justice is the hallmark of many democracies, this argument is surely deserving of “considerable respect”.48 Yet, like the floodgates argument, is ultimately speculative, and was rejected for this reason in relation to the malicious prosecution of criminal proceedings in Savile v Roberts49 and Martin v Watson.50 Equally, the majority in Crawford noted that Lord Neuberger’s analysis of American jurisprudence presented “no evidence” of the supposed “chilling effect” on honest claimants,51 presumably because there was none. Thus, though admirable in spirit, the argument loses force in the absence of empirical evidence. v. Simple justice

At any rate, establishment of the tort of malicious prosecution of civil proceedings will be very difficult, requiring proof of a negative proposition, the “most difficult of evidential requirements”.43 Given the onerousness in proving liability under this tort, it will likely only lie in “exceptional cases”44 and fears of opening the floodgates to fresh litigation about litigation are exaggerated. Further safeguards could also be added to deter the overzealous litigant by extending liability to their solicitors and barristers. This might appear extreme, but Mallen argues that in the United States, where contingent fee arrangements are commonplace and litigation has been characterised as a “national pastime”,45 such preventative steps can function as important deterrents. iv. Deterrence argument Another argument advanced in declining an extension of the tort is that doing so would

41 42 43 44 NLJ 12 45 1880

Garbett and Karagoz (n22) 12 Willers (n2) [46] Crawford Adjusters (n10) [109] (Lord Kerr) James Goudkamp, ‘A tort is born’ (2017) 167/7753 Bickel v Mackie (1978) 447 F Supp 1876 (N.D. Iowa)

Overall, the most compelling argument in favour of allowing an extension of the tort is that the law should provide a remedy where injustice has been suffered. As the “cornerstone” of any system of justice,52 the dictates of “simple justice”53 thus impose a high standard that should not yield to anything short of “strict and cogent”54justifications. Where the injustice is substantial, and has been facilitated by the legal system, this is all the more true. Thus, the equitable maxim ubi jus ibi remedium – where there is a wrong, there must be a remedy – acquires particular resonance in this context. Conclusion That the legal landscape is radically different now than it was 400 years ago is undeniable. The establishment of the Public Prosecutor’s office has consolidated criminal proceedings 46 47 48 49 50 51 52 53 54

Tat Development (n1) [117] Willers (n2) [45] Crawford Adjusters (n10) [72] Savile v Roberts [1698] 1 Ld Raym 374 Martin v Watson [1996] 1 AC 74 Crawford Adjusters (n10) [72] Jones v Kaney [2011] UKSC 13 (Dyson J) Willers (n2) [57] Jones (n49)

ARTICLES

and rendered the tort in its original capacity obsolete. At the same time, the volume and complexity of civil claims have increased, as has media exposure of such claims. Against this backdrop a fundamental reassessment of this ancient tort of malicious prosecution is well overdue. Where objections to the extension of this tort are premised on anachronistic assumptions, they should be critically and cautiously applied. It is clear that malicious civil suits in today’s legal landscape have the potential to cause irreparable damage and that existing remedies are inadequate to this end. At the same time, forward-looking policy concerns are no doubt well-intentioned yet inherently speculative. Thus, as Todd asserts, these concerns may prove “largely unfounded or at least manageable”.55

anticipated that the issue will be addressed with equal judiciousness.

It is possible that the introduction of this tort might cause uncertainty in its application. The question of what damages will be recoverable, as well as defining what constitutes malice or a “reasonable and probable cause” will all be questions that the courts must answer. Equally, it seems possible that the tort may extend to include a malicious defence. But these challenges are neither insurmountable, nor are they principled grounds to deny a remedy to one who has suffered a substantial injustice facilitated by the law. Ultimately, the careful path traced by the Court of Appeal in Tat Development is undeniably apposite. On the facts of that case, there was clearly no claim to be had. Given the recent extension of the tort in the U.K. to include civil proceedings and the implicit acceptance in other common law jurisdictions such as Hong Kong56, Australia57 and New Zealand,58 Singaporean courts will be afforded ample opportunity in coming years to analyse the implications of such an extension. Should a future opportunity arise for the issue to be reconsidered, it is 55 Todd (n3)130 56 Chua, Grace Gonzales v Sobrevilla, Rhennie Boy Fernandez [2017] HKDC 1029 57 Little v Law Institute of Victoria [1990] VR 257 58 Robinson v Whangarei Heads Enterprises Ltd [2015] 3 NZLR 734

ARTICLES

In Light Of The Technical Impracticality Of The Right To Erasure, What Answers Can Actor-Network Theories Provide? Justin Kwik, University of Bristol Introduction From representation as “Oblivion”1 to “Delisting”2 to “Erasure”,3 the Right To Be Forgotten (RTBF) has long reflected a societal want to control one’s identity and informational flows. In the UK, this is represented by the GDPR’s Article 17 “right to erasure”,4 which this essay argues is ill adapted to robustly protect an individual’s RTBF in an Artificial Intelligence (AI) era. This right is given direct effect in the UK through s.2(1) of the European Communities Act 1972, which directly applies European Union Regulations in the UK. In Singapore, this right does not have any explicit legal grounding, and its closest representation may be found in the Personal Data Protection Act 2012 (PDPA). First, this essay will examine the legal and conceptual history of the right, as well as the necessary definitions to contextualise a contemporary technological environment that features AI and Blockchain. Secondly, this essay will attempt to illustrate the ways in which the GDPR’s right to erasure fails to meaningfully protect an individual’s personal data rights, particularly in: (i) Machine Learning AI, (ii) Blockchain systems, and (iii) jurisdictions outside the EU. Finally, this essay will frame the right to erasure through the perspectives of different cyberspace

1 S. Creese, In Search Of Oblivion? How The ‘Right To Be Forgotten’ Could Undermine Web-based Corpora” (2015) Procedia – Social and Behavioral Sciences, vol. 198, p. 95-102 2 M.D. Rosnay, A. Guadamuz, “Memory Hole Or Right To Delist?” (2017) RESET, Association Recherches en Sciences Sociales Sur Internet, 6 | 2017, DOI: 10.4000/ reset.807 3 EU General Data Protection Regulation (GDPR): Regulation (EU) 2016/679, Article 17 4 ibid.

regulation theories, using them to assert firstly that the aforementioned problems were caused by a lack of recourse to said theories, and secondly that a combination of Murray’s and Vranaki’s Actor-Network Theories may provide valuable insights into potential solutions for cyber regulation. We then review Singapore’s recently published proposed framework for AI governance in light of these theories. 1. Contextual Information 1.1 The Right To Be Forgotten From the outset, this essay notes the phrase, the “Right To Be Forgotten” (RTBF), is often interchangeably used by academics5 and professionals6 alike to refer to two separate rights: a “right to de-listing” and a “right to erasure” respectively. Although some authors7 interpret the right to erasure” as simply having the same powers as the right to de-listing, albeit with a larger scope, this essay concurs with the Article 29 Working Party,8 Ambrose,9 Bunn10 5 L. Edwards, “Data Protection: Enter The General Data Protection Regulation” in L. Edwards, Law, Policy And The Internet (Hart Publishing, Oxford 2019 6 IAPP, “IAPP-EY Annual Privacy Governance Report 2017” (2017), International Association Of Privacy Professionals, available at https://iapp.org/media/pdf/ resource_center/IAPP-EY-Governance-Report-2017.pdf, accessed on 21st March 2019 7 L. Edwards, “Data Protection: Enter The General Data Protection Regulation” in L. Edwards, Law, Policy And The Internet (Hart Publishing, Oxford 2019) p. 107 8 J. Ausloos, B.V. Alsenoy, “Implementing “The Right To Be Forgotten”: The Article 29 Working Party Speaks Up” (2017) LSE Media Project, available at http://eprints.lse.ac.uk/80436/1/ Implementing%20the%20%E2%80%9Cright%20to%20 be%20forgotten%E2%80%9D_%20the%20Article%20 29%20Working%20Party%20speaks%20up%20_%20LSE%20 Media%20Policy%20Project.pdf, accessed on 22nd March 2019 9 M. L. Ambrose, J. Ausloos, “The Right To Be Forgotten Across The Pond” (2013) Journal Of Information Policy, Vol. 3, p. 1-23 10 A. Bunn, “The Curious Case Of The Right To Be Forgotten” (2015) Computer Law & Security Review, Volume

ARTICLES

and Graux11 in stating that it is imperative not to conflate these two separate rights, as they have accord drastically different obligations and implications, especially in today’s digital environment. The legal basis of the RTBF was the “Le droit a l’oubli”, or “Right to Oblivion”,12 which emerged with the delineation of “personality rights” in continental European countries like Spain, Switzerland and France in the 19th century.13 Founded on the general principle of a right to “dignity” and a “private life”,14 this right primarily entitled individuals to have their published (typically judicial) history erased once a “substantial” amount of time had elapsed so as to remove it from the sphere of public interest15 and was exercised in Germany (Lebach)16 and Switzerland (Societe Suisse).17 There has been some debate over the conceptual basis of this right: while the RTBF is most commonly referred to as a right to privacy,18 Andrade19 and Gutwirth20 have insisted it is instead a “right to identity”, a view that MayerSchönberger reiterates.21 This essay finds 31, Issue 3, June 2015, Pages 336-350 11 H. Graux, J. Ausloos, P. Valcke, “ICRI Working Paper Series: The Right To Be Forgotten In The Internet Era” (2012) Interdisciplinary Centre For Law And ICT, K.U.Leuven, ICRI Working Paper 11/2012 12 H. Deschenaux, P.H Steinauer, Personnes Physiques et Tutelles (Berne, 2001) p. 159 – 173 13 ibid. 14 R. Weber, “The Right To Be Forgotten: More Than A Pandora’s Box?” (2011) JIPITEC 120, para.1 15 A. Bunn, “The Curious Case Of The Right To Be Forgotten” (2015) Computer Law & Security Review, Volume 31, Issue 3, June 2015, Pages 336-350 16 BVerfGE 35, 202., Bundesverfassungsgericht (First Division) 5 June 1973 17 X v. Societe Suisse de Radio et de Television (1983) BGE 109 11 353 18 A. Giurgiu, “Challenges Of Regulating A Right To Be Forgotten With Particular Reference To Facebook” (2013) 7 Masaryk U. J.L. & Tech. 361 19 N. N. G. Andrade, “Oblivion: The Right To Be Different From Oneself - Reproposing The Right To Be Forgotten” (2012) The Ethics Of Memory In A Digital Age, p. 65-81 20 S. Gutwirth, “Beyond Identity?” (2009) Identity In The Information Society, Vol. 1, No. 1, p. 122-133. 21 V. Mayer-Schönberger, Delete, The Virtue Of Forgetting In The Digital Age (Princeton University Press, Oxford, 2009) p. 169

favour with Ausloos’s22 middle ground of a proprietary right to privacy and control of one’s own information, as it matches the right’s “erga omnes” nature to restrict access.23 However, regardless of whichever conceptual basis one favours, the RTBF had an undeniably strong grounding in European personality rights, and was seen as a “justifiable limit to the freedom of expression”.24 As Gasser and Movius both note, this “precautionary”,25 rights-based regulation is typical of European Commission legislation regarding technology, and is starkly contrasted against the “laissez-faire” style of American innovation that prizes innovation and applies technological state-of-the-art.26 This essay will explore this contrast in cyber-regulatory theory, as well as Singapore’s regulatory approach, in sections 1.3 and 4.3. 1.2 United Kingdom Prior to the GDPR’s release and in the absence of an explicit RTBF, the ECJ in Google Spain v. AEPD, Mario Cojesta27 relied on an expansive interpretation of the EU’s 1995 Data Protection Directive28 to “invent”29 a right to “de-listing”. 22 J. Ausloos, “The ‘Right To Be Forgotten’ – Worth Remembering?” (2012) Computer Law & Security Review, Vol. 28, Issue 2, Pages 143-152 23 B.J. Koops, “Forgetting Footprints, Shunning Shadows: A Critical Analysis Of The Right To Be Forgotten” (2011) SCRIPTed, vol. 8, no. 3, pp. 229-256 24 E. Gratton, J. Polonetsky, “Droit A L’oubli: Canadian Perspective On The Global ‘Right To Be Forgotten’ Debate” (2017) Colorado Technology Law Journal, Vol. 15, Issue 2, p. 1 - 45 25 Infocomm Media Development Authority (IMDA), “Singapore’s Governing Framework for Artificial Intelligence”, available at https://www.nytimes.com/paidpost/imda/ singapores-governing-framework-for-artificial-intelligence. html, accessed 21st June 2016 26 L. Movius, N. Krup, “U.S and EU Privacy Policy: Comparison of Regulatory Approaches” (2009) International Journal of Communication, Vol. 3, p. 169-187 27 Case C-131/12, Google Spain SL and Google Inc. v Agencia Española de Protección de Datos (AEPD) and Mario Costeja González. (2014) 28 Directive 95/46/EC, referred to in Case C-131/12, Google Spain SL and Google Inc. v Agencia Española de Protección de Datos (AEPD) and Mario Costeja González. (2014) 29 S. Breheny, “The Right To Be Forgotten Online Sets a Dangerous Precedent” (2014) The Age, available at http:// www.theage.com.au/comment/the-right-to-be-forgotten-

ARTICLES

This controversial decision drew a slew of different criticisms. For example, Bunn argues that the ECJ pointedly “circumvented” any meaningful deliberation of the general public interest argument elucidated by the Advocate General.30 Also, the House of Lords accused the ECJ of failing to account for the (then) “current state of communications service provision” and not interpreting the directive “stringently” enough.31 Tréguer32 asserted that the right to delisting deviated so much from the le droit a l’oubli that it became a “far cry” from an “actual right to be forgotten” such that the data could still be found indexed elsewhere. Indeed, this became the CNIL’s contention with Google in the recent case of CNIL v Google.33 Although this right has had a significant practical impact (over 100 000 requests in 3 months, and 3 million deleted URLs by Jan 201934) and is integral to the European understanding of the RTBF, this essay aims to dissect the GDPR’s right to erasure and its short-comings. Today, the latest iteration of the RTBF is the “right to erasure”, enshrined in Article 17 of the GDPR.35 Originally proposed by Viviane Reding in 2010,36 this right was intended to

online-sets-a-dangerous-precedent-for-freedom-of-speech20140715-zt7w7.html, accessed 20th March 2019 30 A. Bunn, “The Curious Case Of The Right To Be Forgotten” (2015) Computer Law & Security Review, Volume 31, Issue 3, June 2015, Pages 336-350 31 Parliament, House of Lords “2nd Report - EU Data Protection Law: A ‘Right To Be Forgotten’? – European Union Committee | Chapter 5: The Views Of The Committee” (2014) HL Paper 40 32 F. Treguer, “Right To Be Forgotten: With Free Expression Under Threat, Europe Needs A ‘Marco Civil Moment’” (2014) Global Voices, available at https:// globalvoices.org/2014/09/11/right-to-be-forgotten-with-freeexpression-under-threat-europe-needs-a-marco-civil-moment/, accessed 21st March 2019 33 M. Finck, “Google v. CNIL: Defining The Territorial Scope Of European Data Protection Law” (2018) Oxford, Faculty of Law Blog, available at https://www.law.ox.ac.uk/ business-law-blog/blog/2018/11/google-v-cnil-definingterritorial-scope-european-data-protection-law, accessed on 23rd March 2019 34 Google Transparency Report, available at https:// transparencyreport.google.com/eu-privacy/overview, accessed on 22nd March 2019 35 EU General Data Protection Regulation (GDPR): Regulation (EU) 2016/679, Article 17 36 V. Reding, “The Upcoming Data Protection Reform

re-enact la droit a’ l’oubli as a “well-meaning attempt to allow children to hide the mistakes of their youth…on social networks”37 by allowing individuals to erase their data, subject to certain conditions. This essay agrees with Zanfir’s finding that this right was merely an interpretation and expansion of existing DPD rights (art. 1238 & 1439), albeit an expansive one40. Although academics such as Koops41 and Rosen warned of its potential to become a “grave threat to free speech”,42 the right was not made “automatic”43 as the right to de-listing was. Rather, it came with a comprehensive set of conditions by which individuals might exercise their right: by withdrawing consent when: (i) the personal data is no longer necessary for its original purpose (ii) objecting to a lack/ expiration of a legitimate interest (iii) the data has been unlawfully processed or (iv) the consent was collected from a child.44 Alongside this right, the GDPR also mandated compulsory notification of erasure, and separate rights to “restrict”45 and “object to”46 processing - the ICO recommends the former as an “alternative to data erasure”.47 These clauses must all be For The European Union” (2010) International Data Privacy Law, Vol. 1, Issue 1, p. 3-5 37 L. Edwards, “Data Protection: Enter The General Data 38 EU Directive 95/46/EC, Article 12 39 ibid., Article 14 40 G. Zanfir, “Tracing The Right To Be Forgotten In The Short History Of Data Protection Law: The New Clothes Of An Old Right” (2015) Reforming European Data Protection Law, available at https://www.academia.edu/6103947/ Tracing_the_right_to_be_forgotten_in_the_short_history_ of_data_protection_law_The_new_clothes_of_an_old_right, accessed 22nd March 2019 41 B.J. Koops, “Forgetting Footprints, Shunning Shadows: A Critical Analysis Of The Right To Be Forgotten” (2011) SCRIPTed, Vol. 8, No. 3, pp. 229-256 42 J. Rosen, “Free Speech, Privacy, And The Web That Never Forgets” (2011) 9. J., Telecomm 43 A. Bunn, “The Curious Case Of The Right To Be Forgotten” (2015) Computer Law & Security Review, Volume 31, Issue 3, June 2015, Pages 336-350 44 EU General Data Protection Regulation (GDPR): Regulation (EU) 2016/679, Article 17 45 ibid., Article 18 46 ibid., Article 21 47 Information Commissioner’s Office, “Guide To The General Data Protection Regulation (GDPR), Individual Rights”, available at https://ico.org.uk/for-organisations/ guide-to-data-protection/guide-to-the-general-data-protectionregulation-gdpr/individual-rights/right-to-object/, accessed on

ARTICLES

read in light of the principles of lawfulness,48 adequate transfer protection,49 purpose50 and storage51 limitation, and data minimisation,52 which provide a first-level safeguard of reducing data retention. As Mayer-Schönberger53 and Villaronga et. al.54 reiterate, applying a human understanding of “forgetting” to erasure is misguided – data erasure in algorithmic memory requires explicit removal or inaccessibility.55 In lieu of direct guidance from the ECJ or EC in interpreting “erasure” as specified in the GDPR,56 the ICO directs that while live systems must have their data erased immediately, backup data must be “put beyond use”57 if it must remain within the controller’s data systems for technical or practical reasons. Based on their guidance on the DPD,58 this requires that controllers can no longer use the data to inform a decision in a manner that affects the individual; does not give any other organisation access to that data; surrounds the data with the appropriate technical/organisational security, and commits 21st March 2019 48 EU General Data Protection Regulation (GDPR): Regulation (EU) 2016/679, Article 5(1) 49 ibid., Chapter V 50 ibid., Article 5(1)(b) 51 ibid., Article 5(1)(e) 52 ibid., Article 5(1)(c) 53 V. Mayer-Schönberger, Delete, The Virtue Of Forgetting In The Digital Age (Princeton University Press, Oxford, 2009 54 E. Villaronga, P. Kieseberg, T. Li, “Humans Forget, Machines Remember: Artificial Intelligence And The Right To Be Forgotten” (2018) Computer Law & Security Review, Vol. 34, Issue 2, p. 304-313 55 E. Esposito, “Algorithmic Memory And The Right To Be Forgotten On The Web” (2017) Big Data & Society, p. 1-11 56 I. Wilson, “Forthcoming ‘Right To Be Forgotten’ Cases And The Interplay With The GDPR” (2018), available at https://www.brettwilson.co.uk/blog/forthcoming-rightforgotten-cases-interplay-gdpr/, accessed: 23rd March 2019 57 Information Commissioner’s Office, “Guide To The General Data Protection Regulation (GDPR), Individual Rights”, available at https://ico.org.uk/for-organisations/ guide-to-data-protection/guide-to-the-general-data-protectionregulation-gdpr/individual-rights/right-to-object/, accessed on 21st March 2019 58 Information Commissioner’s Office, “Deleting Personal Data: Data Protection Act” (2014) 20140226, Ver 1.1, available at https://ico.org.uk/media/for-organisations/ documents/1475/deleting_personal_data.pdf, accessed on 23rd March 2019

to the permanent deletion the data when possible.59 The International Data Consortium (IDC) provides several methods for effective data “sanitisation”, inter alia: (i) physical destruction of the hard disk (ii) cryptopgraphic erasure (whereby the encryption key for the data is removed, rendering it inaccessible) or (iii) data erasure through overwriting software. However, it is unclear whether this standard can be applied to all EU member states.60 Beyond this, anonymisation, pseudoanonymisation and functional encryption have been proven to create sufficient de-identification to constitute functional “erasure”, 61 but have been dismissed as lacklustre responses to erasure requests, due to the damage it does to the accuracy and capability of AI actants trained on affected databases.62 These latter options will be discussed later. 1.3 Singapore In Singapore, there exists no explicit representation of the RTBF,63 although the rights to identity and privacy are protected in the PDPA. This includes s.16, which enshrines an individual’s right to withdraw consent and prevent the further collection, use and disclosure of their personal data by a data controller.64 This provision mirrors the GDPR’s article 17 right to erasure consent requirements. Likewise, 59 ibid. 60 International Data Sanitization Consortium, “Data Sanitization Regulations”, available at https://www. datasanitization.org/data-sanitization-regulations/, accessed 23rd March 2019 61 A. Cavoukian, D. Castro, “Big Data And Innovation, Setting the Record Straight: Anonymization Does Work” (2014) Information and Privacy Commissioner, Ontario, Canada and Information Technology and Innovation Foundation, available at http://www2.itif.org/2014-big-datadeidentification.pdf, accessed 24th March 2019 62 E. Villaronga, P. Kieseberg, T. Li, “Humans Forget, Machines Remember: Artificial Intelligence And The Right To Be Forgotten” (2018) Computer Law & Security Review, Vol. 34, Issue 2, p. 304-313 63 B. Zeller, L. Trakman, R. Walters, S. Rosadi, “The Right To Be Forgotten – The EU and Asia Pacific Experience (Australia, Indonesia, Japan and Singapore)” (2019) European Human Rights Law Review, Vol. 1, p. 23 64 Personal Data Protection Act (PDPA), S.16(1)

ARTICLES

s.22 PDPA mirrors the GDPR’s article 16 right to rectification, which enables individuals to better represent themselves online.65 Although PDPA s.25 requires controllers to destroy and de-identify personal data once its retention is no longer necessary for legitimate business/legal purposes,66 this is more similar to GDPR article 5.1(c) principle for data minimisation than a full right to erasure. This lack of similar data rights is explained by Zeller et. al. as the result of distinctly different legal and economic cultures that underpin European and Singaporean legislation.67 As Kumar68 and Chew69 both elucidate, the history and nature of the developing economies in Asia over the 20th century has led to a tradition of strong government intervention in the pursuit of economic growth, taking precedence over the concept of fundamental human rights such as privacy. Governmental perceptions of the otherwise “unchecked freedoms of the individual” led to “restrictive limits” created to ensure economic growth were not impeded, 70 best encapsulated by a quote from the former Prime Minister Lee Kuan Yew: “[Singapore] wouldn’t be here, we would not have made economic progress, if [the government] had not intervened on very personal matters – who your neighbour is, how you live, the noise you make, how you spit, or what language you use.”71

65 ibid., S. 22 66 ibid., S. 23 67 B. Zeller, L. Trakman, R. Walters, S. Rosadi, “The Right To Be Forgotten – The EU and Asia Pacific Experience (Australia, Indonesia, Japan and Singapore)” (2019) European Human Rights Law Review, Vol. 1, p. 23 68 S. Kumar, S. Siddique, “Beyond Economic Reality: New Thoughts On The Growth Triangle” (1994) Southeast Asian Affairs, p. 47-56 69 M. Chew, “Human Rights in Singapore: Perceptions and Problems” (1994) Asian Survey, Vol. 34, No. 11, p 933-948 70 ibid. 71 Lee Kuan Yew, quoted in Ministry of Foreign Affairs Overseas Missions, “The Statesman Who Talked Tough” (2015), available at https://www.mfa.gov.sg/content/mfa/ overseasmission/ abu_dhabi/embassy-events-and-pressstatements/2015/201503/the-statesmanwho-talked-tough-. html, referenced in S. Chesterman (ed.), Data Protection Law In Singapore: Privacy and Sovereignty in an Interconnected World

Likewise, Chesterman notes that while privacy and human rights lie at the core of European data protection legislation, Singaporean data protection has focussed primarily on economic issues and the proportionate balance of the needs of businesses against an individual’s rights72 – indeed, the PDPA was drafted primarily with Singapore’s economic “competitiveness” and “processing services” in mind,73 and does not even use the word privacy once.74 However, Kumar also notes that such a model is not a “static design”.75 As Singapore approaches economic maturity and develops its own culture and counterculture towards fundamental rights, there is a growing need for Singapore to acknowledge the rights of its citizens.76 As an economic example, both the US (General System of Preferences, Most-Favoured Nation status, etc.) and the EU (GDPR Adequacy Requirements) carry rights-based protections as conditions for trade. Similarly, the growing international pressure from groups such as Liberty and Privacy International,77 combined with an internal rise in rights-based discourse (2nd ed., Singapore Academy of Law Academy Publishing, Singapore, 2018) 72 S. Chesterman (ed.), Data Protection Law In Singapore: Privacy and Sovereignty in an Interconnected World (2nd ed., Singapore Academy of Law Academy Publishing, Singapore, 2018) p.3 73 Singapore Parliament Reports (Hansard) (15 October 2012) “Personal Data Protection Bill” vol 89 (Assoc Prof Dr Yaacob Ibrahim, Minister for Information, Communications and the Arts), referenced in S. Chesterman (ed.), Data Protection Law In Singapore: Privacy and Sovereignty in an Interconnected World (2nd ed., Singapore Academy of Law Academy Publishing, Singapore, 2018) 74 S. Chesterman (ed.), Data Protection Law In Singapore: Privacy and Sovereignty in an Interconnected World (2nd ed., Singapore Academy of Law Academy Publishing, Singapore, 2018) p.3 75 S. Kumar, S. Siddique, “Beyond Economic Reality: New Thoughts On The Growth Triangle” (1994) Southeast Asian Affairs, p. 47-56 76 S. Kumar, “East Asia’s Economic Lessons for the World” (1994) Straits Times, referred to in M. Chew, “Human Rights in Singapore: Perceptions and Problems” (1994) Asian Survey, Vol. 34, No. 11, p 937 77 Privacy International, “The Right To Privacy In Singapore” (2015) Stakeholder Report, Universal Periodic Review, 24th Session – Singapore, available at https:// privacyinternational.org/sites/default/files/2017-12/ Singapore_UPR_PI_submission_FINAL.pdf

ARTICLES

in Singapore, have led to strong demands for a clearer delineation and protection of rights for Singaporean citizens. On these two fronts, academics such as Chik78 and Zeller79 have stated that there is clear room for adaption and creation of such a right. However, this essay argues that such a right must be cleanly adapted to the Artificial Intelligence Era using cyber regulation theories, lest it fall prey to same failings as the GDPR’s article 17, as elucidated below. 2. The Artificial Intelligence Era 2.1 Big Data & Data Analytics We begin with a brief discussion of the environment in which these data controllers handle personal data: namely, “Big Data” and its related functions. Even as early as 2011, Koops identified the significant growth of both our “digital shadows” and “footprints”80 – with the rise of Web 2.0, the amount of personal data generated about and by us was up to 4.8 zettabytes (zb) in 2014 alone, with 2.9zb created by consumers’ actions, and of which 85% have been engaged with by enterprises such as Google and Facebook.81 This number continues to grow, as more people create a presence online. For example, in 2009, 500 million people were on Facebook. In 2019, this became 2.32 billion, with 1.52 billion active users every day,82 and each “like” and photo 78 W. Chik, “The Singapore Personal Data Protection Act and an Assessment of Future Trends in Data Privacy” (2013) Computer Law and Security Review. 29 (5), 554-575 79 B. Zeller, L. Trakman, R. Walters, S. Rosadi, “The Right To Be Forgotten – The EU and Asia Pacific Experience (Australia, Indonesia, Japan and Singapore)” (2019) European Human Rights Law Review, Vol. 1, p. 23 80 B.J. Koops, “Forgetting Footprints, Shunning Shadows: A Critical Analysis Of The Right To Be Forgotten” (2011) SCRIPTed, vol. 8, no. 3, pp. 229-256 81 EMC^2, “The Digital Universe Of Opportunities: Rich Data & The Increasing Value Of The Internet Of Things” (2014) Infographic available at https://uk.emc.com/ infographics/digital-universe-2014.htm, accessed on 22nd March 2019 82 D. Noyes, “The Top 20 Valuable Facebook Statistics – Updated March 2019” (2019) Zephoria, available at https:// zephoria.com/top-15-valuable-facebook-statistics/, accessed 22nd March 2019

leaving a trail of personal data. This volume is encapsulated by Laney’s definition of big data with the “3 V’s” – volume, velocity, and variety.83 As AIs have an “insatiable appetite”84 for large swathes of “personal, even sensitive”85 data, a symbiotic relationship forms: the collection of such large amounts of raw personal data allows companies to monetise it through the use of AI or AI-as-a-Service (AIaaS), specifically Machine Learning (ML) and Deep Learning (DL), by not only parsing this data to train the AI, but also identifying and selling trends and models within the data.86 2.2 Artificial Intelligence Artificial Intelligence is notoriously hard to define, with numerous variations of parameters or requirements – some like Boullart even postulate that it has “no strict definition”.87 However, to regulate AI one must conceptualise it: while John McCarthy advocated a broader interpretation of simply “intelligent machines” capable but not limited to human intelligence,88 a narrower view of AI espoused by prominent

83 D. Laney, “3D Data Management: Controlling Data Volume, Velocity, and Variety” (2001) Gartner, Blog Post available at https://blogs.gartner.com/doug-laney/ files/2012/01/ad949-3D-Data-Management-ControllingData-Volume-Velocity-and-Variety.pdf, accessed 22nd March 2019 84 C. Kuner, F. Cate, O. Lynskey, C. Millard, N. Loideain, Dan. Svantesson, “Expanding The Artificial Intelligence – Data Protection Debate” (2018) International Data Privacy Law, Vol. 8, Issue 4, p. 289-292 85 Information Commissioner’s Office, “Big Data, Artificial Intelligence, Machine Learning and Data Protection” p. 11 (Version 2.2 – 2017) available at https://ico.org.uk/ media/for-organisations/documents/2013559/big-data-ai-mland-data-protection.pdf, accessed 22nd March 2019 86 M. Poon, “Corporate Capitalism And The Growing Power Of Big Data” (2016) Science, Technology, & Human Values, Vol. 41, Issue 6, p. 1088–1108. DOI: 10.1177 / 0162243916650491 87 L. Boullart, A. Krijgsman, R.A. Vingerhoeds, Application Of Artificial Intelligence In Process Control (Pergamon Press, Oxford, 1992) 88 J. McCarthy, “What Is Artificial Intelligence?” (2007) Stan. U. available at: http://www.formal.stanford.edu/jmc/ whatisai/whatisai.html [https://perma.cc/N5YZ-QYS7], accessed on 20th March 2019

ARTICLES

specialists like Minksy89 and Feigenbaum90 has a common theme of “intelligent behaviours”, or the replication of “human intelligent behaviours” such as reasoning, perception and design.91 While other conceptualisations of AI exist, such as Guihot et. al.’s “risk-based” regulatory ranking,92 these definitions all accord their own questions of practical implementation and are outside of this essay’s purview Instead, this essay will adopt Russel and Norvig’s “rational agent approach”, where machines “operate autonomously… perceive their environment, persist… adapt… create and pursue”,93 as it is strikingly similar to the European Commission’s (EC) definition of task and analysis-based behaviour.94 In the context of the RTBF, apart from parsing incredible volumes of personal data, AI and their algorithms are directly responsible for creating “social memory”95 in today’s digital environment – for example, Google’s Hummingbird search algorithm (itself an AI) uses several ML AIs like Rankbrain to make autonomous decisions about the personal data that exists about us online, and how to best to present said information.96

form models out of data through Machine Learning. Both the Royal Society97 and MATLAB Deep Learning98 textbook describes ML as a form of AI where an algorithm is given “training” data sets and taught how to identify its own models from these data sets through experience, so as to effectively and autonomously analyse large quantities of data without constant reference to an external analytical model or rule set99 – essentially, it “learns to think”.100 The processing of personal data can happen with reinforced, supervised and unsupervised machine learning, and so we dispense with that distinction for the purposes of this essay. Further, the concepts of neural networks and Deep Learning are central to this essay, as most of the AI that process and store personal data are complex algorithms that are required to make autonomous decisions and value judgements. A neural network “imitates the mechanisms of the brain”,101 with nodes (representing neurons) that apply “weight values”, or attributes, to data that passes through it (fig. 2).102 In DL, the machine creates its own “learning” or “classificatory rules”103 (fig. 1),104 by parsing data through a deep neural network

2.3 Machine Learning AIs are only able to autonomously process and

89 L. Boullart, A. Krijgsman, R.A. Vingerhoeds, Application Of Artificial Intelligence In Process Control (Pergamon Press, Oxford, 1992) 90 ibid. 91 G.F. Luger, Structures And Strategies For Complex Problem Solving (5th edn, Pearson, Edinburgh, 2005) 92 M. Guihot, A. Matthew, N. Suzor, “Nudging Robots: Innovative Solutions To Regulate Artificial Intelligence” (2017) Vanderbilt Journal of Entertainment & Technology Law, Forthcoming 93 S.J. Russell, P. Norvig, Artificial Intelligence: A Modern Approach (3rd edn, Pearson, Oxford, 2010) p. 1-5 94 European Commission, “Digital Single Market, Artificial Intelligence For Europe” (2019), infographic available at https://ec.europa.eu/digital-single-market/en/artificialintelligence, accessed 21st March 2019 95 E. Esposito, “Algorithmic Memory And The Right To Be Forgotten On The Web” (2017) Big Data & Society, p. 1-11 96 MOZ, “Google Hummingbird”, blog article available at https://moz.com/learn/seo/google-hummingbird, accessed 22nd March 2019

97 The Royal Society, “Machine Learning: The Power And Promise Of Computers That Learn By Example” (2017) ISBN: 978-1-78252-259-1, available at https:// royalsociety.org/~/media/policy/projects/machine-learning/ publications/machine-learning-report.pdf, accessed 23rd March 2019 98 P. Kim, MATLAB Deep Learning – With Machine Learning, Neural Networks and Artificial Intelligence (2017), DOI 10.1007/978-1-4842-2845-6_5, p.103 99 ibid. p.4 100 D. Landau, “Artificial Intelligence And Machine Learning: How Computers Learn” (2016) iQ, available at https://iq.intel.com/artificial-intelligence-and-machinelearning/, accessed 22nd March 2019 101 P. Kim, MATLAB Deep Learning – With Machine Learning, Neural Networks and Artificial Intelligence (2017), DOI 10.1007/978-1-4842-2845-6_5, p.103 102 ibid. 103 B. Malle, P. Kieseberg, E. Weippl, A. Holzinger, “The Right To Be Forgotten: Towards Machine Learning On Perturbed Knowledge Bases” (2016) International Federation For Information Processing, CD-ARES 2016, LNCS 9817, p. 251-266 104 P. Kim, MATLAB Deep Learning – With Machine Learning, Neural Networks and Artificial Intelligence (2017), DOI 10.1007/978-1-4842-2845-6_5, p.103

ARTICLES

comprised of many nodes and creating values for each node.105 The AI is then able to apply these classificatory rules to new input data and evaluates the data accordingly.

Figure 1: Different Data Sets in a Deep Learning Set Up

possibility of decryption and the identifiable nature of public keys mean that most Blockchain users will likely be uploading their personal data onto the blockchain.108 In essence, a blockchain is a database, comprising individual data “blocks” (or ledgers) connected in a distributed network, where rules agreed by all users of the chain determine the type of data added in. Each block contains a cryptographic image of the previous block, which cannot be altered once the data is entered and “completed” (through hashing, etc.) into the block. This process follows a “consensus procedure” resulting in an exact and irrefutable copy of the entire chain across all individual servers.109 This allows Blockchain users to authenticate transactions as these copies verify each other and are “immutable” (explained below).110 Opinion 05/2014 of the Article 29 Working Party111 and Case C-582/14112 of the CJEU have ascertained that all data processed by blockchains are considered “pseudonymous” data, which is still personal data but for which its controller is given greater leeway for processing.113 At present, there is still considerable debate as to who or what in the blockchain hierarchy (or lack thereof) constitutes a data controller and processor.114

Figure 2: Deep Learning Neural Network Illustration

2.4 Blockchain Systems Finally, blockchain systems are also relevant to the RTBF, due to its use of personal data, its immutable nature, and the growing presence of AI in blockchain technology.106 The prevalence of blockchain cannot be understated– the World Economic Forum predicted that by 2025 the world will reach a “tipping point” where 10% of the global GDP will be predicated on Blockchain innovations.107 Additionally, the 105 ibid. 106 P. Treleaven, B. Bogdan, “Algorithmic Regulation: Automating Financial Compliance Monitoring and Regulation Using AI and Blockchain,” (2017) Journal of Financial Transformation, Capco Institute, Vol. 45, p. 14-21. 107 World Economic Forum, “The Global Competitiveness Report 2015-2016) (2016) available at http://repor ts.wefor um.org/global-competitiveness-

report-2015-2016/, accessed 23rd March 2019 108 W. Maxwell, J. Salmon, “A Guide To Blockchain And Data Protection” (2017) Hogan Lovells, available at https://www.hlengage.com/_uploads/ downloads/5425GuidetoblockchainV9FORWEB.pdf, accessed 22nd March 2019 109 M. Abrmaowicz, “Cryptocurrency-Based Law” (2016) Arizona Law Review, Vol. 58, Issue 359, p. 360 110 X. Xu, I. Weber, M. Staples, L. Zhu, L. Bass, C. Pautasso, P. Rimba, “A Taxonomy Of Blockchain-based Systems For Architecture Design” (2017), excerpt from ICSA’17: IEEE International Conference On Software Architecture 2017 DOI: 10.1109 111 Opinion 05/2014 of Article 29 Working Party 112 Case C-582/14, Patrick Beyer v Bundesrepublik Deutschland (2016) 113 M. Mourby, et. al., “Are ‘Pseudonymised’ Data Always Personal data? Implications Of The GDPR For Administrative Data Research In The UK” (2018) Computer Law & Security Review, Vol. 34, Issue 2, April 2018, p. 222-233 114 L.D. Ibanez, K. O’hara, E. Simperl, “On Blockchains And The General Data Protection Regulation” (2018) EU Blockchain Forum And Observatory p. 13

ARTICLES

3. “The Edge Of Impossibility” Although considerable ink has been spilled discussing the conceptual issues surrounding the RTBF, (e.g. censorship versus protection115 and the semantics of the right116), these fall outside the purview of the essay. Instead, this essay will focus on implementational issues that, as Villaronga et.al. posits, places the RTBF in an AI age on the “edge of impossibility”117: primarily, issues with data removal in AI and Blockchain systems, and the territorial limits of the GDPR’s right against transnational data flow. This list is by no means exhaustive, but these two issues showcase an excellent intersection between the RTBF and AI, demonstrating how the latter creates both challenges and opportunities to enforce this right. 3.1 Data Removal in Machine Learning On this topic, the articles by Malle et. al.118 and Villaronga et. al.119 are particularly illuminative, and describe the damage done to the accuracy and capabilities of artificial intelligences when information is “deleted” from the training sets. The damage varies depending on the type of back-end database affected. While a customer facing “front-end” database that handles the bulk of day-to-day transmissions and customer data would likely only suffer “benign”

115 S. Breheny, “The Right To Be Forgotten Online Sets A Dangerous Precedent For Freedom Of Speech” (2014) The Age, news article available at: https://www.theage.com.au/ opinion/the-right-to-be-forgotten-online-sets-a-dangerousprecedent-for-freedom-of-speech-20140715-zt7w7.html, accessed 21st March 2019 116 M. Douglas, “Questioning The Right To Be Forgotten” (2015) Alt LJ., Vol 40, Issue 2, p. 112 117 E. Villaronga, P. Kieseberg, T. Li, “Humans Forget, Machines Remember: Artificial Intelligence And The Right To Be Forgotten” (2018) Computer Law & Security Review, Vol. 34, Issue 2, p. 304-313 118 B. Malle, P. Kieseberg, E. Weippl, A. Holzinger, “The Right To Be Forgotten: Towards Machine Learning On Perturbed Knowledge Bases” (2016) International Federation For Information Processing, CD-ARES 2016, LNCS 9817, p. 251-266 119 E. Villaronga, P. Kieseberg, T. Li, “Humans Forget, Machines Remember: Artificial Intelligence And The Right To Be Forgotten” (2018) Computer Law & Security Review, Vol. 34, Issue 2, p. 304-313

repercussions, databases prepared for machine learning or statistical databases are likely to be much more severely impacted. This is because they do not hold personal data in full, but rather relevant fragments of the original data, which are hard to separate from each other in that state.120 Further, Villaronga states that databases that are “ACID-compliant”121 (adhering to Atomicity, Consistency, Isolation, and Durability requirements) suffer the risk of instability when faced with the permanent deletion of even some of its data.122 Assuming that the data controller intends to keep a usable data set as a training database, we can rule out the IDC’s proposed methods, leaving us with the “pseudo-deletion” methods, namely functional encryption and anonymisation. Without engaging the procedural mathematics, these essentially render the data unrecognisable, either through cryptographical functions or generalising sensitive attributes.123 Villaronga et. al. rules out encryption as too inefficient,124 and Malle and Kieseberg’s experiment125 on k-anonymisation is instructive: throughout all four classifier algorithms tested, there was an increasing distortion in the algorithm’s accuracy corresponding to the volume of data anonymised, as the AI could discern fewer classificatory rules with each corresponding k-anonymity grade increase.126 Mayer-Schönberger elegantly summarises the implications: ‘the quality of [AI applications] would deteriorate absent sufficient data, 120 B. Malle, P. Kieseberg, E. Weippl, A. Holzinger, “The Right To Be Forgotten: Towards Machine Learning On Perturbed Knowledge Bases” (2016) International Federation For Information Processing, CD-ARES 2016, LNCS 9817, p. 251-266 121 T. Haerder, A. Reuter, “Principles Of TransactionOriented Database Recovery” (1983) ACM Computing Surveys 15(4) 287, DOI: 10.1145/289.291 122 E. Villaronga, P. Kieseberg, T. Li, “Humans Forget, Machines Remember: Artificial Intelligence And The Right To Be Forgotten” (2018) Computer Law & Security Review, Vol. 34, Issue 2, p. 304-313 123 ibid. 124 ibid. 125 B. Malle, P. Kieseberg, E. Weippl, A. Holzinger, “The Right To Be Forgotten: Towards Machine Learning On Perturbed Knowledge Bases” (2016) International Federation For Information Processing, CD-ARES 2016, LNCS 9817, p. 251-266 126 ibid.

ARTICLES

leading to inefficient transactions and reduced consumer welfare”.127 Compounding this, there is evidence to suggest that “re-anonymisation” could render this method obsolete as well,128 although this would require a regulator’s strict interpretation to rule it out. 3.2 Data Removal in Blockchain As aforementioned, one of the key tenets and attractions of blockchain is the very fact that its data is “relatively immutable”,129 in that its data cannot be altered without the consensus of every party in the blockchain – a difficult task, given that Bitcoin’s blockchain, for example, has 581,664 blocks at the time of writing.130 Additionally, if a single block were to be removed from the chain, the next and subsequent block’s hashing key would not match, and the chain would fail. As such, both authorities such as the CNIL131 and industry professionals132 treat the GDPR’s right to erasure as “impossible”133 to implement in a blockchain environment. Workarounds thus far have included the use of “off-chain storage”,134 deletion of encryption 127 V. Mayer-Schönberger, T. Range, “A Big Choice For Big Tech: Share Data Or Suffer The Consequences” (2018) Foreign Affairs, p.52 128 P. Ohm, “Broken Promises Of Privacy: Responding To The Surprising Failure Of Anonymisation” (2012) UCLA Law Review, Vol. 57, p. 1701 129 L.D. Ibanez, K. O’hara, E. Simperl, “On Blockchains And The General Data Protection Regulation” (2018) EU Blockchain Forum And Observatory p. 13 130 BitInfoCharts, “Bitcoin (BTC) Price Stats and Information” live-graph, available at https://bitinfocharts. com/bitcoin/, accessed at 2019-06-21, at 02.47a.m. (GMT+2) 131 Commission Nationale Informatique & Libertes, “First Elements: Analysis of the Blockchain“ (2018) Infographic available at https://www.cnil.fr/en/blockchain-and-gdprsolutions-responsible-use-blockchain-context-personal-data 132 B. Beck, Dr. U. Worm, “GDPR Implications For Blockchain And Distributed Ledger Technologies” (2018) Mayer Brown, Lexology, available at https://www.lexology. com/library/detail.aspx?g=303d2d14-1af9-4631-898ad307bace3e3d, accessed on 24th March 2019 133 Commission Nationale Informatique & Libertes, “First Elements: Analysis of the Blockchain“ (2018) Infographic available at https://www.cnil.fr/en/blockchain-and-gdprsolutions-responsible-use-blockchain-context-personal-data 134 X. Xu, I. Weber, M. Staples, L. Zhu, L. Bass, C. Pautasso, P. Rimba, “A Taxonomy Of Blockchain-based Systems For Architecture Design” (2017), excerpt from ICSA’17: IEEE International Conference On Software

keys, 135 and smart-contracts to revoke all access to the blockchain.136 Though these methods do render the content unavailable to all parties, they also strip the user of the benefits of using blockchain.137 As blockchain begins to become implemented in industries like healthcare138 and the Internet-Of-Things,139 there is a real risk of an individual having no recourse to their personal data trapped within a blockchain system – conversely, others have accused the blunt invasion of data protection laws of “ruining”140 the libertarian141 ideal and technological landscape of blockchain. 3.3 International Jurisdiction The third issue we dissect is the “classic opposition between networks and territories”,142 an issue related to both the right to de-list and the right to erasure. In 2000, the French and American courts had to contend with the 135 P. Townsend, “GDPR, Right Of Erasure (Right To BE Forgotten), And Encryption Key Management” (2018), Blog available at https://info.townsendsecurity.com/gdprright-erasure-encryption-key-management, accessed 20th March 2019 136 A. Ouaddah, A.A. Elkalam, A.A. Ouahman, “Towards A Novel Privacy-Preserving Access Control Model Based On Blockchain Technology In IoT” (2016) Europe and MENA Cooperation Advances In Information and Communication Technologies, p. 523 137 W. Maxwell, J. Salmon, “A Guide To Blockchain And Data Protection” (2017) Hogan Lovells, available at https://www.hlengage.com/_uploads/ downloads/5425GuidetoblockchainV9FORWEB.pdf, accessed 22nd March 2019 138 A. Azaria, A. Ekblaw, T. Vieira, A. Lippman, “MedRec: Using Blockchain For Medical Data Access And Permission Management” (2016) 2016 2nd International Conference On Open And Big Data (OBD) DOI: 10.1109/ OBD.2016.11 139 A. Dorri, S. Kanhere, R. Jurdak, “Towards An Optimised Blockchain For IoT” (2017) IoTDI 17 Proceedings Of The Second International Conference On The Internet-OfThings Design and Implementation, p. 173-178 140 M. Beedham, “Here’s How Blockchains Could Be Ruined By Personal Data” (2019) TNW Hard Fork, Blog available at https://thenextweb.com/hardfork/2019/03/25/ moonday-mornings-twitch-drops-bitcoin-most-btc-tradingcould-be-fake-and-more/, accessed 24th March 2019 141 I. Eyal, “Blockchain Technology: Transforming Libertarian Cryptocurrency Dreams To Finance And Banking Realities” (2017) Computer, Vol. 50, Issue 9, p. 38-49 142 Y. Padova, “Is The Right To Be Forgotten A Universal, Regional, Or “Glocal” Right?” (2019) International Data Privacy, 2019, Vol. 0, No. 0, p. 15

ARTICLES

application of national laws to the internet in LIRCA v Yahoo143 – 19 years later the same questions rear their heads in CNIL v Google.144 In the latter case, CNIL argued that to refuse to enforce an obligation on Google to de-list from all relevant domains would be to hamstring the right to de-listing, in that information deindexed in one domain could still be found in other domains, and that even Google’s highly effective response of “geo-blocking” was “insufficient”.145 In response, Google countered that that would mean extending the jurisdiction of the right to de-listing to all countries that Google operates in, which the Advocate General recently declared beyond the ambit of the GDPR.146 The problem hereto is that both parties were right, and this Gordian knot lies at the “core” of international cyberlaw theory, where the regulation of a space that transgresses existing geopolitical boundaries is not easily informed by existing laws in differing jurisdictions that treat privacy rights differently, and too often places quasi-public regulatory responsibilities on private companies.147 This problem is exacerbated for the right to erasure, which has no alternative solution such as geo-blocking to simulate the effects of erasure for EU members. To contextualise the problem, any cloud-based service or small online business that cannot afford a Europebased data centre or servers would process, store and utilise personal data it receives globally in its own country, which might not have an equal right to erasure (and indeed may actively resist one).148 Even if consent is withdrawn, there is

143 Ligue contre le racisme et l’antisémitisme et Union des étudiants juifs de France c. Yahoo! Inc. et Société Yahoo! France (2000) 144 Case C-507/17, Google Inc. v. Commission Nationale de I’informatique Et Des Libertes (CNIL) 145 Commission Nationale de I’informatique Et Des Libertes’s deliberation no 2016-054 (n 31) 146 Case C-507/17 Google Inc. v. Commission Nationale de I’informatique Et Des Libertes (CNIL) Opinion of AG Spuznar 147 P. Stephenson, “Defining a Cyber Jurisprudence” (2017) Annual ADFSL Conference on Digital Forensics, Security and Law, available at https://commons.erau.edu/cgi/ viewcontent.cgi?article=1373&context=adfsl 148 R. Atani, “The Resistance Of Memory: Could The

no guarantee that the controller would delete the data, or even refrain from challenging the jurisdiction of the enforcement. As aforementioned, when the CNIL argued the application of Google’s geo-blocking technique, while 99% effective, it asserted that the solution was “as a principle, insufficient”149 as “the protection of a fundamental right cannot vary depending on the data’s recipient’.150 While the proportionality of that claim may be questionable,151 this essay postulates that as a result of the technical difficulties aforementioned, there could potentially be a non-insignificant number of scenarios whereby there exists no technical solution to give effect to the GDPR’s right to erasure, or else one that does not damage to society’s general use of AI or Blockchain. As such, this essay asserts that the GDPR fails to provide a full and futureproof protection of an individual’s right to identity and privacy, and further that this failure is the result of the EC’s adherence to traditional “command and control” regulation,152 as well as a lack of recourse to contemporary cyberspace regulation theories when crafting the GDPR. 4. Cyber Regulation From as early as Reidenberg’s Lex Informatica,153 cyber-paternalistic154 theories have advocated a regulatory shift away from the direct control of cyberspace actors towards “influencing changes to its architecture”155 – code. This European Union’s Right To Be Forgotten Exist In The United States?” (2015) Berkeley Technology Law Journal, Vol. 30, No. 4, Annual Review, p. 1173-1210 149 Commission Nationale de I’informatique Et Des Libertes’s deliberation no 2016-054 (n 31) 150 ibid. 151 Y. Padova, “Is The Right To Be Forgotten A Universal, Regional, Or “Glocal” Right?” (2019) International Data Privacy, 2019, Vol. 0, No. 0 152 T.C. Dainith, “The Techniques of Government” in J. Jowell & D. Oliver, The Changing Constitution (3rd edn, Oxford University Press, Oxford, 1994) 153 J. Reidenberg, “Lex Informatica: The Formation Of Information Policy Rules Through Technology” (1998) 76 Texas Law Review 553 154 A. Murray, Information Technology Law: The Law And Society (3rd ed, Oxford University Press, Oxford, 2016) p. 66 155 ibid. p. 67

ARTICLES

was encapsulated by Lessig’s seminal book “Code”156 and his (Cyber-Regulatory Theory (CRT)157 model: that there existed “the four modalities” of regulation (architecture, market, norms, and law); that people, labelled “pathetic dots”, had their behaviour controlled invariably and inevitably by any combination of these four modalities, and; that in an online environment, code as architecture provided the foremost solution to enacting regulation.158 As Lessig himself declared: “We can build… or code cyberspace to protect values we believe are fundamental. Or we can… code cyberspace to allow those values to disappear. There is no middle ground. There is no choice that does not include some kind of building.”159 Lessig’s theory quickly came to influence American approaches to regulating technology, and combined the pursuit of innovation with an appreciation of the regulatory capabilities and limitations of the technology itself.160 In contrast, the EU’s approach, as aforementioned, has traditionally focussed on the historic rights of the individual rather than the current state of technology, and this essay posits that the GDPR fell foul of the same fault that the CJEU did in Google Spain161 – it failed to properly account for the “current state of communications service provision”.162 In the GDPR, the only direct attempt to address code as regulatory architecture lies in Article 25, which requires 156 L. Lessig, Code And Other Laws Of Cyberspace (Basic Books, 1999) 157 L. Lessig, Code: Version 2.0 (Basic Books, 2006) p. 4 158 L. Lessig, Code And Other Laws Of Cyberspace (Basic Books, 1999) 159 L. Lessig, Code: Version 2.0 (Basic Books, 2006) p. 6 160 D. McCullagh, “What Larry Didn’t Get” (2009) in “Ten Years of Code: A Reassessment of Lawrence Lessig’s Code and Other Laws of Cyberspace”, available at https:// www.cato-unbound.org/2009/05/04/declan-mccullagh/whatlarry-didnt-get, accessed 6th June 2019 161 Case C-131/12, Google Spain SL and Google Inc. v Agencia Española de Protección de Datos (AEPD) and Mario Costeja González. (2014) 162 European Union Committee, 2nd Report - EU Data Protection Law: A ‘Right To Be Forgotten (2010 HL 40) at “Chapter 5: The Views Of The Committee”

organisations to enact “data protection by design and default” by “implement[ing] appropriate technical and organisational measures”.163 Indeed, prior to 2018, the EU did not even (and arguably still does not164) have a cohesive plan to engender the development of such “technical and organisational measures”, and funding for AI technology was severely lacking: $4b compared to the $23b and $12b of North America and Asia respectively (the highest estimate per market).165 We also note that the UK Parliament’s recommendation of the creation of an overarching “digital authority” to help with “digital capabilities”166 to comply with the new GDPR obligations suggests that the EC has yet to undertake a cohesive regulatory body or action towards creating technological solutions to implementing GDPR rights. Although the EC has made promising overtures towards enabling a “European Approach to AI” such as increasing AI investment under the research programme Horizon 2020 by up to 70%,167 and dedicating a “High-level Expert Group”168 to produce proposals by June 2019, this essay notes that these were done in response to the drastic gap in AI progress between the EU and US/Asia, aforementioned above, as opposed to pre-emptive measures taken in tandem with the drafting of the GDPR, to enable technology to 163 EU General Data Protection Regulation (GDPR): Regulation (EU) 2016/679, Article 25 164 C. Minksy, “Kaifu Lee: No Hope For Europe’s Artificial Intelligence Sector” (2018), Interview available at https://sifted.eu/articles/interview-kaifu-lee-artificialintelligence/, accessed 22nd March 2019 165 European Commission, “Digital Transformation Monitor: USA-China-EU Plans For AI: Where Do We Stand?” (2018) Infographic available at: https://ec.europa.eu/growth/ tools-databases/dem/monitor/sites/default/files/DTM_ AI%20USA-China-EU%20plans%20for%20AI%20v5.pdf 166 Parliament, House Of Lords, “2nd Report – Regulating In A Digital World” (2019) Select Committee On Communications HL Paper 299 167 European Commission, “Digital Single Market Policy: Artificial Intelligence” , available at https://ec.europa. eu/digital-single-market/en/artificial-intelligence#AEuropean-approach-to-Artificial-Intelligence, accessed 14th June 2019 168 European Commission, “Digital Single Market Policy: High-Level Expert Group on Artificial Intelligence”, available at https://ec.europa.eu/digital-single-market/en/ high-level-expert-group-artificial-intelligence, accessed 14th June 2019

ARTICLES

evolve in line with regulations, and vice-versa. Indeed, this is reflected in Reding’s introduction of the GDPR’s draft proposals, where she speaks only of “regulating” the “astounding capabilities of modern technology”,169 as opposed to the adoption of these capabilities to frame new regulatory paradigms. Although regulations can invariably affect a cyberspace environment,170 and the GDPR’s comprehensive scope has enormous ramifications for data protection,171 the stark lack of attention to “code as architecture” betrays the fatal flaw in the EC’s perspective and reliance on traditional “command and control”,172 “law-based” regulation, which has in turn created “regulatory gaps”173 in legislation such as the DPD and the GDPR. However, the “classical cyberpaternalist”174 CRT itself had conceptual limitations, and does not provide a robust solution to these problems. For example, Vranaki argues that CRT views law in an “Austinian” manner, characterising law as “sanctions backed by a sovereign”, which does not reflect the “heterogenous ways” in which laws manifests itself (elsewise as rights, etc).175 Although Murray does also acknowledge the importance of code as a regulating medium, he found CRT failed to account for the “complexity of information flows”,176 and did not appreciate the intricate relationship between human and

169 V. Reding, “The Upcoming Data Protection Reform For The European Union” (2010) International Data Privacy Law, Vol. 1, Issue 1, p. 3-5 170 P. Pattberg, ““The Institutionalisation Of Private Governance: How Business And Non-profit Organisations Agree on Transnational Rules” (2005) Governance, Volume 18, Issue 4, October 2005 171 L. Edwards, Law, Policy and the Internet (Hart Publishing, 2019) p. 91 172 A. Murray, The Regulation Of Cyberspace: Control In The Online Environment (Routledge, 2007) 173 A. Murray & C. Scott, “Controlling The New Media: Hybrid Responses To New Forms Of Power” (2002) 65(4) MLR 491 174 A. Murray, Information Technology Law: The Law And Society (3rd ed, Oxford University Press, Oxford, 2016) p. 73 175 A. Vranaki, “Social Networking Site Regulation: Facebook, Online Behavioral Advertising, Power And Data Protection Laws (2017) Rutgers Computer & Technology Law Journal, 43(2), 169-210 176 A. Murray, The Regulation Of Cyberspace: Control In The Online Environment (Routledge, 2007)

non-human actants. Too, Murray argued that Lessig’s conceptualisation of the individual as a “pathetic dot”,177 at the whim of regulatory forces around them, was misguided. In practice, this would be to presume individuals in a society as perfectly malleable, who would simply accept and perfectly execute all regulations placed upon them; such a paradigm does not reflect our reality. 4.1 Network Communitarianism As such, Murray proposed his own theory, Network Communitarianism, that combined Latour’s Actor-Network Theory (ANT)178 with Luhmann’s Social Systems Theory (SST).179 Network Communitarianism reconceptualised the individual as an “active dot” that, when working in tandem with a “social system” comprised of other “dots”, actively affected regulation by electing lawmakers, exercising market support, and deciding social norms – what he collectively called “socially mediated modalities”.180 He also characterised intermediaries such as Google and Facebook as the nexus of informational flows, which increased their ability to represent dots and affect regulation. Although Leiser does criticise both Murray and Lessig for assuming individuals represent “rational” dots,181 we are able to characterise the Google Spain case182 as an example of Murray’s “active dot” engendering “a discourse… between the individual and society”183 to enact regulation. Due to the lack 177 L. Lessig, Code: Version 2.0 (Basic Books, 2006) p. 6 178 B. Latour, Reassembling The Social: An Introduction to Actor-Network Theory (Oxford University Publishing, Oxford, 2007) 179 N. Luhmann, Law As A Social System (Oxford University Publishing, Oxford, 2008) 180 A. Murray, The Regulation Of Cyberspace: Control In The Online Environment (Routledge, 2007) 181 M. Leiser, “The Problem With ‘Dots’: Questioning The Role Of Rationality In The Online Environment” (2016) International Review Of Law, Computers & Technology, Vol. 30, Issue 3, p. 191-210 182 Case C-131/12, Google Spain SL and Google Inc. v Agencia Española de Protección de Datos (AEPD) and Mario Costeja González. (2014) 183 A. Murray, Information Technology Law: The Law And Society (3rd ed, Oxford University Press, Oxford, 2016) p. 243

ARTICLES

of an explicit RTBF in the DPD, there existed a regulatory gap in which an “active dot” (Mario Cotesja, represented by the AEPD) brought an action against Google Spain and, through the court, effectively “invented”184 the common law right to de-listing (affecting the socially mediated law), which was then implemented by Google (as the intermediary and representation of the market and architectural modalities). Subsequently, the 786,757 requests made so far further represent the societal want for this norm.185 Unfortunately, Network Communitarianism also lacks conceptual rigour when it addresses the non-human actants in the actor network. This is particularly relevant in an AI context, that provides both challenges and opportunities specific to itself. To this end, we find that Vranaki’s criticisms of the Network Communitarianism elucidate its shortcomings. Firstly, Network Communitarianism’s incongruity with the ANT tenets of agnosticism186 (not privileging one actant over another) and analytical symmetry (actively treating types of actors similarly) leads it to over-emphasise the role of human actors over non-human counterparts,187 which ironically mirrors similar criticism of CRT’s “top-down” approach and fixation with the state.188 Additionally, Murray fails to adopt the 184 S. Breheny, “The Right To Be Forgotten Online Sets a Dangerous Precedent” (2014) The Age, available at http:// www.theage.com.au/comment/the-right-to-be-forgottenonline-sets-a-dangerous-precedent-for-freedom-of-speech20140715-zt7w7.html, accessed 20th March 2019 185 Google Transparency Report, “Search Removals Under European Privacy Law: Overiew”, available at https:// transparencyreport.google.com/eu-privacy/overview, accessed 24th March 2019 186 M. Callon, “Some Elements Of A Sociology Of Translation: Domestication Of The Scallop And The Fishermen of St Brieuc Bay” in, Action and Belief; A New Sociology of Knowledge (ed. J. Law, 1986) p. 196, referenced in A. Vranaki, “Social Networking Site Regulation: Facebook, Online Behavioral Advertising, Power And Data Protection Laws (2017) Rutgers Computer & Technology Law Journal, 43(2), 169-210 187 A. Vranaki, “Social Networking Site Regulation: Facebook, Online Behavioral Advertising, Power And Data Protection Laws (2017) Rutgers Computer & Technology Law Journal, 43(2), 169-210 188 E. Dommering, “Regulating Technology: Code Is Not Law” in E. Dommering & L. Asscher, “Coding Regulation:

social constructivist approach that is central to ANT, and attributes regulatory change primarily to “settlements” driven by society, rather than the specific and individual factors that contributed deictically.189 Lastly, this essay notes that Network Communitarianism’s fixation with the visualisation of “regulation” as prohibitive legal rules (e.g. child pornography example, etc.190) Lastly, this essay notes that Network Communitarianism’s fixation with the visualisation of “regulation” as prohibitive legal rules (e.g. child pornography example, etc.) similarly does not reflect the heterogenous nature of law. 4.2 ANT-Foucauldian Power Lens As such, we find the terminology and concepts used in Vranaki’s ANT-Foucauldian Power Lens191 (ANT-FPL) theory are most suited to tackle a discussion of data protection rights, especially since it was specifically constructed for this purpose. Of particular relevance is ANTFPL’s use of Foucauldian power terminology: (i) the conceptualisation of the “actant” that recognises the power in non-human actants like legal knowledge and hyperlinks; (ii) the conceptualisation of power as a “dynamic”, “enabling and productive phenomenon” that can transforms how relations are organised; (iii) that power exists only “in action” as an effect generated by the connection of a chain of actants, and that latent potential in actants without action is worthless, and; (iv) that material conditions can affect the power generated by said connections.192

Essays On The Normative Role Of Information Technology” (2006) T.M.C Asser Press, The Hague 189 Muniesa, F., . “Actor-Network Theory” (2015) in James D. Wright (Ed.), The International 190 A. Murray, Information Technology Law: The Law And Society (3rd ed, Oxford University Press, Oxford, 2016) 191 A. Vranaki, “Social Networking Site Regulation: Facebook, Online Behavioral Advertising, Power And Data Protection Laws (2017) Rutgers Computer & Technology Law Journal, 43(2), 169-210 192 A. Vranaki, “Social Networking Site Regulation: Facebook, Online Behavioral Advertising, Power And Data Protection Laws (2017) Rutgers Computer & Technology Law Journal, 43(2), 169-210

ARTICLES

Through this understanding of power and relations between actants, we are able to explain the trend of reliance on technological and market-based solutions, predominantly led by individual research teams or market actors, to these problems. For example, Privacy-Aware Machine Learning (PA-ML) is a school of ML that is currently undergoing heavy research and experimentation193 to enable privacy by creating change-resistant databases,194 and allowing selective sharing, adaptive use,195 and data integration196 through improved sharing techniques, all of which engender a safer, less damaging protection of personal data when used for ML training. Similarly, advances in the field of Memory Optimised & Flexible BlockChain (MOF-BC) have shown promise in allowing for personal data to be selectively removed from active blockchains through techniques such as SP-Initiated Memory Optimisation197 or hierarchical deterministic wallets that tolerate “key leakage”.198 In these scenarios, although the law (i.e. GDPR article 17), as an actant might accord for the right to be forgotten, the data right is not realised until a new privacy-enabled technological actant interacts with both it and a user, and in doing so effectuates a transformation in how the law can be utilised and the right can have effect. By modelling it as such, it becomes apparent

193 A. Holzinger, P. Kieseberg, “Current Advances, Trends And Challenges Of Machine Learning And Knowledge Extraction: From Machine Learning To Explainable AI” (2018), International Cross-Domain Conference For Machine Learning And Knowledge Extraction 194 B. Malle, P. Kieseberg, S. Schrittwieser, A. Holzinger, “Privacy Aware Machine Learning And The ‘Right To Be Forgotten’” (2016) ERCIM News, 107 195 I. Bilogrevic, “A Machine-Learning Based Approach To Privacy-Aware Information-Sharing In Mobile Social Neworks” (2016) Pervasive And Mobile Computing, Vol. 25, p. 125-142 196 C. Clifton, G. Schadow, “Privacy Preserving Data Integration And Sharing” (2004) DMKD 04, 9th ACM SIGMOD Workshop, p. 19-26 197 A. Dorri, S. Kanhere, R. Jurdak, “MOF-BC: A Memory Optimised And Flexible Blockchain For Large Scale Networks” (2019) Future Generation Computer Systems, Vol. 92, p. 357-373 198 G. Gutoski & D. Stebila, “Hierarchical Deterministic Bitcoin Wallets That Tolerate Key Leakage” (2015) International Conference On Financial Cryptography and Data Security, Springer, p. 497-50

that future cyberspace regulators must not only account for a legal right but also pay particular attention to the flow of power between existing human and non-human actants to pre-empt potential technological and so regulatory gaps. As such, this essay ventures that a combination of Vranaki’s ANT-FPL and Murray’s Network Communitarianism would best provide for future solutions to AI and cyberspace regulation. This is because Network Communitarianism’s conceptualisation of intermediaries such as Facebook and Google as the nexus of information flow199 does correctly characterise their influence in regulatory changes, and fits with the “flow of power”200 so envisioned by ANT-FPL, as intermediaries are the convergence points of several chains of actants (e.g. for Google, that could include: GDPR article 17, individual users and their personal data, privacy control buttons, Rankbrain’s indexing AI and its classificatory rules, etc.). For example, a potential technical solution to the problem of international jurisdictions might lie in the revival of Mayer-Schönberger201 “self-destructing data”. Although Graux et. al.202 had already dismissed the idea as being too onerous on the average user to implement, a revisualisation of the model might prove useful: the transformative power of an assembly of actants [such as the GDPR Article 17 RTBF,203 self-deleting data techniques (created by market forces such as Vanish),204 web-extensions that 199 A. Murray, The Regulation Of Cyberspace: Control In The Online Environment (Routledge, 2007) 200 A. Vranaki, “Social Networking Site Regulation: Facebook, Online Behavioral Advertising, Power And Data Protection Laws (2017) Rutgers Computer & Technology Law Journal, 43(2), 169-210 201 V. Mayer-Schönberger, Delete, The Virtue Of Forgetting In The Digital Age (Princeton University Press, Oxford, 2009) p. 169 202 H. Graux, J. Ausloos, P. Valcke, “ICRI Working Paper Series: The Right To Be Forgotten In The Internet Era” (2012) Interdisciplinary Centre For Law And ICT, K.U.Leuven, ICRI Working Paper 11/2012 203 EU General Data Protection Regulation (GDPR): Regulation (EU) 2016/679, Article 17 204 R. Geambasu, “Vanish: Increasing Data Privacy With Self-Destructing Data” (2017) 18th USENIX Security Symposium, available at: https://www.usenix.org/legacy/ event/sec09/tech/full_papers/sec09_crypto.pdf, accessed

ARTICLES

allow for ease of access to privacy controls,205 a specific “safe-harbour” style legal agreement,206 and a “realistic” understanding of erasure by regulators like the ICO207 would be best combined by an intermediaries such as Google, who would link their own actants (such as political and economic power, and control over TCP/IP domain identifiers208) to this chain. The result, with luck, might circumvent the uncertain legal and technological barriers to this puzzle, and transform regulation to enact this comprehensive right. 4.3 Frameworks For The Future This essay therefore posits that this combined theory should inform Singaporean regulators should they attempt to enact a Singaporean version of the RTBF, either as to work in compliance with the EU’s “Adequacy Decisions” to allow for better economic and digital privacy ties, or to simply create a robust right to identity for Singaporeans, a right thus far seriously lacking. Although an explicit measure to create such a right has yet to be tabled, Singapore’s Personal Data Protection Commission’s (PDPC) recently published “Proposed Model Artificial Intelligence Framework”209 (the “Framework”) may provide insight into Singapore’s approach to AI regulation generally, especially when viewed with the precepts of ANT-FPL and Network 24th March 2019 205 H. Graux, J. Ausloos, P. Valcke, “ICRI Working Paper Series: The Right To Be Forgotten In The Internet Era” (2012) Interdisciplinary Centre For Law And ICT, K.U.Leuven, ICRI Working Paper 11/2012 206 A. Morrison, “Case Law, Systemic Law, And A Very Modest Suggestion” (2014) Statute Law Review 35(2), p. 159180 207 Information Commissioner’s Office, “Deleting Personal Data: Data Protection Act” (2014) 20140226, Ver 1.1, available at https://ico.org.uk/media/for-organisations/ documents/1475/deleting_personal_data.pdf, accessed on 23rd March 2019 208 M. Alani, “Guide To OSI and TCP/IP Models” (2014), Online ISBN: 978-3-319-05152-9 209 Personal Data Protection Commission, “A Proposed Model Artificial Intelligence Governance Framework” (2019) available at https://www.pdpc.gov.sg/-/media/Files/PDPC/ PDF-Files/Resource-for-Organisation/AI/A-ProposedModel-AI-Governance-Framework-January-2019.pdf, accessed on 22nd June 2019

Communitarianism theories in mind. It is important to note that the Framework is still in its second round of public consultation (closing 30th June 2019) and focuses primarily on the fair and transparent use of AI in a business environment,210 more akin to the GDPR Art. 13-15’s “Right to Explanation”211 rather than the Art.17 RTBF.212 Still, while the Framework itself does not promise an enforceable RTBF or even directly address issues of privacy, it makes promising overtures regarding the PDPC’s attitude towards AI regulation, and lays the groundwork necessary for the safe enforcement of the GDPR’s RTBF, if not the creation of a Singaporean RTBF. From the outset, the proposed Framework seems to adhere with the tenets of ANT-FPL as well as Network Communitarianism, even if was not explicitly intended as such. First and foremost, the framework unambiguously addresses both the human and non-human actants that underpin AI operations, especially at a larger scale. For example, it not only distinguishes between the various modes of AI decision-making models, identifying the various levels of human involvement in an AI’s ability to generate output (e.g. Human-inthe-loop v.s. Human-over-the-loop, e.t.c.), but also focuses on internal governance structures and operational management strategies. These regulate not just the linear use of AI between developer and end-user, but more specifically the “flow” of action between human and nonhuman actants. This comprehensive method of applying regulation across actants not only provides for more transparent practices for AI end-users and business management teams, but represents a far more nuanced, sensitive approach to regulating nascent technology – what Kalmanath believes will become an 210 A. Kamalnath, “The Perennial Quest for Board Independence – Artificial Intelligence To The Rescue?” (2019) Albany Law Review, Forthcoming 211 EU General Data Protection Regulation (GDPR): Regulation (EU) 2016/679, Article 13-15 212 S. Chesterman, “Should We Regulate A.I.? Can We?” (2019) available at SSRN: https://papers.ssrn.com/sol3/ papers.cfm?abstract_id=3357618

ARTICLES

“important guide globally”.213 Of particular relevance to this essay is the Framework’s focus on the datasets used for AI Model Development and its call for higher standards of data quality, data lineage records, and the minimisation of inherent bias within the data. Kuner214 commends the PDPC’s identification of data as the “basic building block of the digital economy”,215 which is strikingly similar to Lessig’s own visualisation of data,216 as opposed to the GDPR’s treatment of personal data as a liability to be minimised.217 As identified above at section 3.1 of this essay, a legislator’s ignorance of these specific operational requirements in ensuring robust AI development would lead to the technical problems explored by Villaronga218 and Kieseberg.219 The PDPC’s decision to focus the Framework on these issues represents a necessary step towards creating a regulatory environment that both allows for innovation and yet is primed for the implementation of future “rights of the individual” like a RTBF. Further, the identification and clear definition of the “organisation” as the nexus of different operational actants and the actuator of ethical

213 A. Kalmanath, “The Perenial Quest for Board Independence – Artificial Intelligence to the Rescue?” (2019) Albany Law Review, Forthcoming 214 C. Kuner, F. Cate, O. Lynskey, C. Millard, N. Loideain, D. Svantesson, “Expanding the Artificial Intelligence – Data Protection Debate” (2019) International Data Privacy Law, Vol. 8, Issue 4, p. 289-292 215 Singapore Personal Data Protection Commission, ‘Discussion Paper on Artificial Intelligence (AI) and Personal Data—Fostering Responsible Development and Adoption of AI’, (5 June 2018) p 2, <https://www.pdpc.gov.sg/-/media/ Files/PDPC/PDF-Files/Resource-for-Organisation/AI/ Discussion-Paper-on-AI-and-PD—050618.pdf>. 216 L. Lessig, Code: Version 2.0 (Basic Books, 2006) p. 6 217 EU General Data Protection Regulation (GDPR): Regulation (EU) 2016/679, Article 17art 5(1)(c). 218 E. Villaronga, P. Kieseberg, T. Li, “Humans Forget, Machines Remember: Artificial Intelligence And The Right To Be Forgotten” (2018) Computer Law & Security Review, Vol. 34, Issue 2, p. 304-313 219 B. Malle, P. Kieseberg, E. Weippl, A. Holzinger, “The Right To Be Forgotten: Towards Machine Learning On Perturbed Knowledge Bases” (2016) International Federation For Information Processing, CD-ARES 2016, LNCS 9817, p. 251-266

and responsible management policies is in line with Network Communitarianism. This focus distinctly acknowledges the shift in the locus of power from government regulators to private/quasi-public companies in effectuating regulation in the digital sphere, especially when it comes to more nascent technology such as AI. The Framework’s use and incorporation of consultation feedback from organisation such as Facebook, Element AI, BSA, and IBM Asia Pacific, as well as service providers such as Grab, DBS and Mastercard,220 is also welcome progress, as it shows deliberate consideration of the unique positions and technological capabilities that these intermediaries and service providers have. These developments in turn have the effect of clearing any regulatory ambiguity, which allows for investors and developers alike to begin greater investment and research in Singapore’s AI ecosystem.221 However, that is not to say that the Framework is entirely without flaws. While an application of ANT-FPL and Network Communitarianism should inform any AI regulator’s decisions, it is pertinent to remember that their roots stemmed from Lessig’s theory, itself grounded in the innovation-driven, laissez-faire style of American cyberlaw jurisprudence. As Movius warns, a strict adherence to this paradigm might result in an over-reliance in intermediaries for regulation.222 Chesterman further cautions that this in turn may lead to regulatory arbitrage,223 whereby private companies may cut regulatory corners to eke out competitive advantages and 220 Personal Data Protection Commission, “A Proposed Model Artificial Intelligence Governance Framework” (2019) available at https://www.pdpc.gov.sg/-/media/Files/PDPC/ PDF-Files/Resource-for-Organisation/AI/A-ProposedModel-AI-Governance-Framework-January-2019.pdf, accessed on 22nd June 2019 221 D. Helbing, B. Frey, G. Gigerenzer, et. al., “Will Democracy Survive Big Data and Artificial Intelligence?” in D. Helbing (eds.) Towards Digital Enlightenment (2019) (Springer, Cham, 2019), p. 73-98 222 L. Movius, N. Krup, “U.S and EU Privacy Policy: Comparison of Regulatory Approaches” (2009) International Journal of Communication, Vol. 3, p. 169-187 223 S. Chesterman, “Should We Regulate A.I.? Can We?” (2019) available at SSRN: https://papers.ssrn.com/sol3/ papers.cfm?abstract_id=3357618

ARTICLES

shift risk, liabilities and losses onto consumers,224 as was the case with Google’s illegal tax benefits in Ireland,225 and Facebook’s personal data scandal with Cambridge Analytica.226 As such, while the PDPC’s approach has much to be commended for, it must take caution to not disregard the personal privacy rights of Singaporeans in the pursuit of a businessfriendly AI environment. 5. Conclusion This essay hopes to have illustrated the specific technical and legal challenges created by AI actants, and that regulators disregarding contemporary cyberspace theories and technological actants will invariably suffer from regulatory gaps. While authors like Polanski227 and Horne228 might suggest more traditional legal solutions to the problem of regional jurisdiction across cyberspace such as a reliance on ICANN or a re-imagination of Safe Harbour provisions to include a right to erasure, this essay posits that it is imperative to consider all potential actants that have the potential transformative power to enact regulatory change. In this regard, Vranaki’s ANT-FPL and Murray’s Network Communitarianism are particularly helpful. As technology progresses faster than Moore’s law could have predicted, regulators cannot forget to acknowledge and embrace every tool at their disposal. 224 B. Clark, A. Ebrahim, “Risk Shifting and Regulatory Arbitrage: Evidence from Operational Risk” (2017) Available at SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_ id=2991789 225 European Commission, “State Aid: Ireland gave illegal tax benefits to Apple worth up to €13 billion” (2016), European Commission Press Release, Brussels, available at http://europa.eu/rapid/press-release_IP-16-2923_en.htm, accessed 22nd June 2019 226 C. Cadwalladr, E. Graham-Harrison, “Revealed: 50 Millions Facebook Profiles Harvested for Cambridge Analytical in Major Data Breach” The Guardian (London 17th March 2018), available at https://www.theguardian.com/news/2018/ mar/17/cambridge-analytica-facebook-influence-us-election, accessed on 22nd June 2019 227 P. P. Polanksi, Customary Law Of The Internet: In The Search For A Supranatural Cyberspace Law (T.M.C Asser Press, The Hague, 2007) p. 347 228 J. Horne, Cross-border Internet Dispute Resolution (Cambridge University Press, New York, 2009)

In a Singaporean context, the RTBF represents both a unique techno-legal and cultural challenge to the existing regulatory regime. While the PDPC’s proposed model Framework does show that Singaporean regulators do seem primed to tackle future regulatory challenges, especially regarding AIs, regulators must take care not to leave rights by the wayside – a stark contrast to the issues faced by the GDPR. In this sense, the RTBF and the other rights of the individual may not pose a challenge but rather an opportunity for regulators; balanced between the technology-first approach represented in the Framework and the safeguarding of the individual rights to privacy, Singaporean regulation would be well-poised to become a global leader in AI regulation.

ARTICLES

Abandoning Old Banks and Brooking New Streams: A Case for the Reformation and Statutory Codification of the English Law on A Bank’s Duty of Confidentiality Filbert Lam, University of Edinburgh

I. INTRODUCTION The banker’s duty of confidentiality originates from payment systems of antiquity.1 This precipitated from when temples stored worldly and otherworldly treasures to the Civil Codes and lex mercatoria.2 The development of the English law on bank confidentiality culminated in the landmark decision of the Court of Appeal of England and Wales in Tournier v National Provincial and Union Bank of England (Tournier),3 which held this duty is implied in the bankercustomer contract.4 This duty is subject to four qualifications. These qualifications are: (a) Where disclosure is under compulsion by law; (b) where there is a duty to the public to disclose; (c) where the interests

1 R-M B Antoine, Confidentiality in Offshore Financial Law (2nd edn, OUP 2014) para 2.04. 2 ibid.; J O’Donovan and V Priskich, Lender Liability (2nd edn, Lawbook Co 2016) para 3.10; W Meier, ‘Bank Secrecy in Swiss and International Taxation’ (1973) 7 International Lawyer 16, 17; F Neate, ‘Bank Secrecy’ (1979) 7 International Business Law 230. 3 Tournier v National Provincial and Union Bank of England (Tournier) [1924] 1 KB 461 (CA); The banker-customer relationship is a contractual one, as between debtor and creditor: Foley v Hill (1848) 2 HLC 28; 9 ER 1002, at 35-6 (HLC); See also, European Bank Ltd v Citibank Ltd (2004) 60 NSWLR 153; [2004] NSWCA 76 [21] (Court of Appeal of the New South Wales, Australia); See also P Hood, Principles of Lender Liability (OUP 2012) para 2.14; EP Ellinger, et al, Ellinger’s Modern Banking Law (5th edn, OUP 2011) 176; G Godfrey and S Elcock, “Chapter 13: England” in G Godfrey and FW Neate, Neate and Godfrey Bank Confidentiality (5th edn, Bloomsbury 2011) para 13.2; K Stanton, ‘Chapter 12: The United Kingdom’, in S Booysen and D Neo (eds), Can Banks Still Keep a Secret? Bank Secrecy in Financial Centres around the World (CUP 2017) 342; RM Goode, ‘The Banker’s Duty of Confidentiality’ [1989] JBL 269; O’Donovan and Priskich (n 2) para 3.90; However, the origins of the duty in English law dates back to the 19th Century: Taylor v Blacklow 132 ER 401; [1836] 11 WLUK 33 (CCP); Tassell v Cooper (1850) 9 Common Bench Reports 509; 137 ER 990 (CCP); Foster v Bank of London (1862) F & F 214; 176 ER 96 (Assizes); Hardy v Veasey (1868) LR 3 Ex 107 (Ex Ct); See, for a historical analysis of the law, R Stokes, ‘The Genesis of Banking Confidentiality’ (2011) 32(3) Journal of Legal History 279. 4 Tournier (n 3), per Scrutton LJ, at 480.

of the bank require disclosure; (d) where the disclosure is made by the express or implied consent of the customer.5 However, recent legislative developments have narrowed the scope of this duty. These include the tightening of anti-money laundering and counter-terrorist financing (“AML/CTF”) and anti-tax evasion rules.6 This paper argues that the common law duty of confidentiality is unduly tempered by statutory incursions. This paper suggests that, in order to protect the wide scope of the bank’s duty of confidentiality, this duty and its exceptions must be codified and put on a statutory footing These propositions will be advanced in three stages. Part II critically analyses the decision in Tournier. This section studies the law of implied terms and suggests that the Tournier case may be decided differently today, given the recent developments in contract law. Part III discusses the statutory incursions into the common law. The two primary statutory limitations on the banker’s duty of confidentiality originate from AML/CTF and anti-tax evasion statutes. Finally, Part IV argues for the codification and substantive reformation of the duty. This paper primarily argues for a rights-based approach of codification which allows for incursions into the banker’s duty of confidentiality in statutorily-defined circumstances and insofar as the measures used to achieve the objectives are necessary and proportionate means to the ends.

5 Tournier (n 3), per Bankes LJ, at 473 (emphasis added). 6 Review Committee on Banking Services Law (“Jack Committee”), Banking Services: Law and Practice Report by the Review Committee (“Jack Committee Report”) (CM 622, 1989) Ch 5.

ARTICLES

II. CRITICAL ANALYSIS OF TOURNIER The coherence of the Tournier decision must be questioned in light of developments in the law of implied terms. Moreover, the line of cases following Tournier reveal important lacunae in the law. These arguments support the case for the clarification of this duty in a statute. 1. Tournier and Implied Terms In Tournier, the impugned disclosure arose when the plaintiff-customer indorsed a cheque, drawn by another customer, to third parties. When the cheque was returned defendantbank, its manager discovered the third parties were bookmakers. This was disclosed to the plaintiff ’s employers who refused to renew his employment.7 The Court of Appeal of England and Wales held the bank violated its implied contractual duty of confidentiality, which extends to any “information was obtained arose out of the banking relations of the bank and its customers”.8 This author argues the legal bases for the implication of this contractual duty into the contract in Tournier were unclear. English law today recognises that a term may be implied in law but not implied in fact.9 This affects the scope and contents of the implied term. Implied terms are sometimes used to impose a legal duty on a party to a dispute. However, at the time Tournier was decided, English courts did not distinguish between the two main categories of implied terms.10 In the 1920s, the leading case on implied terms was In re Comptoir Commercial Anversois and Power.11 At the time, the test of implied terms 7 Tournier (n 3), per Bankes LJ, at 468-70. 8 Tournier (n 3), per Atkin LJ, at 485; Cf, Ibid., per Scrutton LJ, at 481, dissenting on this point 9 Liverpool City Council v Irwin [1977] AC 239, at 283-90; Barrett v Lounava (1982) Ltd [1990] 1 QB 348. 10 See, for instance, Lister v Romford Ice and Cold Storage Co Ltd [1957] AC 555, at 574. 11 In re Comptoir Commercial Anversois and Power

applied consistently by the English courts was that such a term: must be such a necessary term that both parties must have intended that it should be a term of the contract, and have only not expressed it because its necessity was so obvious that it was taken for granted12 The fundamental distinction between terms implied in law and fact did not exist then. Their Lordships in Tournier held conflicting opinions on the nature of the implied term. Bankes LJ suggests that this contractual duty was, in contemporary terms, implied in law when his Lordship held, “the contractual duty of secrecy [was] implied in the relation of banker and customer”.13 However, both Atkin and Scrutton LLJ held the term was to be implied in fact.14 For instance, Scrutton LJ held, for a term to be implied into a contract, it “must necessarily have been in the contemplation of the parties in making the contract”.15 Both approaches further departed from previous decisions on bank confidentiality, which considered that the term was implied by trade custom and usage.16 In contrast, the contemporary law of implied terms distinguishes between terms implied in fact and law.17 Terms implied in fact “rest on the parties’ objective intentions”18 determined by (“Comptoir”) [1920] 1 KB 868. 12 Comptoir (n 11), per Scrutton LJ, at 899-900. 13 Tournier (n 3), per Bankes LJ, at 473 (emphasis added). 14 Tournier (n 3), per Scrutton LJ, at 480; per Atkin LJ, at 483. 15 Tournier (n 3), per Scrutton LJ, at 480 (emphasis added). 16 Foster v Bank of London (n 3) 217; O’Donovan and Priskich (n 2) para 3.70; This approach was expressly rejected in Turner v Royal Bank of Scotland (“Turner”) [1999] 2 All ER (Comm) 664; [1999] Lloyd’s Rep: Banking 231; On terms implied by custom and trade usage: E Peel, Treitel on The Law of Contract (“Treitel”) (14th ed, Sweet & Maxwell 2015) paras 6-048 to 6-050. 17 The landmark case on this point is Reid v Rush & Tompkins Group Plc [1990] 1 WLR 212; But see also, Shell UK Ltd v Lostock Garage Ltd [1976] 1 WLR 1187 (CA), per Lord Denning MR, at 1196, on the two different categories of implied terms. 18 A Burrows, A Restatement of the English Law of Contract (OUP 2013) 93 (emphasis added); Treitel (n 16) paras 6-032 to 6-042; BP Refinery (Westernport) Pty Ltd v Shire of Hastings (1978) 52 ALJR 20, per Lord Simon, at 26 (PC); This does not refer to

ARTICLES

applying either the “officious bystander”19 or “business efficacy”20 tests.21 Conversely, terms are only implied in law “where ‘necessary’ to the particular type of contract or relationship”22 without inference to the parties’ objective intentions.23 In this author’s view, the duty of confidentiality is one which is implied by law in the bankercustomer contract. This is because such a term is implied given the nature of the bankercustomer relationship, not the specific intention of the parties, embodied in the contract and its terms. The ambiguity in the legal bases for the implication of the duty of confidentiality in bank-customer contracts has led to controversies concerning the proper scope of the duty and the Tournier qualifications. 2. Lacunae in the Qualifications A bank’s duty of confidentiality is a qualified, not absolute, not absolute duty.24 Where a Tournier qualification applies, the bank is under no duty of confidentiality.25 However, the scope of these qualifications is often unclear. Earlier decisions suggest that “necessity, not reasonableness” governs the implication of

the actual intention of parties: Equitable Life Assurance Society v Hyman [2002] 1 AC 408, per Lord Steyn, at 459. 19 Shirlaw v Southern Founderies (1927) Ltd [1939] 2 KB 206, per MacKinnon LJ, at 227. 20 The Moorcock (1889) 4 PD 64 (CA). 21 Marks and Spencer plc v BNP Paribas Securities Services Trust Co (Jersey) Ltd and another (“Marks and Spencer”) [2015] UKSC 72, [2016] AC 742, per Lord Neuberger of Abbotsbury, at [21]; Attorney General of Belize v Belize Telecom Ltd (“Belize”) [2009] 1 WLR 1988, per Lord Hoffmann, at [27]. 22 Burrows (n 18) 94 (emphasis added); Treitel (n 16) paras 6-043 to 6-047; Liverpool CC v Irwin [1977] AC 239 (HL); Scally v Southern Health and Social Services Board (“Scally”) [1992] 1 AC 294, at 302 (“the search, based on wider considerations, for a term which the law will imply as a necessary incident of a definable category of contractual relationship”, emphasis added) (HL); Mahmud v Bank of Credit and Commerce International SA [1998] AC 20 (HL), per Lord Steyn, at 45 (a “standardised term implied by law”, emphasis added). 23 Treitel (n 16) para 6-044. 24 Tournier (n 3), per Bankes LJ, at 472-3; and per Atkin LJ, at 484; See also, Robertson v Canadian Imperial Banking Corpn [1994] 1 WLR 1493 (PC); Lipkin Gorman v Karpnale Ltd [1989] 1 WLR 1340 (CA); Turner v Royal Bank of Scotland (n 16). 25 Barclays Bank plc v Taylor [1989] 1 WLR 1066, per Lord Donaldson of Lymington MR, at 1074; El Jahwary v Bank of Credit and Commerce International SA [1993] BCLC 396, per Sir Donald Nicholls V-C, at 400; Hood (n 3) para 2.15.

terms in law.26 However, terms are implied in law “based on wider policy grounds”.27 Dyson LJ held the implied term must be fair, reasonable and balance competing policy considerations.28 Thus, the considerations of “justice and policy” are the contemporary requirements for contractual terms to be implied in law.29 These considerations of “justice and policy” in turn affect the scope of the Tournier qualifications. a. Legal Compulsion Bankers may be required by statute or a court order to disclose a customer’s information.30 Parties cannot contract out of their duty “to comply with the law of the land”.31 However, several questions linger about the scope of this qualification. Where a banker is unable to contact their customer about a legally compelled disclosure, is he under an implied obligation to inform the court? Courts have not hitherto held that banks are required to do so.32 Banks will also have to beware that information which is not subject of the proceedings should remain confidential.33 Further, it is also unclear if the bank should obtain a customer’s approval before the disclosure.34 Thus, in this author’s view, this 26 Scally (n 21), at 307; Cf Tai Hing Cotton Mill Ltd v Liu Chong Hing Bank Ltd [1986] AC 80, at 104-5; Spring v Guardian Insurance Plc [1995] 2 AC 296, at 339 and 354; Reid v Rush & Tompkins Group Plc (n 17), at 220. 27 Burrows (n 18) 94; Shell UK Ltd v Lostock Garage Ltd (n 17), per Lord Denning MR, at 1196. 28 Crossley v Faithful & Gould Holdings Ltd [2004] EWCA Civ 293, [2004] 4 All ER 447, per Dyson LJ, at [36]; See also, E Peden, ‘Policy concerns behind implication of terms in law’ (2001) 117 LQR 459. 29 The Star Texas [1993] 2 Lloyd’s Rep 445, at 491; Crossley v Faithful & Gould Holdings Ltd (n 28), per Dyson LJ, at [36]; Treitel (n 16), para 6-045. 30 Hood (n 3) para 2.29 to 2.31. 31 Parry-Jones v Law Society of England and Wales [1969] 1 Ch 1, per Diplock LJ, at 9 (CA). 32 Robertson v Canadian Imperial Bank of Commerce [1994] 1 WLR 1493, per Lord Nolan, at 1499-1500 (PC). 33 Hood (n 3) para 2.33; Robertson v Canadian Imperial Bank of Commerce (n 32), per Lord Nolan, at 1499; Pharaon v Bank of Credit and Commerce International SA (in liq) [1998] 4 All ER 455, per Rattee J, at 465-6. 34 Robertson v Canadian Imperial Bank of Commerce (n 32),

ARTICLES

qualification is not well defined in law, which leads to very real concerns given that legal compulsion is perhaps the qualification with the sharpest teeth. b. Public Duty The scope of this qualification is vast.35 An issue of public interest is “something which is of serious concern and benefit to the public”, not merely something “the public would … find interesting”.36 These include the prevention of wrongdoing and miscarriages of justice and maintenance of public safety.37 The most important justification warranting disclosure of the information is bank supervision.38 This may be enforced through non-statutory public inquiry bodies39 and private proceedings over allegations of fraud.40 This is difficult to “apply with certainty”41 and is not, as the Jack Committee Report suggests,42 merely a matter of looking at statutory provisions in many cases.43 A balance must be struck between private interests in keeping the information confidential and the “countervailing public interest” in disclosure.44 per Lord Nolan, at 1499. 35 Hood (n 3) para 2.51 (“a nebulous and changing concept”). 36 British Steel Corpn v Granada Television Ltd [1981] AC 1096, per Sir Robert Megarry V-C, at 1113; See also, National Irish Bank Ltd v Radio Telefis Eireann [1998] 2 IR 465, per Keane J (dissent), at 486 (Irish Supreme Court). 37 Price Waterhouse v BCCI Holdings (Luxembourg) SA (n 44), per Millett J, at 601. 38 Hood (n 3) paras 2.52 ff. 39 See generally Price Waterhouse v BCCI Holdings (Luxembourg) SA (n 44), where the disclosure was made to the Bingham Inquiry on the Bank of England’s performance of statutory supervisory functions concerning the BCCI group. 40 See generally Pharaon v Bank of Credit and Commerce International SA (in liq) (n 33). 41 Hood (n 3) para 2.48. 42 Jack Committee Report (n 6) paras 5.30 and 5.41. 43 Hood (n 3) para 2.52 (“[T]he public interest exemption sometimes taking on a gap-filling role, where there is no relevant legislation, though such cases will be rare.”); Price Waterhouse v BCCI Holdings (Luxembourg) SA (n 44), per Millett J, at 598. 44 Attorney General v Observer Ltd (“Spycatcher”) [1990] 1 AC 109; [1988] 3 WLR 776, per Lord Keith of Kinkel, at 256 (HL); Ibid., per Lord Griffiths at 269; Ibid., per Lord Goff of Chieveley, at 282; Price Waterhouse v BCCI Holdings (Luxembourg) SA [1992] BCLC 583, per Millett J, 596 and 598.

There are several uncertainties concerning the exact scope of this qualification. A party requesting disclosure to be made to them must have a public interest in receiving the confidential information.45 Disclosures may not be made to the public generally,46 but in a limited form to law enforcement or regulatory authorities.47 Disclosures of such confidential information also cannot exceed what is “reasonably necessary to achieve the purpose of that public interest”.48 However, as the term remains ambiguous within this area of banking law, the public interest qualification, like the notion of “public policy”, runs the risk of turning into an “unruly horse”.49 c. Lenders’ Interest Banks are allowed to disclose information to prevent financial loss and reputational damage,50 but may not do so in order to obtain a benefit.51 Disclosures may not be made to credit agencies52 or through marketing campaigns without their customers’ consent.53 Permitted disclosures may include the enforcement of legal rights, protection from reputational damage,54 and assignment of debts.55 This qualification may permit disclosure in legal proceedings for payment of an overdraft56 and dishonouring 45 Initial Services v Putterill [1968] 1 QB 396, per Lord Denning MR, at 405. 46 Spycatcher (n 44), per Lord Griffiths, at 269. 47 ibid.; National Irish Bank Ltd v Radio Telefis Eireann (n 36), per Keane J (dissent), at 483-5. 48 Pharaon v Bank of Credit and Commerce International SA (in liq) (n 33), per Rattee J, at 465; Price Waterhouse v BCCI Holdings (Luxembourg) SA (n 44), per Millett J, 602; National Irish Bank Ltd v Radio Telefis Eireann (n 36), per Lynch J, at 495; Ibid., per Keane J (dissent), at 485. 49 Richardson v Mellish (1824) 2 Bingham 229; 130 ER 294, per Burrough J, at 303, CCP (“[P]ublic policy … is a very unruly horse, and when once you get astride it you never know where it will carry you.”) 50 Tournier (n 3), per Atkin LJ, at 486; X AG v A bank [1983] 2 All ER 464, per Leggatt J, at 477-9. 51 ibid. para 2.79. 52 Hood (n 3) paras 2.106 to 2.117. 53 Hood (n 3) para 2.79. 54 Sunderland v Barclays Bank Ltd (1938) 5 Legal Decisions Affecting Bankers 163. 55 Hood (n 3) para 2.82. 56 Tournier (n 3), per Bankes LJ, at 473; Ibid., per Scrutton LJ, at 481. May

ARTICLES

cheques due to insufficient funds.57

this industry practice.65

However, disclosures made under a subpoena order in foreign proceedings are not allowed under this qualification.58 In X AG v A Bank, Leggatt J held a foreign court subpoena was also “totally different” from the original circumstances contemplated under the Tournier exception.59 Since foreign court orders do not bind English courts, the legal compulsion exception also does not apply to disclosures made in purported compliance of foreign court subpoenas.60 However, Leggatt J did not provide a clear indication of the circumstances which were contemplated by the court in Tournier to be included under this qualification. Thus, whereas one can be fairly certain that disclosures made in compliance with foreign court orders do not constitute legitimate disclosures made in the interests of lenders, neither Tournier nor X AG v A Bank definitively identify the scope of this qualification.

Further, intra-group disclosures made for marketing purposes may violate the principle that “personal data … must not be processed in a manner that is incompatible with the purpose for which it was collected” under the Data Protection Act 2018.66 They may be permitted if the data processing for “internal administrative purposes” constitutes a legitimate interest.67 However, since this is governed by statutory data protection laws, we do not know the exact scope of this common law implied term.

Two outlying uncertainties concerning the application of the lender’s interests qualification which remain unaddressed by the courts concern the disclosures made within a “banking group”61 and to credit rating agencies (“CRAs”).62 First, subsidiaries in a group are separate legal entities and sharing information from one to another, or their parent is unlawful.63 Lenders may want to share information to advertise financial services and products to their customers.64 It is unlikely that customers could have implicitly consented to such disclosures since most are unaware of

57 ibid., per Scrutton LJ, 481. 58 Hood (n 3) paras 2.84 to 2.86. 59 X AG v A bank (n 50), per Leggatt J, at 479. 60 British Nylon Spinners Ltd v Imperial Chemical Industries Ltd [1953] Ch 19. 61 Defined in Companies Act 2006, s 1164(4), (5). 62 Hood (n 3) paras 2.90 ff. 63 Bank of Tokyo Ltd v Karoon [1987] 1 AC 45n, per Ackner LJ, at 53-5 (CA); Ibid., per Goff LJ, at 63-4; Bhogal v Punjab National Bank [1988] 2 All ER 296, per Bingham LJ, at 305 (CA); R v Grossman (1981) 73 Cr App Rep 302 (CA); On separate legal personality under general company law: Saloman v A Salomon & Co Ltd [1897] AC 22 (HL); Adams v Cape Industries plc [1990] Ch 433 (CA); Prest v Petrodel Resources Ltd [2013] UKSC 34, [2013] 2 AC 415 (UKSC). 64 Hood (n 3) para 2.91.

Finally, the extent to which information may be disclosed to CRAs was also unclear.68 Previously, under the Lending Code,69 banks should obtain their customer’s consent before disclosing any information to CRAs. While this removed the distinction between “white” and “black” information,70 it does not change the general perception the customer’s consent must be obtained for the bank to disclose information to CRAs.71 Unfortunately, the courts have left both these issues undecided.72 The Jack Committee thought disclosures should be allowed where 65 Turner v Royal Bank of Scotland (n 16); Jack Committee Report (n 6), paras 5-12 and 5-15. 66 Data Protection Act 2018 (“DPA”), s 36(1); The DPA implements and supplements the Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation) (“GDPR”) [2016] OJ L 119. 67 GDPR, Recital 48 and Article 6(1)(f); P Voigt and A von dem Bussche, The EU General Data Protection Regulation (GDPR): A Practical Guide (Springer 2017) 107. 68 Tournier (n 3), per Scrutton LJ, at 481; ibid., per Atkin LJ, at 486; Hood (n 3) paras 2.106 ff. 69 British Bankers’ Association and The UK Cards Association, The Lending Code: Setting standards for banks, building societies, credit card providers and their agents (“Lending Code”) (March 2011, Revised September 2015), available at https://www. lendingstandardsboard.org.uk/wp-content/uploads/2016/06/ The-Lending-Code-Mar-2011-revised-2015.pdf, accessed 3 Dec 2018, Section 3, paras 36-7, 41 and 48. 70 Hood (n 3) paras 2.107 to 2.113. 71 Hood (n 3) para 2.113. 72 Turner v Royal Bank of Scotland (n 16); Hood (n 3) para 2.114.

ARTICLES

it was in the lender’s interests, but there was no public interest in such disclosures.73 Today, the Standards of Lending Practice stipulate that banks may share information with CRAs under the principles of reciprocity.74 While there was some indication that the courts thought such disclosures would promote “responsible lending” under the old data protection regime,75 it is not clear if the common law adopts the same position. Thus, it is this author’s view that the lenders’ interests qualification requires some much-needed clarifications, especially concerning how it interacts with “soft” law principles and codes of conduct. d. Consent Banks often also justify disclosing information to CRAs based on the banks’ customers’ implied consent.76 This view is controversial within academic circles.77 However, in Turner,78 the Court of Appeal held the customer must know of the matter to which they have impliedly agreed.79 Disclosures must be accompanied by the customer’s knowledge.80 Finally, customers may explicitly override the implied duty of confidentiality.81 Difficulties arise over disclosures made to guarantors, which may amount to an unlawful disclosure.82 Lenders have no duty to disclose 73 Jack Committee Report (n 6), para 5.15. 74 Lending Standards Board, The Standards of Lending Practice: Personal Customers (7 December 2018), available at https://www.lendingstandardsboard.org.uk/wp-content/ uploads/2017/12/Standards-of-Lending-Practice-Personal7-Dec.pdf, accessed 16 December 2018, at 7, para 7; Lending Standards Board, The Standards of Lending Practice: Business Customers (28 June 2018), available at https://www. lendingstandardsboard.org.uk/wp-content/uploads/2018/09/ Standards-of-Lending-Practice-Business-28-Jun-18.pdf, accessed 16 December 2018, at 9, para 5. 75 McGuffick v Royal Bank of Scotland Plc [2009] EWHC 2386 (Comm); [2010] Bus LR 1108. 76 Tournier (n 3), per Bankes LJ, at 473; Ibid., per Atkin LJ, at 486. 77 Hood (n 3) para 2.134 and fn 438 for academic commentary on this point. 78 See n 16 79 Turner (n 16), per Judge LJ, at 235-7. 80 Turner (n 16), per Judge LJ, at 235. 81 Hood (n 3) para 2.148. 82 Hood (n 3) paras 2.149 to 2.154; RBS v Etridge (No 2)

the debtor’s financial information to the guarantor since they are not bound by uberrimae fidei obligations.83 In any event, full disclosure would be commercially impractical.84 However, the lender must inform the guarantor of matters which do not relate to the banker-customer relationship.85 Finally, when a guarantor enquires about the debtor’s financial information before signing the guarantee, lenders should always answer honestly.86 3. Reformulating the Implied Term The scope of the implied term in a “definable category of contractual relationship[s]” is determined by the relevant public policy considerations.87 The banker’s duty of confidentiality applies when he deals with the customer qua banker.88 This author proposes to distinguish between rights-based and interestsbased qualifications.89 The rights-based qualifications deal with disclosures with the customer’s consent qualifications. The bank has a right to share information within the limits of the consent given. Customers may restrict to whom disclosures may be made, how far the information may be shared, and in what form such disclosures should take. Interests-based qualifications allow for disclosures which are legally compelled, in the public interests, and furtherance of the [2001] UKHL; [2002] AC 773, per Lord Nicholls of Birkenhead, at [79(2)]. 83 Hamilton v Watson (1854) 12 Cl & Fin 109, at 119; Seaton v Heath [1899] 1 QB, per Romer LJ, at 792; National Provincial Bank of England v Glanusk [1913] 3 KB 335, 338; Etridge (No 2) (n 82), per Lord Nicholls of Birkenhead, at [79]; North Share Ventures Ltd v Anstead Holdings Inc [2011] EWCA Civ 320; [2011] 3 WLR 628, at [21] and [29]. 84 North Share Ventures Ltd v Anstead Holdings Inc (n 83), at [31]. 85 Hamilton v Watson (n 83), at 119; Etridge (No 2), per Lord Nicholls of Birkenhead, at [81] and [188]; Levett v Barclays Bank plc [1995] 1 WLR 1260 (HC); North Share Ventures Ltd v Anstead Holdings Inc (n 83), at [21] and [28]. 86 Hamilton v Watson (n 83). 87 Scally (n 21), at 302; Treitel (n 16) para 6-045. 88 Tournier (n 3), per Bankes LJ, at 474; Barclays Bank plc v Taylor (n 25), per Lord Donaldson of Lymington MR, at 1070. 89 See, for instance, Jack Committee Report (n 6), paras 5.06 to 5.14.

ARTICLES

banks’ interests. Statutorily-required disclosures address overriding public interests, which are codified.90 Since most statutorily compelled disclosures are in the public interests, some argue this distinction is negligible.91 This author disagrees with the Jack Committee’s proposal to abolish the public interest qualification.92 The common law should be left to develop where a statute does not protect such interests. Parliament may subsequently opt to codify such rules in statute if there is a public interest for such legislation. Banks have the right to disclose such information where it is in the public interest to do so.93 In contrast, disclosures in the banks’ interests are primarily made to “shield” banks from suffering financial loss or reputational damage. These qualifications do not involve a right to disclose confidential information, but some underlying public or private interests for such disclosures. One may criticise this distinction as arbitrary since the Tournier qualifications are not excuses, but quasi-exceptions.94 Where any of them apply, the bank has the right to disclose the information within limits. However, this author maintains that the distinction is useful in delineating the jurisprudential bases for such rights. This distinction allows courts to determine the precise scope of the implied term and assists legislators in codifying the obligation. III. STATUTORY INCURSIONS INTO THE COMMON LAW The common law applies where statutes are silent.95 However, through a slew of statutory

90 See Part III. 91 Jack Committee Report (n 6), paras 5-30 and 5.48; Ellinger (n 3) 190. 92 Jack Committee Report (n 6), para 5.48. 93 Ellinger (n 3) 190; Hood (n 3) paras 2.54 to 2.56; Price Waterhouse v BCCI Holdings (Luxembourg) SA (n 44), per Millett J, at 601. 94 See fn 25. 95 On the interaction between common law and statutory law, see Action at Common Law (1774) Lofft 493,

enactments, banking confidentiality has been rendered nugatory in key areas of modern banking practice. This section seeks to highlight key features of the AML/CTF and anti-tax evasion legal frameworks. The limited space means that the treatment of these issues will be ambulatory to the central proposition advanced by this paper. 1. AML/CTF There is a significant overlap between the AML and CTF regimes. The former is mainly governed by the Proceeds of Crime Act 2002 (“POCA”);96 and the latter by the Terrorism Act 2000 (“TA”). However, both regimes are similar since the TA prohibits “terrorist money laundering”.97 This section will analyse the AML regime and highlight the parts of the CTF which differ from the former. a. AML Money laundering is a criminal offence characterised by (a) dealings with “criminal property”;98 (b) engaging in “an arrangement” which “facilitates” the dealing with “criminal property”;99 or (c) acquiring, using or

at 493; Johnson v Unisys Ltd [2001] UKHL 13, per Lord Millett, at [80]; In Re McKerr (AP) (Northern Ireland) [2004] UKHL 12, per Lord Hoffmann, at [73]; Deutsche Morgan Grenfell Group plc v Inland Revenue Comrs [2007] 1 AC 558, per Lord Hoffmann, at [19]; R (on the application of Rowley) v Secretary of State for Work and Pensions [2007] EWCA Civ 598, per Dyson LJ, at [73]; R (on the application of the Child Poverty Action Group) v Secretary of State for Work and Pensions [2010] UKSC 54, per Lord Dyson, at [32] and [34]; Moohan and another v The Lord Advocate [2014] UKSC 67, per Lord Hodge, at [34]; Keyu and others v Secretary of State for Foreign and Commonwealth Affairs and another [2015] UKSC 69, per Lord Mance, at [150]; Darby v Richmond Upon Thames LBC [2017] EWCA Civ 252, per Thirlwall LJ, at [20]; Investment Trust Companies (In Liquidation) v Revenue and Customs Commissioners [2017] UKSC 29, per Lord Reed, at [86]. 96 See Part 7 of the Proceeds of Crime Act 2002 (“POCA”), as amended by the Serious Organised Crime and Police Act 2005 (“SOCPA 2005”). 97 Neate and Godfrey (n 3) para 13.28 98 POCA, s 327, “conceal, disguise, convert, transfer or remove criminal property … from the UK”; “Criminal Property” is defined in s 340(3). 99 POCA, s 328, “an arrangement which … facilitates the acquisition, retention, use or control of criminal property”.

ARTICLES

possessing “criminal property”.100 The POCA creates criminal offences of failure to disclose information in the “regulated sector”.101 The POCA creates criminal offences of failure to disclose information in the “regulated sector”. Three offences are worth highlighting. First, persons employed in the regulated sector commit an offence when they obtain information which provides reasonable grounds for suspecting money laundering and fails to report it.102 Second, a money laundering reporting officer (“MLRO”) in a regulated sector commits an offence if the officer receives such information but does not make a suspicious transaction report to the National Crime Agency.103 Such “suspicion” must be “more than fanciful” or “vague”.104 Finally, it is an offence for officers to “tip-off ” another person by informing them a disclosure has been made to the relevant authorities.105 Besides, the Money Laundering Regulations 2007106 sets out guidelines applicable to banks, where non-compliance attracts criminal sanctions. This includes customer identification procedures, staff training and internal reporting procedures, including identifying the MLRO. It is submitted these represent substantial incursions on the duty of confidentiality, albeit for legitimate aims. The criminalisation of nondisclosure enables law enforcement authorities to investigate the potential commission of a predicate offence.107 Further, the provisions set 100 POCA, s 328. 101 POCA, Schedule 9; The “statutory obligations are functional rather than institutional”: R Fortson, “Money Laundering Offences under POCA 2002” in W Blair and R Brent, Banks and Financial Crime: The International Law of Tainted Money (2nd edn, OUP 2017) paras 7.35 and 7.37. 102 POCA, s 330. 103 POCA, s 331. 104 Neate and Godfrey (n 3) para 13.24; R v Da Silva [2006] EWCA Crim 1654; Shah v HSBC Private Bank (UK) Ltd [2012] EWHC 1283; Iraj Parvizi v Barclays Bank Plc [2014] EWHC B2 (QB). 105 POCA, s 333A. 106 SI 2007/ 2157. 107 POCA, ss 327 (“Conceals, disguises; converts, transfers”), 328 (“Concerned in an arrangement”), and 329

out a wide range of defences, such as the lack of training108 alternatively, reasonable belief that the activity took place outside the UK and was lawful.109 Thus, there is a legitimate public interest in legislating for such disclosures to be mandatory. b. CTF The TA largely mirrors the POCA regime in criminalising “terrorist money laundering”.110 The Act includes offences of (a) money laundering;111 (b) failing to report terrorist money laundering;112 and (c) tipping-off another person about a disclosure.113 However, when the TA was amended in 2001,114 two new offences and one law enforcement power were included which differ from the POCA regime. First, it is an offence where businesses in the regulated sector fail to report such unlawful activity.115 This differs from the POCA regime because Section 21A applies both when there was a failure (i) to report actual knowledge or suspicion, and (ii) by an institution or person to identify and report “reasonable grounds” for such knowledge or suspicion.116 This introduces an objective standard for criminal liability, which broadens the CTF regime’s scope.117 Second, section 21D introduces a tippingoff offence applicable to regulated sectors.118 Whereas the general tipping-off offence criminalises the disclosure of information “which is likely to prejudice the investigation”,119 (“acquisition, use or possession”); Some academics argue that this applies only to section 330 of POCA, but this perhaps also applies to section 331: Banks and Financial Crime (n 101) para 7.158. 108 POCA, s 330(7). 109 POCA, s 330(7A) as amended by section 102(5) of SOCPA 2005. 110 See fn 97. 111 TA, s 18. 112 TA, s 19. 113 TA, s 39. 114 Amended by the Anti-Terrorism Crime and Security Act 2001. 115 TA, s 21A. 116 ibid. (emphasis added). 117 Neate and Godfrey (n 3) para 13.28. 118 TA, s 21D. 119 TA, s 39 (emphasis added); POCA, s 333A(2)

ARTICLES

section 21D further criminalises the disclosure of “any matter” reported to a law enforcement authority or employer.120 It is submitted this undesirably broadens the scope of the offence since disclosure “any matter” by a bank’s employees may incur criminal penalties. Finally, the police now have reinforced powers to conduct investigations. They may obtain monitoring orders to compel banks to provide customers’ information. They also have powers to seize, freeze and forfeit suspicious funds.121 Thus, the disclosure of customers’ information involves further intrusive investigations and restrictions on personal information and property. 2. Anti-Tax Evasion Tax authorities may gain access to bank account information. Schedule 36 of the Finance Act 2008 confers on the authorities wide-ranging powers to compel disclosure of bank account information.122 This may also be shared with foreign tax authorities.123 This demonstrates the extensive reach of the disclosure requirements. Further, the authorities may compel disclosure of a wide gamut of information. This includes both “past, present and future”124 tax liabilities and “foreign tax”,125 where the UK has a tax information exchange agreement with the foreign jurisdiction. Additionally, they may access Suspicious Activity Reports made by banks to tackle tax evasion.126 However, this is

(emphasis added). The tipping-off offence under the AML regime also includes an addition offence where the tip-off relates to the disclosure that “an investigation into allegations … is being contemplated or carried out”: POCA, s 333A(1). 120 TA, s 21D(1) and (2) (emphasis added). 121 See also Counter-Terrorism Act 2008; Terrorist Asset-Freezing etc. Act 2010; Terrorism Prevention and Investigation Measures Act 2011. 122 Finance Act 2008, Schedule 36 juncto Finance Act 2011, Schedule 23; See especially, Finance Act 2008, Schedule 36, para 2. 123 Commissioners for Revenue and Customs Act 2005, ss 17-23. 124 Finance Act 2008, Schedule 36, para 64(1)(a). 125 Finance Act 2008, Schedule 36, para 63(4). 126 Stanton (n 3) 354.

not a carte blanche for obtaining any confidential information outside of what is necessary for “assessing or collecting taxes”.127 As these mischiefs become more extensive and insidious, legislative intrusions into confidential bank information will increase. This leads to further diminutions of the duty of confidentiality. Not only may banks disclosure confidential information, but there are also more circumstances under which they must do so. This author argues that without a formal codification of the duty of confidentiality, such legislative intrusions may erode the trust and confidence customers place in the financial system. IV. PROPOSALS FOR REFORM Codification crystallises a common law rule and allows the legislature to strengthen it.128 This author suggests the banker’s duty of confidentiality should be codified. This argument will be advanced in three stages. First, there is a strong public interest in the codification of the duty. Next, the rule should be reformed to allow only for legitimate and proportionate interferences into the banker’s duty. Finally, this author considers and rejects the view that bankers should face criminal liability for the breach of this duty. 1. Public Interest in Codification The books are filled with terms implied in law which have been put in the statutory form.129 The public interest considerations underpinning this implied term support the codification of the duty of confidentiality.130 The lacunae in the 127 Stanton (n 3) 357-8. 128 See generally, MA Hogg, ‘Codification of Private Law: Scots Law at the Crossroads of Common and Civil Law’, in K Barker, et al. (eds), Private Law in the 21st Century (Hart Publishing, 2017) 107 ff. 129 Treitel (n 16), para 6-043; See for instance Sale of Goods Act 1979, ss 12-15, especially implied terms about title (s 12) and quality or fitness (s 14); Similar provisions as the Sale of Goods Act 1979 as applied to contracts between a “trader” and “consumer”: See Consumer Rights Act 2015, Part I. 130 Indeed, some commentators suggest that the Tournier

ARTICLES

law involve public policy issues which can be “resolved only by the legislature”.131 Codification also prevents this critical obligation from turning into “something of a shibboleth”.132 There are two critical flaws in leaving the existence and scope of this duty subject to the common law of contractual implied terms. First, this duty may be excluded by express terms,133 even if customers do not expressly consent to the disclosure. This undesirably prejudices their rights to determine how their information is protected or disseminated. Second, third parties will not be subject to the obligation due to the principle of privity.134 These arguments support the proposal for an unambiguous statutory duty of confidentiality. For instance, the Singapore Parliament recognised the strong public interest arguments for legislating on the duty of confidentiality.135 This is imposed by s 47 of the Banking Act (“BA”),136 which aims at “maintaining the confidence of customers in [Singapore’s] banking system”.137 However, the Singapore Government also recognised the pre-2001 regime “impeded banks wanting to take advantage of potential operational benefits and savings”.138 The BA was amended to include exceptions to the statutory duty139 in the Third Schedule.140 qualifications have a “statute-like quality”: See, e.g. Hood (n 3) para 2.17. 131 Treitel (n 16), para 6-046; See also, Reid v Rush & Tompkins Group Plc (n 17); Scally (n 21); Crossley v Faithful & Gould Holdings Ltd (n 28), per Dyson LJ, at [46]; Jack Committee Report (n 6), para 5-48. 132 O’Donovan and Priskich (n 2) para 3.140. 133 Stanton (n 3) 343; See also Primary Group (UK) Ltd v The Royal Bank of Scotland plc (“Primary Group”) [2014] EWHC 1082 (Ch). 134 Stanton (n 3) 343; Primary Group (n 133). 135 See generally A Yeo and J Tan, “Chapter 33: Singapore”, in Neate and Godfrey (n 3). 136 Singapore Banking Act (Cap 19, 2008 Rev Ed), as amended by the Banking (Amendment) Bill 2001. 137 Parliament of Singapore 16 May 2001, Vol 73, Col 1689 (BG Lee Hsien Loong (then Deputy Prime Minister)). 138 Ibid.; Note that this is not a problem under the Tournier exceptions because disclosures in the lenders’ interests can never justify making a profit or operational benefits in general: See Hood, at fn 51. 139 Neate and Godfrey (n 3) para 33.7. 140 BA, Third Schedule.

This allows for disclosures when made for a “lawful purpose” to a “lawful recipient” only when the prescribed “conditions” are complied with.141 This author submits this approach is doctrinally sound and pragmatic and therefore should be implemented in English law. However, the Singapore courts had to grapple with the interaction between the statute and the Tournier decision. VK Rajah JA in Susilawati v American Express Bank Ltd (Susilawati)142 held s 47 sets out an exclusive regime governing the duty of confidentiality and Tournier no longer applies.143 This was based on a plain reading of s 47(1), which provides, “[c]ustomer information shall not … be disclosed … except as expressly provided in this Act”.144 This author respectfully disagrees with the ratio of the Susilawati judgment. When tabling the 2001 amendments, BG Lee Hsien Loong said in Parliament:145 “[A]ll banks will provide a basic level of customer confidentiality to all their customers. Beyond this legal minimum, individual banks and customers may reach their own contractual arrangements offering higher standards of confidentiality.” Section 47 does not hinder the imposition of a higher express or implied contractual duty under common law by the parties. It does, however, impose a statutory de minimis duty which banks must abide by. Having consulted Hansard, this was the intention of the Singapore Parliament in introducing these amendments. In this author’s view, the Singapore Parliament did not, and cannot be seen to have, intended for Section 47 of the BA to govern all other contractual agreements which bind the banks to a duty to provide higher levels of confidentiality. These 141 ibid.; Neate and Godfrey (n 3) para 33.9. 142 Susilawati v American Express Bank Ltd [2009] 2 SLR(R) 737, per VK Rajah JA, at [66]-[67] (Court of Appeal, Singapore). 143 Neate and Godfrey (n 3) para 33.23. 144 BA, s 47(1) (emphasis added). 145 Parliament of Singapore 16 May 2001, Vol 73, Col 1690 (BG Lee Hsien Loong (then Deputy Prime Minister)) (emphasis added in italics and underline).

ARTICLES

protections remain governed by the common law as expressed in Tournier. Nevertheless, there are strong public interest arguments in favour of codification, and the proposed statute may expressly repeal the common law. 2. Substantive Reform This author suggests the statutory duty should mirror the existing jurisprudence on the right to privacy. This may be formulated along these lines: 1. If the information concerned is confidential, there would be a breach of the duty of confidence when such information was disclosed; 2. There must be a legitimate aim pursued in its disclosure; and 3. The means of disclosure must be necessary and proportionate to the ends pursued. This proposed statutory scheme is based on the three-tiered test for establishing a person’s right to privacy as summarised in Goodwin v NGN Ltd (No 3).146 First, the statutory duty of confidentiality should be interpreted and applied in the same way as the general law,147 but under the veneer of a banker-customer relationship. The duty is breached if the information to (a) has “the necessary quality of confidence”; (b) was imparted under the circumstances giving rise to a duty of confidence, and (c) was disclosed without permission or misused.148 Information is confidential if the confider

146 Goodwin v NGN Ltd (No 3) [2011] EMLR 27, per Tugendhat J, at [62(b)]. 147 J Rushbrook and A Speker, ‘Chapter 4: Breach of Confidence’, NA Moreham and M Warby (eds), Tugendhat and Christie The Law of Privacy and the Media (OUP 2016) paras 4.14, 4.15 and 4.63 to 4.102. 148 Saltman Engineering v Campbell Engineering [1948] 65 RPC 203, per Lord Greene MR, at 215; Coco v AN Clark (Engineers) Ltd [1969] RPC 41, per Megarry J, at 47, approved in AG v Observer Ltd [1990] 1 AC 109, per Lord Griffiths, at 168 and Douglas v Hello! Ltd [2007] UKHL 21; [2008] 1 AC 1, at

reasonably believed it was confidential under the “usages and practices of the particular trade or industry concerned”.149 Such information must not have been in the “public domain”.150 Information protected from “prying eyes” may also be confidential.151 Further, while the information need not be novel,152 it should not be trivial153 and must be detailed and identifiable.154 Most importantly, the information must have come into the bank’s cognisance qua banker. Second, at first blush, most of the Tournier exceptions, except for the lender’s interest exception, may be different species of legitimate aims. Upon closer inspection, however, the difficulty with applying the lender’s interest exception to intra-group disclosures or to CRAs may be side-stepped by merely obtaining the customer’s consent. Thus, it is suggested that all the Tournier exceptions are different legitimate aims which potentially justifies disclosure. Finally, the proportionality test is a novel development in this area of the law. According to Bank Mellat v HM Treasury (No 2),155 courts must consider if the objective is sufficiently important and the means are rationally connected to accomplishing it, the least reasonably intrusive and strikes a fair balance [307]; R v Dept of Health, ex p Source Informatics [2001] QB 424, per Simon Brown LJ, at [14]; Force India Formula One Team Ltd v 1 Malaysia Racing SDN BHD (“Force India”) [2012] EWHC 616 (Ch), per Arnold J, at [215]-[224] (affirmed in [2013] EWCA Civ 780; [2013] RPC 36). 149 Rushbrooke and Speker (n 147) para 4.68; Thomas Marshall v Guinle [1979] Ch 227, per Sir Robert Megarry V-C, at 248. 150 Observer Ltd (n 148), per Lord Goff, at 282; Saltman (n 148), per Lord Greene MR, at 214; Coco v Clark (n 148), per Megarry J, at 47; Force India (n 148) 151 Rushbrooke and Speker (n 147) para 4.71; Lansing Linde v Kerr [1991] 1 WLR 251, per Staughton LJ, at 260 (“[T]he owner must limit the dissemination of it or at least not encourage or permit widespread publication.”). 152 Rushbrooke and Speker (n 147) para 4.80. 153 Rushbrooke and Speker (n 147) paras 4.81 to 4.84; Coco v Clark (n 148), per Megarry J, at 48; Observer Ltd (n 148), per Lord Goff,at 284; Church of Scientology v Kaufman [1973] RPC 635, per Goff J, at 658. 154 Rushbrooke and Speker (n 147) paras 4.85 to 4.89; Theakston v MGN Ltd [2002] EMLR 398, per Ouseley, at [75]. 155 Bank Mellat v HM Treasury (No 2) [2013] UKSC 39, per Lord Sumption, at [20].

ARTICLES

between individual rights and public interests. A potential problem is that proportionality analyses are usually restricted to the area of human rights violations. As Lord Steyn held, the Wednesbury test,156 , but not the proportionality analysis, applies outside the human rights sphere, thought the comments were made within the public law context.157 However, this author suggests there has been an ossification of a common law proportionality analysis outside of human rights law, where the proportionality analysis is increasingly accepted and applied by the English courts.158 There is no conceptual difficulty in legislating for a statutory proportionality test for justifiable breaches of banking confidentiality obligations in private law.

penalties. In Singapore, individuals who breach the statutory duty may be fined up to S$125,000 and/or imprisoned for up to 3 years.160 Banks may also be subject to a fine up to S$250,000.161 Similarly, under Swiss law, breach of bank confidentiality is a crime, which attracts penalties of imprisonment for up to 3 years or a fine up to CHF250,000.162 The criminalisation of breaches of bank confidentiality serves a deterrent purpose.

Further, the proportionality analysis proposed is not a vague one. Instead, this proportionality analysis is restricted to any disclosure of any “information was obtained arose out of the banking relations of the bank and its customers”159 must be made in a form which is necessary for pursuit of the legitimate aim. The means of disclosure must also be one which breaches the duty of confidentiality in the most limited way. This ensures that banks are bound only to disclose what is strictly necessary for the purposes stated in the statutory scheme.

However, in this author’s view, the UK should not introduce similar criminal sanctions. Banks are currently subject to a broad panoply of criminal and regulatory sanctions. These also include various mandatory disclosure obligations which run the gamut of reasonable belief and actual knowledge of underlying crimes being committed. If criminal sanctions are further imposed for breaches of bank confidentiality, banks and their employees will be put between a rock and a hard place. This is especially so when the information lies at the penumbrae of suspicions that a crime may have been committed. Since money laundering attempts are never always visible at the outset, threats of criminal sanctions will undesirably lead to delayed disclosures. Thus, the criminalisation of breaches of bank confidentiality risks usurping the objectives of both the mandatory disclosure and statutory regimes.

Case Against Criminal Liability

V. CONCLUSION

In some jurisdictions, a breach of the statutory duty of confidentiality attracts criminal

The law on implied terms in contracts has developed significantly since the Tournier decision. Terms implied in law must address important public interest considerations rather than be merely necessary for contractual efficacy. The common law, as it stands, does not satisfactorily protect the bank’s duty of confidentiality from statutory incursions. Moreover, the scope and application of this duty and its qualifications, as set out in Tournier,

156 Associated Provincial Picture Houses Ltd. v Wednesbury Corporation (“Wednesbury”) [1948] 1 KB 223; See also, Council of Civil Service Unions v Minister for the Civil Service (“GCHQ”) [1983] UKHL 6; [1984] 3 All ER 935, per Lord Diplock, at [410] (HL) (“So outrageous in its defiance of logic or accepted moral standards that no sensible person who had applied his mind to the question to be decided could have arrived at it.”). 157 R (Daly) v SoS for Justice [2001] UKHL 26, per Lord Steyn, at [26]. 158 See Kennedy v Charity Commission [2014] UKSC 20, per Lord Mance, at [51]; Pham v SSHD [2015] UKSC 19, per Lord Mance, at [98]; R (UNISON) v Lord Chancellor [2017] UKSC 51, per Lord Reed, at [89]. 159 Tournier (n 3), per Atkin LJ, at 485; Cf, Ibid., per Scrutton LJ, at 481, dissenting on this point.

160 BA, s 47(5). 161 ibid. 162 Swiss Federal Banking Law (“BkL”), 8 November 1934, as amended, RS 954.1, Art 47.

ARTICLES

are vague and at times difficult to apply. Apart from reformulating the implied term to take this into account, the duty should be codified in statute. This proposed statutory duty should allow disclosures only where necessary and proportionate to the pursuit of defined legitimate aims. This approach makes rules on permitted disclosure clearer and engenders an approach to bank confidentiality, which is condign to the objective of maintaining the integrity of and customersâ&#x20AC;&#x2122; trust in the banking system.

ARTICLES

Divergent Paths: The Doctrine on Equivalents in Singapore and the United Kingdom Joel Sherard Chow, King’s College London Introduction To determine whether a patent has been infringed, the scope of the protection accorded to it must firstly be ascertained based its claims. Courts around the world have adopted varying approaches between two opposing extremes – a strictly literal approach and the use of claims as guidelines. These have important implications on legal certainty, protection for patentees and encouraging innovation. The apex courts of Singapore and the United Kingdom recently considered this question and adopted diverging approaches. The Singapore Court of Appeal in Lee Tat Cheng v Maka GPS Technologies Pte Ltd1 (“Lee Tat Cheng”) favoured a purposive approach while the UK Supreme Court in Actavis UK Ltd v Eli Lilly2 (“Actavis”) adopted the doctrine of equivalents. adopted the doctrine of equivalents. This article seeks to examine the different approaches adopted by the UK and Singapore and evaluate their effectiveness in attempting to preserve the delicate balance between ensuring adequate patentee protection and certainty for third parties in constructing claims. This article will only consider issues of construction and not those concerning prosecution history or validity. The Evolution of Claim Construction preActavis in the UK The scope of protection accorded to a patent is governed by section 125 of the Patents Act 1977 (“UK Patents Act”), which is modelled after Article 69 of the European Patent Convention

1 2

[2018] 1 SLR 856; [2018] SGCA 18. [2017] UKSC 48.

(“EPC”).3 Prior to the 1977 Act, the common law approach to claim construction was traditionally narrow. The House of Lords in EMI v Lissen held that the claims marked the outer limits of the scope of protection:4 The function of the claims is to define clearly and with precision the monopoly claimed, so that others may know the exact boundaries of the area within which they will be trespassers. Their primary object is to limit and not to extend the monopoly. What is not claimed is disclaimed. (emphasis own) However, in 1981, the House of Lords in Catnic v. Hill and Smith5, applying the 1977 Act, held that claims should be construed ‘purposively’. The claimant’s claim in Catnic was for a lintel with a ‘vertical’ backplate (i.e. 90 degrees). The defendant’s lintel had a backplate at an angle of 84 degrees, resulting in a marginal 0.6% reduction in load-bearing capacity. While this fell outside the scope of a literal interpretation, the defendant’s lintel was found to be infringing under a purposive construction of the claimant’s lintel. Lord Diplock laid out the test as follows:6 whether a person skilled in the art would have understood that strict compliance with the claim was intended by patentee to be an essential requirement, such that any variant would fall beyond the scope even though it could have no material effect on the way the invention worked. 3 Section 125(1), UK Patents Act (1977); Convention on the Grant of European Patents (5 October 1973) (“EPC”). 4 (1939) 56 RPC 23, 39. 5 [1982] RPC 183. 6 [1982] RPC 183, 243.

ARTICLES

As such, construing the patent so narrowly to exclude lintels not at 90 degrees (without making any material difference to its function) would “would render [the] monopoly for practical purposes worthless, since any imitator could avoid it and take all the benefit of the invention by the simple expedient of positioning the back plate a degree or two from the exact vertical”.7 The purposive approach was further clarified in Improver v. Remington.8 The case concerned a patent for an epilator which removed hairs by trapping them in a “coiled helical spring”. The defendant’s device functioned identically except that the spring was replaced by a slotted rubber rod. The device thus fell outside of the literal scope of the claim, however, Hoffmann J (as he then was) laid out three questions to determine whether the variant would fall under a purposive interpretation of the claim:9

might have been difficult to overcome.10 As such, the slotted rubber rod did not infringe the claimant’s patent. The first two Improver questions are questions of fact, while the third is a question of construction.11 However, the Improver questions proved difficult in their application where construction of the claims was required to answer the three questions. This was the case in Kirin-Amgen Inc v Hoechst Marion Roussel Ltd,12 where the claimant’s patent related to the production of the hormone erythropoietin (“EPO”). The claims within the patent related to production through an “exogenous” DNA sequence. The issue before the court was whether the defendant’s production of EPO through an “endogenous” DNA sequence fell within the scope of the claims.

2. Would this (ie that the variant had no material effect) have been obvious at the date of publication of the patent to a reader skilled in the art? If no, the variant is outside the claim. If yes?

The challenge arose as, prior to applying the Improver questions, the Court had to construe the claims – determining whether they related to the discovery of EPO or the method of creating the hormone. Further, the application of the second question proved difficult in situations where the technology involved was unknown at the priority date, as the person skilled in the art would not know how it worked at all, making a comparison inappropriate.

3. Would the reader skilled in the art nevertheless have understood from the language of the claim that the patentee intended that strict compliance with the primary meaning was an essential requirement of the invention? If yes, the variant is outside the claim.

Given the difficulties associated with the Improver approach, Lord Hoffmann, in applying the Improver questions, distilled the process down to a single question – “what a person skilled in the art would have understood the patentee to have used the language of the claim to mean”.13

Hoffmann J found that the defendant’s device did not have a material effect on the way the invention worked, and that this would be obvious to the person skilled in the art. However, on the third questions, he found that it would be obvious to a person skilled in the art that the “rubber had problems…which

The House of Lords also considered the amended Protocol on the Interpretation to Article 69 of the EPC in 2000. Article 2 of the Protocol required the Court to take “due account” of equivalents.14 . Lord Hoffmann

1. Does the variant have a material effect upon the way the invention works? If yes, the variant is outside the claim. If no?

7 8 9

[1982] RPC 183, 244. [1990] FSR 181. [1990] FSR 181, 189; Lee Tat Cheng, [30].

10 11 12 13 14 69, EPC.

[1990] FSR 197. Lee Tat Cheng, [31]. [2005] RPC 9 [2005] RPC 9, [69]. Article 2, Protocol on the Interpretation of Article

ARTICLES

found that the Protocol “did not expressly lay down any principle for the construction of claims” but rather supported a purposive interpretation.15 The Court also considered the question of equivalents from American jurisprudence. He expressed concern that “once the monopoly had been allowed to escape from the terms of the claims, it is not easy to know where its limits should be drawn”.16 Further, he concluded that Article 69 of EPC 2000 “firmly shuts the door on any doctrine which extends protection outside the claims”.17 Actavis: The Present Position in the UK Twelve years following Kirin-Amgen, the Supreme Court in Actavis considered the case law on claim construction and adopted the doctrine of equivalents. In his leading judgment, Lord Neuberger placed Art 69 of EPC 2000 front and centre. He noted that the Protocol on the Interpretation of Art 69 bore all the hallmarks of a compromise agreement and that it was “not at all clear how far a court is permitted to move away from the literal meaning”.18 However, he considered the approaches of other EPC States and found that “it is sensible for national courts at least to learn from each other and to seek to move towards, rather than away from, each other’s approaches”19. Lord Neuberger’s approach required the consideration of two issues: (i) does the variant infringe any of the claims as a matter of normal interpretation, and; (ii) does the variant nonetheless infringe because it varies from the invention in a way or ways which is or are immaterial? If the answer is yes to either, there is an infringement.20

15 [2005] RPC 9, [47]-[49]. 16 [2005] RPC 9, [39]. 17 [2005] RPC 9, [42]-[44]. 18 [2005] RPC 9, [31],[33]. 19 Schütz (UK) Ltd v Werit (UK) Ltd [2013] Bus LR 565; [2013] RPC 16, [40]. 20 Actavis, [54].

Lord Neuberger’s second question limits the extent of equivalents only to “immaterial variants” and thus quells concerns about broadening the scope of the monopoly accorded to patentees. He also found that Lord Hoffman’s Improver questions (as subsequently considered in Kirin-Amgen) conflated the two questions of direct infringement and immaterial variants and should be dealt with separately. The first question would be determined based on normal rules of interpretation, as laid out in Wood v Capita Insurance Services Ltd.21 Lord Neuberger further found that the Improver questions were in need of reformulation as they proved unsatisfactory in determining what constituted an immaterial variant. He found, citing Sir Hugh Laddie, that the central inquiry (in accordance with the Protocol on the Interpretation of Art 69 EPC 2000) was not interpretation, but determining the scope of protection accorded to a claim.22 A key change was made to the second Improver question which Lord Neuberger considered to impose “too high a burden on the patentee”.23 This is because they would have to prove that it would have been obvious to the person skilled in the art (the addressee) that the variant had no material effect on the way the invention worked. Often, the addressee would find difficulty in determining what the patentee intended without knowing how the patent works – as was the case in both Kirin-Amgen and Actavis. As such, the second question would have to be asked assuming that the addressee knows how the variant works, taking into account developments from the priority date.24 Similarly, Lord Neuberger clarified the third question and found that it had to be “properly applied”. Specifically, he found that it was important to consider whether the addressee 21 [2017] 2 WLR 1095 22 Laddie, H. ‘Kirin-Amgen - The End of Equivalents in England? ’ (2009) 40 IIC 3, [68]. 23 Actavis, [61]. 24 Actavis, [63].

ARTICLES

would find that the component was an essential part of the invention (as opposed to the larger product or process of which the inventive concept is part). Thus, the reformulated Protocol questions are thus: 1. Notwithstanding that it is not within the literal meaning of the relevant claim(s) of the patent, does the variant achieve substantially the same result in substantially the same way as the invention, i.e. the inventive concept revealed by the patent? 2. Would it be obvious to the person skilled in the art, reading the patent at the priority date, but knowing that the variant achieves substantially the same result as the invention, that it does so in substantially the same way as the invention? 3. Would such a reader of the patent have concluded that the patentee nonetheless intended that strict compliance with the literal meaning of the relevant claim(s) of the patent was an essential requirement of the invention? The Development of Claim Construction in Singapore Construction of patents in Singapore is governed by section 113 of the Patents Act, which is in pari materia to section 125(1) the UK Patents Act.25 It is thus no surprise that the Singapore Court of Appeal has endorsed the purposive approach in various judgments, affirming the Improver/Catnic line of cases.26

disregard the “clear and unambiguous words of the claim”.27 However, the Court of Appeal firmly closed the door on the doctrine of equivalents in Lee Tat Cheng, and laid out several key reasons for rejecting the Actavis approach. Primarily, the Court found that as a matter of statutory interpretation, the Protocol on the Interpretation of Art 69 and the EPC did not apply. The statutory regime made it clear that patent protection would be determined by the claims. Any departure from this approach could thus only be made through Parliament. Further, the Court found that a purposive approach favoured certainty for third parties, given that the assessment would be made at the priority date, rather than at the date of infringement. The Court found that this approach was needed to “provide a reasonable degree of certainty to third parties” who rely on patent claims as delimiting the scope of patent protection.28 The Court also found that the doctrine of equivalents would give rise to undue certainty given the limits of the claims are unclear and the assessment would be made ex post facto. 29 Analysis of the Actavis and Lee Tat Cheng approaches A question of statutory interpretation?

While the Court never explicitly rejected the doctrine of equivalents, it did so implicitly in Bean Innovations v Flexon, where the Court declined to

It is clear that the statutory regimes in the UK and Singapore formed the basis of the differing approaches. In Actavis, Lord Neuberger was bound to give effect to Art 69 EPC 2000 and considered European jurisprudence heavily in arguing for a more unified approach. Similarly, the Court of Appeal Lee Tat Cheng held that the text of section 113, read with section 25(5) (a) of the Patents Act differed substantially from Art 69 EPC 2000 and its Protocol on Interpretation. As such, the Court of Appeal

25 Cap 221, 2005 Rev Ed (Sing). 26 FE Global Electronics Pte Ltd v Trek Technology (Singapore) Pte Ltd [2006] 1 SLR(R) 874, [14]; First Currency Choice v MainLine Corporate Holdings Ltd [2008] 1 SLR(R) 335, [25]; Bean Innovations Pte Ltd v Flexon (Pte) Ltd [2001] 2 SLR(R) 116, [19].

27 Bean Innovations Pte Ltd v Flexon (Pte) Ltd [2001] 2 SLR(R) 116; B. Ching, “Retaining the Catnic/Improver Approach in Patent Law: Why Singapore Should Not Adopt the Doctrine of Equivalents” (2018) 30 SAcLJ 871, 884. 28 Lee Tat Cheng, [52]. 29 Lee Tat Cheng, [53].

ARTICLES

rightly held in Lee Tat Cheng that the question of whether equivalents should apply would be one for Parliament to consider. The differing approaches thus appear to be consistent with their respective statutory regimes, which have diverged with the adoption of the Protocol on the Interpretation of Art 69 EPC 2000. Extent of the notional addressee’s knowledge in construing the claims One key development in Actavis was Lord Neuberger’s departure from construing claims by the notional addressee at the priority date. This was in response to challenges in particular fact scenarios (such as in Actavis and KirinAmgen) where an understanding of how an invention worked was required before the claims could be construed. As noted by Floyd LJ in the English Court of Appeal judgment in Actavis, “predicting in advance whether any particular counter-ion would work was not possible, and therefore that the second Improver test could not be answered yes”.30 The Court of Appeal in Lee Tat Cheng found that an ex post facto analysis involving the state of developing scientific knowledge at the date of infringement would have a “material impact on the protection afforded to the patentee” and thus would give rise to undue uncertainty. However, it is the author’s view that such concerns are overstated given the importance of according fair protection in new and developing fields. Lord Neuberger addressed this concern in Actavis, where he noted with approval the extrajudicial writings of Judge Kalden, who argued that: Variants that are not foreseeable at the priority date may well, due to later developments, become an obvious variant at a later date. This may happen in case of a pioneer invention, where

Actavis, [61].

at the priority date the full breadth of the possible applications could or has not been fully recognised and therefore was not sufficiently taken into account when drafting a claim. The author argues that this approach would encourage innovation, research and development, particularly in developing areas of pharmaceutical products and technology and would not conflict with the statutory regime. One of Singapore’s key goals is to be a regional IP hub and as noted by Senior Minister of State Indranee Rajah in Parliament, speaking on amendments to the Patents Act:31 Enterprises need to be able to effectively protect and commercialise their intellectual property (IP) to preserve their competitive edge. The Committee on the Future Economy (CFE) recognised this when it stated that a strong IP ecosystem is needed to support innovation and technology adoption. As such, it is the author’s view that even when considering a purposive approach, Courts should consider the developing scientific knowledge at the date of infringement, rather than considering the claims solely at the priority date. This is unlikely to affect the vast majority of claims and does not significantly broaden the scope of protection beyond the claims themselves. Imbuing the notional addressee with such knowledge does not ipso facto mean that the second Protocol question would be answered in the affirmative – the key question would be one of foreseeability. As noted by Lord Neuberger:32 while the notional addressee may answer the reformulated second question affirmatively even where the variant was unforeseeable at the priority date, he is 31 Parliamentary Debates Singapore: Official Report vol 94 (28 February 2017) (Indranee Rajah) 32 Actavis, [63].

ARTICLES

less likely to do so than in relation to a variant which was unforeseeable as at that date. It is evident that the Court intended this to be a narrow category and, as such, is unlikely to significantly impact the question of certainty for third-parties. Conclusion The decisions in Actavis and Lee Tat Cheng demonstrate two different approaches that are in conformity with their respective statutory regimes. However, the Actavis approach to â&#x20AC;&#x2DC;interpretationâ&#x20AC;&#x2122; does include welcome improvements, most notably in the form of the revised second Protocol question. This approach would not substantially depart from the purposive approach espoused by the Court of Appeal, but it remains to be seen whether construing the claims by a notional addressee with the knowledge at the date of infringement would be a step too far for the Singapore courts.

ARTICLES

Have the Rules Governing the Interpretation of Pari Passu Clauses in Sovereign Bonds Developed in Equal Step Under English and New York Law? Filbert Lam, University of Edinburgh

I. Introduction International or international bonds are issued by states on capital markets to raise funds. States prefer to issue bonds vis-à-vis obtain syndicated loans1 as bonds are more flexible, liquid and easier to assign ratings to and price.2 Consequently, sovereign bonds attract a panoply of investors from retail investors seeking lowrisk and secure investments to vulture funds trading in distressed debt.3 The pari passu clause is a ubiquitous and controversial clause often found in sovereign bonds.4 As states cannot go bankrupt, bondholders replicate protections conferred by domestic statutory insolvency regimes through contractual mechanisms.5 Until the 1990s,6 the pari passu clause provided for equal ranking of holders’ unsecured debt with other external debt of the same nature (“equal ranking simpliciter” interpretation).7 This stemmed from the “equal and fair” treatment assurances which governments provided to foreign creditors.8 However, since the early 1990s, some clauses contained an additional modifier which prima facie contained an equal ranking in the creditors’ right of payment (“equal payments” interpretation).9

1 Yvonne Wong, Sovereign Finance and the Poverty of Nations: Odious Debt in International Law (Edward Elgar 2012) 28. 2 Wong (n 1) 29; See also, Jill E Fisch and Caroline M Gentile, ‘Vultures or Vanguards?: The Role of Litigation in Sovereign Debt Restructuring’ (2004) 53 Emory Law Journal 1043, 1069. 3 Wong (n 1) 28. 4 Benjamin Chabot and G Mitu Gulati, ‘Santa Anna and his Black Eagle: the origins of pari passu?’ (2014) 9(3) CMLJ 216, 216. 5 Chabot and Gulati (n 4), 217; Lee C Buchheit and Jeremiah S Pam, ‘The Pari Passu Clause in Sovereign Debt Instruments’, (2004) 53 Emory LJ 869, 873. 6 Chabot and Gulati (n 4), 222 (The first sovereign bond was issued by Mexico in 1843). 7 Chabot and Gulati (n 4), 217. 8 Chabot and Gulati (n 4), 228. 9 ibid.

This multiplicity of interpretations has introduced uncertainty in the sovereign bond markets and increased transactional costs and default risks for states. The US Second Circuit Court of Appeals held, in NML v Argentina (NML),10 if a pari passu clause contained both the equal ranking and payments elements, they conferred separate but related rights on the bondholders.11 Since NML, pari passu clauses have been the subject of a Herculean body of academic literature.12 This paper considers how courts in the main sovereign debt restructuring jurisdictions – English and New York (NY) – will interpret pari passu clauses.13 In any event, there are solutions available to sovereign issuers to circumnavigate the legal effects of NML. This author seeks to advance these propositions in four stages. Part II discusses pari passu clauses generally. This explores the interests which creditors seek to protect and the pari passu clauses themselves. Part III analyses the NML judgment and the criticisms of the decision. This paper will investigate the background of pari passu clauses, analyse the NML decision and study subsequent decisions in New York 10 NML Capital v Argentina (NML), 699 F3d 246 (2d Cir 2012). 11 Rodrigo Olivares-Caminal, ‘An Introduction to Sovereign Debt Restructuring’, in Rodrigo Olivares-Caminal et al, Debt Restructuring (2nd ed, OUP 2016) para 10.13. 12 See, for instance, Joseph Cotterill, ‘Pari Passu and the Litigators of the Lost Clause’ (2013) 9 CMLJ 18; Anna Gelpern, ‘Contract Hope and Sovereign Redemption’ (2013) 8 CMLJ 132; W Mark C Weidemaier, ‘Sovereign Debt After NML v. Argentina’ (2013) 8 CMLJ 123; Natalie A Turchi, ‘Restructuring a Sovereign Bond Pari Passu Work-around: Can Holdout Creditors Ever Have Equal Treatment?’, (2015) 83(4) Fordham Law Review 2171; Lee C Buchheit and G Mitu Gulati, ‘Restructuring sovereign debt after NML v Argentina’ (2017) 12(2) CMLJ 224. 13 Rodrigo Olivares-Caminal, ‘Litigation Aspects of Sovereign Debt’, in Rodrigo Olivares-Caminal et al, Debt Restructuring (2nd ed, OUP 2016) para 11.05.

ARTICLES

which considered the NML case. Part IV considers how English courts would have decided NML. This section considers the rules governing the interpretation of boilerplate clauses under English law and whether English courts would have arrived at the same decision as the Second Circuit did in NML. Finally, Part V critically considers three solutions advanced by the commentators – contractual solutions, international arbitration and sovereign insolvency regimes. II. Pari passu clauses Pari passu clauses are one of the main boilerplate clauses through which sovereign bondholders protect their interests.14 This section seeks to highlight the primary interests of creditors and how pari passu clauses protect them. 1. Creditors’ Interests

maturities, lower par values and lower interest rates (“haircuts”).17 More creditors now hold a wider variety of debt instruments and involving more legal jurisdictions.18 Governments and bondholders prefer to pursue negotiated solutions. States do not benefit from the rules protecting them from an individual creditor’s legal actions under domestic insolvency rules.19 States also hold very little property abroad,20 which are protected by state immunity against creditor attachment.21 However, holdout creditors may pursue litigation to publicly pressure governments to pay them despite the difficulties of legal enforcement.22 States opt to pay because failure to do so might trigger creditor flight in highly liquid sovereign bond markets, denial of trade credits, freeze on International Monetary Fund (IMF) assistance and denial of access to primary capital markets in London and New York.23

Sovereign debt restructuring has become increasingly complex with rapid economic growth and increasing interconnectedness within the global economy. With the rise of “too-big-to-fail” financial institutions, there is evidence that the banking crises invariably lead to sovereign debt crises and vice versa. This is the so-called “doom loop”,15 where states which are increasingly dependent on debt for liquidity expose themselves to the risks of spiralling debt crises.16

The paths for holdout creditors have been smoothed by several contractual provisions. For instance, states may include express waivers of their immunity from suits and of their assets from execution.24 Claims may be founded on pari passu, negative pledge or cross-default clauses.25 States include these terms to entice investors to invest or exchange their bonds.26 This led to the prominence of the pari passu clause in holdout litigation.

States restructure their debt when it becomes unsustainable. They exchange old debt instruments for new ones with longer

17 Olivares-Caminal (n 14) 166. 18 Olivares-Caminal (n 14) 166. 19 Lee C Buchheit and Elena L Daly, ‘Minimizing Holdout Creditor: Carrots’, in Rosa M Lastra and Lee C Buchheit (eds), Sovereign Debt Management OUP 2014) para 1.03. 20 Buchheit and Daly (n 19) para 1.04. 21 Buchheit and Daly (n 19) para 1.04. 22 Buchheit and Daly (n 19) para 1.08; W Mark C Weidemaier and Ryan McCarl, ‘Creditors’ Remedies’, in Rosa M Lastra and Lee C Buchheit (eds), Sovereign Debt Management OUP 2014) para 11.02. 23 Marcus H Miller and Dania Thomas, ‘Sovereign Debt Restructuring: The Judge, the Vultures, and Creditor Rights’, in Robert W Kolb (eds), Sovereign Debt: From Safety to Default (John Wiley & Sons, Inc 2011) 213-215. 24 Weidemaier and McCarl (n 22) para 11.02; W Mark C Weidemaier, ‘Disputing Boilerplate’ (2009) 82(1) Temple Law Review 1. 25 Olivares-Caminal (n 14) 166. 26 Olivares-Caminal (n 14) 169.

14 Rodrigo Olivares-Caminal, ‘The Definition of Indebtedness and the Consequent Imperilling of the Pari Passu, Negative Pledge and Cross-Default Clauses in Sovereign Debt Instruments’ (2017) 12(2) Capital Markets Law Journal 164, 165-6. 15 Emmanuel Farhi and Jean Tirole, ‘Deadly Embrace: Sovereign and Financial Balance Sheets Doom Loops’ (2016) NBER Working Paper No 21843. 16 Irina Balteanu and Aitor Erce, ‘Linking Bank Crises and Sovereign Defaults: Evidence from Emerging Markets’ (2017) European Stability Mechanism Working Paper Series No 22; Russell Cooper and Kalin Nikolov, ‘Government debt and banking fragility: the spreading of strategic uncertainty’ (2018) ECB Working Paper Series No 2195.

ARTICLES

2. Pari Passu Clauses Pari passu means “in equal step”.27 In the insolvency context, this rule requires “all unsecured creditors [to] be paid … in equal proportions between themselves”.28 This was derived from the equitable maxim – aequitas est aequalitas.29 In the context of bonds, when debtors become insolvent, they must respect the “equal legal ranking of unsecured claims”.30 The clause appears in different formulations in bonds – corporate and sovereign. Some read: The Notes rank and will rank, pari passu in right of payment with all other present and future unsecured and unsubordinated External Indebtedness of the Issuer.31 In contrast, other clauses may state: The borrower’s obligations under the loan agreement will rank pari passu with all its other unsecured liabilities.32 Despite their different iterations,33 pari passu clauses incorporate both a representation and warranty element and a covenant element.34 They affirm both the equal ranking of all unsecured debtors and the promises this position is

27 Buchheit and Pam (n 5) 871; Olivares-Caminal (n 13) para 11.44. 28 Stuart Bridge, ‘Payment of Debts’, in John McGhee (eds), Snell’s Equity (33rd ed, Sweet & Maxwell 2015) para 32005 (parentheses omitted, emphasis added in italics). 29 Re Lehman Brothers International (Europe) (In Administration) [2010] EWCA Civ 917, [76] (per Arden LJ); TC01627: Martin Hedley Rogers [2011] UKFTT 791 (TC), [117]. 30 Philip R Wood, The Law and Practice of International Finance: International Loans, Bonds, Guarantees, Legal Opinions, Vol 2 (2nd ed, Sweet & Maxwell 2007) para 5-026. 31 Buchheit and Pam (n 5) 871 (emphasis added in underline). 32 Wood (n 30) para 5-026 (emphasis added in underline). 33 See, on the effect of different formulations, Stephen J Choi et al, ‘The Black Hole Problem in Commercial Boilerplate’ (2017) 67 Duke LJ 1. 34 Wood (n 30) para 5-026; Lee C Buchheit, How to Negotiate Eurocurrency Loan Agreements (2nd ed, International Financial Law Review 2004) 82.

maintained in perpetuity, respectively.35 Pari passu clauses ensure issuers have and will not allow the subordination of the bondholders’ claims to other unsecured debt.36 Both de jure and de facto subordination of unsecured debts are prohibited.37 They assure bondholders that there is no preferential treatment of specific classes of unsecured creditors to the formers’ detriment,38 unless otherwise required by law,39 and protects them against the “involuntary subordination” of their claims.40 Other clauses may even state that the debt “ranks, and will rank, pari passu … and will be paid as such”, which prohibits the debtor from paying bondholders preferentially.41 Pari passu clauses are not all homogenously drafted or interpreted.42 Consequently, bondholders must seek to carefully negotiate and accurately define the legal scope and effect of the clause in the loan agreement.43 At a minimum, pari passu clauses demonstrate that parties’ intended for newly-issued bonds will rank equally with other unsubordinated unsecured debt.44 However, boilerplate clauses often may contain additional obligations.45 Thus, this paper now turns to consider if pari passu clauses in sovereign bonds may contain obligations additional to equal ranking obligation.

35 Buchheit (n 34) 82; Olivares-Caminal (n 11) para 10.13, fn 19; Georges Affaki, ‘Revisiting the Pari Passu Clause’, in Rosa M Lastra and Lee C Buchheit (eds), Sovereign Debt Management OUP 2014) para 4.08. 36 Buchheit (n 34) 83. 37 Buchheit (n 34) 83; Wood (n 30) para 5-029. 38 Buchheit (n 34) 84. 39 Wood (n 30) para 5-029. 40 Buchheit (n 34) 41 Buchheit (n 34) 84; Olivares-Caminal (n 13) para 11.48. 42 See, for example, Buchheit (n 34) 84 (“Accuracy presents something of a challenge in the absence of any uniform view as to what the pari passu undertaking is intended to cover”, emphasis added); Cf Wood (n 30) para 5-027 (“The clause has a settled meaning in corporate credits. It is considered that it does not have a different meaning in sovereign credits.”, emphasis added). 43 Buchheit (n 34) 84. 44 Buchheit (n 34) 84. 45 Choi et al (n 33) 8-15.

ARTICLES

III. NML and New York (‘NY’) Law

2. Interpretation of Boilerplate Clauses under NY Law

1. Background In NML, the disputed clause (the “Clause”) in the Fiscal Agency Agreement (FAA Bonds) reads: t]he Securities will constitute … direct, unconditional, unsecured and unsubordinated obligations of the Republic and shall at all times rank pari passu without any preference among themselves. [“first sentence”] The payment obligations of the Republic under the Securities shall at all times rank at least equally with all its other present and future unsecured and unsubordinated External Indebtedness … [“second sentence”]46 When Argentina defaulted, the President imposed a temporary moratorium on the payment of both the principal and interest payments on its public external debt. The legislation was passed to effectuate this.47 This caused the Plaintiffs approximately $1.33 billion in unpaid debt.48

Under NY law, a boilerplate clause must be interpreted based on the market’s understanding of what the term means.53 In Sharon Steel Corp v Chase Manhattan Bank, NA (“Sharon Steel”),54 the US Second Circuit Court of Appeals held boilerplate clauses are not the consequence of the relationship of particular borrowers and lenders and do not depend upon [particularised] intentions of the parties to [a contract].55 They must be interpreted in a “general, uniform and unwavering” way according to the “custom and usage” of the trade.56 This rule of interpretation seeks to promote market efficiency.57 It is thought that inconsistent interpretations of boilerplate clauses increase the costs of borrowing without countervailing financial benefits and do not further the “administration of justice”.58 This background informed the approach taken by the Second Circuit in the NML decision. 3. NML decision

Argentina then offered new bonds (“Exchange Bonds”), which were worth a quarter of the defaulted bonds in 2005 and 2010, to allow FAA bondholders.49 Further, Argentina maintained a policy of ensuring hold-out creditors “remain in default indefinitely”.50 Finally, the Argentine government passed a statute called the “Lock Law” which prohibited settlements with holdout creditors.51 Thus, Argentina had de jure and de facto subordinated the FAA bonds to the Exchange Bonds.52 46 NML (n 10) 251 (emphasis original in italics and added in bold and underline). 47 NML (n 10) 251. 48 NML (n 10) 251. 49 NML (n 10), 252; For a detailed analysis of the exchange offer, see Rodrigo Olivares-Caminal, ‘Sovereign Bonds: A Critical Analysis of Argentina’s Debt Exchange Offer’ (2008) 10(1) Journal of Banking Regulation 28. 50 NML (n 10), 252, citing the Prospectus. 51 NML (n 10), 252. 52 NML (n 10), 260.

Bonds are contracts.59 The Court was bound by the rules of contractual interpretation to determine the meaning of the Clause.60 . Argentina’s argued that the Clause was 53 Turchi (n 12), 2190 ff; William W Bratton, ‘Pari Passu and a Distressed Sovereign’s Rational Choices’, (2004) 53 Emory LJ 823, 863-64. 54 Sharon Steel Corp v Chase Manhattan Bank, NA (“Sharon Steel”) 691 F 2d 1039 (2d Cir 1982). 55 Sharon Steel (n 54), 1048 (emphasis added). 56 NML (n 10), 258, Citing with approval Law Debenture Trust Co of NY v Maverick Tube Corp, 595 F 3d 458, 466 (2d Cir 2010). 57 Sharon Steel (n 54), 1048; See also, the decision of the Fifth Circuit, Broad v Rockwell International Corp, 642 F 2d 929, 943 (5th Cir), cert. denied, 454 US 965, 102 SCt 506, 70 L Ed 2d 380 (1981). 58 Sharon Steel (n 54), 1048. 59 NML (n 10), 257; Citing with approval, Arch Ins Co v Precision Stone, Inc, 584 F 3d 33, 39 n 4 (2d Cir 2009) (“In New York, a bond is a contract …”). 60 NML (n 10), 258; Citing with approval, EM Ltd v Republic of Argentina, 382 F 3d 291, 292 (2d Cir.2004) (“simple

ARTICLES

“universally understood” only meant unsecured and unsubordinated creditors ranked pari passu with external creditors.61 Conversely, the Plaintiffs argued the preferential payments made to Exchange bondholders meant the FAA Bondholders was de facto subordinated to the Exchange bonds without any express provision for subordination.62 Nevertheless, the Second Circuit Court agreed with the Plaintiffs. It was held that there was no evidence to suggest the “equal ranking simpliciter” interpretation constituted any settled market understanding of pari passu clauses,63 which do not have a clear and consistent meaning in the sovereign bonds market.64 Moreover, the Clause encompassed two sets of rights and obligations in separate sentences to which Argentina failed to give its full effect.65 Contracts must be interpreted without “leav[ing] one of its provisions substantially without force or effect”.66 The Court held the first sentence restricts the issuer from “formally subordinating” the FAA Bonds.67 The second sentence restricts Argentina qua debt payor from repaying similarly-situated creditors unequally.68 This interpretation was also adopted by the Brussels Court of Appeal holding that “the various creditors benefit from a pari passu clause that in effect provides that the debt be repaid pro rata among all creditors”.69 Thus, the Clause prohibits both de facto and de jure or formal

question of contract interpretation.”). 61 NML (n 10), 258. 62 NML (n 10), 258. 63 ibid. 64 NML (n 10), 258-60, citing with approval, Lee C Buchheit, ‘The Pari Passu Clause Sub Specie Aeternitatis’ (1991) 10(12) International Financial Law Rev 11, 11; G Mitu Gulati and Kenneth N Klee, ‘Sovereign Piracy’, (2001) 56 Business Law 635, 646; Stephen Choi and G Mitu Gulati, ‘Contract As Statute’, (2006) 104 Michigan Law Review 1129, 1134; Philip R Wood, Project Finance, Subordinated Debt and State Loans (Sweet & Maxwell 1995) 165. 65 NML (n 10), 258-60. 66 Singh v Atakhanian, 31 AD 3d 425, 818 NYS 2d 524, 526 (NY App Div 2d Dep’t 2006). 67 NML (n 10), 259 (emphasis added). 68 NML (n 10), 259. 69 Translation provided by Brief for the United States of America as Amicus Curiae, in NML (n 10), 14; Original: Elliot Associates, LP v Banco de la Nacion, General Docket No

and informal subordination of the FAA Bonds or otherwise similarly-situated, unsubordinated and unsecured external debt. Consequently, Argentina was held to have breached its obligations under the Clause.70 Argentina’s moratoriums against the payment on FAA Bonds and the legislation of the Lock Law constituted informal and formal subordination, respectively. These frustrated the Plaintiff ’s right to “direct, unconditional, unsecured and unsubordinated obligations”.71 However, the Court held that even if Argentina’s equal ranking simpliciter interpretation was adopted, it would still have breached their obligations under the Clause.72 While the Court does not elaborate on this holding, the Court read the two sentences separately. Their accompanying rights and obligations were held to be distinct, though symbiotically connected. This means that even if Argentina had only violated the first sentence, that was sufficient to constitute a breach of the Clause. Accordingly, the Court upheld the injunctive relief ordered by the lower court. Monetary damages and acceleration were considered “ineffective” because Argentina would refuse to pay and leave judgment-debtors without further recourse to reclaim their loans.73 As injunctions do not attach to the state’s property,74 As injunctions do not attach to the state’s property,75 Further, there were no legislative prohibitions on injunctions against states.76 Finally, the Court found no evidential basis for suggesting the injunctions would impose undue burdens on the state by causing another financial crisis or usurping the restructuring plan.77 Thus, 2000/QR/92 (Court of Appeals of Brussels, 8th Chamber, 26 September 2000) (Unreported) 3. 70 NML (n 10), 260. 71 NML (n 10), 260. 72 NML (n 10), 260. 73 NML (n 10), 262. 74 Stephens v Nat’l Distillers & Chem Corp, 69 F.3d 1226, 1229 (2d Cir.1995). 75 NML (n 10), 262. 76 NML (n 10), 262-3. 77 NML (n 10), 263-4.

ARTICLES

the Second Circuit remanded the case to the District Court which ordered the injunctions to bind specific financial intermediaries. 4. Analysis Since the interpretation of boilerplate clauses requires courts to consider what the market understands them to mean, commentators have conducted historical analyses of the pari passu clause.78 a. Historical Context Sovereign debt restructuring frequently involved the state halting payments on the existing bonds, renegotiates for repayment terms in their favour and then pays those who agree to restructure their debt.79 There was no evidence that highly influential committees representing bondholders’ interests in the UK and the US read similarly-worded pari passu clauses to include an “equal payments” obligation.80 Thus, this restructuring approach was never treated as a violation of the pari passu clause.81 However, this author considers this historical context to be of limited assistance when interpreting pari passu clauses. Boilerplate clauses are terms which are “standard in a certain genre of contracts”.82 Pari passu clauses did not contain implicit or explicit references to payment until the 1990s.83 There are three distinct types of pari passu clauses – (a) equal ranking simpliciter; (b) explicit equal payments; and (c) implicit equal payments clauses.84 The latter two gained prominence in the 1990s.85 This author suggests these are three distinct 78 Chabot and Gulati (n 4), 235; Lee C Buchheit, ‘A Note on Contract Palaeontology’ (2014) 9(3) CMLJ 251. 79 Weidemaier and McCarl (n 22) para 11.14. 80 Weidemaier and McCarl (n 22) para 11.16. 81 Weidemaier and McCarl (n 22) para 11.14. 82 Sharon Steel (n 54), 1048. 83 Weidemaier and McCarl (n 22) para 11.18; See also, W Mark C Weidemaier et al, ‘Origin Myths, Contracts, and the Hunt for Pari Passu’ (2013) 38(1) Law and Society Inq 84. 84 Weidemaier and McCarl (n 22) para 11.19. 85 Weidemaier and McCarl (n 22) para 11.20, Figure 11.1.

species of pari passu clauses, even though they share common characteristics. Thus, there is no legal prohibition on interpreting the Clause as requiring further obligations apart from the equal ranking of unsecured creditors. b. NML Decision Three primary criticisms have been advanced by critics of the NML judgment. First, some argued the Court erred in placing too much emphasis on “payment obligations”. They suggest that the reference to “payment obligations” merely extended the equal ranking protection to other unsecured creditors, whereas the first sentence referred only to other bondholders.86 This author finds this unpersuasive. A distinction was made between bondholders “among themselves” vis-à-vis other future unsecured lenders and also between “the Securities” and “the payment obligations” under them.87 There is simply no basis for suggesting the second sentence merely reinforces the “equal ranking simpliciter” obligation.88 Next, critics and Second Circuit Court itself highlighted that the third-party payment intermediaries might be unfairly treated by being bound by the injunction. This is because of the delays it potentially causes to payments and settlements unrelated to the Exchange Bonds.89 While unfairness is a recognised objection against the granting of an injunction,90 the Court considered Argentina’s arguments91 “speculative” and “hyperbolic” 86 Weidemaier and McCarl (n 22) para 11.13, fn 17; Philip R Wood, ‘Pari Passu Clauses – What Do They Mean?’ (2003) 18(10) Journal of International Banking and Financial Law 371, 373. 87 NML (n 10) 251. 88 Cf Weidemaier and McCarl (n 22) para 11.13, fn 17. 89 NML (n 10) 264; Weidemaier and McCarl (n 22) para 11.22; Buchheit and Pam (n 5) 880; See also, Grain Traders, Inc v Citibank, NA, 160 F3d 97, 102 (2d Cir1998). 90 NML Capital Ltd v Republic of Argentina 727 F3d 230, 246 (2nd Cir 2013); Nemer Jeep–Eagle, Inc v Jeep–Eagle Sales Corp, 992 F2d 430, 436 (2d Cir 1993); Winter v Natural Res Def Council, Inc, 555 US 7, 24, 129 S Ct 365, 172 L Ed 2d 249 (2008) (Courts of equity should pay attention to the public consequences of injunctions). 91 NML (n 10),263-4.

ARTICLES

and, if the consequences materialised, they were “almost entirely of the Republic’s own making”.92 This author agrees with the Second Circuit’s decision because all parties were aware of the default risks at the time they entered into the bonds, which are debt contracts. Thus, the proper mechanisms to account for such risks are the forces and price-forming mechanisms of the free sovereign debt market, not judicial ambivalence on granting injunctive remedies for reasons of financial expediency alone. Finally, other critics have suggested that the NML approach will make payment and settlement systems sluggish and sovereign debt markets illiquid because few investors and payment intermediaries will want to deal with sovereign debt.93 This author agrees that this criticism is “speculative”.94 Similar prohibitions have been legislated against paying and settling transfers in breach of international sanctions without accompanying delays in payments and settlements.95 Thus, in this author’s view, these three criticisms of the NML decision are unpersuasive. c. Subsequent Decisions Efforts have been made to distinguish NML as an unusual case in NY law. Argentina has been described as a “uniquely recalcitrant debtor” by the Second Circuit.96 Commentators argue this is why their actions attracted stern treatment.97 This author disagrees. The Court held, in White Hawthorne, LLC et al v Argentina (White Hawthorne), that Argentina’s recalcitrance was characterised by the “executive declarations and legislative enactments” and the “entire course of conduct” of the state.98 This gave the Court

92 NML (n 90) 246. 93 Weidemaier and McCarl (n 22) para 11.27. 94 NML (n 90) 246. 95 NML (n 90) 246. 96 NML (n 90) 270; White Hawthorne, LLC et al v The Republic of Argentina (White Hawthorne) 16-cv-1042 (TPG) (2016), 5 (Unreported, written decision on file with author). 97 Weidemaier and McCarl (n 22) paras 11.29 and 11.3511.39; Hans Tjio, ‘Restructuring the bond market in Singapore’ (2019) 14(1) Capital Markets Law Journal 16, 31. 98 NML (n 10) 246, cited with approval in White

little confidence that it would pay the holdout creditors monetary damages had the remedy been awarded over injunctions.99 Further, this author argues the Second Circuit’s holding in NML was narrower than its critics suggest. In White Hawthorne, the Court held that mere payment to other creditors, including payment encouraged by this court to settle claims on other defaulted debt, [was “inadequate” to constitute] a violation of the pari passu clause100 On this view, a violation of the “equal payments” obligation occurs only when the state has also breached its “equal ranking simpliciter” obligation through both de facto and de jure subordination. This author suggests that a court must first determine the type of obligation(s) the specific pari passu clause in question embodies. If the clause embodied both the “equal ranking simpliciter” and “equal payments obligation”, a breach of the former through both de facto and de jure subordination are conditions precedent to finding a breach of the latter obligation. This reading is consistent with the reasons the Second Circuit Court considered Argentina to be “uniquely recalcitrant”101 and criticisms that the NML decision was an outlier.102 Thus, this author suggests that the criticisms have been misdirected because commentators have not considered subsequent decisions by the Second Circuit which cited and applied the ratio of NML, which inform how the Second Circuit decided the NML case itself. III. English Law Approaches The English rules governing contractual interpretation fills chapters and books.103 When Hawthorne (n 96) 5. 99 NML (n 10), 262. 100 White Hawthorne (n 96) 6-7 (emphasis added). 101 NML (n 90) 270; White Hawthorne (n 96) 5. 102 Weidemaier and McCarl (n 22) paras 11.35-11.39; Tjio (n 97) 31. 103 See, for instance, Andrew Burrows, A Restatement of

ARTICLES

English courts interpret contracts, they must “find the contextual meaning of the language of the text, ie, what the words would convey to the reasonable person circumstanced as the parties were”.104 Lord Steyn sets out four general principles of interpretation to be considered in his Lordship’s seminal extra-judicial commentary.105 First, a term must not be interpreted in isolation of its “contextual scene”.106 Next, courts must “derive a meaning from [the text’s] nature and contents”.107 Third, they must take consider contractual terms both literally and purposively, depending on the specific contract.108 Finally, when faced with “feasible” alternative interpretations, courts may look at “[b]roader policy considerations” to determine the “best contextual interpretation”.109 These principles remain fundamental to contractual interpretation today. 1. Contractual Interpretation under English Law a. Objective Determination of Parties’ Intentions Courts must identify the intention of the parties,110 reading the contract as a whole, when seeking to interpret the meaning of a prima facie ambiguous term used in a contract.111 In The English Law of Contract (OUP 2016) 84-90; Kim Lewison, The Interpretation of Contracts (6th ed, Sweet & Maxwell 2015); Ewan G McKendrick, ‘Express Terms’ in Hugh G Beale (ed), Chitty on Contracts (32nd ed, Sweet & Maxwell 2015, Vol I). 104 Johan Steyn, ‘The Intractable Problem of The Interpretation of Legal Texts’ (2003) 25(1) Sydney Law Review 5, 6 (emphasis added). 105 Steyn (n 104) 7-8. 106 Steyn (n 104) 7. 107 Steyn (n 104) 7. 108 Steyn (n 104) 7-8. 109 Steyn (n 104) 8. 110 Rainy Sky SA v Kookmin Bank [2011] UKSC 50, [2012] 1 All ER (Comm) 1, [14] (per Lord Clarke of Stone-cumEbony); Pink Floyd Music Ltd v EMI Records Ltd [2010] EWCA Civ 1429, [2011] 1 WLR 770n, [17]. 111 Wood v Capita Insurance Services Ltd [2017] UKSC 24, [2017] AC 1173, [10] (per Lord Hodge); Chamber Colliery Co Ltd v Twyerould [1915] 1 Ch 268 (Note), [1893] 7 WLUK 97, 272 (per Lord Watson); Charter Reinsurance Co v Fagan [1997] AC 313, 384 (per Lord Mustill); Summit Investment Inc v British Steel Corpn (The Sounion) [1987] 1 Lloyd’s Rep 230, 235 (per Lloyd LJ); Similarly, see also, Arbuthott v Fagan (Unreported) 30 July 1993, CA, cited

Investors Compensation Scheme Ltd v West Bromwich Building Society (ICS),112 Lord Hoffmann held, courts must objectively construct and interpret contractual terms, taking into account the commercial context and purpose of the transactions.113 This basic rule has been hitherto consistently adopted by English courts in subsequent decisions.114 b. Commercial Context The commercial context of the whole document must be taken into account in contractual interpretation.115 The UK Supreme Court in Marley v Rawlings held, judges must “identify the intention of the … parties … by interpreting the words used in their documentary, factual and commercial context”.116 This includes, “the genesis with approval in Globe Motors, Inc (a corporation incorporated in Delaware, USA) v TRW Lucas Varity Electric Steering Ltd (Globe Motors v TRW) [2016] EWCA Civ 396, (2016) 168 Con LR 59, [59] (per Beatson LJ); See also, Robert Goff, ‘Commercial Contracts and the Commercial Court’ [1984] LMCLQ 382, 3878; Thomas Bingham, ‘A New New Thing Under the Sun? The Interpretation of Contracts and the ICS Decision’ (2008) 12 Edinburgh LR 374, 376. 112 Investors Compensation Scheme Ltd v West Bromwich Building Society, Investors Compensation Scheme Ltd v Hopkin & Sons (a firm), Alford v West Bromwich Building Society, Armitage v West Bromwich Building Society (ICS) [1998] 1 WLR 896, 912-13 (per Lord Hoffmann). 113 In the Matter of the Nortel Networks UK Pension Plan (Lewis v The Pensions Ombudsman) [2005] EWHC 103 (Ch), [53]; Globe Motors v TRW (n 111) [56] (Beatson LJ); Sirius International Insurance Co (Publ) v FAI General Insurance Ltd [2004] UKHL 54, [2004] 1 WLR 3251, [19] (per Lord Steyn) (“The tendency should therefore generally speaking be against literalism.”); Gerard McMeel, McMeel on the Construction of Contracts: Interpretation, Implication, and Rectification (3rd ed, OUP 2017) para 1.33; Cf NLA Group Ltd v Bowers [1999] 1 Lloyd’s Rep 109, 112 (per Timothy Walker J). 114 Rainy Sky (n 110) [21] (per Lord Clarke of Stonecum-Ebony); Marley v Rawlings [2014] UKSC 2, [2015] AC 129, [19] (per Lord Neuberger of Abbotsbury). 115 Mannai Investment Co Ltd v Eagle Star Life Assurance Co Ltd [1997] AC 749, 771 (per Lord Steyn); See, for an overview, Patrick S Hodge, ‘Can Judges Use Business Common Sense in Interpreting Contracts?’, in Larry A DiMatteo and Martin A Hogg (eds), Comparative Contract Law: British and American Perspectives (OUP 2016) 272, 274-5. 116 Marley v Rawlings (n 114) [20] (per Lord Neuberger of Abbotsbury, emphasis added), citing with approval Kirin-Amgen Inc v Hoechst Marion Roussel Ltd [2004] UKHL 46, [2005] 1 All ER 667, [64] (per Lord Hoffmann) (“No one has ever made an acontextual statement. There is always some context to any utterance, however meagre.”); Arbuthnott v Fagan [1995] CLC 1396, 1400 (per Sir Thomas Bingham MR) (“courts will never

ARTICLES

of the transaction, the background, the context, the market in which the parties are operating”.117 Thus, a proper consideration of the contract’s commercial context is integral to determining the parties’ intentions.

are not public documents and are assignable.123 Allowing extensive consideration of the background context may “potentially prejudice third parties” who were unaware of such background information.124

There is, however, a limit to the scope of the “factual matrix” which courts may consider.118 In the UK Supreme Court’s decision in Rainy Sky, the Lord Clarke of Stone-cum-Ebony held the “factual matrix” ought to be determined according to what a person

Courts should also give “a business sense” to contracts.125 This means judges should not emphasise semantic or syntactical differences in priority over applying “rules easily learned and easily retained”126 according to the “expectations of many business people”.127 Where plausible rival meanings were proffered, courts may adopt the construction which is “more consistent with business common sense”.128 The quality of drafting may affect how courts consider “the language and the implications of the competing constructions”.129 For instance, courts may favour a textual approach if the contract was drafted meticulously with the assistance of professionals.130In contrast, if certain provisions were unclear, then courts may pay more attention to the relevant commercial context of the agreement.131

who has all the background knowledge which would reasonably have been available to the parties in the situation in which they were at the time of the contract, would have understood the parties to have meant.119 Courts can only take into account those facts which all parties knew or reasonably ought to have been aware of,120 insofar as they relate to the impugned document(s).121 Further, the courts should be primarily concerned with the provision’s language and not replace the parties’ intentions for what the court thinks should be agreed.122 Unlike legislation, contracts

construe words in a vacuum”); See also, Prenn v Simmonds [1971] 1 WLR 1381, 1384-6 (per Lord Wilberforce); Bank of Credit and Commerce International SA (in compulsory Liquidation) v Ali (BCCI v Ali) [2001] UKHL 8, [2002] 1 AC 251, [8] (per Lord Bingham of Cornhill). 117 Reardon Smith Line Ltd v Hansen-Tangen [1976] 1 WLR 989, 995-6 (per Lord Wilberforce, emphasis added). 118 Reardon Smith Line Ltd v Hansen-Tangen (n 117) 997 (per Lord Wilberforce). 119 Rainy Sky (n 110) [21] (per Lord Clarke of Stonecum-Ebony, emphasis added in italics and bold). 120 Rainy Sky (n 110) [21] (per Lord Clarke of Stone-cumEbony); ICS (n 112) 912-13 (per Lord Hoffmann); BCCI v Ali (n 116) [39] (per Lord Hoffman, Dissenting Speech); Absalom v TRCU Ltd [2005] EWHC 1090 (Comm), [2005] 2 Lloyd’s Rep 735, [25(ii)] (per Aikens J, affirmed [2005] EWCA Civ 1586, [2006] 2 Lloyd’s Rep 129, [7] (per Longmore LJ); Egan v Static Control Components (Europe) Ltd [2004] EWCA Civ 392, [2004] 2 Lloyd’s Rep 429. [39] (per Arden LJ); Inglis v Buttery & Co (1878) 5 R (HL) 87, 103 (per Lord Blackburn); Bank of Scotland v Stewart (1891) 18 R 957, 960 (per Lord President Inglis); Hodge (n 115) 274. 121 Charrington & Co Ltd v Wooler [1914] AC 71, 80 (per Lord Kinnear) (“prove the relation of the document to the facts”). 122 Arnold v Britton [2015] UKSC 36, [2016] 1 All ER 1, [17] and [20] (per Lord Neuberger of Abbotsbury); The Antaios

c. Interpretation of Boilerplate Clauses Boilerplate clauses are commonly-used clauses in commercial contracts. They have occasionally attracted the attention of English courts. Under English law, they “deal with how the contract itself operates, as opposed to the rights of Compania Naviera SA v Salen Rederierna AB [1985] AC 191, 201 (per Lord Diplock); See also, Thomas Bingham, ‘A New Thing Under the Sun? The Interpretation of Contracts and the ICS decision’ (2008) 12 Edinburgh LR 374. 123 Burrows (n 103) 85-6. 124 Burrows (n 103) 86. 125 Glynn v Margetson & Co [1893] AC 351, 359 (per Lord Halsbury). 126 Hamilton v Mendes (1761) 2 Burr 1198, 1214 (per Lord Mansfield). 127 Hodge (n 115) 283; See also, The Antaios Compania Naviera SA v Salen Rederierna AB (n 122) 201 (per Lord Diplock). 128 Wood v Capita Insurance Services Ltd (n 111) [11] (per Lord Hodge). 129 Wood v Capita Insurance Services Ltd (n 111) [11] (per Lord Hodge). 130 Wood v Capita Insurance Services Ltd (n 111) [13] (per Lord Hodge). 131 Wood v Capita Insurance Services Ltd (n 111) [13] (per Lord Hodge, emphasis added).

ARTICLES

the parties … embodied in the substantive clauses”.132 Boilerplates may “regulate, control, and … modify” substantive terms and their enforcement.133 Further, draftspersons adapting standard form contracts are presumed by courts to have been aware of past decisions interpreting the clauses.134 Consequently, in cases which concern similar legal issues, they must be interpreted to embody “similar legal rights and obligations”.135 In the interests of justice and commercial efficacy, parties must be able to rely on “a uniform construction being given to standard terms”.136 This provides “certainty and consistency” in commercial affairs.137 Even if a provision is not a boilerplate clause, this approach to interpreting contracts helps to shed light on the transaction’s commercial context. In Wood v Capita Insurance Services Ltd, the Supreme Court held “the purpose of similar provisions in contracts of the same type” may be considered if a provision was lacked clarity or coherence.138 This does not necessarily mean that all similar contractual provisions are necessarily boilerplate clauses. Nevertheless, English courts may consider previous decisions when determining the contract’s commercial context.139 This approach of applying commercial context has been consistently adopted.140 Thus, courts may consider similar 132 Richard Christou, Boilerplate: Practical Clauses (6th ed, Sweet & Maxwell 2012) para 1-002. 133 Christou (n 132) para 1-002. 134 Beaufort Developments (NI) Ltd v Gilbert-Ash NI Ltd [1999] AC 266, 274 (per Lord Hoffmann); Niobe Maritime Corp v Tradax Ocean Transport SA [1995] 1 Lloyd’s Rep 579; MDIS Ltd v Swinbank [1999] Lloyd’s Rep IR 516, 521 (per Clarke LJ); McMeel (n 113) para 5.59. 135 Pioneer Shipping Ltd v BTP Tioxide Ltd, The Nema (The Nema) [1982] AC 724, 737 (per Lord Diplock). 136 The Nema (n 135) 737 (per Lord Diplock). 137 GSO Credit v Barclays Bank Plc [2016] EWHC 146 (Comm), [2017] 1 ER (Comm) 421, [27] (per Knowles J). 138 Wood v Capita Insurance Services Ltd (n 111) [13] (per Lord Hodge). 139 Seadrill Management Services Ltd v OAO Gazprom [2010] EWCA Civ 69, [17] (per Moore-Bick LJ); McMeel (n 113) para 5.59; Cf John W Carter, ‘Commercial Construction and Contract Doctrine‘ (2009) 25 Journal of Contract Law 83, 85-6 (Only general legal principles, not specific precedents, should be admissible when considering the commercial context). 140 Hodge (n 115) 283; See also, The Antaios Compania Naviera SA v Salen Rederierna AB (n 122) 201 (per Lord Diplock);

or boilerplate clauses to determine the meaning of the disputed clause or the “factual matrix” underpinning the transaction. 2. NY Law: A Comparison There are two discernible differences between the English law and NY law approaches to interpreting boilerplate clauses and contractual provisions. First, they differ in how they define boilerplate clauses. Under English law, boilerplates govern how contracts operate rather than the parties’ substantive rights and obligations. In contrast, NY law considers boilerplates to include the latter as well. Thus, whereas the pari passu clause may be considered a boilerplate clause under NY law, English courts would likely not consider it to be a boilerplate clause. Next, NY courts take a different approach where the common interpretation adopted in the market constitutes an alternative to parties’ intentions. In Utica City National Bank v Gunn,141 Cardozo J held the choice between the “meaning of the parties” vis-à-vis the “meaning in the language of business life” depends on “the genesis and aim of the transaction”. Whereas NY law considered commercial context and parties’ intentions as discrete approaches, English law takes an iterative approach142 which requires both to be considered. These differences suggest the English courts may take a different approach to how the Clause in NML should be interpreted. 3. Application of English law to the NML Clause This author argues that English courts would have arrived at the same decision as the Second Circuit in NML, albeit for different reasons. First, the pari passu clause is not a boilerplate See above, Part IV(1)(b). 141 Utica City National Bank v Gunn (1918) 118 NE 607, 608. 142 Wood v Capita Insurance Services Ltd (n 111) [13] (per Lord Hodge); Sigma Finance Corpn, In re [2009] UKSC 2, [2010] 1 All ER 571, [12] (per Lord Mance).

ARTICLES

clause in English law. It seeks to establish rights and obligations rather than merely modify them or regulate their enforcement. Thus, the rules of interpretation peculiar to boilerplate clauses do not apply. Second, English courts will likely conclude that there were two separate but related sets of rights and obligations the parties intended to create. The first sentence affirms the equal ranking of all FAA Bonds, whereas the second sentence promises payment obligations will be met equally vis-à-vis all future unsecured debt. A literal approach to interpreting contracts will lead English courts to conclude that there were two sets of rights and obligations captured in the NML Clause. Even on a contextual analysis, this interpretation is commercially sound because the state may be signalling their creditworthiness to the market to obtain lower interest rates and increased market access.143 It is not for a court to determine what contracting parties should have agreed to, let alone what representations a sovereign state should make to capital market investors. These issues were considered but left undecided in the past,144 but the Court of Appeal held against granting injunctions due to the disruptive effects on payments already made and problems with monitoring and compliance.145 Instead, English courts are likely to grant monetary relief.146 In the absence of an express provision suggesting otherwise, the textual and contextual analyses of the Clause leads to an outcome which is similar to the approach was taken in NML. Argentina may, on hindsight, regard interpretation as unfavourable to their restructuring strategy. However, this is no reason to suggest that, when they entered into the agreements, they

143 Weidemaier and McCarl (n 22) para 11.24. 144 Kensington International Ltd v Republic of the Congo (Kensington (HC)) [2003] EWHC 2331, [91] (per Tomlinson J). 145 Kensington (HC) (n 144) [92]-[96] (per Tomlinson J), affirmed Kensington International Ltd v Republic of the Congo (Kensington (CA)) [2003] EWCA (Civ) 709, [9]-[14] (per Waller LJ). 146 Kensington (HC) (n 144) [100] (per Tomlinson J), affirmed Kensington (CA) (n 145).

did not or could not have intended to pay all unsecured creditors equally. One could equally suggest that Argentina never intended to default at all. Both these statements, in this author’s view, are nothing more than mere one-sided speculations which are inadmissible under the parole evidence rule. Thus, an English court may reasonably arrive at the same conclusion as the Second Circuit in NML. IV. Proposals Substantive contractual and procedural reform proposals which may alleviate the risks of holdout litigation. Contractual and procedural reforms would be more effective in English and NY law, respectively, because they represent the preferred approach of each jurisdiction in restructuring sovereign debt.147 1. Contractual Solutions Three contractual mechanisms have been included in sovereign bonds to give creditors a greater sense of certainty and security – collective action clauses (CACs), pari passu clauses and negative pledges.148 a. Collective Action Clauses CACs are the most potent contractual tools a debtor-state can utilise to forestall any holdout litigation. The rapid trading of sovereign bonds in the secondary markets result in a panoply of different creditors with variegated economic interests holding similar rights against the debtor-state.149 There are two problems which 147 Rodrigo Olivares-Caminal, ‘Transactional Aspects of Sovereign Debt Restructuring’, in Rodrigo Olivares-Caminal et al, Debt Restructuring (2nd ed, OUP 2016) para 12.12; See generally, Marcus Miller, ‘Sovereign Debt Restructuring: New Articles, New Contracts – or No Change?’ International Economics Policy Briefs, No PB02-3 (April 2002) 3 < https:// piie.com/sites/default/files/publications/pb/pb02-3.pdf > accessed 20 April 2019; Lee C Buchheit and G Mitu Gulati, ‘Exit Consents in Sovereign Bond Exchanges’ (2000) 48 UCLA Law Review 59. 148 Alison Wirtz, ‘Bilateral Investment Treaties, Holdout Investors, and Their Impact on Grenada’s Sovereign Debt Crisis’ (2015) 16(1) Chicago Journal of International Law 251. 149 David Billington, ‘European Collective Action

ARTICLES

holdout creditors may pose to a state’s finances. First, they may cause deadlocks in debt restructuring negotiations.150 Second, creditors may, in a self-preservation effort, cause a “rush for the exit” and cause creditor flights.151 Over 90% of newly issued sovereign bonds include CACs.152 Most CACs contain three features – (a) collective representation through bondholders’ meetings; (b) majority vote requirement for restructuring; and (c) minimum vote requirement for litigation.153 These procedural mechanisms can be used for varying coupon payment obligations by restricting how bondholders can consent or otherwise to the debtor-state’s repayment variation proposals.154 Where a vote is required to vary the payment obligations, modern CACs require the consent of a “supermajority” vote, typically of threequarters,155 to bind all bondholders.156 Further, a minimum of 25% of bondholders is required for litigation to proceed.157 However, these may not reflect customary sovereign debt market practice. Debtor-states are free to structure CACs differently, but face a “Goldilocks dilemma”.158 This dilemma arises where a high threshold Clauses’, in Rosa M Lastra and Lee C Buchheit (eds), Sovereign Debt Management (OUP 2014) para 25.02. 150 Billington (n 149) para 25.04. 151 Lee C Buchheit and G Mitu Gulati, ‘Sovereign Bonds and the Collective Will’ (2002) 51 Emory LJ 1317, 13245; Billington (n 149) para 25.05. 152 Eric Helleiner, ‘Filling a Hole in Global Financial Governance? The Politics of Regulating Sovereign Debt Restructuring’, in Walter Mattli and Ngaire Woods (eds), The Politics of Global Regulation (Princeton University Press 2009) 105. 153 United Nations Conference on Trade and Development (UNCTAD), ‘Sovereign Debt Restructuring and International Investment Agreements’, IIA Issues Note, No 2, (July 2011) (UNCTAD/WEB/DIAE/PCB/2011/3) 6 < https://unctad.org/en/Docs/webdiaepcb2011d3_en.pdf > accessed 4 April 2019. 154 Stephen J Choi et al, ‘Pricing terms in sovereign debt contracts: a Greek case study with implications for the European crisis resolution mechanism’ (2011) 6(2) Capital Markets Law Journal 163, 170 (“the conditions under which the payment terms for a bond can be modified”). 155 Wirtz (n 148) 258. 156 Billington (n 149) para 25.04. 157 Wirtz (n 148) 258. 158 Billington (n 149) para 25.25.

dilutes the utility of the CAC, whereas a low threshold may be perceived by investors to be a risky investment. If a bondholder’s hands are bound too tightly, the porridge may be too cold as the bonds are less attractive to investors. This makes the bonds less liquid and hence, less attractive on the debt market. Conversely, if a bondholder’s hands are bound too loosely, the debtor-state opens itself up to higher holdout risks. Thus, this demonstrates the difficulties of balancing the interests of debtor-states and bondholders when drafting CACs in sovereign bonds. There are three different approaches to the denominator of the aggregate principal amount required to pass an amendment. The NY lawgoverned CACs typically require a specific threshold number of votes connected with all outstanding bonds.159 In contrast, English law-governed CACs require a supermajority of the bondholders represented at a quorate meeting.160 Finally, the EU Model CAC161 adopt a bifurcated approach with different supermajority thresholds for written resolutions and those passes at a bondholder meeting.162 This is because written resolutions are becoming increasingly prevalent.163 Further, the Model CAC also involves voting at both the aggregate and individual series levels.164 Thus, CACs mitigates holdout litigation risks by through 159 Billington (n 149) para 25.26. 160 Billington (n 149) para 25.26. 161 European Union Economic and Financial Committee (EFC), ‘Euro Area Model CAC 2012’ (Model CAC) (17 February 2012) < https://europa.eu/efc/sites/efc/files/ docs/pages/cac_-_text_model_cac.pdf > accessed 4 April 2019; EFC, ‘Supplemental Provisions’ (17 February 2012) < https://europa.eu/efc/sites/efc/files/docs/pages/cac_-_ supplemental_provisions.pdf > accessed 4 April 2019; See also, EFC, ‘Explanatory Note’ (26 July 2011) < https://europa. eu/efc/sites/efc/files/docs/pages/explanatory_note_draft_ on_the_model_cac_-_26_july.pdf > accessed 4 April 2019; EFC, ‘Model CAC Supplemental Explanatory Note’ (26 March 2012) < https://europa.eu/efc/sites/efc/files/supplemental_ explanatory_note_on_the_model_cac_-_26_march_2012.pdf > accessed 4 April 2019. 162 Billington (n 149) para 25.33. 163 Billington (n 149) para 25.32. 164 Deborah Zandstra, ‘A possible work plan for further reform of sovereign debt restructuring’ (2018) 13(3) Capital Markets Law Journal 356, 361.

ARTICLES

qualified voting mechanisms which prevent blockholders from initiating litigation. However, CACs are not panaceas. The specific form which CACs take may vary between issuances and need to be carefully scrutinised for their exact legal effect.165 They also do not prevent holdouts from occurring when the debtor-state defaults on multiple bond issues or there is a cross-default.166 Given the wide publicity of the voting thresholds, creditors may intentionally engineer a holdout by buying controlling holdings.167 The modified CACs, therefore, do not eliminate the problem of creditor holdouts.168 Finally, modified CACs will not eliminate holdout problem because it will take at least a decade for legacy bonds to be amended to include such CACs.169 The IMF estimates about a third of sovereign bonds which do not contain modified CACs are maturing in over a decade.170 The inertia is caused by the high costs involved in enacting sweeping contractual reforms across all bonds or even tranches or bonds issued by states.171 In this author’s view, while modified CACs are viable contractual solutions to the holdout problem in the long term, the holdout will not be effectively ameliorated by the modification of CACs alone, at least within the next decade, ceteris paribus. b. Pari Passu A relatively simple way for states to contractually reduce the risk of holdout litigation post-NML is to exclude the “equal payments” obligation from the pari passu clause. This should be

165 Wirtz (n 148) 258-9. 166 Wirtz (n 148) 259. 167 UNCTAD (n 153) 7; Wirtz (n 148) 259. 168 Zandstra (n 164) 362. 169 Zandstra (n 164) 361. 170 International Monetary Fund, ‘Third Progress Report on Inclusion of Enhanced Contractual Provisions in International Sovereign Bond Contracts’ (December 2017) < https://www.imf.org/en/Publications/Policy-Papers/ Issues/2017/12/15/pp113017third-progress-report-on-cacs > accessed 5 April 2019. 171 Zandstra (n 164) 361.

expressed in clear terms for the benefit of investors and future courts interpreting and applying the clauses.172 For instance, Belize issued new bonds in 2013 which explicitly excluded any obligation to “pay all items of its bond indebtedness on a ratable basis”.173 This expressly sought to disapply the rateable payments interpretation of the pari passu clause to the Belize bonds. However, this author argues that this does not go far enough to exclude liability under NML-type clauses, which imposes “equal payments” between bondholders and other external unsecured debt rather than just between bondholders.174 While it may seem attractive to attempt to bind the hands of external debtors by modifying the pari passu clauses, this is unworkable. This is because, save in cases of unilateral contracts made to the world at large, only parties to the contract are bound by its terms.175 c. Negative Pledges Negative pledges restrict the creditors’ ability to grant future security which ranks equal or higher, on insolvency, to the instrument in question.176 In sovereign bonds, they may be modified to require prior debt to be secured equally with the new bonds granted.177 This seeks to preserve the existing creditors’ rank.178 Thus, this prevents holdouts from inhibiting the restructuring process.179 However, this author regards both the pari passu and negative pledge proposals as ineffective because they raise the same problems of inertia as CACs. Any such changes will likely be put in force in the longterm. Thus, a consideration of other strategies to mitigate the holdout creditor risks is required.

172 173 174 175 176 177 178 179

Affaki (n 35) para 4.46 ff; Zandstra (n 164) 365. Affaki (n 35) para 4.47 (emphasis added). See above, Part III(4)(b). Carlill v Carbolic Smokeball Co [1893] 1 QB 256. Wirtz (n 148) 259. Choi et al (n 154) 14. Wirtz (n 148) 259-60. Wirtz (n 148) 260.

ARTICLES

2. Investment Treaty Arbitration (ITA) Argentina’s default on its coupon payments have also triggered a trilogy of investment disputes – Abaclat v Argentine Republic (Abaclat),180 Ambiente Ufficio v Argentine Republic (AU)181 and Alemanni v Argentine Republic (Alemanni).182 This generated a significant body of commentary183 and attracted the attention of other investors to use ITA in order to enforce their “investment” claims.184 These arbitrations arose from the bondholders’ refusal to participate in the exchange offer proposed by Argentina in 2010.185 While the decisions hitherto concernthe only jurisdiction of the tribunals and admissibility of the claims,186 three key issues which were ventilated give the investors insightinton the viability of ITAs.187 180 Abaclat and others v Argentine Republic (Abaclat), ICSID Case No ARB/07/5, Decision on Jurisdiction and Admissibility (4 August 2011). 181 Ambiente Ufficio SpA and others v Argentine Repulic (AU), ICSID Case No ARB/08/9, Decision on Jurisdiction and Admissibility (8 February 2013). 182 Alemanni and others v Argentine Republic (Alemanni), ICSID Case No ARB/07/8, Decision on Jurisdiction and Admissibility (17 November 2014). 183 See, for example, Michael Waibel, Sovereign Defaults before International Courts and Tribunals (CUP 2011); Ellie Norton, ‘International Investment Arbitration and the European Debt Crisis’ (2012) 13 Chicago JIL 291; Jessica Beess und Chrostin, ‘Sovereign Debt Restructuring and Mass Claims Arbitration before the ICSID, The Abaclat Case’ (2012) 53 Harvard International LJ 505; Joanna Simões, ‘Arbitration as a Method of Settling Disputes Arising under Sovereign Bonds’ (2012) XI(33) Revista Brasileira de Arbitragem 9; S I Strong, ‘Mass Procedures as a Form of “Regulatory Arbitration” – Abaclat v Argentine Republic and the International Investment Regime’ (2013) 38(2) Journal of Corporation Law 259; Felipe Suescun de Roa, ‘Investor-State Arbitration in Sovereign Debt Restructuring: The Role of Holdouts’ (2013) 30(2) Journal of International Arbitration 131; Youngjin Jung and Sangwook Daniel Han, ‘Sovereign Debt Restructuring under the Investor-State Dispute Regime’ (2014) 31(1) Journal of International Arbitration 75; Ryan McCarl, ‘ICSID Jurisdiction over International Mass Investment Arbitrations – Due Process and Default Rules’ (2015) 51(2) Stanford Journal of International Law 173; Wirtz (n 148) 260; Anna O Mitsou, ‘Greek Debt Restructuring and Investment Treaty Arbitration: Jurisdictional Stumbling Blocks for Bondholders’ (2016) 33(6) Journal of International Arbitration 687. 184 Wirtz (n 148) 260; See generally, Strong (n 183). 185 Norton (n 183) 297. 186 Alemmani (n 182) para 252; Wirtz (n 148) 261. 187 Norton (n 183) 298; Wirtz (n 148) 261; See also, Karen H Cross, ‘Sovereign Arbitration’, in Rosa M Lastra and Lee C Buchheit (eds), Sovereign Debt Management (OUP 2014)

a. “Investment” The first issue concerns whether the sovereign bonds fell within the definition of “investment” within the bilateral investment treaty (BIT) and the Convention on the Settlement of Disputes Between States and Nationals of Other States (ICSID Convention), where applicable.188 The Abaclat tribunal held the sovereign debt must satisfy with the conditions in both the Italy-Argentina BIT and ICSID Convention to constitute a protected “investment”.189 The tribunal held the BIT’s broad definition of investment includes “obligations, private or public title or any other right to performance or services having economic value”,190 including financial instruments.191 The only criterion for the bonds to qualify as an “investment” is that they must “create the value” which State Parties intended to protect under the BIT.192 Thus, purchasing sovereign bonds were considered an “investment” under the Italy-Argentina BIT and ICSID Convention. However, this approach to defining “investment” is not uncontroversial. In his dissenting opinion, Prof Abi-Saab highlighted two reasons why the sovereign bonds were not “investments” for the purposes of the Italy-Argentina BIT and the ICSID Convention. First, he considered the term “investment” to be capable of a much narrower “core” meaning which cannot be contracted out of by means of a BIT.193 AbiSaab also dissented on the basis that there was no territorial jurisdictional link between the sovereign bonds bought by funds on the secondary market and Argentina.194 He opined that since para 12.22 (highlighting only the first two issues). 188 Convention on the Settlement of Disputes Between States and Nationals of Other States (ICSID Convention) 575 UNTS 159. 189 Abaclat (n 180) para 344. 190 Abaclat (n 180) para 352 (emphasis added). 191 Abaclat (n 180) para 353, citing Article 1(1)(c) of the Italy-Argentina BIT. 192 Abaclat (n 180) para 365. 193 Abaclat and others v Argentine Republic (Abaclat), ICSID Case No ARB/07/5, Decision on Jurisdiction and Admissibility, Dissenting Opinion of Professor Georges AbiSaab (28 October 2011) para 46. 194 Abaclat, Dissenting Op (n 193) para 75, citing the

ARTICLES

the transactions were deliberately engineered to fall outside Argentina’s territorial jurisdiction, they could not constitute “investment[s]” under the BIT.195 Second, the “double-barrelled test” applied in Abaclat represents a departure from the approach other tribunals have adopted. For instance, Schreuer suggested that “an investment”196 must display five characteristics – (a) sufficient duration; (b) regularity of profits and returns; (c) risk sharing; (d) substantial commitment of capital; and (e) significance for the host state’s development (the “Salini” test).197 Thus, commentators have considered, but rarely agreed with, the possibility that the majority in Abaclat applied the incorrect test.198 In this author’s view, neither objection is defensible. Abi-Saab’s dissent relies on an unreasonably narrow test, which does not reflect approaches taken by arbitral tribunals. Contrary to Abi-Saab’s dissent, the “doublebarrelled test” has been repeatedly adopted by arbitral tribunals.199 Further, the tribunal in Fedax v Venezuela, a decision concerning sovereignissued promissory notes, held the fact that the investors’ identity changesares germane to the issue of whether they are “investments” so long as the “investment itself will remain constant” for the period which the sovereign-debtor “enjoy[s] continuous credit benefit until the time the notes become due”.200 The Alemanni tribunal also held that what matters is, “where and/or for the benefit of whom the funds are ultimately used, and not the place where the funds were paid

100

Preamble of the Italy-Argentina BIT. 195 Abaclat, Dissenting Op (n 193) para 78-87. 196 Art 25(1) ICSID Convention. 197 Named after the landmark decision of Salini Costruttori SpA and Italstrade SpA v. Kingdom of Morocco [I] (Salini), ICSID Case No. ARB/00/4 Decision on Jurisdiction (English translation) (2003) 42 ILM 609; Schreuer’s fivecriterian were set out in Christoph H Schreuer et al, The ICSID Convention: A Commentary (2nd ed, CUP 2011) 128-9. 198 Cross (n 187) para 12.24; Suescun de Roa (n 183) 145. 199 Schreuer et al (n 197) 117. 200 Fedax NV v The Republic of Venezuela, ICSID Case No ARB/96/3, Decision of the Tribunal on Objections to Jurisdiction (11 July 1997) para 40 (emphasis added).

out or transferred”.201 Thus, Abi-Saab’s dissent departs from well-established approaches by ICSID tribunals. This author further suggests that the lack of a unified or consistent approach to defining the term “investment” will not prove fatal to future bondholders contemplating ITAs asa means of pursuing their claims. Moreover, there is no significant distinction between the Abaclat and Salini approaches to defining “investment”. Schreuer has acknowledged that the “double-barrelled” test applies to Article 25(1) ICSID Convention.202 As with all ratione materiae jurisdictional claims, what matters is whether the parties were ad idem about the definition of substantive terms in the treaties.203 Finally, bonds are financial instruments and qualify as “investment[s]”.204 This was affirmed in Alemanni.205 Thus, the lack of any significant differences between the approaches does not introduce sufficiently significant uncertainties to deter investors from pursuing their claims through ITAs. b. Mass Claims The second issue concerns the admissibility of mass claims. As Abaclat noted, the ICSID Convention neither explicitly provides for 201 Alemanni (n 182) para 374. 202 Lanco v Argentina, ICSID Case No ARB/97/6, Decision on Jurisdiction (8 December 1998) para 48; Ceskoslovenska Obchodni Banka, AS v The Slovak Republic, ICSID Case No ARB/97/4, Decision on Jurisdiction (24 May 1999) para 55; Salini (n 197) paras 36 and 44; Joy Mining Machinery Limited v Arab Republic of Egypt, ICSID Case No ARB/03/11, Award (6 August 2004) paras 42-50; Aguas del Tunari v Bolivia, ICSID Case No ARB/02/3, Decision on Jurisdiction (21 October 2005) para 278 (“double keyhole”); Jan de Nul NV and Dredging International NV v Arab Republic of Egypt, ICSID Case No ARB/04/13, Decision on Jurisdiction (16 June 2006) para 90; Patrick Mitchell v Democratic Republic of the Congo, ICSID Case No ARB/99/7, Decision on Annulment (1 November 2006) para 31; Malaysian Historical Salvors SDN BHD v Malaysia, ICSID Case No ARB/05/10, Award, (17 May 2007) para 55; Ioannis Kardassopoulos v The Republic of Georgia, ICSID Case No ARB/05/18, Decision on Jurisdiction, 6 July 2007, para 113; MCI Power Group LC and New Turbine, Inc v Republic of Ecuador, ICSID Case No. ARB/03/6, Award (31 July 2007) paras 57—160.; See also, Schreuer et al (n 197) 117-8. 203 Schreuer et al (n 197) 118. 204 Schreuer et al (n 197) 126. 205 Alemanni (n 182) para 296.

ARTICLES

nor prohibits mass claims.206 The tribunal had to consider if this represented an “intended silence” which signals an implied prohibition or an unintended “gap” which the tribunal could fill.207 It concluded that the former interpretation would be contrary to the “object and purpose”208 of the BIT.209 Finally, the tribunal held tha, under Article 44 ICSID Convention210 and Rule 19 ICSID Arbitration Rules,211 it would not exceed their powers to allow mass claims. Thus, the ICSID Convention does not preclude mass claims and bondholders are certainly not inhibited from pursuing their claims collectively if they wish to do so. c. Treaty or Contractual Claims? The third and final issue concerns whether the claims were treaty or contractual claims.212 The latterclaims were not intended to be protected by the BI, and thus the tribunal would have no jurisdiction over purely contractual matters.213 The Abaclat tribunal held that the sovereign bonds and Argentina’s default share as their legal bases Argentina’s exercise of its “sovereign power”.214 Thus, claims arising out of sovereign bonds and their default fall squarely within the sovereign powers of states. d. ITAs as Means of Enforcing Pari Passu Clauses ITAs are not without their limitations. There has been a decline in the popularity of ITAs as means of dispute resolution on the international plane.215 Oft-cited reasons for this decline 206 Abaclat (n 180) para 517. 207 Abaclat (n 180) para 517. 208 Vienna Convention on the Law of Treaties, adopted 22 May 1969, entered into force 27 January 1980, 18232 UNTS 331 (VCLT), Article 31(1). 209 Abaclat (n 180) paras 519 and 528. 210 Art 44 ICSID Convention. 211 Rule 19 ICSID Arbitration Rules < https://icsid. worldbank.org/en/Documents/resources/2006%20CRR_ English-final.pdf > accessed 5 April 2019. 212 Beess und Chrostin (n 183) 509. 213 Abaclat (n 180) para 316. 214 Abaclat (n 180) para 323; Beess und Chrostin (n 183) 509-10. 215 See for instance, Lucy Greenwood, ‘The Rise, Fall

include, but are not limited to, inconsistent decisions and the moral hazard of parties appointing arbitrators. This has led to the creation and burgeoning popularity of specialist commercial courts, such as the Singapore International Commercial Court, which has the jurisdiction to decide on foreign disputes applying foreign law. However, it remains to be seen whether bondholders are willing to take their claims in such courts, given that issues of state immunity under the lex fori may arise to prevent such cases from being adjudicated upon. The ambiguity as to the future role of ITAs encourages parties to incorporate modified CACs and pari passu clauses into the bonds sooner rather than later.216 Moreover, even assuming that investors increasingly pursue their claims through ITAs, Norton argues this may not significantly increase, or otherwise, the legal force of pari passu and other contractual clauses.217 Bondholders may pursue claims based on “fair and equitable” treatment and nonexpropriation obligations where tribunals are more likely to grant damages than injunctions.218 This reduces the effect on third parties and payment and settlement systems. Further, this author argues that while recent decisions have sharpened the legal bite of pari passu contractual clauses, the market will not be amenable to swift changes with unclear legal consequences. 3. Sovereign Bankruptcy Regime A unified sovereign bankruptcy has been mooted as a formal restructuring and insolvency mechanism. The primary purpose of this regime and Rise of International Arbitration: A View from 2030’ (2011) 77 Arbitration 435; Michael Waibel et al, ‘The Backlash Against Investment Arbitration: Perceptions and Reality’ (2011) < https://commons.allard.ubc.ca/cgi/viewcontent. cgi?article=1193&context=fac_pubs > accessed 27 June 2019; Susan D Franck, ‘The Legitimacy Crisis in Investment Treaty Arbitration: Privatizing Public International Law Through Inconsistent Decisions’ (2005) 73(4) Fordham Law Review 1521. 216 Wirtz (n 148) 264. 217 Norton (n 183) 302. 218 Simões (n 183) 11; Suescun de Roa (n 183) 133, fn 12.

101

ARTICLES

is to allow the bankrupt state to apply for a judicial stay against creditor actions.219 They also have the additional benefit of preventing bailouts and reducing moral hazard.220 This author suggests that this solution is attractive but not feasible in the short term. The IMF began developing but shelved proposals for a Sovereign Debt Restructuring Mechanism (SDRM).221 Two features of the proposed SDRM directly addresses the problems posed by holdout creditors. First, it allows a qualified majority of creditors to bind the minority to a restructuring agreement.222 This operates like a scheme of arrangement under English law.223 Second, the SDRM proposes a mechanism to stay creditor litigation, which prevents a rush by creditors to “grab” the state’s assets through enforcement actions.224 The SDRM will also impose obligations to negotiate in good faith and disclose information.225 Though this risk is less prominent in the sovereign debt context, such litigations threatened to slow down negotiations. The main advantage of these mechanisms over CACisre that, theoretically, it binds secured and internal creditors apart from unsecured external lenders. A common dispute settlement forum226 prevents regulatory arbitrage and creditors racing to attach assets in their jurisdictions. Thus, A harmonised regime facilitates the prompt and orderly restructuring of unsustainable debt.227

102

219 Hal S Scott, International Finance: Transactions, Policy, and Regulation (17th ed, Foundation Press 2010) 1081. 220 Scott (n 219) 1081. 221 International Monetary Fund (IMF), ‘Report of the Managing Director to the International Monetary and Financial Committee on a Statutory Sovereign Debt Restructuring Mechanism’ (8 April 2003) < https://www.imf.org/external/ np/omd/2003/040803.htm > accessed 19 April 2019; Anne O Krueger, A New Approach to Sovereign Debt Restructuring (IMF, 2002) < https://www.imf.org/external/pubs/ft/exrp/sdrm/ eng/sdrm.pdf > accessed 19 April 2019; See also, Rodrigo Olivares-Caminal (n 147) paras 12.17 ff. 222 Krueger (n 221) 14. 223 Companies Act 2006, Parts 26 and 27. 224 Scott (n 219) 1081; Krueger (n 221) 15. 225 IMF (n 221); Rodrigo Olivares-Caminal (n 147) para 12.22. 226 IMF (n 221); Rodrigo Olivares-Caminal (n 147) para 12.22. 227 Krueger (n 221) 1; IMF (n 221) Part A.2; Rodrigo

However, the SDRM has several weaknesses. The lack of a “cramdown” mechanism which can be invoked by the debtor will leave them at the mercy of the supermajority of creditors.228 Further, secured lenders are excluded from the SDRM, which limits the creditors’ interests it protects.229 The lack of an insolvency regime to wind down a state’s debt when restructuring fails diminishes the viability of the SDRM. The IMF will be put in a conflicted position because it provides both the lending and restructuring mechanisms.230 The invocation oan a SDRM may lead to illiquidity in the debt capital markets as investors become more ambivalent about lending.231 Finally, due to the heavy political overtones of the sovereign debt markets,232 both states and investore will be unamenable to the SDRM, without the force of binding international law,233 because it increasee borrowing costs.234 All in all, none of the proposals for a unified sovereign debt restructuring mechanism hitherto mirrors the contractual creditor and debtor protection mechanisms currently embodied in sovereign bonds. It may even be argued that the use of a singular unified SDRM approach, without taking into consideration how the different levels of socioeconomic development may invariably differentiate the types of creditor and debtor protection mechanisms required by different states. Moreover, an arbitrarily differentiated SDRM based on legally ambiguous labels, such as “developed country” or “developing country”, may introduce opportunities for arbitrage and rent-seeking by vulture funds. This will exacerbate, not ameliorate, the holdout problem. Thus, it seems that the SDRM will not Olivares-Caminal (n 147) paras 12.17 and 12.19. 228 Scott (n 219) 1084. 229 IMF (n 221). 230 Scott (n 219) 1085. 231 Scott (n 219) 1085. 232 Rodrigo Olivares-Caminal, ‘Statutory Sovereign Debt Resolution Mechanisms’, in Rosa M Lastra and Lee C Buchheit (eds), Sovereign Debt Management OUP 2014) para 22.80; Anna Gelpern, ‘What Iraq and Argentina Might Learn from Each Other’ (2005) 6(1) Chicago Journal of International Law 414. 233 Rodrigo Olivares-Caminal (n 147) paras 12.41 and 12.43. 234 Rodrigo Olivares-Caminal (n 147) paras 12.41 and 12.43.

ARTICLES

offer additional benefits which the use of CACs does not.235 V. Conclusion A review of the literature suggests that the existing approaches to interpreting pari passu clauses are flawed. There are three, not one, categories of pari passu clauses, contrary to the NML decision, and only one contain the bare “equal rankings simpliciter” obligation. The Clause in NML clearly included an additional “equal payments” obligation. Next, this author suggests that English and NY law adopt different approaches to considering boilerplate clauses generally and the NML Clause specifically. Whereas the Clause would constitute a boilerplate under NY law, English courts will not consider it a boilerplate. However, even when applying different rules of interpretation, this author suggests that they will arrive at the same decision. Nevertheless, NY courts are likely to grant injunctions whereas English courts will only grant monetary damages.

235 Rodrigo Olivares-Caminal (n 147) para 12.50; John B Taylor, ‘Sovereign Debt Restructuring: A US Perspective’ (2 April 2002) < https://piie.com/commentary/speeches-papers/ sovereign-debt-restructuring-us-perspective > accessed 20 April 2019.

103

ARTICLES

The Murky Waters of Treaty Law: the 1962 Malaysia-Singapore Water Agreement Joel Sherard Chow, King’s College London Introduction Treaties are agreements between States governed by international law.1 These are typically concluded and negotiated by organs of the State. However, it has become increasingly common for entities such as statutory boards or municipal governments to conclude international agreements, or “arrangements administratifs”. One such agreement is the 1962 Johor River Water Agreement (“1962 Agreement”). It was concluded between the State of Johor in southern Malaysia and the City Council of the State of Singapore concluded an agreement to allow Singapore to draw 250 million gallons of raw water per day,2 providing half of Singapore’s national water supply.3 However, the price of water sold under the agreements has been a thorn in bilateral relations between both states.4 The status of the 1962 Agreement is an important preliminary question that affects its interpretation. This article will, thus, explore the legal implications of the 1962 Water Agreement on the long-standing dispute between Singapore and Malaysia. Arrangements Administrafs Arrangements administratifs are inter-State interdepartmental agreements, typically relating

106

1 Art 2(1)(a), Vienna Convention on the Law of Treaties (“VCLT”) (1969) 1155 UNTS 331 2 Johor River Water Agreement (1962) 3 Singapore-Malaysia Water Agreemnts (National Library Board, Singapore 2011) http://eresources.nlb.gov.sg/ infopedia/articles/SIP_1533_2009-06-23.html (last accessed 23 Jun 2019) 4 T. Sukumaran, “Singapore-Malaysia relations threaten to boil over as Mahathir makes splash about water prices” (4 Mar 2019, SCMP Online) https://www.scmp.com/week-asia/ geopolitics/article/2188586/singapore-malaysia-relationsthreaten-boil-over-mahathir-makes (last accessed 23 Jun 2019), R. Sim, “Singapore ‘clear, consistent’ in position that Malaysia has lost right to review water price under 1962 agreement” (13 Mar 2019, Straits Times).

to technical matters falling within the sphere of the department concerned.5 These may include financial agreements, water agreements or postal agreements to be negotiated and discussed by the department officials.6 They are becoming increasingly common given the bureaucratic red tape and lengthy negotiating periods often associated with treaties. However, as these agreements fall beyond the scope of official diplomatic channels, the question of whether international law applies is important. Under English law, the position is that not all inter-department arrangements will result in a binding treaty obligation. The Judicial Committee of the Privy Council in The Blonde and other Ships observed that:7 “although no doubt consensus ad idem is fundamentally necessary to an international agreement as it would be to a private offer and acceptance under municipal law it does not follow that in the intercourse of sovereign States every interchange of messages, some formal and some informal, should be deemed to have resulted in a new and binding agreement as soon as the parties have reached the stage of affirming identical propositions.” The key test is thus whether the department (objectively) has the administrative competence to bind the State and whether the agreement satisfies the requirements of a treaty under international law.8 On the present facts, 5 JM Jones, Full Power and Ratification (CUP, 1946), p.54. 6 See for instance, Anglo-French Postal Treaty (1856), Relatif à l’importation de grains dee vers à soie en Bulgarie (1906); B. Sen, A Diplomat’s Handbook of International Law and Practice (Springer, 2012), pp. 463-464. 7 The Blonde and other Ships [1922] 1 AC 313 at 322 8 JM Jones, Full Power and Ratification (CUP, 1946), p.61.

ARTICLES

both Singapore and Malaysia acknowledge the existence of obligations under the 1962 Agreement. As such, the administrative competence requirement is not in doubt. The sole question is thus whether the Agreement satisfies the definition of a treaty. Definition of a Treaty The starting point is Article 2(1)(a) of the Vienna Convention on the Law of Treaties (VCLT), which reflects customary international law. It defines a treaty as an ‘international agreement concluded between States in written form and governed by international law’. The two key elements in contention are whether the 1962 Agreement was concluded between “States” and whether it is governed by international law. These will be analysed in turn. 1. An Agreement Between States The 1962 Agreement was concluded between the State of Johor and the City Council of Singapore. At the time, Johor was a state within the Federation of Malaya while Singapore was a British Crown colony. Under international law, the ordinary meaning of the term “state” refers to “an independent country with standing in international law such as by membership in the United Nations, recognition by other states and adherence to international treaties”.9 Neither entity falls within the aforementioned definition. However, recourse may be made to the travaux preparatoires to confirm the meaning of the term “state”.10 The commentary to the VCLT notes that this definition seeks to exclude “other subjects of international law” such as individuals, international organizations or belligerents which had received de facto recognition.11 Both parties to the agreement do 9 Art 31(1) VCLT; L. Duhaime, Duhaime’s Law Dictionary (Online). 10 Art 32 VCLT. 11 M.E. Villiger, Commentary on the 1969 Vienna Convention on the Law of Treaties (BRILL, 2009), p. 80, Statement by Sir Humphrey Waldock in the ILC, Yearbook of the International Law Commission (1962) Vol I, 172, [32].

not fall within this category. In fact, in 1965, Singapore obtained independence and affirmed its obligations under the 1962 Water Agreement. This is confirmed by Article 14 to Annex B of the Separation Agreement.12 This “express agreement” thus allows Singapore to succeed to the obligations of the Crown colony of Singapore.13 Given that neither party to the Agreement falls within the categories that the definition sought to deliberately exclude and both parties subsequent affirmed their obligations in a later treaty, both parties are likely to be considered “States” within the context of the VCLT.14 2. Governed by International Law The second requirement is that a treaty be governed by international law. It is trite that agreements between international persons are subject to public international law.15 However, the 1962 Agreement includes an express clause to subject the agreement to Malaysian laws and jurisdiction.16 As such, the proper law governing the 1962 Agreement must be determined. The doctrine of proper law places the intention of the parties front and centre. This is similarly reflected in the drafting history of the VCLT. The second revision of the draft articles by Special Rapporteur Sir Humphrey Waldock defined a treaty as an agreement “intended to be governed by international law”.17 However, the word “intended to” was removed by the drafting committee in response to concerns from representatives that it would cause confusion. 12 Agreement relating to the separation of Singapore from Malaysia as an independent and sovereign State, 1965 UNTS 8206. 13 Article 24(a), Vienna Convention on Succession of States in respect of Treaties (1978) 17 ILM 1488. [This does not apply to independence, because it entered into force after Singapore’s independence.] 14 Art 31(1), VCLT. 15 F.A. Mann, The Proper Law of Contracts Concluded by International Persons (1960) 35 BYBIL 34. 16 Art 15, 1962 Agreement. 17 Yearbook of the International Law Commission (1962) Vol I, p. 46.

107

ARTICLES

Nonetheless, the Committee affirmed that the phrase “governed by international law” comprised the element of an intention to create legal obligations on the international plane by States. On the other hand, some academics, such as Professor Roberto Ago, have argued that parties are not free to decide whether or not a treaty is governed by international law.18 However, others have argued that the phrase would merely refer to how an agreement should be interpreted, rather than reflecting a choice of law.19 For example, the loan agreements between the International Bank for Reconstruction and Development state that the agreement would be “interpreted in accordance with the State of New York, United States”.20 Sommers argues that this agreement does not exclude the application of public international law, but merely incorporates the New York rules of contractual interpretation. However, the Privy Council in Vita Foods construed a similar clause more strictly. The Court held that the clause “to be construed in accordance with English law” reflected the intention of the parties to be bound by English law.21 On the balance, it is likely that a tribunal or court would find that Article 15 reflected the intention of the parties to be governed by Malaysian law at the time. However, interpreting the Agreement on that basis would fail to consider the dramatic changes in the relationship between the parties after Singapore gained independence. This has been emphasised by Kaplan and Katzenbach, who consider the “relationship of the parties” and the “subject matter” to be dispositive.22 Relying

108

18 Yearbook of the International Law Commission (1962) Vol I, p. 52. 19 Sommers, Broches and Delaume, Conflict Avoidance in International Loans and Monetary Agreements, Law and Contemporary Problems, 21 (1956), p. 463. 20 ibid.p. 463. 21 Vita Food Products Inc v Unus Shipping Co Ltd (In Liquidation) [1939] AC 277 at p.298. 22 Kaplan and Katzenbach, Political Foundation of International Law (John Wiley & Sons, 1964), p.236; see also Maritime Delimitation and Territorial Questions between Qatar and

on this analysis, the relationship of the parties would suggest an intention to create obligations on the international plane. First, the obligations under the 1962 Agreement were affirmed in the 1965 Separation Agreement.23 This agreement sought to outline the relationship between the newly independent state of Singapore and the Federation of Malaysia and was deposited with the United Nations. The inclusion of the water agreements in the 1965 Separation Agreement suggests that both states viewed the issue of water as central to their relationship as independent states, thereby creating obligations on the international plane. Secondly, the subject-matter also suggests an intention to create international obligations. Water is a resource of fundamental importance and, under the doctrine of permanent sovereignty over natural resources, Malaysia has exclusive right to use its water resources as it sees fit.24 There is therefore a case to be made that such international agreements relating the water should be construed as creating obligations on the international plane. However, this analysis would have to demonstrate that the subject matter and relationship between the parties would displace the text of the 1962 Agreement itself – a threshold that a court may be unwilling to cross. Without a Treaty However, even if the formal requirements of a treaty are not met, this does not mean that the 1962 Agreement is void of significance under international law. It may be argued that Singapore’s independence from Malaysia in 1965 and the continued affirmation of the Agreement by both States Bahrain (Qatar v. Bahrain) [1994] (Decision on Jurisdiction and Admissibility) ICJ Rep. 7. 23 Agreement relating to the separation of Singapore from Malaysia as an independent and sovereign State, 1965 UNTS 8206. 24 UNGA Res. 17/1803 Permanent Sovereignty over Natural Resources (1962) UN Doc. A/RES/17/1803

ARTICLES

would estop either party from asserting that the Agreement is void of significance under international law. A similar factual scenario was considered by the Permanent Court of Arbitration (“PCA”) in Chagos Marine Protected Area Arbitration (“Chagos Arbitration”).25 In that case, the agreement concerning the detachment of the Chagos archipelago was concluded between the United Kingdom and Mauritius (whilst the latter was a British colony). The Tribunal agreed that in the period of time when Mauritius was a British colony, British constitutional law would govern an agreement between the British government and a non selfgoverning territory. However, it found that the independence of Mauritius in 1968 and the commitment by both States to honouring the Agreement in their post-independence relations had effectively elevated the “package deal…to the international plane and … transforming the commitments made in 1965 into an international agreement”.26

Agreement would estop either party from asserting that no obligations existed on the international plane. First, it follows from the Chagos Arbitration analysis that the obligations under the 1962 Agreement were likely “elevated” to the international plane, purely by virtue of Singapore’s independence from Malaysia in 1965. Secondly, the inclusion of the water agreements within the 1965 Separation Agreement suggests that both parties intended to continue the arrangement, as secured by a treaty on an international level.

The PCA in the Chagos Arbitration found that the “elevation” of the Agreement to the international plane effectively estopped the United Kingdom from asserting that there were no obligations under the Lancaster House Undertakings. The principle of estoppel under international law protects the legitimate expectations of States acting in reliance on the representations of another.27 Where there is a “clear and unequivocal representation” by an “authorised official”,28 a State may not renege on a promise if there has been detrimental reliance by the other State.29

Significance of the Agreement

It may be argued that the conduct of both Malaysia and Singapore in relation to the 1962 25 Award in the Arbitration regarding the Chagos Marine Protected Area between Mauritius and the United Kingdom [2015] XXXI UNRIAA 359. 26 Chagos Arbitration, p. 540. 27 Chagos Arbitration, p. 543. 28 Temple of Preah Vihear (Cambodia v Thailand), [1962] (Separate Opinion of Vice President Alfaro) ICJ Rep. 39; Delimitation of the Maritime Boundary in the Gulf of Maine Area (Canada v U.S.A.), [1984] (Merits) ICJ Rep. 165, [307]. 29 North Sea Continental Shelf Cases (Germany v Denmark/ Germany v Netherlands), [1969] (Merits) ICJ Rep. 3, [30].

However, one counter-argument is that the inclusion of the water agreements within the 1965 Separation Agreement demonstrated that the 1962 Agreement itself had no significance under international law, thereby demanding a separate treaty. However, the existence of a separate agreement does not necessarily preclude the 1962 Agreement from creating international obligations, ipso facto.

The 1962 Agreement is likely to be the start point for any potential legal dispute between Singapore and Malaysia. Thus, its classification a treaty, or otherwise, would impact the subsequent analysis on the substantive dispute – the price of water. As a treaty, the terms of the 1962 Agreement would be interpreted in accordance with the principles in Article 31(1) of the VCLT. Clauses relating to the price of water would thus have to be interpreted in “good faith” and in its context. Singapore’s position is that Malaysia “lost” its right to amend the price of the water in 1987 unless the treaty was to be subsequently amended by both parties. However, if a court gave effect to the literal terms of the 1962 Agreement, any dispute would be governed by Malaysian law, which could potentially lead to a vastly different outcome. Thus, the preliminary question as to the status of the 1962 Agreement is pivotal to the understanding and interpretation of the water agreement. 109

ARTICLES

Conclusion The rise in arrangement administratifs between states means that the question of determining the “intention” of the parties is crucial. The Singapore-Malaysia water dispute thus presents an opportunity for clarity on the matter. The two competing positions – adopting a literal approach, versus construing the intentions of the parties otherwise – would have a significant impact on how the 1962 Agreement would be interpreted. Even if it is established that the 1962 Agreement was not initially governed by international law, both parties may be estopped from asserting otherwise, given the relationship between both states, particularly the conclusion of the 1965 Separation. Agreement.

110

ARTICLES

Paternalism vs Autonomy: Examining the moral concerns of Criminalising Suicide and Physician-Assisted Suicide Carol Yin Yihui, King’s College London Introduction Following the recent second reading of the Criminal Law Reform Bill1 in Singapore, the Parliament passed some key proposals made in the Penal Code Review.2 These include the decriminalisation of attempted suicide. It is an indication that the United Kingdom and Singapore will share an identical approach in legalising suicide, which goes against the doctrine of paternalism. However, this does not imply the states will not interfere with an individual’s personal autonomy where selfinduced harm is concerned. Certain restrictions are still implanted by both the UK and Singaporean legal authorities on attempted suicide and assisted suicide. This article seeks to comparatively analyse the legal prospects of decriminalising suicide in the UK and Singapore. In particular, it is argued that individuals ought to have the option and freedom to seek physician-assisted suicide (“PAS”). While PAS is criminalised in both the UK and Singapore, the differences in their legal responses and level of discretion on PAS will be examined comparatively. The doctrine of paternalism will be examined in conjunction with the concept of suicide. This provides a foreground for comparative analysis between the UK and Singapore on their legal measures towards attempted suicide. Specifically, the recent change to decriminalise attempted suicide by the Singapore government will be discussed. Subsequently, PAS is examined. Criminalising PAS has been subjected to prolonged debate due to its conflict with moral 1 Criminal Law Reform Bill, Singapore Parliament Bill, (06052019) [6/2019]. 2 Criminal Law Reform Bill, Singapore Parliament Bill, Second Reading Bills, Vol 94 (06052019), pp. 2-3.

concerns such as the restrictions of individual autonomy. It is argued that terminally-ill patients should be provided with an alternative to end their lives. As such, the respective legal paternalistic measures adopted by the UK and Singapore against PAS will be comparatively discussed. Suicide and the Doctrine of Paternalism The law regulates wrongful behaviour which are negative constraints on the society.3 Such conduct includes unjustifiably inflicting or threatening substantial harm to other individuals, which go against public interests. Contention lies between the law and individual liberty when the government adopts paternalistic measures that take away an individual’s freedom to choose.4 This drives the need to secure a balance between state interference and liberty of citizens. The doctrine of paternalism, as Dworkin argues, seeks to justify this state interference when the welfare, good, happiness, needs, interests, or values of the citizens5 are compromised. Paternalism refers to the state limitations on the liberties of individuals for their own interests as well as for the liberty of the general population.6 This doctrine goes against the concept of autonomy where an individual is entitled to make relevant choices over his personal life, therefore leading to over protection against harm at the expense of upholding personal liberty. It is argued that legal paternalistic 3 Mills, John Stuart and Stefan Collini, On liberty (15. print. edn Cambridge Univ. Press, Cambridge 2008), pp. 16-18. 4 Feinberg, Joel, The Moral Limits of the Criminal Law Volume 3: Harm to Self (Oxford University Press, 1989), pp. 2. 5 Dworkin, Gerald, “Paternalism,” Morality and the Law, ed. R. A. Wasserstrom, Belmont, Calif., Wadsworth Publishing Co., (1971), pp. 107-120. 6 Kleinig, John, ‘The Paternalistic Principle’ 10(2) Criminal Law, (2016), Philosophy 315, pp. 5-7

111

ARTICLES

measures do not compromise individuality, but rather promotes freedom by correcting their cognitive bias that certain harms are distant and impossible.7 Such controls include the requirement to wear a seatbelt while driving, where the law assumes that if the driver were to be warned of the potential harm, he would surely give consent to the imposed obligation of wearing a seatbelt. This is an illustration of the hypothetical consent test used to justify the paternalistic criminalisation of conducts in the situation8 where the possibility of serious harm outweighs the inconvenience of the behaviour and the autonomy for free choice on that concerning issue. Protection against harm and promoting individual welfare therefore justifies the paternalistic measures in criminalising such conduct. However, this may not be applicable for suicide. As Mills suggests, an individual may be harmed by self-induced conduct such as the failure to wear a seatbelt, yet the individual should have the right to choose to take the risk.9 In particular, a competent adult who seeks to commit suicide has the right to exercise his own autonomy and develop his individuality by choosing to end his life the way he intends. a. Comparison between Singapore and the UK In Singapore, attempting suicide was a criminal offence in section 309 of the Penal Code.10 The government believed that individuals who attempt suicide are severely distressed individuals who are unable to make rational decisions, such that state interference is deemed necessary. The criminalisation of attempted suicide also serves to promote social welfare by preventing citizens from taking their own lives. However, following the recommendations made in the Penal Code

112

7 J. D. Trout, ‘Paternalism and Cognitive Bias’ (2005) 24(4) Law Philos 393, pp. 394-395 8 ibid 9 Mills, John Stuart and Stefan Collini, On liberty (15. print. edn Cambridge Univ. Press, Cambridge 2008), pp. 105112. 10 Penal Code, Section 309.

Review on August 2018, Singapore has recently passed the Criminal Law Reform Bill11 to decriminalise attempted suicide. Prior to the repeal, Singapore’s paternalistic position to criminalising suicide was justified in preventing individuals from making a single free choice that subsequently deprives them from all their future choices.12 This choice to restrict the individual’s freedom sends the normative signal that death is not the solution to life problems. As such, instead of restricting individual liberty, the paternalistic legal measure is believed to deter suicidal conduct and contingently support individual welfare.13 However, Singapore also realised that the forceful means implemented through the criminal justice system is not the most suitable method to resolve the social problem of suicide. In fact, distressed individuals are treated like criminals by the law, which worsens their emotional states.14 Thus, it is clear that a punitive approach to control individuals in distress is inappropriate. Instead of bringing charges to those who have attempted suicide, decriminalising it will bring the focus towards rehabilitation and support for these individuals in Singapore.15 Similar to Singapore’s current legal position, Section 1 of the Suicide Act 196116 decriminalises suicide in the UK. Consistent with Mills’ harm principle, coercive interference by the state should be justified in preventing harm or the risk of harm on other individuals,17 which is not applicable for suicidal acts. The UK’s liberal assertion on suicidal conduct, as opposed to Singapore’s previous paternalistic position, 11 Criminal Law Reform Bill, Singapore Parliament Bill, Second Reading Bills, Vol 94 (06052019). 12 Cholbi, Michael, ‘Kantian paternalism and suicide intervention’ in Paternalism (Cambridge University Press, Cambridge 2013), pp. 121-125. 13 Danny Scoccia, ‘In Defense of Hard Paternalism’ 27(4), (2008) Law and Philos 351, pp. 352, 354-356. 14 Criminal Law Reform Bill, Singapore Parliament Bill, Second Reading Bills, Vol 94 (06052019), pp. 3, 17, 26. 15 ibid 16 Suicide Act 1961, Section 1. 17 Mill, John Stuart and Stefan Collini, On liberty (15. print. edn Cambridge Univ. Press, Cambridge 2008), pp. 22.

ARTICLES

indicates a limited restriction on behaviour that is primarily concerned with the individual alone. From the UK’s legal perspective on suicide, the domain of private morality and immorality is not within the responsibility of the law.18 Even if the act of suicide is thought to be wrongful, it does not breach the threshold of criminal law in causing harm to others. The UK legal authority thus respects the individual’s autonomy to commit suicide and places emphasis on individual freedom, while it seems that Singapore’s new legal position prioritises individual morality by destigmatising suicide.19

the clarity to make rational decisions. These free choices allow the individual to pursue their actual preference, promoting one’s autonomy and personal welfare. Criminalisation of suicide prevents distressed individuals from seeking necessary assistance in fear of prosecution, which aggravates the problem through a build-up of psychological pressure.23 This also explains Singapore’s new legal position to decriminalise suicide. Alternatively, individuals who are determined to commit suicide are less likely to be deterred by legal prosecution,24 thus the respective humanitarian approaches adopted by the UK and Singapore which focus on providing social support will serve to be more appropriate for those individuals under severe distress.

Besides prosecution, social welfare should be promoted through the provision of mental treatment to individuals with suicidal intent. Notably, while attempted suicide used to be a criminal offence in Singapore, prosecutions under Section 309 of the Penal Code were rare in practice. As a less coercive measure, concerning individuals are often referred to a hospital or the Institute of Mental Health for assessments.20 Mandatory charges are sometimes issued to compel the concerning individual to seek necessary treatment, for which the charges will be withdrawn when the individual arrives in a stable mental state.21 This discretion emphasises treatment instead of prosecution on of a distressed individual, which is operated similarly under the UK’s legal jurisdiction. In the UK, rather than a complete withdrawal of state control over suicidal conduct, decriminalisation serves to transfer the responsibility to the medical institutions such as the Mental Health Foundation for more effective control.22 By counselling and restoring the individual to a sane mental state, one regains

Individuals should have the right to die provided their intent is clear.25 The intent to have an alternative beyond insufferable medical treatment while maintaining their dignity through death indicates their desires to control their own lives. Attempted suicide is criminalised in certain jurisdictions because the law assumes that individuals involved in such behaviours have acted irrationally. Criminalising suicide in Singapore, for instance, serves as a paternalistic measure to deter individuals from committing suicides under the influence of psychological instability.26 As such, treatment is imposed in Singapore and the UK in order to restore a sense of rationality in the concerning individuals. However, individuals should not be deprived of their ability to make autonomous choices on their own deaths.27 For instance,

18 Tariq Teja, Live and Let Die: Bringing, PhysicianAssisted Suicide to the UK, 2 King’s Student, (2010) L. Rev. 7, pp. 8. 19 Criminal Law Reform Bill, Singapore Parliament Bill, Second Reading Bills, Vol 94 (06052019), pp. 3, 17, 26. 20 Deborah L. Kahn, David Lester, Efforts to Decriminalize Suicide in Ghana, India and Singapore, Suicidology Online 2013, (2013), pp. 101. 21 ibid 22 Hawton, K., Fagg, J., Suicide, and other causes of death, following attempted suicide, British Journal of Psychiatry, vol. 152, (1998), pp. 359-366.

23 Moore, Sheila. ‘The decriminalisation of suicide’ (ProQuest Dissertations Publishing 2000), pp. 40-43. 24 Moore, Sheila. ‘The decriminalisation of suicide’ (ProQuest Dissertations Publishing 2000), pp. 143-147. 25 Toh, Puay San and Stanley Yeo, ‘Decriminalising Physician-assisted Suicide in Singapore’ 22(1) Singapore Academy of Law Journal 379, (2010), pp. 380-381. 26 Cholbi, Michael, ‘Kantian paternalism and suicide intervention’ in Paternalism (Cambridge University Press, Cambridge 2013), pp. 116-117. 27 Toh, Puay San and Stanley Yeo, ‘Decriminalising Physician-assisted Suicide in Singapore’ 22(1) Singapore

Physician-Assisted Suicide

113

ARTICLES

the decriminalisation of suicide in the Suicide Act 1961 set the basis to lobby for legalizing Physician-Assisted Suicide (PAS).

It also reinforces the patient’s clear intention to end his life as well as the desire to take control over the time and the manner of his death.33

It is argued that dying patients ought to have the option to seek PAS in order to minimise their suffering and honour their dignity.28 Following the advancements in modern medical technology, new ways have been developed to prolong life and delay death. However, this may not be a blessing. Instead of enriching a patient’s life meaningfully, these treatments sometimes serve to only prolong the physical existence of patients who are terminally ill.29 They experience mental and emotional torment despite the provision of palliative care that reduces their physical pain.30 Hence, these patients should have the freedom to seek PAS, in order to assert control over the time and way of their own deaths.

a. Comparison between Singapore and the UK

PAS involves a licensed physician prescribing the means of suicide. The patient can subsequently choose to adopt this prescription and end his life on his own accord. Another lethal means that is discussed in conjunction to PAS is voluntary active euthanasia (VAE), where the distinction lies in the physicians delivering the death instead of the patients themselves.31 Both forms of medical interventions involve a patient with a clear intention to die and a physician who intends to aid the patient in doing so. However, it is argued that PAS should be decriminalised. In PAS, the patient makes the final decision to perform the fatal act. This conserves the patient’s autonomy to have the freedom and the right to self-determination.32

114

Academy of Law Journal 379, (2010), pp. 380-381. 28 ibid 29 E Jackson, “Death, Euthanasia and the Medical Profession” in Death Rites and Rights (B Brooks-Gordon et al eds) (Hart Publishing, 2007), pp 44-46. 30 T Cipriani, “Give Me Liberty and Give Me Death”,3 Journal of Law and Medicine, (1996) pp. 177-183. 31 Toh, Puay San and Stanley Yeo, ‘Decriminalising Physician-assisted Suicide in Singapore’ 22(1) Singapore Academy of Law Journal 379, (2010), pp. 382-385. 32 Antje Pedain, Assisted Suicide and Personal Autonomy, 61 Cambridge L.J. 511 (2002), pp. 512-513.

Abetment or assisting a complete suicide is a criminal offence in both Singapore and the UK. In Singapore, an individual abetting the suicide of another face prosecution under Section 30534 and 30635 of the Penal Code. The criminal culpability for abetting suicide is in line with that of homicide and causing death. Similarly, the UK prosecutes individuals who are involved in the suicide or attempted suicide of another under Section 2 of the Suicide Act 1961.36 In both jurisdictions, the individual who attempts suicide is not morally culpable.37 This is because the irrational act of attempting suicide indicates that the individual is under severe distress. In practice, these acts are rarely prosecuted even under the Singapore law, where attempted suicide is an offence. However, the abetter who voluntarily facilitates the suicidal conduct is morally culpable for ending the life of another,38 which justifies the consequent criminal charges for abetment. Physicians who are involved in PAS are not exempted from moral culpability as they have clear intention to cause death to the patient. In Singapore, Section 88 of the Penal Code39 provides that an act will not be construed as intending to cause death if done by consent in good faith for the benefit of a person. Despite operating under the consent of the patient and in good faith, the physician remains liable for the death due to their direct intentions to 33 ibid 34 Penal Code, Section 305. 35 Penal Code, Section 306. 36 Suicide Act 1961, Section 2. 37 Wilson, William, Criminal Law: Doctrine and Theory, Chapter 2 - Principles and Ideas Informing Decisions to Criminalise, Pearson Education UK, White Plains, (2003), pp. 37-39. 38 James Chalmers, Assisted Suicide: Jurisdiction and Discretion, 14 Edinburgh L. Rev. 295 (2010), pp. 296-298. 39 Penal Code, Section 88.

ARTICLES

assist with the suicide of the patient and cause death.40 As an alternative to the legal limitation on the operation of PAS, Singapore introduced the Advance Medical Directive (AMD)41 in 1997 that allows a terminally ill patient of sound mind to terminate extraordinary lifesustaining treatment.42 In contrast to PAS, it does not authorise assisted suicide, medication is halted to only hasten death but not directly result in death. This indicates Singapore’s respect for the dignity of terminally-ill patients. It also demonstrates Singapore’s legal position in promoting and prioritising individual autonomy.43 Consistent with the legal position in Singapore, prosecuting cases of PAS in the UK requires a breach in law, determined by assessing the intention of the physician. Under the Full Code Test, the consent to the prosecution is dependent on the discretion of the Director of Public Prosecutions (DPP). This is indicated in R v DPP (Purdy)44 where the DPP has “to clarify what his position is as to the factors that he regards as relevant for and against prosecution”.45 It requires the execution of both the evidential and the public interest stages.46 The former requires the physician to have performed an act that is capable of assisting the suicide of the patient and proof that the physician intended to assist in the suicide. For the latter, the physician is less likely to be prosecuted if the patient had a voluntary, clear and informed intention to commit suicide; the physician was wholly motivated by compassion under the determined demand of the patient and the conduct only pertains to minor assistance. 40 Toh Puay San and Stanley Yeo, ‘Decriminalising Physician-assisted Suicide in Singapore’ 22(1) Singapore Academy of Law Journal 379, (2010), pp. 382-385. 41 ibid 42 ibid 43 Toh Puay San and Stanley Yeo, ‘Decriminalising Physician-assisted Suicide in Singapore’ 22(1) Singapore Academy of Law Journal 379, (2010), pp. 386-387. 44 R (Purdy) v DPP [2009] UKHL 45, para 55. 45 ibid 46 Tariq Teja, Live and Let Die: Bringing PhysicianAssisted Suicide to the UK, 2 King’s Student L. Rev. 7 (2010), pp. 1, 19-20

For instance, relying on established guidelines, the physician in R v Cox47 was convicted for PAS by injecting potassium chloride into a suffering patient but was not sentenced. Furthermore, in R v Adams,48 Devlin J submitted that doctors are entitled to relieve pain and suffering even if the measures may incidentally shorten the patient’s life.49 The processes provide guidance in prosecuting physicians who are responsible for PAS. It indicates UK’s humanitarian attitude towards PAS. Despite the differences in the legal proceedings and measures adopted by Singapore and the UK, it is suggested that the notion of human rights and patient autonomy are paramount in criminal law.50 In both jurisdictions, consent given by a fully informed patient to be killed may reduce the liability of the physician but does not exculpate the charges completely. In a sense, this serves to protect the patient’s autonomy. Conclusion In the final analysis, this essay concludes that Singapore’s new legal position to decriminalise attempted suicide is justified to deter attempted suicide. Individuals who attempt suicide are often severely distressed and the criminal charges are not effective to sway their emotional state of mind.51 Previously, Singapore adopted a very lenient position in prosecuting individuals who attempted suicide and prioritized treatment over the consequent criminal liabilities.52 With the upcoming new legislation, more emphasis will be placed on individual morality by destigmatising suicide. The comparative 47 R v Cox [1992] 12 BMLR 38 48 R v Adams [1957] Crim LR 365, pp. 171-179. 49 ibid 50 Wilson, William, Criminal Law: Doctrine and Theory, Chapter 2 - Principles and Ideas Informing Decisions to Criminalise, Pearson Education UK, White Plains, (2003), pp. 32, 37-39. 51 Moore, Sheila. ‘The decriminalisation of suicide’ (ProQuest Dissertations Publishing 2000), pp. 143-147. 52 Deborah L. Kahn, David Lester, Efforts to Decriminalize Suicide in Ghana, India and Singapore, Suicidology Online 2013, (2013), pp. 101.

115

ARTICLES

analysis of the UK and Singapore indicates that Singapore is following the global shift to decriminalise suicide53 and individual autonomy by moving away from pure paternalism. On the other hand, it is noted that Singapore and the UK adopt similar legal positions on the criminalisation of PAS. While these paternalistic legal measures against PAS serve to protect patients against self-harm,54 it is also motivated for the protection of patients in general. By criminalising PAS, it would prevent doctors from readily ending lives through mercy killing. It also protects patients from any instances of malicious culpable motives some doctors may have.55 Even though legalising PAS risks creating uncertainty in the law on homicide, it is argued that patients remain in control over their lives and their intent should justify the conduct of PAS. As such, legalising PAS should be advocated.

116

53 L Mishara, Brian & N Weisstub, David, The legal status of suicide: A global review. International Journal of Law and Psychiatry, (2015), pp. 1-3. 54 James Chalmers, Assisted Suicide: Jurisdiction and Discretion, 14 Edinburgh L. Rev. 295 (2010), pp. 295-297. 55 Tariq Teja, Live and Let Die: Bringing PhysicianAssisted Suicide to the UK, 2 Kingâ&#x20AC;&#x2122;s Student L. Rev. 7 (2010), pp. 1, 13.

ARTICLES: WRITTEN MOOT 1

Why Carol Ann Armstrong does not introduce uncertainty: a comparative analysis Tan Yin Theng Sarah, University of Bristol Introduction In order to fully appreciate the dilemma posed by Gregg v Scott1 and Armstrong, Carol Ann v Quest Laboratories Pte Ltd,2 it is necessary to first define the meaning of loss of a chance. This concept is typically seen in cases of where the defendant’s negligence results in a missed opportunity to recover damages. The lost chance doctrine can be traced back to the case of Chaplin v Hicks,3 where the plaintiff, Ms Chaplin was allowed to claim for her lost opportunity in presenting herself to gain employment due to the defendant, Mr Hicks’ failure to notify her amounting to a form of breach. In this case, Vaughan Williams LJ ruled that “the fact that damages cannot be assessed with certainty does not relieve the wrong-doer of the necessity of paying damages for his breach of contract”.4 The first appearance of loss of a chance in medical negligence cases is seen in the case of Hotson v Berkshire Area Health,5 where loss of chance was not applied to personal injury in the case of negligence causing a 25 percent lowered chance of recovery. In this case, the “all or nothing” approach was applied where it is the duty of the claimant to prove that (of a 51% likeliness) that the defendant’s breach had caused the loss incurred by them. The current ruling on loss of a chance of recovery has been laid out in the case of Gregg where it was held that the prior ruling in Hotson should still apply, resulting in the claimant being unable to claim for his loss of a chance at recovery. 1 Gregg v Scott [2005] 2 WLR 268 (“Gregg”) 2 Armstrong, Carol Ann v Quest Laboratories Pte Ltd (“Armstrong”) [2018] SGHC 66 3 Chaplin v Hicks [1911] 2 KB 786 4 Chaplin v Hicks [1911] 2 KB 786 at 792 (per Vaughan Williams LJ) 5 Hotson v East Berkshire Area Health Authority [1987] AC 750

This article aims to compare and contrast the details of that case in the United Kingdom to the case of Armstrong, Carol Ann v Quest Laboratories Pte Ltd6 in Singapore that allowed for the plaintiff to claim for damages in their claim for loss of a chance. This essay argues that the Armstrong ruling does not further complicate the ruling of loss of a chance in the United Kingdom despite the apparently contradictory result. This is due to the fact that both cases, while similar in nature, can still be differentiated, allowing for the law to be flexible in dealing with loss of a chance in other medical diseases rather than confining all future claimants to one particular ruling. Therefore, it would be fair to say that the ruling in Armstrong, Carol Ann would add clarity to the current law on lost chance. This article would start by providing a description of both cases followed by an analysis on the ruling of Armstrong and how it can be seen as an extension of Gregg v Scott and finally the possibility of its application in the United Kingdom. Gregg v Scott: Facts and rulings of the case The claimant, Mr Gregg, had developed a lump under his left arm in 1994 and sought the professional help of the defendant, Dr. Scott, a general practitioner who informed him that the lump was benign and no further action was needed, advice which the claimant followed. The claimant subsequently relocated in 1995 and registered himself with another general practitioner, who, upon seeing the lump on his arm, referred him to a hospital where he was diagnosed with non-Hodgkin’s lymphoma. At this point, the tumor had spread to his chest and 6 This essay makes its arguments in light of the Singaporean High Court judgment, whilst being aware that the Singaporean Court of Appeal judgment is still pending at time of publication.

117

ARTICLES: WRITTEN MOOT 1

he proceeded to suffer suffered relapses despite treatment, rendering his chances of survival extremely low. Accordingly, the claimant brought an action against the defendant, claiming that had he been referred to a hospital back in 1994, he would have had a high chance of recovery. i. Court of Appeal In the Court of Appeal, the claimant asserted that the courts ought to consider whether reducing the risk of an adverse outcome as a result of medical advice or medical intervention, should be recognised as damage giving rise to a claim in negligence. Also, the claimant submitted that the court should consider as a matter of policy whether or not such a claim could properly succeed. However, this appeal was ultimately dismissed. Mance LJ held that “English law should not contemplate what would be essentially speculative actions for loss of life expectancy, based on no more than general statistical evidence. Even where a claimant had suffered physical injury and it could be shown that there had been some negligence which might have contributed to that injury, a claimant could not succeed simply on the basis that there was a chance that the negligence caused the injury”7 therefore highlighting the reluctance of the Court of Appeal to take a stance on how far the courts are allowed to provide damages based on lost chances. It was also found that even if the claimant was treated promptly, he would only have had a 42% chance of survival.8 It was thus held that while the defendant was negligent in not considering that the growth could be malignant and was guilty of breach of duty; the claimant could not prove that his outcome would be materially different if he had been treated sooner. ii. Supreme Court The claimant’s appeal was still dismissed by a 3:2 majority with their Lordships holding that the prior ruling of Hotson would continue to

118

7 8

Gregg v Scott [2002] EWCA Civ 1471 at [80]-[81] ibid. at para [14]

apply. The reasoning for this can be found in Baroness Hale’s speech explaining the potential difficulties that can arise should the courts accept claims for loss (of less than 50%) chance of recovery. She held that accepting a loss of a chance approach that allows claimants to recover for the lost opportunity of recovery “would cause far more problems in the general run of personal injury claims than the policy benefits are worth”.9 This is largely due to the fact that there is little to no basis for calculating the odds due to the fact that in any case, the defendant would always be held liable for something in any event and would have also lost the benefit of chance that causation could not prove in his favour. Furthermore, this creates further confusion on the side of claimants as those who would be able to make a full recovery may not be able to in the future should proportionate recovery be put in place. This would create further confusion and unpredictability for future cases. This was supported by Lord Nicholls in his dissent holding that these cases are often difficult to prosecute due to the outcomes of each individual case being “inherently uncertain”10 however his Lordship continues to point out that despite the supposed difficulty, ultimately the duty of the law is to “protect medical reality”11 and in cases of a breach, the law should recognise the need to remedy the loss of the thing that was his doctor’s duty to protect. Applying the “all or nothing” approach in cases like this would not serve to protect the interests of patients as having them prove on the balance of probabilities that the breach had resulted in reduced chances of recovery are often difficult to prove and may result in individuals being unable to claim for damages in situations which they should be able to claim for. This was highlighted by Lord Nicholls who cited the example that “a patient with a 60% chance of 9 Gregg v Scott [2005] 2 WLR 268 at [225] (Per Baroness Hale) 10 Gregg v Scott [2005] 2 WLR 268 at [26], (Per Lord Nicholls) 11 Gregg v Scott [2005] 2 WLR 268 [42], (Per Lord Nicholls)

ARTICLES: WRITTEN MOOT 1

recovery reduced to a 40% prospect by medical negligence can obtain compensation. But he can obtain nothing if his prospects were reduced from 40% to nil.”12 However, this does not fully address the loss that the patient has suffered. In addition, it can be observed that the line of reasoning put forth by Baroness Hale, that accepting loss of a chance as actionable damage could increase the floodgates of future cases is lacking in practical substance due to the fact that courts should be able to take into account the individual situations and apply different remedies accordingly. This is supported by the fact that loss of a chance has been accepted as a form of actionable damage in several states of the United States.13 One case that puts forth a strong argument for loss of a chance is Dickhoff v Green.14 In this case, the parents of a premature infant who discovered a lump was reassured by doctors that it was benign. However, it was discovered after a year that the child was found to have metastatic rhabdomyosarcoma and that the delay in diagnosis reduced the child’s odds of survival from 60% to 40% and the courts recognised that this breach amounted to compensable damage. The courts also held that recent advancement in medical science have allowed us to effectively gauge the patient’s chances of survival to reasonable degree of medical certainty making it possible to prove certainty,15 countering Baroness Hale’s statement that proving loss of a chance would be too difficult.

Tan (the second defendant) two days later, ruled that the specimen sent over was benign with “no malignancy”. Two years later, swelling occurs on the right armpit of Peter Traynor, prompting him to seek the medical advice of a Dr Ang who calls for a further review on the report by Quest Laboratories by another pathologist. This time, the report dated in January 2012 revealed the specimen to be “malignant melanoma with ulceration”.16 As the first report was incorrect, Peter Traynor subsequently died due to malignant cancer cells spreading throughout his body during the time from 2009 to 2013. This action was brought by Carol Ann, his widow against Dr Tan and Quest Laboratories for their negligence in reporting the malignant specimen as benign and failing to diagnose the cancer. i. The Singapore High Court Ruling

In 2009, a piece of skin from a mole of the back of Peter Traynor was extracted and sent to Quest Laboratories (the first defendant) for examination. A pathology report signed by Dr

The judgement was put forth by Justice Choo Han Teck who ruled that Dr Tan’s negligence had caused Mr Traynor to die years earlier than he would have had. Justice Choo had drawn upon the dissenting judgement in Gregg v Scott particularly that of Lord Nicholls as to determine whether a doctor’s negligence could sound in damages on the basis only of a loss of chance. In that particular case, Lord Nicholls had suggested that the courts should consider “leaping the evidentiary gap when overall fairness plainly so requires”.17 Justice Choo also recognised that while it may be hard to say how long Mr Traynor would have lived had the disease been discovered sooner18 due to opposing medical opinions, it does not mean that there cannot be any compensation put in place.19 That being said, Justice Choo had also disagreed20 with the defendants claim that Peter Traynor had a 68% chance of surviving ten years but merely extrapolated from the facts of the case to say that he might have lived twice as long (8 years) as he had lived had the diagnosis

12 ibid at [46] 13 B Sonny Bal and Lawrence H Brenner, “Medicolegal Sidebar: The law and Social Values: Loss of Chance”, (2014), Clinical Orthopedics and Related research 472(10), 2923–2926. 14 Dickhoff v Green [2013] 836 NW 2d 321 15 ibid at 335

16 Armstrong at [2] 17 Armstrong at [16] 18 ibid. 19 Armstrong at [18] 20 Armstrong at [19]

Armstrong, Carol Ann v Quest Laboratories Pte Ltd: Facts and ruling of the case Facts

119

ARTICLES: WRITTEN MOOT 1

been accurate.21 Problematic ruling? The problematic nature of this judgement by the High Court is due to the fact that its end result directly contradicts that of Gregg v Scott. Justice Choo’s acknowledgement of the uncertainty as to how long Mr Traynor would have lived and his refusal to place a percentage on his final judgement is similar to that of Mance LJ’s holding in the Court of Appeal. It was said that one could only claim damages if one could prove on the balance of probabilities of 51% and being unable to place a number as to how much Dr Tan’s act caused Mr Traynor’s death in in direct conflict with the usually prescribed rule. This could therefore open up more litigation in the area of medical claims where individuals who may have experienced some form of negligence that may or may not have caused their loss. The lack of a clear guideline that was previously held in Gregg v Scott cases ambiguity in this area of law and could result in a variety of different outcomes that could boil down to factors such as timing, the type of disease etc. Thus fulfilling the prediction put forth by Baroness Hale in her obiter comments that more claimants would come forth seeking damages and it could possibly be an area of public policy and not that of tort law to ascribe the type of damages one could receive in situations that may not fulfill the balance of probabilities requirement.22 Extending Gregg v Scott Cane argues that “there are certain aspects of human wellbeing which can be regarded as fundamental or core moral interests that setbacks to these interests constitute harm”23 Such interests could relate to that of life and bodily integrity.24 This is recognised in the

120

21 ibid. 22 Gregg v Scott, at [79]-[80] 23 Peter Cane, “Tort Law and Public Functions” in John Oberdiek, Philosophical Foundations of the Law of Torts (OUP 2014) 24 ibid.

area of medical negligence in tort where if a doctor’s negligent act results in the claimant’s suffering that would not have happened had the negligent act not occurred. Thus it would be fair to compensate one for a lost chance when the defendant’s one negligent act set the claimant on a radically different course of recovery with lowered chances and has an equally relevant place in the area of tort and compensation. While the arguments against this advanced by Baroness Hale and Mance LJ are sound, one need only look further into the judgement of Carol Ann Armstrong to identify the fundamental differences between both cases. These differences include the fact that the diseases suffered by Mr Gregg and Mr Traynor are very different in their nature and manifestation which warrants distinction between them and the way they are treated. This was noted by Justice Choo where he found that “the prognosis for non-Hodgkins lymphoma is not the same as melanoma”.25 Therefore the way melanoma progresses and how it manifests in its patients would differ from that of nonHodgkins lymphoma and it would not be fair to have Gregg v Scott as a true precedent for Armstrong as the diseases are wholly different to begin with. He had also stressed on the need to focus on Mr Traynor and not get caught up in the mechanics of Gregg v Scott.26 The next distinction between both cases is the simple fact that Mr Traynor had passed away and that Mr Gregg was still alive. This means that the argument of “lost years” due to the defendant’s mistake more compelling due to the fact that we can truly ascertain how much was “lost” by counting from the years between his death and the age which he would have lived had the diagnosis been accurate. As opposed to Mr Gregg who was alive, it is comparatively harder to estimate how much Mr Gregg had lost in terms of years. With these differences, it is fair for the law to be flexible to accommodate differing situations and claims as opposed to completely closing the door on any claims 25 Armstrong at [16] 26 ibid.

ARTICLES: WRITTEN MOOT 1

relating to loss of a chance when it may be fair to do so depending on the situation. Application in the United Kingdom It is this author’s opinion that this case could be applied in the United Kingdom. With such a subjective area as loss of a chance, having the need to prescribe a percentage as to how much the defendant’s negligent act had resulted in the claimant’s loss of a chance can be challenging and overly restrictive. Justice Choo’s choice to eschew utilizing percentages, citing their purpose to be “for doctors to advise their patients”27 as they did not reflect the finer detail of Peter Traynor living for a further 4 years is therefore justified. By simply focusing on the details on the claimant at hand, judges would be able to have greater flexibility in awarding compensation when it is due, rather than establishing an arbitrary 51% which might not result in the most equitable outcome. This can be seen in Gregg v Scott where Lord Hoffman posits that everything is determined by causality” and that “knowledge” which the law deals with “by the concept of the burden of proof ” is lacking.28 This displays a failure to set apart “an act which causes a claimant’s damage and one which causes a claimant to lose a better than even chance of avoiding that damage”,29 therefore resulting in a lack of compensation30. The reasoning in Carol Ann Armstrong could be utilised when the courts preside over cases where a loss of a chance could be compensated. This is as the judgement lays out how damages could be distributed for loss of a chance due to the fact that the United Kingdom currently does not have any case that tackles the area of compensation for loss of a chance cases due to its history of not recognizing loss of a chance as a claimable tort. In addition, the ruling of Wright v Cambridge Medical Group,31 a case similar 27 Armstrong at [19] 28 Gregg v Scott [2005] 2 AC 176 at [79] 29 Margaret Fordham, “Loss of Chance—A Lost Opportunity?”, (2005) Singapore Journal of Legal Studies at p 205 30 ibid. 31 Wright v Cambridge Medical Group (a partnership) [2011]

to that of Armstrong in which the defendant was negligent in referring his 9 month old patient 4 days late to the hospital resulting in the patient being unable to receive proper medical treatment when she most needed it causing her to suffer permanent hip bone damage. In that case, it was held that the defendant was liable for the claimant’s losses and that the appeal was allowed on a majority of 2:1. Lord Neuberger held that “by negligently failing to refer the patient promptly, he deprived her of the opportunity to be treated properly by the hospital, and, if they had not treated her properly, that opportunity would be reflected by the fact that she would have been able to recover damages from them”32 This demonstrates a willingness to accept that the defendant’s negligent act had caused the claimant’s lowered chance of recovery resulting in her suffering permanent damage. However, he declined to ascribe any damages to the claimant citing the need for consistency in the difficult area and therefore chose to follow the ruling in Gregg v Scott33 where the claimant would not be entitled to damages. That being said, he had also expressed a wish for this issue to be addressed by the Supreme Court in a future case. In other words, the reasoning in Armstrong should be applied in the UK due to the fact that there has not been a case where damages have been allocated in the area of loss of a chance in medical negligence. Conclusion This case demonstrates the ability of the Singapore High Court to forgo certain established principles of causation to award damages where it was found to be necessary. By departing from the strict rule that was held in Gregg v Scott, the Court established a more Med LR 496, [2011] EWCA Civ 669, [2013] QB 312, [2012] 32 Wright v Cambridge Medical Group, [2011] EWCA Civ 669 at [61] 33 ibid. at [84]

121

ARTICLES: WRITTEN MOOT 1

flexible rule that allows justice to be served unencumbered by the difficult requirement for loss of a chance to be regarded as a compensable loss. Should the United Kingdom be able to apply Armstrong, the families of claimants who have passed due to being reduced chances of survival following a medical practitionerâ&#x20AC;&#x2122;s negligent act will be able to claim for damages. This would increase the protection for future claimants against negligent malpractice. That being said, in the course of medical treatment, practitioners are often exposed to a range of options, all of which remains open to future criticism by experts who might claim that they would have done otherwise and yield better outcomes which can expose the medical profession to the risk of payments in future litigation. Nonetheless, the strict rule put forth by Gregg v Scott is one of inflexibility and the inclusion of Armstrong provides a sense of balance to this issue and lends further thought into the issue of loss of a chance to ensure a more equitable outcome. Therefore, it is this authorâ&#x20AC;&#x2122;s opinion that Carol Ann Armstrong does not wholly contradict the ruling in Gregg v Scott due to the key differences between both cases that warrant their differing outcomes.

122

ARTICLES: WRITTEN MOOT 1

On the basis of fairness? Taking a chance on recognising lost chances in Singapore: a comparative analysis Lau Yi Hang, Singapore Management University I. Introduction “Our lives are defined by opportunities, even the ones we miss.”1 Roth could not have put it better in expressing the importance of every opportunity that is presented before an individual – a chance for change of one’s life. And should this chance be unfairly taken away, there would always be a profound impact on life. The loss of chance doctrine has plagued the tort of negligence in common law with uncertainty2 While it is settled in pure economic loss cases concerning negligent legal advice,3 the debate has not yet been foreclosed regarding physical injury arising from medical negligence. While most cases across common law jurisdictions seem to be settled on not recognising this doctrine,4 it would appear that Singapore intends to go in a different direction. In particular, the Singapore High Court5 in Armstrong Carol Ann v Quest Laboratories Pte Ltd (“Armstrong”)6 saw Justice 1 Eric Roth, The Curious Case of Benjamin Button Screenplay. 2 See Andrew Burrows, ‘Uncertainty about Uncertainty: Damages for Loss of a Chance’ [2008] Journal of Personal Injury Law 31 and Harvey McGregor, ‘Loss of Chance: Where Has It Come From and Where Is It Going?’ [2008] Professional Negligence 1. 3 See for example Allied Maples Group Ltd v Simmons & Simmons (a firm) [1995] EWCA Civ 17, [1995] 1 WLR 1602. The application of the loss of chance doctrine encompasses both making of a financial gain and avoidance of financial loss. 4 See for example Gregg v Scott [2005] UKHL 2, [2005] 2 AC 176 (HL) where the UK House of Lords by a 3-2 majority denied a loss of chance claim in the context of medical negligence; See also Tabet v Gett [2010] HCA 12, [2010] 240 CLR 537 (HCA) where the loss of chance doctrine was rejected in Australia’s law of negligence; cf. the recognition of the Loss of Chance in US’s tort jurisprudence in Matsuyama v Birnbaum, 890 N.E. (2d) 819 (2008) (Mass. Sup. Jud. Ct.). 5 It is noted that Armstrong has since gone on appeal and the hearing was completed earlier this year in January, although the judgment was ultimately reserved. Therefore, this paper is written taking into account the context in which only the High Court’s judgment was available for commentary. 6 Armstrong, Carol Ann (executrix of the estate of Peter

Choo Han Teck’s support for the minority dissenting voices of Lords Nicholls and Hope in Gregg v Scott in recognising a loss of chance claim in the context of medical negligence. The claimant was unable to fulfil the balance of probabilities requirement in demonstrating causation.7 In a few paragraphs,8 Choo J’s judgment profoundly changed the law of negligence by recognising loss of chance as an actionable head of damage in the context of physical harm. The broad objective of this essay thus seeks to understand Choo J’s differing approach and his Honour’s basis for recognising a loss of chance claim in Armstrong, and to compare his approach vis-a-vis the UK House of Lords in the seminal decision of Gregg v Scott (“Gregg”).9 Whether loss of chance is an actionable head of damage, proving causation on a balance of probabilities and quantification of loss are all important issues which have been hotly debated.10 There has also been a failure to realise that each element calls for its own separate inquiry.11 To answer these questions, Part II examines the judgment in Armstrong and seeks to demonstrate that Choo J’s approach is a policy-driven one. This is characterised by apparent uncertainty in the causal process as a result of a lack of evidence. Part III then seeks to reconcile the theoretical and legal basis of the traditional causation test in negligence. It also explores why courts in certain instances resort to a Traynor, deceased and on behalf of the dependents of Peter Traynor, deceased) v Quest Laboratories Pte Ltd and another [2018] SGHC 66. 7 ibid.,[16] and [17]. 8 ibid., [16] to [19]. 9 Gregg (n 4) 10 Andrew Burrows (n 1) 37; McGregor (n 1) 6. 11 Gemma Turton, Evidential Uncertainty In Causation In Negligence (Hart Publishing 2016) 125.

123

ARTICLES: WRITTEN MOOT 1

modification of the causation test to establish liability. This analysis will demonstrate that policy driven reasoning is due to evidential gaps allowing for circumvention of the traditional causation test in attributing liability. Part IV then examines whether a loss of chance claim should be recognised on the basis of policy reasons which function to circumvent the traditional causation test. This is remembering that policy, an “unruly horse”12 if left untamed would give rise to uncertainties in the law. This essay argues against such an approach, primarily because of the differences between the nature of the evidential uncertainty between deterministic and indeterministic causal processes. To ignore this distinction would be to render the law uncertain regarding when policy arguments can be invoked to circumvent traditional causation. Part V will examine strong policy reasons for and against the recognition of a loss of chance claim, further arguing that the validity of these policy arguments should perhaps be left to the Legislature to resolve. II. The judgment in Armstrong In 2009, the deceased, Mr Peter Traynor, developed an ulcerous skin lesion. His doctor removed and sent a sample piece of skin to the first defendant, Quest Laboratories Pte Ltd, where it was analysed by the second defendant, Dr Tan Hong Wui, for signs of malignancy. A pathology report signed off by Dr Tan concluded that there was “no malignancy”. Three years later, Mr Traynor developed a swelling under his armpit. This was diagnosed as cancer, which had metastasised. A further review of the sample extracted from the deceased 3 years ago returned with the diagnosis of malignancy. Mr Traynor died eighteen months later, and his wife brought an action against the defendants. It was clear that Dr Tan had breached his duty was established for failing to undertake further examination when the first specimen “did not

124

12 Richardson v Mellish [1824] 2 Bing 229, 252, [1824-34] 130 ER 294 (Burrough J).

appear to be a normal healthy cell”.13 ”. It was established that “at the very least, there was a loss of an early opportunity for treatment.”14 The issue, was whether Dr Tan’s breach of duty could have been an effective cause of Mr Traynor’s death.15 The expert witnesses for both sides had opposite conclusions.16 The expert witnesses’ evidence were crucial because whether Dr Tan was to be found liable would depend very much on the conclusion which was accepted by Choo J. If it were accepted that the cancer in Mr Traynor’s body had spread to a point where it could not be cured, then it was the cancer, and not Dr Tan’s breach of duty, which factually caused Mr Traynor’s death. The plaintiff ’s expert witness relied heavily on statistics which demonstrated that Mr Traynor had a 68% to 80% probability of surviving for 10 years if there had been proper diagnosis in 2009.17 On the other hand, the defendant’s expert witness claimed that there was less than 50% probability for a 10-year survival in 2009.18 Admittedly, Choo J stated that the statistics did not reflect whether Mr Traynor fell in the majority 80% that would have survived for another 10 years had he received immediate treatment, or would he be among the 20% “black swan of melanoma”.19 The evidential uncertainty thus precluded any support to the definitive finding on whether, on a balance of probability, Dr Tan’s negligence did cause Mr Traynor’s death. The plaintiff ’s case should have, unfortunately, ended here. 13 Armstrong (n 6) [7] and [8]. 14 Armstrong (n 6) [8]. Emphasis added. Notice that Choo J had framed the loss as an “opportunity for early treatment” which is in substance largely similar to a loss of chance for a better outcome, since early treatment implies the possibility that Mr Traynor’s cancer would stand a better chance of being cured. 15 Although this was not explicitly mentioned by Justice Choo to be the plaintiff ’s claim, it was mentioned at [9] of the judgment that one of the issues considered was whether Dr Tan’s negligence had caused the death of Mr Traynor. 16 Armstrong (n 6) [9] to [12]. 17 Armstrong (n 6) [12]. 18 Armstrong (n 6) [12]. 19 Armstrong (n 6) [15].

ARTICLES: WRITTEN MOOT 1

Nevertheless, Choo J, who was inspired by Lord Nicholls in Gregg,20 held that the High Court should “leap an evidentiary gap when overall fairness plainly so requires”21 and decided that, notwithstanding the medical uncertainty in the evidence, Dr Tan must be found negligent for causing Mr Traynor to “lose a fighting chance”,22 and that his untimely death be compensated since Mr Traynor’s loss “is still a loss”.23 In other words, even if his Honour could not come to a conclusion that Dr Tan’s negligence was causally connected with Mr Traynor’s death, the argument on fairness dictates that the causal connection be established through making a determinative holding that Mr Traynor was a survivor, and that Dr Tan’s negligence had at the very least caused Mr Traynor’s chance of survival to be extinguished. The first part (i.e. the assumption that Mr Traynor was a survivor) relates to the problem of proof. The basis of this assumption is the inability to prove whether Mr Traynor was a survivor is limited not by the plaintiff ’s capacity but by medical scientific knowledge. It is debatable whether the court should have used the policy reason of fairness to overcome the evidential gap.24 The focus is on the second part of this process (i.e. the extinguishing of a chance of survival). A recharacterization of the physical injury suffered as a loss of a chance for a 10-year survival would remove the plaintiff ’s burden to prove that Dr Tan’s negligence, and not the cancer, has to be proved, on a balance of probabilities, to be an effective cause of death, thus overcoming the evidential uncertainty faced in proving causation. This process assumes that 20 Gregg (n 4) [25] – [31] (Lord Nicholls). 21 Armstrong (n 6) [15] (citing Lord Nicholls in Gregg (n 3) [25] and [31]). 22 Armstrong (n 6) [19]. 23 Armstrong (n 6) [18]. 24 See Gregg (n 4) (Lord Phillips). His Lordship pointed out that the limitations of statistical evidence and its fallacies in using a general to prove a particular is a matter of policy, with Lord Nicholl’s judgment based on the policy of fairness. See also Margaret Fordham, ‘Loss of Chance - A Lost Opportunity’ [2005] Singapore Journal of Legal Studies 204, 213. Professor Fordham submits arguments in favour of upholding this policy of fairness.

Dr Tan’s negligence did in fact cause a loss, and once the loss has been identified, all that has to be shown, on a balance of probabilities,25 is the counterfactual that Mr Traynor did have a chance of survival, which Dr Tan’s breach has caused to be extinguished in the factual reality. However, on what basis did Justice Choo permit the claim for lost chance? The answer is that it was impossible to conclusively prove a causal nexus between Dr Tan’s negligent act and the death of Mr Traynor.26 Thus, Justice Choo’s belief that justice be done, and to see that Mr Traynor has had his day in the court, no doubt forms the basis for which a loss of chance claim ought to be recognised to circumvent having to proof causation in the initial way where the head of damage pleaded is the death of Mr Traynor. In short, causation was the problem, fairness was the motivation and a loss of chance claim was the solution. Justice Choo’s belief is certainly admirable. After all, it is no better exemplification of the maxim ubi jus ibi remedium,27 and it would appear that Dr Tan’s negligent diagnosis has done some wrong upon Mr Traynor so as to take away his chance of a longer lifespan, which certainly deserves compensation – there was, clearly, a dereliction of his duties as a medical professional. An appeal to policy reasons of fairness to circumvent causation in establishing liability would prevent the materialization of an empty duty28 and incentivise medical professionals to take greater care when lives are at stake. However, inherent in Justice Choo’s reasoning is the assumption that there was indeed a problem of evidential uncertainty which led 25 See Jeremy Liang Shi Wei and Low Kee Yang, ‘Recognising Lost Chances in Tort Law’ [2014] Singapore Journal of Legal Studies 98, 114-117. It was argued that a loss of chance claim would still preserve the balance of probabilities standard of proof, and the only issue lay in the reliability of evidence. 26 Armstrong (n 6) [16] (citing Gregg (n 4) [25] (Lord Nicholls)). 27 The latin translation is as follows: For every wrong, the law provides a remedy. 28 Gregg (n 4) [4] (Lord Nicholls).

125

ARTICLES: WRITTEN MOOT 1

to the inability to prove causation on a balance of probability, and when combined with the need to ensure that fairness be upheld, thus facilitated an alternative approach. Whether this is indeed the case turns on our understanding of causation which may explain the need for this alternative approach. III. Causation in negligence A. The traditional requirements of causation 1. Standard of proof – the Balance of Probabilities In civil claims, the burden of proof lies on the claimant to establish his case on the balance of probabilities. This requires establishing that the facts require an application of the law.29 This also applies to establishing causation, where the burden lies with the plaintiff to prove, on a balance of probabilities, that the defendant’s breach of his duty has caused the loss.30 2. Factual and legal causation – the “But For” test and effective cause The starting point for analysis is the “but for” test. Claimants must demonstrate that harm would not have resulted had the defendant acted in a non-negligent manner.31 However, the “but for” test may be insufficient in certain cases since it only filters out irrelevant causes.32 In addition, legal causation in the form of an “effective

126

29 See Steve Gold, ‘Causation in Toxic Torts: Burdens of Proof, Standards of Persuasion, and Statistical Evidence’ [1986] 96 Yale Law Journal 376, 395. Cited with approval in Sienkiewicz v Greif [2011] UKSC 10, [2011] 2 AC 229, 285 (per Lord Rodger), 296 (per Lord Brown) and 302-03 (per Lord Dyson). It is Steve Gold’s position that when one talks about the “more likely than not” test, it refers mainly to the requisite strength of the fact-finder’s belief and not for the fact-finder to prove a fact probability greater than 50%. 30 Gary Chan Kok Yew, The Law of Torts in Singapore (2nd edn, Academy Publishing 2016) 07.002, 07.013 and 07.023. In short, the court must be satisfied that it is more likely than not that the defendant’s breach caused the loss. 31 Steve Hedley, ‘Negligence: Causation and Damage’ in Ken Oliphant (gen ed), The Law of Tort (2nd edn, LexisNexis Butterworths 2007) 784. See also Sarah Green, Causation in Negligence (Hart Publishing, 2015) 9. 32 JSI Shipping (S) Pte Ltd v Teofoongwonglcloong (a firm) [2007] SGCA 40, [2007] 4 SLR(R) 460 [141].

cause” for the loss must be demonstrated.33 After all, correlation of a breach of duty and the injury suffered does not equate to causation. In Barnett v Chelsea and Kensington Hospital Management Committee, it was found that notwithstanding the failure of the doctor to treat the plaintiff, the patient would have died. Therefore, the doctor’s negligence did not cause the death of the plaintiff.34 Likewise in Yeo Peng Hock v Pai Lily,35 the issue before the Singapore Court of Appeal was whether the plaintiff would have gone to a hospital, had she been properly advised by the defendant doctor, and whether her eyesight would have been saved.36 Ultimately, the Court found that while the plaintiff would have gone to the hospital immediately, had she been properly advised by the defendant,37 the plaintiff did not discharge her burden of proof in proving that the defendant’s negligence had caused or materially contributed to the loss of her vision.38 In this case, the defendant’s negligence did not effectively cause the harm to materialize since there could have been other factors that could have equally gave rise to the plaintiff ’s injury. Therefore, negligence was not established. It is therefore clear that the requirement for both factual and legal causation in establishing the effective link between the defendant’s harm caused and the injury suffered by the plaintiff is essential before a claim in negligence can be established.39 33 Gary Chan Kok Yew (n 30) 07.023. See also JSI Shipping (S) Pte Ltd (n 33) [141] (citing Anthony M Dugdale & K M Stanton, Professional Negligence (3rd edn, Butterworths 1998) 18.01; Chitty on Contracts (29th edn, Sweet & Maxwell 2004) 26029; Sarah Green (n 30) 15; Gemma Turton (n 10) 34). 34 Barnett v Chelsea and Kensington Hospital Management Committee [1969] 1 QB 428, [1968] 2 WLR 422. See also Vasuhi d/o Ramasamypillai v Tan Tock Seng Hospital Pte Ltd [2001] SGHC 45, [2011] 1 SLR(R) 303 [52] and Tong Seok May Joanne v Yau Hok Man Gordon [2012] SGHC 252, [2013] 2 SLR 18 [170]. 35 Yeo Peng Hock v Pai Lily [2001] 3 SLR(R) 555. 36 Yeo Peng Hock (n 35) [19] and [20]. 37 Yeo Peng Hock (n 35) [20]. 38 Yeo Peng Hock (n 35) [52]. 39 Gary Chan Kok Yew (n 30) 07.017. This is subjected to the fulfilment of the last requirement of remoteness, which is not included in the scope of this paper.

ARTICLES: WRITTEN MOOT 1

B. Limitations of the traditional test However, certain factual scenarios preclude establishing causation of damage, despite a finding of a breach of duty.40 There are cases where it is impossible to demonstrate that either (or both) factual and legal causation was fulfilled, either because the evidence does not show any effective links between the negligent act and the injury suffered and that it was a mere coincidence,41 or because evidential uncertainty precludes the courts from drawing any links between the negligent act and the damage suffered.42 Nevertheless, in cases where fairness demands that the injuries suffered must be compensated because there has been a correlating breach, courts are willing to depart from the traditional principles of causation and create a modified test to allow the plaintiff ’s claim to succeed and for liability to be established. In Chester v Afshar (“Chester”),43 the defendant doctor failed to properly advise the plaintiff regarding the risk of surgical procedures. Notwithstanding the doctor’s competence in executing the medical procedure, the risk materialised. Although the plaintiff could not show that she would have refused the surgery had she been aware of the risk,44 a bare majority of the House of Lords held that the defendant 40 Gary Chan Kok Yew (n 30) 07.023. 41 See for example cases such as Wilsher v Essex Area Health Authority, [1987] QB 730, [1998] AC 1074, Hotson v East Berkshire Area Health Authority [1988] UKHL 1, [1987] AC 750 and Chester v Afshar [2004] UKHL 41, [2004] 3 WLR 927 (Chester will be discussed below). 42 Gary Chan Kok Yew (n 30) 07.026 and 07.027. See also Steve Hedley, ‘Negligence: Causation and Damage’ in Ken Oliphant (gen ed) The Law of Torts (2nd edn, LexisNexis Butterworths 2007) 786 and 787. In the case of evidential uncertainties precluding establishing causation, see cases such as Fairchild v Glenhaven Funeral Services Ltd [2002[ UKHL 22, [2002] 3 WLR 89, Barker v Corus [2006] UKHL 20, [2006] 2 WLR 1027, Bonnington Castings Ltd v Wardlaw [1956] UKHL 1, [1956] 2 WLR 707 and McGhee v National Coal Board 1973] 1 WLR 1 (HL). 43 Chester v Afshar [2005] 1 AC 134 44 Instead of arguing that she would have gone on to see another doctor to perform the operation or not have done the operation at all, the plaintiff merely said that she would have

was liable for his failure to warn her of the risk. Even though factual causation was satisfied,45 legal causation was not.46 The majority relied heavily on policy factors concerning patient autonomy in modifying the requirements of the traditional causation test.47 Chester therefore exemplifies an instance where the UK court was motivated by policy reasons to recognise an exception in the traditional test of causation.48 delayed the operation to seek advice on any alternatives. 45 See Tamsyn Clark and Donal Nolan, ‘A Critique of Chester v Afshar’ [2014] 34(4) Oxford Journal of Legal Studies 659, 662-663. While it has been debated whether factual causation was not proven, Professors Clark and Nolan argued against both the dissenting judges (Lord Bingham in particular) and other academics that this was not the case, since the factual causation stage is concerned with the historical connection between the breach of duty and the physical injury suffered by the plaintiff and not the plaintiff undergoing surgery. Therefore, it was clear that the “but for” test was fulfilled since the failure of the defendant doctor to warn the plaintiff of the risk related to the surgery has led to the patient undergoing the surgical procedure and for the risk to materialise. The legal question to be answered was whether the doctor’s failure to warn was an effective cause of the injury. 46 Chester (n 43) (Lord Hope) [61] and [81]. Lord Hope had pointed out that the defendant doctor’s negligence was not the effective cause of the injury, since “to expose someone to a risk to which that person is exposed anyhow is not to cause anything”. 47 See for example Lara Khoury, ‘Chester v Afshar: Stepping Further Away from Causation’ [2005] Singapore Journal of Legal Studies 246, 248-251. Professor Khoury pointed out that the plaintiff ’s claim was largely justified based on policy reasons, as the majority Law Lords were in agreement that, based on the traditional principles of causation, it was not demonstrated in this case, since it was found that even if the risk had been disclosed, the patient could have undertaken the same procedure but at a later time, which would still expose the plaintiff to the same amount of risks. Nevertheless, it was the majority Law Lord’s position that legal causation should give way to legal policy and that the “normal approach to causation be modified”: See Chester (n 42) (Lord Hope)[85] and [86]. 48 See Jane Stapleton, ‘Occam’s Razor Reveals an Orthodox Basis for Chester v Afshar’ [2006] 122 LQR 426, 438439. Professor Stapleton first argues that coincidental outcome refers to consequences that are connected to the breach but does not increase in probability of happening on the occurrence of breach. Therefore, she argues that the damage suffered by the plaintiff in Chester was not entirely coincidental, since the negligent failure to warn has the potential affect, to a certain degree, the patient’s participation in the treatment in question, because some patients would refuse the treatment if warned, and hence that it could affect the probability in which the risk eventuates; cf. Clark and Nolan (n 46) 672. Professors Clark and Nolan that coincidental damage is properly understood as damage which does not represent the materialization of one of the risks that made the defendant’s conduct negligent and therefore, it is clear that the risk in this case was incidental to the defendant’s conduct and hence coincidental.

127

ARTICLES: WRITTEN MOOT 1

In Fairchild v Glenhaven Funeral Services Ltd,49 the plaintiffs contracted mesothelioma through exposure to asbestos dust while working for multiple employers. It was not possible to determine which employer each plaintiff had been working for when the condition was triggered. In this case, it was neither possible to establish factual nor legal causation. The medical and scientific evidential uncertainty involved50 made it impossible to discern which employer had caused the disease. Nevertheless, the employers were held jointly and severally liable, on the basis that the defendants had each contributed materially to the increased risks of harm. Again, the traditional rule of causation was modified to uphold fairness with various policy reasons submitted to justify this change.51

128

this is truly the case turns on our understanding of the basis for a lost chance claim, which will be explored in the next part. IV. The validity of the basis for justifying a loss of chance claim Given that a loss of chance claim in the context of economic loss is recognized in law, should the same principles be applicable in the context of physical harm? A. The legal basis for loss of chance

In Armstrong, the evidential uncertainty of medical science precluded the court from establishing whether it was the defendant’s negligence or the plaintiff ’s disease that was the effective cause of death. It appears that the justice of the case is a strong reason for introducing a loss of chance claim to circumvent traditional causation rules and impose liability for the breach. Justice Choo would thus appear to be correct in his decision. Yet, one should not jump to conclusions too quickly, since whether

The doctrine of loss of chance was first espoused in Chaplin v Hicks,52 invoked as a form of loss when assessing damages, upon breach of contract. The basis for this was because the “object and scope of the contract was to give the plaintiff the chance of being selected a prize winner.”53 McGregor suggests the focus is on the defendant’s duty to provide the plaintiff the chance, the breach of which has deprived the plaintiff of a chance or opportunity to secure a favourable outcome.54 Instead of dealing with past facts, one deals with a hypothetical event of the future, where the uncertainty lies in determining whether this outcome could have materialized at all.55

49 Fairchild v Glenhaven Funeral Services Ltd [2002] 3 WLR 89 50 It was found from the medical evidence on the aetiology of mesothelioma that it did not necessarily build up gradually but might be triggered suddenly, through one individual asbestos fibre causing a single cell in the lung to become malignant, or possibly by several such fibres. Therefore, while each successive employer and moment of exposure to asbestos increased the risk of mesothelioma, none could actually be said to have been a contributory cause of the disease, neither could it be proven which employer(s) were effectively responsible for causing that fatal fibre(s). 51 Fairchild (n 49) [33], [43] and [61]. It was Lord Bingham’s opinion that the injustice of imposing liability on the negligent employer who may not have been the one to cause damage was not as significant as the injustice of denying a redress for the victim; Lord Nicholls held that policy reasons may constitute good reason for departing from the normal threshold ‘but for’ test where a court has to decide whether the difficulties of proof confronting the plaintiff justify taking this exceptional course; Lord Hoffmann identified several important factors including a duty to protect employees from the risk of incurring a particular disease and the scientific uncertainty behind the contracting of mesothelioma.

52 Chaplin v Hicks [1911] 2 KB 786. In this case, the plaintiff was a participant in a beauty contest who was unable to make it for her appointment where she would be assessed on her eligibility to be one of the winners, despite making it into the top 50 candidates with the greatest number of votes, and thus did not obtain an opportunity to have herself considered as a potential winner. On this basis, the plaintiff sued the defendant in contract for not giving her a reasonable opportunity to present herself for selection. While she could not show that she would have gone on to win the contest, the court held that since the average chance of each competitor was about one in four, the plaintiff was awarded 25% of the amount that she would have won, had she been selected. 53 Chaplin (n 52) at 795. 54 Harvey McGregor (n 2) 6. However, it is acknowledged that this reason by itself is insufficient to establish the legal basis for loss of chance as an actionable head of damage, and more has to be shown (which, unfortunately, is not the focus of this paper): See for example Jane Stapleton, ‘The gist of negligence: Part 1 Minimal actionable damage’ [1988] 104 Law Quarterly Review 213 and Donal Nolan, ‘Damage in the English Law of Negligence’ [2013] 4(3) JETL 259. 55 Andrew Burrows (n 2) 33.

ARTICLES: WRITTEN MOOT 1

1. Loss of chance in economic loss context Allied Maples Group Ltd v Simmons & Simmon56 held that where the defendant’s negligence results in the loss of a real or substantial chance of a financial benefit or avoidance of a financial loss that would have been conferred by a third party, the plaintiff merely has to prove that the third party might have acted in the manner, as opposed to on a balance of probabilities.57 The injury suffered is the loss of a hypothetical outcome, which can no longer be realised as the chance for it occurring has been extinguished by the breach.58 A plausible explanation for why the law recognises a loss chance in this context is the unpredictability of third party actions. No scientific evidence can measure the probability of an outcome attributed to a third party’s conduct. Therefore, it would have been impossible to prove that had the defendant not been negligent, let alone on a balance of probabilities. As Lord Diplock held in assessing damages concerning a lost chance claim, the court must make an estimate as to what are the chances that a particular thing will or would have happened…59 56 Allied Maples Group Ltd (n 3). 57 Allied Maples Group Ltd (n 3) 1610, 1611 and 1614. It was Stuart-Smith LJ’s position that where the defendant’s negligence consist of an omission, the inquiry, although based on a hypothetical counterfactual as to what the third party would had done if there had been no negligence, is nevertheless one that would have to be proved, on a balance of probabilities and from an inferential process taking into account all the circumstances of the case. Therefore, the plaintiff still has to prove, on a balance of probabilities, that he would have taken the benefit or avoided the risk. On the other hand, where the loss dependent on the hypothetical actions of a third party, then the plaintiff is only required to demonstrate the existence of a real or substantial chance that the third party would have acted so as to confer the benefit or avoid the risk to the plaintiff. The plaintiff does not have to prove, on a balance of probabilities, that the third party would have acted as such. See also Asia Hotel Investments Ltd v Starwood Asia Pacific Management Pte Ltd [2005] 1 SLR(R) 661 (SGCA) [47] where Stuart-Smith LJ’s approach towards loss of chance claim in the context of pure economic loss was endorsed in Singapore. 58 Sarah Green (n 30) 157. 59 Mallet v McMonagle [1970] AC 166 (HL) at 176 (Lord Diplock). See also Davis v Taylor [1874] AC 207 (HL) at 213 (Lord Reid): “You can prove that a past event happened, but you cannot prove that a future event will happen and I do not think that the law is so foolish as to suppose that you can.

Therefore, it would be impossible for the plaintiff to prove the probable actions of a third party, and a true evidential gap is recognised.60 Recognising a lost chance is the only way to bridge this evidential gap. 2. Loss of chance in physical harm context One would expect the same principles to apply in medical negligence cases which cause the worsening of the plaintiff ’s physical condition or his death. After all, the uncertainty surrounding the plaintiff ’s claim is similar to that surrounding third party actions in that both prevent an understanding of the true causal process for the loss. In Armstrong, the question of whether Mr Traynor would have fallen in the category of surviving for another 10 years or not, is prima facie unable to be determined, notwithstanding the advancement of medical science. This is notwithstanding that Mr Traynor’s death may still occur even with prompt treatment due to the cancer that was unstoppable at that instant. However, Helen Reece would beg to differ. In her seminal article that has been cited with approval in Gregg,61 Reece argues that the distinction lies in understanding the causal process between an economic loss dependent on a third party’s action and a physical injury that is dependent on either the defendant or the plaintiff.62 On one hand, the former is “indeterministic”, meaning it occurs randomly and cannot be predicted even with unlimited knowledge of the past. On the other hand, the latter is “deterministic”, meaning given sufficient knowledge of the past, a “hypothetical chain of events is fully determined by the events which have occurred” and the cause of anything can be discovered.63 Reece further points out that in All that you can do is to evaluate the chance. Sometimes it is virtually 100 per cent; sometimes virtually nil. But often it is somewhere in between.” 60 Gemma Turton (n 11) 143. 61 Helen Reece, Losses of Chances in the Law (1996) The Modern Law Review 188. 62 Helen Reece, Losses of Chances in the Law (1996) The Modern Law Review 188. 63 Helen Reece (n 61) 192.

129

ARTICLES: WRITTEN MOOT 1

indeterminism (or quasi-indeterminism),64 the likelihood of a future event occurring has an objective probability that is part an external operation beyond our knowledge.65 In contrast, determinism entails no objective probability of an occurrence. The occurrence of anything deterministic can be predicted based on our limited knowledge available. Reece describes this as an epistemological probability.66 She then provides examples to contrast an example of a deterministic event67 with that of an indeterministic event68 to demonstrate that the balance of probabilities approach cannot be practically nor fairly applied in cases of true indeterminacy or quasi-indeterminacy, and cases such as Chaplin and Allied Maples, where the outcome depends on third-party human conduct, can be considered as valid loss of chance claims under this class due to a true evidential gap that would have precluded the

130

64 Helen Reece (n 61) 193. Professor Reece acknowledges that there are instances where it is difficult or impossible to know whether an event is truly indeterministic, since our inability to predict the occurrence may be either because the event is truly unpredictable, or because we have not yet found the complete set of necessary and sufficient causes. A good example would be Quantum Mechanics and various associated phenomenon such as Heisenberg’s Uncertainty Principle or Schrödinger’s Box. To overcome this, Professor Reece’s definition of an indeterministic event is where if we could not have been predicted at any point in the past, we cannot be predicted in the present even given unlimited time, resources and evidence, and we cannot imagine how it would become predictable in the future, even given the success of present research programmes. Professor Reece refers to this type of event as a quasi-indeterministic event, and the chance as a quasi-objective chance. 65 Helen Reece (n 61) 193. 66 Helen Reece (n 61) 193. 67 Helen Reece (n 61) 194-95. Professor Reece gives the example about a road accident where a motorist injures a pedestrian to illustrate a deterministic event. It was stated that the question of whether the pedestrian would not have suffered the injury without the negligence was deterministic. The pedestrian’s bones break because a force is applied to them, and bones do not break in the same way on a random basis. 68 Helen Reece (n 61) 195. Professor Reece gives an example about the decay of a Uranium atom to demonstrate indeterminism. Uranium decays into Thorium at a random rate (termed a stochastic process), and according to quantum theory, it is impossible to predict when a particular atom will decay, regardless of how long the atom has existed. However, for a collection of Uranium atoms, the expected decay rate is measured by the decay constant within an exponential distribution equation, otherwise known as half-lives.

understanding of the causal process.69 On the other hand, medical negligence cases such as Hotson v East Berkshire Area Health Authority are considered deterministic,70 since it was possible for the condition contributing to the cause of the injury suffered by the plaintiff to have been discovered in the past, which helps determine the true cause of the plaintiff ’s injury.71 The inability to determine this at trial was merely an epistemic probability related to an issue of proof and thus an evidential difficulty. As far as the causal process is concerned, there is no uncertainty. In summary, where a breach relates to an uncertain hypothetical outcome dependent upon a third party’s conduct, that outcome is an objective probability that is indeterministic and gives rise to a true evidential gap. On the other hand, where the breach relates to the existence of an uncertain past fact, that past fact would have already set in motion the chain of events that will occur. The only problem here is that of evidential difficulty. Given that Reece’s position received judicial endorsement and application in Gregg, it is pertinent to revisit this case and understand why a loss of chance claim was rejected from the perspective of causation. B. Loss of chance - Gregg v Scott In Gregg,72 the appellant, Mr Gregg, suffered from a rare kind of non-Hodgkin’s lymphoma where he had a 42% chance of surviving for 10 years if treated promptly. However, the respondent, Dr Scott, misdiagnosed Mr Gregg’s condition as benign, resulting in a nine-month treatment delay, leading to the spread of cancer in Mr Gregg which diminished his recovery prospects to 25% at the date of trial. The issue was whether Dr Scott’s negligence had caused Mr Gregg to suffer a premature death by losing 10 years of his lifespan. 69 Helen Reece (n 61) 197-98 and 204-05. See also Gemma Turton (n 10) 143. 70 Hotson v East Berkshire Area Health Authority [1987] AC 750 (HL). 71 Helen Reece (n 61) 196. 72 Gregg (n 4).

ARTICLES: WRITTEN MOOT 1

At trial, the judge found that, if treated promptly, Mr Gregg’s initial treatment would have achieved remission without an immediate need for more intensive treatment, since this would have prevented the spreading of the cancer.73 Nevertheless, it was not possible to conclude that, on the balance of probability, Mr Gregg would have been able to survive for an additional 10 years,74 since the chance of recovery would still have been less than 50%. In short, Mr Gregg could not prove that a better outcome could happen, had Dr Scott not been negligent. One of the arguments employed by Mr Gregg’s counsel was that the reduction in Mr Gregg’s chances of survival was itself an actionable head of damage,75 and was employed to circumvent the inability of the plaintiff counsel to prove causation given the uncertainty in the statistics of the medical evidence. In Gregg therefore, the basis for which loss of chance was employed was similar to that in Armstrong – in both cases, the supposed inability to prove causation as a result of evidential gaps prompts both counsel and judge to reformulate the injury into a loss of chance claim. However, Lord Hoffmann held that the lack of indeterminacy of the evidence on the facts precludes the recognition of a loss of chance claim to circumvent the general rules of causation. His Lordship held that: the law regards the world as in principle bound by laws of causality. Everything has a determinate cause, even if we do not know what it is. … The fact that proof is rendered difficult or impossible … makes no difference. There is no inherent uncertainty about what caused something to happen in the past or about whether something which happened in 73 Gregg (n 4) [105]. 74 Gregg (n 4) [105] – [106]: This was because Mr Gregg already had a less than evens chance (at 42%) of achieving an additional 10 year lifespan with proper treatment. The delay, therefore, merely reduced the chance by half instead of extinguishing it. 75 Gregg (n 4) [66].

the past will cause something to happen in the future. Everything is determined by causality. What we lack is knowledge and the law deals with lack of knowledge by the concept of the burden of proof.76 The uncertainty surrounding the medical evidence which precludes any accurate judgment on the causal connection between Dr Scott’s negligence and the loss of opportunity suffered by Mr Gregg’s is not, as Reece terms it, an indeterministic objective probability that gives rise to a true evidential gap.77 What caused Mr Gregg’s loss was a fact of the past (either the cancer or the negligent diagnosis, both of which were relevant causes) that set in motion the outcome of whether Mr Gregg lived or died. The only problem relates to the difficulty of proof. Given the limitations of human knowledge regarding the aetiology of cancer, the only issue relates to the fact that the plaintiff is unable to prove his case, not because it was truly unprovable, but the evidence was difficult to gather. Therefore, unlike economic loss cases where there was a genuine extinguishment of a chance, the fact that no third-parties exist in this case means that it would still be incumbent on Mr Gregg to prove causation on a balance of probabilities. A counter-argument against the distinction made could be made, namely that whether Mr Gregg would have then opted for the treatment or alternative forms of treatment with different probability of success is also an exercise of freewill. Moreover, the presence of side-effects cannot be predicted such as to constitute indeterminacy.78 This does not detract from the argument that a line has to be drawn between 76 Gregg (n 4) [79]. 77 Helen Reece (n 61) 194. 78 It is worth mentioning that besides the initial treatment that Mr Gregg would have undergone, it may also perhaps be possible for Mr Gregg to undertake different kinds of treatment or take any alternative actions such as getting a second opinion. As provided by the National Cancer Institute’s website, there are 9 different types of standard treatment for

131

ARTICLES: WRITTEN MOOT 1

a situation where a likely outcome is truly uncertain in itself and therefore impossible of being proven, because of the impossibility of obtaining evidence, versus an outcome that is merely more difficult to prove due to insufficient evidence. In this case, statistical data, if altered to be more precise in its composition, could produce more persuasive results.79 Instead, Lord Hoffmann acknowledged the apparently arbitrary distinction between requiring a plaintiff to prove an outcome that is dependent on himself or the defendant on a balance of probability, versus what a third party would have done on a matter of chance, which can only be justified on grounds of policy. The inability to prove that a third party would have done something does not mean that it should not be proved that the third party was more likely than not to have done it.80 However, McGregor argues that the distinction is not arbitrary since the onus is always on the plaintiff to prove his case and therefore his or the defendant’s actions, or both. On the other hand, there is no such onus on third parties.81

132

patients with non-Hodgkin lymphoma, as well as new types of treatment tested in clinical trials. Since Mr Greg is suffering from ALK Negative Anaplastic large cell lymphoma, which may relapse despite being responsive to chemotherapy treatment, it may be possible for Mr Gregg to have opted for other forms treatment if diagnosed and advised. It is also noted that different side effects may following these treatments, some of which may include fatal side effects such as a second cancer. Taking all these into consideration, the permutations of outcomes are endless as to whether Mr Gregg ultimately lives another 10 years or die, and can be said to be indeterministic: see Adult Non-Hodgkin Lymphoma Treatment (PDQ®)–Patient Version (National Cancer Institute) <https://www.cancer.gov/types/lymphoma/ patient/adult-nhl-treatment-pdq#_190> accessed 19th June 2019; Understanding Non-Hodgkin Lymphoma (Lymphoma Research Institute, July 2018) <https://www.lymphoma.org/ wp-content/uploads/2018/11/NHL-Booklet-2019.pdf> accessed 19th June 2019. 79 Gemma Turton (n 11) 125. 80 Gregg (n 4) [85]. One of the major precedent case to be overruled would be Hotson v East Berkshire Area Health Authority [1987] AC 750 (HL), where the House of Lords had made it clear that it would not recognise any claims for loss of chance nor modify the traditional rules of causation where the plaintiff is unable to satisfy the balance of probability test. Another would have been Wilsher v Essex Area health authority [1988] AC 1074. 81 Harvey McGregor (n 1) 7. See also Andrew Burrows

More fundamentally, to prove something on a balance of probabilities presupposes that there is a degree of probability as to the occurrence of that action. As Professor Reece has argued, not everything can be proved on probabilities, and hypothetical future human actions cannot be engaged in a question of epistemological probabilities, but only on possibilities or chance.82 C. Determinacy of evidence in Armstrong and Gregg In Armstrong, the loss of a “fighting chance” suffered by Mr Traynor cannot be determined by Choo J to have been caused by Dr Tan’s breach, since it was claimed that medical uncertainty precluded the court from establishing whether Mr Traynor had indeed fallen in the 68-60% category where he would have an actual chance of living for another 10 years. This gave impetus for Justice Choo’s recognition of the lost chance doctrine on the basis of perceived uncertainty in evidence that precluded the establishing of a causal nexus between the breach of duty and the death of Mr Traynor. However, it is now clear that there is no basis for a loss of chance claim in Armstrong since the plaintiff ’s failure to establish liability was not due to any evidential gaps arising from an indeterministic causal process, but rather merely an issue of evidential difficulty with a deterministic cause. What made the deterministic nature stronger was that Mr Traynor was already dead before the action was brought whereas Mr Gregg was still alive at the time his case was heard before the House of Lords. In this case, the possibility in Armstrong had already materialized – Mr Traynor had fallen in the category where he did not have a chance for a 10-year survival, and thus Dr Tan should arguably be held negligent for reducing Mr Traynor’s survival prospects. On the other hand, evidence before the House of Lords indicated that Mr Gregg’s probability (n 1) 36. Professor Burrows argues that since third parties may not be before the court, there is generally speaking greater uncertainty as regards the hypothetical actions of third parties so that a balance of probabilities approach seems less appropriate. 82 Helen Reece (n 61) 197-8.

ARTICLES: WRITTEN MOOT 1

for surviving 10 years was very high, and as Lord Phillips noted, the closer Mr Gregg came to being a survivor, the smaller the likelihood of delay in his treatment had actually affected his expectation of survival.83 However, this fact would also mean that Gregg was a deterministic case as well – Mr Gregg simply appears to have, more likely than not, fallen into the survivor category.

resorting to public policy arguments, however, is a reality best expressed by Justice Burrough’s protest: “it is a very unruly horse, and once you get astride it you never know where it will carry you’.86 Bearing this in mind, we now turn to examining the validity of policy reasons in recognising a lost chance claim in the context of medical negligence giving rise to physical injury.

As such, the legal basis for recognising a loss of chance claim in economic loss contexts are not transferrable to physical injury contexts. This warrants examining the validity of reasoning advanced in Armstrong. This means scrutinising the policy reasons which justify to recognising the loss of chance doctrine, addressed in the next section.

A. Policy reasons in Gregg

V. Policy reasons as an “unruly horse” Lady Hale in her recent speech explored whether doctrine or policy should guide such developments.84 Particularly in the realm of tort, judges frequently have recourse to public policy to apply or develop the law,85 when faced with a novel situation where established principles do not provide an answer. The problem of 83 Gregg (n 4) [169]. 84 Brenda Hale, President of the UK Supreme Court, ‘Principle and Pragmatism in Developing Private Law’ (Speech at the Cambridge Freshfields Lecture, Cambridge, 7 May 2019) <https://www.supremecourt.uk/docs/speech-190307.pdf> accessed 22 June 2019 . 85 See for example the applicable law for establishing a duty of care in the law of negligence, as encapsulated by the “fair, just and reasonable” stage of the UK Caparo test in Caparo Industries plc v Dickman [1990] 2 AC 605 (HL) and the “policy considerations” stage of the Singapore Spandeck test in Spandeck Engineering (S) Pte Ltd v Defence Science & Technology Agency [2007] 4 SLR(R) 100 (SGCA), where the judicial inquiry in ascertaining whether a duty of care should be imposed on the parties are encapsulated by policy considerations; see also the wrongful birth cases such as in UK, where Rees v Darlington Memorial Hospital NHS Trust [2004] 1 AC 309 (HL) saw the application of public policy to overrule the Court of Appeal decision and to deny the plaintiff ’s claim for upkeep costs, and to devise a new remedy in recognition of the policy of individual autonomy, and also in Singapore’s context, ACB v Thomson Medical Pte Ltd and others [2017] SLR 918 (SGCA), where upkeep costs as a head of damage was no recognised in Singapore for policy reasons, but the court then recognised a loss of “genetic affinity” head of damage on the basis of alternative policy reason.

In Gregg, both the majority and minority judges resorted to various policy reasons in buttressing their reasoning. These will be analysed in turn. 1. The minority in Gregg Lord Nicholls’ arguments, which Lord Hope agreed with,87 were similar to those of Justice Choo in that they focused on achieving justice, noting that “ the loss of a 45% prospect of recovery is just as much a real loss for a patient as the loss of a 55% prospect of recovery” Thus to deny a claim here would render the doctor’s duty a “hollow” one.88 Most pertinently, “justice requires that”, just as where the law has recognised as actionable damage a professional’s negligence that resulting in a loss of an economic opportunity, a doctor’s negligence which resulted in the chance of health or life should also constitute actionable damage.89 In short, fairness and justice was the primary basis on which he sought to counter the principled arguments that precludes recognition of a loss of chance claim. 2. The majority in Gregg The main policy concern for Lord Hoffmann was that recognizing the lost chance doctrine would have “enormous consequences for insurance companies and the National Health Service” which is arguably Parliament’s 86 87 88 89

Richardson (n 12) 252. Gregg (n 4) [92]. Gregg (n 4) [4]. Gregg (n 4) [25].

133

ARTICLES: WRITTEN MOOT 1

prerogative.90 On the other hand, Baroness Hale’s arguments were formulated in response to Mr Gregg’s “policy approach” and lend support for the distinction. Firstly, allowing for a loss of chance approach would entail finding liability even if the harm had a less than even chance of occurring. This disincentivises medical practitioners to take proper care of their patients, which is antithetical to their professional ethos.91 Secondly, Baroness Hale’s stronger argument is that where the plaintiff will be able to reformulate any claim for loss of outcome as a loss of chance of that outcome, plaintiffs would in any case be able obtain full damages where he is able to prove the causal relationship between breach and loss of an outcome, but also a proportionately lesser amount where the causal relationship established is between breach and loss of chance.92 This is a policy factor that weighs heavily in the plaintiff ’s favour without giving due regard to the defendant. This introduces unwarranted uncertainty in personal injury claims.93 Finally, Lord Phillips’ highlighted the technical difficulties and uncertainties in considering medical evidence based on clinical statistics,94

134

90 Gregg (n 4) [90]. 91 Gregg (n 4) [216]. 92 Gregg (n 4) [224]. Nevertheless, even if damages were apportioned according to the chance lost, Baroness Hale also points out that this would make the entire area of damages less predictable by redefining quantification of recovery in almost every single case, which would cause far more problems than the policy benefits are worth. 93 Gregg (n 4) [233] and [225]. 94 Gregg (n 4) [147]. The key to Mr Gregg’s arguments that related to expert evidence submitted by one Professor Goldstone regarding the statistical chances of Mr Gregg’s survivability had he received the treatment earlier. In short, Professor Goldman’s model shows that out of a group of 100 patients suffering from the same disease as Mr Gregg, only 42 would survive. What was also important to note was that for patients who received initial treatment but did not achieve complete remission, 4 will be brought for further treatment to achieve complete remission. After the initial complete remission, for those who relapse, responded to further treatment and then relapsed again, only 1 will survive out of 6 patients. However, Lord Phillips was of the opinion that certain fallacies inherent in the statistical mode made it an inaccurate on. What Lord Phillips first pointed out was that the assumption made by the judges

which he claimed was a “policy-factor” that weighs against the introduction of a loss of chance claim,95 a view shared by Baroness Hale.96 The inherent evidential difficulties, in drawing of logical deductions from statistical data compounded the practical difficulties of determining the true causal relationship between the breach and the loss of chance.97 A “robust test that produces rough justice may be preferable to a test that on occasion will be difficult, if not impossible, to apply with confidence in practice”.98 B. The adequacy of policy reasoning in Armstrong While Choo J’s focus on the policy reason of ensuring that justice be done is admirable, it is with the greatest of respect, insufficient to persuade convincingly why loss of chance should be recognised. This leaves us wanting of greater judicial insight as to why fairness offers a compelling reason to circumvent causation and the recognition of lost chance as actionable damage. Attempting to distinguish Gregg on the facts,99 without acknowledging and addressing below (that the statistical evidence comprises patients who are at the same stage of the disease as Mr Gregg) was questionable, since it is medical knowledge that the duration and intensity of treatment, as well as the probability of survival differs between different stages of cancer, and Professor Goldstone’s model comprises persons with “unrevealed personal characteristics… [such as] the stage of the disease at diagnosis”. Secondly, it was also pointed out that the statistical evidence was a mere reflection of the outcome of the 100 patients that was present in the model and did not provide an accurate assessment of the kinds of factors that would determine the likelihood of attaining a specific outcome by the patient. It was therefore not possible to conclude that Mr Gregg would have achieved initial complete remission had he received prompt treatment because there could have been some other factors that would have prevented Mr Gregg from achieving complete remission (and thus fall into the category of persons that would require further treatment notwithstanding prompt treatment). 95 Gregg (n 4) [170]. 96 Gregg (n 4) [214]. 97 Gregg (n 4) [170]. 98 Gregg (n 4) [170]. 99 Armstrong (n 6) [16]. Choo J pointed out, as part of his reasons for not following the UK Supreme Court decisions of Gregg v Scott, that the prognosis for non-Hodgkins lymphoma (which was the disease suffered by the plaintiff in Gregg) is not the same as melanoma (which was Mr Traynor’s condition) and further, that since the case dealt with Peter Traynor and not Malcolm Gregg, the difference in the individual’s physical traits ought to be taken into account. As for the Australian High

ARTICLES: WRITTEN MOOT 1

the strong reasons provided by the respective judges against the introduction of loss of chance claims in physical injury contexts, does little to offer persuasive arguments as to why Singapore’s approach should differ from the UK’s position. This may result in greater uncertainty in the realm of medical practice and negligence law whilst engendering increased medical litigation and defensive medical practice that is antithetical to doctor-patient relationships.100

inadequate expression of public policy as a firm basis in recognising lost chance claims would result in uncertainty in the law of negligence, principles of causation, and detrimental effects to the community. Further, given the nature of policy, it would perhaps be best left to the legislature to conduct a cost-benefit analysis in considering the introduction of a lost chance claim, given that they are a democratic and far more institutionally competent branch of government.

In the words of Lord Kerr:

VI. Conclusion

“…a decision which is not the product of, in the words of Lord Oliver, “any logical process of analogical deduction” holds less sway, particularly if it does not accord with what the subsequent decisionmaker considers to be the correct instinctive reaction to contemporaneous standards and conditions…”101 Given the inherent weaknesses in policy arguments, more ought to be done if the loss of chance claim ought to be recognised on the policy grounds of fairness. As aptly put by Menon CJ, “Public policy arguments … all require the court to have regard not only to the interests of the parties to the dispute, but also to those of the community at large.”102 Therefore, an Court’s position as regards loss of chance in Tabett v Gett. 100 See for example Kumaralingam Amirthalingam, ‘Commentary: Are complaints against doctors destroying the doctor-patient relationship?’ (Channel News Asia Online, 28 February 2019) <https://www.channelnewsasia.com/news/ commentary/commentary-are-complaints-against-doctorsdestroying-the-doctor-11292942> accessed 19 June 2019; Wong Chiang Yin, ‘The toxic environment that breeds defensive medicine’ (The Straits Times, 12 April 2019) < https:// www.straitstimes.com/opinion/the-toxic-environment-thatbreeds-defensive-medicine> accessed 19 June 2019; Kelly Ng, ‘Concerns in Parliament over ‘defensive medicine’ after doctor’s suspension’ (TODAY Online, 02 August 2017) <https:// www.todayonline.com/singapore/concerns-parliament-overdefensive-medicine-after-doctors-suspension> accessed 19 June 2019; Dr Desmond Wai, ‘Fear drives doctors to practise defensive medicine’ (TODAY Online 27 August 2017) <https://www.todayonline.com/daily-focus/health/feardrives-doctors-practise-defensive-medicine> accessed 19 June 2019. 101 Michael v Chief Constable of South Wales [2015] AC 1352 (HL) [161]. 102 Sundaresh Menon, Chief Justice of the Singapore Supreme Court, ‘Taming The Unruly Horse: The Treatment Of

At first blush, the decision of the Singapore High Courts in Armstrong would certainly be a welcome relief for Mrs Armstrong and many in similar positions as the plaintiff. Nevertheless, a deeper insight into the realm of causation, the doctrine of lost chance, and policy reasons against recognising this claim, exposes the inadequacies in the theoretical, legal and policy bases respectively in permitting lost chance claims in medical negligence contexts. More fundamentally, Baroness Hale illuminates that the question whether loss of chance should be recognised assumes the truth of its premise – that a lost chance is a recognizable damage in the first place. This was why the first inquiry ought to be the concept of actionable damage.103 In this case, if a lost chance cannot be recognised as actionable damage, then it would be meaningless to even talk about other related issues such as causation and quantification.104 Public Policy Arguments In The Courts’ (Lecture at the High Court of Sabah and Sarawak, 19 February 2019) <https://www. supremecourt.gov.sg/Data/Editor/Documents/Public%20 Policy%20Lecture%20-%2019Feb2019.pdf> accessed 20 June 2019. Although it is acknowledged that Menon CJ’s speech was primarily on the constitutional issues of using public policy in the application of the law, it is recognised that they are equally applicable in all cases where public policy is invoked as judicial reasoning. 103 Gregg (n 4) [217] and [223]. 104 Because of the nature of this inquiry and the fact that this paper is, at the current moment, not capable of addressing this issue, the inquiry as to whether a loss of chance can be legally recognized as an actionable head of damage in the tort of negligence shall be left to a subsequent academic article to be dealt with.

135

ARTICLES: WRITTEN MOOT 1

Therefore, rather than analysing the recognition of lost chance claims from a perspective of causation, perhaps future courts can address whether lost chance claims can be, and ought to be recognised as a new head of actionable damage. After all, as Baroness Hale in Gregg and Justice of Appeal Andrew Phang I ACB v Thomson Medical rightly pointed out, damage is the gist of negligence.105 The sanctity of life is such that where illness befalls our loved ones, we turn to the medical profession for help. Where the medical profession fails us, it is only human to feel anguish and pain, wanting to seize every reason we can conjure to explain why the doctor has failed to provide an opportunity for a timely cure. Alas, the lawâ&#x20AC;&#x2122;s mechanism of requiring proof seeks to, in this context, strike a balance between upholding the rights of patients that has been done wrong with the need to ensure certainty and fairness for the medical profession â&#x20AC;&#x201C; where only those who have been proven to have done wrong ought to be punished. In the absence of proof, the law cannot provide a remedy. And thus, whether a missed opportunity truly defines our life can only be known with further reflection, not mere intuition.

136

105 Gregg (n 4) [217]. See also ACB v Thomson Medical Pte Ltd and others [2017] SLR 918 (SGCA) [47].

ARTICLES: WRITTEN MOOT 2

The United Kingdom’s deferred prosecution agreement regime five years on: is it an effective tool in addressing economic crime perpetrated by companies? Wee Toh Loo, University of Bristol

Introduction A company may perpetrate economic crime, like fraud and bribery, by being a “means by and through which” such misconduct is committed by its employees, directors, agents or other associated individuals.1 Concerned that such offending often went without redress in the United Kingdom (“UK”)2 but having regarded it to be effectively dealt with in the United States (“US”) through deferred prosecution agreements (“DPAs”),3 the UK Government proposed the introduction of DPAs here, specifically in England and Wales.4 This took effect when the Crime and Courts Act 2013 (“CCA”)5 came into force in February 2014,6 with the stated aim of providing prosecutors with an additional tool to address corporate economic crime.7 Accordingly, this essay will analyse the effectiveness of DPAs as such. Although the infancy of the UK DPA regime means that a definitive conclusion may not be reached, it will be argued that, in practice, its ability to effectively address corporate economic crime 1 Celia Wells, ‘Corporate Crime: Opening the Eyes of the Sentry’ (2010) 30 Legal Studies 370, 371; see also James Gobert and Maurice Punch, Rethinking Corporate Crime (CUP 2003) 11, 13. 2 Ministry of Justice, Consultation on a New Enforcement Tool to Deal with Economic Crime Committed by Commercial Organisations: Deferred Prosecution Agreements (Cm 8348, 2012) (DPA Consultation) para 2. 3 ibid paras 59, 69-70. 4 ibid para 5. 5 Crime and Courts Act 2013 (CCA), s 45 and Schedule 17. 6 Criminal Procedure (Amendment No. 2) Rules 2013, SI 2013/3183, r 2. 7 DPA Consultation (n 2) paras 5, 23; Ministry of Justice, Deferred Prosecution Agreements: Government Response to the Consultation on a New Enforcement Tool to Deal with Economic Crime Committed by Commercial Organisations (Cm 8463, 2012) (Government Response to the DPA Consultation) 3-4; Michael Bisgrove and Mark Weekes, ‘Deferred Prosecution Agreements: A Practical Consideration’ (2014) 6 Crim LR 416, 416.

is likely to be limited. Additionally, it will be posited that the UK DPA is even more unlikely to address such wrongdoing effectively because it is fundamentally constrained by its inherent characteristics. Part II briefly summarises the UK’s DPA regime before setting out how DPAs may, in theory, be an effective tool in tackling corporate economic crime. First, they may improve the ability of prosecutors to hold corporate offenders to account by allowing for enforcement flexibility, especially against large companies, as well as by incentivising selfreporting. Second, when a company enters into a DPA, prosecutors may be able to change its culture, through the deterrent effect of financial penalties and by enforcing compliance reforms, such that future wrongdoing is prevented. Part III evaluates these claims. It will be contended first that insofar as the DPA is said to address economic crime effectively because it provides for flexible enforcement, the necessity of this flexibility is questionable. It will then be submitted that the UK DPA as applied in reality is unlikely to effectively address corporate economic crime. First, the ability to hold corporate offenders accountable is potentially compromised because self-reporting is likely to be disincentivised, especially after the Rolls-Royce DPA.8 Second, the prospect of influencing corporate culture is limited because financial penalties are likely to be of inadequate deterrence. Even if they are accompanied by requirements to enhance compliance programmes, cultural change is unlikely to be realised as companies can and are likely to implement compliance reforms without substantively changing their underlying practices. Part IV advances the proposition that 8 Serious Fraud Office v Rolls-Royce Plc [2017] Southwark Crown Court Case No. U20170036, [2017] Lloyd’s Rep FC 249 (Rolls-Royce).

137

ARTICLES: WRITTEN MOOT 2

the UK DPA is even more unlikely to effectively address corporate economic crime because it is fundamentally limited in its ability to do so. First, its ability to enhance the accountability of corporate offenders intrinsically depends on the efficacy of prosecution, which at present is largely toothless in the UK. Further, as a mechanism targeted only at companies and other similar bodies, it is beset with legitimacy issues which hinder its ability to foster a genuinelyœ compliant corporate culture in companies which have entered into DPAs. Therefore, it will be concluded that the UK DPA is unlikely to be effective in addressing corporate economic crime. I. The DPA in theory A. The UK DPA regime To date, four DPAs have been entered into.9 A DPA in the UK is an agreement between a prosecutor – primarily the Serious Fraud Office (“SFO”) or Crown Prosecution Service10 – and a person under consideration for prosecution11 to suspend and eventually discontinue criminal proceedings,12 subject to the latter’s compliance with certain requirements.13 These may include paying a financial penalty14 and the implementation of robust compliance programmes.15 An invitation to negotiate a DPA is solely based on the prosecutor’s discretion,16 having satisfied itself as to the evidence of offending and the public interest of entering into a DPA instead of prosecuting.17 However, a DPA can only be entered into with companies

138

9 Serious Fraud Office, ‘Deferred Prosecution Agreements’ (Serious Fraud Office Guidance, Policy and Protocols) <https://www.sfo.gov.uk/publications/guidancepolicy-and-protocols/deferred-prosecution-agreements/> accessed 13 March 2019. 10 CCA (n 5), Schedule 17 para 3(1). 11 ibid, Schedule 17 para 1(1). 12 ibid, Schedule 17 paras 2(2), (3) and 11(1), (2). 13 ibid, Schedule 17 para 1(2). 14 Serious Fraud Office and Crown Prosecution Service, Deferred Prosecution Agreements Code of Practice: Crime and Courts Act 2013 (DPA Code of Practice) para 7.8. 15 ibid para 7.10.iii. 16 ibid paras 1.1, 2.1; CCA (n 5), Schedule 17 para 3(2). 17 ibid para 1.2.

or similar bodies, like partnerships,18 and is only in respect of “basically economic or financial crimes”.19 Additionally, a DPA must include a statement of facts concerning the alleged offence20 and specify when it expires if it has not been terminated beforehand for breach.21 The court must at a preliminary hearing, before terms are agreed, declare entering into a DPA “likely to be in the interests of justice” and its proposed terms “fair, reasonable and proportionate”.22 It must then, at a final hearing after terms have been agreed, declare the negotiated DPA to be in the “interests of justice” and its terms “fair, reasonable and proportionate”.23 Only then will the DPA come into force:24 the prosecutor will proceed with an indictment, which will thereafter be immediately suspended pending the satisfactory performance of the DPA.25 If satisfactorily performed and the expiry date is reached, the prosecutor has to then give the court notice that prosecution has been discontinued.26 Conversely, the courts are also involved in determining if the DPA has been breached,27 which may lead to its termination.28 If so, prosecution normally resumes.29 In theory, the DPA regime appears able to effectively address corporate economic crime because it improves prosecutors’ ability to hold errant companies to account and influence the culture of companies which have entered into 18 CCA (n 5), Schedule 17 para 4(1). 19 Federico Mazzacuva, ‘Justifications and Purposes of Negotiated Justice for Corporate Offenders: Deferred and Non-Prosecution Agreements in the UK and US Systems of Criminal Justice’ (2014) 78(3) JCL 249, 252; see ibid, Schedule 17 para 1(1) and Part 2. 20 CCA (n 5), Schedule 17 para 5(1). 21 ibid, Schedule 17 para 5(2). 22 CCA (n 5), Schedule 17 para 7(1). 23 ibid, Schedule 17 para 8(1). 24 ibid, Schedule 17 para 8(3). 25 DPA Code of Practice (n 14) para 1.6. 26 CCA (n 5), Schedule 17 para 11(1). 27 ibid, Schedule 17 para 9(1), (2). 28 ibid, Schedule 17 para 9(3)(b). 29 DPA Code of Practice (n 14) para 12.7.

ARTICLES: WRITTEN MOOT 2

DPAs.30 These aspects will be considered in turn. B. Improving accountability The DPA has been suggested to allow prosecutors to hold corporate offenders to account in circumstances where these companies would otherwise escape enforcement action,31 thereby effectively addressing corporate economic crime. This is due to the fact that as an additional enforcement tool, it seems to provide prosecutors with flexibility in terms of enforcement options, thus filling purported enforcement gaps.32 Further, it may incentivise self-reporting, which facilitates the detection and investigation of such wrongdoing.33 i. Enforcement flexibility By entering into DPAs, prosecutors can hold corporate offenders, especially large companies, accountable when a successful prosecution is not possible, or where prosecution itself is inappropriate or subject to time and resource constraints.34 This is because DPAs ensure that enforcement action with an arguably similar result as a successful prosecution can still be taken – an outcome satisfying similar purposes as a sentence following conviction35 and the “acceptance of the existence of [material] facts” constituting the alleged offence, through the statement of facts,36 can both be secured.

30 Bisgrove and Weekes (n 7) 417. 31 DPA Consultation (n 2) paras 3, 5, 7 and 30; Government Response to the DPA Consultation (n 7) para 6; Bisgrove and Weekes (n 7) 417; Constantino Grasso, ‘Peaks and troughs of the English deferred prosecution agreement: the lesson learned from the DPA between the SFO and ICBC SB Plc’ (2016) 5 JBL 388, 389. 32 DPA Consultation (n 2) para 30; Government Response to the DPA Consultation (n 7) paras 8-9. 33 DPA Consultation (ibid) para 34. 34 ibid para 42. 35 ibid para 76. 36 Polly Sprenger, Deferred Prosecution Agreements: The Law and Practice of Negotiated Corporate Criminal Penalties (Sweet and Maxwell, 2014) 333; see also DPA Code of Practice (n 14) para 6.3 and Criminal Procedures Rules 2015, SI 2015/1490, r 11.3(3).

Corporate prosecution, especially against “large… and sophisticated” companies, for most economic crimes is unlikely to succeed because liability is difficult to establish.37 Corporate liability for most such offences is premised upon the ‘identification doctrine’.38 This requires a person who (a) is the “directing mind and will” of the company, such as a director or senior officer, and (b) possesses the mens rea required for that offence (e.g. knowledge or intention) to be identified,39 , which is difficult especially as regards large companies.40 This is because decisions in these organisations are usually taken collectively at board level rather than individually.41 Further, even if ‘directing minds’ may be identified, “complex management structures” effectively insulate them from claims that they knew of or were complicit in the crimes committed,42 particularly because they tend to focus on strategic rather than operational issues.43 Indeed, the SFO has 37 DPA Consultation (n 2) para 26; Stephen F Copp and Alison Cronin, ‘New models of corporate criminality: the development and relative effectiveness of “failure to prevent” offences’ (2018) 39 Comp Law 104, 107. 38 Ministry of Justice, Corporate Liability for Economic Crime: Call for Evidence (Cm 9370, 2017) (Call for Evidence 2017) 5-6, 11; Jeremy Horder, Ashworth’s Principles of Criminal Law (8th edn, OUP 2016) 168; Richard Card and Jill Molley, Card, Cross and Jones’ Criminal Law (22 edn, OUP 2016) 795. 39 Tesco Supermarkets Ltd v Nattrass [1972] AC 153, [1971] 2 WLR 1166 (HL), 171 (Lord Reid); followed in R v St Regis Paper Co Ltd [2011] EWCA Crim 2527, [2012] PTSR 871 [12] (Moses LJ) and Attorney-General’s Reference (No 2 of 1999) [2000] QB 796, [2000] 3 WLR 195 (CA) 816 (Rose LJ). 40 DPA Consultation (n 2) para 26; Call for Evidence 2017 (n 38) 7; Ernest Lim, ‘A Critique of Corporate Attribution: Directing Mind and Will and Corporate Objective’ [2013] JBL 333, 333; Celia Wells, ‘A Quiet Revolution in Corporate Liability for Crime’ (1995) 145 NLJ 1326, 1327. 41 Call for Evidence 2017 (ibid) 13; Chris MV Clarkson, ‘Kicking Corporate Bodies and Damning Their Souls’ (1996) 59 MLR 557, 561; Nicholas Ryder, ‘“Too scared to prosecute and too scared to jail?” A critical and comparative analysis of enforcement of financial crime legislation against corporations in the USA and the UK’ (2018) 82 JCL 245, 258. 42 Call for Evidence 2017 (ibid) 13; Celia Wells, ‘Criminal Responsibility of Legal Persons in Common Law Jurisdictions’ (Paper prepared for OECD Working Group on Bribery in International Business Transactions Consultation on “The Responsibility of Legal Persons” 2000) 9; Olivia Dixon, ‘Corporate Criminal Liability: The Influence of Corporate Culture’ in Justin O’Brien and George Gilligan (eds), Integrity, Risk and Accountability in Capital Markets: Regulating Culture (Hart Publishing 2013) 256. 43 Horder (n 38); Brent Fisse and John Braithwaite,

139

ARTICLES

publicly attributed its inability to press charges against banks for fraud arising from the LIBORfixing scandal to these evidential hurdles.44 DPAs, which do not require criminal liability to be established as they divert from prosecution,45 thus circumvent these difficulties while holding corporate offenders accountable. Even where the identification doctrine is satisfied or where prosecution is otherwise successful, pursuing criminal proceedings may be an inappropriate mechanism to hold corporate offenders to account because, primarily with large corporations, it may cause collateral harm to innocent thirdparties.46 Measures resulting from conviction, such as debarment from public contracts,47 and even the reputational damage following an indictment, may critically compromise a company’s financial viability,48 which in turn may put at risk the jobs and investment of many innocent employees and shareholders respectively.49 This may consequently destabilise

140

Corporations, Crime and Accountability (CUP 1994) 47; Law Commission, Criminal Liability in Regulatory Contexts (Consultation Paper No 195, 2010) para 5.87. 44 David Green, ‘Cambridge Symposium 2016’ (Speech at the Cambridge Symposium on Economic Crime, Cambridge, 5 September 2016) <https://www.sfo.gov.uk/2016/09/05/ cambridge-symposium-2016/> accessed 13 March 2019; see also Jeremy Wright, ‘The Attorney General on the threat faced today from economic crime and how the UK Government is addressing it’ (Speech at the Cambridge Symposium on Economic Crime, Cambridge, 5 September 2016) <https:// www.gov.uk/government/speeches/attorney-general-jeremywright-speech-to-the-cambridge-symposium-on-economiccrime> accessed 13 March 2019. 45 Liz Campbell, ‘Corporate Liability and the Criminalisation of Failure’ (2018) 12 LFMR 57, 65. 46 DPA Consultation (n 2) para 27; VK Rajah, ‘Prosecution of financial crimes and its relationship to a culture of compliance’ (2016) 37 Comp Law 122, 125; David Green, ‘Ethical Business Conduct: an Enforcement Perspective’ (Speech at PricewaterhouseCoopers, 6 March 2014) <https:// www.sfo.gov.uk/2014/03/06/ethical-business-conductenforcement-perspective/> accessed 13 March 2019; Jed S Rakoff, ‘Justice Deferred is Justice Denied’ The New York Review of Books (New York, 19 February 2015) <https://www. nybooks.com/articles/2015/02/19/justice-deferred-justicedenied/> accessed 13 March 2019. 47 Grasso (n 31) 394; see for e.g. Public Contracts Regulations 2015, SI 2015/102, reg 57 and Utilities Contracts Regulations 2016, SI 2016/274, reg 80. 48 Mazzacuva (n 19) 250. 49 Paul Raudnitz, ‘Deferred prosecution agreements:

the wider economy.50 Such consequences have been termed the ‘Arthur Andersen effect’ after the collapse of Arthur Andersen, following its (eventually overturned) conviction for complicity in the Enron accounting scandal, which caused the loss of 28,000 jobs.51 Hence, in cases where prosecutors may be dissuaded from pursuing criminal proceedings against such commercial organisations due to concerns over such effects,52 the DPA option ensures that enforcement action can still take place. This is because it diverts from criminal proceedings, thus avoiding its alleged attendant consequences.53 In fact, such consequences are expressly set out in the UK’s DPA Code of Practice as a basis for which prosecutors can exercise their discretion against prosecution and in favour of entering into a DPA.54 Furthermore, corporate prosecution often involves high “resource and financial costs”55 as well as protracted investigations and court proceedings thereafter.56 For example, criminal proceedings resulting in late guilty pleas on bribery’, Insight (7 June 2017) <Westlaw> accessed 13 March 2019; Peter Reilly, ‘Justice Deferred is Justice Denied: We Must End Our Failed Experiment in Deferring Corporate Criminal Prosecutions’ (2015) 2015 BYU L Rev 307, 316; Simon Bronitt, ‘Regulatory Bargaining in the Shadows of Preventive Justice’ in Tamara Tulich et al (eds), Regulating Preventive Justice: Principle, Policy and Paradox (Routledge 2017) 215. 50 Bronitt (ibid) 215; Grasso (n 31) 393. 51 Bronitt (ibid) fn 21; Bob Ferguson, ‘Sanctions, Incentives and Better Behaved Banks’ in Justin O’Brien and George Gilligan (eds), Integrity, Risk and Accountability in Capital Markets: Regulating Culture (Hart Publishing 2013) 103. 52 Harry Wilson, ‘Banks are Too Big to Prosecute Says FSA’s Andrew Bailey’ Daily Telegraph (London, 14 December 2012) <https://www.telegraph.co.uk/finance/newsbysector/ banksandfinance/9743839/Banks-are-too-big-to-prosecutesays-FSAs-Andrew-Bailey.html> accessed 13 March 2019; Justin O’Brien, ‘The Sword of Damocles: Deferred Prosecutions and the Search for Accountability’ Justin O’Brien and George Gilligan (eds), Integrity, Risk and Accountability in Capital Markets: Regulating Culture (Hart Publishing 2013) 164. 53 Government Response to the DPA Consultation (n 7) 3; Chen Siyuan and Eunice Chua, ‘2018 Changes to the Evidence Act and Criminal Procedure Code’ (2018) 30 SAcLJ 1064, 1083. 54 DPA Code of Practice (n 14) para 2.8.2.vii; see also Crown Prosecution Service, ‘Corporate Prosecution: Legal Guidance’ para 30 <https://www.cps.gov.uk/legal-guidance/ corporate-prosecutions> accessed 13 March 2019. 55 DPA Consultation (n 2) para 25. 56 Mazzacuva (n 19) 251.

ARTICLES

average take 8 years and cost the SFO £1.6 million to conclude.57 This limits “the number of cases that can be pursued to an outcome”.58 Comparatively, DPAs may “provide a complete resolution to the wrongdoing” in a shorter time and with fewer resources expended,59 while obtaining a purportedly similar outcome as a successful prosecution. This is because a lower standard of evidence is required before commencing DPA negotiations than before starting criminal proceedings:60 prosecutors need only a “reasonable suspicion” that an offence has been committed and “reasonable grounds for believing that a continued investigation would provide further admissible evidence within a reasonable period of time”61 as compared to a “realistic prospect of conviction”.62 Prima facie, DPAs allow corporate offenders to be held accountable where the cost and duration of prosecution may be prohibitive.63 ii. Incentivising self-reporting Additionally, the UK DPA regime may incentivise companies suspected of wrongdoing to selfreport.64 By expressly emphasising extensive and timely self-reporting as a public-interest factor that weighs the exercise of prosecutorial discretion in favour of DPAs instead of prosecution, the DPA allows prosecutors to “capitalise” on the aversion companies have towards prosecution.65 This may arise 57 DPA Consultation (n 2) para 41. 58 ibid para 25. 59 Vicky Comino, ‘The GFC and beyond - how do we deal with corporate misconduct?’ (2018) 1 JBL 15, 35; see also Green (n 46) and Ryder (n 41) 259. 60 Ben Allen, ‘Deferred prosecution agreements – a new weapon in the anti-fraud and corruption armoury?’ (2014) 66(5) Governance Directions 285, 286. 61 DPA Code of Practice (n 14) para 1.2(i)(b). 62 Crown Prosecution Service, The Code for Crown Prosecutors (8th edn, October 2018) para 4.6. 63 Allen (n 60) 286. 64 Government Response to the DPA Consultation (n 7) 3; Rajah (n 46) 125; Copp and Cronin (n 37) 116. 65 Jonathan Grimes, Rebecca Niblock and Lorna Madden, ‘Corporate criminal liability in the UK: the introduction of deferred prosecution agreements, proposals for further change, and the consequences for officers and senior managers’ (2013) <Practical Law Multi-Jurisdictional Guide 2013/14: Corporate Crime, Fraud And Investigations> accessed 13

from the expenses involved, the financial losses resulting from measures that follow conviction like debarment (alluded to above)66 and its uncertainty of outcome which disrupts profitability.67 Hence, arguably, companies would be incentivised to self-report in the hopes of obtaining a DPA.68 This prospect is perhaps enhanced by the fact that a DPA seems to provide comparatively “greater certainty of outcome”69 and therefore a more commerciallyfriendly resolution to wrongdoing.70 This is because, inter alia, the likelihood of a DPA being approved at the final hearing may be indicated at the preliminary hearing, where ‘approval’ is subject to grounds similar to those at the former.71 Self-reporting allows prosecutors to better hold corporate offenders to account.72 This is because, in addition to difficulties relating to identifying a ‘directing mind’ with the necessary mens rea in companies with complex corporate structures,73 economic crime itself is difficult to detect and investigate without cooperation from within the company, given their “hidden, specialist or technical” nature.74 Not only do prosecutors often lack the requisite specialist knowledge and skills, the inner workings of large commercial organisations are “black boxes”,75 opaque to March 2019 referring to DPA Code of Practice (n 14) para 2.9. 66 Grimes, Niblock and Madden (ibid). 67 DPA Consultation (n 2) para 35. 68 Grimes, Niblock and Madden (n 65). 69 Comino (n 59) 35; Mazzacuva (n 19) 255. 70 n 67. 71 Oliver Heald, ‘The mechanics of Deferred Prosecution Agreements in the UK’ (Speech to the C5 7th Advanced Forum on Anti-Corruption, London, 26 June 2013) <https://www.gov.uk/government/speeches/the-mechanicsof-deferred-prosecution-agreements-in-the-uk> accessed 13 March 2019; Government Response to the DPA Consultation (n 7) para 74. 72 Comino (n 59) 35. 73 Mazzacuva (n 19) 251-2. 74 DPA Consultation (n 2) para 3; Bisgrove and Weekes (n 7) 417; Katalin Ligeti and Vanessa Franssen, ‘Current Challenges in Economic and Financial Criminal Law in Europe and the US’ in Katalin Ligeti and Vanessa Franssen (eds) Challenges in the Field of Economic and Financial Crime in Europe and the US (Hart Publishing 2017) 2. 75 Ligeti and Fransson (ibid); see also William Laufer, ‘Secrecy, silence, and corporate crime reforms’ (2010) 9 Criminology & Public Policy 455, 455-6, 458 and Henk van de

141

ARTICLES: WRITTEN MOOT 2

the outside world.76 Therefore, by incentivising self-reporting, the DPA allows prosecutors to be better placed not just to uncover evidence satisfying the identification doctrine, but also to detect the predicate offences, thus expanding the range of circumstances in which corporate offenders can be held to account.

the “pressure to do whatever it takes to meet business targets” and a belief that “their code of conduct is not [to be] taken seriously”.83 Particularly, it has been noted that a culture of irresponsibility84 and “cynical greed”85 precipitated the reckless practices at the heart of the LIBOR-fixing scandal.

B. Influencing corporate culture to prevent future wrongdoing

Corporate culture may be influenced by the DPA-imposed financial penalty86 because, by affixing a cost to uncompliant practices, these undermine the target company’s profitability,87 thus deterring the perpetuation of the poor underlying culture that in the first place encourages such wrongdoing.88 Additionally, a compliant culture may also be fostered by enforcing the adoption of more robust compliance programmes in the company,89 the failure of which would likely result in the resumption of prosecution,90 which companies seek to avoid (above). These programmes may shape compliance-friendly behaviour within the company,91 which when repeated over time may instil a more compliant corporate culture.92

The UK DPA is also arguably able to “encourage and enforce” changes in the culture, which is the “spirit that guides behaviour in the absence of rules”,77 of an entity under a DPA.78 Cultivating a corporate culture of compliance is necessary79 because where this is instilled, future corporate economic crime may be prevented.80 This is due to the fact that the “root cause” of most offending in this area is a deficient corporate culture.81 Such a culture creates conditions in which misconduct can take place,82 such as

142

Bunt, ‘Walls of secrecy and silence: the Madoff case and cartels in the construction industry’ (2010) 9 Criminology & Public Policy 435, 438-440. 76 Laufer (ibid) 455-6; van de Bunt (ibid); Anna Zalewska, ‘Twenty years after Cadbury, ten years after SarbanesOxley’ (2013) 27 Journal of Empirical Finance 1, 2. 77 Rajah (n 46) 129. 78 Bisgrove and Weekes (n 7) 417, 419; Serious Fraud Office, ‘UK’s first Deferred Prosecution Agreement, between the SFO and Standard Bank, successfully ends’ (30 November 2018) <https://www.sfo.gov.uk/2018/11/30/uksfirst-deferred-prosecution-agreement-between-the-sfo-andstandard-bank-successfully-ends/> accessed 13 March 2019. 79 Rajah (n 46) 129; Bisgrove and Weekes (n 7) 417. 80 Comino (n 59) 35; Jake A Nasar, ‘In Defense of Deferred Prosecution Agreements’ (2017) 11 NYU JL & Liberty 838, 849; Attorney-General’s Department, Improving enforcement options for serious corporate crime: consideration of a Deferred Prosecution Agreements scheme in Australia Public Consultation Paper (March 2016) 9, 21. 81 Cristie Ford and David Hess, ‘Corporate Monitorships and New Governance Regulation: In Theory, in Practice, and in Context’ (2011) 33 Law & Pol’y 509, 512; Justin O’Brien and George Gilligan, ‘Regulating Culture: Problems and Perspectives’ in Justin O’Brien and George Gilligan (eds), Integrity, Risk and Accountability in Capital Markets—Regulating Culture (Hart Publishing 2013) xxviii. 82 Nicholas Lord, Regulating Bribery in International Business: Anti-Corruption in the UK and Germany (Ashgate 2014) 3; Ford and Hess (ibid); see generally Christof Miska, Günter K Stahl and Matthias Fuchs, ‘The Moderating Role of Context in Determining Unethical Managerial Behavior: A Case Survey’ (2018) J Bus Ethics 157.

83 Ford and Hess (ibid) citing KPMG, KPMG Forensic: Integrity Survey 2008-2009 (2008) 6 <http://www.kpmg.com. br/publicacoes/forensic/Integrity_Survey_2008_2009.pdf> accessed 13 March 2019. 84 Ferguson (n 51) 92 citing Will Hutton, ‘Bank RateFixing Scandals Reveal the Rotten Heart of Capitalism’ The Guardian (23 December 2012) <https://www.theguardian. com/commentisfree/2012/dec/23/banking-reform-ubs> accessed 13 March 2019. 85 Ferguson (ibid) citing Adair Turner, ‘Banking at The Cross-Roads: Where Do We Go from Here?’ (Speech at Bloomberg, London, 24 July 2012) <http://www.fsa.gov. uk/library/communication/speeches/2012/0724-at.shtml> accessed 13 March 2019. 86 Bisgrove and Weekes (n 7) 417. 87 Ferguson (n 51) 93, 99, 110. 88 Bisgrove and Weekes (n 7) 417; Nasar (n 80) 867; O’Brien and Gilligan (n 81) fn 43. 89 Comino (n 59) 35; Bronitt (n 49) 217; Ryan D McConnell et al, ‘Plan Now or Pay Later: The Role of Compliance in Criminal Cases’ (2011) 33 Hous J Intl L 509, 584-86. 90 Mazzacuva (n 19) 261 citing Leonard Orland, ‘The Transformation of Corporate Criminal Law’ (2006) 1 Brooklyn Journal of Corporate, Financial & Commercial Law 45, 84-85. 91 Christine Parker and Vibeke L Nielsen, ‘Corporate Compliance Systems Could They Make Any Difference?’ (2009) 41 Administration & Society 3, 9. 92 Financial Reporting Council, Corporate Culture and the Role of Boards: Report of Observations (July 2016) 25 <https:// www.frc.org.uk/getattachment/3851b9c5-92d3-4695-aeb2-

ARTICLES: WRITTEN MOOT 2

Hence, in positively affecting corporate culture, future wrongdoing may be averted. II. The DPA in practice A. Overstated flexibility

need

for

enforcement

This essay acknowledges that DPAs have indeed made enforcement against corporate economic crime more flexible in the UK as an additional tool in the prosecutorial arsenal. Indeed, the fact that the Standard Bank DPA took merely 2.5 years to conclude93 highlights that the DPA regime provides prosecutors with a more efficient means to hold corporate offenders to account. However, it is submitted that the need for flexibility may be overstated because the enforcement gap which DPAs purport to fill is in reality narrower than recognised. This is for two reasons. First, the notion that prosecution is inappropriate especially against large companies because it brings about disproportionate collateral consequences on innocent thirdparties is empirically unsupported. These consequences are said to flow from the initial collapse of a company, especially a large one; yet, the claim that prosecution causes such companies to go out of business in the first place is unsubstantiated. In fact, studies in the US indicate that not only is the causative relationship between conviction and ‘corporate death’ absent in publicly-listed companies, the vast majority of those which have been convicted remain in operation as well.94 Further, while the UK Government pointed to Arthur

87c9052dc8c1/Corporate-Culture-and-the-Role-of-BoardsReport-of-Observations.pdf> accessed 13 March 2019. 93 Jan Miller, ‘First DPA negotiated with SFO’ (2015) 165 NLJ 7680, 4. 94 Gabriel Markoff, ‘Arthur Andersen and the Myth of the Corporate Death Penalty: Corporate Criminal Convictions in the Twenty-First Century’ (2013) 15 U Pa J Bus L 797, 82325, 827; Gordan Bourjaily, ‘DPA DOA: How and Why Congress Should Bar the Use of Deferred and Non-Prosecution Agreements in Corporate Criminal Prosecutions’ (2015) 52 Harvard J on Legis 543, 556.

Andersen’s collapse in support of this claim,95 this is explainable by the firm’s anomalous characteristics.96 As a ‘big-five’ accounting firm, its “core product was not simply accounting services” but “the trustworthiness of its name”.97 Since this very ‘trustworthiness’ was shattered by its public connection with Enron’s accounting fraud, Arthur Andersen was left without a “product to offer”.98 Therefore, it is contended that its collapse had less to do with its involvement in criminal proceedings per se but more this aforementioned publicity.99 Second, a successful prosecution is not as unlikely as has been argued, albeit in the context of bribery and tax evasion. This is because the introduction of the failure to prevent bribery100 offence as well as the failure to prevent the facilitation of UK tax evasion101 and foreign tax evasion102 offences makes prosecuting corporate offenders in relation to such conduct a more successful prospect.103 This is evident in the convictions of Sweett Group104 and Skansen Interiors105 for failing to prevent bribery. Indeed, it is even possible for a company to be convicted where none of its officers knew of the bribes 95 DPA Consultation (n 2) para 28. 96 Ellis W Martin, ‘Deferred Prosecution Agreements: Too Big to Jail and the Potential of Judicial Oversight Combined with Congressional Legislation’ (2014) 18 NC Banking Inst 457, 469-70; Brandon L Garrett, Too Big to Jail: How Prosecutors Compromise with Corporations (HUP 2014) 42. 97 Markoff (n 94) 834. 98 ibid. 99 David Uhlmann, ‘Deferred Prosecution and NonProsecution Agreements and the Erosion of Corporate Criminal Liability’ (2013) 72 Md L Rev 1295, 1321; Cindy Schipani, ‘Trends in prosecutions for corporate crime in the US’ (2018) 39 Comp Law 43, 46. 100 Bribery Act 2010 (BA), s 7. 101 Criminal Finances Act 2017 (CFA), s 45. 102 ibid, s 46. 103 Campbell (n 45) 61; Celia Wells, ‘Corporate failure to prevent economic crime - a proposal’ (2017) 6 Crim LR 426, 434; Carey Lynn and Neil Warlow, ‘The Criminal Finances Act 2017 - what does it mean for directors and officers insurance?’ (2017) 6 Comp & Risk 10, 12. 104 R v Sweett Group plc (2016) Southwark Crown Court (unreported) noted Marie-Anne Denicolo, ‘First conviction for Bribery Act corporate offence’ (2016) 5 CRJ 17. 105 R v Skansen Interiors Limited (2018) Southwark Crown Court (unreported) noted Omar Qureshi, Amy Wilkinson and Iskander Fernandez, ‘Failure to prevent bribery: first contested prosecution’ (2018) 29(3) PLC 6.

143

ARTICLES: WRITTEN MOOT 2

paid, or the facilitation of tax evasion committed by, an ‘associated person’,106 which includes any person performing services for or on behalf of the company.107 This is because liability for these offences is “semi-strict”:108 a company is prima facie liable for the bribes paid for its benefit by an ‘associated person’,109 or for the facilitation of tax evasion by an ‘associated person’ acting in that capacity,110 unless it has implemented adequate procedures at the time of the bribery111 or reasonable prevention procedures at the time of tax evasion facilitation.112 In other words, prosecutors can circumvent the identification doctrine and its evidential difficulties when prosecuting companies in these areas.113 Therefore, in the context of these offences, the likelihood of successful prosecution (along with the unsupported nature of the ‘Arthur Andersen effect’) undercuts the necessity of the flexibility provided by DPAs.114 Hence, it is arguable that the DPAs entered into in relation to the failure to prevent bribery offences, i.e. Standard Bank,115 XYZ (partially),116 and RollsRoyce (partially),117 have been to some extent unnecessary because enforcement action via prosecution could have been taken instead. Notwithstanding the above, this essay also

144

106 Tim Pope and Thomas Webb, ‘Legislative Comment – the Bribery Act 2010’ (2010) 25 JIBLR 480, 482; Campbell (n 45) 60. 107 BA (n 100), ss 7-8; CFA (n 101), s 44. 108 Celia Wells, ‘Corporate criminal liability: a ten year review’ (2014) 12 Crim LR 849, 872. 109 BA (n 100), s 7(1). 110 CFA (n 101), ss 45(1), 46(1). 111 BA (n 100), s 7(2). 112 CFA (n 101), ss 45(2), 46(3). 113 Call for Evidence 2017 (n 38) 8, 18; Rita Cheung, ‘Money laundering - a new era for sentencing organisations’ (2017) 1 JBL 23, fn 4; Ryder (n 41) 259; Karl Laird, ‘The Criminal Finances Act 2017 - an introduction’ (2017) 12 Crim LR 915, 933. 114 Colin King and Nicholas Lord, Negotiated Justice and Corporate Crime: The Legitimacy of Civil Recovery Orders and Deferred Prosecution Agreements (Palgrave Macmillan 2018) (Negotiated Justice and Corporate Crime) 137. 115 Serious Fraud Office v Standard Bank Plc [2015] Southwark Crown Court Case No. U20150854, [2016] Lloyd’s Rep FC 102 (Standard Bank) [10]. 116 Serious Fraud Office v XYZ Ltd [2016] Southwark Crown Court Case No. U20150856, [2016] Lloyd’s Rep FC 509 (XYZ) [12]. 117 Rolls-Royce (n 8) [77].

furthers the argument that the UK DPA is unlikely to be an effective tool in addressing corporate economic crime because it is unlikely to improve corporate accountability and change corporate culture in practice. B. Accountability is likely limited because self-reporting is potentially disincentivised DPAs are unlikely to improve prosecutors’ ability to hold corporate offenders accountable in practice because the incentives to self-report are potentially undermined following the RollsRoyce DPA. This would deny prosecutors the evidential benefits that DPAs purportedly facilitate. The Rolls-Royce DPA was approved by Sir Leveson P even though Rolls-Royce had failed to self-report; instead, SFO investigations had commenced only after becoming aware of corruption allegations involving the company online.118 This was despite the fact that the previous two DPAs (Standard Bank119 and XYZ)120 appeared to have been approved on the ‘precondition’ of a self-report.121 Indeed, in approving those, His Lordship had held that the “core of the interests of justice test”, which the court has to consider when approving a DPA, “lies in the promptness of the self-report”.122 In addressing this inconsistency, Leveson P accepted the SFO’s argument that Rolls-Royce’s “extraordinary” cooperation,123 like its voluntary disclosure of internal investigations and waiver of privilege over internal investigation 118 Rolls-Royce (n 8) [16]; noted Rachel Nicolson et al, ‘The Rolls-Royce bribery case and its implications in Australia’ (2017) 69 Governance Directions 117, 118. 119 Standard Bank (n 115) [14]. 120 XYZ (n 116) [16]. 121 OECD, Implementing the OECD Anti-Bribery Convention Phase 4 Report: United Kingdom (OECD, 2017) 16 <http://www. oecd.org/corruption/anti-bribery/UK-Phase-4-Report-ENG. pdf> accessed 13 March 2019. 122 Rita Cheung, ‘Deferred Prosecution Agreements: Cooperation and Confession’ (2018) 77 Camb LJ 12, 13 citing Standard Bank (n 115) [14] and XYZ (n 116) [16]. 123 Rolls-Royce (n 8) [22], [23].

ARTICLES: WRITTEN MOOT 2

memoranda,124 was indistinguishable from the self-reporting by Standard Bank and XYZ.125 In so doing, His Lordship effectively equated cooperation with self-reporting.126 However, this is likely to undermine the incentives to self-report127 because it represents that selfreporting is unnecessary. A company can respond to wrongdoing passively without “proactively” notifying the authorities and still escape criminal sanction128 because it would nevertheless remain eligible for a DPA if it subsequently cooperates.129 Moreover, given that at least some of the alleged conduct was known to Rolls-Royce two years before SFO investigations had taken place,130 Rolls-Royce’s DPA signals that even a “deliberate” failure to self-report may not be fatal to a company’s DPA prospects.131 Furthermore, His Lordship eventually granted a 50% discount on Rolls-Royce’s financial penalty to reflect its considerable cooperation.132 While this discount was not unprecedented since it followed the same deduction granted to XYZ,133 the basis for which the discount had been applied in XYZ was to “encourage others to conduct themselves as XYZ ha[d] when confronting criminality”, i.e. self-report.134 Thus, by applying the approach taken in XYZ to Rolls-Royce, where there was no self-report, the incentives for “rational, amoral corporates” to

124 ibid [121]. 125 ibid [22]. 126 Cheung (n 122) 13. 127 Corruption Watch, ‘A Failure of Nerve: The SFO’s Settlement with Rolls Royce’ (Corruption Watch, 19 January 2017) <https://www.cw-uk.org/single-post/2017/01/19/AFailure-of-Nerve-The-SFO%E2%80%99s-Settlement-withRolls-Royce> accessed 13 March 2019; Wells (n 103) 431. 128 Nicholas Lord and Colin King, ‘Negotiating NonContention: Civil Recovery and Deferred Prosecution in Response to Transnational Corporate Bribery’ in Liz Campbell and Nicholas Lord (eds) Corruption in Commercial Enterprise: Law, Theory and Practice (Routledge 2018) 241. 129 Corruption Watch (n 127). 130 Rolls-Royce (n 8) [16]-[17]. 131 Corruption Watch (n 127). 132 Rolls-Royce (n 8) [120]-[123]; OECD (n 121) 16. 133 XYZ (n 116) [23]. 134 XYZ (ibid) noted King and Lord, Negotiated Justice and Corporate Crime (n 114) 108.

self-report are potentially further undermined.135 Companies who fail to self-report not only have recourse to a DPA but can even take the benefit of a “full discount” by cooperating extensively once caught.136 Therefore, with incentives to self-report likely to be undermined in reality, the DPA regime is unlikely to contribute towards effectively addressing corporate economic crime because the difficulties in detecting and investigating corporate economic crime are likely to persist. Although, to some extent, this may be mitigated by expanding prosecutors’ “intelligence capabilities”, this requires sufficient funding,137 an issue which lies beyond the scope of this essay. C. Unlikelihood of effecting changes in corporate culture Additionally, real changes in the culture of companies subject to a DPA, which goes towards preventing future wrongdoing, are unlikely to take place in reality because the deterrent effect of financial penalties is likely to be muted and compliance programmes are unlikely to effect substantive improvements in corporate behaviour. Even if the causal connection between deterrence and cultural change can be made out, for which there appears to be an unexplained assumption in the academic commentary, it is questionable that deterrence can be achieved in the first instance. Financial penalties imposed by a DPA cannot generally be of sufficient deterrence as regards corporations, especially large ones. This is because they may “virtually pay a penalty of any amount, considering it just as an additional operational cost.”138 For example, despite suggestions by some that the financial penalty imposed on Standard 135 136 137 138 469.

King and Lord (ibid). King and Lord (ibid) 109, 111; OECD (n 121) 17. King and Lord (ibid) 106; OECD (ibid). Grasso (n 31) 405; Ryder (n 41) 256; Martin (n 96)

145

ARTICLES: WRITTEN MOOT 2

Bank was adequate139 or even too large,140 it is submitted otherwise. The Industrial and Commercial Bank of China, which acquired a majority stake in Standard Bank after the events central to the DPA took place,141 had assets totalling US$3,000,000 million and a net profit of approximately US$42,000 million in the year before the DPA was approved.142 It is “evident” that the penalty of US$16.8 million “cannot plausibly” be of significance to it.143 The lack of deterrence is further perpetuated by the very fact that these financial penalties are not the result of criminal proceedings, which means that they do not carry “moral condemnation” 144 and hence are of much lower “weight and stigma” than a criminal fine.145 This entrenches the view that these penalties are simply another cost of doing business,146 a “risk factor to consider on the road to profits”.147 Therefore, financial penalties are unlikely to be of sufficient deterrence.

146

Nevertheless, even if financial penalties do indeed have a deterrent effect, it has been suggested that they need to be accompanied

by enforced compliance reforms to achieve cultural change.148 To this end, it should be noted that all four UK DPAs required compliance programmes to be reviewed and changes implemented.149 Yet, it is unlikely that genuine cultural change that eliminates the causes of misconduct would take effect. This requires substantive, “warranted commitment to moral restraint” by these companies.150 However, this is unlikely to take place because boards generally do not “sufficiently value compliance”,151 regarding it as a “cost to be minimized”.152 Indeed, it has been observed that a ‘box-ticking’ approach to compliance is prevalent among companies, not least in the UK:153 companies simply comply with “a specific provision in a literal sense without observing the spirit of the principle on which the provision is based”.154 It is thus possible that compliance measures are implemented and formally adhered to without delivering substantive changes to a company’s practices, let alone culture.155 For example, banks have been observed to extend credit to associated companies for 364 days in order to comply with the Basel I regulation requiring

139 Nicola Padfield, ‘Deferred Prosecution Agreements’ (2016) 7 Crim LR 449, 449. 140 Ellen Gallagher, ‘The Standard Bank DPA—The First of Many?’ (International Bar Association, 9 June 2016) <https:// www.ibanet.org/Article/Detail.aspx?ArticleUid=70d2382ae94e-4d81-a12c-3bef0730747b> accessed 13 March 2019. 141 Standard Bank (n 115) [14]. 142 Grasso (n 31) 406; see ICBC, ‘Annual Report 2015’ (ICBC, 2015) 6 <http://v.icbc.com.cn/userfiles/Resources/ ICBC/haiwai/ICBCLondon/download/2015/ICBC_ Group_2015_annual_report.pdf> accessed 13 March 2019. 143 Grasso (ibid). 144 Randall Eliason, ‘We Need to Indict Them’ Legal Times (22 September 2008) <http://www3.wcl.american.edu/ faculty/eliason/LegalTimes_9_22_08.pdf> accessed 13 March 2019. 145 Martin (n 96) 468; Corruption Watch, Out of Court, Out of Mind: do Deferred Prosecution Agreements and Corporate Settlements Fail to Deter Overseas Corruption (Corruption Watch, 2016) 13 146 O’Brien and Gilligan (n 81) xvii; Fisse and Braithwaite (n 43) 42; Eliason (n 144); Miriam F Weismann, Christopher A Buscaglia and Jason Peterson, ‘The Foreign Corrupt Practices Act: Why It Fails to Deter Bribery as a Global Market Entry Strategy’ (2014) 123 J Bus Ethics 591, 597. 147 Editorial, ‘Too Big to Indict’ New York Times (New York, 11 December 2012) <https://www.nytimes. com/2012/12/12/opinion/hsbc-too-big-to-indict.html> accessed 13 March 2019; see also Weismann, Buscaglia and Peterson (ibid) 599.

148 O’Brien (n 52) 164; Ferguson (n 51) 110. 149 Standard Bank (n 115) [13]; XYZ (n 116) [14]; RollsRoyce (n 8) [67]; Serious Fraud Office v Tesco Stores Ltd [2017], Southwark Crown Court Case No. U20170287 (Tesco) [97]-[101]. 150 O’Brien and Gilligan (n 81) xxi. 151 Parker and Nielsen (n 91) 10. 152 Ford and Hess (n 81) 515; William S Laufer, ‘Corporate liability, risk shifting and the paradox of compliance’ (1999) 52 Vanderbilt Law Review 1341, 1404, 1415; Steven Sampson, ‘The “Right Way”: Moral capitalism and the emergence of the corporate ethics and compliance officer’ (2016) JBA Special Issue on Business Ethics 65, 77; Economist Intelligence Unit, ‘Beyond box-ticking: A new era for risk governance’ (The Economist, 2009) 4 <http://graphics.eiu.com/marketing/pdf/ beyondboxticking.pdf> accessed 13 March 2019. 153 Bobby V Reddy, ‘Thinking Outside the Box – Eliminating the Perniciousness of Box-Ticking in the New Corporate Governance Code’ (2019) MLR (Early View) 2-6 <https://doi.org/10.1111/1468-2230.12415> accessed 13 March 2019; Marie McKendall, Beverly Demarr and Catherine Jones-Rikkers, ‘Ethical Compliance Programs and Corporate Illegality: Testing the Assumptions of the Corporate Sentencing Guidelines’ (2002) 37 J Bus Ethics 367, 380; Campbell (n 45) 63 (‘ritualism’). 154 Reddy (ibid) 2; see also Doreen McBarnet and Christopher Whelan, ‘The elusive spirit of the law: formalism and the struggle for legal control’ (1991) 54 MLR 848, 848-849. 155 Parker and Nielsen (n 91) 9-10; McKendall, Demarr and Jones-Rikkers (n 153) 379-380 and Laufer (n 152) 1407, 1415 (‘cosmetic’ compliance).

ARTICLES: WRITTEN MOOT 2

such loans for at least 365 days to be backed up by capital.156

A. Inherent dependence on the strength of prosecution

Therefore, with cultural change that prevents future wrongdoing unlikely to take effect through deterrence or enforced compliance measures, it is submitted that the DPA is unlikely to be an effective tool in addressing corporate economic crime. This is further supported by the recidivist behaviour of companies which have entered into DPAs in the US,157 most notably Pfizer Inc., which was involved in bribery despite previously entering into three DPAs for similar conduct.158

As discussed, DPAs may be said to allow prosecutors to hold corporate offenders to account by enhancing the detection and investigation of corporate economic crime through incentivising self-reporting and providing scope for enforcement action to be taken in lieu of formal criminal proceedings. As regards the former, corporate offenders are fundamentally unlikely to self-report in the first place. This is because the “absence of a credible threat of corporate prosecution”159 arising from the evidential limitations of the identification doctrine, which is the principal mechanism for establishing corporate criminal liability for economic crimes (above), makes the risk of “getting caught” too low to justify “coming forward and accepting liability”.160 Similarly, for a DPA to come into force as an enforcement mechanism, it requires the target company to voluntarily agree to its terms and statement of facts, which it is not obliged to do.161 Yet, because of the weakness of the corporate criminal liability regime, the likelihood of being convicted at trial would also similarly be insufficiently significant to justify accepting

III. Fundamental limitations of the UK DPA Moreover, it is further submitted that UK DPAs are even more unlikely to effectively address corporate economic crime because they are fundamentally limited by their inherent characteristics. First, as alluded to in Part II, their ability to hold corporate offenders to account is intrinsically dependent upon the strength of the UK’s corporate crime enforcement response. This being largely weak, their capacity to ensure accountability is thus constrained. Second, being a procedure targeting corporations or similar bodies, the DPA is beset with legitimacy issues that restrict its potential to institute substantive changes in corporate culture.

156 Grant Kirkpatrick, ‘The Corporate Governance Lessons from the Financial Crisis’ [2009] 1 Financial Market Trends 61, 70; Adrian Blundell-Wignall and Paul Atkinson, ‘The Sub-Prime Crisis: Causal Distortions and Regulatory Reform’ in Paul Bloxham and Christopher Kent (eds) Lessons from the Financial Turmoil of 2007 and 2008 (Reserve Bank of Australia 2008) 78; Julia Black, ‘Paradoxes and Failures: ‘New Governance’ Techniques and the Financial Crisis’ (2012) 75 MLR 1037, 1040. 157 Weismann, Buscaglia and Peterson (n 146) 596; Eunice Chua, ‘Deferred Prosecution Agreements in Singapore?’ (Singapore Law Blog, 30 January 2018) <http://www. singaporelawblog.sg/blog/article/205> accessed 13 March 2019. 158 Nasar (n 80) 874-875; Rakoff (n 46).

159 King and Lord, Negotiated Justice and Corporate Crime (n 114) 76, 82, 119, 132; Mike Levi and Nicholas Lord, ‘WhiteCollar and Corporate Crimes’, in Alison Liebling, Shadd Maruna, and Lesley McAra (eds), Oxford Handbook of Criminology (6th edn, OUP 2017) 734; Call for Evidence 2017 (n 38) 21, 23. 160 Marialuisa Taddia, ‘Economic crime: corruption conundrum’ The Law Society Gazette (London, 27 October 2014) quoting Jeremy Summers <http://www.lawgazette.co.uk/law/ economic-crime-corruption-conundrum/5044646.fullarticle> accessed 13 March 2019; James Mulholland, Alexandra Ward and Eleanor Mawrey, Joint Response of the Criminal Bar Association And Law Reform Committee of the General Council of the Bar of England and Wales to the Ministry Of Justice Consultation On A New Enforcement Tool To Deal With Economic Crime Committed By Commercial Organisation: Deferred Prosecution Agreements (July 2012) para 7 <https://www.barcouncil.org.uk/media/160666/ joint_cba___lrc_deferred_prosecution_agreements_response. pdf> accessed 13 March 2019; ‘Lessons Learnt from Recent Deferred Prosecution Agreements and What’s in Store for 2018? Exiger Integrity Forum’ (Exiger, 13 December 2017) citing Edward Garnier’s speech at the Exiger Integrity Forum <https://www.exiger.com/perspectives/lessons-learnt-recentdeferred-prosecution-agreements-and-what%E2%80%99sstore-2018-exiger> accessed 13 March 2019. 161 DPA Code of Practice (n 14) para 3.3.

147

ARTICLES: WRITTEN MOOT 2

the burdens of the DPA’s terms.162 This is perhaps emphasised by the recent acquittals of Tesco’s ‘directing minds’ due to the lack of evidence establishing mens rea for, inter alia, false accounting,163 even after a DPA had been entered into with Tesco for that offence.164 This demonstrates that the threshold for imposing criminal liability on corporations is high, since a ‘directing mind’ with the appropriate mental element has to be identified first. Ergo, this strongly suggests that Tesco’s entry into a DPA along with its accompanying £129 million penalty is not justifiable in hindsight.165 Hence, it has been argued that “the success of DPAs in the UK is linked to reforming corporate criminal liability across the board”.166 In this regard, as discussed above, it should be noted that the other three DPAs entered into have been for the ‘failure to prevent’ bribery offence, which enables liability to be more easily established. Yet, there have been concerns about the lack of prosecution for this offence, given that there have only been two so far.167 Indeed, the enforcement of the ‘failure to prevent’ offences have been characterized as demonstrating a “preference for settlement over contention”;168 resultingly, it appears that prosecution will only be commenced against

148

162 Corruption Watch (n 145) 20; David Green, ‘Cambridge Symposium 2015’ (Speech at the Cambridge Symposium on Economic Crime, Cambridge, 7 September 2015) <https://www.sfo.gov.uk/2015/09/07/cambridgesymposium-2015/> accessed 13 March 2019; Jonathan Fisher et al ‘The Global Financial Crisis: The Case for a Stronger Criminal Response’ (2013) 7 LFMR 159, 162. 163 R v Bush [2019] EWCA Crim 29 [121]-[131]; see Serious Fraud Office, ‘Deferred Prosecution Agreement between the SFO and Tesco published’ (Serious Fraud Office News Releases, 23 January 2019) <https://www.sfo.gov. uk/2019/01/23/deferred-prosecution-agreement-betweenthe-sfo-and-tesco-published/> accessed 13 March 2019. 164 Tesco (n 149) [7]. 165 Joanna Dimmock, ‘Tesco case highlights fundamental flaws with current DPA regime’ (White & Case, 1 February 2019) <https://www.whitecase.com/publications/alert/tescotrial-collapse-highlights-dangers-early-deferred-prosecutionagreement> accessed 13 March 2019. 166 Taddia (n 160); see also Green (n 162). 167 Law Society of England and Wales, the City of London Law Society and the Fraud Lawyers Association, House of Lords Committee on the Bribery Act: Responses To Call For Evidence (31 July 2018) para 2.1; Frederick Saugman, ‘The Bribery Act 2010 – Post-Legislative Scrutiny’ (WilmerHale, 2 July 2018) 168 Campbell (n 45) 65.

a commercial organisation, especially a large one, in very limited circumstances.169 This is underscored by the Rolls-Royce DPA itself, which was entered into and subsequently approved despite the egregiousness of the bribery offences involved.170 Therefore, because it appears that the corporate criminal enforcement regime in the UK is overall weak, it is unlikely that companies would accept a DPA or even self-report, thus rendering the UK DPA fundamentally unlikely to improve the accountability of corporate offenders for economic crime. B. Legitimacy issues Additionally, this essay contends that UK DPAs are fundamentally unlikely to foster changes in corporate culture, even through enforcing compliance reform, because it is in the first place unable to effect the implementation of compliance measures that results in genuine change in compliance practice. This is because, by design, DPAs suffer from legitimacy issues.171 An important feature of legitimacy is “normative validity”, which, where applied to the DPA, entails that it must be justifiable according to accepted standards.172 This includes notions of procedural fairness,173 for which “equal treatment before the law” is foundational.174 The UK DPA regime, however, contradicts this principle,175 since it only extends 169 Justin O’Brien, ‘Just when will a corporation be prosecuted to a judicial conclusion?’ (2017) 11, LFMR 1, 2; Robert Barrington, ‘What Can We Learn about the Future of DPAs from the Rolls-Royce Case?’ (Transparency International UK, 18 January 2017) <https://www.transparency.org.uk/ what-can-we-learn-about-the-future-of-dpas-from-the-rollsroyce-case/> accessed 13 March 2019; King and Lord, Negotiated Justice and Corporate Crime (n 114) 102. 170 Rolls-Royce (n 8) [4], [61]. 171 Bronitt (n 49) 218. 172 King and Lord, Negotiated Justice and Corporate Crime (n 114) 28, 30; see generally David Beetham, The Legitimation of Power (2nd edn, Palgrave Macmillan 2013). 173 King and Lord (ibid) 25; Christopher Hodges, Law and Corporate Behaviour Integrating Theories of Regulation, Enforcement, Compliance and Ethics (Hart Publishing 2015) 27; see generally Tom R Tyler, ‘Procedural Justice, Legitimacy, and the Effective Rule of Law’ (2003) 30 Crime and Justice 283. 174 King and Lord (ibid) 120. 175 Chua (n 157); Bronitt (n 49) 168; Comino (n 59) 32.

ARTICLES: WRITTEN MOOT 2

to companies or similar bodies.176 This “leaves the lasting impression of a two-tier criminal system”177 with companies being too big to prosecute. Indeed, the DPA Code of Practice, by permitting prosecutors to consider the effect of a company’s criminal conviction on thirdparties when exercising their discretion to enter into a DPA,178 arguably encourages lawyers to argue that their corporate clients should be treated differently from individuals.179 For corporations then, DPAs appear merely to be “state-sanctioned ‘legal bribes’” to avoid prosecution.180 When a DPA is perceived as illegitimate, the company subject to it is unlikely to possess a long-term, normative commitment to its demands,181 viewing them mercenarily rather than valuing them “for [their] own sake”.182 As such, this fundamentally prevents compliance programmes mandated by a DPA from fostering a genuine commitment to change actual compliance behaviour,183 resulting in an unchanged culture and hence the persistence of wrongdoing in future.184 Moreover, this 176 Bronitt (ibid). 177 Matt Trome, ‘UK Bribery Prosecutions and the Rule of Law’ (The Global Anticorruption Blog, 24 August 2017) <https://globalanticorruptionblog.com/2017/08/24/guestpost-uk-bribery-prosecutions-and-the-rule-of-law/> accessed 13 March 2019; Chen and Chua (n 53) 1084; King and Lord, Negotiated Justice and Corporate Crime (n 114) 30; Nicola Padfield, ‘Deferred Prosecution Agreements’ (2012) 7 Arch Rev 4, 5. 178 Bronitt (n 49) 217 referring to DPA Code of Practice (n 14) para 2.8.2.vii. 179 Bronitt (ibid). 180 Bronitt (ibid) 218; see also King and Lord, Negotiated Justice and Corporate Crime (n 114) 30 and Padfield (n 177) 5. 181 Christine Parker, ‘The “Compliance” Trap: The Moral Message in Responsive Regulatory Enforcement’ (2006) 40 Law & Society Review 591, 592; King and Lord, Negotiated Justice and Corporate Crime (n 114) 24-25, 121; Hodges (n 173) 27-28; Tom R Tyler et al, ‘The Ethical Commitment to Compliance: Building Value-Based Cultures’ (2008) 50 California Management Review 31, 35. 182 Bourjaily (n 94) 554. 183 Bronitt (n 49) 218 citing N Richard Janis, ‘Deputising company counsel as agents of the federal government’ (White Paper, Cato Institute, 2000) 4 <https://object.cato.org/sites/ cato.org/files/pubs/pdf/janis_deputizing.pdf> accessed 13 March 2019 and Kurt Eichenwald, Serpent on the Rock (Harpercollins 1995) xiii-ix; see also O’Brien (n 52) 161, 164. 184 King and Lord, Negotiated Justice and Corporate Crime (n 114) 24.

is exacerbated by the fact that compliance programmes perceived to be illegitimate conversely facilitate behaviour which “compliance is intended to prevent”.185 This is because they encourage individuals acting within a company to rationalise the misconduct that they commit through it,186 a process which is critical in economic crime.187 This erodes any “normative ‘check’” against wrongdoing that may previously have existed.188 IV. Conclusion This essay has examined the arguments in support of the UK DPA being an effective tool for addressing economic crime perpetrated by companies. To the extent that UK DPAs may be said to improve prosecutors’ ability to hold corporate offenders to account by providing enforcement flexibility, this essay argues that this is less necessary than perceived. While acknowledging that DPAs side-step the resource and time constraints of prosecution, it is submitted that corporate prosecution has not been empirically demonstrated to result in collateral consequences on innocent thirdparties and is in fact likely to be successful (if pursued) in the context of the ‘failure to prevent’ offences. Hence, prosecution cannot conclusively be said to be inappropriate or unlikely to be successful, rendering the argument as regards enforcement flexibility less convincing. In addition, it has also been argued that DPAs improve the ability of prosecutors to hold corporate offenders to account by incentivising self-reporting so that prosecutors can detect 185 Todd Haugh, ‘The Criminalisation of Compliance’ (2017) 92 Notre Dame L Rev 1215, 1252; see Campbell (n 45) 64. 186 Haugh (ibid). 187 ibid 1255 citing, inter alia, Michael L Benson, ‘Denying the Guilty Mind: Accounting for Involvement in a White-Collar Crime’ (1985) 23 Criminology 583, 591-598 and Petter Gottschalk, ‘Rotten Apples Versus Rotten Barrels in White Collar Crime: A Qualitative Analysis of White Collar Offenders in Norway’ (2012) 7 International Journal of Criminal Justice Sciences 575, 580–81. 188 Haugh (ibid) 1261.

149

ARTICLES: WRITTEN MOOT 2

corporate economic crime better. Yet, this essay avers that self-reporting is likely to have been disincentivised in practice following Rolls-Royceâ&#x20AC;&#x2122;s DPA. Furthermore, this essay challenges the assertion that the DPA improves the ability of prosecutors to change the culture of companies who have entered into DPAs through the deterrent effect of the financial penalty as well as by enforcing compliance reforms. This is because, in practice, these financial penalties are unlikely to be of sufficient deterrence to large commercial organisations and enforced compliance reforms are unlikely to be genuinely implemented. Moreover, this essay also advances the proposition that UK DPAs are even more unlikely to effectively address corporate economic crime because they are fundamentally constrained in their ability to do so. First, this is because its ability to enhance the accountability of corporate offenders for economic crime is intrinsically dependent on the threat of prosecution, which is, at present, weak. Second, its ability to change corporate culture through effecting genuine improvements in compliance practice is hamstrung by legitimacy issues arising from its structural focus on companies and other similar organisations. Therefore, in conclusion, this essay submits that the UK DPA regime is unlikely to be an effective tool in addressing economic crime perpetrated by companies.

150

ARTICLES: WRITTEN MOOT 2

Evaluating Deferred Prosecution Agreements in the Context of Singapore Tan Yann Xu, National University of Singapore Introduction Deferred Prosecution Agreements (“DPAs”) are agreements between the Public Prosecutor and any of (i) a body corporate, (ii) a limited liability partnership, (iii) a partnership or (iv) an unincorporated association,1 whereby the Public Prosecutor agrees to suspend2 and eventually discontinue3 prosecution of a corporate entity if it complies with specific conditions. Singapore, adopting the United Kingdom (“UK”) position, has chosen to confine the application of the DPA to the context of corporate crimes, although historically, DPAs were originally implemented in the United States (“US”) through the Speedy Trial Act4 as “alternative solutions to rehabilitate individuals charged with non-violent offences and other low-level crimes”.5 Nevertheless, even in the U.S., the use of DPAs in the context of corporate crime has “become so prevalent that the former chief of the Department of Justice (“DOJ”)’s criminal division, Lanny Breuer, referred to these agreements as a ‘mainstay of white-collar criminal law enforcement’”.6 The question whether DPAs would benefit the Singaporean public at large7 was answered by 1 Criminal Procedure Code (Cap 68, 2012 Rev Ed) (“CPC”), s 149D(1). 2 CPC, s 149C(a), where a discharge not amounting to an acquittal is given. 3 If the conditions prescribed under CPC, s 149I(3)(a), (b) and (c) are satisfied, the High Court may, on the application of the Public Prosecutor, grant the subject a discharge amounting to an acquittal in relation to the alleged offence. 4 88 Stat 2080. 5 Eunice Chua, ‘Deferred Prosecution Agreements in Singapore?’ (Singapore Law Blog, 30 January 2018) <http://www. singaporelawblog.sg/blog/article/205> accessed 23 June 2019. 6 Paola C Henry, ‘Individual Accountability for Corporate Crimes after the Yates Memo: Deferred Prosecution Agreements & Criminal Justice Reform” (2016) 6 Am U Bus L Rev 153, 157. 7 Ministry of Law, Strengthening Singapore’s Criminal Justice Framework – Implementation of Criminal Justice Reform Act 2018 and Evidence (Amendment) Act 2018

Parliament in the affirmative when the Criminal Justice Reform Act 2018 (No. 19 of 2018) was passed on 19 March 2018. The relevant provision8 came into effect from 31 October 2018, with the stated rationale being, among other things, to “better [equip] Singapore to investigate large-scale, complex corporate crimes, and bring the truly culpable individuals to justice”,9 as well as to “[reduce] the risk of future offending by [a] corporation”10 through the imposition of conditions under the DPA framework mandating the relevant company to undertake corporate reform measures. However, it remains to be seen exactly how the DPA framework will work in practice given that there has as yet been no reported case dealing with the implementation of a DPA locally.11 A comparison with the UK and US experience with DPAs may shed some light on its operational aspects, and may assist in evaluating Singapore’s DPA framework, which has clearly borrowed from both jurisdictions. This comparison helps address the following interconnected questions: (i) Are the stated rationales for the Singapore DPA framework defensible? (ii) Does the legislative regime give due regard to the principle of separation of powers? (iii) Is this initiative consistent with upholding and enforcing the rule of law? While the dearth of local experience with DPAs means that the answers to these questions have Provisions <https://www.mlaw.gov.sg/content/minlaw/ en/news/press-releases/Strengthening-Singapore-CriminalJustice-Framework-Implementation-of-Criminal-JusticeReform-Act-2018-and-Evidence-Amendment-Act-2018Provisions.html> accessed 23 June 2019. 8 Criminal Justice Reform Act 2018 (No. 19 of 2018), s 35. 9 Supra note 7. 10 ibid. 11 The relevant legislation has only been in effect for less than 10 months as at the time of writing this article.

151

ARTICLES: WRITTEN MOOT 2

yet to be fully explored, the position adopted here is that Singapore’s DPA framework is defensible. This is because the key advantage of the enforcement flexibility conferred on the Prosecution by DPAs outweigh the corresponding disadvantages, which may include potentially compromising the principle of the separation of powers and rule of law and transparency concerns. On deeper analysis, these disadvantages may in fact be mitigated by the legislative framework. Firstly, the primary issue to be addressed is the difficulty of bringing individuals party to corporate crime to justice. The DPA is a useful augment to the tools at the Prosecution’s disposal and may, in difficult cases, be a practical solution. This is because it grants enforcement flexibility to the Prosecution, who might then escape from the binary choice of whether to pursue criminal charges or not.12 This therefore avoids certain inherent disadvantages associated with the criminal prosecution of corporate entities, which may range, inter alia, from the ‘Arthur Andersen effect’13 to the unnecessary punishment of innocent stakeholders like employees and shareholders of the corporate entity.14 Secondly, the Singapore legislative regime, somewhat similar to the UK, provides for a far greater role for the Judiciary than in the US. As a result, there is far less concern that the Public Prosecutor could intrude on the Judiciary’s purview. While the separation of powers could therefore be preserved, much depends on the operation of DPAs in practice. Thirdly, arguments raising the rule of law

152

12 Mike Koehler, ‘Measuring the Impact of NonProsecution and Deferred Prosecution Agreements on Foreign Corrupt Practices Act Enforcement’ (2015) 49 U.C. Davis Law Review 497, 500 (in the US context). 13 Ellis Martin, ‘Deferred Prosecution Agreements: Too Big to Jail and the Potential of Judicial Oversight Combined with Congressional Legislation’ (2013) 18 NC Bank Inst 457, 462-463; See below section 2 for discussion. 14 Singapore Parliamentary Debates, Official Report (19 March 2018) vol 94 (Indranee Rajah, Senior Minister of State for Finance and Law).

concern that using DPAs for large-scale corporate crime may create the impression of a “two-tier criminal justice system” and thereby undermine the deterrent effect of the law15 should be considered against the key advantage of enforcement flexibility. Depending on the circumstances of the case, the deterrence argument may even be wholly misconceived. Furthermore, while the rule of law concern may arguably apply in greater force to Singapore because of the relative opacity as regards the operation of DPAs,16 this concern is mitigated somewhat by the strong checking role played by the Judiciary as provided under the legislative framework. However, care must be taken to avoid sweeping justifications. There is academic commentary which suggests that countries such as the UK17, Canada18 and Australia19 have justified their positions through “the same flawed and hollow rhetoric the US has used to justify alternative resolution vehicles – namely that alternative resolution vehicles are necessary to avoid the perceived ‘Arthur Andersen effect’ and achieve deterrence”.20 Many similar arguments prefaced the introduction of Singapore’s DPA regime. While Singapore’s position may be defensible, there are legitimate concerns which may not have been fully addressed. These include the fear of inadequate judicial supervision or lack 15 Matt Trome, ‘UK Bribery Prosecutions and the Rule of Law’ (The Global Anticorruption Blog, 24 August 2017) <https://globalanticorruptionblog.com/2017/08/24/guestpost-uk-bribery-prosecutions-and-the-rule-of-law/> accessed 23 June 2019 ; See also Ellis Martin, ‘Deferred Prosecution Agreements: Too Big to Jail and the Potential of Judicial Oversight Combined with Congressional Legislation’ (2013) 18 NC Bank Inst 457, 468-469; See also Chen Siyuan and Eunice Chua, ‘Comment: 2018 Changes to the Evidence Act and Criminal Procedure Code’ (2018) 30 SAcLJ 1064 at [31]. 16 Chen Siyuan and Eunice Chua, ‘Comment: 2018 Changes to the Evidence Act and Criminal Procedure Code’ (2018) 30 SAcLJ 1064 at [36]. 17 Crime and Courts Act 2013 (c 22) (UK) Sch 17. 18 Criminal Code, RSC 1985, c. C-46, Part XXII.1, “Remediation Agreements”. 19 Crimes Legislation Amendment (Combatting Corporate Crime) Bill 2017, Schedule 2. 20 Mike Koehler, ‘Measuring the Impact of NonProsecution and Deferred Prosecution Agreements on Foreign Corrupt Practices Act Enforcement’ (2015) 49 U.C. Davis Law Review 497, 564.

ARTICLES: WRITTEN MOOT 2

of transparency regarding the exercise of prosecutorial discretion in implementing DPAs. In the final analysis, Singapore may benefit from minor tweaks to the existing DPA framework, in order to better meet the demands of public interest. 1.

Singapore’s DPA Framework

A few basic points regarding Singapore’s DPA framework are reproduced here to contextualise the discussion. DPAs are entered into between the Public Prosecutor and a “subject”, who may be a body corporate, a limited liability partnership, a partnership or an unincorporated association.21 Individuals may not be the subject of a DPA.22 The range of offences in respect of which a DPA may be entered into is set out in the Sixth Schedule to the Criminal Procedure Code; this is somewhat less extensive than the range of offences to which a UK DPA is applicable.23 A “person” has the choice whether to enter into a DPA or not,24 and such DPA comes into effect only pursuant to a High Court declaration. The Public Prosecutor will have to apply to obtain the High Court’s declaration via criminal motion,25 and such a declaration can only be obtained if the High Court is of the view that the DPA is in the interests of justice,26 and that the terms of the DPA are fair, reasonable and proportionate.27 Once a DPA is approved by the court, the Public Prosecutor must give public notice of such DPA, the corresponding declaration, and any reasons given by the High Court.28 This public notice may be postponed by the High Court if it appears that the postponement is necessary “to avoid substantial risk of prejudice to the administration of justice” in, inter alia, 21 22 23 2. 24 25 26 27 28

CPC, s 149D(1). ibid. Crime and Courts Act 2013 (c 22) (UK) Sch 17, part CPC, s 149B(4). CPC, s 149F(1). CPC, s 149F(1)(a). CPC, s 149F(1)(b). CPC, s 149F(5).

any investigation under the Criminal Procedure Code.29 If the Public Prosecutor believes that the subject that entered into a DPA has failed to comply with the terms thereof, he may apply to the High Court to terminate the DPA. The Public Prosecutor and the subject of the DPA may at any time when a DPA is in force agree to vary the terms of the DPA30 Any such variation only takes effect when the High Court approves the variation through a declaration, such declaration only being made if the variation is in the interests of justice and the terms of the DPA as varied are fair, reasonable and proportionate.31 The requirements as to the giving of public notice by the Public Prosecutor of the DPA applies here as well, subject to any prohibition pursuant to any written law or order of the High Court. The following decisions of the High Court under the DPA regime are appealable: (i) a decision not to approve a DPA, (ii) a decision that the subject has failed to comply with the terms of the DPA, (iii) a decision that the subject did not fail to comply with the terms of the DPA and (iv) a decision not to approve a variation of the terms of a DP.A32 Only the Public Prosecutor may appeal against a decision in (i) or (iv).33 2. Scrutinising the rationale underlying the DPA regime in the Context of Singapore As noted, the Singapore DPA framework is applicable only to corporate entities. This reflects the use of DPAs in the UK context, from which Singapore adapted many of its provisions, and in the US context34 (at least since the global financial crisis of 2008) where DPAs have increasingly been used to “defer prosecuting corporations that commit crimes 29 30 CPC, s 149H(1). 31 CPC, s 149H(2). 32 CPC, s 149M(1). 33 CPC, s 149M(2). 34 Crime and Courts Act 2013 (c 22) (UK) Sch 17, paras 1(1) and 4(1).

153

ARTICLES: WRITTEN MOOT 2

in exchange for large fines paid to the federal government, and sometimes agreements to be monitored for a set period of time”.35 There are various theoretical justifications for the introduction of a DPA regime in Singapore, but assessing the cogency of such proffered justifications is the first step to evaluating its likely impact (or lack thereof), and how the questions posed above36 may be answered. The common thread running through the various justification is that of enforcement flexibility. Much turns on the idea that liberating the Prosecution from the confines of the traditional criminal justice system would lead to better outcomes – in terms of avoiding “lengthy and costly trials without unduly compromising the rule of law, and hopefully leading to a positive change in corporate behaviour”.37 The “Arthur Andersen effect” In her Second Reading speech on the Criminal Justice Reform Bill 2018, the former Senior Minister of State for Law Indranee Rajah38 asserted that the DPA tool “is particularly useful where prosecution of a corporate offender unnecessarily punishes not just the management, but also the employees and shareholders of the corporate entity for what might be the acts of a small number of employees or officers of the company”.39 This is redolent of what has been termed the “Arthur Andersen effect”, named after the collateral consequences suffered by Enron’s long-time auditor Arthur Andersen LLP in the aftermath of charges brought by the US DOJ.40 Arthur Andersen LLP was charged with obstruction of justice and was criminally

154

35 Paola C Henry, “Individual Accountability for Corporate Crimes after the Yates Memo: Deferred Prosecution Agreements & Criminal Justice Reform”, 6 Am U Bus L Rev 153 (2016) at 157. 36 See Introduction. 37 Supra note 16 at [29]. 38 Currently Minister in the Prime Minister’s Office and Second Minister for Finance and Education. 39 Supra note 14. 40 Larry Thompson, ‘Deputy Attorney General, Transcript, News Conference — Arthur Andersen Indictment’ (Mar. 14, 2002), available at <http://www.justice.gov/archive/ dag/speeches/2002/031402newsconferncearthurandersen. htm> accessed 23 June 2019.

convicted in 2002. The collateral consequences suffered by the business included “the loss of its certified public accounting license and the resulting inability to audit public companies”, which led to “Arthur Andersen [laying] off thousands of employees and effectively [going] out of business in 2002”.41 Even in the US, however, it has been recognised that the “Arthur Anderson effect”, which is the DOJ’s primary policy justification underpinning the DPA regime, has been debunked. This view is attributed to Gabriel Markoff who found that “much in opposition to the warnings of extreme collateral consequences that are continually repeated in both the popular and academic literature[,] no publicly traded company went out of business as the result of a federal criminal conviction in the years 2001 to 2010”.42 Markoff ’s study certainly does not invalidate the former Senior Minister of State for Law’s stated rationale. Just because a company does not go out of business does not mean innocent stakeholders such as employees and shareholders do not suffer from criminal prosecution, for example because of a corresponding fall in share price. This, however, does not resolve the issue, which is whether preferring DPAs to criminal convictions is supportable by evidence that employees and shareholders would needlessly suffer otherwise. Fortunately, Singapore has not yet faced the problem of an “excessive” use of DPAs, a problem prevalent in the context of the US DOJ’s use of alternative resolution vehicles. This may, however, be a function of the fact that the legislation has only just come into force, which also means that evidence of any deficiencies in the use of DPAs would not 41 Supra note 12, 501-502. 42 Mike Koehler, ‘Measuring the Impact of NonProsecution and Deferred Prosecution Agreements on Foreign Corrupt Practices Act Enforcement’ (2015) 49 U.C. Davis Law Review 497, citing Gabriel Markoff, ‘Arthur Anderson and the Myth of the Corporate Death Penalty’ (FCPA PROFESSOR, 12 Aug 2012) <http://www.fcpaprofessor.com/arthurandersonand-the-myth-of-the-corporate-death-penalty> accessed 23 June 2019.

ARTICLES: WRITTEN MOOT 2

yet be forthcoming. Depending on the way in which the Singapore Public Prosecutor wields this new discretionary power, the DPA could prove greatly beneficial to reduce the collateral and unintended consequences of criminal prosecution. The difficulty in prosecuting corporations In his speech supporting the Criminal Justice Reform Bill 2018, the Member of Parliament for the Holland-Bukit Timah GRC Mr Christopher de Souza suggested that the DPA framework “seeks to resolve the difficulties in punishing the legal personality or the separate legal personality of a company”, and that “[i] t does this by facilitating investigations so that there can be a more meaningful attribution of culpability”.43 This must certainly be true given the difficulties inherent in proving corporate liability. Criminal offences generally require the Prosecution to prove beyond reasonable doubt (a higher standard of proof than in civil cases) that the company (i) committed the act which is prohibited by the relevant offence (the actus reus), and (ii) that the company had the requisite guilty mind (mens rea; the specific mens rea to be proven would depend on the nature of the relevant offence). However, the doctrine of the separate legal personality of companies44 means that the law had to develop an approach to attribute the relevant mens rea of company employees (or other relevant company officers or agents) to the corporate entity – this was termed the “identification doctrine”.45 Neil Campbell and John Armour describe the scope of the identification doctrine thus:46 As a matter of precedent, the identification 43 Singapore Parliamentary Debates, Official Report (19 March 2018) vol 94 (Christopher de Souza, Member of Parliament for Holland-Bukit Timah GRC). 44 This is the fundamental principle of independent corporate identity laid down in Aron Salomon (Pauper) v A Salomon and Company Limited [1897] AC 22, 51. 45 Lennard’s Carrying Co Ltd v Asiatic Petroleum Co Ltd [1915] AC 705, 713-714; Tesco Ltd v Nattrass [1972] AC 153, 170-171. 46 N Campbell & J Armour, ‘Demystifying the Civil Liability of Corporate Agents’ (2003) 62 CLJ 290, 295.

principle was developed solely to attribute the actions or knowledge of corporate agents to a company. Clearly the Prosecution would face evidentiary obstacles in establishing the requisite mens rea due to collective decision-making and complex management structures, which may insulate the relevant decision-makers from claims. Therefore, the enforcement flexibility which DPAs promise would serve as a practical addition to the Prosecution’s arsenal, as a means to encourage cooperation with the Prosecution especially as regards investigations with the object of holding individual offenders to account. In this respect, the Member of Parliament for Sembawang GRC Mr Vikram Nair correctly identified the commercial incentives that the DPA regime intends to promote: “[DPAs are] basically to support and encourage whistle-blowing”.47 Resolving systemic corporations

problems

It is arguably a key advantage of the DPA regime, relative to criminal prosecution, that the Prosecution may impose conditions (which subject corporations must comply with) which may oblige the subject corporation to implement or make changes to a compliance programme.48 The objective is to prevent repeat offending by, among other things, reforming corporate culture so that the circumstances which led to the alleged offence in respect of which the DPA was first entered into would be rectified. The criticism of corporate compliance systems is that corporate management merely use them for “external perception management or ‘window dressing’ – a symbolic or halfhearted implementation of compliance systems which business hopes regulators and other stakeholders will take as delivering truly effective compliance 47 Singapore Parliamentary Debates, Official Report (19 March 2018) vol 94 (Vikram Nair, Member of Parliament for Sembawang GRC). 48 CPC, s 149E(3)(e).

155

ARTICLES: WRITTEN MOOT 2

– to avoid making substantive changes to company practices”.49 In other words, corporate management usually does not sufficiently value compliance to effectively and substantively implement compliance; furthermore, regulators (or for our purposes, the Public Prosecutor) may not have the expertise or requisite information to monitor proper adherence to the prescribed conditions. This criticism is somewhat mitigated by Singapore’s legislative framework. Under CPC, s 149G(1), the Public Prosecutor has the discretion to apply to the High Court to terminate the DPA if he believes that the subject corporation has failed to comply with the terms of the DPA. Pursuant to CPC, s 149G(2), the Public Prosecutor must prove on a balance of probabilities that the subject corporation has breached the DPA, thus making it easier to enforce compliance measures. In fact, even in the context of the US, where judges are disinclined to strike down DPA settlements50 corporate compliance efforts have been argued to have facilitated “unequivocally, far greater accountability for corporate wrongdoing”.51 Furthermore, the DPA regime has an in-built flexibility as regards monitors appointed under the relevant compliance programmes; as the former Senior Minister of State for Law stated, “[t]he conditions that can be attached to a DPA are deliberately left flexible, subject to the High Court’s approval” and that “[i]t is open to parties to negotiate and set out details concerning the role and supervision of monitors in the DPA”.52 Presumably, the DPA regime would also be

156

49 Christine Parker and Vibeke L Nielsen, ‘Corporate Compliance Systems Could They Make Any Difference?’ (2009) 41 Administration & Society 3, 9-10. 50 To the point where they have been accused of “rubber-stamping” such settlements; see Corruption Watch, ‘Out of Court, Out of Mind: do Deferred Prosecution Agreements and Corporate Settlements Fail to Deter Overseas Corruption’ (Corruption Watch, 2016), 15. 51 Lanny A. Breuer, U.S. Dep’t of Justice, Address at the New York City Bar Association (Sept. 13, 2012) < https:// www.justice.gov/opa/speech/assistant-attorney-general-lannybreuer-speaks-new-york-city-bar-association> accessed 23 June 2019. 52 Supra note 14.

open to ensuring that monitors appointed under a DPA be subject to oversight by the Court53 or imposing strict reporting obligations in order to increase enforcement effectiveness. Increasing the enforcement accorded to the Prosecution

flexibility

That the Prosecution would benefit from an increase in enforcement flexibility undergirds the other stated rationales for introducing the DPA regime. As has already been noted, the Prosecution previously had a binary choice: either charge the relevant corporation, or do not. But this typified an all or nothing regime wherein the Prosecution had to essentially balance between the possible collateral consequences affecting innocent employees, shareholders and/or management54 against the possible injustice of letting corporations escape prosecution. In the most extreme situation (ie, where the “Arthur Andersen effect” would materialise), the need for enforcement flexibility would be incontrovertible. However, given that the chances of such extreme collateral consequences occurring may have been overstated,55 the question is whether enforcement flexibility still confers concrete advantages. Ellis Martin summarised the advantages of increased flexibility as follows:56 Fundamentally, DPAs arm prosecutors with formidable weapons to ensure corporate compliance without disastrous collateral consequences. Under the appropriate circumstance, DPAs can correct and restore a company’s operating practice and preserve its viability into the future. While DPAs do not come with the stigma of a criminal indictment, 53 Thus addressing Mr Murali Pillai’s concern raised during the Parliamentary Debates, that monitors may fall afoul of a moral hazard, “in that the external monitor may have a financial interest in having an expanded scope of work”; Singapore Parliamentary Debates, Official Report (19 March 2018) vol 94. 54 Supra note 39. 55 Supra note 42. 56 Supra note 13, 471.

ARTICLES: WRITTEN MOOT 2

they allow prosecutors to achieve many of the same goals while leaving out the problems inherent with criminal indictments. The flexibility rationale is also supported by the fact that, unlike criminal fines, the financial penalty that may be imposed under a DPA is not subject to a statutory maximum. One reason for this, as expressed by the former Senior Minister of State for Law, is that “the harm brought about by corporate offending may eclipse any pre-determined statutory maximum that may be prescribed”.57 DPAs therefore increase the options available to the Prosecution so that they may tailor a more nuanced response to the unique circumstances of individual cases. This is not solely dependent on collateral consequences arising, but is merely “particularly useful” in such a situation.58 In this respect, the “Arthur Andersen effect” does not exhaust the range of collateral consequences which may materialise. As has been observed, prosecution may also result in serious disruption to the business of a company and to its constituencies. This may also potentially make businesses overly cautious, deterring “appropriate and needed entrepreneurial spirit and risk taking”.59 To reiterate the point, the criminal justice system alone may not appropriately address the legal and practical difficulties associated with prosecuting corporate entities.60 In the absence of evidence about the success or lack thereof of the DPA regime in Singapore, an evaluation of its potential impact must therefore address the various general concerns raised against such a policy.

57 Supra note 14. 58 Supra note 39. 59 Pamela H. Bucy, ‘Corporate Criminal Liability: When Does it Make Sense?’ (2009) 46 Am. Crim. L. Rev. 1437, 1440 60 Supra note 12.

3. The Usefulness of DPAs – Balancing Advantages and Disadvantages We turn to balancing the advantages of Singapore’s DPA regime, which stem ultimately from enforcement flexibility, against the various concerns which have been raised in the literature. The DPA framework has been said to reflect an inconsistent application of laws to large corporations, inadequate judicial supervision, the encouragement of irresponsible behaviour by corporations, a lack of transparency regarding the exercise of prosecutorial discretion and undermining the deterrent effect of the law. In this respect regard must be had to: (i) how specific legislative provisions attempt to mitigate these issues, (ii) how such provisions are intended to operate in practice and (iii) Singapore’s constitutional culture in general. The Rule of Law Issue – a “two-tiered criminal system”? The DPA regime running contrary to the enforcement of the rule of law against both individuals and corporate entities has been considered a “key issue”61 in the United States. The concern is that offering DPAs (and the similar Non-Prosecution Agreements (“NPAs”)) solely to corporate offenders “makes it appear that companies can buy their way out of the justice system in a way that no other person can”. As Senator Elizabeth Warren puts it, the failure to prosecute corporations and their executives “has a corrosive effect on the fabric of democracy and our shared belief that we are all equal in the eyes of the law”.62 A study of enforcement actions under the US Foreign Corrupt Practices Act (“FCPA”) by Mike Koehler bears out these fears in the 61 Corruption Watch, ‘Out of Court, Out of Mind: do Deferred Prosecution Agreements and Corporate Settlements Fail to Deter Overseas Corruption’ (Corruption Watch, 2016), 1112. 62 ibid, citing Senator Elizabeth Warren, ‘Rigged Justice: How Weak Enforcement lets Corporate Offenders Off Easily’ (January 2016) <http://www.warren.senate.gov/files/ documents/Rigged_Justice_2016.pdf> accessed 23 June 2019.

157

ARTICLES: WRITTEN MOOT 2

US context. Since the US DOJ first utilised alternative resolution vehicles (DPAs and NPAs) in December 2004, there have been 84 criminal FCPA enforcement actions against business organisations in the period spanning 2004 to 2014, 70 of which involved an alternative resolution vehicle.63 Of the 84 criminal FCPA enforcement actions, 64 (76%) of these enforcement actions did not involve any related criminal prosecution of company employees. Most significantly, this gap between criminal FCPA enforcement actions against business organizations and related individual enforcement actions did not exist prior to the DOJ introducing NPAs and DPAs to the FCPA context, where about 83% of enforcement actions against business organisations did involve related criminal prosecutions of company employees from the period spanning 1977 (when the FCPA was enacted) to December 2004.64 The fear is that Singapore may experience the same result as in the US – that DPAs granted for corporate crimes would not been accompanied with any charges against responsible individual employees (or other responsible officers or agents),65 thus becoming a vehicle by which large corporations may game the system, resulting in a “two-tiered criminal system”.66 This may have the unfortunate effect of encouraging irresponsible corporate behaviour since a DPA will always be available to escape prosecution. Parliament has noted this. As Mr Christopher de Souza stated during the debate on the Criminal Justice Reform Bill 2018, “it is important that the introduction of the deferred Prosecution Agreement does not create the misperception that laws will not be consistently applied”.67 Fundamental rule of law concerns vs potential rule of law concerns

158

63 Supra note 12, 521. 64 Supra note 12, 541. 65 Supra note 35, citing Brandon L. Garrett, Too Big to Jail (1st ed. 2014), 295. 66 Supra note 15. 67 Supra note 43.

At this point we must carefully distinguish fundamental rule of law concerns from potential rule of law concerns. The DPA regime does allow corporate entities to escape prosecution in return for compliance with conditions which may include a corporate reform programme or furnishing evidence of individual wrongdoing. However, Senator Warren’s concern applies only insofar as the DPA regime allows individual wrongdoing executives to escape prosecution, and is not so convincing when applied to the separate legal personalities of corporate entities.68 In this respect, some academics have even suggested that “there is no theoretical justification for corporate criminal liability”69 and that it is “logically incoherent to attribute moral responsibility to corporations”.70 While we may avoid becoming mired in the philosophical underpinnings of corporate criminal liability, we cannot escape the intuitive appeal that the corporation as a separate legal personality does not have moral agency. Adopting this perspective would suggest that there is no fundamental rule of law concern to allowing corporations to escape prosecution through the DPA regime, 68 In this respect, it is certainly an interesting legal question as to whether Article 12(1) of the Singapore Constitution, which states that “[a]ll persons are equal before the law and entitled to the equal protection of the law” applies to corporate entities as well as individuals, especially since section 2 of the Interpretation Act (Cap 1, 2002 Rev Ed) defines “person” to include “any company or association or body of persons, corporate or unincorporate”. However, in interpreting Article 12(1) we must be sensitive to its context: Article 12 appears in Part IV of the Constitution which enumerates Fundamental Liberties as applicable to individuals, for example, Article 9(1) which protects the “life and personal liberty” of persons, Article 10(1) which prohibits slavery of persons and Article 12(2) which prohibits certain grounds of discrimination against “citizens of Singapore”. These provisions make sense only if we confine the application of certain Part IV liberties to natural persons. In any case, while it may be plausible to interpret “person” in Article 12(1) to include corporate entities, this would likely not be of any substantive help to any person attempting to impugn the DPA regime given that the applicable test for constitutionality under Article 12(1) is the “reasonable classification test” (Public Prosecutor v Taw Cheng Kong [1998] 2 SLR 410; Lim Meng Suang v Attorney-General [2015] 1 SLR 26 at [57]), which has been described only as a “threshold test” (Lim Meng Suang v Attorney-General [2015] 1 SLR 26 at [62]). 69 John Hasnas, ‘The Centenary of A Mistake: One Hundred Years of Corporate Criminal Liability’ (2009) 46 Am. Crim. L. Rev. 1329, 1329. 70 ibid at 1331.

ARTICLES: WRITTEN MOOT 2

because there is no “person” whose moral agency justifies the equal subjection to the rule of law. Yet the potential rule of law concern still exists, in the form of allowing the individual controllers of the corporate entity (or the individual wrongdoers, which may be employees or other agents of the corporation) to escape prosecution. As Koehler’s study shows, this poses a serious problem in the US where lack of information regarding how Federal Prosecutors really exercise their discretion forces academics to construct a “working hypothesis” justifying why so few criminal charges are brought against individuals.71 But it is a fallacy to suggest that the same issues do not arise in the absence of DPAs. In fact, it has already been noted that a key rationale for introducing the DPA regime is the evidential difficulty (among other things) in prosecuting corporations.72 It is plausible that in the absence of the DPA regime, wrongdoing individuals would have at least the same chance and possibly a greater chance of escaping criminal prosecution. Lack of transparency in the exercise of prosecutorial discretion One problem stemming from a “two-tiered criminal system” is that corporations may be incentivised to game the system by paying their way out of tight situations. Depending on one’s perspective, however, the non-publication of prosecutorial guidelines73 as regards the use of DPAs74 may actually solve this potential issue, by ensuring that “prosecutorial guidelines do not become a tool for criminals to refer to in manipulating the criminal justice system 71 Supra note 63. 72 This being mitigated by the flexibility given to the Prosecution to impose conditions on a subject corporation pursuant to CPC, s 149E(3). 73 Unlike the UK’s Code of Practice, <https://www. cps.gov.uk/publication/deferred-prosecution-agreementscode-practice> accessed 23 June 2019. 74 CPC, s 149B read with CPC, s 11(1) and Singapore Constitution, Article 35(8).

to escape punishment”.75 Indeed, this was expressed to be a deliberate departure from the UK’s position where a Code of Practice for the use of DPAs is published. In this respect, Singapore departs from the US practice as well, where informal guidelines on the use of DPAs are periodically promulgated.76 This lack of transparency becomes problematic when considering the tremendous protection already accorded prosecutorial discretion, as pointed out by Chen and Chua.77 This is because Article 35(8) of the Singapore Constitution, which states that “[t]he AttorneyGeneral shall have power, exercisable at his discretion, to institute, conduct or discontinue any proceedings for any offence”, has been interpreted to confer a constitutional status on the prosecutorial function which is coequal with that of, inter alia, the Judiciary.78 Hence, “neither may interfere with each other’s functions or intrude into the powers of the other”.79 Furthermore, challenging the exercise of prosecutorial discretion entails discharging a “near impossible burden”80 of establishing a prima facie case that the discretion was exercised arbitrarily or in bad faith.81 But the solution to an unduly broad prosecutorial discretion may not be so simple as publishing DPA guidelines in the same vein of the UK. Not only may it trade opacity for a greater risk that corporations would game the system, a set of guidelines may actually undermine the legitimacy of the whole legislative framework (and result in confusion) if applied improperly or if departed from without good reason. This 75 Supra note 14. 76 Supra note 16 at [32]. 77 ibid at [30]. 78 The judicial power of Singapore is vested into the Supreme Court by virtue of the Singapore Constitution, Article 93. 79 Law Society of Singapore v Tan Guat Neo Phyllis [2008] 2 SLR(R) 239 at [144]; see also Ramalingam Ravinthran v AttorneyGeneral [2012] 2 SLR 49 at [44]-[74]. 80 Supra note 77. 81 Ramalingam Ravinthran v Attorney-General [2012] 2 SLR 49 at [17], citing Law Society of Singapore v Tan Guat Neo Phyllis [2008] 2 SLR(R) 239 at [149].

159

ARTICLES: WRITTEN MOOT 2

was the issue experienced in the UK in Serious Fraud Office v Rolls-Royce plc82 where the Director of the Serious Fraud Office (“SFO”) applied for a DPA despite the fact that the investigation was “not triggered by a self-report”.83 In fact, the SFO was only notified in early 2012 by internet postings raising concerns about the operation of Rolls-Royce’s civil business in China and Indonesia,84 and evidence suggested that Rolls-Royce had known about such questionable conduct since 2010. Arguably this was a departure from the promulgated Code of Practice,85 and this departure was only reinforced by Sir Brian Leveson P’s approving the DPA application, on account of RollsRoyce’s purported “extraordinary” assistance.86 As a result, there have been criticisms that the UK’s DPA regime has been “significantly weakened by the Rolls Royce precedent”.87 The UK experience therefore shows that reliance on published guidelines is not a perfect answer, and may have serious drawbacks. In Singapore, the crucial check on prosecutorial discretion is provided by the High Court’s oversight regarding entering,88 varying89 or terminating the DPA.90 Under this framework,91 the High Court only approves entering into or varying DPAs if it is satisfied that the DPA satisfies the interests of justice and the terms of the DPA are fair, reasonable and proportionate.92 This ensures a substantive supervision of prosecutorial discretion. Furthermore, transparency concerns may be mitigated by the requirement that public notice be given of relevant court approvals, DPA expiries and DPA breaches unless prohibited by

160

82 [2017] Southwark Crown Court Case No. U20170036. 83 ibid at [22]. 84 ibid at [16]. 85 Serious Fraud Office, Deferred Prosecution Agreements Code of Practice, para 2.9.2. 86 Supra note 83. 87 Corruption Watch, ‘A Failure of Nerve: The SFO’s Settlement with Rolls Royce’ (Corruption Watch, 19 January 2017). 88 CPC, s 149F(3). 89 CPC, s 149H(3). 90 CPC, s 149G(3). 91 Which is in pari materia with the UK position. 92 CPC, s 149F(1).

law or by an order of the High Court.93 It is appreciated that legislative mechanisms may give way to the culture of judicial deference we see manifested in the constitutional94 and administrative law context,95 such that the proper functioning of the DPA regime may rest entirely with the Public Prosecutor. Ultimately, only time will tell if court oversight would adequately check prosecutorial discretion. Separation of powers Since the legislative framework expressly provides for court oversight over DPAs in various respects,96 the argument that “giving the prosecutor the ability to essentially find guilt and impose a sentence”97 results in an unacceptable compromise of the separation of powers holds less weight in the Singaporean context. While much relies on how the court exercises its powers in practice,98 the intention at this early stage of implementation is that “in assessing the proportionality of the conditions proposed, the High Court will balance the extent of the wrongdoing, including the revenue or profits attributable to the wrongdoing, with the ability of the corporation to comply”.99 This contemplates a substantive role for the Judiciary, who “should not grant or approve the DPA unless it has met [the requisite] threshold”;100 furthermore, even if approval is granted, such decision is appealable.101 93 See CPC, ss 149F(5), 149G(4) and (5), 149H(4) and (5), 149I(1)(a)(ii) read with 149J(1). 94 Li-ann Thio & Kevin YL Tan eds., The Evolution of a Revolution: 40 Years of the Singapore Constitution (Routledge Cavendish, 2009) 221. 95 The “green-light approach”, CJ Chan, ‘Judicial Review – From Angst to Empathy’, (2010) 22 SAcLJ 469, 480. 96 See previous section; this adapts the UK position and may be contrasted with the US position. 97 Supra note 16 at [31]. 98 There is always the possibility that judges defer to a large extent to prosecutorial discretion so that judges simply “rubber-stamp” DPAs brought before them. 99 Supra note 14, scoping out the contemplated operation of CPC, s 149F(1). 100 Ie, the threshold under CPC, s 149F(1) that approval will only be granted if the DPA is in the interests of justice, and its terms are fair, reasonable and proportionate. 101 CPC, s 149M(1).

ARTICLES: WRITTEN MOOT 2

Undermining the deterrent effect of the law Those suggesting that the deterrent effect of the law is compromised by DPAs usually assert that (i) any financial penalty could not plausibly be of any significance to corporations, especially large ones102 and/or (ii) that the moral condemnation aspect of a criminal conviction would be lost.103 As regards the first assertion, one may counter-argue that the flexibility inherent in the Singaporean DPA framework could plausibly allow for a financial penalty large enough to matter (subject to court approval). Ultimately, this boils down to the interaction between court and Prosecution. If the court discharges its oversight function properly, it will assess a financial penalty commensurate with the magnitude of the wrongdoing.104 It will accordingly find the need for deterrence if doing so is in the interests of justice.105 As regards the second assertion, it is a mistake to attribute the moral responsibility that comes with a criminal prosecution to a corporation.106 Granted, moral condemnation as a tool of deterrence matters to a corporation insofar as it is translated into financial penalty,107 but beyond that, it is too much of a blunt instrument to use against complex corporations comprising innocent and responsible individuals alike. It 102 Constantino Grasso, ‘Peaks and troughs of the English deferred prosecution agreement: the lesson learned from the DPA between the SFO and ICBC SB Plc’ (2016) 5 JBL 388, 394. 103 Supra note 61, 13. 104 Supra note 99. 105 CPC, s 149F(1)(a). 106 Supra note 13, 469; supra note 69. 107 In this respect, Randall Eliason suggests that “Fines and negotiated agreements may be portrayed as business as usual, but it’s hard to spin a federal indictment. The moral condemnation of the community that accompanies a criminal conviction makes these sanctions a uniquely powerful deterrent”, but this may be understood another way: the moral condemnation of the community hurts the corporation economically, and insofar as it does so, deters wrongdoing. it makes no sense to think that a corporation will care for See Randall Eliason, ‘We Need to Indict Them’ Legal Times (22 September 2008) <http://www3.wcl.american.edu/faculty/ eliason/LegalTimes_9_22_08.pdf> accessed 23 June 2019.

makes no sense108 to impose criminal liability on an entity that is “not capable of morally blameworthy behaviour”.109 What is needed is for responsible individuals to be held liable, and it is that which the DPA facilitates. Finally, to reiterate, deterrence is only one aspect of the myriad considerations which the Prosecution must assess. The DPA regime better allows the Prosecution to explore what is in fact a multifaceted inquiry, so that they may give due weight to the collateral consequences affecting key corporate stakeholders, or even to consider the imposition of financial penalties greater than the statutory maximum which may arguably redound more to the public benefit. 4. Conclusion Chen and Chua suggests that the Singapore framework “lies somewhere between the UK and the US”, except for “the exercise of prosecutorial discretion more transparent where the Singapore position is, unlike the US and UK, completely opaque”.110 I therefore consider that while the Singapore DPA regime has many strengths and rests on firm theoretical foundations, minor tweaks increasing the level of transparency surrounding DPAs would protect this regime from its most salient weakness – the lack of any indication on the implementation of DPAs. While the publication of prosecutorial guidelines has its drawbacks,111 complete opacity may not only facilitate abuse of prosecutorial discretion, it may impede the court’s oversight function because of inadequate guidance regarding the factors which the Prosecution must consider. Given the inclination of Singapore’s judiciary to defer to executive authority,112 this may 108 “logically incoherent”, supra note 63. 109 ibid. 110 Supra note 16 at [36]. 111 Supra note 87. 112 For Example, in Lee Hsien Loong v Review Publishing [2007] 2 SLR(R) 453, in the context of judicial review, Sundaresh Menon JC observed that: [96] ... within the span of executive decisions that are immune

161

ARTICLES: WRITTEN MOOT 2

compromise the separation of powers. In this respect, MPs have observed that “making it mandatory for the High Court to publicly justify why it approved a DPA can only help to instill confidence in the system”.113 I agree, but only insofar as it does not disincentivise corporations from entering into such DPAs in the first place.114 While the former Senior Minister of State for Law justified the position that courts are not required to give reasons with reference to the fact that “the DPA is not a judgement[,] [s]o, it does not have a binding effect”115 and that “it is open to the court to give grounds if the court wishes to do so”,116 this, with respect, does not adequately

162

from judicial review are those involving matters of “high policy”. This includes such matters as dissolving Parliament, the conduct of foreign affairs, the making of treaties, matters pertaining to war, the deployment of the armed forces and issues pertaining to national defence. These are what the American courts call “political questions” and the reasons underlying the deference accorded to the executive branch of government in such areas have been articulated in the cases I have referred to. In my judgment, cases concerning international boundary disputes or the recognition of foreign governments comfortably fall within this class of cases... [97]... apart from issues of foreign affairs or national defence, there are other areas that the court will no doubt find unjusticiable. These include cases concerning the interpretation of international treaties operating solely on the international plane, or where the legislature has made it clear that the question is reserved to the executive to answer... [emphasis mine] While in this case section 149F(1) Criminal Procedure Code clearly demarcates the High Court’s purview in terms of assessing whether the DPA is in the interests of justice and whether the terms of the DPA are fair, reasonable and proportionate, the lack of guidance may lead to the conclusion that certain factors considered by the Public Prosecutor would be non-justiciable. 113 Singapore Parliamentary Debates, Official Report (19 March 2018) vol 94 (Ms Sylvia Lim). 114 For example, the former Senior Minister of State for Law has clarified that a key reason that court hearings for the approval of a DPA are in camera is that “if a DPA approval proceedings was held in the public view, and if the DPA did not eventually come into force, the company negotiating the DPA would be prejudiced, or could be prejudiced, in any future criminal proceedings based on the same facts”. If this is the case, and “if companies have reason to fear that their willingness to negotiate a DPA will prejudice their defence later on, then they may be unwilling to volunteer information or cooperate and the benefits of the DPA scheme will then be undermined.” (Supra note 14). In the same vein, any requirement for public justification must be tailored to avoid prejudicing subject corporations. 115 Supra note 14. 116 ibid.

address the public interest (which may apply with greater force) in ensuring the proper discharge of the prosecutorial discretion as against large corporations for crimes with consequences the magnitude of which dwarfs conventional crimes.117 This may warrant departing from the “current practice on judgments”118 where courts have discretion in giving reasons, with the concern about prejudicing subject corporations being met by Chen and Chua’s eminently sensible suggestion that “the grounds of decision of the court can be redacted to remove any identifying information”.119 This minor change may go a long way toward safeguarding the integrity of the DPA regime from both internal and external pressures – and upholding Singapore’s reputation for efficiency and incorruptibility.

117 118 119

Supra note 61, 2. Supra note 14. Supra note 16 at [39].

ARTICLES: WRITTEN MOOT 3

Nothing gold can stay: No better ‘substitute judgment’ against the ‘best interests’ of dying patients Edwin Teong Ying Keat, University of Bristol Introduction “Nature’s first green is gold, her hardest hue to hold. Her early leaf ’s a flower; but only so an hour. Then leaf subsides to leaf. So Eden sank to grief, so dawn goes down to day. Nothing gold can stay.1” In his illuminative poem, Frost reminds us that the essence of life is its evanescence. In the United Kingdom, the statutory ‘best interests’ test for dying patients was conceived to ensure quality living.2 Much debate have centered on what constitutes ‘best interests’. Three main considerations surface. Whose wishes should be prioritized?3 On what basis?4 Why should the Courts be allowed to dictate how dying patients should live?5 This essay argues that the courts should be allowed to decide in the best interests of dying patients for three reasons. Firstly, the Courts themselves have developed the ‘best interests’ test. It follows that they should have the institutional capacity to apply the best interests test. Secondly, the merits of the ‘best interests’ test, chief amongst which, a delicate balancing exercise between autonomous wishes and paternalistic concerns, is best applied by the Courts, an arbiter of concerns between laymen and hospitals. Thirdly, the alternative test - the substituted judgment approach pales in comparison due to its limitations. This essay will make its arguments primarily based on the jurisdiction of the United Kingdom where the ‘best interests’ test has been 1 Robert Frost, ‘New Hampshire’ (Dover Publications 2019) at p. 84 2 In Re A [2000] 1 FLR 549, the patient’s mental, emotional and welfare issues were concerned 3 Nell Munro, ‘Taking wishes and feelings seriously: the views of people lacking capacity in Court of Protection decision-making’, Journal of Social Welfare and Family Law 2014, Vol 36, No.1, pp. 59-75 at p. 60 4 ibid. 5 ibid.

applied and refined. It will make references to the Singaporean jurisdiction where applicable. The limited reference to the Singaporean jurisdiction is informed by the understanding that the Mental Capacity Act6 was only recently took effect in March 2010.7 As such, the impact of this legislation has yet to take root deeply enough to spur debates as regards the layered considerations in the best interests test. 1. Ethical debates We cannot analyse the ‘best interests’ of patients without considering first, the ethical conflicts of this test. Doing so informs an understanding of why the Courts are best placed to apply the ‘best interests’ test. Discussing ‘ethics’ potentially opens a Pandora’s box given the subjective connotations8 of ‘autonomy’ and ‘paternalism’. Yet, a genuine study of bioethics discourages simply accepting one ideology.9 However, establishing certain goal posts is the only way ethical considerations can be meaningfully incorporated into medical law debates.10 Therefore, the ethical debates here will be investigated from the perspective that the policymakers strive toward preserving sanctity of life.11 An objective approach characterized by utilitarianism - effective use 6 Mental Capacity Act (Chapter 177A) 7 Sumytra Menon, ‘The Mental Capacity Act: Implications for Patients and Doctors faced with difficult choices’ Ann Acad Med Singapore 2013; 42, pp. 200-202 at p.200 8 Ian Jennings, ‘Wolf and Christman on autonomy: Two objective views’, South African Journal of Philosophy (2000), Volume 19, pp. 151-167 at p. 152 9 Gerald Dworkin, The theory and practice of autonomy (First published 1988, CUP 2012) at p.5 10 Jonathan Herring, Vulnerable adults and the law (OUP 2016) at p. 6 11 Jukka Varelius, ‘Medicine, Health Care and Philosophy’, Health and Autonomy (2005), Vol 8, pp. 221-230, at p. 221

163

ARTICLES: WRITTEN MOOT 3

of limited resources to preserve life - should therefore be favoured over any deontological approach which may prescribe subjective views onto meaningful living. This prevents what Karlsen and Solbakk deem as imposing moral judgment ‘high and low.’12 Yet, wading into the moral morass is not necessary. Critics cite emotionally charged cases such as Charlie Gard13 base their arguments on Courts as ‘institutions designed to pronounce on criminal cases14’ but eventually concede there is no ‘right’15 moral answer. The allegations regarding the institutional capacity of courts being limited to deciding on criminal cases must be addressed. Firstly, it is a misconception that the Courts are designed to pronounce only on criminal cases. There is no validity in such claims other than progressing a narrow viewpoint of the Court’s institutional capacity. Secondly, given that there is no ‘right’ moral answer and given the presence of emotional interference with judgments made, the Courts as a neutral third party is thus most suited to decide in this respect. Lastly, the reason for adducing a broadsheet article as evidence of ‘critics’ illustrates the extent to which debate on this issue is amplified by invalid arguments but emotionally driven by the mainstream media. In almost every case concerning dying patientsfrom Alfie Evans,16 Conway17 and

164

12 Jukka Varelius, ‘Medicine, Health Care and Philosophy’, Health and Autonomy (2005), Vol 8, pp. 221-230, at p. 221 13 Great Ormerod Street Hospital for children NHS Foundation Trust v Yates [2017] EWHC 1909 14 Kenan Malik, ‘A court is an unfit place for a case like Alfie Evans and its moral complexity’ (29 Apr 2018) <https:// www.theguardian.com/commentisfree/2018/apr/29/courtunfit-place-for-case-like-alfie-evans-moral-complexity/> 15 ibid. 16 ‘Alfie Evans parents ‘feared’ they would resent new baby’, 3 September 2018 https://www.bbc.co.uk/news/ukengland-merseyside-45402094/ Here the title is an obvious choice to focus on embellishing the emotional element of Alfie’s suffering. There is no introduction of the substantive considerations weighing for or against Alfie’s predicament. 17 “Terminally ill Noel Conway loses challenge to assisted dying ban”, 27 June 2018 <https://www.theguardian. com/society/2018/jun/27/assisted-dying-uk-court-rejectsnoel-conway-challenge> , the article cherry-picked Master of the Rolls, Sir Terence Etherton’s dicta that the Court was not well-placed to decide as compared to Parliament on the blanket ban without analyzing its reasoning regarding the necessity and

Nicklinson,18 the media tends to contribute nothing substantive to this debate besides claiming the Courts should not decide.19 Yet they do not illuminate us on who then should decide for patients if parents are grieving and doctors have deferred to higher judgment. Also, while it is humane to express sentiments on one’s passing, the degree to which these emotions are portrayed in the media tends to inevitably undermine, the substantive considerations that the Courts undertake. The epitome of this is the 132 page Nicklinson judgment20 which has explored complexities which cannot possibly be fully encapsulated by a newspaper article media.21 For instance in Charlie Gard, the Courts had to contend with layered considerations. The case concerned whether or not it was in Charlie’s best interests to switch off a life machine given that he was terminally ill22. Firstly and most self-evidently, the Courts contended with the parents’ grief vis-à-vis the need to determine a course of treatment ‘most compatible with maintaining Charlie’s dignity’.23 Secondly, the Courts had to consider S1 of the Children’s Act 1989, where the child’s welfare is the ‘paramount consideration’.24 Thirdly, the court considered expert evidence from the doctor as regards what would be the most ‘effective treatment’.25 It was in the light of these considerations that the Court held in absence of new evidence, the previous declarations made to withdraw artificial ventilation would be upheld.26 As such, as a preliminary point, the Courts in being neutral, avoid wading into the moral morass and instead adjudicates based on facts and policy considerations. Within ‘policy proportionality requirements. 18 Sarah Boseley, ‘Tony Nicklinson dies six days after losing ‘right to die’ case” (22 Aug 2012) https://www. theguardian.com/uk/2012/aug/22/tony-nicklinson-right-todie-case/ accessed 10 March 2019 19 ibid. 20 Nicklinson v Ministry of Justice [2014] UKSC 38 21 ibid. 22 Charlie Gard at [1]-[3] 23 Charlie Gard at [3] 24 Charlie Gard at [18] 25 Charlie Gard at [13] 26 Charlie Gard at [18]

ARTICLES: WRITTEN MOOT 3

considerations’, the tension between paternalism and autonomy is explored. 1.1 ‘Paternalism’ and ‘Autonomy’ Paternalism is defined as the ‘doctor knows best’,27 connoting respect for medical professionals, whose specialist knowledge forms their authority.28 Conversely, autonomy means our individual right to self-determination.29 Yet there are three different ways to conceive this ‘right’, set out as follows30:a. Ideal Desire Autonomy (‘IDA’) – what persons should want31 b. Best Desire Autonomy (‘BDA’) – ‘overall desire.’32 c. Current Desire Autonomy (‘CDA’) – based on ‘immediate inclinations’,33 possibly impulsive.34 Coggon argues that different conceptions of autonomy should be used to test for the courts’ inconsistent application of ‘autonomy’.35 However, while using multiple conceptions of autonomy aids in identifying inconsistency, it does not actually resolve the tension. As such, in line with the need to preserve sanctity of life, this essay proposes that IDA should be ‘preferred36 to avoid ‘confused and 27 Sidaway v Board of Governors of the Bethlem Royal Hospital [1985] AC 871 at 899 28 “Perhaps the least disputed is the knowledge gap that separates patient and doctor. It motivates the former to seek medical attention in the first place and underlies, in part, the professional authority of the latter.” Mark S Komrad, ‘A defence of medical paternalism: maximizing patients’ autonomy’, Duke University School of Medicine, Journal of Medical Ethics, 1983,Volume 9, p.p. 38-44, at p.38 29 “The courts repeatedly proclaim the value that the law gives to the individual’s right to self-determination”, Alasdair Maclean, Autonomy, Informed Consent and Medical Law, a relational challenge (1st Edition, CUP 2009) at p.149 30 John Coggon, ‘Varied and principled understandings of Autonomy in English Law: Justifiable Inconsistency or blinkered moralism?’, Health Care Analysis, September 2007, Vol 15, Issue 3, 235-255, at p. 240 31 ibid. 32 ibid. 33 ibid. 34 Coggon (n.30) at p. 241 35 ibid. 36 Coggon (n.30) at p. 253

inconsistent37’ principles.38 On this note, we can see that an objective assessment of one’s IDA requires neutrality and objectivity. Furthermore, there is scientific basis to the claim that a third-party decision will be most free from cognitive and metaphysical bias. Tversky and Kahneman argues that all people experience intuitive bias.39 When faced with a set of possible outcomes, they will usually be ruled by intuitive judgments.40 These are often undetected41 and applies even to the ‘statistically sophisticated42’. In deciding for a dying loved one, there is high probability of emotional intrusion reducing the objectivity of assessing potential choices and judgment.43 This rules out family or doctors which the alternative test, the substituted judgment test empowers to make decisions on behalf of the patients.44 It follows that the Courts in being neutral and objective, possess institutional capacity to decide in the best interests of patients. 2. Institutional Capacity While judges are human, the probability of experiencing bias is reduced from a third-party perspective. One might then argue Parliament, in being a third party, should be the deciding organ. However, in the context of the United Kingdom, given the relations between the ruling 37 Coggon (n.30) at p. 240 38 “There is evidence that aspects of the law are moving towards an approach that is more reflective of the promotion of best desire autonomy.” Emma Cave, ‘Protecting patients from their bad decisions: Rebalancing rights, relationships and risk’ Medical Law Review (2017), Vol 25, Issue 4, pp. 527-553 at p. 539 39 Amos Tversky and Daniel Kahneman, Judgment under Uncertainty: Heuristics and biases (Cambridge University Press 2013) at p. 18 40 ibid. 41 ibid. at p.19 42 ibid. (n.40) 43 Justin Welby, ‘Archbishop of Canterbury: my heart breaks for Charlie Gard’s family’, 31st July 2017 <https:// www.theguardian.com/uk-news/2017/jul/31/archbishopof-canterbury-my-heart-breaks-for-charlie-gards-family/> accessed 10th March 2019 44 Michael Ardagh, ‘Resurrecting autonomy during resuscitation – the concept of professionally substituted judgment’ Journal of Medical Ethics, 1999, Vol 25, pp. 375-378 at p.378

165

ARTICLES: WRITTEN MOOT 3

government and the NHS, often the Defendant in such proceedings, it is best to avoid any conflict of interest. After all, the Courts are the final arbiters of truth,45 which necessarily requires determining a dying patient’s best interest. This argument is buttressed by the fact that the Courts are ‘fiercely independent of the state46’ and will adjudicate in accordance to evidence surfaced and the law. Similarly, in Singapore, the Minister of Law, K. Shanmugam has reiterated the fact that the Courts will be the ‘final arbiters’47 of truth. While this comment was made in the context of the Protection of Online Falsehoods and Manipulation Act, this essay argues that it is the essence of its reasoning that should be extracted for the purposes of determining the Court’s institutional capacity. On this note, we examine the genesis of the best interests test in the UK and why it should be best applied by the Courts. 3. Development of the ‘best interests’ test

Figure 1: How the law in UK approves medical treatment48

Having established the need for neutrality and objectivity, which the Court represents, we examine the reasons why it is best placed to apply the ‘best interests’ test.

166

45 Lord Neuberger, ‘Twenty Years a Judge: Reflections and Refractions’, Neill Lecture 2017, Oxford Law Faculty, 10 February 2017 at [12], ‘I believe that it is fitting that the trial judge is, within the parameters of rationality, the final arbiter of fact’. 46 Charlie Gard at [18] 47 Tham Yuen-C, The Straits Times, 3rd April 2019, ‘Courts will be final arbiters of truth, says Shanmugam’, Accessed 15th June 2019, “The courts decide ultimately what is true and what is false, and they will be the final arbiters” 48 Parliamentary Office of Science and Technology, ‘Mental Capacity and Healthcare’, Post note, Number 381, June 2011, at p.2 < https://www.parliament.uk/pagefiles/504/ postpn_381-mental-capacity-and-healthcare.pdf> Accessed on 9th January 2019

In Bolam,49 it was held that the decisions made for patients were deemed valid so long as the method adopted was in accordance with ‘a responsible body of medical men skilled in that particular art.’50 Re A51 established for the first time in departing from the paternalistic Bolam test,52 the need to account for the patient’s medical and emotional welfare. The case concerned a 63 year old man with down syndrome, who had an application for vasectomy made on behalf of him for fear that his unbridled sexual tendencies would threaten the female residents of a local authority care he was moving into. Thorpe LJ rejected the application on the ground that it was necessary to focus on A’s best interests not the female residents.53 This holistic assessment in accordance with the patient’s IDA was picked up on and implemented in the Mental Capacity Act 2005.54 This exists in two main strands. i. S4(1) underscores the need to decide in the patient’s best interests, free from any biased assumptions based on age. ii. The statutes give guidance regarding what factors should be considered in the holistic assessment55 of the patient’s condition, modelling the approach in Re A. This includes:a. Past and present wishes and feelings of the patient56 b. Beliefs and values of the patient57 c. If patient is mentally incapacitated, will the person have capacity at some time in relation to the medical decision?58 d. Views of the patient’s loved ones if practical59 49 Bolam v Friern Hospital Management Committee [1957] 1 WLR 582 50 ibid. at 587 51 Re A (Medical Treatment: Male Sterilisation) [2000] 1 FCR 193 52 “The Bolam test does not provide the answer which remains best interests” Re A 53 Re A at p.6 54 Mental Capacity Act 2005, S4 55 Mental Capacity Act 2005, S4(2) 56 Mental Capacity Act 2005, S4(6)(a) 57 Mental Capacity Act 2005, S4(6)(b) 58 Mental Capacity Act 2005, S4(3) 59 Mental Capacity Act 2005, S4(7)

ARTICLES: WRITTEN MOOT 3

The Courts have been applying this test to cases, answering the difficult questions and therein developing the principles which undergird ‘best interests’. Therefore, they are best placed to apply the ‘best interests’ test given their ability to balance paternalistic and autonomous concerns. For example in Aintree,60 the court held that the ‘best interests’ test had to consider the patient’s wishes but not to have them prevail absolutely.61 This respects the knowledge of doctors and their right to prescribe treatment while balancing the wishes and beliefs of patients. A NHS Trust applied Aintree and offered some clarity in holding that doctors were not liable if they had ‘reasonable belief62’ that treatment was in patient’s best interests. Therefore, the Courts in being the gatekeeper between the medical profession and patients ensure that while the patient’s IDA is enshrined, doctors also contribute to the dialogue of ‘best interests’. For instance, Burke63held that advance directives demanding medical treatment have no legal force.64 In allowing but limiting the operation of advance directives, the judgments of doctors are respected, but the voices patients and their loved ones are heard. As such, the Courts have in fact proven that by virtue of being arbiters of truth, they have been more than capable in applying the ‘best interests’ test. This lays to rest any allegation that the Courts should not be allowed to decide what constitutes ‘best interests’ for dying patients. 3.1 Merits of the Courts applying the ‘Best Interest’ test The best interests test should be applied by the Courts instead of Parliament for the following 3 reasons, set out as follows:i. The Courts prove to be more objective 60 Aintree University Hospitals NHS Trust Foundation Trust v James [2013] UKSC 67 61 ibid. at [45] 62 An NHS Trust and others v Y [2018] UKSC 46 at [92] from hereon known as ‘A NHS Trust’ 63 R(on the application of Burke) v GMC [2005] EWCA Civ 1003 64 ibid. at [35]

than parliament in deciding the patient’s best interests ii. The Courts are best placed to ventilate intricate concerns of mentally capacitated and incapacitated patients iii. ‘Best interests’ test proves more robust than the ‘substituted judgment’ test Firstly, besides being the arbiter between paternalistic and autonomous concerns, the Courts prove to be more objective in deciding in the patient’s actual best interests given the lack of vested interest. While the duty of Parliament is to make or unmake laws,65 it may not be best placed to decide for patient’s best interests. This is as the merits of the best interest test require total neutrality and objectivity. No doubt, parliamentary mechanisms such as written questions and debate can engender some discussion over ‘best interests’. However, the perennial presence of political parties and different political identities mean even the most objective debates may be tinted with politicized responses.66 Furthermore, other pressing issues may feature more prominently in certain parliaments.67 Therefore, the Courts are better placed to rule on the ‘best interests’ test in deciding the balance between interests of the patient and doctors’ professionalism. This is especially so in circumstances involving life and death.68 In this respect, neutrality is especially important in interrogating submissions purporting to set aside medical decisions made by the local health authority. In Alfie Evans,69 the Court of Appeal was tasked with the heavy duty of considering whether the parents’ contention to set aside the local health authority’s decision 65 A.V. Dicey, “Introduction to the study of the law of the constitution” at p.3 < http://files.libertyfund.org/ files/1714/0125_Bk.pdf> Accessed 10 March 2019 66 Lord Patten admits that the issues of life and death are invariably politicised when they are ‘brought into politics’, Mental Capacity Bill, House of Lords Debate 17 March 2005, vol 670, cc1441-76 at 1460 67 Such as BREXIT for the United Kingdom 68 Nicklinson v Ministry of Justice [2014] UKSC 38 at [315] 69 Mr Thomas Evans, Ms Kate James v Alder Hey Children’s NHS Foundation Trust, Alfie Evans [2018] EWCA 984 (‘Alfie’)

167

ARTICLES: WRITTEN MOOT 3

was valid70. Alfie was a young child who suffered from a degenerative condition that had resulted in total disintegration of his brain. The core issue was whether it was in the best interests of Alfie to prolong his life in the absence of treatment.71 In considering the submissions of the appellants, Mcfarlane LJ had to interrogate the medical evidence submitted without making personal conjectures.72 Indeed, any finding made by the Court must be established with the support of medical evidence.73 This tenet is set out in great brevity by Macfarlane LJ:-

the claim for the Court serving to ventilate these concerns. This streamlines the way in which mentally capacitated patients are treated in their best interests in accordance with the IDA approach. For example in Montgomery,76 Lady Hale affirmed that informed consent, not consent, is now required.77 A patient’s ‘informed’ approval is obtained before doctors commence particular modes of treatment.78 ‘Informed’ meaning doctors are required to disclose material risks, which involves two considerations:

“The question of what the fact that a child is breathing tells us about tomorrow, this evening, next week, is not a matter for us to entertain conjecture about. If this does represent a change in the medical understanding about this condition and the medical viability and the best interests of the child in terms of undergoing the palliative care regime, then there is a need for that submission to be supported by some medical evidence.”74

i. whether a reasonable patient would find that risk significant or if ii. the doctor reasonably appreciated that the patient would find such risks significant.79

This is where the Courts are better placed to rule on such matters than Parliament. In light of politicized debate, and the need to nitpick on the ruling political party, the same level of objective interrogation may not materialize. Secondly, concerns of both mentally capacitated and incapacitated patients can be ventilated in the course of legal proceedings. While the ‘best interests’ test pertains only to mentally incapacitated patients in the UK, this is not to say that decisions are not made in the best interests of mentally capacitated patients. However, given that they possess capacity to consent, there is a statutory presumption pursuant to the Mental Capacity Act 2005 that any decisions made will likely be informed decisions.75 Yet concerns still linger as to what constitutes informed consent which buttresses

168

70 ibid. at [1] 71 ibid. at [20] 72 ibid. at [27] 73 ibid. 74 ibid. 75 Mental Capacity, S1(2) and S1(3)

These considerations are predicated on one’s autonomy to decide when doctors can interfere with one’s bodily integrity.80 However, the need to draw boundaries for informed consent saw the Courts hold that 2 qualifications exist:i. The ‘therapeutic exception81’ ii. Applicability limited to mentally capacitous patients.82 The therapeutic exception gives doctors rights to withhold risk-related information if disclosure was detrimental to the patient’s health, when the patient is subconscious or is unable to appreciate it.83 Cave argues that this exception subjects the ‘objective approach’ of the doctor’s standard of care to a ‘subjective exception’,84 a privilege for doctors. Therefore, in doing so, the Courts have effectively ensured 76 Montgomery v Lanarkshire Health Board [2015] UKSC 11 77 ibid at [107] 78 ibid.at [87] 79 ibid. at [72] and [73] 80 ibid. at [108] 81 Montgomery at [85] 82 See Figure 1. 83 Montgomery at [88] 84 Emma Cave, ‘The ill-informed: Consent to medical treatment and the therapeutic exception’, Common Law World Review 46(2), 140-168 at p. 143

ARTICLES: WRITTEN MOOT 3

that while patients are respected as autonomous consumers, doctors are given due respect for their specialization. Also, the application of informed consent to patients with mental capacity respects their use of information, made available by technology.85 However, the law also recognizes that ‘many online roads lead back to the clinic’.86 While the patient has access to medical knowledge online, this has to be verified by doctors. In this regard, the doctor’s medical opinion is still accorded due respect. This prevents ‘informed consent’ from being too onerous on doctors. In principle, Duce,87 which applied Montgomery, clarified that prior to applying the materiality test, the first ‘hurdle’88 to be crossed by the claimant is whether the doctor should have reasonably appreciated the significance patients attach to risks. This must be supported by expert evidence.89 Thus, where mentally capacitated patients are concerned, the possibility of autonomy encroaching onto paternalism is ventilated and rectified to a certain extent by the Courts. For mentally incapacitated patients, the Courts have adopted nuanced reasoning to deal with concerns that doctors blatantly impose treatment.90 The possibility of blatantly imposing treatment is illustrated in A NHS Trust, which held that it was not mandatory to obtain a court order to withdraw clinically assisted nutrition.91 Brazier and Cave argue that patients’ beliefs and values should be the ‘predominant interest’ as it alleviates blatant impositions of treatment by doctors in practice.92 Accordingly, the Courts 85 “A burgeoning market in digital medical information is developing, with 43,000 medical apps available in 2016 and reports that 500 million people would use a medical app in that year”, Sarah Devaney and Soren Holm, ‘The transmutation of deference in medicine: An Ethico-Legal Perspective’, Medical Law Review, Vol 26, No 2, pp. 202-234 86 ibid. at p.211 87 Gail Marie Duce v Worcestershire Acute Hospitals NHS Trust [2018] EWCA Civ 1307 88 ibid. at [42] 89 ibid. 90 ibid. 91 ibid. at [126] 92 Margaret Brazier and Emma Cave, Medicine, Patients

in A NHS Trust also affirmed that other persons taking an interest in the patient must be considered too.93 This recognizes the role of family as an added party in the doctor-patient interlocution,94 ensuring that familial concerns are addressed. Thirdly, the application of the ‘best interests’ test by the Courts subsumes the merits of its alternative - the ‘substituted judgment’ test. In Aintree, Lady Hale in citing the Law Commission Report No 128 on their report on Mental Incapacity illuminated that the Court has incorporated the merits of the ‘substituted judgment’ test in applying ‘best interests’.95 This is fulfilled by considering his or her ‘past and present wishes’ and ‘altruistic sentiments’96 for others. In fact, Gutheil and Appelbaum argue that the ‘substituted judgment’ test is a poor disguise of the ‘best interests’ test given the lack of a universal basis underpinning substituted judgment.97 It is on this note, that we take apart the flaws of the alternative test - substituted judgment, which empowers family members or doctors to decide what is in the best interests of patients. 4. ‘Substituted Judgment’ Interests’

versus

‘Best

4.1 United States In the United States, the ‘substituted judgment’ test is employed in place of the best interests test. It is defined as the appointment of a guardian who makes life decisions for the and the Law (Manchester University Press 6th edition) at p. 192 93 Gail at [125] 94 “Recognising our vulnerabilities means acknowledging that each of us are profoundly dependent on others and on a range of social provisions.” Jonathan Herring, Vulnerable adults and the law (OUP 2016) at p. 264 95 Aintree at [24] 96 ibid. 97 “The court places great stress on the decision maker’s ‘objectivity and selflessness’..then went on to elabourate a complicated six factor structure for decision making by the lower court judges in similar cases”, Thomas G. Gutheil and Paul S. Appelbaum, ‘Substituted Judgment: Best Interests in Disguise’, The Hastings Center Report, Vol 13, No. 3 (Jun 1983), pp. 8-11 at p. 9

169

ARTICLES: WRITTEN MOOT 3

incompetent, affirmed in Strunk v Strunk98 to be ‘recognised in American Courts since 1844.’99 However, this test has been criticized on the basis that while it claims to preserve a ‘purely personal right’, this right is decided by a third party, the Courts.100 Weber argues this forms the test’s ‘critical theoretical flaw’.101 In practice, 2 further issues arise:- i) On what basis and standard of measurement do we apply personal rights to incompetents?102 ii) How do we ascertain the incompetent’s intent which is not absolutely determinable?103 For example, in Re Quinlan,104 Karen Quinlan, a 22 year old lady was in a ‘moribund’, chronic vegetative state.105 Despite the lack of evidence pertaining to the patient’s feelings, the Court granted permission to remove artificial ventilation on 2 bases. Firstly, that the father was qualified to determine what Karen would have chosen (despite lack of hard evidence pertaining to Karen’s feelings).106 Secondly, that the overwhelming majority of people in such a situation would opt to do so.107 Therefore, this decision fails to clarify the issues above by essentially looking to how a random individual would respond.108 While one may counter-argue that there was however, no other option given Karen’s comatose state, the fact of the matter is in the absence of a universal framework establishing guiding principles, the substituted judgment test fails to illuminate an already complex area of law. In this regard, the best interests test fares better. Similarly in Saikiewicz,109 Joseph Saikiewicz was a sixty seven year old man with the mental capacity of a

170

98 Strunk v Strunk, 445 S.W.2d 145(1969) 99 ibid. at 148 100 Walter M. Weber, ‘Substituted judgment doctrine: a critical analysis’, Issues in Law and Medicine, Volume 1, No. 2, September 1985, pp.131-162 at p. 138 101 ibid. 102 Weber (n 100) at p.141 103 ibid. 104 Re Quinlan, 355 A.2d 647 (1976) 105 ibid. at 647 106 Re Quinlan at 653 107 Re Quinlan at 664 108 Weber (n 101) at p.143 109 Superintendent of Belchertown State School v Saikiewicz 370 N.E 2d 417 (1977) (‘Saikiewicz’)

two and a half year old child.110 The Supreme Court affirmed the request made by Joseph’s guardian to withdraw chemotherapy relying on a substituted judgment.111 This was again without clear evidence of Joseph’s wishes and feelings, instead, choosing to impute an intention of what the incompetent would have decided, set out as follows:“If that person were competent, but taking into account the present and future incompetency of the individual as one of the factors which would necessarily enter into the decision making process of the competent person.”112 This confuses the law given that a ‘substituted judgment’ is meant to be a substitute thinking and deciding in the shoes of the patient, not based on a third person’s perspective. On this note, we consider the comparatively more objective ‘best interests’ test in the United Kingdom and Singapore. 4.2 United Kingdom Brazier and Cave highlights the limited operative scope of the ‘substituted judgment’ test being that if one possessed mental capacity all his life, it becomes a ‘myth’.113 The Law Commission Report in the United Kingdom cited two main reasons for rejecting a wholesale interpretation of the ‘substituted judgment’ approach. Firstly, it cited the lack of a basis for judgment for the substitute nominees who have to make such complex decisions.114 To buttress this claim, there is empirical proof provided by Torke, Alexander and Lantos that laymen change their minds on health decisions every two years.115 Secondly, it rejected the 110 ibid. at 417 111 Saikiewicz at 430-431 112 Saikiewicz at 431 113 Margaret Brazier and Emma Cave, Medicine, Patients and the law (Manchester University Press 6th edition) at p. 165 114 Law Commission Report No. 128 at para 2.14 115 Alexia Torke, G. Caleb Alexander and John Lantos, ‘Substituted judgment: the limitations of autonomy in surrogate

ARTICLES: WRITTEN MOOT 3

binary view of ‘substituted judgment’ and ‘best interests’, arguing that it would be feasible to incorporate some degree of substituted judgment.116 Therefore, in its development of the ‘best interests’ test, the Courts have given life to a consistent line of reasoning and principles.117 The approach to rejecting substituted judgment is justified. Despite its ostensible merits, the prevalence of multiple irreconcilable normative standards of judgment makes it difficult to formulate any professional guidance.118 This difficulty is protracted further given the prevalence of ‘more unfavourable’ conditions for competent patients who become incompetent. This is because it is difficult to answer the question regarding on what basis does one decide if one only has on record, the person’s wishes when he or she was competent. It is highly improbable that laymen can answer this question conclusively, and if they do, the judgment will be most probably be made based on those wishes made when the person was competent. This, Berghmans argue, is ‘seriously flawed’119 as it ignores looking at things from a mentally incapacitated patient’s point of view. In this regard, a clear statutory framework such as that stipulated in the Mental Capacity Act is favoured and required. The accounts for why the Law Commission does not brook the substituted judgment test.

decision making’, Journal of General Internal Medicine at p.1514 116 ibid. 117 John Coggon, ‘Mental Capacity Law, Autonomy, and Best Interests: An argument for conceptual and practical clarity in the Court of Protection’, Medical Law Review, Vol 24, No 3, pp 396-414 at 414 118 Linus Brostrom, Mats Johansson and Morten Klemme Nielsen, ‘What the patient would have decided: a fundamental problem with the substituted judgment standard’, Medicine, Health Care and Philosophy Journal, 2007, Volume 10, pp. 265-278 at p.266 119 Ron L.P. Berghmans, ‘Ethical hazards of the substituted judgment test in decision making concerning the end of life of dementia patients’, International Journal of Geriatric Psychiatry, Vol 12, pp. 283-297, at p. 285

4.3 Singapore In a similar vein, in Singapore, in the absence of case law illuminating whether the courts are best placed to decide in patients’ ‘best interests’, we look at the statutory framework. The Mental Capacity Act 2010120 stipulates the ‘best interests’ test as being objective, and cannot be made on the basis of appearance,121 age,122 or an aspect of his behaviour that may result in unjustified assumptions made regarding his best interests.123 This framework is similar to that of s4(1) of the Mental Capacity Act 2005 in the United Kingdom. However, all persons appointed as deputies, people who are legally authorized to make all decisions pertaining to the patient’s property and affairs124 are to be appointed by the Court.125 As regards the assessment of the patient’s medical condition, the Courts have the power to accept or reject the doctors’ conclusions.126 As such, it is evident that the courts in Singapore oversee and decide who should be appointed to decide in the best interests of patients. However, since judicial decisions cannot satisfy all parties, one may ask is it even possible to determine a universal definition of ‘best interests’? 5. Harking back to what constitutes ‘best interests’?

actually

This presupposes no conclusive answer given the complexity of case facts,127 the limitations of the law, the embellishment of mainstream media and the perennially sensitive topic of life 120 Mental Capacity Act (Chapter 177A) 121 Mental Capacity Act (Chapter 177A), s6(1)(a) 122 ibid. 123 Mental Capacity Act (Chapter 177A), s6(1)(b) 124 Mental Capacity Act (Chapter 177A), s24(1)(a) 125 Mental Capacity Act (Chapter 177A), s20(2)(b) 126 Peng Soon Ng et al, ‘Importance of mental capacity: time for greater attention and action’, Singapore Medical Journal 2015, 56(12), pp. 646-648 at p. 646 127 No one case is the same in terms of context. For example, Re A concerned a specific scenario of moving to a local housing authority regarding the question of best interests. This is as opposed to right to assisted dying cases. All of which presuppose a general universal test to be applied.

171

ARTICLES: WRITTEN MOOT 3

and death. Yet the question of ‘best interests’ of whom is worth pursuing. One can argue that the dissatisfaction targeted at ‘unfavourable’ judicial decisions is merely a projection of our conception of ‘best interests’. Accordingly, it follows that any other objections voiced by the media represents their projection of the patient’s best interests, not necessarily an objective projection. Harking back to the IDA conception, what defines ‘best desire autonomy’ should ultimately be rooted in an objective assessment of the patient’s welfare128 While it is difficult to pull apart emotional attachment to a person from influencing this assessment, one must desist from projecting subjective notions of ‘best interests’. Doing so defeats the entire purpose of having the best interests test. The Courts in adjudicating difficult cases rule in an institutional capacity on the patient’s predicament. In such cases, critics argue that the Courts may serve the best interests of Parliament especially when they defer in right to assisted dying cases. For example, in Nicklinson,129 the Court held that it was for Parliament to change the law on assisted dying. One can see the value of this criticism when the same strand of reasoning was conveniently repeated in Conway.130 However, despite inability to make new laws (which is rightly so), the Courts have still made substantive rulings regarding the patients’ best interests. Within their judicial powers, the Courts still ruled on the merits of the case before coming to their decision. While they cannot authorize assisted dying, it does not

172

128 “Also, ideal desire autonomy might include simple reference to a purportedly objective system of ideals that for some other reason would lead to an agent’s being considered ‘irrational’ were he to ignore it.” Coggon (n.31) at p.241 129 “We have no jurisdiction to impose anything: that is a matter for Parliament alone” Nicklinson at [325], per Lady Hale 130 R(on the application of Conway) v the Secretary of State for Justice at [205], “They explained (at [109]) why, in the circumstances, there are powerful constitutional reasons for the court to respect Parliament’s assessment of the necessity of maintaining section 2 and (at [110]) why Parliament is better placed than the court to make the relevant assessment regarding the likely impact of changing the law”

mean they are effete. By ruling on the principles of necessity and proportionality, they have provided justification to Parliament to account for its current laws.131 For instance, in Nicklinson, while the Court deferred to Parliament to amend laws, there was healthy debate regarding whether the current laws contravened Mr Nicklinson’s Article 8 right and whether it was necessary and proportionate to impose a blanket ban for suicide (assisted dying). In fact, 5 out of the 9 justices held that the courts had the authority to issue a declaration of incompatibility for the blanket ban on assisted dying.132 However, Lord Neuberger, Lord Mance and Lord Wilson declined to grant the declaration while Lady Hale and Lord Kerr would have. Lord Neuberger’s reasoning was twofold. Firstly, it was that the blanket ban was necessary to protect the ‘weak and vulnerable’,133 with reference to those who may be in Mr Nicklinson’s position but whom do not want to die.134 Secondly, it was not ‘institutionally appropriate’ for the Courts to rule on such issues.135 Conversely, Lady Hale reasoned that regarding infringement of Article 8, protection of morality such as autonomy and human dignity was equally important and should not be contravened by the blanket ban.136 Minimally, this conflict of opinion ensured an informed conclusion regarding whether it was in the ‘best interests’ of the two appellants to pursue assisted dying. Therefore, where the ‘best interests’ of the patient are concerned, the Courts have done their due diligence within constitutional limits. 6. Conclusion The courts in being objective and neutral are best placed to decide in the patient’s best interests. Despite the appeal of having loved ones decide for patients, the prospect of a loved one dying often engenders emotions leading to bias clouding judgment. A longer life may not 131 HC Deb 11 September 2015,Vol 599, col 656 132 Suicide Act 1961, S2 133 Nicklinson at [85] 134 ibid. 135 Nicklinson at [148] 136 Nicklinson at [311]

ARTICLES: WRITTEN MOOT 3

mean a better life. In other circumstances, a swift death beats prolonged suffering. If the most ideal decision is to be reached, the courts are best placed to adjudicate on such matters for the established reasons. In this regard, there is no other institution or alternative test that fares better against the Court’s objective judgment. The greatest paradox is that for all who criticize the Courts as not being best positioned to decide for dying patients are projecting a certain conception of best interests. It could be a selfish conception, rooted in emotion or the fear of letting go. In an empirical study conducted by Suhl et al, a sample size of 100 patients were given questionnaires detailing life support scenarios to document their response.137 In general, only 35% of the patients surveyed expressed desire for life support.138 One question in particular was if after 3 weeks of coma, a breathing machine is required to keep them alive, would they want it continued?139 Only 26% answered in the affirmative.140 Sometimes, wishing the best for a dying patient is not equivocal to a valid decision in their ‘best interests’. Suhl et al provides empirical proof that surrogates, supposedly the ones closest to the suffering patient, correctly guessed their wishes for 59.6% of the time.141 Admittedly, the Courts issue certain decrees that may amount to prolonging life when death is ‘preferred’. Yet in the absence of a better and definitive answer or institution to judge these highly intricate matters, the Courts who look to uphold sanctity of life, are best placed to decide the best interests of dying patients.

will be to find whatever medical treatment possible, at whatever cost to save them. This is indubitably, a delicate situation. There is never a right or wrong answer, but a spectrum of opinions: the doctor’s, the family’s and the patient’s (where he is lucid). Yet, the ‘best interests’ test does not seek to be a universal solution, neither does it promise to be a panacea to the patient’s medical condition. However, from an objective point of view, when applied robustly, the ‘best interests’ test upholds the sanctity of life. After all, the age old chinese adage: “拿得起，放得下”142 when used in this context reminds us to let go of what we cannot control; to be stoic when confronting these calamities. The essence of life is after all, its evanescence. As such, in reply to Robert Frost, this essay concludes on the following note:Nothing Gold Can Stay. Yet, surely, it is enough, even for a day.143

Sometimes, it is better to cherish the sanctity of life. This means appreciating the good health of our loved ones. When afflictions, immense pain and suffering befall them, our natural instincts 137 Suhl et al, ‘Myth of substituted judgment: Surrogate decision making regarding life support is unreliable’, Arch International Medical Journal, 1994, Vol 154, pp. 90-96 at p.90 138 ibid. 139 Suhl et al, ‘Myth of substituted judgment: Surrogate decision making regarding life support is unreliable’, Arch International Medical Journal, 1994, Vol 154, pp. 90-96 at p. 92 140 ibid. 141 Suhl et al (n 139) at p. 93

142 Liwei Jiao and Benjamin Stone, ‘500 Common Chinese Proverbs and Colloquial Expressions: an annotated frequency dictionary’ (Routledge Publishing 2014) at p. 29. Literal translation means being able to pick up and put down things in life. On a connotative level, this means going with the flow and accepting what happens to us. 143 Poem my own. This article is dedicated to my mother.

173

ARTICLES: WRITTEN MOOT 3

Not my decision to make: The courts, best interests, and legally incompetent children* Soh Kian Peng, Singapore Management University “Some things we have only as long as they remain lost, some things are not lost only so long as they are distant”1 Introduction Death is just another journey, one that we all must make.2 Advancements in medical technology, however, has lengthened this journey considerably. It has enabled people to be kept alive, long after their bodies have failed them.3 Naturally, this has created enormous problems; especially when the patient in question is a child and both the medical staff and loved ones disagree as to whether life support should be withdrawn. Who then, is best placed to resolve this dispute? This issue rose to prominence in the UK following the case of Alder Hey Children’s NHS Foundation Trust v Thomas Evans (“Alfie Evans”).4 Alfie Evans, the child at the heart of that case, suffered from

174

* I would like to thank Edwin Teong for his comments. However, any errors, as usual, remain mine. 1 Rebecca Solnit, A Field Guide to Getting Lost (Canongate Books, 2017), 41. 2 See the Goodreads website: <https://www. goodreads.com/quotes/701843-pippin-i-didn-t-think-itwould-end-this-way-gandalf> (accessed 25 February 2019). 3 Adrian Owen, Into the Grey Zone (Guardian Books, 2017); Airedale NHS Trust v Bland [1993] 2 WLR 316, 381. See also, Atul Gawande, Being Mortal (Profile Books, 2014), 157. Gawande writes: “Technology can sustain our organs until we are well past the point of awareness and coherence. Besides, how do you attend to the thoughts and concerns of the dying when medicine has made it almost impossible to be sure who the dying even are?” [emphasis added] 4 [2018] EWHC 308 (Fam). This case received worldwide attention as well: ‘Alfie Evans: World’s reaction to sick toddler’s case’ (BBC, 25 April 2018) <https://www.bbc. com/news/uk-england-merseyside-43892684> accessed 5 June 2019. See also, the Charlie Gard Foundation Website <https:// www.thecharliegardfoundation.org/charlies-law/> (accessed 27 June 2019). The case of Charlie Gard preceded that of Alfie Evans. They were largely similar on the facts, except that Charlie’s parents eventually agreed to withdraw life support. However, they have since been attempting to develop draft legislation that would prevent such cases from reaching the court.

a “degenerative neurological condition”.5 He eventually became comatose and the hospital applied to the court for a declaration that continuing to keep Alfie on life support was not in his best interests. The court agreed, granting the declaration sought.6 Notably, in Singapore, this issue has yet to be adjudicated before our courts.7 This piece considers the approach Singapore should adopt in cases where legally incompetent children are concerned. We begin by examining the UK’s approach, viz, the “best interest” test. We then turn to examine local precedent in the case of Re LP (adult patient: medical treatment).8 Our attention will then turn to examining various criticisms of the “best interest” test. In particular, it will be argued that as far as possible, a legal solution should not be adopted for two reasons. First, the “best interest” test is doctrinally problematic. Second, the fluidity of the factual matrix adds to the difficulty of applying the “best interest” test. In that light, it is preferable that such disputes be resolved through nonlegal means where possible. Litigation should only be turned to as a last resort. The author makes three small suggestions in that regard. First, better communication between doctors and the parents may prevent such disputes from arising. Second, if such a dispute should arise, mediation might be a better alternative. Finally, the court should only find that it has 5 Alder Hey Children’s NHS Foundation Trust v Thomas Evans [2018] EWHC 308 (Fam), [25]. The first indications that Alfie was not well was first noted in July 2016 when he was noted to have a divergent squint. When he was 6 months of age, he showed marked signs of developmental delay. His condition deteriorated precipitously after he was admitted to hospital in December 2016. 6 ibid, [66]. 7 The only case where such a court declaration has been sought was the case of Re LP (adult patient, medical treatment) [2006] 2 SLR 13. We discuss this case below. 8 [2006] 2 SLR 13.

ARTICLES: WRITTEN MOOT 3

the jurisdiction to hear such cases where (a) the dispute is intractable, (b) the dispute poses a threat to the safety of medical staff and other patients or (c) where the possibility of abuse exists. Scope Before we begin, it should be noted that different scenarios can raise different end-of-life issues.9 For the purposes of this piece, we are not concerned with issues of euthanasia or a child’s right to consent to medical treatment.10 Our interest is in an Alfie Evans or Charlie Gard11 type scenario involving a legally incompetent child.12 We illustrate this accordingly.13 The patient, a child, is admitted to hospital. His condition deteriorates rapidly. The doctors do what they can, to no avail. At this stage, the patient is only being kept alive through the assistance of life support machines. Doctors recommend that life support should be withdrawn or that no attempt be made to resuscitate the child. In their opinion, the prognosis is poor; the patient has little chance of recovery.14 Disagreeing with the doctor’s assessment, both his parents refuse to consent to the withdrawal of life support.15 Subsequently, the hospital applies to the court for a declaration that would either (a) allow them to withdraw the patient from life-support,

9 There must be a common basis for comparison; in comparative law this is known as the tertium comparationis. See Jan M. Smits et al, Elgar Encyclopaedia of Comparative Law (Edward Elgar, 2006), 442. 10 Gillick v West Norfolk and Wisbech Area Health Authority [1986] AC 112. 11 Great Ormond Street Hospital v Yates, Gard and Gard [2017] EWHC 972 (Fam). 12 Lynn Hagger, ‘Children, Law and a Good Death’ in Lynn Hagger and Simon Woods (eds), A Good Death? Law and Ethics in Practice (Ashgate 2013), 124. 13 Alder Hey Children’s NHS Foundation Trust v Thomas Evans [2018] EWHC 308 (Fam); Portsmouth NHS Trust v Wyatt [2004] 1 FLR 21. 14 Airedale NHS Trust v Bland [1993] 2 WLR 316, 323; Portsmouth NHS Trust v Wyatt [2004] 1 FLR 21, [1]. 15 Alder Hey Children’s NHS Foundation Trust v Thomas Evans [2018] EWHC 308 (Fam). See also R v Portsmouth Hospitals NHS Trust ex parte Glass [1999] 2 FLR 905, 906 – 907. In this case, a “Do Not Resuscitate” order was put in the patient’s medical notes without consulting his mother.

or (b) withhold life-saving treatment.16 Notably, such a scenario raises ethical concerns. For instance, Wilkinson and Savulescu noted that the Charlie Gard case raised the following ethical questions: 1. Would life for Charlie in his impaired state be a life worth living?17 2. Should we judge this based on a subjective or objective accounts of well-being?18 3. How should diverging views about medical facts be taken into account?19 4. Should such decisions should be made through the courts or “some extra-judicial process”?20 While the question of who should decide in the child’s best interests is an ethical one, for the purposes of this piece, our focus is on the legal aspect of the “best interest” test. In particular, we are concerned with its conceptual clarity and how it has fared in practical application. In doing so, we adopt an internal perspective of law.21 While law may be viewed through an external perspective, in this case, through ethical theory,22 it is important that the law is judged “on its own terms”.23 One should bear in mind what the law is capable of achieving and what it is not.24 This sets the stage for our evaluation as to the approach Singapore should adopt in relation to such cases. With this context in mind, we now turn to consider the applicable legal principles as applied by the UK courts. UK – Best Interest test The paramount consideration underlying cases 16 ibid. 17 Dominic Wilkinson & Julian Savulescu, “Hard lessons: learning from the Charlie Gard case” (2018) 44 Journal of Medical Ethics 438, 440. 18 ibid. 19 ibid. 20 ibid. 21 Richard Huxtable, Law, Ethics and Compromise at the Limits of Life (Routledge, 2013), 14. See HLA Hart, The Concept of Law (2nd Ed, Oxford University Press), 203. 22 ibid. 23 ibid. 24 ibid.

175

ARTICLES: WRITTEN MOOT 3

in this area is the welfare of the child.25 This is enshrined in section 1(1) of the Children Act 1989.26 It applies regardless of whether the decision is taken by “parents, doctors or the court”.27 In cases of intractable disputes between parents and medical professionals, the court must resolve the dispute, bearing in mind the welfare of the child.28 In doing so, the court “considers what is in the child’s best interests from the child’s perspective”.29 The “best interest” test has its roots in the medical profession.30 It refers to treatment which, in the doctor’s opinion, would be in the patient’s best interests.31 The genesis of the “best interest” test qua legal test is owed to Bolam v Frien Hospital Management Committee (“Bolam”).32 The court in Bolam held that a doctor “would not be guilty of negligence if he has acted in accordance with a practice accepted as proper by a responsible body of medical men skilled in that particular art” [emphasis added].33 Use of the Bolam test to resolve disputes centered around disagreement as to treatment options where consent was unable to be provided transformed the “best interest” test into a legal one.34 Essentially, “if a responsible body of medical opinion thought that treatment was in the

176

25 ibid, 50. 26 Emma Cave and Emma Nottingham, “Who Knows Best (Interests)? The Case of Charlie Gard” (2017) 26(3) Medical Law Review 500, 504. 27 ibid. 28 ibid. 29 ibid. 30 Rob Heywood, “Parents and Medical Professionals: Conflict, Cooperation, and Best Interests” (2012) 20 Medical Law Review 29, 35. 31 ibid, 36. Heywood notes that “[t]raditionally, best interests was dictated by the medical profession”. 32 The court in Bolam held that a doctor “would not be guilty of negligence if he has acted in accordance with a practice accepted as proper by a responsible body of medical men skilled in that particular art” [emphasis added]. 33 ibid, 587. See also Bolitho v City and Hackney Health Authority [1998] AC 232. Bolitho modified the Bolam test. The judge must be satisfied that the body of expert opinion can be logically supported before it can be used to assess whether the defendant had indeed fallen short of the standard of care owed. 34 Rob Heywood, “Parents and Medical Professionals: Conflict, Cooperation, and Best Interests” (2012) 20 Medical Law Review 29, 36.

patient’s best interests, the law would agree”.35 The “best interest” test, however, has evolved since. The Children Act 1989 provides a “range of non-exhaustive factors that must be taken into account when determining the welfare of a child”.36 Courts have recognised that what constitutes “best interests” is a multi-faceted inquiry, “encompass[ing] medical, emotional and all other welfare issues”.37 In this regard, “there is a balancing exercise to be performed in assessing the course to be adopted in the best interests of the child”.38 However, it should be noted that the courts still find it “incredibly difficult to distance themselves from medical opinion”.39 Mention should also be made, at this juncture, of the “substituted judgement” approach to the “best interest” test. Application of the substituted judgement test involves “donning the mental mantle of the incompetent and striving to make a decision that equates to, or closely resembles that which the patient himself would make, if competent”.40 This approach has been adopted by the UK courts. In Portsmouth Hospital NHS Trust v Wyatt,41 the court concluded that in deciding what is in the child’s best interests, the judge “must look at the question from the assumed point of view of the patient”.42 Having briefly sketched out the approach UK 35 ibid. See also Paul Tan, “When Death Knocks, Who Will Answer and What Will They Say?” (2006) 18 Singapore Academy of Law Journal at [10]. 36 ibid citing the Children Act 1989, s 1 (3). Some factors the court has to consider include: (a) ascertainable wishes and feelings of the child concerned, (b) his physical, emotional and educational needs, (c) the likely effect on him of any change in his circumstances, (d) his age, sex background and any other characteristics of his life the court may find relevant. 37 ibid. See Re A (Male Sterilisation) [2000] 1 FLR 549, 555. 38 Re J (A Minor) (Wardship: Medical Treatment) [1991] 1 FLR 366, 375. 39 Rob Heywood, “Parents and Medical Professionals: Conflict, Cooperation, and Best Interests” (2012) 20 Med Law Rev 29, 39. 40 Richard Huxtable, Law, Ethics and Compromise at the Limits of Life (Routledge, 2013), 87. 41 [2005] EWCA Civ 1181. 42 ibid, [87].

ARTICLES: WRITTEN MOOT 3

courts have adopted in such cases, we turn now to examine the local position on this matter. Singapore – Best interest There has yet to be, as far as this author is aware, any case directly on all fours with Alfie Evans that has been brought before the Singapore courts.43 To date, the only reported case of a hospital seeking a court declaration is that of Re LP (adult patient: medical treatment).44 That case concerned an application by the Mount Alvernia Hospital which sought a court declaration that a proposed operation to amputate both legs of a patient was lawful. The patient in question (“Mdm LP”) was in a comatose state. She had informed her doctors to “save her legs at all costs”.45 However, it had not been made known to her that she would die if her legs were not amputated as she had entered a coma before the doctors could discuss this with her.46 Mount Alvernia Hospital subsequently made an urgent application to court, seeking a declaration that the proposed operation to amputate Mdm LP’s legs was lawful.47 In deciding whether the proposed operation was in Mdm LP’s best interest, Choo J took into account the views of her son (“L”) and range of medical outcomes the proposed amputation would have.48 He concluded that the proposed operation was in Mdm LP’s best interest and allowed the application.49 As a preliminary point, it should be noted that Choo J held that the court did have the jurisdiction to hear the application.50 It is trite that courts do not hear cases in the absence of

43 A search on LawNet showed no such cases. 44 [2006] 2 SLR 13. 45 Re LP (adult patient: medical treatment) [2006] 2 SLR 13, [2]. 46 ibid. 47 ibid, [1]. 48 ibid [3], [8]. 49 ibid [11]. 50 ibid, [6]. See also Paul Tan, “When Death Knocks, Who Will Answer and What Will They Say?” (2006) 18 Singapore Academy of Law Journal, [7] and [52].

a real dispute.51 However, as Choo J pointed out, given that the wishes Mdm LP expressed conflicted with the proposed operation, it would be “prudent for the doctors to place the matter for an independent evaluation of their opinion regarding the patient’s best interests”.52 The same reasoning is likely to apply in a case similar to Alfie Evans where the wishes of the child’s parents conflict with that of the doctor’s clinical judgement and the disagreement cannot be resolved.53 Therefore, Singapore courts will likely have the jurisdiction to hear such an application if a case similar to Alfie Evans arises. The question then is; should the court hear such cases, notwithstanding the fact that it has the jurisdiction to do so. Criticisms of the Best interest test The “best interest” test has received criticism in academic circles. As Huxtable points out, it “fails to measure up to the standards of rational law”.54 Huxtable established two criteria on which to judge the rationality of a law. First, laws must be consistent; “fundamental norms should not directly compete, and similar cases should be dealt with similarly”.55 Second, laws must be capable of serving as a guide to human action.56 This requires laws to be, inter alia; clear, prospective and non-contradictory.57 51 See the remarks made in Tan Eng Hong v Attorney General [2012] 4 SLR 476, [132]. The court highlighted that: “the function of the courts [is] to adjudicate on and determine disputes between parties. Without a lis, the courts may find themselves being called on to give advisory opinions on abstract, hypothetical and/or academic questions instead of deciding on real disputes”. 52 ibid, [6]. 53 Lynn Hagger, The Child as Vulnerable Patient; Protection and Empowerment (Ashgate, 2009), 63. 54 Richard Huxtable, Law, Ethics and Compromise at the Limits of Life (Routledge, 2013), 80. 55 ibid, 30. 56 ibid. See HLA Hart, The Concept of Law (3rd Ed, Oxford University Press, 2012), 11 - 12. Hart writes that “there is, over and above the clear ascertainable facts of group behaviour and predictable reaction to deviation, our own powerful feelings of compulsion to behave in accordance with the rule and to act against those who do not”. See also Scott J. Shapiro, “What is the internal point of view?” (2006) 75(3) Fordham Law Review 1157. 57 Lon Fuller, The Morality of Law (Yale University Press,

177

ARTICLES: WRITTEN MOOT 3

Arguably, the best interest test fails on both counts; it clashes with established principles in criminal law and its inherent flexibility creates uncertainty, meaning that it fails to serve as a guide to human action. We discuss both points seriatim. Clashing with criminal law? Huxtable opines that “it is difficult to see how [cases where withholding or withdrawing lifesupporting treatment is allowed] can occupy the same terrain as rulings which insist that the failure to care for people who are incapable of caring for themselves can amount to a homicide offence when such a person dies”.58 While decisions regarding the continuation of life-support are commonly viewed through the lens of medical law, it is difficult to ignore the fact that withholding treatment can constitute a criminal offence.59 At criminal law, there is a duty to care for someone who is unable to care for themselves. This was established as early as the case of The Queen v Instan.60 In that case, the defendant was the niece of the deceased. They lived together. The deceased, shortly before she died, suffered from gangrene.61 It rendered her unable to look after herself. The defendant failed to give or procure medical treatment for her and continued to live in the house at the deceased’s expense.62 In affirming the defendant’s conviction, Lord Coleridge noted:

178

One, however, might argue that an exception can be made where doctors are concerned.64 But can this exception be meaningfully carved out? After all cases with the same material facts should, applying the principle of stare decisis, produce the same outcome.65 Lord Mustill grappled with this question in Airedale NHS Trust v Bland.66 In that case, Anthony Bland was a victim of the Hillsborough disaster.67 In a persistent vegetative state, Bland had no awareness of the world around him. There was virtually no hope of recovery or improvement.68 His doctors, fearful of criminal proceedings, applied for a declaration that it was lawful to withdraw life-support.69 Lord Mustill pointed out: “A doctor who kills his patient even with the consent of the patient is guilty of murder. Plainly a second doctor who kills his patient in circumstances where the obtaining of consent is impracticable cannot be in a better position than the first, even if the termination of life is in the best interests of the patient; for the combination of necessity and best interests is no more than a replacement for consent. How then can best interests legitimate the conduct proposed in the present case? [emphasis added]”70

“[I]t was the clear duty of the prisoner to impart to the deceased so much as was necessary to sustain life of the food which she from time to time took in.”63

Going further, Lord Mustill asked how the present case was distinguishable from cases such as Gibbins v Proctor.71 In that case, the appellant “had a helpless person in her care” who depended on her.72 It was found that the appellant intended to and did cause the

1964), 39. Fuller’s 8 desiderata are generality; promulgation; prospectivity; clarity; lack of contradiction; possibility of performance; persistence and congruence between announced rules and their administration. 58 Richard Huxtable, Law, Ethics and Compromise at the Limits of Life (Routledge, 2013), 76. 59 Airedale NHS Trust v Bland [1993] 2 WLR 316, 383. 60 [1893] 1 QB 450. See also Gibbins v Proctor (1918) 12 Cr App Rep 134. 61 ibid, 451. 62 ibid. 63 ibid, 454.

64 Richard Huxtable, Law, Ethics and Compromise at the Limits of Life (Routledge, 2013), 77. 65 A.L. Goodhart, “The Ratio Decidendi of a case” (1959) 22(2) Modern Law Review 117, 119. See also, J. Montrose, “The Ratio Decidendi of a Case” (1957) 20(6) Modern Law Review 587; Julius Stone, “The Ratio of the Ratio Decidendi” (1959) 22(6) Modern Law Review 597. 66 [1993] 2 WLR 316. 67 ibid, 323. 68 ibid. 69 ibid, 324. 70 ibid, 395. 71 (1918) 12 Cr App Rep 134. 72 Airedale NHS Trust v Bland [1993] 2 WLR 316, 396.

ARTICLES: WRITTEN MOOT 3

death of the helpless person by withholding nourishment.73 The cases, Lord Mustill noted, were “miles apart from an ethical standpoint” but could hardly be distinguished on the essential facts.74 Therefore, it would seem that end-of-life cases where the best interest test is applied appears to contradict principles well established in criminal cases such as Gibbins.75 On that ground, one might argue that the “best interest” test is merely, as Lord Mustill pointed out, an attempt to legitimise such conduct. Best interest – creating uncertainty In deciding what is in the child’s best interests, the court, in its judgement, signals to medical practitioners the circumstances in which it would be appropriate to withhold treatment. However, while it is easy to state that the “best interest” test requires the judge to consider the patient’s medical, emotional and all other welfare issues, this belies the difficulty of applying the test in practice. Courts have explicitly pointed out that a balancing act must be performed when determining “best interests”.76 But how does one strike this balance? As one author has pointed out, many of the factors that judges have to consider in this balancing exercise are “impossible to measure or compare”.77 For instance, one can hardly compare the pain suffered by a child to the wishes of their parents.78 Absent any concrete guiding principles or general rules as to how this balancing exercise is to be conducted,79 the “best interest” test risks becoming shorthand for whatever the judge 73 ibid. 74 ibid. 75 See Joel Feinberg, Harm to Others (Oxford University Press, 1984), 126 – 186. 76 Rob Heywood, “Parents and Medical Professionals: Conflict, Cooperation, and Best Interests” (2012) 20 Medical Law Review 29, 38 citing Re OT [2009] EWHC 633 (Fam). 77 ibid, 39. 78 ibid. 79 Richard Huxtable, Law, Ethics and Compromise at the Limits of Life (Routledge, 2013), 84. Huxtable notes that “there is a lack of sufficiently general rules” and that courts have generally placed “a strong emphasis on the particularities of the case at hand”.

thinks is appropriate in a particular situation. One might argue that the “vast array of situations that might arise in the life of a child” demands that the “best interest” test be sufficiently flexible.80 For example, another aspect of “best interests” that should be considered is the view of the child. That said, some disputes may either (a) “not involve infants too young to have any (even imputed) views of their own”81 or (b) involve a child that is “slightly older, albeit not Gillick competent”.82 Applying the “best interest” test to both situations requires a great amount of flexibility. However, this may be a double-edged sword. The inherent flexibility of the “best interest” test creates uncertainty amongst medical practitioners who are left to second guess when the court will place a greater emphasis on “wider social and quality factors” instead of medical considerations.83 The “best interest” test is therefore problematic; the lack of clarity and predictability in its application means that it fails to serve as a guide to human action. This could, in turn, encourage the practice of defensive medicine. Bereft of guidance from the law,84 doctors would invariably want to seek a judicial declaration that their proposed course of treatment is lawful, lest they be hit with a lawsuit or disciplinary proceedings. Practical difficulties Even if we shut our eyes to the jurisprudential quagmire the “best interest” test represents, such cases are immensely difficult to deal with, both for the judge assigned to hear the case and the court in general. For one, deciding what is in the 80 ibid, 85. 81 Rob Heywood, “Parents and Medical Professionals: Conflict, Cooperation, and Best Interests” (2012) 20 Medical Law Review 29, 41. 82 ibid. 83 ibid, 42. 84 See Singapore Medical Council v Dr Lim Lian Arn [2018] SMCDT 9. There was a huge uproar amongst the medical community following the conclusion of this case. See also the website of an anonymous doctor who commented on the case for a summary: <https://hobbitsma.blog/2019/01/25/aboutthat-100000-fine-for-an-injection/> (accessed 12 February 2019).

179

ARTICLES: WRITTEN MOOT 3

child’s “best interests” is complicated because the medical condition of the child may not always be stable.85 Miracles do happen. Charlotte Wyatt’s case illustrates this perfectly.86 She was born three months early and was beset with a whole host of medical problems.87 The hospital approached the court seeking a declaration that it would be lawful to withhold artificial ventilation if she needed to be resuscitated again. Her parents opposed the application as they did “not believe that she [was] yet ready to die”.88 However, Hedley J granted it, concluding that it was not in Charlotte’s best interests to receive any further aggressive treatment, even if necessary to prolong life.89 Just a few months later, Hedley J was asked to reverse his decision as Charlotte’s condition was improving.90 This case would take many more twists and turns, with Hedley J reversing his decision several more times before this legal saga ended.91 Miraculously, Charlotte defied the expectations of her doctors. She was eventually discharged from hospital and was living in foster care where she was learning to walk.92

Uncertainty arising over the condition of the child in question makes the judge’s application of the “best interests” test difficult.93 Moreover, the shifting factual matrix further complicates a judge’s assessment of the wider social and emotional factors. Charlotte Wyatt’s case further demonstrates that such cases impose an enormous strain on the legal system. It would involve repeated applications to court by both the hospital and the parents. These applications might often be made ex parte and at the eleventh hour. Such cases, unsurprisingly, are protracted affairs. Moreover, the court may have to monitor the child’s condition to determine if any variation of the order granted is necessary. Doing this places the court in an difficult position – after all the court is meant for dispute resolution, it is not a “general advice centre”.94 Finally, such hearings are usually acrimonious (hardly surprising given what is at stake) and subject to intense media scrutiny. The way forward

Given that courts are still unable to distance themselves from medical opinion in the assessment of best interests, cases such as Charlotte Wyatt are problematic. This is because the constantly changing factual matrix makes the rendering of a medical opinion difficult.

180

85 Re Wyatt (a child) (medical treatment: continuation of order) [2005] EWHC 693, [3]. 86 Portsmouth NHS Trust v Wyatt [2004] 1 FLR 21. 87 ibid, [1]. Charlotte had chronic respiratory and kidney problems coupled with the most profound brain damage that left her blind, deaf and incapable of voluntary movement or response. 88 ibid, [12]. 89 ibid, [38]. 90 Portsmouth NHS Trust (No 1) v Wyatt [2005] EWHC 117. 91 Re Wyatt (a child) (medical treatment: continuation of order) [2005] EWHC 693. Hedley J’s decisions were upheld by the Court of Appeal. See Portsmouth Hospital NHS Trust v Wyatt [2005] EWCA Civ 1181. This however, did not conclude the saga. See also Portsmouth NHS Trust v W [2005] EWHC 2293; Re Wyatt [2006] EWHC 319. 92 Stewart Payne, ‘Christmas with foster parents for Charlotte’ (The Telegraph, 21 Dec 2006) < https://www. telegraph.co.uk/news/uknews/1537483/Christmas-withfoster-parents-for-Charlotte.html> accessed 5 June 2019.

It is clear then that the courts are not the best forum by which to resolve such issues.95 While the judge undertakes the role of the mediator, his role is to ultimately decide what is to be done in order to resolve the dispute.96 The outcome is a binary one, it is either in the patient’s best interest to continue to provide life supporting treatment, or not.97 There is 93 See Margot Brazier, “An Intractable Dispute: When Parents and Professionals Disagree” 2005 (13) Med Law Review 412, 418. She notes that the more Charlotte proves “the experts wrong in terms of the quantity of her life, the harder it will be to persuade her parents that [the] doctors were right in their judgement of its quality”. 94 Portsmouth Hospital NHS & Anor v Wyatt & Anor [2005] EWCA Civ 1181, [117] – [118]. 95 Paul Tan, “When Death Knocks, Who Will Answer and What Will They Say?” (2006) 18 Singapore Academy of Law Journal, [58]. See also Thaddeus Pope & Ellen A. Waldman, “Mediation at the End of Life: Getting Beyond the Limits of the Talking Cure” (2007) 23(1) Ohio State Journal on Dispute Resolution 143, 147. 96 Margot Brazier, “An Intractable Dispute: When Parents and Professionals Disagree” (2005) 13 Medical Law Review 412, 417. 97 Richard Huxtable, Law, Ethics and Compromise at

ARTICLES: WRITTEN MOOT 3

no middle ground. This may be particularly devastating if the parents are on the losing end, to which no amount of kind words in the judgement can “assuage the anguish of [their] bereavement”.98 Litigation is also expensive and costly.99 Moreover, the adversarial nature of litigation may make it more difficult for doctors and parents to come to an agreement.100 We therefore turn to consider two alternatives to litigation in resolving such disputes; (a) preserving the partnership of care between the medical team and parents and (b) mediation. Partnership of care Where the treatment of children is concerned, there needs to be a “partnership of care”101 between medical professionals and parents.102 Absent this “partnership of care”, resolving disputes between the doctors and parents through non-legal means may be difficult, if not impossible.103 This is evident from the

bitter disputes between the parents and medical professionals which are characteristic of cases in this field.104 Improving communication between the doctors and parents along with the provision of counselling services may help preserve this “partnership of care”, pre-empting any disputes before they arise. Managing unrealistic patient expectations through effective communication may also be useful.105 Doctors should communicate effectively with the parents to avoid giving them unrealistic expectations as to their child’s medical prognosis. Bad news should be broken with tact, so as to ensure cooperation as to future treatment options for the child.106 This, however, does not mean that doctors should bend over backwards and accommodate unreasonable demands. There can be room for debate and disagreement whilst allowing collaborative decisions concerning treatment options to be made.107 Mediation

the Limits of Life (Routledge, 2013), 78 citing Kennedy I. and Grubb A. (1993) “Withdrawal of artificial hydration and nutrition: incompetent adult”, Medical Law Review 359, 364. The authors pointed out: “Surely continuing or withdrawing is in the patient’s best interests: not both!”. 98 This quote originated from the Bixby letter which Abraham Lincoln supposedly penned to Lydia Bixby expressing condolences that she lost five sons to the Civil War. See “A Century-Old Abraham Lincoln Mystery May Finally Have an Answer” (Time, 19 July 2017) <https://time.com/4855857/ abraham-lincoln-john-hay-bixby-letter/> (accessed 27 June 2019). 99 Dominic Wilkinson & Julian Savulescu, “Hard lessons: learning from the Charlie Gard case” (2018) 44 Journal of Medical Ethics 438, 440. 100 ibid. 101 Margot Brazier, “An Intractable Dispute: When Parents and Professionals Disagree” (2005) 13 Medical Law Review 412, 418. See also Mark A. Hall, “Law, Medicine and Trust” (2002) 55 Stanford Law Review 463. He points out that “trust sets the stage for, even if it does not resolve, most of the major issues in health care law”. Arguably, the foundation of the partnership of care, is trust. 102 Paul Tan, “When Death Knocks, Who Will Answer and What Will They Say?” (2006) 18 Singapore Academy of Law Journal, [58]. See R v Portsmouth Hospitals NHS Trust ex parte Glass [1999] 2 FLR 905. In that case, the parent’s relationship with the medical team had deteriorated to the extent that Ms Glass applied to court for a declaration regarding the treatment her son could expect if he was admitted to hospital. 103 See remarks made in Portsmouth Hospital NHS & Anor v Wyatt & Anor [2005] EWCA Civ 1181, [119]. The court noted that “[i]deally, Charlotte’s treatment (using that term to include the decision not to ventilate) should be a matter of agreement between the Trust and Mr and Mrs Wyatt. The

As one author put it; “the gentler touch for matters that ha[s] such great impact on human lives and emotions evidently requires a holistic and multi-dimensional approach” [emphasis added].108 The acrimonious proceedings observed in the strains on Charlotte’s parents have been substantial, but it has to be said that their conduct in the past has been less than helpful. Reporting the doctors in this case to the police achieves nothing. Everybody sympathises with the parents of a disabled child, but there are limits to that sympathy when parental conduct ceases to bear any relation to the child’s welfare”. 104 ibid. 105 See Jennifer Fong and Nancy Longnecker, ‘DoctorPatient Communication: A Review’ (2010) 10 The Ochsner Journal 38, pp 40 - 41. 106 See Linda Beecham, ‘Learning to break bad news’ (1994) 314 British Medical Journal 1501. Beecham notes that “[m]any patients’ complaints are due to poor communication between doctors and patients, and this leads to frustration, dissatisfaction, and anxiety”. Learning to break bad news is also essential for nurses. See Margaret Rosenzweig, ‘Breaking bad news: A guide for effective and empathetic communication’ (2012) 37(2) The Nurse Practitioner. 107 Sabine Vanacker, ‘The Story of Isabel’ in Lynn Hagger and Simon Woods (eds), (Ashgate 2013), 173. 108 Dorcas Quek Anderson, “Medical negligence proceedings in Singapore: Instilling a gentler touch” (2018) International Academy of Comparative Law 20th General Congress 2018, July 22-28.

181

ARTICLES: WRITTEN MOOT 3

cases of Alfie Evans and Charlie Gard may be avoided through the use of alternative dispute resolution mechanisms. Mediation might be one way of ensuing an “amicable resolution of conflict or disagreement between family members or between families and hospitals”.109 Locally, there have been attempts to move away from the “conventional adversarial approach” in medical negligence cases.110 This could be extended to cases where a dispute has arisen as to the proper course of treatment in an Alfie Evans type scenario. For instance, instead of seeking a declaratory order from the court, the hospital could utilise the healthcare mediation scheme pioneered by the Ministry of Health.111 There are particular advantages in preferring mediation over litigation where a child is concerned in end-of-life decisions. For one, mediation affords the family of the patient privacy as compared to an open trial.112 Moreover, the flexibility of mediation would allow developments in the child’s medical condition to be better accounted for.113 In addition, mediation is conducted in a “more civil and less costly atmosphere” as compared to litigation.114 Finally, mediation allows for a wider range of options which might be more conducive in getting both parties to compromise.115

182

109 ibid. 110 Dorcas Quek Anderson, “Medical negligence proceedings in Singapore: Instilling a gentler touch” (2018) International Academy of Comparative Law 20th General Congress 2018, July 22-28, p 8. 111 ibid, citing MOH Holdings, “Healthcare Mediation Scheme” <http://www.mohh.com.sg/hms/home> accessed 5 June 2019. 112 ibid. 113 J. Daniel Breen, “Mediation and the Magistrate Judge” (1996) 26 University of Memphis Law Review 1007, 1014. 114 ibid, 1008. 115 See Emma Cave and Emma Nottingham, “Who Knows Best (Interests)? The Case of Charlie Gard” (2017) 26(3) Medical Law Review 500, 511. The authors point out that the Great Ormond Street Hospital’s failed attempts at mediation was because there was “a stark choice between two options: allow the unproven therapy or withdraw treatment and allow Charlie to die”.

Higher threshold for jurisdiction Whilst alternatives to litigation seem attractive, in certain cases, judicial oversight or intervention may still be needed. Under what circumstances, then, would judicial oversight be necessary? Three situations come to mind. First, mediation is by no means a silver bullet in resolving such disputes. There must be a willingness to cooperate and compromise between both parties116. Absent this and mediation will, in all likelihood, fail. It is in such cases where one party stubbornly refuses to compromise that the court’s function as the final arbiter is useful.117 Second, in such cases, emotions often run high. The actions of families or general members of the public may threaten the safety of the medical staff, or other patients warded at the same hospital.118 Finally, judicial oversight may be necessary if there is a risk of an abuse of process. Consider a case where the legally incompetent child in question is an orphan and the doctors have decided that withdrawing life supporting treatment is the best course of action. Judicial oversight would arguably be needed to determine if the proposed course of treatment is indeed in the child’s best interests.119 Therefore, the court should be slow to find that it has the jurisdiction to hear the dispute in cases that are not similar to the ones outlined above. The threshold that must be met before the court has jurisdiction should be a high one – airing such cases in court should be the exception and not the norm. 116 Machteld Pel, “The Ingredients for a Justified and Successful Referral to Mediation: Balance And Trust” (Wolter Kluwer Mediation Blog, 9 October 2011) <http:// mediationblog.kluwerarbitration.com/2011/10/09/ the-ingredients-for-a-justified-and-successful-referral-tomediation-balance-and-trust/> (accessed 27 June 2019). 117 Richard Huxtable, Law, Ethics and Compromise at the Limits of Life (Routledge, 2013), 8. 118 This was the case in Alfie Evans where protestors stormed the hospital where Alfie was warded at. See ‘Alfie Evans: Protesters try to storm Alder Hey hospital’ (BBC, 23 April 2018) <https://www.bbc.com/news/uk-englandmerseyside-43867132> accessed 5 June 2019. 119 Paul Tan, “When Death Knocks, Who Will Answer and What Will They Say?” (2006) 18 Singapore Academy of Law Journal at [57]. See also R v Portsmouth Hospitals NHS Trust ex parte Glass [1999] 2 FLR 905.

ARTICLES: WRITTEN MOOT 3

Conclusion Where an end-of-life decision has to be made concerning a legally incompetent child, perhaps a legal solution is not the best option. The ubiquitous â&#x20AC;&#x153;best interestâ&#x20AC;? test, well known in family law, is riddled with problems, both doctrinal and practical. Alternative ways of resolving such matters should therefore be considered. In that regard, alternative dispute resolutions such as mediation may provide a better alternative to resolving such disputes should they arise. Undoubtedly, doing so would greatly ease the enormous emotional burden on parents and doctors alike, sparing both parties from lengthy and acrimonious court proceedings. Above all, it may afford parents the time and space to come to terms with the grief of their loss.

183

ARTICLES: GUEST WRITER 1

Lift Not the Painted Veil Professor Walter Woon, Dean, Singapore Institute of Legal Education Lift not the painted veil which those who live Call Life; though unreal shapes be pictured there, And it but mimic all we would believe … Percy Bysshe Shelley For some reason that is not entirely clear, candidates for the Company Law examination in Part A of the Bar Examination seem to have reams of notes on “lifting the corporate veil”. These are duly trotted out (sometimes cut and pasted) at the slightest opportunity, even when the examination question does not even raise the the issue. For those who have never sat for an open book examination there is a misconception that flinging a mass of material at the examiner is the way to pass. This misconception should be firmly exorcised. In a Part A Company Law examination the candidate is put in the position of a legal assistant who is given a file by the Senior Partner of his firm. His task is to prepare a brief for the Senior Partner. Candidates fail because they do not identify the crucial issues, much less present a useful brief. Clients pay for solutions, not erudition. The blunderbus approach to answering Part A examination questions essentially is a plea to the Senior Partner to pick out what is relevant from the unfiltered deluge of words. This is not what an associate is paid for; nor is it the way to pass the Part A examination. Back to lifting the corporate veil. It is essential to appreciate what corporate personality entails. One starts with Aron Salomon v A Salomon & Co Ltd [1897] 1 AC 22 (House of Lords). This has repeatedly been accepted as good law in Singapore. Mr Aron Salomon set up the company and ran the business. Effectively, it was his business. The other shareholders were his wife, a daughter and

his four sons. Salomon had 20,001 shares out of 20,007. Debentures were issued to a creditor to secure a loan. The business later failed. It went into liquidation. After paying off the debentures, there wasn’t enough left to pay the unsecured creditors. The liquidator sued Salomon, seeking an indemnity from him against the whole of the unsecured debt; in other words, to make him responsible for the debts. The case worked its way through the courts until it reached the chill and distant heights of the House of Lords, where it was held that Aron Salomon was not liable for the debts of his company. It is therefore permissible to set up a company to shield oneself from the consequences of business failure. This is perfectly legitimate. Thus, for example, in Cavanagh Investment Pte Ltd v Kaushik Rajiv [2013] SGHC 45 (High Court), the Lee Tat ‘group’ consisted of several companies of which Cavenagh Investment was one.1 Cavenagh owned two properties in Pebble Bay condominium. It had no employees. CMF was its sole shareholder and director. The properties were managed by Lee Tat Property Management Pte Ltd, which was run by CMF. Among its employees was one R. R defrauded D by forging CMF’s signature on a Tenancy Agreement. He collected the rent. This fraud was later discovered and Cavenagh sued D for damages for trespass to land. The primary defence was that R was authorised, but this was rejected. The fallback defence was that Cavenagh was vicariously liable for R’s acts. The problem was that R was an employee of Lee Tat Property, not Cavenagh. It was therefore argued that the corporate veil should be lifted so as to treat the two companies as effectively the same entity. Chan Seng Onn J rejected this argument. 1 This is an example of the common mistake of treating companies sharing a common shareholder as a group. Strictly speaking, a group consists of a holding company and its subsidiaries.

185

ARTICLES: GUEST WRITER 1

CMF was entitled to structure her companies so that each one held different properties. In the absence of ‘fraud or serious abuse of the corporate form’ the courts would not treat a group as a single entity. The defence therefore failed on this point. However, the separate entity doctrine is often abused in ways that were not envisaged in the Salomon case. The courts therefore are often invited to ignore the Salomon rule that a company is an entity distinct from the corporators. It is this process that is picturesquely referred to as “lifting the corporate veil”. It is essential to appreciate that lifting the veil actually encompasses three separate submissions. The first is that the liability of the company is the liability of the corporators. The second is that the company’s property belongs to the corporators. The third is that the corporators and the company are one and the same. These are distinct situations and should not be confused. To maintain intellectual rigour, it is vital to distinguish situations where no lifting of the veil is necessary. Firstly, a person is always liable for any tort or crime committed by himself. The fact that a company may also be liable does not absolve him. The company is liable in addition to, not in lieu of the malefactor. This must be so on principle. Consider the following scenario. Abel cheats Baker using his company, Abel Pte Ltd, in which he is the sole member and director. All the invoices are issued by Abel Pte Ltd. It would be appalling if Abel could squirm out of civil and criminal liability by pleading that the deed was done by the company and not by him. Thus, for example, in TV Media Pte Ltd v De Cruz Andrea Heidi [2004] SGCA 29 (Court of Appeal) A sued the defendants for damages caused by Slim 10, a slimming drug. D1, the Chinese manufacturer, could not be served with the writ. D2 Health Biz Pte Ltd imported 186

and sold Slim 10. D3 was the director and principal shareholder of Health Biz. D4 TV Media was the sole distributor of Slim 10 in Singapore. Tay Yong Kwang J found Health Biz (importer) and TV Media (distributor) liable for A’s liver damage. He also found D3 (controller of Health Biz) personally liable for authorising and/or procuring Health Biz’s negligent acts of importing and selling Slim 10. In dismissing the appeal Yong Pung How CJ held that D3 was clearly the ‘controlling mind and spirit of Health Biz.’ He was the one that had procured and directed the wrong to be done. He was therefore also liable along with Health Biz. If a company’s business is carried on for a fraudulent purpose, the people who were knowingly party to this may be ordered to pay the debts of the company wholly or in part without limitation of liability. Section 340(1) of the Companies Act provides: 340.—(1) If … in any proceedings against a company, it appears that any business of the company has been carried on with intent to defraud creditors of the company or creditors of any other person or for any fraudulent purpose, the Court, on the application of … any creditor or contributory of the company, may, if it thinks proper to do so, declare that any person who was knowingly a party to the carrying on of the business in that manner shall be personally responsible, without any limitation of liability, for all or any of the debts or other liabilities of the company as the Court directs. Thus, a creditor may sue a company and join as co-defendants any persons who are ‘knowingly a party to the carrying on of the business’ in a fraudulent manner. These persons can be ordered by the court to pay the debts of the company, wholly or in part. Secondly, when it is necessary to attribute a state of mind to a company, of necessity one must identify the humans who are in control.

ARTICLES: GUEST WRITER 1

This is a question of attribution, not lifting the corporate veil. Thus, for instance, in New Line Productions Inc v Aglow Video Pte Ltd [2005] SGHC 118 (High Court) New Line owned the copyright in three films. It sued the defendants essentially for breach of copyright. Most of the defendants were companies in the “TS Group”. Each company operated one outlet and kept separate accounts. All had a common registered office with TS Laser and TS Entertainment. The marketing manager of TS Laser testified that the TS Group were independent companies operating together in a loose arrangement (in other words, not a group in the legal sense). Each company had its own directors and shareholders and maintained its own accounts. The group was started by D20, who expanded by buying other companies; each company operated one outlet. The directors and staff reported to D20 and took instructions from him. Most of the companies were owned by D20, his wife and brother. Tay Yong Kwang J found that far from being a loosely-knit group, the TS Group were ‘little pieces of mosiac forming a complete mural, glued together by the four directing minds’. TS Laser was the parent or nerve centre. The companies shared many common directors and shareholders, and took instructions from D20, who appeared to be beneficial owner of practically all of the companies. Given these facts, Tay J held that the entire group were liable for dealing with infringing products. Although Tay J treated this as an instance of lifting the corporate veil, it is better to see it as an example of attribution. All the companies “knew” that the videos were counterfeit, because the people who controlled those companies had that knowledge. One has to look to human beings to determine what a corporate entity knows. Thirdly, where the “company” does not exist. This is often referred to as a sham. A distinction must be made between a sham company and a sham transaction between existing companies. A fraudster may purport to contract in the name of a company; anyone can run off a letterhead on a good printer. If the company has not been

incorporated anywhere in the world, then clearly it is the fraudster who is liable for whatever transactions have been entered into. Fourthly, a company may hold assets on trust for its controllers. This is a question of evidence. Companies are often used to hide assets from the tax authorities (almost invariably illegal) or wives (not very nice, but usually not illegal per se). Obviously, the intent of the controller is to use the assets as if they were his own, despite the legal fiction that the company owns them. In such a case, a trust may be inferred. Needless to say, it is unnecessary to ignore the Salomon rule, as a company may be trustee of assets for the benefit of its controller. Fifthly, a company may act as an agent for its controller. Often a businessman negotiates with professional (an architect, for instance) for the provision of services. He does this because the professional has expertise and a reputation. But when it comes to billing, the professional sends the invoice in the name of a company that he controls. This does not change the essential nature of the contract, which is between the professional and the client. It does not absolve the professional of liability for misrepresentation or negligence. The company in such a case is only an agent for collection of money for the professional. It is unnecessary to ignore the separate legal personality of the company. With this in mind, one can consider the situations in which the Salomon may be ignored by the courts. In the Singapore context, there are three broad situations in which a court may be induced to ignore the Salomon rule: • Purposive interpretation of a statute • Court exercising an equitable or analogous discretion • Company acting as controller’s alter ego

187

ARTICLES: GUEST WRITER 1

Section 9A of the Interpretation Act (Cap 1) states: 9A.—(1) In the interpretation of a provision of a written law, an interpretation that would promote the purpose or object underlying the written law (whether that purpose or object is expressly stated in the written law or not) shall be preferred to an interpretation that would not promote that purpose or object. Companies are often used as a means to evade uncomfortable consequences of a statute or to take advantage of certain benefits. The classic case would be taxing statutes. In the interpretation of a statute a Singapore court will consider whether treating the company as an entity separate from the corporators will defeat the purpose of the legislation. If the answer is ‘yes’, then the court will ignore the Salomon rule. From an academic’s point of view this is unsatisfyingly vague. From a practitioner’s point of view the advice one gives to a client is that if he is incorporating a company for the purpose of avoiding a statutory provision or exploiting a statutory benefit, this may not work if the client in reality controls and owns the company. In other words, if the reality is that the company is not a separate entity, the court may ignore it when purposively interpreting a statute. In the second class of cases, remember that the eye of equity is not blind to reality. If the reality is that the company is not independent of the corporators, a court of equity may decline to grant an equitable remedy like specific performance of a contract or an injunction. This applies also in situations where the court is exercising a discretion akin to an equitable one. So, for instance, in Raffles Town Club Pte Ltd v Lim Eng Hock Peter [2012] SGCA 62 (Court of Appeal) RTC was marketed as an exclusive premier club. Its membership drive was wildly successful and it recruited over 19000 members (the members of the club were not shareholders of RTC). 4885 members began a representative action against RTC for breach of 188

contract. The suit was dismissed by the High Court but allowed on appeal. As RTC had not enough cash to pay the damages a scheme of arrangement was agreed which cost RTC $53m. Subsequently, some of the former directors sued the current directors for defamation in relation to statements made in relation to the scheme of arrangement. The current directors were found liable. In the present litigation RTC sued the former directors for breach of directors’ duties. The claim was dismissed and so was the appeal. The Court of Appeal held that the suit against the former directors was essentially an action by the new shareholders/ directors using the company as a vehicle. They were ordered to personally bear the costs on an indemnity basis. The third class of cases is the most amorphous. One traces the origin of the alter ego doctrine to dicta by V K Rajah JA in two cases. In Singapore Tourism Board v Children’s Media Ltd [2008] SGHC 77 (High Court) [affirmed by the Court of Appeal in Children’s Media Ltd v Singapore Tourism Board [2008] SGCA 45] STB contracted with Children’s Media (D1, UK company) and Tribute Third Millenium (D2, UK company limited by guarantee, holding company of D1) to stage Listen Live. D3 was CEO and sole shareholder of D2. D3 claimed to have expertise in organising such events. He represented to STB that Listen Live would be the culmination of a 180-day worldwide campaign that would raise substantial funds and feature film and music stars and reach 500 million people in 80 countries. The contract was signed by STB with Children’s Media, a special purpose vehicle. Children’s Media failed to provide the necessary artistes, broadcasters and financing, ostensibly due to events beyond its control (the 2004 tsunami). STB accepted this and agreed to a second agreement. Children’s Media again failed to secure the artistes etc by the stipulated deadline. D3 represented to STB that if the event could be postponed, they could still proceed. STB then entered a third agreement. Children’s Media still couldn’t live up to its obligations and STB terminated

ARTICLES: GUEST WRITER 1

the agreement and sued the three defendants. Third Millenium and D3 claimed that STB only contracted with Children’s Media and therefore could not sue them. They were nonetheless held liable and the appeal was dismissed by the Court of Appeal. In the course of the Court of Appeal’s judgment V K Rajah JA said that D3 was ‘not just the controlling mind of these entities ... he was in short the alter ego of these entities’. They were no more than corporate puppets dancing to his tune. Their liabilities were therefore his liabilities. In Alwie Handoyo v Tjong Very Sumito [2013] SGCA 44 (Court of Appeal) one of the issues was whether money paid to O, a company incorporated in the British Virgin Islands, was to be treated as having been paid to A. A was the sole director and shareholder of O. He admitted in his affidavit that O was controlled by him and that he was its directing mind and will. He stated in cross-examination that he was entitled to the money paid to O. The manner in which he procured the payments suggests that he made no distinction between O and himself. O was incorporated in the BVI solely to receive payment. V K Rajah JA held that the ground of alter ego is different from that of sham or façade. The key question that must be asked is whether the company is carrying on the business of its controller.

record and claimed expertise, the interposition of a company Y specially set up for the purpose does not absolve X of liability. There would be two contracts: one with X and a collateral one (usually for billing purposes) with Y. Nor does the interposition of Y give X immunity from the consequences of misrepresentation or fraud. A final word: there are literally hundreds of cases from all over the world on lifting the corporate veil. Gallons of ink have been spilt and forests of trees cut down for the sake of commentary on these cases. One cannot be dogmatic, nor should one expect consistency of doctrine. There is no “theory of everything” to account for all instances of lifting the corporate veil. The general rule is that the separate entity principle in Salomon’s case is the norm and that any argument to ignore it will rarely succeed. When advising clients, it would be best to address the problem in a manner that does not require a frontal attack on Salomon. In litigation it is well to take Shelley’s words to heart and not attempt to lift the veil unless absolutely necessary.

The Alwie Handoyo case is probably better treated as an example of the company acting as an agent to collect money for its controller or as trustee, neither of which requires the Salomon rule to be ignored. Indeed, in Aron Salomon v A Salomon & Co Ltd [1897] 1 AC 22 the business was in reality that of Aron Salomon. This leaves the Children’s Media case as a tantalizing instance of treating the controller as responsible for the company’s debts. That case could indeed be treated as an example of a person (D3) being liable for his own wrongs (fraudulent misrepresentation). However, moving forward, perhaps the best approach is to look at the commercial reality. If a client clearly wants to contract with X by reason of his purported track 189

ARTICLES: GUEST WRITER 2

Liability for Road Accidents Caused by Driverless Cars Ken Oliphant* Head of Law School, University of Bristol Introduction Sally was coming down the lake road, so I waved to her and called her by name. I always liked to see Sally. I liked all of them, you understand, but Sally’s the prettiest one of the lot.1 Before you ask: Jake, our narrator, isn’t talking about a woman. Sally is a car. And not just any car. To resume Jake’s narration: ‘Sally was a 2045 convertible with a… positronic motor. She had the cleanest, finest lines I’ve ever seen on any model, but none. For five years, she’d been my favourite…. In all that time, there’d never been a human being behind her wheel. Not one.’2 Sally, you’ll have guessed, is a driverless car, a creation of science fiction (‘SF’) - in fact, the eponymous heroine of Isaac Asimov’s short story Sally, published in 1953. That makes it possibly the first self-driving vehicle in the SF canon, the precursor to more celebrated examples like the Batmobile, Knight Rider’s KITT and Total Recall’s Johnny Cab.3 Cars with a computer ‘brain’ that can drive themselves have become a familiar SF trope.

cameras and an on-board computer. Admittedly, it took 15 to 20 minutes to plan and execute every one-metre move, but it was a start. Significant further development had to wait until the new millennium, when the US military (through its Defense Advanced Research Projects Agency) established a challenge race for AVs, open to all comers, across the Mojave Desert. In the first challenge, none of the entrees made it to the finish line; second time around, five vehicles completed the course. In 2010, an adapted mass-market hatchback, a Prius, succeeded in self-navigating 1000 miles of California roads, fulfilling a commission by Google founder, Larry Page. The market potential of AVs was now becoming obvious, and manufacturers began to sell models with more and more autonomous driving features, as well as to design new fully autonomous vehicles like Toyota’s ePalette, conceived as being able to provide a range of functions other than transportation, including mobile office or retail space, mobile medical clinics and mobile restaurants.5 Developmental Accidents

But autonomous vehicles (‘AVs’) can no longer be considered as just the product of speculative fiction.4 In 1961, the ‘Stanford Cart’ was able to navigate around obstacles through the use of

190

* Professor of Tort Law and Head of School, University of Bristol Law School, United Kingdom. Email: ken. oliphant@bristol.ac.uk. 1 Isaac Asimov, ‘Sally’ (1953) 2:3 Fantastic 34 at 35. 2 Ibid 36. 3 See respectively online: https://en.wikipedia. org/wiki/Batmobile; https://en.wikipedia.org/wiki/KITT; https://talkingpointz.com/johnny-cab/. All online materials referenced in this article were available to access on 21 June 2019. 4 The examples in the text are sourced from Alex Davies, ‘The WIRED guide to self-driving cars’ (13 December 2018) online: https://www.wired.com/story/guide-selfdriving-cars/.

Putting these AVs on the roads, and especially into situations of potential conflict with other road users, has proved far from straightforward. The process of development and testing has been beset with problems, with a succession of accidents, near-misses and other types of failure. In June 2016, a Tesla test driver died in Florida in the first fatal crash involving a vehicle in autopilot mode. The autopilot sensors on the vehicle were not designed to, and did not, identify a truck crossing the highway ahead and drove into it without reducing velocity 5 See https://www.toyota.ca/toyota/en/ connect/2000/toyota-e-palette-concept-vehicle-ces-2018.

ARTICLES: GUEST WRITER 2

or applying the braking system. The National Transportation Safety Board accident report recorded that the probable cause of the accident was the truck driver’s failure to yield the right of way to the car, combined with the car driver’s inattention due to overreliance on vehicle automation, which resulted in his lack of reaction to the truck’s presence. The report found that the car’s operational design contributed to the driver’s overreliance on the automation because it permitted his prolonged disengagement from the driving task and his use of the automation in ways inconsistent with guidance and warnings from the manufacturer. 6 In March 2018, a self-driving Uber vehicle killed a woman on a public street in Arizona, in what was reported as the first fatal crash involving a self-driving vehicle and a pedestrian.7 The vehicle sensed the woman’s presence six seconds before the impact, but (as a federal investigation subsequently found), its emergency braking system was not enabled, ostensibly to avoid erratic vehicle behaviour. Uber stated that it relied upon operators to intervene and take action, though the system itself is not designed to alert them, and in this instance, the driver failed to brake until after the impact had occurred.8 It was reported little than a week later that the dead woman’s family had reached a settlement with Uber, pre-empting possible legal proceedings, on terms that have not been disclosed. Around the same time, it was 6 See National Transportation Safety Board, ‘Highway Accident Report: Collision Between a Car Operating With Automated Vehicle Control Systems and a Tractor-Semitrailer Truck Near Williston, Florida May 7, 2016,’ NTSB/HAR17/02 (12 September 2017) online: https://www.ntsb.gov/ investigations/AccidentReports/Reports/HAR1702.pdf. Another Tesla driver was killed in an accident in Mountain View, California on 23 March 2018. National Transportation Safety Board, ‘Preliminary Report: Highway HWY18FH011’ (7 June 2018) online: https://www.ntsb.gov/investigations/ AccidentReports/Reports/HWY18FH011-preliminary.pdf. 7 ‘Self-driving Uber kills Arizona woman in first fatal crash involving pedestrian’, The Guardian, 19 March 2018, available at https://www.theguardian.com/technology/2018/ mar/19/uber-self-driving-car-kills-woman-arizona-tempe. 8 National Transportation Safety Board, ‘Preliminary Report: Highway HWY18MH010’ (24 May 2018) online: https://www.ntsb.gov/investigations/accidentreports/ reports/hwy18mh010-prelim.pdf.

revealed that Uber’s testing program in Arizona was authorised by the state governor, without informing the public. Uber subsequently suspended its AV testing in Arizona and other locations in the US, as well as in Toronto.9 The tragic story provides a foretaste of the further litigation surrounding AV collisions that is surely inevitable, even if the direct causes here were human choice and human inadvertence. The Wider Context For now, AVs are hardly ever to be found on our roads – and not at all, so far as I am aware, on public roads in Europe, at least if we speak only of vehicles with ‘high’ or ‘full’ automated driving capacity (denoted Levels 4 and 5 on the now standard scale).10 Even Level 3 vehicles – with ‘conditional’ automation, meaning that drivers can choose whether they drive or not - are rare, as can be seen from the graphical representation below of AV production penetration in the UK. However, aspects of lower level automation are now familiar features of new cars (cruise control, ABS, lane-keeping assistance, self-parking etc) and it is projected that all vehicles produced in the UK by 2027 will have at least L3 technologies embedded in them and that there will be a 25% penetration of fully autonomous vehicles by 2030.

9 Mark Harris, ‘Exclusive: Arizona governor and Uber kept self-driving program secret, emails reveal’, The Guardian, 28 March 2018, available at https://www.theguardian.com/ technology/2018/mar/28/uber-arizona-secret-self-drivingprogram-governor-doug-ducey. 10 SAE International, ‘Taxonomy and Definitions for Terms Related to On-Road Motor Vehicle Automated Driving Systems,’ Recommended Practice 3016_201806 (June 2018)..

191

ARTICLES: GUEST WRITER 2

Table: forecast of UK production of autonomous vehicles.11

the Government introduced a Bill in Parliament in 2017 to establish a new regulatory and liability regime14 and revived it under a different name after the General Election of that year, passing in July the year after and becoming the Automated and Electric Vehicles Act 2018. Policy Choices

Though the market for AVs is so far undeveloped, the attraction of investing in further AV development – for both manufacturers and governments – are obvious. As summarised in a recent policy paper by the UK Department of Transport, The Pathway to Driverless Cars, the key advantages of AVs are to create more free time for persons generally (it being claimed that the average driver can save up to six working weeks of driving time each year), to improve health and safety, inasmuch as 94% of road injuries and deaths are currently attributed to human error, to reduce emissions and ease congestion, and to increase access to vehicles for everyone.12 Necessary steps in this process are to promote the UK as a place to test, sell and use AVs by creating a regulatory landscape that is appropriately permissive, clarifying liability rules and giving victims easy access to compensation.13 With these objectives in mind,

192

11 KPMG, ‘Connected and Autonomous Vehicles – The UK Economic Opportunity ‘ (March 2015), p 9 (chart and table), available at https://www.smmt.co.uk/wp-content/ uploads/sites/2/CRT036586F-Connected-and-AutonomousVehicles-%E2%80%93-The-UK-Economic-Opportu...1.pdf. 12 Department of Transport, ‘The Pathway to Driverless Cars: A detailed review of regulations for automated vehicle technologies’ (February 2015), paras 1.2–1.8 and Figure 13.1, available at https://assets.publishing.service.gov. uk/government/uploads/system/uploads/attachment_data/ file/401565/pathway-driverless-cars-main.pdf 13 Department of Transport, The Pathway to Driverless Cars (0000). The UK Government’s policy response was developed by the Department in a series of research and policy papers with the recurring title ‘The Pathway to Driverless Cars’.

Before we look at the details of the Act, what are the basic policy choices that the legislator must make as regards the liability issues that are the focus of this paper? Who should bear the risk of damage resulting from the use of an AV? In what circumstances should the victim be granted compensation?15 Conceivably, the loss could simply be left to lie with the victim: casum sentit dominus (let the loss lie where it falls). That is the basic starting point in all legal systems,16 though it is also universally recognised that there may be good reason for shifting the loss through the imposition of a liability to compensate. The initial response is generally to do this in the case of fault – the negligence of the driver, as it most commonly is. But with an AV, there may be no driver, so it is necessary to ask whether other parties can be made to bear the risk and on what basis: the carowner, its insurer, the car-maker, the software designer, an at-fault third party, or perhaps the State? Fault, of course, remains an option – but it is not the only possible basis of allocation.17 Further, even where in a first step the loss is shifted from the victim to another person, it may remain possible for the latter to pass 14 Bill 43, Vehicle Technology and Aviation Bill, UK Parl, Sess 2016-17, introduced 22 February 2017. 15 See further David C. Vladeck, ‘Machines Without Principals: Liability Rules and Artificial Intelligence,’ (2014) 89 Wash. L. Rev. 117 at 125–129. 16 See further Helmut Koziol, Basic Questions of Tort Law from a Germanic Perspective (Vienna: Jan Sramek Verlag 2012) at para 1/1; Ken Oliphant, ‘Basic Questions of Tort Law from the Perspective of England and the Commonwealth,’ in Helmut Koziol, ed., Basic Questions of Tort Law from a Comparative Perspective (Vienna: Jan Sramek Verlag 2015) at paras 5/8–5/10. 17 As to liability in the absence of either fault or defect, see Vladeck, supra note 15 at 146–47.

ARTICLES: GUEST WRITER 2

the loss on through further litigation so as to achieve the goal of loss distribution. The first defendant then becomes simply the conduit through which the loss is ‘channelled’ to the ultimate risk-bearer or group of risk-bearers. The rationale for doing so may be that the latter are better able to bear the risk, or to distribute it further without additional litigation (eg through product pricing), or because they are in the best position to control the risk. The possible justifications, and their possible combinations, are various.18 We therefore have to ask the following questions. First, on what basis should the law shift the loss from the victim to another person? Second, to whom should the loss be shifted? And third, what further loss distribution should there be, and on what basis? The Automated and Electric Vehicles Act 201819 Accidents caused by self-driving cars may be dealt with under the established negligencebased regime of tortious liability and/or by other applicable liability regimes. If the ‘operator’ of a self-driving car did not pay due attention in advance of the accident, and could have avoided its occurrence, then he or she would potentially be liable in negligence to anyone who suffered damage as a result in the usual way. But if there is no actual driver, the focus in many road accident cases will necessarily shift to some alternative defendant. One possible target is the producer of the vehicle, with a concomitant shift of focus from the general fault liability regime governed by the law of negligence to the strict liability for damage caused by defective products under the Consumer Protection Act 1987. However, proving a defect in the vehicle may not be a simple matter.20 18 As to distributive justice through tort law, see generally Tsachi Keren-Paz, Torts, Egalitarianism and Distributive Justice (Ashgate, 2007). 19 Automated and Electric Vehicles Act 2018 (UK), 2018, c 18. 20 As to the difficulties, see Kenneth S. Abraham and Robert L. Rabin, ‘Automated Vehicles and Manufacturer

The Automated and Electric Vehicles Act 2018 introduces a new direct claim against the insurers of an automated vehicle (defined us one to be listed as such by the Secretary of State: s. 1). The proposal follows an official consultation in the second half of 201621 and agreement on the main contours of the reform from the UK insurance industry, a key stakeholder.22 After the first attempt to introduce legislation was foiled by the calling of a General Election in 2017 (as noted above), a revised Bill was introduced to the new Parliament and successfully negotiated the legislative process. The ensuing Automated and Electric Vehicles Act was given Royal Assent on 19 July 2018. Basic liability provisions The key provision is section 2(1), which imposes strict liability on the vehicle insurer for damage the vehicle causes in a self-driving accident: Where (a) an accident is caused by an automated vehicle when driving itself..., (b) the vehicle is insured at the time of the accident, and (c) an insured person or any other person suffers damage as a result of the accident, the insurer is liable for that damage. This may be termed the primary liability provided for in the Act. The insurer bears liability for Responsibility for Accidents: A New Legal Regime for a New Era’ (2019) 105 Va. L. Rev. 127, 143 f. 21 Department for Transport/Centre for Connected and Automatous Vehicles, ‘Pathway to Driverless Cars: Proposals to support advanced driver assistance systems and automated vehicle technologies ‘(July 2016), available at https://assets.publishing.service.gov.uk/government/uploads/ system/uploads/attachment_data/file/536365/driverless-carsproposals-for-adas-and_avts.pdf; eidem, ‘Pathway to driverless cars: Consultation on proposals to support Advanced Driver Assistance Systems and Automated Vehicles. Government Response’ (January 2017), available at https://assets.publishing. ser vice.gov.uk/government/uploads/system/uploads/ attachment_data/file/581577/pathway-to-driverless-carsconsultation-response.pdf. 22 Association of British Insurers/Thatcham Research, ‘Pathway to Driverless Cars: Proposals to support advanced driver assistance systems and automated vehicle technologies. Response of the Association of British Insurers and Thatcham Research’ (undated, late 2016).

193

ARTICLES: GUEST WRITER 2

damage suffered by any person, including the insured person, as a result of the accident, provided the vehicle is insured at the time. The vehicle must be ‘driving itself ’ in the sense that ‘it is operating in a mode in which it is not being controlled, and does not need to be monitored, by an individual’ (s. 8(1)(a)). This requirement makes clear that the liability is intended to apply to vehicles on Levels 4 and 5 of the widely recognised scale, but not those at Level 3. The strictness of the liability is apparent inasmuch as it may arise where neither the owner of the vehicle was at fault nor the vehicle defective, for example, where it is hacked by a third party.23 This primary liability is complemented by an alternative liability on the vehicle owner. if the vehicle is not insured. Section 2(2) provides: Where (a) an accident is caused by an automated vehicle when driving itself..., (b) the vehicle is not insured at the time of the accident […], and (d) a person suffers damage as a result of the accident, the owner of the vehicle is liable for that damage. This liability is also a strict liability; there is no requirement that either the owner was at fault or that the vehicle or its automated systems were in any way defective. The provision ensures that the victim of an AV accident will still have someone to sue even if the owner has failed to comply with the mandatory insurance requirement. The scheme thus created means this: if the automated vehicle is insured at the time, the insurer is liable for the damage (s. 2(1)); there is no requirement of fault or even defect. If the automated vehicle is not insured at the time, the owner is liable for the damage (s. 2(2)).

arising under existing tort law. Section 2(7) preserves the liability of other persons: The imposition by this section of liability on the insurer or vehicle owner does not affect any other person’s liability in respect of the accident. This would include persons who are ‘at fault’ (eg the negligent driver of another vehicle or a negligent cyclist or pedestrian) and the manufacturer (for product defects). Other liable persons may be sued either directly by the victim or by the insurer/the owner seeking indemnity in respect of damages they have paid to the injured party. The latter is addressed in section 5(1) of the Act, which preserves the insurer/the owner’s right to seek an indemnity from other responsible persons: Where (a) s 2 imposes on an insurer, or the owner of a vehicle, liability to a person who has suffered damage as a result of an accident (‘the injured party’), and (b) the amount of the insurer’s or vehicle owner’s liability to the injured party in respect of the accident ... is settled, any other person liable to the injured party in respect of the accident is under the same liability to the insurer or vehicle owner. The insurer/owner must pay the injured party any excess in the amount recovered from another liable person over the damages paid (s. 5(3)). Taken in the round, these provisions manifest an intention to protect the injured person from the financial consequences of the damage suffered, while channelling the loss to any person bearing responsibility for it under existing principles of tortious liability. Putting the risk back on the victim

These liabilities are additional to liabilities

194

23 Oliver Jeffcott and Rose Inglis, ‘Driverless cars: ethical and legal dilemmas’ [2017] Journal or Personal Injury Law 19 at 24.

There are a number of mechanisms in the Act that limit or exclude the liabilities for which it provides, thereby putting the risk of an

ARTICLES: GUEST WRITER 2

AV accident back on the person who suffers damage in it. First, a defence of contributory negligence applies (s. 3(1)), which provides that ‘the amount of the liability is subject to whatever reduction under the Law Reform (Contributory Negligence) Act 1945 would apply to a claim in respect of the accident brought by the injured party against a person other than the insurer or vehicle owner.’ We are asked to imagine then the application of a reduction in the damages by the same reduction as would have been appropriate if it had been possible to bring the claim against an at-fault driver or an on-defect producer. This presents a difficult logic problem, as the appropriate reduction would depend on (amongst other things) the degree of the at-fault driver’s negligence – and possibly the degree of the product defectiveness – which cannot be specified if we are talking in the same abstract terms as the statute. This will no doubt prove an interpretative difficulty in the application of the Act. More stringently, there is a 100% exemption from liability under the Act where the claim is brought by the person in charge of the vehicle at the relevant time and ‘the accident… was wholly due to the person’s negligence in allowing the vehicle to begin driving itself when it was not appropriate to do so’ (s. 3(2)). It is not immediately clear why there should be a total defence in such circumstances, rather than the proportional reduction in damages appropriate in cases of contributory negligence generally. It must be said that the statutory provision seems rather harsh. Additionally, it may be noted that the insurer may exclude or limit liability to an insured for prohibited software alterations or the failure to install critical updates (s. 4(1)): An insurance policy in respect of an automated vehicle may exclude or limit the insurer’s liability under s 2(1) for damage suffered by an insured person arising from an accident occurring as a

direct result of (a) software alterations ... that are prohibited under the policy, or (b) a failure to install safety-critical software updates... This is subject to conditions in the case of an insured person who is not the policy holder. In such case, the exclusion or limitation of liability in relation to software alterations is effective only to alterations which, at the time of the accident, the person knows are prohibited under the policy (s. 4(2)). The Act makes no provision for any policy exclusion or limitation of liability to other persons, who can thus enforce the liability against the insurer even if the accident is caused by prohibited software or the failure to install critical updates. The liability cannot otherwise be limited or excluded by a term of the insurance policy or in any other way (s. 2(6)). Lastly, there are limits on the damages recoverable under the Act. The Act includes restrictions on the property damage that is included within its definition of ‘damage’ (s. 1(3)), excluding damage to (a) the AV itself, (b) goods carried for hire or reward in or on the vehicle or a trailer drawn by it, and (c) property in the custody, or under the control, of the insured person (where the insurer is liable) or the person in charge of the automated vehicle (where the owner is liable). The quantum of damages payable in respect of property damage in any one is also limited to the limit of compulsory insurance for property damage (as specified in s. 145(4)(b) of the Road Traffic Act 1988). An obvious concern is that these various limitations on the liabilities created by the Act could result in unfairness in individual cases. Their application will have to be carefully monitored and, if experience proves it necessary, modifications will have to be introduced to avoid disproportionate consequences for victims.

195

ARTICLES: GUEST WRITER 2

Closing observations At the conclusion of the Asimov short story referenced at the start of this piece, ‘Sally’ and the other cars combine to see off the nefarious Mr Gellhorn, who was trying to steal their motors and had already used violence against their keeper, Jake. Gellhorn’s body is discovered shortly afterwards, a distance away and covered with tire tracks. Jake begins to wonder whether the cars can be trusted not to turn against him, notwithstanding his good intentions towards them. He finds he is even beginning to avoid Sally. The concerns that actually surround the introduction of AVs to our roads are somewhat less sinister. The benefits of AVs are widely trumpeted, especially in reducing accident rates, improving mobility and increasing leisure time. But the story of their development has not been one of unalloyed success and tragic accidents have occurred. To ensure the public is receptive to their presence, it is important that they have reassurance that they will be compensated in the event they suffer damage in an AV accident. Existing liabilities will not suffice as there maybe no ‘driver’ to sue for negligence, while proving a ‘defect’ in the vehicle may be too onerous.

it is certainly noteworthy in the UK, where strict liability is treated as anomalous. It comes just a few years after the abolition in 2013 of strict liability for workplace injuries,26 and without any attempt to identify a principled distinction between the two contexts. Further, though the notion of a ‘risk community’ of road-users, all of whom pose risks to each other, while at the same time benefitting individually from their ability to do so, may provide a justification for a general regime of strict liability for road traffic accidents,27 it is hard to see a principled reason why a strict liability should be introduced that applies only to self-driving – and not to conventionally driven – motor vehicles. As AVs begin to be seen on our roads, it may be questioned whether the public will be satisfied with the lack of a similar strict liability – and thus comparable rights to compensation – for road traffic accidents caused by conventionally driven vehicles. The result may well be political demands for a uniform regime of strict liability applicable to all road traffic accidents.

The 2018 Act thus creates a regime of strict liability for damage caused by self-driving AVs. To that extent it shifts the loss from the victim and places it on the insurer or (in the absence of insurance) the owner. The liable party has the ability to seek recourse from at-fault persons or an on-defect producer, which preserves the deterrent effect of tort law – such as it is.24 Perhaps the most notable thing about liability under the Act is that it is a strict liability. This might not seem that remarkable to observers in Continental Europe, where strict liability for road accidents in quite widely recognised,25 but

196

24 As to the deterrent effect of tort law, see Don Dewees, David Duff and Michael Trebilcock, Exploring the Domain of Accident Law: Taking the Facts Seriously (OUP, 1996). 25 See Wolfgang Ernst (ed), The Development of Traffic

Liability (Cambridge University Press, 2010; Ernst Karner, ‘A Comparative Analysis of Traffic Accident Systems’ (2018) 53 Wake Forest L. Rev. 365. 26 Enterprise and Regulatory Reform Act 2013, s. 69. 27 See Koziol, Basic Questions of Tort Law from a Germanic Perspective, supra note 16 at paras 6/179–6/181.

HERE TO GIVE YOU HOME ADVANTAGE

We are as ambitious as you are At Rajah & Tann, we're never satisfied with what we've achieved and we are not afraid to aim high. Currently, we have 12 offices across 10 countries in South East Asia and in China. But, ultimately, our vision for the firm extends even further afield. Rajah & Tann has represented clients in major cities around the globe; and for further proof of our ability to compete at the highest level internationally, we have been named as the Most Innovative Law Firm in the Business of Law at the Financial Times (FT) Asia Pacific Innovative Lawyers Awards. Quite simply, we believe there is no limit to what we can achieve. And exactly the same applies to the individuals we look for.

CAMBODIA

CHINA

INDONESIA

LAOS

MALAYSIA

MYANMAR

PHILIPPINES |

recruit.rajahtann.com www.rajahtannasia.com

SINGAPORE

THAILAND | VIETNAM

CASE NOTES

Noor Azlin Bte Abdul Raman v Changi General Hospital Pte Ltd and Others [2019] SGCA 13 Suchetra Sivakumar, Queen Mary University of London INTRODUCTION Noor Azlin Bte Abdul Rahman1 successfully sued Changi General Hospital Pte Ltd (‘CGH’), the first Respondent, for systemic negligence, a ground that has proven difficult to succeed at. Noor Azlin Binte Abdul Rahman, the Appellant, alleged the delay in detecting her cancer was caused by the negligence of three doctors. Dr Imran, the second respondent, a specialist respiratory physician in the hospital’s specialist outpatient clinic (‘SOC’). Dr Yap and Dr Soh, the third and fourth Respondents respectively, were from the Accident and Emergency (‘A&E’) department.2 There were two levels of claims in this case. The first is against CGH, by way of primary liability for implementing an inadequate system. The second is against Dr Yap, Dr Soh and Dr Iman for being negligent during their individual consultations.3 1. FACTS Since 2007, the Appellant visited CGH for various medical conditions but was only clinically diagnosed with lung cancer in 2012. Three visits formed the core of her contention in medical negligence. These events are set out chronologically below.

198

open date for follow-up and advised for her to return if unwell.4 On 29th April 2010, the Appellant went to the A&E department complaining of right lower chest pain. She denied having any respiratory symptoms. Dr Yap, a locum medical officer, ordered an ECG (April 2010 X-ray) revealing an opacity. At that time, no radiological reports had been made on the 2007 X-rays. Comparing the 2007 X-rays with the April 2010 X-rays, Dr Yap saw that the opacity has been present since 2007 and noted that it appeared stable. The Appellant also informed Dr Yap of Dr Imran’s diagnosis. These observations led Dr Yap to take the view that the presenting symptoms of the Appellant on this visit were not related to the opacity. The opacity was therefore deemed an ‘incidental finding’. This was partly because her lower chest pain had only begun an hour prior to the Appellant’s examination. It is important to note that Dr Yap did not have Dr Imran’s notes and relied solely on the Appellant’s account of what Dr Imran had told her.5

On 15th November 2007, the appellant visited Dr Imran for a specialist appointment. Referring to a previous X-ray and the opacity that it had identified, Dr Imran ordered a repeat chest X-ray in two views. Dr Imran concluded that the opacity appeared to be resolving or had resolved on its own. He gave the Appellant an

On 31st July 2011, the Appellant visited the A&E complaining of a left lower ribcage pain which had persisted for a month. Dr Soh, a medical officer, attended to her. He ordered two chest x-rays which focused on the left chest. Upon seeing that that the ECG tracing came back as normal, Dr Soh told the Appellant to monitor her symptoms and consult a General Practitioner or return should her symptoms worsen or persist. He did not notice the opacity at that time.6 The X-rays were sent for reporting and both times, the radiologist recommended follow up for the opacity. However, none

1 Noor Azlin Bte Abdul Rahman v Changi General Hospital Pte Ltd [2019] SGCA 13. 2 ibid [2] - [4]. 3 ibid [6].

4 5 6

ibid [12] - [15]. ibid [16] – [20]. ibid [21] – [24].

CASE NOTES

was carried out.7 On 28th November 2011, the Appellant visited Raffles Medical Centre complaining of cough, breathlessness and blood in her sputum. On 16th February 2012, the biopsy confirmed that the nodule was malignant and the Appellant was subsequently diagnosed with Stage IIA non-small cell lung cancer, which eventually spread to her brain.8 2. FIRST INSTANCE The Appellant’s central argument was that the Respondents’ cumulative negligence (i.e CGH’s inadequate system and the failure to diagnose during individual consultations) caused a delay in the diagnosis and the treatment of her lung cancer. As a result, the Appellant claimed that the cancer progressed from Stage 0 to I to Stage IIA before it was detected. Had the opacity been assessed sooner, the outcome would not have been so adverse.9 The Respondents’ denied these allegations by contending that even if there had been any breach, there was no causation. This was substantiated by the fact that Appellant did not have cancer between October 2007 and July 2011. Moreover, the Appellant had received treatment for her cancer at the earliest, because the drugs had only been made available in 2013 and 2014.10 At the first instance, the Appellant’s claim was dismissed on the basis that causation was not proven.11 Even if one or more defendants had behaved negligently, it was difficult to make a causal link between the negligence and the delay in diagnosis. Hence, “the key finding … was that …. on a balance of probabilities, the Appellant was not afflicted with cancer as at July 2011, [and that] such negligence could not have caused the delay in diagnosis.”12 The judge also held that the Appellant was diagnosed at the earliest possible stage of her cancer and received the full treatment available to her at the 7 8 9 10 11 12

ibid [25]. ibid [27] – [37]. ibid [38]. ibid [40]. ibid [41]. ibid [7].

earliest time. She did not undergo any treatment that could have been avoided.13 With regards to Dr Yap and Dr Soh, no breach of duty of care was found as the opacity was only an incidental finding. Hence, it was per the reasonable standard for Dr Yap to not have followed up on it and refer the Appellant to a specialist after deeming it to be stable. He had also relied on CGH’s radio reporting to alert him on any ‘significant abnormalities’ to come to this decision. Likewise, it was reasonable for Dr Soh to have missed the opacity since the Appellant had complained of pain in her left chest.14 On the other hand, Dr Imran was found to have breached his duty of care towards the claimant for not following up on the Appellant to ensure that the opacity was resolved. It was not sufficient for him to have set an open date.15 Also, the judge found that CGH had not fulfilled its standard of care in failing to have a routine reporting system for X-rays until 2010. CGH was also found to not have been negligent for failing to route radiological reports to specific doctors who had ordered the reports. However, CGH was in breach for not sending or communicating the April 2010 and July 2011 X-ray reports to the Appellant, postconsultations with Dr Yap and Dr Soh. The breach stood regardless of both the doctors deciding that there was no need for a follow up as a patient is entitled to be informed of her condition in order to decide to either seek a second opinion or return to CGH for another consultation.16 3. ISSUES The issues that were raised on appeal are as follows:17 i. What was the requisite standard of care in “pure diagnosis” cases; 13 14 15 16 17

ibid [43]. ibid [46]. ibid [47]. ibid [48]. ibid [58].

199

CASE NOTES

ii. Whether the Respondents had breached their duty of care towards the Appellant; and iii. If there was a breach, did it cause any loss or damage to the Appellant. 4. JUDGMENT i. “Pure diagnosis” cases The present case is one of ‘pure diagnosis’ as there were no issues pertaining to treatment. The Appellant contended a different fact-based test applied for cases of ‘pure diagnosis’, not the Bolam-Bolitho test. This argument relied on the English Court of Appeal case of Penney v East Kent Health Authority18 (“Penney”) and the English High Court decision in Muller v King’s College Hospital NHS Foundation Trust19 (“Muller”).20 In Penney, Lord Woolf MR held that what was evident from the smear slides was factbased. Whether a reasonable screener is likely to have misread it and whether a reasonable screener would have deemed the test to be negative despite the fault in the machinery was a question subject to standards professional practice.21 Likewise, in Muller, the misdiagnosis was not negligent in the eye of an expert per the Bolam test, prompting the claimant to argue that the “correct approach was for the court to determine the objective facts of what the biopsy showed and then to decide for itself whether the misdiagnosis was one which a histopathologist acting with reasonable skill and care could have made”.22 Kerr J viewed such a case, where experts had a contrasting view, was not objective (“cannot be both right and wrong”), therefore deeming it to be an issue for the courts to ponder upon.23 The above ambiguities led the court, in this case, to believe that neither of the English precedents meant for the Bolam-Bolitho test to not be applied to

200

18 19 20 21 22 23

[2000] PNLR 323. [2017] 2 WLR 1595. Noor Azlin (n 1) [50]. ibid [60]. ibid [61]. ibid [62].

cases of ‘pure diagnosis’,24 concluding that the first instance judge had employed the correct approach to reach the decision.25 ii. Breach of standard of care Where Dr Yap and Dr Soh were concerned, a lower standard of care was imposed given their intense working conditions - high-volume of cases which required immediate attention.26 Making quick decisions are a necessity, achieved by a ‘targeted approach’, which deals with the ‘treatment of the patient’s presenting symptoms and the elimination of life-threatening conditions’.27 That being said, incidental findings cannot be disregarded, especially when it is grave. Depending on the incidental finding, seriousness, and the patient’s medical history, an A&E doctor should either make a referral to the appropriate department or conduct relevant tests. However, the execution of these orders is reliant on the system of the hospital and not on the particular doctor.28 Upon establishing this threshold, the court concluded that both Dr Yap and Dr Soh did not breach their duty of care. Dr Yap was found to not have ignored the opacity, upon comparing previous X-ray reports and instructing the Appellant.29 Similarly, Dr Soh was also not found to be negligent for not noticing the opacity. His priority, which was executed, was to treat the presenting symptoms through a ‘targeted approach’.30 The standard of care is returned to its normal (higher) threshold with doctors in other departments. Where Dr Imran is concerned, the first instance decision was upheld in that he was negligent in allowing an open appointment when he was unsure if the opacity had been resolved. Being a specialist, he was also found 24 25 26 27 28 29 30

ibid [63]. ibid [64]. ibid [68]. ibid [69]. ibid [75]. ibid [76] – [78]. ibid [80] – [82].

CASE NOTES

to be last in line for diagnosis and therefore higher expectations were placed in terms of diagnosing the opacity.31 However, the court took a different stance from the first instance decision where the claim against CGH for systemic negligence was concerned. The appeal against the finding that CGH had not conducted proper followups on the radiological reports was successful. On the balance of probabilities, it was proven that the system was not adhered to as there was no proper follow-up.32 In addition, it was ordered that the reports of an incidental finding should be routed to a specialist, even if it were to be made under the instructions of an A&E doctor.33 CGH would not have been able to perform this task as it did not have an efficient mechanism to consolidate patient records. If this system had been implemented, both the A&E doctors, upon reviewing the radiological reports and Dr Imran’s notes, would have realised that the opacity was not resolved/in the process of resolving. This unavailability of vital information rendered CGH negligent for not having a database of patient records across different departments.34 CGH was also found to not have had a system to record decisions of A&E doctors, especially when acting contrary to a radiologist’s recommendation. This lack of accountability was factored into the hospital’s negligence.35 iii. Causation and Loss Where Dr Imran was concerned, though he had breached his duty of care, it was found that there was no causation between the breach and the diagnosis36 CGH, on the other hand, was found to have caused the delay in diagnosis by not imposing adequate systems in place, especially when it came to consolidating patient records. But for the delay, it was held that there would 31 32 33 34 35 36

ibid [85] – [89]. ibid [95] and [97]. ibid [98]. ibid [99]. ibid [100]. ibid [104].

not have been a development of cancer from Stage 1 to Stage IIA, the growth of the nodule, and the spread of the cancer.37 This delay was deemed to have caused significant loss to the Appellant. 5. IMPACT The case encapsulates the adaptability of law to complex systems such as public hospitals operating with many tiered departments. Following this decision, an ease of flow is to be established between the diagnosis of different doctors, and disseminating information to the patient. This allows for better treatment (by relying on more accurate information) within the hospital and more informed decisions made. Crucially, whilst the latter is not monumental in deciding this case, the fact that the judgement pushes for an efficient system of communication between doctor and patient is noteworthy, as it is pre-emptive of causative issues that can arise out of current discrepancies in the system. Moreover, the decision embraces the heights that can be reached by today’s state of technology. This compels all institutions, not just hospitals, to reconsider their usage of it in light of consumer welfare, especially considering liability being imposed in light of failure to do so. While the present case is known for its successful suit against CGH, it is also significant for clarifying the law with regards the treatment of A&E doctors. This issue was pondered upon in the UK case of Mulholland v Medway NHS Foundation Trust.38 The case highlighted that the standard of care should be lowered where A&E doctors are concerned, given the time constraints that they are placed in.39 Despite this finding, it was held that an A&E doctor should not blindly rely on a specialist’s error, especially when the “error was “barn door” … which … meant something big and obvious.”40 37 38 39 40

ibid [117]. [2015] EWHC 268 (QB). ibid [101]. ibid [96].

201

CASE NOTES

The balance that is struck here is similar to that seen in the present case. This is achieved by imposing that A&E doctor should attend an ‘incidental finding’ if it is viewed to be grave, regardless of the ‘targeted approach’ that is generally employed. The prudent approach in both cases deserves merit as the law should not foster complacency even where exceptions are found. This case also illustrates how medical institutions, not just in the areas of systemic negligence can evolve, in order to provide better patient care. In this respect, the exception made in cases of emergency doctors can be reconsidered in cases where reservations were not made. A particular reference is drawn to junior doctors (per Wilsher v Essex AHA41) and their relative inexperience. In the present case, the courts placed the burden on the hospital’s system to ensure necessary follow-ups are being made, given the hectic schedule of A&E doctors. Likewise, why should that not be the case for junior doctors in order to make up for their inexperience? Why should an additional onus not be placed on hospital systems to ensure that a junior doctor treats under the supervision of a senior doctor/nurse, or takes on cases with lower risks where there is lack of supervision? The justification for such a standard of care is obvious. Higher stakes involved warrants a higher level of responsibility taken. However, in my opinion, the law at present imposes liability on an individual (being the inexperienced doctor), forcing the system to remain stagnant especially where there is potential to evolve in terms of reducing the rate of overall negligence. Sir Nicolas Browne-Wilkinson in Wilsher42 (Court of Appeal), stated that ‘… one of the chief hazards of inexperience is that one does not always know the risks which exist. In my judgement, so long as the English law rests liability on personal fault, a doctor who has properly accepted a post in a hospital in order to gain necessary experience should only be held liable for acts or omissions which a careful

202

41 42

[1988] AC 1074. [1987] QB 730, 777(F).

doctor with his qualifications and experience would not have done or omitted.’ Where such a point of view is accepted by the legal systems, it will force hospitals to think of avenues where both an inexperienced doctor is less likely to commit errors during complicated treatments/diagnosis and enable to gain experience at the same time, in order to grow into their roles. The system at present, unfortunately, cultivates a hostile environment wherein innocent parties are held liable without much of an effort from the institution itself. This landmark decision in many ways has the potential to transform the method in which healthcare is delivered in this country. This is especially crucial in light of rising numbers of elderly43 and overcrowding of the A&E in public hospitals,44 pushing hospitals into greater efficiency.

43 Janice Tai, ‘Old and home alone in Singapore’ The Straits Times (Singapore, 17 August 2015). 44 ‘Singapore public hospitals cave under pressure of overcrowding’ (Singapore Business Review, 08 August 2018) <https://sbr.com.sg/healthcare/news/singapore-publichospitals-cave-under-pressure-overcrowding-0> accessed 09 June 2019.

CASE NOTES

Prabagaran a/l Srivijayan v Public Prosecutor [2016] SGCA 67: Approach to Prosecutorial Discretion Tan Yan Shen, University of Oxford Introduction The Certificate of Substantive Assistance (“CSA”) under s 33B(2)(b) of the Misuse of Drugs Act 19731 (“MDA”) gives a judge hearing a capital drug trafficking case the discretion to substitute capital punishment for life imprisonment and caning if the accused has proved that on a balance of probabilities, his role in the offence was restricted to that of a courier and that the prosecution has certified that the accused had substantively assisted the Central Narcotics Bureau (“CNB”) in disrupting drug trafficking activities. The CSA is a legislative prescription for the exercise of judicial power to be conditional upon the exercise of executive power.2 Prabagaran a/l Srivijayan v Public Prosecutor [2016] SGCA 67 Facts In Prabagaran, the accused was convicted of importing not less than 22.24g of diamorphine into Singapore on 12 April 2012, which attracted the mandatory death penalty. The certificate of substantive assistance under s 33B(2)(b) of the MDA was not issued and the mandatory sentence of death was imposed by the High Court judge.3 No finding was made as to whether the applicant satisfied the Courier Requirement. Issues The accused in Prabagaran challenged the constitutionality of ss 33(1), 33B(2)(b) and 1 Misuse of Drugs Act 1973. 2 Prabagaran a/l Srivijayan v Public Prosecutor [2016] SGCA 67 (“Prabagaran”) at [2] 3 Public Prosecutor v Prabagaran a/l Srivijayan [2014] SGHC 222 at [16]

s 33B(4) of the MDA on the basis that these sections are in breach of the constitutional principle of separation of powers and that these sections are not “law” capable of depriving the applicants’ lives and liberty under Art 9(1) of the Constitution of the Republic of Singapore (1985 Rev Ed, 1999 Reprint). This article will focus on the court’s interpretation of s 33B(2) (b) and s 33B(4). Judgment The court in Prabaragan held that s 33B(2)(b) and s 33B(4) MDA both prescribe a “subjective assessment of an objective condition”.4 This assessment is conducted not on the basis of whether an offender had cooperated in good faith with the CNB, but whether the offender’s assistance had yielded actual results in disrupting drug trafficking activities.5 This involves a “multi-faceted inquiry” taking into account a “multitude of extra-legal factors such as the upstream and downstream effects of any information provided, the operational value of any information provided to existing intelligence, and the veracity of any information provided when counterchecked against other intelligence sources”.6 The court further reasoned that considerable deference should be accorded to the Attorney-General as the operational effectiveness of the CNB will be otherwise hampered.7 4 Prabaragan at [65] 5 Muhammad Ridzuan bin Mohd Ali v Attorney-General [2015] SGCA 53 at [45] (“Ridzuan at CA”). These factors include the “upstream and downstream effects of any information provided, the operational value of any information provided to existing intelligence, and the veracity of any information provided when counterchecked against other intelligence sources”. 6 Muhammad Ridzuan bin Mohd Ali v Attorney-General [2014] SGHC 179 (“Ridzuan at HC”) at [50]. 7 If the prosecution had to justify every certification decision in court every time allegations are made by accused persons, CNB’s modus operandi may end up in the public

203

CASE NOTES

The court in Prabaragan held that s 33B(2)(b) and s 33B(4) MDA both prescribe a “subjective assessment of an objective condition”.8 This assessment is conducted not on the basis of whether an offender had cooperated in good faith with the CNB, but whether the offender’s assistance had yielded actual results in disrupting drug trafficking activities.9 This involves a “multi-faceted inquiry” taking into account a “multitude of extra-legal factors such as the upstream and downstream effects of any information provided, the operational value of any information provided to existing intelligence, and the veracity of any information provided when counterchecked against other intelligence sources”.10 The court further reasoned that considerable deference should be accorded to the Attorney-General as the operational effectiveness of the CNB will be otherwise hampered.11 The court in Prabagaran also found that the requirement of “substantive assistance” is not contrary to the rule of law as this standard was neither unstable nor uncertain.12 The court in Prabagaran also found that the requirement of “substantive assistance” is not contrary to the rule of law as this standard was neither unstable nor uncertain. They also found that even though “substantive assistance” may depend on the “operational necessities of the CNB”,13 these

204

domain. This would have severe detrimental effects on CNB’s enforcement capabilities and consequently on the broader public interest: at [39] of Ridzuan at CA. 8 Prabaragan at [65]. 9 Muhammad Ridzuan bin Mohd Ali v Attorney-General [2015] SGCA 53 at [45] (“Ridzuan at CA”). These factors include the “upstream and downstream effects of any information provided, the operational value of any information provided to existing intelligence, and the veracity of any information provided when counterchecked against other intelligence sources”. 10 Muhammad Ridzuan bin Mohd Ali v Attorney-General [2014] SGHC 179 (“Ridzuan at HC”) at [50]. 11 If the prosecution had to justify every certification decision in court every time allegations are made by accused persons, CNB’s modus operandi may end up in the public domain. This would have severe detrimental effects on CNB’s enforcement capabilities and consequently on the broader public interest: at [39] of Ridzuan at CA. 12 Prabaragan at [97]. 13 Prabaragan “and other factors which an offender would not know of and may not be within his control” at [96]

were not valid considerations under the “multifaceted inquiry” and was “purely speculative”.14 R (on the application of Purdy) v Director of Public Prosecutions [2009] UKHL 45 In R (on the application of Purdy) v Director of Public Prosecutions [2009] UKHL 45 (“Purdy”), the appellant wished to undergo assisted suicide aided by her husband. Under s 2(1) of the Suicide Act 1961 (“s 2(1)”), a person is liable to be prosecuted at the discretion of the Director of Public Prosecutions (“DPP”) if he assists another to travel to a place where assisted suicide is lawful.15 The Appellant sought guidance and information from the DPP on the factors taken into account when deciding if a prosecution is in the public interest but the DPP declined to do so. The court held that the law was neither sufficiently accessible to the affected individual nor sufficiently precise for him to regulate his conduct without breaking the law.16 The DPP was thus ordered to promulgate an offencespecific policy identifying the facts and circumstances which will be taken into account in deciding whether to consent to a prosecution under s 2(1).17 Analysis of the Prabagaran and Purdy approaches In Purdy, the court ordered the DPP to promulgate a set of guidelines to identify the facts and circumstances where prosecutions will be carried out despite the actual decision

14 Prabaragan at [97]. 15 s 2(4) Suicide Act 1961. 16 The court noted that “law” is to be understood substantively and not formalistically, which implies qualitative requirements such as accessibility and foreseeability. Accessibility means that an individual must know from the wording of the provision and with the assistance of the court’s interpretation what acts and omissions will make him criminally liable whereas foreseeability means that the individual is able to foresee the consequences which an action may entail: at [41] of Purdy. 17 For example, where it has been resorted to in bad faith or in a way that is not proportionate: at [40] of Purdy.

CASE NOTES

being highly fact-sensitive and controversial.18 On the other hand, in Prabagaran, the court stated that there is no need for the AG to lay out the circumstances in which “substantive assistance” may be provided as it involves a “multi-faceted inquiry” taking into account a “multitude of extra-legal factors”19 and release of the guidelines may hamper the operational effectiveness of the CNB. There is thus a rift in the approaches of both courts even though the cases share several similarities.20 There is thus a rift in the approaches of both courts even though the cases share several similarities. Admittedly, it is arguable that there are a few differences between the cases as in Purdy and in Prabaragan. In the United Kingdom, the DPP’s actions are governed by the Human Rights Act21 and the European Convention of Human Rights (“ECHR”).22 In Purdy, Lord Hope recognised a principle of legality derived from the ECHR23 which requires, inter alia, the dual elements of accessibility and foreseeability24. In Singapore, there is no equivalent of the ECHR that requires that rules need to be accessible and foreseeable. However, in Public Prosecutor v Hue An Li [2014] 4 SLR 661, when deciding on the application of the doctrine of prospective ruling in Singapore, Menon CJ noted that people should be able to conduct their affairs on the basis of what they understand the law 18 Purdy at [55] 19 Ridzuan at HC at [50]. 20 Ridzuan at CA.at [39]. 21 S 6 of the Human Rights Act 1998: it is unlawful for a public authority to act in a way which is incompatible with a Convention right. 22 Art 8(2) of the ECHR: public authority (of which the Director was part of) may not interfere with the exercise of right to respect for an individual’s private and family life except such as is in accordance with the law. 23 Purdy at [40]. 24 Principle of legality from the ECHR: whether there is a legal basis in domestic law for the restriction, whether the law or rule in question is sufficiently accessible to the individual who is affected by the restriction, and sufficiently precise to enable him to understand its scope and foresee the consequences of his actions so that he can regulate his conduct without breaking the law and whether, assuming that these two requirements are satisfied, it is nevertheless open to the criticism that it is being applied in a way that is arbitrary because, for example, it has been resorted to in bad faith or in a way that is not proportionate.

to be25 and upheld the formalist conception of the rule of law as propounded by theorists such as Joseph Raz and Fuller. This means that in Singapore, it is recognised that the law should, as far as possible, be foreseeable and accessible. It may also be argued that the nature of the provisions in question in both cases are different. S 33B was a “subjective assessment of an objective condition” whereas in s 2(1), there was no objective element to the prosecutor’s discretion. To this end, the court in Prabaragan opined that the objective condition is neither “unclear” nor “unstable”. However, even if it were accepted that the objective condition itself may be clear, the subjective assessment is by nature unclear. It is entirely possible for an authority to employ their discretionary subjective assessment to give disproportionate weight to a particular set of facts and ignore other sets of facts in order to justify an outcome that may have been derived through other means. For instance, the prosecution could have decided against awarding the certificate on the basis of a dice roll and publicly justify that outcome by using his “subjective assessment” of the facts at hand to reach an objective conclusion that there was no substantive assistance rendered. Even though the prosecutor’s decision is reviewable if the accused proves a prima facie case of reasonable suspicion of breach of the relevant standard, it has already been recognised that it would be “virtually impossible” to adduce evidence to discharge the evidentiary burden in some cases especially if there are no co-accused.26 S 33B is thus unclear and unstable without knowing what is taken into account in subjective assessment. This goes against the rule of law as the accused is unable to decide whether to cooperate and therefore admit to trafficking or importing drugs, thereby precluding him from running a defence that he was “no more than a courier” and putting the accused “in a bind”.27 25 works of 26 27

Hue An Li at [109]. Menon CJ explicitly cited the Hayek, Raz and Fuller in his judgment. Ridzuan at CA at [40] – [43] [I]n order to make the claim that he was no more

205

CASE NOTES

The court in Prabagaran also opined that the CNB’s job will be significantly hampered if the prosecution had to justify every certification decision in court every time allegations are made by accused persons. However, it is submitted that if guidelines are narrow enough to define what “substantive” means, cases challenging the prosecutor’s decision will be few and far between. Even if every of the prosecution’s certification decision is questioned, this approach is consistent with current treatment of death penalty cases. Under s 313 of the Penal Code,28 persons sentenced to the death penalty by the trial judge has a right of appeal to the Court of Appeal. This highlights the emphasis placed on ensuring that there are no grounds of doubt that the person suffering the death penalty is truly deserving of the punishment. In the same vein, there should be no reason why the prosecution should not be made to justify their decision given the wide discretion they currently have and the disproportionate impact the certification decision bears on the accused. It was held in Purdy that just because it is primarily for the Director to decide whether it is appropriate or necessary to issue guidance “cannot mean that the courts have no role to play”.

facie case of reasonable suspicion of breach of the relevant standard. It is submitted that the prosecution should promulgate a policy that outlines the factors taken for the subjective assessment of the objective condition, and this policy should count towards the “relevant standard”.29 This would still ensure deference towards the prosecution while still being able to hold them to account. Above all, this will greatly support the rule of law as recognised in Singapore.

Impact on prosecutorial discretion in Singapore Ultimately, it is unlikely for the prosecution to be completely bound by a set of guidelines if Parliament had intended for the prosecution to have discretion over certification. However, the prosecutor’s decision still can be challenged on the basis of bad faith or malice under s 33B(4) of the MDA. Currently, the prosecutor’s discretion to refuse to award the certificate may be challenged if the accused proves a prima

206

than a “courier”, the accused must first admit that he was trafficking or importing or exporting drugs. Choosing not to admit this might subsequently preclude him from arguing that he was no more than a “courier” should the court convict him nonetheless: Public Prosecutor v Chum Tat Suan [2013] SGHC 221 at [6]. See also Singapore ‘s New Discretionary Death Penalty for Drug Couriers: Public Prosecutor v Chum Tat Suan by Chen Siyuan. 28 Penal Code Cap 224.

Ridzuan at CA at [51]-[52]

CASE NOTES

B2C2 Ltd v Quoine Pte Ltd [2019] SGHC(I) 03 Evan Ng, University of Birmingham INTRODUCTION In the first cryptocurrency-related dispute that has come before the Singapore Court1 the Singapore International Commercial Court (SICC) found cryptocurrency exchange operator Quoine Pte Ltd (Quoine) liable for breach of contract and breach of trust for reversing a customer’s Bitcoin (“BTC”) / Ethereum (“ETH”) trades. Regarding knowledge of mistake, the court held that the state of mind of the programmer at the time the program was written must be considered.2 The trades in dispute were executed automatically by the deterministic software programs of the Plaintiff and the Defendant without any human intervention. The issues that arose included the recognition of cryptocurrencies as property as well as matters of segregation and trust over cryptocurrencies. Although the nature of the dispute arose from the trading of cryptocurrency, the underlying principles surround the issues were still contractual. Potential dangers of modern-day transactions involving minimal or close to zero human interaction – specifically deterministic algorithm trading – were examined to determine its ramifications on the application of the doctrine of mistake. A. FACTS The defendant, Quoine Pte Ltd (“Quoine”), operated a currency exchange platform (“The Platform”) which allowed third parties to trade cryptocurrencies as well as fiat currencies including Singapore or US Dollars. The plaintiff, B2C2 Ltd (“B2C2”) was an electronic market maker providing liquidity on exchange platforms,3 including Quoine. 1 2 3

B2C2 Ltd v Quoine Pte Ltd [2019] SGHC(I) 03. ibid [106]. ibid [1] – [2].

In the daily operation of Quoine and B2C2, complex software programs are utilised for the purpose of market making to ensure required liquidity levels are met. Quoine achieves this requirement with a software called the “Quoter Program”. This software produces new orders in the platform based on the market prices of the cryptocurrencies it retrieves from other exchanges.4 It is responsible for 98% of market making trades on The Platform.5 A technical fault prevented the Quoter program from generating new orders. Besides resulting in low liquidity, it also placed some of Quoine’s customers in a margin call position with attempts to have their position closed out through the buying of ETH with BTC at the best rate available.6 On the other hand, the software used by B2C2 in this case produces orders based off an algorithm which evaluates the first 20 market prices and other factors on The Platform before deriving an optimum price for the setting of new purchase or sell order.7 Due to the Quoter program’s failure to generate new orders, there was a lack of market data prices. This made it difficult for B2C2’s program to derive an optimum trading price. However, B2C2’s software was programmed with a “deep quote” price for reference in these situations. Yet, this set a rate approximately 250 times higher than the market rate, to which seven trades of ETH to BTC were executed. These trades were executed even though the Buyer’s account under Quoine did not possess the required 3092.517116 BTC. B2C2 was then 4 5 6 7

ibid [18]. ibid [73]. ibid [25]. ibid [83].

207

CASE NOTES

credited with 3092.517116 BTC and 309.2518 ETH was debited.8 On the following day, this matter came to the attention of Quoine’s Chief Technology Officer, which resulted in the decision to reverse them. In the following month, B2C2 commenced proceedings against Quoine.9 B. ISSUES B2C2 claimed that Quoine had no contractual entitlement to reverse the trades once they were carried out. In defence, Quoine counter-argued on the basis of express and implied terms of their contract, unjust enrichment, the holding of cryptocurrencies on trust by segregation and unilateral mistake.10 These issues were addressed in the judgment delivered by international judge, Simon Thorley QC set out below. C. JUDGMENT i. Express and Implied terms of the contract On the topic of express and implied terms, Quoine argued that it was entitled to reverse the trades based on an implied term of their contract. However, this was rejected by the SICC as implying such a term contradicts against the express term of their contract, which provided certainty on the irreversibility of filled trades.11 ii. Unjust enrichment Quoine also argued that they were entitled to carry out the reversal of trades on the basis of unjust enrichment. However, the SICC rejected that defence, and agreed with B2C2’s argument, stating that in the circumstance where enrichment is granted pursuant to a valid contract, the doctrine of unjust enrichment should not operate, irrespective of the presence

208

8 9 10 11

ibid [4]. ibid [5] – [9]. ibid [133]. ibid [152]-[154].

of an unjust factor. Additionally, the enrichment obtained by B2C2 was a result of Quoine’s failure to protect its platform and its users, especially since the Defendants did not take sufficient steps to ensure that their beliefs were correct. In other words, this was not a scenario where the enrichment received is unjustified, but rather a result of the manner, which the parties decided to trade with one another.12 iii. Unilateral Mistake Among all the issues discussed in the case, the topic of unilateral mistake was identified to be one of the most complex issues addressed in the judgment of the SICC. In a contractual setting, a unilateral mistake occurs when one party misunderstands the terms of a contract and when that misunderstanding leads to a breach of contract. For the contract to be void under the doctrine of unilateral mistake, there must be a presence of actual knowledge. However, in the scenario where there is presence of constructive knowledge, the contract will then be voidable in terms of equity. In an attempt to understand the comprehensive on the law of unilateral mistake, the SICC revisited the case of Chwee Kin Keong v Digilandmall. com Pte Ltd,13 which the defendant had incorrectly listed the price of a printer for sale at $66 instead of $3854 on its website. Following that mistake, several thousand purchase orders were made for that printer. However, the defendant later informed that those orders will would not be met due to the error in price. The courts dismissed the plaintiff ’s claim in that case, based on the reasoning that even though parties are bounded to the terms of contract which they entered into, one exception is when either party knows that the other party did not intend for the terms of the offer. 12 13

ibid [252]. [2005] 1 SLR(R) 502.

CASE NOTES

Knowledge and Unilateral Mistake in the Context of Algorithmic Trading As the trades involved algorithmic trading without human intervention, attempting to apply the law on unilateral mistake raised the issue of knowledge. The SICC held that in the situation where the actions were carried out by deterministic computer software, the issue of identifying and assessing knowledge would be to the state of mind of the programmer of the software, at the time it was written.14 On addressing the Defendant’s defence of unilateral mistake in equity, it was found that the software programmer of B2C2 had neither actual nor constructive knowledge of mistake. Therefore, the SICC held that the defence of unilateral mistake in equity has failed.15 As established, Quoine’s defences based on express or implied terms of the contract, unjust enrichment, and unilateral mistake, were rejected by the SICC and it was further held that Quoine was liable to pay damages to B2C2. D. IMPACT Even though this case involved around the trade of cryptocurrency, the dispute in the matter could be analogous to cases that involved around the trade of fiat currency. There is no need to touch on the intricacies of cryptocurrency but rather the fact that the trades were made through a deterministic software which applied algorithmic trading. The SICC judgment pointed out the risk that follows the advancement of technology. Algorithmic trading is not only applied in the trading of cryptocurrency, it also contributes to a large amount of modern-day financial trading. The advantage that such a software brings is the removal of the user’s need to monitor the market. The software is programmed to do it for the user. The downside would be that users 14 15

B2C2 Ltd (no 1) [211]. ibid [232] – [236].

take full responsibility for their utilisation of the software and its consequences should there be an instance of financial loss due to any software faults/glitches or failure. In this case, the SICC held that it was required to consider the mindset of a human actor, namely the programmer and his knowledge of the software. This was achievable as the software was created mostly by one individual. However, challenges might arise in the context where a trading software program is created by more than one programmer. The complexity of the software might bring about more difficulty in identifying which programmer is responsible for the code. It would also be a challenge to determine the knowledge at the time of programming that code. At this instance, the test applied by the SICC was only workable because the software employed was programmed only by one individual and was deterministic – carrying out trades only as they were programmed. If the software involved machine learning, it might be possible that an advanced artificial intelligence software could learn how to trade. The possibility of it developing its own trading algorithm would then result in a situation where there will be no relevant human actor’s mindset to consider. The judgment of the SICC on this case highlights the increasing risks and potential challenges which may arise with the advancement of technology. It seems to suggest that, in order to cope with these challenges in a fair manner, the law must adapt and develop at a similar pace; this might include the reviewing of existing law as well as the development of new law.

209

CASE NOTES

Easing the law onto unchartered terrain: Regency Park Villas Title Ltd and others v Diamond Resorts (Europe) Ltd and others [2018] UKSC 57

Edwin Teong Ying Keat, University of Bristol

INTRODUCTION Gray and Gray posit that the idea of property is the ‘relationship which one has with a thing’1 rather than black letter law. Context is crucial, especially when analysing the law on easements. In this respect, rigid application of the law compels individuals to purchase entire freehold or leasehold estates to enjoy access to land if his right of way is not recognised.2 In Re Ellenborough Park,3 Eversheds MR set out the following 4 requirements which constitute an easement4: (i) presence of dominant and servient tenement (ii) existing as different persons (iii) easement must accommodate the dominant tenement (iv) right claimed must be capable of forming the subject matter of a grant. These 4 requirements were introduced to prevent uncertainty in the law but should be construed flexibly and in context. This means recognising the utility and benefit when new methods of land use surface. In Regency Park Villas Title Ltd v Diamond Resorts (Europe) Ltd (hereafter known as ‘Regency’),5 Lord Briggs held that recreational easements can be recognised. This was considering the ‘service, utility and benefit’6 in the context of timeshare apartments. While this may engender allegations of introducing uncertainty, this case note hopes to advocate a

210

1 Kevin Gray and Susan Francis Gray, ‘The idea of property in land’, Land Law: Themes and Perspectives (Oxford University Press 1998, pg 15-51), at p. 15 2 Kevin Gray and Susan Francis Gray, ‘Land Law’ (Oxford University Press 7th edition) at p. 234 3 Re Ellenborough Park [1956] Ch 131 from hereon known as “Ellenborough Park”. 4 ibid. at 140. 5 Regency Park Villas Title Ltd v Diamond Resorts (Europe) Ltd [2018] 3 W.L.R. 1603 6 ibid. at [53].

more flexible approach bearing in mind the idea of property. This will be achieved by setting out the facts, issues and judgment before proffering an analysis of Regency. 1. FACTS The dominant tenement in question was Broome Park and its surrounding land which had included one Elham House while the servient tenement was the rest of Broome Park.7 In or before 1979, the park was acquired by Gulf Investments which developed a timeshare and leisure complex consisting of 18 timeshare apartments on the upper 2 floors of Broome Park and a communal club house including an 18-hole golf course, swimming pool and garden.8 Amidst multiple transactions, Elham house was transferred to Barclays Bank Trust Co Ltd to be held on trust for Regency Villas Owners Club (‘RVOC’).9 The first claimant is the freehold owner of Elham house. The remaining claimants are individual members of the RVOC. The claimants contended that they were entitled to easements over all the sporting and recreational facilities.10 On that basis, the claimants sought an injunction restraining interference with their use of the recreational facilities and damages for interference with their easement.11 On the other hand, the Defendants contended that the claimants had no benefit of easements and sought a quantum-meruit, in order to seek compensation for providing recreational facilities during and after 2012.12 7 Regency at [3]. 8 Regency at [4]. 9 Regency at [7]. 10 Regency at [16]. 11 ibid. 12 Regency at [17].

CASE NOTES

2. ISSUES The following issues were revisited in the Supreme Court judgment in order to ascertain whether an easement existed :i. Whether the grant was capable of forming an easement? ii. Whether recreational use of easement constitutes accommodation of dominant ten-ement? iii. What boundaries govern the ouster principle? iv. What constitutes more than mere passivity such that the servient tenement acknowl-edged rights conferred was capable of forming a grant? 3. JUDGMENT Whether grant constituted an easement? The Supreme Court, per Lord Briggs, held that parties intended for the grant to be an easement rather than a ‘purely personal right’13 This conclusion was informed by 3 strands of reasoning. Firstly, that the grant was undertaken for a common purpose by two associated companies.14 Secondly, they shared a common conveyancing solicitor and knew about the leasehold structure pertaining to the development of timeshare units.15 Thirdly, both parties appreciated that the transfer of interest of the grantee to a successor in title was to facilitate a sale of the lands.16 In doing so, Lord Briggs rejected the Court of Appeal judgment which had regarded the absence of express wording as indication that the use of facilities was being confined to use of existing facilities.17 Whether recreational use of easement constitutes accommodation of dominant tenement? 13 14 15 16 17

Regency at [25]. Regency at [22]. Regency at [23]. Regency at [24]. Regency at [26]-[27].

This facet was the main area of controversy.18 Lord Briggs held that recreational use can constitute accommodation of dominant tenements for the purposes of constituting an easement.19 This finding is premised on the overarching need for the law to accommodate societal change and new uses of land such as timeshare apartments which he equated, in terms of ‘service, utility and benefit’,20 to the grant of rights over a communal garden.21 In revisiting Ellenborough Park, Lord Briggs observed that Eversheds MR concluded that the test should be if the garden, to which the houses were adjacent to, constituted enjoyment in a ‘real and intelligible sense’.22 This citation was used as a premise to criticise the Court of Appeal judgment which did not consider the private nature of rights enjoyed.23 Furthermore, Hill v Tupper24 was distinguished on the basis that it was impossible to conclude an easement for a pleasure boat business annexed to a tiny portion of the canal and the grant was itself given regardless of whether it accommodated the supposed dominant tenement.25 Conversely, in Regency, there was actual use of the 18 timeshare apartments in Broome Park. However, Lord Carnwarth dissented, disputing the extension of the Ellenborough Park principle. Firstly, he counterargued that an easement by definition is a right to do something on another’s land not to ‘have something done’.26 In context, it was impossible for enjoyment of rights without active participation of the servient tenement. He substantiated this point by highlighting the fact that neither principle nor the 70 odd cited authorities accrued over 350 years across common law jurisdictions support this conception of an easement.27 Secondly, he singles out Lord Scott’s second qualification 18 Regency at [44]. 19 Regency at [53]. 20 ibid. 21 Regency at [52]. 22 Regency at [51]. 23 Regency at [52]. 24 Hill v Tupper 2 Hurl & C 121. 25 Regency at [56]. 26 Regency at [94]. 27 Regency at [96].

211

CASE NOTES

in Moncrieff v Jamieson28 that easements cannot arise from the grant of a right that requires some ‘positive action to be undertaken by the owner of the servient land’29 which includes the right to use a neighbour’s swimming pool. This qualification was formulated on the basis that easements cannot be granted to dominant tenements when active participation by the servient tenement forms an ‘intrinsic part of that right’.30 Lastly, he argues that there are ‘other and better legal procedures’ for dealing with such novel methods of enjoying land.31 What boundaries govern the ouster principle and what constitutes ‘more than mere passivity’ for a servient owner? Ellenborough Park established that rights conferred must be capable of forming the subject matter of a grant but must not be too exclusive of the servient owner.32 Given the increasing shift towards balancing the rights of both the dominant and servient tenement, Lord Briggs considered both sides of the coin. Firstly, what boundaries govern the ouster principle, i.e when does the conduct of the dominant tenement oust the servient tenement from his use of land? Secondly, what constitutes more than mere passivity such that the servient tenement acknowledged that rights conferred were capable of forming a grant. Regarding the boundaries governing the ouster principle, it was unfortunate that Lord Briggs found it unnecessary to resolve it.33 Instead, he took issue with the Court of Appeal’s approach which was premised on evaluating what the dominant tenement may do by way of ‘step-in’ rights if the servient owner failed to maintain the facilities.34 He had two reasons for this. Firstly, that the ouster principle is to be gauged in accordance with the ordinary expectations

212

28 Moncrieff v Jamieson [2007] 1 WLR 2620. 29 ibid. at [47]. 30 Regency at [98]. 31 Regency at [96]. 32 Ellenborough Park at p.6. 33 Regency at [64]. 34 ibid.

of parties. In fact, the Court of Appeal had found that the servient tenements being owners of the leisure complex were responsible for maintenance of facilities.35 Secondly, such an approach would confuse the distinction between rights to reasonable access to maintenance and rights that were capable of forming a grant. It is not axiomatic that the former gives rise to the latter.36 Regarding what is needed beyond mere passivity, Lord Briggs concluded that the ‘common understanding’ between parties pertaining to maintenance and repairs of facilities formed the ‘express basis’37 for the timeshare units that were sold to the public. This line of analysis is wholly inadequate as he had merely reiterated the principles governing mere passivity.38 In doing so, he missed an opportunity to clarify an area of uncertainty plaguing the law. This issue was disputed by Lord Carnwarth. He counterargued that there were no actual factual findings by the court to justify such a conclusion.39 4. IMPACT We discuss whether Lord Briggs’ reasoning represents a welcome or unjust extension of the Ellenborough Park principles. Insofar as the need for flexibility in recognising new forms of land use, Lord Briggs starts on the right foot.40 However, given the novelty conferred by flexibility and the law’s perennial penchant for certainty, it is axiomatic that Lord Carnwarth is unable to find authorities justifying this in the previous cases in the last 350 years. Yet, certainty in principles must prevail over certainty in categories. The dispute regarding recreational use is hinged on the fact that the law did not previously recognise it, but this must be evaluated in context vis a vis the ‘real and intelligible’ test in Ellenborough Park. As such, 35 ibid. 36 Regency at [65]. 37 Regency at [70]. 38 Regency at [66]. 39 Regency at [102]. 40 Regency at [76].

CASE NOTES

there was a real sense of the facilities being used by the dominant tenement by virtue of the fact that the Defendants were actually counterclaiming for use of the recreational facilities after 2012. Therefore, Lord Briggs’ reasoning is justifiable and should be commended. However, uncertainties remain regarding the extent of the ouster principle. One can argue that Lord Briggs in explicitly avoiding the question of extent has dodged the bullet and has left loopholes in the law. Bevan rebuts that there are 3 grounds to resisting this ‘overly hysterical reaction’.41 Firstly, recreational purposes are still governed by a contextual assessment, governed by a ‘factual matrix’.42 Secondly, it remains unlikely that recreational easements can arise from implied creation.43 Lastly, the ‘final brake’ pertains to the Law Commission’s scheme to discharge easements, extending the powers under S84, LPA 1925 to easements.44 This mirrors the effect of S105A of the Land Titles Act in Singapore which gives the Court the power to vary or extinguish easements.45 As such, Bevan’s argument conclusively lays to rest any allegations regarding the introduction of uncertainty. In this day and age with new and emerging uses of land, it will only be fatalistic to cling onto the ‘innately conservative tradition’46 of property law. In Singapore, in Muthukumaran s/o Varthan,47 it was established that the law in Re Ellenborough Park as regards the constituents of an easement is trite law,48 affirmed and applied in Fragrance Realty49 and Botanica Pte.50 While there has been 41 Chris Bevan, ‘Opening Pandora’s Box? Recreation pure and simple: easements in the Supreme Court Regency Villas Title Ltd v Diamond Resorts Europe Ltd’, Conv 2019,1 55-70 at p. 68. 42 ibid. at p.68. 43 ibid. at p. 69. 44 ibid at p.69. 45 S105A, Land Titles Act (Chapter 157). 46 Kevin Gray and Susan Francis Gray (n.2) at p. 240. 47 Muthukumaran s/o Varthan and another v Kwong Kai Chung and others [2015] SGCA 69. 48 ibid. at [40]. 49 Fragrance Realty Pte Ltd v Rangoon Investment Pte Ltd [2013] 2 SLR 1007 at [30]. 50 Botanica Pte Ltd v MCST Plan No 2040 [2012] 3 SLR

no case in Singaporean jurisdiction that aligns itself with Regency, it is humbly submitted that the Singaporean courts will likely rule in context since Re Ellenborough Park prohibited recreational easements only to avoid conferring capricious and personal benefits and promote clarity. In this regard, ‘mere recreation and amusement’51 in the absence of ‘utility and benefit’52 will not be recognised. If the issue is determining when do acts fall into the category of ‘ousting’ the servient tenement’s use of land, Lord Briggs reminds us that the ‘intention’ test is crucial. If parties clearly intended for the use of land to be in such a way, there can be no claim for ‘ousting’. This is an objective test inferred from the conduct of parties53 and should pose no problems if applied in principled fashion. Therefore, this judgment is welcome in its flexible approach. The law has not perambulated onto precarious land but in widening its scope of reasoning, eases the minds of dominant tenements while preventing vexatious claims. While one may be tempted to lapse into conservative and outmoded reasoning, this will only prove fatal. With emerging methods of land use, the increasing prominence of timeshare apartments is not unexpected.54 As such, the law must not remain mired in a traditionalist mindset but embrace change where necessary. In this respect, Lord Briggs’ perspicacious reasoning and principled judgment in Regency deserves commendation. Having said that, in Singapore, it remains to be seen whether the Courts will accept and apply the reasoning in Regency. This author humbly concludes there is no reason not to.

476 at [46(a)]. 51 Ellenborough Park at p.142 52 ibid. 53 Regency at [70]. 54 Rachel Gordon, The Guardian, 20 July 2001, ‘Timeshares and holiday clubs explained’ < https:// w w w. t h e g u a r d i a n . c o m / m o n e y / 2 0 0 1 / j u l / 2 0 / t r ave l . budgetingforyourholiday1/> Accessed 20th June 2019

213

CASE NOTES: GUEST WRITER

The future of third party (non-spousal) property interest being adjudicated in relation to divorce proceedings: BUE v TZQ Margaret Yeow Tin Tin, HOH Law Corporation INTRODUCTION This article serves to discuss the recent line of cases on issue of the ancillary powers of Family Justice Court in relation to adjudication of third party beneficial rights and interest in properties co-owned with the parties to the divorce. I. FACTS In BUE v TZQ1 (“BUE”), this involved two sons whose names were added as joint owners to a Housing Development Board property (“Property”) that was initially held in their biological father’s sole name after the former stepmother had left the Property. The Property was initially purchased by the father and the sons’ mother during his first marriage and the sons had lived in the Property all their lives. After the father’s 1st divorce, the Property was transferred to the father’s sole name. Eventually, the father remarried the former stepmother. However, the former stepmother was not included as a co-owner to the said Property even after marriage. The father and stepmother’s marriage broke down and eventually she left the Property. After her departure, the sons’ names were added as co-owners to the title of the property. As their father and their former step-mother were involved in divorce proceedings, it was determined at the Family courts that the interests of the 2 sons were to be disregarded as the former step-mother had submitted that no payment was made by the 2 brothers. The decision of the Family Courts was made in the absence of any documentary evidence of their financial contributions.

v TZR2 (“TZQ”), leave was granted for their father to admit evidence of the 2 sons’ Central Provident Fund (CPF) contributions to prove their financial contributions to the Property, amidst other evidence. For the avoidance of doubt, the name “TZQ” that appears in both cases refer to different parties. The sons had initially filed an application to the Family Courts to set aside the ancillary order made on the premise that their legal and beneficial interest had not been taken into account but this was unsuccessful. The sons had then intervened in the appeal in TZQ and applied for a stay pending an application in High Court for a determination of their interest in the HDB property (“the Intervener’s Application”). The sons were added as Interveners in respect of the Appeal. The sons had then filed an application in the High Court for a determination of their interest in the Property and sought a declaration that they would be entitled to 33.3% of the value of the Property. Both the Application and the Appeal were then heard before the Honourable Judicial Commissioner, Mr Tan Puay Boon. II. ISSUES a. Procedural Issue – Women’s Charter3 S112 At the time of the hearing, a preliminary objection was raised by the former stepmother on the necessity of the Application since the sons’ interest as interveners would be bound in the outcome of the Appeal. It was submitted on behalf of the sons that that the application should be heard first to preserve their interest in the HDB property. The father’s submissions was that a determination of the sons’ interest would predicate the amount of matrimonial assets subject to division between

Upon appeal of the divorce proceedings in TZQ 214

BUE v TZQ [2018] SGHC 276

2 3

TZQ v TZR [2019] SGHCF 3. Women’s Charter (Cap 353, 2009 Rev Ed)

CASE NOTES: GUEST WRITER

the father and their former step mother. It was further submitted that the jurisdiction of the Family Courts was limited to the division of matrimonial assets under S112, which referred to rights ascribed to the parties to the marriage. As such, the appellate court could not exercise the powers which the Family Court did not have in determining third party’s interest, as set out in UDA v UDB4 (“UDA (HC)”).

constructive trust; and 3. Alternatively, was there any advancement in favour of the sons. In its exposition of the 3 issues, the court had relied on Chan Yuen Lan v See Fong Mun9 (“Chan Yuen Lan”) in considering beneficial interest in property.10 This is set out in a chart for reference:-

The Court of Appeal in UDA v UDB5 (“UDA (CA)”) also provided clarity that the approach taken in ABX v ABY6 (“ABX”) though practical, was not intended to be within the jurisdictional ambit of S112 which provides that the division of assets referred to the parties to the marriage. CA Justice Judith Prakash held that7: “Section 112 supports the philosophy of marriage by enabling the court to treat all assets acquired during marriage as the communal assets of married partners…… It cannot be enlarged by general jurisdiction-conferring provisions in the SCJA or FJA.” In that regard, the court had adopted the approach as set out in UDA (HC) which was also endorsed on appeal in UDA (CA) to stay the appeal on the division of matrimonial assets pursuant to S112 of the Women’s Charter8 (“S112 proceedings”) pending a determination of the beneficial interest of the 2 sons in the Application. b. Substantive Issues – Resulting trust/CICT On the Application, the issues before the court were:1. Was there a resulting trust in favour of the sons and/or the former step-mother; 2. If so, whether there was common intention on the factual matrix to amount to a 4 5 6 7 8

UDA v UDB [2018] 3 SLR 1433 [28]. [31], [33] UDA v UDB [2018] 1 SLR 1015 ABX v ABY and others [2014] 2 SLR 969 Supra note 5, [27] Supra note 3

III. JUDGMENT The first step: whether there were any resulting trusts a. In respect of the sons’ CPF contributions towards the mortgage The Court in BUE held that the CPF monies paid by the sons pursuant to an agreement made to discharge the mortgage could not amount to a loan to the father when he had more than 9 10

Chan Yuen Lan v See Fong Mun [2014] 3 SLR 1048 ibid, [35]

215

CASE NOTES: GUEST WRITER

sufficient CPF funds to discharge the loan. In that regard, their CPF contributions could amount to a resulting trust of 12.2% in favour of the sons’ interest in the property.11 b. In respect of the former step mother’s contribution of $20,000, which is disputed, towards renovations of the Property The Court further held that “the long time lag between the acquisition of the Property and the alleged expenditure” could not be regarded contribution towards the acquisition of the Property.12 The fact that the alleged expenditure was not made pursuant to any agreement between the father and her when he acquired the property or when it was transferred in his sole name was indicative that there was no intention for her to have any beneficial interest in the Property. Gleaning from the facts, the court held that her disputed contributions of $20,000.00, if any, did not constitute a resulting trust in her favour. However, this issue could still be considered in respect of her indirect contributions within the ambit of S112. The second step: If so, whether there was common intention on the factual matrix to amount to a common intention constructive trust (CICT); On the evidence, the Husband therefore added the Sons as joint tenants to the Flat because he wanted to ensure that their interest in the Flat was protected and crystallised, especially so after his brush with death.

constructive trust.14 The High Court in Ng So Hang held that adding a person as a joint owner, without more, does not support a common intention to share the beneficial interest equally.15 Furthermore, it was ruled that there was a relationship of, at best, some closeness between the parties.16 In this regard, the nature of their relationship could not assist in the determination of whether there was a CICT.17 b. Right of survivorship did not equate with beneficial interest The Court in BUE relied on Ng So Hang which drew a distinction between intent behind right of survivorship and intent to grant beneficial interest when the parties were alive18:“46 Even if the Plaintiff and Defendant intended for the right of survivorship to operate upon the demise of one party in deciding to hold the Property as joint tenants, this did not mean that the parties also intended for the beneficial interest to be held in equal shares, or in any particular proportion while both parties were alive. The former discloses an intention that on the death of one party, the other would be entitled automatically to the property, rather than an intention to share the beneficial interest in the property equally while both parties are alive (emphasis is ours). The following passage from Neo Hui Ling v Ang Ah Sew [2012] 2 SLR 831, at [39], is instructive: “The intended consequences of the rule of survivorship operate after the death of one tenant, and say nothing whatsoever about what should happen while both tenants are alive. This is not to say that the rule of survivorship is incompatible with an intention that the joint tenants should also hold

a. Relationship of the parties The Court in BUE had directed the parties’ attention to Ng So Hang v Wong Sang Woo13 (“Ng So Hang”) in relation to the right of survivorship and its relationship with the common intention

216

11 Supra note 1, [47] 12 Supra note 1, [51]; Tan Chui Lian v Neo Liew Eng [2006] SGHC 203 [24] 13 Ng So Hang v Wong Sang Woo [2018] SGHC 162

14 15 16 17 18

ibid, [44] – [49] ibid, [48] ibid, [30] ibid, [54] ibid, [46] – [47]

CASE NOTES: GUEST WRITER

the property beneficially during their lifetimes. What this means is that the rule of survivorship sheds no light on the tenants’ intentions as to their beneficial interests in the property while both are alive. In addition, the right of survivorship may entitle the surviving owner to a share of the property, but what is gained is only at law; it may be that even where the right of survivorship operates to make the sole surviving owner the sole registered proprietor of the property, the survivor may not be entitled to the whole of the beneficial interest. Such would be the case for instance where a resulting trust operates in favour of the deceased’s estate, which is not displaced by a common intention constructive trust, intention to gift or presumption of advancement (see Lim Chen Yeow Kelvin v Goh Chin Peng [2008] 4 SLR(R) 783 at [116]). Where co-owners have contributed unequal amounts towards the purchase price of a property, how the beneficial interest in the property is to be apportioned upon the death of one party is still to be determined by undertaking the steps set out in Chan Yuen Lan (see above at [23]).” On the evidence, the Court in BUE construed that the father only wanted to give brothers a share in the property, and for the brothers to make financial contributions through discharge of the mortgage using CPF savings so that they could feel a sense of ownership. As he was afraid that the stepmother would evict them, he therefore made them contribute financially to protect their interest. The former stepmother’s arguments that the father had intended it as a loan to the children when he had more than sufficient CPF contributions to redeem the balance of the mortgage was therefore untenable. In essence, in reliance on Ng So Hang, the Court held that the father did intend for the sons to inherit upon his death but not for the sons to benefit equally when he was still alive. At this juncture, it would appear that the outcome of this is indeed a compelling bifurcation: the

right of survivorship grants the survivor a legal claim in the property which vests in his name after death. This is a separate process from that laid out in Chan Yuen Lan as the Chan Yuen Lan inquiry mainly focuses on the agreement at the time of acquisition of property, or a short time afterwards. The third step: Whether a gift was intended For the same reasons as set out above, the Court was of the view that the intention to gift, if at all, was only upon death and not while the father was still alive. The fourth step: Whether the Presumption of advancement applies The Court held that based on the current factual matrix, the parent–child relationship falls into the well-recognised categories of advancement from father to child. Notwithstanding that, the same reasons proffered for the 2nd step had negated any intention to advance each of them an immediate gift of a one- third share of the beneficial interest in the Property.19 An interesting analysis of the categories of relationship was also discussed in Ng So Hang as to whether cohabitees should be accorded with the same status as married couples. The Court was doubtful that such extension should be warranted.20 In comparison, as seen in TZQ, the period of cohabitation between the Husband and the former step mother was taken into account for the purposes of assessing the length of marriage as one of the factors under S112.21 The fifth step: Whether there was a subsequent change in common intention There was no suggestion that there had been a subsequent change in common intention on the part of any of the other parties. It is also 19 20 21

Supra note 1, [72] Supra note 13, [114] Supra note 2, [87]

217

CASE NOTES: GUEST WRITER

pertinent to note that since the court had held that there was no common intention between the parties, this notional step was not necessary. IV. IMPACT Joint tenancy on the title of a property does not equate with beneficial interest as the myriad of permutations as to whether a resulting trust, or common intention constructive trust or a gift or even a change of a subsequent common intention to vitiate such constructive trust. As the concepts of trusts are constantly evolving, these equitable notions that seek to perfect any trust created or dispel any hope of the gifts in the midst of the procedural hoops with third party encumbrances are the litigator’s dream of multiplicitous actions. To the sandwich class where the social needs of three-generation families are registered as joint owners in an HDB property,22 the jurisdictional difficulties of S112 may not be the most financially palatable perspective in the event of a divorce. Perhaps, any enhancement of the jurisdictional provisions akin to the provisions in S25 of the Family Justice Act23 in respect of S112 would be a greeted with a warm welcome.

218

22 As at 2017, an estimated 81% the Singapore resident population reside in HDB flat. Statistics obtained from Data. gov.sg, ‘Estimated Singapore Resident Population in HDB flats’ https://data.gov.sg/dataset/estimated-resident-populationliving-in-hdb-flats?resource_id=a7d9516f-b193-4f9b-8bbf9c85a4c9b61b” 23 Section 25 of the Family Justice Act (No. 27 of 2014) sets out “the purpose of S25 is to be a gap-filing provision for the purpose of assisting the Family Division in exercising its primary jurisdiction. It confirms that the Family Division has power to deal with civil issues when they arise in the course of matters in which the Family Division’s jurisdiction has been properly invoked.”

COMMENTARY

Ready for Repeal: Lessons from the Hart-Devlin Debate Alexander Yean*, University of Oxford Introduction

three arguments:

Section 377A of the Penal Code of Singapore (hereafter ‘377A’),1 which criminalises private sexual acts between males (hereafter ‘homosexual behaviour’ for short), was most recently thrust into the limelight of public debate in the wake of Navtej Singh Johar v Union of India,2 decided on 6 September 2018, in which the Supreme Court of India declared sections of the Indian equivalent of 377A to be unconstitutional.

i. Homosexual behaviour is an inherent or prima facie wrong. ii. Repeal will lead to a normalisation of homosexuality that threatens the cohesion, stability, and well-being of society. iii. The moral status quo, supported by the majority, is a thing of intrinsic value to be defended from change.

The scope of the debate over 377A has been wide, straddling questions that range from the constitutional to the theological. Prominent advocates of 377A’s repeal include Professor Tommy Koh, who called for a class-action lawsuit by the LGBT community,3 and V K Rajah, former Judge of Appeal and AttorneyGeneral, who considered the statute to be “anachronistic” and “misaligned”.4 Supporters of 377A include NUS Law Professor Thio Li-ann, who commended the statute’s “contemporary relevance and substantive importance”,5 and Christian and Muslim community leaders.6 Supporters of 377A have broadly relied on

220

* BA (Hons) Jurisprudence Candidate, Exeter College, University of Oxford. I am grateful to Ms. Vivian Leong for her contributions towards an early draft of this paper. Any errors and infelicities remain my own. 1 Penal Code of Singapore, s 377A. 2 Navtej Singh Johar & Ors. v. Union of India thr. Secretary Ministry of Law and Justice W. P. (Crl.) No. 76 of 2016. 3 Yasmine Yahya, ‘Tommy Koh’s Facebook Comment Fuels Heated Debate on Section 377A’ The Straits Times (Singapore, 8 September 2018). 4 V K Rajah, ‘Section 377A: An Impotent Anachronism’ The Straits Times (Singapore, 30 September 2018). 5 Thio Li-ann, ‘Section 377A: A Contemporary, Important Law’ The Straits Times (Singapore, 7 October 2018). 6 Melody Zaccheus, ‘Church Groups say They do not Support Repeal of 377A’ The Straits Times (Singapore, 13 September 2018); Toh Ting Wei, ‘Islamic Scholars, Teachers Oppose Section 377A Repeal’ (Singapore, 20 September 2018).

The first argument, which for the ease of reference will be referred to as the “orthodox thesis”, is favoured by religious groups and has been espoused for millennia, with the Emperor Justinian having believed that homosexuality was the cause of earthquakes.7 The National Council of Churches of Singapore’s suggestion that “the homosexual lifestyle is… harmful for individuals” on the basis of being a “perversion” against God and nature8 is thus far from novel, but the orthodox thesis itself remains beyond the ambit of this paper in that it is addresses a theological dialectic that holds little jurisprudential weight in a secular and religiously plural state. Rather, this paper is concerned with the purported detriment that repeal will bring to proponents of the orthodox thesis, which falls under the second and third arguments that supporters of 377A have put forth. The second and third arguments, situated at the intersection of the criminal law and societal conceptions of morality, bring to mind the Hart-Devlin exchange of the 1960s, “perhaps the most celebrated jurisprudential debate of the twentieth century”,9 which was ignited in 7 Russell Hittinger, ‘The Hart-Devlin Debate Revisited’ (1990) 35 Am. J. Juris. 47, 49. 8 Zaccheus (n 7). 9 Gregory Bassham, ‘Legislating Morality: Scoring the Hart-Devlin Debate after Fifty Years’ (2012) 25(2) Ratio Juris 117.

COMMENTARY

response to the Wolfenden Report of 1957 that called for the decriminalisation of private homosexual behaviour between consulting adults.10Lord Devlin, a Lord of Appeal in Ordinary and staunch Catholic, launched a swift attack of the Report through his Maccabaean Lecture to the British Academy in 1958, self-explanatorily entitled The Enforcement of Morals. He asserted that it was “wrong to talk of private morality or of the law not being concerned with immorality as such,”11 but in what was then a revolutionary step, supported naturalistically moral laws on “primarily harmbased grounds.”12 Within Devlin’s complex argument are two theses that are of immediate relevance to the 377A debate in Singapore, each of which corresponds to one of the arguments propounded by supporters of 377A. First, Devlin’s “disintegration thesis” asserts that society has a prima facie right to legislate against immorality because society might disintegrate, or be substantially weakened, if its common morality is lost: if immoral behaviour is lawful, it may well cause some to perceive the law as “condoning” immorality, and thus lose respect for not only morality, but also the wider structure of the law and the state.13 This idea would no doubt resonate with supporters of 377A, and indeed premises their second argument that repeal would condone homosexual behaviour as a social norm, and in so doing bring “dreadful consequences for the stability of our families, the well-being of our children, and… risks to the common good.”14 Hart and Ronald Dworkin later suggested a further facet of Devlin’s argument, which Hart labelled the “conservative thesis”: broadly, “society has the right to enforce its morality by 10 Brian Lewis, Wolfenden’s Witnesses: Homosexuality in Postwar Britain (Palgrave Macmillan 2016) 275. 11 Patrick Devlin, The Enforcement of Morals (Oxford University Press 1965) 14. 12 Bassham (n 10) 121. 13 Devlin (n 12) 11. 14 Tan Tam Mei, ‘Section 377A Should Not be Repealed under the Present Circumstances: Archbishop William Goh’ The Straits Times (Singapore, 19 September 2018).

law because the majority have the right to follow their own moral convictions that their moral environment is a thing of value to be defended from change.”15 This idea in turn corresponds to the third argument propounded by supporters of 377A; while not explicitly articulated as such, it certainly undergirds many of their assertions16 and thus merits critical examination. Hart, then Professor of Jurisprudence at the University of Oxford, was the most vocal of Devlin’s contemporary critics, and presented strident rebuttals to both his disintegration and conservative theses. In examining the points raised by Hart and other subsequent jurists, and contextualising them in the 377A debate, this paper hopes to not only demonstrate that many of the contentions raised by supporters of 377A have already been debunked half a century ago, but also advance fresh refutations of Devlin’s original arguments as they apply to 377A. Indeed, the rebuttals against the continued criminalisation of homosexual behaviour remain as relevant and as resonant as they have ever been. The Myth of Disintegration; or, Homosexual Behaviour Doesn’t Cause Earthquakes Among Devlin’s many prongs of attack, the most significant and most discussed argument must be his disintegration thesis: that society has a prima facie right to enforce its shared morality by law because society might disintegrate, or be substantially weakened, if that common morality is lost. Echoes of the disintegration thesis can still be heard, for instance, in Jamiyah’s assertion that repeal “may impair Singapore’s admirable social cohesion, growth and future progress as a nation”,17 or Pergas’ concern for the “fabric of society.”18 15 Bassham (n 10) 122. 16 See n 15. 17 Timothy Goh, ‘377A Repeal would do More Harm than Good: Jamiyah’ The Straits Times (Singapore, 25 September 2018). 18 Toh Ting Wei, ‘Pergas Rejects Repeal of Section 377A as it Can Cause ‘Worrying Implications’’ The Straits Times (Singapore, 20 September 2018). See also Zacchaeus (n 7).

221

COMMENTARY

While seemingly banal to contemporary students of jurisprudence, the disintegration thesis was genius for its time. As far back as since On Liberty in 1859, Mill had rejected the Victorian notion of the state’s role in enforcing morality, instead suggesting that the government’s only use of its coercive powers should be to prevent harm,19 in what is now well-known as the harm principle. Unlike previous naturalists who had asserted that the law must exist simply for morality’s own sake, Devlin dexterously used the liberals’ very premise against them, by suggesting that the prevention of harm was in fact the cardinal function of morality, such that the state was well-justified in using its coercive powers (among them, the criminal law) to enforce common morality, which acts as a sort of glue that keeps society from splitting asunder.20 Devlin goes further: his argument includes a strident criticism of the Millian notion of a “private morality,” suggesting that “it is wrong to talk of private morality or of the law not being concerned with immorality as such.21 For Devlin, there was no such thing as a victimless crime – the corruption of the individual necessarily led to the corruption of society, in three ways: i. The corruption of the individual would render him of less use to society – a “tangible harm.”22 ii. The corruption of the individual may well metastasise, causing a sort of “epidemic” of vice – a villager taking to drink may influence a significant portion of the village to become drunkards.23 iii. The corruption of the individual may cause intangible harm to society by weakening the moral bonds that act as society’s cement. If upright citizens perceive the law as “condoning” vice,

222

19 J S Mill, On Liberty (first published 1859, Yale University Press 2003) 80. 20 Bassham (n 10) 118. 21 Devlin (n 12) 14. 22 ibid 111. 23 ibid 143.

they may well lose respect for the law, the state, and society.24 Each of these statements will be rebutted seriatim. Devlin’s first argument is based on a truism: it essentially boils down to arguing that a person who is rendered of less use to society is of less use to society, and is a non-argument. Certainly, any attempt at using this argument to justify coercion must first establish that the corruption in question actually causes the individual to be of less use to society, but in the context of 377A, it would be absurd to suggest that homosexuals contribute less to society than anyone else. Devlin’s second argument strongly resembles the “broken window theory” in criminology, which suggests that visible signs of civil disorder create an urban environment that encourages further disorder. Despite being controversial, the theory has found some support, such as from former New York City mayor Rudy Giuliani, who based his successful policing policy around it.25 Without delving into an examination of the theory’s merits, it is sufficient to state that when applied to a jurisprudential context, it must be based upon the premise that the harm being propagated is actually a harm: one drunkard is bad; a fortiori, a village of drunkards must be worse. The argument again fails to apply to 377A: there is no prima facie reason why homosexuality is bad (any such assertion would simply beg the question); further, the notion that homosexual behaviour among some segment of the population encourages significant numbers of hitherto non-homosexual people to change their sexual orientation is ludicrous. Thus, supporters of 377A necessarily hang their hat on Devlin’s third argument, which also forms the crux of the disintegration thesis. Certainly, the notion of social collapse from moral weakness was far from novel, for 24 ibid 131. 25 Christina Sterbenz, ‘How New York City Became Safe Again’ Business Insider (New York City, 2 December 2014).

COMMENTARY

instance having been accepted by historians as one reason for the Roman Empire’s decline and fall.26 Immediately after the Second World War, belief in Christianity in Britain was overwhelming, with the public perceiving a strong association between religious belief and moral behaviour; but by the late 1950s, staunch Christians (including Devlin) were wary of increasing secularisation,27 and would have provided Devlin a ready audience. There are two rebuttals against Devlin’s third argument: one principled, and one empirical. The principled rebuttal is that in the context of the 377A debate, the disintegration argument is premised upon the fact that homosexual behaviour is a prima facie harm – there is no reason why the law condoning homosexual behaviour will lead to citizens perceiving the law to condone vice unless homosexual behaviour is itself a prima facie vice. Simply, this paper is unwilling to give merit to the notion that homosexual behaviour is a prima facie vice – suffice it to state that it is not a prima facie vice for the same reason that it does not cause earthquakes. Dworkin explains succinctly: it has “no effects which can be deemed injurious to others.28 If the premise of homosexual behaviour being a prima facie vice is not accepted, then the only way that condoning homosexual behaviour can possibly lead to disintegration is if there exists some other non-principled reason why condoning homosexual behaviour would assail society’s common morality. The empirical rebuttal suggests that for all intents and purposes, no such reason exists. First suggested by Hart, it is beautiful in its simplicity – Devlin’s disintegration argument is simply not true in the context of homosexual behaviour. Devlin offers no empirical evidence in support 26 See, for instance, Edward Gibbon, The History of the Decline and Fall of the Roman Empire (first published 1776-1789, Hamish Hamilton 1936). 27 Peter G. Forster, ‘Secularization in the English Context: Some Conceptual and Empirical Problems’ (1972) 20(2) The Sociological Review 153. 28 Ronald Dworkin, ‘Lord Devlin and the Enforcement of Morals’ (1966) 75 Yale L.J. 986, 988.

of it, no prophesised moral doom befell European countries that had decriminalised homosexual behaviour, and history shows that “society cannot only survive divergences... from its prevalent morality, but profit from them.”29 Since the status quo evidence supported the opposite view, the burden fell on Devlin to adduce evidence that legalising homosexual behaviour would lead to moral collapse, but he could not. In present-day Singapore, Archbishop William Goh similarly cited the “dire consequences” that followed when countries normalised samesex unions,30 but declined to elaborate on what those consequences were. Hart’s rebuttal is thus simple and effective, and certainly no less so half a century later. Just as the United Kingdom in 1967 did not collapse after it decriminalised homosexual behaviour, neither, it seems, would Singapore. A final argument against Devlin’s disintegration thesis returns to its underlying premise: the survival of society, which is a greater good that undergirds the harm principle and justifies coercion. Much of Devlin’s argument relies on his naturalistic conception of a fixed, unchanging moral ideal. Even if this position was accepted, it remains a truism that societal values change (either in reference to a fixed moral ideal, as naturalists suggest, or without any fixed reference point, as positivists suggest) – a simple example would be the rapid secularisation of post-war Britain. In light of changing societal values and standards, the survival of society’s moral fabric must necessarily be based upon changing and adaptable laws. Hart’s deepest criticism against Devlin’s disintegration thesis is thus that it is fundamentally irreconcilable with the survival of society under the harm principle. Legal scholar and theologian Russell Hittinger summarises: “For if survival is the proper end of law 29 H L A Hart, Law, Liberty and Morality (Stanford University Press 1963) 71. 30 Most Rev. William Goh, ‘Pastoral Letter on S377A to Catholics’ (Roman Catholic Archdiocese of Singapore, 18 September 2018).

223

COMMENTARY

and morality, and if justice represents the fair distribution of the goods of survival, then it would be foolish to freeze the law as to its preservation of material values – viz., to enforce the law independent of any changes in the conditions of survival.”31 Hart thus argues that it is not only erroneous, but dangerous, to suggest that the survival of society (based on the integrity of its common morality) must depend upon standing against changing values. Quite the opposite, it depends upon adapting to the inevitability of shifting standards of morality, regardless of whether one subscribes to the naturalistic conception of the existence of a moral ideal to strive towards. Singaporean attitudes towards homosexual behaviour are already changing: a study by the Institute of Policy Studies that surveyed over 4000 Singaporeans reports that a slim majority of 50.4% believe that homosexual behaviour is “always wrong”, down significantly from 61.6% in 2013. Notably, the share of young Singaporeans who believe that “gay sex is not wrong” has nearly tripled in the same period.32 If, and when, the majority is reversed, Hart’s wisdom suggests that repealing 377A would become a matter of necessity. Conservatism, Moral Individual Rights

Positions,

and

The second prong of Devlin’s argument, attributed to him by Hart and Dworkin, is the conservative thesis: the notion that “society has the right to enforce its morality by law because the majority have the right to follow their own moral convictions that their moral environment is a thing of value to be defended from change.”33 It is not difficult to see how the conservative thesis serves as an important consideration for supporters of 377A – for instance, Professor

224

31 Hittinger (n 8) 48. 32 Gilaine Ng, ‘55 per cent of Singapore Residents Support Section 377A: Ipsos Survey’ The Straits Times (Singapore, 10 September 2018). 33 Bassham (n 10) 122.

Thio Li-ann writes of “radical, deleterious social consequences” that would follow any “legal change” relating to 377A.34 At first glance, the conservative thesis seems quite defensible – certainly more so than the disintegration thesis. Philosopher Gregory Bassham summarises the burden of proof:35 i. Certain established institutions, ideals, traditions, social environments, and ways of life, while perhaps imperfect, represent moral positions that are valuable and worth preserving. ii. Certain harmless immoralities threaten to undermine or erode those institutions, and by extension, the moral positions they represent. iii. The costs of making and enforcing laws against such immoralities (loss of freedom, criminal justice costs, risks of blackmail, arbitrary enforcement) are outweighed by the importance of preserving those institutions and moral positions. While Hart did not directly address the conservative thesis, several contemporary liberal jurists have, including Joel Feinberg and Andrew Altman.36 The obvious weakness of the conservative thesis is that it is premised upon inherently preferencing such institutions at the expense and neglect of individual autonomy and rights. It follows that the justifiability of any policy under the conservative thesis must weigh the relative importance of three factors: i. the value of the moral positions that it seeks to preserve ii. the value of the individual’s right to enjoy the “harmless immoralities” that seek to threaten or undermine those moral positions iii. the effect of the “harmless immoralities” on the erosion of those moral positions 34 35 36

Thio (n 6). Bassham (n 10) 122-3. ibid 123.

COMMENTARY

In the context of 377A, justifying the continued lawfulness of 377A would require a costbenefit analysis of intangibles: of one moral position against another. It is beyond the ambit of this paper to consider whether pro-377A moral positions like Catholicism, or general social conservatism, are inherently “valuable and worth preserving”, or indeed are any more so than pro-repeal moral positions like libertarianism. However, the conservative thesis does not require the triumph of one moral position against another to be refuted, or at least significantly weakened, in the context of 377A. Rather, two responses suffice. The first response is that for a moral position to be inherently valuable and worth preserving, it must have some basis that consists of more than just prejudicial, arbitrary, or erroneous reasons. Dworkin gives four examples of reasons that cannot make out defensible moral positions:37 i. Prejudicial reasons (e.g. the belief that homosexuals are not “real men”) ii. Reasons based on emotions or arbitrary preferences (e.g. the idea that homosexuals “make me sick”) iii. Erroneous propositions of fact (e.g. the notion that homosexual behaviour is physically debilitating) iv. Reasons that beg the question (e.g. “everyone knows homosexuality is wrong”) – Dworkin notes that “with the possible (though complex) exception of a deity, there is no moral authority to which I can appeal [that] automatically makes my position a moral one.”38 To illustrate the importance of moral positions, Dworkin raises the example of someone who votes against a candidate for the sole reason that he is homosexual, since the voter believes homosexuality to be profoundly immoral. If the voter’s belief in the immorality of homosexuality is based on a moral position, the voter might be perceived as “eccentric, 37 38

Dworkin (n 29) 995. ibid 996.

puritanical, or unsophisticated,” but “these are types of character and not faults of character.” In such an instance, the irresistible conclusion is that the voter has a “moral right to vote against the homosexual” due to the right (and indeed, the duty) to vote his convictions. Such a moral right or duty would not be admitted, however, if the voter has no moral position on homosexuality, but is rather voting merely for prejudicial, arbitrary, or erroneous reasons.39 In the context of Singapore, it is unclear what proportion of the slim majority of Singaporeans who feel that homosexual behaviour is “always wrong”40 believe so from the basis of genuine and defensible moral positions, and what proportion believe so for prejudicial, arbitrary, or plainly erroneous reasons. Suffice it to say that the latter group’s beliefs are not valuable and not worth conserving. The second response to the conservative thesis in the context of 377A is that it is offensive to individual rights. The inherent value of pro-377A moral positions, and certainly the consequential value of any institutions that reinforce such moral positions, must stem at least partially from their benefit to individuals – for instance, the benefit that staunch Catholics receive not only from their belief in the Catholic faith, but also from an environment that supports and accords with such belief. Rather than attempting to defend the inherent value of one moral position over another, this paper would instead return to Mill, and suggest that in a clash between the preservation or furtherance of an individual right and the preservation or furtherance of a moral position, the winner must be decided by the relative harm or benefit that society will accrue on balance.41 It is clear that the continued lawfulness of 377A, even if unenforced, causes harm to the LGBT community: they are regarded as “less 39 40 41

ibid 995. See n 33. See n 20.

225

COMMENTARY

than full members of society, and often [feel] that they must hide who they are because of their gender identity or sexual orientation.”42 In comparison, the harm of repeal for supporters of 377A, if any, is non-injurious.43 Therefore, under the harm principle, the burden of proof must fall on 377A supporters to demonstrate (i) the value of the moral positions and their supporting “institutions, ideals, traditions, social environments, and ways of life” that they seek to preserve, and (ii) that the effect of repeal would be so deleterious upon these moral positions and institutions, ideals, etc. as to justify the continued repression of individual rights. Certainly, this burden has not been satisfactorily met by either the orthodox thesis or the disintegration thesis discussed earlier. The conservative thesis can thus be reduced to a weighing of rights; examined in that light, and if the liberal emphasis on the importance of individual rights is accepted, then the burden of proof becomes quite onerous indeed. On balance, this point does not seem to fall in favour of the supporters of 377A, all the more so given Singapore’s pride in its purported pluralism. The final, but perhaps most important, thing to note about the conservative thesis is that in practice, moral positions only translate to legislative outcomes inasmuch as they are held by the majority of society. Dworkin gives two reasons for this:44

faith of [a majority of] its members.” Empirically, the proportion of Singaporeans who support 377A is decreasing.45 If, and likely when, advocates of repeal become the majority, the conservative thesis would not only be moot, but even support repeal. Conclusion In 1967, 10 years after the release of the Wolfenden Report and at the conclusion of the Hart-Devlin debate, the Sexual Offences Act 1967 was passed that decriminalised homosexual activity in England and Wales. In the 50 years since, the consensus among jurists is that Hart has emerged the victor:46 having been proven right, if nothing else, by history. In half a century that saw the decline of organised religion in the West and vast attitudinal changes towards sexuality, British society appears no closer to disintegration or moral bankruptcy than it was in Devlin’s day – if anything, what stands out most clearly is the “resilience of social cohesion and moral commitment.”47 Hart’s lessons for Singapore are simple: first, the argument that the normalisation (or indeed, the mere decriminalisation) of homosexual behaviour will lead to societal disintegration is not only divisive, but erroneous. Second, the moral environment that perpetuates such a notion should, if anything, change rather than resist change. We are quite ready for repeal.

i. “In a democracy this sort of issue, above all others, must be settled in accordance with democratic principles.” ii. “It is, after all, the community which acts when the threats and sanctions of the criminal law are brought to bear. The community must take the moral responsibility, and it must therefore act on its own lights – that is, on the moral

226

42 Repeal of 43 44

CNA, ‘Human Rights Group MARUAH Calls for 377A’ CNA (Singapore, 24 September 2018). Dworkin (n 29) 988. ibid 993.

45 46 47

See n 33. Mark Murphy, Philosophy of Law (Blackwell 2007) 103 Bassham (n 10) 131.

COMMENTARY

#MenToo: a response to the Criminal Law Reform Bill 2019 Matthew Choo, King’s College London Introduction The Criminal Law Reform Bill was introduced in the Singapore Parliament on 11 February 2019, as part of the Government’s continued efforts to ensure the relevancy of the criminal law with the changing dynamics and demands of society.1 The Bill, if passed, will make sweeping changes to the criminal law, in particular to the definition of rape and the ambit of sexual assault by penetration. While the reforms are welcome and commendable, this article will argue that the reforms do not go far enough. In particular, “compelled-penetration” should be criminalised as rape and not as sexual assault. This article will first analyse the proposed reforms to the law, the rationale and implications. It will then argue for compelled-penetration to be criminalised as rape, owing to the equivalence in harm by examining misconceptions surrounding compelled-penetration. The Proposed Reforms In July 2016, the Penal Code Review Committee (PCRC) was formed to ‘conduct a thorough and comprehensive review of the Penal Code and make recommendations on reforming the Penal Code’.2 The PCRC concluded their review and submitted their report to the Ministry of Law on 31 August 2018. The report was subsequently published on 9 September 2018. Following a public consultation,3 the 1 Ministry of Home Affairs, Press Release: First Reading of Criminal Law Reform Bill and the Government’s Response to Feedback on it (Ministry of Home Affairs, 2019) <https://www.mha.gov. sg/newsroom/press-release/news/first-reading-of-criminallaw-reform-bill-and-the-government-respo nse-to-feedbackon-it> accessed 10 March 2019. 2 Penal Code Review Committee, Penal Code Review Committee Report, (Ministry of Home Affairs, August 2018) <https://www.mha.gov.sg/docs/default-source/defaultdocument-library/penal-code-review-committee-report3d9709 ea6f13421b92d3ef8af69a4ad0.pdf> accessed 10 March 2019. 3 Ministry of Home Affairs, Press Release: Public

Government introduced the Bill in Parliament on 11 February 2019. The Bill proposed an expansion of the definition of rape, with the introduction of a new subsection 1A to section 375 of the Penal Code,4 which provides:-5 Any man (A) who penetrates, with A’s penis, the anus or mouth of another person (B) (a) without B’s consent; or (b) with or without B’s consent, when B is below 14 years of age, shall be guilty of an offence. In recommending an expansion of the definition of “rape” to include penile-anal penetration, the PCRC was of the opinion that the expansion would provide an ‘appropriate label’ to such acts, given the dangers of sexually transmitted diseases,6 and bring the law in harmony with other jurisdictions.7 The Bill however, went further than the PCRC’s recommendations. In their report, the PCRC acknowledged that other jurisdictions have defined rape as including both penile penetration of the mouth and anus, and noted Consultation on Proposed Amendments to the Penal Code (Ministry of Home Affairs, 2018)<https://www.mha.gov.sg/newsroom/ press-release/news/public-consultation-on-proposedamendments-to-the-penal-code> accessed 10 March 2019. While the responses of the public consultation remain unreported, it was reported that more than 700 stakeholders from the legal, social, religious, financial and education sectors participated in engagement sessions and the Ministry received 60 written feedbacks from individuals and organisations. 4 Rev. Ed. Cap 224. 5 Criminal Law Reform Bill, Bill No. 6/2019, clause 110(a). 6 PCRC (n 2) at section 26.20. 7 ibid. at 26.21. For example, Section 1(1)(a) of the Sexual Offences Act 2003 defines rape as the intentional penetration of the vagina, anus or mouth of another with a penis in England and Wales.

227

COMMENTARY

that the High Court has found forced penileoral penetration ‘more disgusting’ than forced penile-anal penetration.8 Despite this, the PCRC was not ready to recommend the expansion of rape to include both oral and anal penile penetration, on the basis that Singaporeans may not readily equate non-consensual penile-oral penetration with penile-vaginal or penile-anal penetration in terms of its gravity.9

may be, of another person (B); (b) causes a man (B) to penetrate, with B’s penis, the vagina, anus or mouth, as the case may be, of another person including A. (c)causes another person (B), to sexually penetrate, with a part of B’s body (other than B’s penis, if a man) or anything else, the vagina or anus, as the case may be, of any person including A or B,

Following public consultations and receiving feedback from the social sector, the Government decided to expand the definition of rape to include both oral and anal penile penetration.10 Therefore, the proposed reform when effected would mean that males can now be victims of rape by either penile-oral or penile-anal penetration. However, the commission and offenders of rape remains gendered given that the perpetrators of rape are still exclusively and solely male. As will be discussed, females who compel male victims to penetrate them will not be legally deemed a rapist but a sexual assaulter.

shall be guilty of an offence if B did not consent to the penetration or if B is below 14 years of age, whether B did or did not consent to the penetration.

The Bill also proposed an expansion to section 376 of the Penal Code,11 amending the section to sexual assault involving penetration.12 The new amended section (in italics) would provide that:13 (1) Any man (A) who causes another man (B) to penetrate with B’s penis, the anus or mouth of A – (a) without B’s consent; or (b) with or without B’s consent, when B is below 14 years of age, shall be guilty of an offence.

228

Under the current law, compelled-penetration can be prosecuted via four routes: 1. 2. 3. 4.

sexual penetration of minor under 16;14 or sexual penetration of a corpse;15 or sexual penetration with living animal;16 or outrage of modesty.17

Such a state of affairs is unsatisfactory since the penalty of route 4, which captures compelledpenetration of males above 16 by females, involve an imprisonment of term not exceeding 2 years, a fine or caning.18 As the PCRC observed in their report, The principle should be that a woman who violates a man’s sexual autonomy by forcing the man to penetrate her vagina, anus, or mouth with his penis, as the case may be, is guilty of sexual assault.19

(2) Any person (A) who – (a)sexually penetrates, with a part of A’s body (other than A’s penis, if a man) or anything else, the vagina or anus, as the case

Clearly, the PCRC is of the view that compelledpenetration is a grave wrong suffered by the victim and the current regime is insufficient in protecting victims. The PCRC reasoned that since the current regime provides for routes 1 to 3, there is no reason why compelled-penetration

8 [73]. 9 10 11 12 13

14 Penal Code, Cap 224, s. 376A(1)(c). 15 ibid. s. 377(3). 16 s. 377B(3)(a). 17 s. 354(1). 18 ibid. 19 PCRC (n 2) at 26.30.

ibid. at 26.22. See PP v BMD [2013] SGHC 235 at ibid. at 26.23. First Reading (n 1) at para 25-26. Rev. Ed. Cap 224. Criminal Law Reform Bill (n 4) at clause 111(i). ibid. at clause 111(a) – (e).

COMMENTARY

‘should also not be covered by the Code’.20 The upshot of the reform to the section is that a woman who compels a man to penetrate her with his penis, whether vaginally, anally or orally can be found guilty of sexual assault involving penetration. While the proposed reform as a whole is a step in the right direction with the expansion of rape to include both penile-oral and penile-anal penetration, it does not go far enough, leaving much to be desired. In particular, this article argues that female compelled-penetration should be criminalised as rape and not just sexual assault, owing to the equivalence in harm of both the prevailing conception of rape and with compelled-penetration. This will be shown by examining misconceptions surrounding compelled-penetration. Misconception of Males as Rape Victims Misconception #1: Rape can only be carried out by males, they can never be victims of rape. Underlying this misconception are two mistaken assumptions: Underlying this misconception are two mistaken assumptions: 1. that males are never compelled to penetrate a woman; or 2. that they cannot sustain an erection and be compelled to penetrate a woman. First, it is simply untrue that males are never compelled to penetrate a woman. The National Intimate Partner and Sexual Violence Survey 2010 conducted in the United States found that ‘approximately 1 in 21 (4.8%) of men reported that they were made to penetrate someone else during their lifetime’, with 79.2% of the perpetrators being female.21 Similar, smaller 20 ibid. 26.29. 21 Brieding, M.J. Chen and Black M.C. Intimate Partner Violence in the United States - 2010 (Center for Injury Prevention and Control and Centers for Disease Control and Prevention, 2014) <http://www.cdc.gov/violenceprevention/pdf/cdc_ nisvs_ipv_report_2013_v17_single_a.pdf> accessed 11 March 2019 at pp. 2 and pp. 24.

scale studies conducted in Germany found that out of 247 younger men respondents to an anonymous questionnaire, 2.8% reported having been forced to vaginally penetrate a woman, while 5.2% of the 152 older respondents reported the same.22 While the numbers are small, they point to the existence of compelledpenetration. Further, the low percentage of compelled-penetration could be a result of under-reporting precipitated by traditional stereotypes of gender and masculinity,23 leading to a ‘dark figure’ of compelled-penetration that remain undiscovered and unreported. The situation is possibly worse in Singapore, given the regrettable dearth of compelledpenetration victim statistics. Only women can at present be considered victims of rape.24 Additionally, as noted by the PCRC, compelledpenetration is likely to be prosecuted under section 354 of the Penal Code.25 While the percentage of male victims is low,26 it is observed that they are on an increasing trend.27 Further, in a conservative society with deeply held notions of traditional gender roles, the social stigma deterring males from reporting such crimes might be more acute.28 22 Krahe, B., Scheinberger-Olwig, R. and Bieneck , S. ‘Men’s Reports of Non-consensual Sexual Interactions with Women: Prevalence and Impact’, (2003) 33 Arch Sex Behav 165 at pp. 172. 23 S. Weare, “Oh you’re a guy, how could you be raped by a woman, that makes no sense”: towards a case for legally recognising and labelling “forced-to-penetrate” cases as rape (2018) 14 Int. J.L.C. 110 at pp. 123. 24 Therefore, there are no male victims of rape recorded. Ministry of Social and Family Development, ‘Violence: Rape Victims’ (Ministry of Social and Family Development, 2018) <https://www.msf.gov.sg/research-and-data/Researchand-Statistics/Pages/Violence-Rape-Victims.aspx> accessed 12 March 2019. 25 PCRC (n 2) at pp. 326. 26 Ministry of Social and Family Development, ‘Violence: Outrage of Modesty Victims’ (Ministry of Social and Family Development, 2018) <https://www.msf.gov.sg/research-anddata/Research-and-Statistics/Pages/Violence-Rape-Victims. aspx> accessed 12 March 2019. 27 Although it is possible that the numbers are representative of male perpetrated outrage of modesty, until further studies are conducted, it remains a possibility that compelled-penetration victims are part of this group. 28 See for example the survey by AWARE, where in a survey of 500 youths where half of the respondents were male, only 6% sought help, citing ‘embarrassment, shame, family

229

COMMENTARY

Mr Kenny Liew, senior clinical psychologist at the Institute of Mental Health remarked: The difference [between female and male reporting of molestation] may be due to gender stereotypes and myths pertaining to sexual abuse of men, such as ‘boys and men cannot be sexually abused’ or ‘they were lucky to have been touched by a female’.29 Similar sentiments were shared by We Can! Singapore, a social movement aimed at ending violence towards woman, who wrote that: ‘…social stigma remains a serious barrier toward male victims seeking help. This isn’t helped by popular culture treating the rape of male prisoners as a joke or a perverse form of “justice”.’30 Further, even if were the case that males are never compelled to penetrate a woman, it does not necessarily entail that the criminal law should not intervene to protect the rights and dignity against potential violations. Second, detractors suggest that even if cases of compelled-penetration exists, males cannot be regarded as genuine victims. The claim is that if the supposed victims were able to sustain and erection and engaged in sexual activity, then this would be inconsistent with being the notion of being compelled”.31 Therefore, males can never be considered as genuine victims of rape, since the act is consensual or at the very

230

shame, disbelief and self-blame.’ as reasons for not seeking help. Association of Women for Action and Research, ‘Survey: 1 in 3 young people have faced sexual violence; few seek or receive help. (Association of Women for Action and Research, 27 March 2015) < http://www.aware.org.sg/2015/03/survey-1-in-3young-people-have-faced-sexual-violence-few-seek-or-receivehelp/> accessed 12 March 2019. 29 Theresa Tan, ‘Number of male victims of molestation on the rise’ (The Straits Times, 18 December 2018) <https:// www.straitstimes.com/singapore/number-of-male-victims-ofmolestation-on-the-rise> accessed 11 March 2019. 30 We Can! Singapore, ‘Busting Rape Myths’ (We Can! Singapore, 1 July 2017) <https://www.wecansingapore.com/ busting-rape-myths/> accessed 11 March 2019. 31 Weare (n 23) at pp. 114, 125.

least, conducted with the implied consent of the victim evidenced with the sustaining of an erection. Yet the above argument is deeply flawed, ignoring the physiological realities of the male reproductive system. Studies have shown that it is not necessary that an erection can only be brought about in instance of sexual excitement, but can be an involuntary response.32 Male victims have also reported ejaculating during these encounters.33 Thus, it is simply untrue that males are never and can never be victims of compelled-penetration. The sustaining of an erection cannot be viewed as an act of consent and compelled-penetration is essentially nonconsensual, indistinguishable from conventional understanding of rape. Misconception #2: Male ‘rape’ is not ‘real’ rape As with the first misconception, this misconception is similarly based on the false assumption that compelled-penetration is different from conventional understanding of rape: 1. in terms of its nature; and 2. in terms of its harm. Difference in Nature In terms of its nature, it has been argued that male ‘rape’ is different since men are assumed to be stronger and able to fend for themselves in attempts of compelled-penetration.34 Such a line of argument is not novel and have been employed by the defence involving female victims, where a ‘lack of physical resistance and injury are inconsistent with a claim of rape.’35 The argument is archaic and inconsistent with our understanding of rape. It has been long 32 P. Rumney, “Male Rape in the Courtroom: Issues and Concerns” (2001) Crim. L.R. 205, at pp. 207. 33 Fisher & Pina, “An Overview of the Literature on FemalePerpetrated Adult Male Sexual Victimization”(2013) 18 Aggress Violent Behav 54, at pp. 57 34 Weare (n 21) at pp. 115. 35 Rumney (n 30) at pp. 208. See also S. Lees, ‘Carnal Knowledge:Rape on Trial’ (London: Penguin Books, 1997) Ch. 4.

COMMENTARY

recognised that a lack of struggle or explicit denial is not evidence of consent.36

therefore, compelled-penetration should not be labelled and criminalised as rape.

Rape is the violation of an individual’s bodily and sexual autonomy and a ‘total denial of the victim’s agency, will and personhood.’37 This appears to be the principle guiding the law in Singapore, where rape, is fundamentally concerned with the non-consensual element of sexual intercourse, present in both the actus reus and mens rea of the offence.38 Force or injury operates as an aggravating factor that leads to imposition of mandatory sentences on the offender.39 If that is so, there is nothing in principle to deny males from being recognised as victims of rape, since it has been observed that female perpetrators employ similar tactics of physical violence, coercion and intoxication to compel penetration.40

Firstly, one might argue that the physical harm suffered by male victims are incomparable to female victims, since female victims are susceptible to the harms of genital injuries and the risk of pregnancy.44

Further, the principle of fair labelling, as picked up by the PCRC,41 requires that ‘the label applied to an offence ought fairly to represent the offender’s wrongdoing’,42 in terms of its ‘nature and magnitude of the law-breaking.’43 It is hoped that by dispelling the misconceptions surrounding male rape, it will demonstrate that the nature and magnitude of compelledpenetration is not dissimilar from rape by penetration. Difference in Harm

Yet, as Weare has pointed out, ‘the inconsistency of genital harms’ blurs any fundamental distinction that could be drawn between victims of rape and victims of compelled-penetration.45 She further argues that the ‘risk of pregnancy is no longer central to understandings of rape’ since the protection of the law extends to all women; the widening definition of rape; and the existence of contraceptive options.46 While agreeing with Weare in her conclusion, one should exercise caution with her argument of the existence of contraceptive options since the taking of such contraceptives is not strictly independent and is arguably an infringement of autonomy specific to female victims. That said, Weare tries to argue that male victims suffer from a particularised harm with a risk of a female perpetrator becoming pregnant and holding him hostage with the child.47 Physical harm also featured in the PCRC’s consideration of their recommendation for the expansion of the definition of rape:48

Even if one were to accept that compelledpenetration is procured in a similar manner by female perpetrators, it could still be contended that the harm suffered by male victims are different from female victims of rape and

‘Forced penile-oral penetration also carries the dangers of transmission of sexually transmitted diseases, and involves the invasion of the body with the penis.’

36 R v Olugboja [1982] QB 320. 37 A. Cahill, ‘Rethinking Rape’ (Ithaca NY: Cornell University Press, 2001) at pp. 132. 38 Penal Code, Cap 224, s. 375(1)(a). 39 ibid. s. 375(3). 40 Weare (n 23) at pp. 115 – 120. 41 See fn. 5. 42 A. Ashworth, ‘The Elasticity of Mens Rea’ in C.F.H. Tapper (ed), Crime, Proof and Punishment: Essays in Memory of Sir Rupert Cross (London: Butterworth, 1981) 45, at 53. 43 A. Ashworth, Principles of Criminal Law (Oxford: Oxford University Press, 5th Edn 2006) at pp. 88.

44 Weare (n 23) at pp. 120-121. See for example, the argument made by the Criminal Law Revision Committee 1984 that sought to distinguish rape by the risk of pregnancy. Rumney, ‘In Defence of Gender Neutrality within Rape’, (2007) 6 Seattle Journal for Social Justice 481, at pp. 483. 45 ibid. 46 ibid. 47 ibid. Which should be doubted since the harm is further particularised in that it would only occur in situations where the perpetrator is known to the victim, whereas a majority of women suffer from the risk of pregnancy. 48 PCRC (n 2) at pp. 325.

231

COMMENTARY

Yet, it is puzzling that the PCRC failed to consider the equivalency of harm suffered by male victims of compelled-penetration. It appears, as already noted above, the justification for reform stemmed from a desire for consistency in the law.49 Given that males are similarly at risk of being afflicted with sexually transmitted disease with compelled-penetration, it is really unclear why compelled-penetration should not be criminalised as rape.50 Further, male victims of compelled-penetration are susceptible to psychological and emotional traumas as with female victims. In particular, victims were found to display ‘fear, anxiety and distress, as well as changes in their behaviour towards, and relationships with, women’,51 with over 90% of victims being unable to perform sexually, mainly due to the fear of recreating the assault.52 Consistent with female victims, male victims also feel a sense of betrayal by their own bodies, when they sustain an uncontrollable erection and ejection.53 Reform to Reforms? The current proposed reforms are wide-ranging and would make sweeping changes to the criminal law in relation to the sexual offences of rape and sexual assault. It would mean that for the first time, males can be considered as victims of rape, with the expansion of rape to include both oral and anal-penile penetration, moving away from the conventional understanding of rape, which is limited to vaginal-penile penetration. Additionally, compelled-penetration would now be caught under section 376, sexual assault involving penetration, carrying the same penalties as rape.54 However, it has been argued that the reforms

232

49 PCRC (n 2) at pp. 325. 50 See above at fn 19 & 20. 51 ibid. at pp. 123. 52 Fisher & Pina (n 33) at pp. 56. 53 Weare (n 23) at pp. 125. Weare notes that the betrayal is arguably heightened where it is only as a result of their erection that they are victimised. 54 PCRC (n 2) at pp. 324.

do not go far enough, in that the commission of rape remains gendered, with only males capable of being caught as perpetrators of rape. As this article has urged, compelledpenetration should be appropriately labelled as rape, given the equivalence in terms of harm and procurement males can and are subjected to instances of compelled-penetration, which should be properly labelled as instances of rape. The law remains unsatisfactory if females compelling penetration by employing means similar to male rapist and inflicting similar harm to the victim are labelled as one who sexually assaults and not one who rapes.55 Therefore, this articles calls for a reform to the reforms, proposing that section 375 of the Penal Code to be amended as follows: (1) Any person (A) who penetrates the vagina (if a woman), anus or mouth of another person (B) – (a) without B’s consent; or (b) with or without B’s consent, when the person is below 14 years of age, shall be guilty of an offence. (1A) Any person (A) who causes another (B) to penetrate, with B’s penis, the vagina (if a woman), anus or mouth of A – (a) without B’s consent; or (b) with or without B’s consent, when the person is below 14 years of age, shall be guilty of an offence. Subsection (1) would capture instances of conventional understanding of rape, as well as penile-anal and oral penetration. Subsection (2) would capture instances of compelledpenetration of the vagina, anus or mouth of the perpetrator, regardless of the gender of the perpetrator. Consequently, the ambit of section 376 of the Penal Code would significantly be reduced, yet 55 Paul Powlesland, Male rape and the quest for genderneutrality in the Sexual Offences Act 2003, 2005 1(1) CSLR 11, at pp. 15.

COMMENTARY

retained to capture instances of compelledpenetration that do not involve penile penetration. It is proposed that section 376 to be reformed in line with the proposed changes to section 375 as follows: (1) Any person (A) who – (a) Sexually penetrates, with a part of A’s body (other than A’s penis, if a man) or anything else, the vagina, anus or mouth, as the case may be, of another person (B); (b) Causes another person (B), to sexually penetrate, with a part of B’s body (other than B’s penis, if a man) or anything else, the vagina, anus or mouth, as the case may be, of any person including A or B, Shall be guilty of an offence if B did not consent to the penetration or if B is under 14 years of age, whether B did or did not consent to the penetration. The proposed reforms would signal that rape is gender-neutral, providing better protection to victims of non-consensual forced penile-vaginal, anal or oral penetration, regardless of gender. This would allow for greater harmonisation in the aims of the PCRC to modernise the Penal Code, to ‘ensure that it covers circumstances where offences are committed against males and females, to reflect changing societal norms and views on the roles of men and women.’56

PCRC (n 2) at pp. 328.

233

COMMENTARY

Criminal Law (Temporary Provisions) (Amendment) Bill – An unjustified infringement on the right to a fair trial? Noel Low, University of Oxford Introduction The Criminal Law (Temporary Provisions) Act (“CLTPA”)1 is a legislation that, inter alia, allows the detention of suspected criminals without trial.2 Originally intended as a temporary measure when first introduced in 1955, the act has been continuously renewed for 5 years at a time.3 The latest renewal was passed in February 2018, with substantive amendments made. This article examines these changes, taking the view that the amendments have increased the scope of the power to detain without trial. Further, the legislation as a whole cannot be normatively justified as it poses a disproportionate breach of the right to a fair trial, in consideration of Rule of Law and human rights principles. 1. Justification for the CLTPA The legislation originated in the need to tackle criminal activity by secret societies. A normal trial procedure did not allow effective prosecution because witnesses involved with secret societies are fearful or unwilling to testify for fear of reprisal.4 While crime rates have fallen over the years, the Ministry of Home Affairs says that the CLTPA remains critical to maintain public order as the threats posed by secret societies and criminal syndicates still exist.5

234

1 Cap. 67, 2000 Rev. Ed. 2 CLTPA, s30 allows the Minister for Home Affairs, with the consent of the public prosecutor, to order that a person be detained (for a period not exceeding 12 months) if he is satisfied that the detention is necessary in the interests of “of public safety, peace and good order” 3 CLTPA, s1(2). The act expires after 5 years if not renewed, due to a sunset provision 4 The Straits Times, ‘Parliament: Criminal Law (Temporary Provisions) Act that allows detention without trial extended after heated debate’, 6 Feb 2018 5 Ministry of Home Affairs, ‘Criminal Law (Temporary Provisions) Act Factsheet’, 2018; and Speech by Mr K Shanmugam, Minister for Home Affairs and Minister for Law for Second Reading of the Criminal Law (Temporary

Proponents of the CLTPA also point to the safeguards in place to protect against abuse. The Public Prosecutor’s consent has to be obtained before a detention order is made, and the order is reviewed by an independent Criminal Law Advisory Committee comprised of prominent private citizens including Justices of the Peace, former judges and senior lawyers. From March 2018, the committees are chaired by sitting Judges of the Supreme Court of Singapore.6 The detainee or police supervisee will be informed of the grounds for his detention and be given the right to make his own representations to the Committee, and may also be represented by a lawyer of his choice. The Advisory Committee will sit to scrutinize the evidence and may examine the detainee, investigating officers and witnesses on the facts of the case, before making its recommendation to the President, who may cancel, confirm or vary the Orders on advice of the Cabinet.7 A further safeguard is that the detention order is open to judicial review through an Order of Review of Detention (formerly known as Habeas Corpus).8 The scope of such review is however limited, as will be explored below.

Provisions) (Amendment) Bill, 6 Feb 2018 6 The Straits Times, ‘Parliament: Supreme Court judges to review detention orders and act as safeguard, says Shanmugam’, 6 Feb 2018 7 Ministry of Home Affairs, ‘Criminal Law (Temporary Provisions) Act Factsheet’, 2018 8 The Singapore Constitution, Art. 9(2), states: “Where a complaint is made to the High Court or any Judge thereof that a person is being unlawfully detained, the Court shall inquire into the complaint and, unless satisfied that the detention is lawful, shall order him to be produced before the Court and release him.” The High Court is empowered by the Supreme Court of Judicature Act (Cap. 322, 2007 Rev. Ed.), s. 18(2) and para. 1 of the 1st Sch., to issue orders for the review of detention, and the procedure for applying for such an order is specified in O. 54 of the Rules of Court (Cap. 322, R 5, 2006 Rev. Ed.).

COMMENTARY

2. Amendments made in 2018 The 2018 amendment introduced a list of offences under which detention without trial can apply to, where previously it was not specified.9 This list is purported to clarify the act’s scope, and prevent arbitrary use of the power to detain.10 However, its actual effect is likely to increase the scope of the power to detain, in light of the context laid down by Tan Seet Eng v Attorney General11 prior to the 2018 amendment. The Court of Appeal in Tan Seet Eng held that the minister’s power in the CLTPA is restricted to a strict criteria, taking into consideration the act’s purpose, based on parliamentary debate on renewal of the bill.12 First, the CLTPA only covered criminal activities where trials in court could not proceed due to threats of harm against witnesses or their families. Secondly, the detainee must have been involved in a criminal activity of a sufficiently serious nature to justify detention without trial.13 Thirdly, the criminal activity must pose harm to the public order within Singapore.14 The detention of Tan Seet Eng (commonly known as Dan Tan), a ringleader of a global soccer match fixing syndicate, was struck down by the apex court as it did not affect the public order in Singapore. The new forth schedule, however, is likely to broaden the scope of the powers to detain, and change the result of a case with facts similar to that of Tan Seet Eng. As Tan’s alleged crime falls within an item on the Fourth schedule (“participation in, or facilitating, any organised crime activity as defined in the Organised Crime Act 2015”), the requirement that the activity 9 CLTPA Forth Schedule. The list includes: unlicensed moneylending; drug trafficking; involvement in a secret society or as a gangster; human trafficking; robbery with firearms; murder; gang rape; kidnapping; general offences relating to the participation in or facilitation of activities for an organised criminal group; and attempting to carry out, abetting, or being a party to a criminal conspiracy to carry out, any of the listed activities. 10 Ministry of Home Affairs, Press Release, 9 Jan 2018 11 [2016] 1 SLR 779 12 Tan Seet Eng at [119]-[120] 13 Tan Seet Eng at [135] 14 Tan Seet Eng at [137]

must pose a threat to the public order within Singapore might no longer hold. Further, concerns abound that the amendment allows detention for activities of a more minor nature, so long as they fall within a crime specified under the Fourth schedule.15 This would circumvent the second test laid down in Tan Seet Eng, that the criminal activity must be of a sufficiently serious nature to justify detention without trial. It remains to be seen whether the court will continue to apply the rigorous test in Tan Seet Eng, even where a detention order is made on a criminal activity that falls under the Forth Schedule. Another controversial amendment made is s30 (2), that every decision of the minister made on an order of detention or supervision under s30 is “final”. The government has clarified that the finality clause does not affect the courts’ power of judicial review, based on the traditional tests of illegality, irrationality and procedural impropriety.16 Rather, the clause is purportedly meant to crystallize that the minister can “establish the facts of a case, apply those facts to the relevant rules and considerations, and exercise his discretion.”17 This means that the minister has final say over whether a person is linked to criminal activities and whether detention is necessary for reasons of public safety, peace and good order.18 While the government has argued that this clause is mere codification of existing case law,19 this is not clearly reflected in the way the clause is worded and it is hoped that there would be no substantive impact on future cases.

15 Speech by Sylvia Lim in Parliament, 6 February 2018 16 Indranee Rajah S.C., ‘The Criminal Law (Temporary Provisions) (Amendment) Bill’, 20 Mar 2018, p2 17 ibid. 18 The Straits Times, ‘Some lawyers worried about how Act will be used’, 10 Jan 2018 19 Indranee Rajah S.C., ‘The Criminal Law (Temporary Provisions) (Amendment) Bill’, 20 Mar 2018, p2

235

COMMENTARY

3. Arguments against detention without trial under the CLTPA 3.1 Constitutional principle Notwithstanding the 2018 amendments, this article also takes a position against the general grain of the CLTPA as its power of detention without trial is a disproportionate infringement to the principles of Rule of Law, Separation of Powers, and fundamental rights. With the social changes that have taken place since the act was first conceived, both in terms of the threat to public order and of the increasing trend towards the protection of individual rights, there is a need to evaluate whether the legislation still passes muster. As brought up by Mr K Shanmugam (then a backbencher Member of Parliament) in a 1989 Parliamentary speech,20 the original justification of the CLTPA, which was to combat the strong presence of secret society activity, is no longer present. Although the number of detention orders issued under the CLTPA for drug offences have plummeted to zero as of 2017,21 the measure is still kept “as a backup” to the Misuse of Drug Act.22 While the CLTPA is still used for other areas of syndicate activity such as illegal moneylending, it is submitted that there is insufficient justification for the infringement on the fundamental rights that will be explored below. Further, there is a danger of arbitrary abuse in allowing such wide discretion to the executive in determining criminal guilt and detention. While there is no suggestion that the CLTPA powers are exercised with anything other than good faith, the lack of due process which does away with the rigorous procedures of the ordinary criminal law and scrutiny from the judiciary risks the wrongful detention of the innocent. Famously, the late criminal lawyer Subhas Anandan was once detained without trial under the CLTPA for alleged secret society

236

20 Tan Seet Eng at [112] 21 Speech by Mr K Shanmugam, Minister for Home Affairs and Minister for Law for Second Reading of the Criminal Law (Temporary Provisions) (Amendment) Bill, 6 Feb 2018 22 ibid.

activity, for which he was later exonerated.23 The power granted to the Minister of Home Affairs to order a detention without trial under the CLTPA also violates Separation of Powers theory, on which the Singapore Constitution is based upon. Art 93 of the Singapore Constitution vests judicial power in the courts.24 Persecution for criminal activity should thus be adjudicated upon by the judiciary and not the executive. Baron de Montesquieu argued that there can be no liberty if the judiciary is not separated from the legislative and the executive, writing that there would be arbitrary control and oppression.25 It is thus constitutionally inappropriate that the CLTPA allows the executive to perform a judicial function. The act also breaches Rule of Law principles. The first strand of A.V. Dicey’s conception of the Rule of Law is that no individual can be punished except through the process of law and the courts.26 Tom Bingham argues that ‘the core of the existing principle is … that all persons and authorities within the state, whether public or private, should be bound by and entitled to the benefit of laws publicly and prospectively promulgated and publicly administered in the courts’.27 Vesting the power of fact-finding and ordering of a detention to the Minister in place of the courts is thus in violation of fundamental constitutional principles. It is particularly concerning that the fact-finding portion of the detention order cannot be reviewed by the courts through judicial review, as discussed above. While the independent advisory committee adds an additional layer of scrutiny, it does not detract from the fact that the process turns on the discretion of the executive since the committee does not have 23 The Straits Times, ‘Subhas case: CPIB probe”, 22 March 1976 24 Constitution of the Republic of Singapore, Art 93 25 Baron de Montesquieu, The Spirit of the Laws (Thomas Nugent trans) (Hafner Press, 1949), at p 151 26 A. V. Dicey, Introduction to the Study of Law of the Constitution (1885; 10th edn., Macmillan & Co., 1959), pp.187-95 27 Tom Bingham, ‘The Rule of Law’ (2007) 66 CLJ 67, 69

COMMENTARY

authority to quash the detention order. 3.2 Human Rights and the UK perspective Further, the CLTPA raises questions of human rights, which can be examined through the lens of UK case law. The position in the UK, which is bound by the European Convention on Human Rights (ECHR), is that imprisonment without trial is an infringement on liberty and the right to a fair trial. In the aftermath of the 9/11 attacks, the Anti-Terrorism Crime and Security Act 2001 was enacted which allowed for a foreign national suspected of involvement in terrorism to be indefinitely imprisoned without charge or trial. This was held by the House of Lords to be incompatible with the right to liberty under Article 5 of the ECHR28 in A v Secretary of State for the Home Department,29 causing the legislature to abolish the measure. The European Court of Human Rights held in the case of Brogan v United Kingdom30 that Article 5 requires prompt release of a suspect if he is not charged or brought before a court. In the case, detention without trial for over four days without a strong justification for the delay was held to be an infringement of the ECHR right. A proposal to extend the maximum period of detention without charge under the Terrorism Act 2000 from 14 to 90 days was rejected by the House of Commons in 2006. There was also strong opposition from prominent members of the public including Archbishop Desmond Tutu, who likened it to a similar law under Apartheid.31 The maximum period was extended to 28 days instead.32 Taking the European approach further, the CLTPA would also offend the right to a fair trial under Article 6 of the ECHR, which provides, inter alia, that:

and obligations or of any criminal charge against him, everyone is entitled to a fair and public hearing within a reasonable time by an independent and impartial tribunal established by law… (2) Everyone charged with a criminal offence shall be presumed innocent until proved guilty according to law. (3) Everyone charged with a criminal offence has the following minimum rights: (a) to be informed promptly, in a language which he understands and in detail, of the nature and cause of the accusation against him. … (d) to examine or have examined witnesses against him and to obtain the attendance and examination of witnesses on his behalf under the same conditions as witnesses against him. The process under the CLTPA does not provide for the determination of a detention order to be decided by a tribunal independent of the executive, whose proper role is to prosecute rather than to adjudicate. Questions of procedural fairness thus arise, which can be resolved by adopting a judiciary-led approach discussed below. Taking heed from the UK interpretation of Human Rights then, the power to detain without trial for a period of 12 months, as granted by the CLTPA, is an unacceptable interference with the right to liberty and fair trial. While the ECHR does not apply to Singapore, there is no reason why the protection of fundamental rights should be any weaker in Singapore than it is in the UK, especially so when the threats to public order are no less serious in the UK.

(1) In the determination of his civil rights 28 29 30 31 Feb 2006 32

European Convention on Human Rights Article 5 [2004] UKHL 56 11 EHRR 117 1988 The Guardian, ‘Blair: Guantánamo is an anomaly’, 17 Terrorism Act 2006 s23

4. Conclusion: Adopting a middle ground approach from the UK The problems with the CLTPA can be resolved by changing the process to one that is conducted 237

COMMENTARY

by the judiciary. Inspiration can be taken from the Closed Material Procedure33 in the UK, which does not compromise on the identity of witnesses who fear reprisal from testifying. In a CMP, sensitive evidence is heard in a closed court where even the defendant is excluded. The defendant is represented in the closed proceedings by a special advocate, who cannot disclose the confidential information to the defendant without permission but whose role is to take instructions and defend his interests. In Secretary of State for the Home Department v. AF (No 3)34 where the applicant was placed under a control order following a CMP, the House of Lords held that there was no reason why a closed material proceeding was inconsistent with the principles of natural justice per se, but that the detainee has to be given sufficient information about the allegations against him to enable him to give effective instructions to his special advocate. Parallels from CMP can thus be drawn to the existing independent committee review procedure under the CLTPA, except that (1) the procedure is conducted by the courts and not a committee of private citizens; (2) the decision lies with the judiciary and not the executive; (3) The detention order is made by the courts in the first place, rather than the procedure being a safeguard or a review of the Ministerâ&#x20AC;&#x2122;s order; and (4) the presence of special advocate who can see the closed evidence mounted against the defendant. A procedure akin to the CMP is thus normatively justifiable on principles of constitutional theory and likely to create a fairer process for alleged criminals facing the CLTPA.

238

33 34

Justice and Security Act 2013 [2009] UKHL 28

COMMENTARY

True or False or Misleading? “[A]dequate judicial oversight” over Part 3 Directions under the Protection from Online Falsehoods and Manipulation Act Lee Kay Howe, University of Oxford Disinformation, or ‘fake news,’ is a serious threat to democracy, with the Minister for Law asserting that it has “combined, like a battering ram” with “inequality, inequity, [and] political systems not delivering results” to “damage [and] destroy” the foundations of democracy.1 Modern media and technology have given disinformation a “new power,” with “fake accounts; digital advertising; and algorithms” magnifying its impact.2 In response, Parliament has passed the Protection from Online Falsehoods and Manipulation Act (POFMA), in order to, inter alia, “prevent the communication of false statements of fact in Singapore and to… counteract the effects of such communication”.3 This is undoubtedly desirable. Nonetheless, questions remain as to whether certain provisions of the Act will instead harm democracy by creating a “chilling effect”4 on free speech and “curtail important discussion of matters of public interest… including content critical of the government.”5 The government, in response, has promised “adequate judicial oversight”6 of the Act. In 1 Protection from Online Falsehoods and Manipulation Bill (POFMB) Debate 7 May 2019. Per the Minister, the foundations of democracy include “trust, free speech and the infrastructure of fact”. 2 ibid. See Select Committee on Deliberate Online Falsehoods, Report of the Select Committee on Deliberate Online Falsehoods – Causes, Consequences and Countermeasures (Parliament 2018) at p. 5 – 13 and 187 – 188 for details of its impact worldwide, and p. 51 – 58 for the specific threat to Singapore. 3 POFMA, s. 5(a). 4 Cherian George, ‘Singapore’s online falsehoods bill will deepen a culture of self-censorship’ (New Mandala, 26 April 2019) <https://www.newmandala.org/singapores-onlinefalsehoods-bill-will-deepen-a-culture-of-self-censorship/> accessed 5 June 2019 . 5 International Commission of Jurists, ‘RE: Protection from Online Falsehoods and Manipulation Bill 2019’ (International Commission of Jurists, 12 Apr 2019) <https:// www.icj.org/wp-content/uploads/2019/04/Singapore-onlineregulation-bill-letter-advocacy-open-letter-2019-ENG.pdf ? accessed 5 June 2019. 6 POFMB Debate (n 1).

exploring the “adequa[cy]” of this judicial oversight, this article will begin with a discussion of the discretion-conferring provisions in Part I; an examination of the appeals mechanisms of the Act in Part II; an assessment of judicial review procedures in Part III; a review of cost and time-frame issues in Part IV; and a final analysis in Part V which concludes that judicial oversight of the Act is, indeed, adequate. Throughout the analysis, this article pursues two lines of inquiry – whether the judicial oversight mechanisms are adequate to “ensure due process and the proper exercise of power, and give assurance to the public of the integrity of the decision-making process”;7 and whether they are adequate to ensure that free speech is not unduly “chilled”. I. Ministerial discretion under Part 3 of POFMA One of the main concerns raised by the Select Committee on Deliberate Online Falsehoods was the potential dangers of “virality,”8 which can lead to disinformation spreading and causing harm in a very short span of time. As such, the Select Committee recommended that the Government should have the powers, as the first-instance decision-maker, to “swiftly disrupt the spread and influence of online falsehoods”.9 This recommendation has been effected by several discretionary powers in POFMA, addressed to individuals, internet intermediaries, and providers of mass media services. Of these, Part 3 Directions have the most significant impact on individuals. Under Part 3, any Minister has a wide discretion to 7 Select Committee Report (n 2) at [438], but see also [364]. 8 The “relative ease and affordability with which [falsehoods] are transmitted,” ibid. at [3] and [396]. 9 ibid., at [438].

239

COMMENTARY

issue Correction Directions (s. 11) and Stop Communication Directions (s. 12), which direct a person to communicate a correction; or to stop communicating an offending statement, respectively, as long as the Minister determines, per s. 10(1)(a), that “a false statement of fact… has been or is being communicated in Singapore”, and, per s. 10(1)(b), “is of the opinion that it is in the public interest to issue the Direction.”10 Adequate judicial oversight is clearly needed to prevent these Directions from being abused to political ends, or to stifle freedom of speech. II. The appeals mechanism under s. 17 of POFMA The appeals mechanism in s. 17 of POFMA is intended to provide applicants with a check on the Minister’s discretion to issue a s. 11 Correction Direction or a s. 12 Stop Communication Direction. As described by the Law Minister, the process in s. 17 has “deliberately made the test lower as in, it is an appeal rather than a judicial review”11 (emphasis added). This means that any appeal of a Part 3 Direction allows the High Court to examine the “question of fact or falsehood”12 de novo and “substitute its view for that of the Minister”13 if, per s. 17(5)(a), “the person did not communicate in Singapore the subject statement;” or (b), “the subject statement is not a statement of fact, or is a true statement of fact;” or (c), “it is not technically possible to comply with the Direction.” This is a lower bar than the traditional heads of judicial review of ‘irrationality’, ‘illegality’, or ‘procedural impropriety’,14 and it is hoped that this procedure “should be fairly straightforward”.15

240

10 It should also be noted, per ss. 11(4) and 12(4), that a Part 3 Direction even if he or she “does not know or has no reason to believe that the statement is false.” 11 POFMB Debate 8 May 2019. 12 ibid. 13 ibid. 14 Per Lord Diplock’s seminal formulation in Council of Civil Service Unions v Minister for the Civil Service [1984] UKHL 9; [1985] AC 374 (“CCSU”) at 410 – 412. 15 POFMB Debate (n 11), per K Shanmugam, Law Minister.

The grounds of appeal under s. 17(5)(a) and (c) are uncontroversial. The uncertainty lies with s. 17(5)(b), on whether “the subject statement is not a statement of fact, or is a true statement of fact”. This can be split into two sub-grounds: first, whether the impugned statement is a “statement of fact”; and second, if the impugned statement is a “statement of fact”, whether it is “true”. To the draftsman’s credit, s. 2(2)(a) clarifies that “a statement of fact is a statement which a reasonable person seeing, hearing or otherwise perceiving it would consider to be a representation of fact,”16 ,” and s. 2(2)(b) states that “a statement is false if it is false or misleading, whether wholly or in part, and whether on its own or in the context in which it appears.” These, however, are not definitions of the terms in s. 17(5)(b) and the Law Minister has insisted instead that these distinctions are well addressed in the “existing case law.”17 There is, of course, nothing wrong with a casuistic approach – but the inability to provide statutory definitions suggests that these distinctions are not as “clear” as the Minister asserts they are.18 When is a statement a statement of fact? While there is existing jurisprudence on when a statement will count as a “comment” or a “statement of fact”, this distinction is, by the Court of Appeal’s own admission, “not always easy to make”.19 Rather, the objective test can be difficult to apply, and arguably leaves significant uncertainty in the law.20 In Oversea-Chinese Banking

16 This mirrors the objective test in the case law, see Review Publishing v Lee Hsien Loong [2009] SGCA 56; [2010] 1 SLR 52 at [140] which provides a test [1994] SGHC 174; [1994] 3 SLR(R) 410. of “whether an ordinary, reasonable reader on reading the whole article would understand the words as comment[s] or [as] statements of fact”. 17 POFMB Debate (n 1). 18 ibid., see below for examples of difficulties in these distinctions. 19 Oei Hong Leong v Ban Song Long David [2005] SGCA 35; [2005] 3 SLR(R) 608 at [44]; see also Chen Cheng v Central Christian Church [1998] SGCA 51; [1998] 3 SLR(R) 236 (“Chen Cheng”) at [33] – [38]. 20 Review Publishing [2009] (n 16) on the “reasonable reader” test.

COMMENTARY

Corp Ltd v Wright Norman (“OCBC”)21, the plaintiff bank sued the defendant consultant for statements made in the Business Times about the lack of confidentiality in the bank’s search for senior executives. One of the statements read, “but there is a prima facie case of rank amateurism or carelessness at OCBC”.22 Considering how the statement was “expressed, the context in which it [wa]s set out and the content of the entire… passage in question,”23 Chao Hick Tin J held that the term “prima facie” was “not necessarily conclusive that what [was] stated [we]re comments”,24 and Chao J found that the statement was a statement of fact. With respect to the honourable judge, however, the issue with this statement is not with the term “prima facie,” but in the use of the words “rank amateurism” and “carelessness”. These are evaluative words, and may be described as “something which is or can reasonably be inferred to be a deduction, inference, conclusion, criticism, remark, observation, etc.”25 What is “amateur” or “careless” to one person may not be so to another, and neither of these are terms of art in the law, nor in common parlance.26 In any case, if the OCBC case is good law, it is hard to square away with the Law Minister’s examples of “opinions [, that are not] caught by this Bill”,27 including: “the Government is to be blamed for rising inequality”;28 “human rights in Singapore is being curtailed”;29 and “the Government is not giving back my CPF(Savings)”.30 With respect, given the 21 [1994] SGHC 174; [1994] 3 SLR(R) 410. 22 ibid., at [35] – [36]. 23 Chen Cheng (n 19) at [35]. 24 ibid. 25 Patrick Wilmo and W V H Rogers (ed.), Gatley on Libel and Slander (9th Ed., Sweet & Maxwell Ltd, 1998) at para. 12.6, quoted in Chen Cheng (n 19) at [34]. 26 The full context of the three libellous passages in the OCBC case is provided below, in Appendix A. 27 Tham Yuen-C, ‘Government makes initial decision on falsehood but courts are final arbiter of truth: Shanmugam’ The Straits Times (Singapore, 2 April 2019) <https://www. straitstimes.com/politics/govt-makes-initial-decision-onfalsehood-but-courts-are-final-arbiter-of-truth-k-shanmugam> accessed 5 June 2019. 28 ibid. 29 ibid. 30 ibid.

decision in the OCBC case, it is not obvious that the reasonable person would regard these as opinions rather than statements of fact, given that they are categorical and indicative, rather than evaluative. Another difficult example is the controversy over Vietnamese actions in Cambodia in 1979. While some have called it an “invasion,”31 others have said that this is “not true,”32 and “does not reflect history”33 as the Cambodians had instead been “liberate[d].”34 The confusion here arises because words such as “invasion” are usually thought of as statements of fact,35 whereas in certain cases these will merely be opinions (in this case, of Vietnamese actions in 1979).36 The line can be very difficult to draw, and may take on political overtones. In all fairness to Chao J, however, the case law does not provide a determinate test for what is a ‘statement of fact,’ and what is a ‘comment’ or opinion. Some jurisdictions have noted this and moved away from this distinction. In Pearson v Fairbanks Publishing Co,37 the Alaskan Supreme Court held that “[t]he distinction between a fact statement and an opinion or comment is so tenuous in most instances, that any attempt to distinguish between the two will lead to needless confusion.” The court instead decided to extend the privilege of fair comment38 to 31 Lee Hsien Loong, ‘Lee Hsien Loong – Posts’ (Facebook, 31 May 2019) <https://www.facebook.com/ leehsienloong/posts/2475835199145838> accessed 5 June 2019. 32 Ben Sokhean, ‘Tea Banh dismisses Singapore PM’s statement as ‘untruthful’’ Khmer Times (Phnom Penh, 5 Jun 2019) <https://www.khmertimeskh.com/50611039/tea-banhdismisses-singapore-pms-statement-as-untruthful/> accessed 5 June 2019. 33 ibid. 34 ibid. 35 See, for instance, Calvin Cheng, ‘Calvin Cheng – Posts’ (Facebook, 8 June 2019) <https://www.facebook.com/ calvinchengnmp/posts/2383863914997004> accessed 8 June 2019. 36 The Oxford English Dictionary defines “invasion” as “The action of invading a country or territory as an enemy; an entrance or incursion with armed force; a hostile inroad.” The question here is whether Vietnam entered “as an enemy” or was “hostile”, or not. 37 413 P.2d 711 (Ala. 1966) 38 This defence is limited to “comments” in Singapore,

241

COMMENTARY

non-malicious statements of fact, finding in favour of the defendant publication, which asserted that the plaintiff columnist had been described by his colleagues as the “Garbage Man of the Fourth Estate,” and that his column had been discontinued because the publication “did not wish to distribute garbage”.39 Philosophically speaking, it is difficult, if not impossible, to find a principled distinction between fact and opinion; the relationship between the two is more accurately characterised as a “continuum”.40 For instance, one might assert that a fact is something that can be “proven [to be] true”.41 This maps on to the common understanding that facts are indicative and categorical, whereas opinions are evaluative. However, provability can be “audiencerelative,”42 – while one person might find, for example, Anselm’s ontological argument43 to be sufficient proof of God’s existence, others may not. Thus, on this principle, the statement that “[t]he earth was created by an omnipotent God” will be a fact, rather than an opinion.44 The same will result if we adopt a principle that facts are “objective” and “mind-independent” and that opinions are “subjective” and “minddependent” – because whether or not God created the earth is something that happened (or not), independent of whether we believe it.45 Thus, the indeterminacy of the fact-opinion distinction creates difficulties and uncertainties for s. 17(5)(b), placing a heavy dependence on

242

with other criteria to be satisfied, see Chen Cheng (n 19) at [33]. 39 Pearson (n 37), at p. 712. 40 Michaele Sanders, ‘The Fact/Opinion Distinction: An Analysis of the Subjectivity of Language and Law’ (1987) 70 Marquette Law Review 673 at p. 680. See also the judgment of Chief Judge Robinson in the DC Court of Appeal in Ollman v Evans, 750 F.2d 970 (DC Cir 1984), cert. denied, 471 US 1127 (1985) at p. 1022. 41 John Corvino, ‘The Face/ Opinion Distinction’ (The Philosopher’s Magazine, 4 March 2015) <https:// www.philosophersmag.com/essays/26-the-fact-opiniondistinction> accessed 8 June 2019. 42 ibid. 43 ibid. See also s. 9 of Graham Oppy, ‘Ontological Arguments’ (Stanford Encyclopedia of Philosophy, 8 February 1996) < https://plato.stanford.edu/entries/ontological-arguments/> accessed 8 June 2019. 44 ibid. 45 ibid.

the “common sense”46 of judges to tell the difference. When is a statement of fact false or misleading? If the court finds that the impugned statement is a statement of fact, the applicant may still succeed in his or her appeal if that statement of fact is true. Read against s. 2(2)(b), this requirement will be satisfied if the statement is not “false or misleading.” The application of this test is closely linked to what is admitted as a statement of fact, and, in particular, whether it has an evaluative component: The test for whether a statement is false or misleading47 is straightforward if the statement is “stable”48 or ““fact of the matter” at a granular level,”49 such as ‘all cats are blue,’50 or “the rooftop in Punggol Waterway Terraces had collapsed”.51 However, there are many other statements that are not easily provable on a balance of probabilities.52 Indeed, “[m]ost of what we can say about social phenomenon, how we say it, and in what direction, depends deeply on standpoint and interests”53 and “concepts with an interpretive or evaluative component are less straightforward: people may validly perceive different meanings or values in the same action 46 Chen Cheng (n 19) at [33]. 47 See, for example, Freely Pte Ltd v Ong Kaili [2010] SGHC 60; [2010] 2 SLR 1065 and the Banking Act (Cap 19, 2008 rev ed) s. 66(2); Money-changing and Remittance Businesses Act (Cap 187, 2008 rev ed) s. 7(1); Consumer Protection (Fair Trading) Act (Cap 52A, 2009 rev ed) s. 4(a) and Securities and Futures Act (Cap 289, 2006 rev ed) s. 199. 48 Teo You Yenn, ‘Academic Freedom in Singapore and the “Fake News” Law’ (New Naratif, 11 April 2019) <https:// newnaratif.com/journalism/academic-freedom-in-singaporeand-the-fake-news-law/> accessed 5 June 2019. 49 ibid. 50 Example mine. 51 ‘Examples of falsehoods in Singapore and other countries,’ The Straits Times, 4 April 2017, at https://www. straitstimes.com/singapore/examples-of-falsehoods-insingapore-and-other-countries. This statement may actually not be clear-cut in certain circumstances; for instance, how much of a rooftop has to fall in before you can say that the rooftop has collapsed? What if only a few tiles had broken? 52 See, for instance, Lee Tat Development Pte Ltd v Management Corporation Strata Title Plan No 301 [2018] SGCA 50; [2018] 2 SLR 866 at [177]. 53 Teo You Yenn (n 48).

COMMENTARY

or interaction”. However, there are many other statements that are not easily provable on a balance of probabilities. Indeed, “[m]ost of what we can say about social phenomenon, how we say it, and in what direction, depends deeply on standpoint and interests”54 and “concepts with an interpretive or evaluative component are less straightforward: people may validly perceive different meanings or values in the same action or interaction”. Words with multiple meanings To take an example from R v Monopolies and Mergers Commission, ex parte South Yorkshire Transport,55 the House of Lords had to determine what was a “substantial part of the United Kingdom” under s. 64(3) of the Fair Trading Act 1973 (UK). Here, Lord Mustill noted that “as a matter of common language no recourse need be made to dictionaries to establish that “substantial” accommodates a wide range of meanings”, and that the word “substantial” was of a “protean nature”. Although this was a judicial review of an authority’s interpretation of a statutory term, Lord Mustill’s comments are clearly relevant to the current context. It is difficult to assess whether a “protean” word is true, false, or misleading even if one examines the context in which it is made – different people will have different views on whether, for example, certain conduct is “amateur” or “careless”.56 A variant issue is that ostensible statements of fact are sometimes contested. This is most prominent in the field of academic research, or challenges to “conventional wisdom”.57 For 54 ibid. 55 [1993] 1 All ER 289 (HL). 56 OCBC (n 21). 57 POFMB Debate (n 11). The Minister for Education has asserted “how impossible it is, for academic research to run afoul of POFMA” and that while “[s]ome of the works of humanities scholars… can be controversial, highly debatable, even offensive to some…. [t]heir conclusions are in the form of hypotheses, theories and opinions that are not covered by [POFMA].” However, with respect, this statement is merely an indication of the way which the Government intends to use its Part 3 powers, and does not determine, as a matter of law,

instance, while the current fashion is to think of the 1965 Separation Agreement as a “mutual agreement”58 between Singapore and Malaysia, the traditional orthodoxy held that Singapore had been one-sidedly “booted or kicked out of Malaysia,”59 a fact which has been used to support the state’s narrative of “vulnerability” and “survival.”60 One can also conceivably see the Vietnamese invasion/ liberation example above being regarded as a statement of fact, and being evaluated for its truthfulness. Here, the difficulty is different – the court is arguably not the best forum to decide contested, polycentric, academic facts; and if it is used as one, it will be hard to challenge the orthodoxy of the time. Burden of proof Furthermore, if the burden of proof is on the appellant to prove that the impugned statement of fact is true, it may be difficult to do so if there is an “information asymmetry between the Government and individuals”.61 In response, Government MP Murali Pillai, has asserted that “[t]he provisions are quite clear. The legal burden always is rested with the Executive to prove that the Part 3 directions comply with the criteria for falsehood as well as public interest.”62 The veracity of this statement depends on how one reads s. 105 of the Evidence Act,63 that “[t]he burden of proof as to any particular fact lies on that person who wishes the court to believe in its existence, unless it is provided by any law that the proof of that fact shall lie on any particular person.” Is the fact to be whether any such theories can come under Part 3. 58 Edmund Lim, ‘Getting textbooks to accurately depict how Singapore gained independence,’ Today, 21 May 2019, at https://www.todayonline.com/commentary/gettinghistory-textbooks-accurately-depict-how-singapore-gainedindependence. 59 ibid. 60 Loh Kah Seng, ‘Within the Singapore Story: The Use and Narrative of History in Singapore,’ (1998) 12 Crossroads: An Interdisciplinary Journal of Southeast Asian Studies 1 at p. 12. 61 POFMB Debate (n 11) per Sylvia Lim MP. An apposite example would be historiographical challenges such as those mentioned in the paragraph above. 62 ibid., per Murali Pillai MP. 63 (Cap 97, 1997 rev ed)

243

COMMENTARY

proved that the statement is true? Or that the Minister has complied with the provisions in ss. 10(1)(a) and (b)? With respect to Pillai MP, on a reading of s. 17(5)(b), that “The High Court may only set aside a Part 3 Direction on any of the following grounds on an appeal… (b) the subject statement is not a statement of fact, or is a true statement of fact,” the burden of proof seems squarely on the applicant. If this is the case, then the burden on the appellant could potentially be “onerous”.64 III. Judicial review of a Part 3 Direction Objections have been raised that the s. 17(5) appeals mechanism does not allow the aggrieved party to appeal against the Minister’s opinion of the “public interest” under s. 10(1) (b) and s. 4, and so the court cannot “review the minister’s opinion that the correction or takedown is in the public interest”.65 This can be dealt with shortly. That the appeals mechanism under s. 17(5) does not include a “public interest” ground does not preclude judicial review of the Minister’s decision. Indeed, this is contemplated by s. 18, which provides that “[o]ther causes of action [are] not affected” by the Act. Rather, this simply means that the court cannot “substitute its view for that of the Minister”66 on a determination of the public interest. The courts can, however, decide if the Minister’s decision is flawed for “illegality”, “irrationality” or “procedural impropriety”67, and subsequently set it aside. Justiciability and deference The orthodox position on judicial review in Singapore is the “green-light” approach.68 This “sees public administration not as a necessary

244

64 POFMB Debate (n 11) per Sylvia Lim MP. 65 See, for instance, Cherian George, ‘Give judges more leeway under fake news law’ Today (Singapore, 5 Apr 2019) <https://www.todayonline.com/commentary/making-courtsfinal-arbiter-online-truths-and-falsehoods> accessed 66 POFMB Debate (n 1). 67 CCSU (n 14) at 410 – 412. 68 Chan Sek Keong, ‘Judicial Review – From Angst to Empathy,’ (2010) 22 SAcLJ 469 at [29].

evil but a positive attribute, and the objective of administrative law as not (primarily) to stop bad administrative practices but to encourage good ones”. The collaborative “green-light” approach has been given judicial recognition in Jeyaretnam Kenneth Andrew v Attorney-General69 and is well reflected in the case law on justiciability and is well reflected in the case law on justiciability and deference.70 In Lee Hsien Loong v Review Publishing Co Ltd,71 Sundaresh Menon JC (as he was then) held that there were “clearly provinces of executive decision-making that are, and should be immune from judicial review,”72 taking into account “the simple fact that there are certain questions in respect of which there can be no expectation that an unelected judiciary will play any role.”73 Menon JC laid out four principles that militate against a matter being justiciable, including (a) where “the executive has access to the best materials available to resolve the issue”;74 (b) the decision “involves matters of government policy and requires the intricate balancing of various competing policy considerations that judges are ill-equipped to adjudicate because of their limited training, experience, and access to materials”;75 (c) a “judicial pronouncement could embarrass some other branch of government or tie its hands in the conduct of affairs traditionally regarded as falling within its purview”;76 and (d) the subject matter falls within the “areas of prerogative power that the democratically elected Executive and Legislature are entrusted to take charge of ”.77 The issue was further considered by Steven Chong J (as he was then) in Yong Vui Kong v Attorney-General78 – the court would decline to 69 [2013] SGCA 56; [2014] 1 SLR 345 at [48] – [49]. 70 Eugene Tan, ‘Curial Deference in Singapore Public Law,’ (2017) 29 SAcLJ 800. 71 [2007] SGHC 24; [2007] 2 SLR(R) 453 at [95] – [98] (to be distinguished from the CA case at n 16). 72 ibid., at [95]. 73 ibid., at [98]. 74 ibid. 75 ibid. 76 ibid. 77 ibid. 78 [2010] SGHC 235; [2010] 1 SLR 1.

COMMENTARY

conduct judicial review where (a) there was an absence of objective legal standards; (b) the court lacks specialist knowledge; and (c) the issue was polycentric and involved multiple concerns and perspectives.79 It may be that the “public interest” requirement of a Part 3 Direction will not be reviewable as it arguably falls within all the factors mentioned by Menon JC and Chong J above. However, even if the court decides that this matter of “high policy” is justiciable,80 the courts will still have to calibrate the “intensity of review,”81 which may be weak, taking into account the courts’ “[l]imited institutional competence”82 in this area. There will also be significant hurdles to cross in the application for review, which require the applicant to have an arguable case,83 and, typically, for the applicant to have exhausted all alternative remedies,84 including that in s. 17(5). This remedy is clearly meant as a last resort. Nonetheless, given the differences in subject-matter between s. 10(1)(b) and s. 17(5) (b), judicial review is clearly more appropriate than an appeal – the court can still examine the Minister’s decision, and will set it aside if it is clearly not in the public interest to issue it85 – but this will be done in a way that is mindful of the courts’ institutional and democratic constraints.

appeals procedure will also be expedited, as follows: 1. Per s. 19(2)(b), the applicant first has to apply to the Minister to vary or cancel the Part 3 Direction. There will be a “standard form”87 to be sent to a specified e-mail address. The Minister will then have to respond within two working days. 2. If the Minister rejects the application, the applicant can then use an expedited appeals process to the High Court. The applicant has 14 days to file an application using “[s]imple standard forms,”88 and will have to appear before the Duty Registrar to request for an expedited hearing, which will be fixed within 6 days of the application being filed. 3. The applicant will then have to serve documents on the Minister by the next day, following which the Minister will have to file his response within 3 days.

IV. Costs and time-frame of judicial process

Thus, the Law Minister has assured that “a person aggrieved by a Direction will have the opportunity to have his or her case heard in the High Court as early as nine days after he initiates a challenge by writing to the Minister”.89 Despite the concern of lawyers’ fees, the expedited appeals process significantly reduces the burden on the aggrieved person to bring a case to court.

An appeal under s. 17(5)

Judicial review

The Law Minister has stated that an applicant would have “three days free in Court”86 – with hearing costs not borne by the applicant. The

If, however, judicial review is mounted, the costs to the applicant would be substantially higher. Before any hearings begin, there are the costs of applying for leave to make an application;90 and subsequently are court costs, and costs of counsel. If the application fails, the applicant will also have to bear the costs of the respondent, unless “a serious question of constitutional law is raised, [where] the court may in its discretion depart from the usual rule

79 ibid. 80 As it did in Tan Seet Eng v Attorney-General [2015] SGCA 59; [2016] 1 SLR 779. 81 Tan (n 70). 82 ibid., at 817. 83 See, for instance, Anwar Siraj and another v AttorneyGeneral [2010] SGHC 36. 84 See Metal Industry Employees Union v Registrar of Trade Unions [1976] 1 MLJ 220, FC (Malaysia). This was also mentioned by Pillai MP in Parliament, see POFMB Debates (n 11). 85 Presumably under the “illegality” or “irrationality” head of judicial review. 86 POFMB Debates (n 11).

87 ibid. 88 ibid. 89 ibid. 90 Rules of Court (Cap 322, R 5, 2006 Rev Ed), O 53.

245

COMMENTARY

that costs follow the event”.91 V. Conclusion The s. 17(5) appeal process and judicial review jointly “ensure due process and the proper exercise of power, and give assurance to the public of the integrity of the decision-making process”92 in two distinct ways. The first is by ensuring that innocent statements, once impugned, will be vindicated in a court of law. The second is by ensuring that Ministers exercise prudence in the exercise of their Part 3 discretions as there is a “sword hanging over” the Minister, whose “personal reputation… [is] at stake every single time a decision is made.”93 These checks are strengthened by the expedited and subsidised appeal process under s. 17(5), which entails greater scrutiny of executive action, and lowered barriers of access to the courts. Seen in toto with judicial review – which, despite its lower level of scrutiny, provides a constitutionally and institutionally appropriate check on the Minister’s discretion – the judicial oversight mechanisms under Part 3 collectively seem to be adequate for purpose. There, of course, remain uncertainties over what will be regarded as a statement of fact; and how evaluative statements and academic research will be treated by the courts. Indeed, it is not difficult to see how POFMA can be applied to domestic historiography, and, in particular, revisionist accounts of sensitive issues such as Operation Spectrum and Operation Cold Store.94 A future government may not be reticent in using POFMA for political ends. Nonetheless, although these uncertainties mean that much depends on the “common sense”95 of judges, a casuistic approach by an independent judiciary will arguably be adequate to prevent an abuse of power by the Minister. Judicial mistakes may be made, but, in a system of

246

91 [94] 92 93 94 95

Wong Souk Yee v Attorney-General [2019] SGCA 25 at Select Committee Report (n 2). POFMB Debates (n 11), per the Law Minister. For details of these events, see Loh (n 60). Chen Cheng (n 19) at [33].

precedent and hierarchy, they will be eventually resolved as long as the judiciary is not complicit in executive abuse of power. Furthermore, these uncertainties work both ways. The same uncertainties mean that a Minister, in order to avoid “public embarrassment,”96 will be discouraged from using a Part 3 Direction for improper means, if he or she is unsure if such a direction will not be upheld by the court. Thus, notwithstanding the potential for abuse under Part 3, the judicial oversight mechanisms are arguably adequate in playing a “crucial balancing role” in ensuring the “propriety of the Executive’s exercise of discretion.”97 On the other hand, adequacy in ensuring that freedom of speech is not unduly “chilled,” is a function of uncertainty in the law – and in particular, whether legal and legitimate statements will be self-censored out of fear of falling afoul of POFMA.98 Due to the uncertainties inherent in s. 17(5)(b), there will potentially be some “chilling effect” on free speech. This, however, should not distract from the clarity that an objective test will bring to most cases. It should also be noted that the “chilling effect” under the present government is further tempered by reassurances given by the Education Minister that academic research will not be subject to the Act; and by examples of fact and opinion by the Law Minister. These help clarify how Part 3 discretions will be exercised, and so provide a “calibration in the powers deployed”99 that arguably adequately ensures that freedom of speech is not unduly “chilled”. The word “adequate” is, of course, an evaluative one. Its meaning can range from “barely reaching an acceptable standard,” to “quite sufficient,” to “[f]ully satisfying what is required.”100 On 96 POFMB Debates (n 11), per the Law Minister. 97 Select Committee Report (n 2) at [364]. 98 ibid., at [423]. See also New York Times Co v Sullivan 376 U.S. 254 (1964) in the US Supreme Court. 99 ibid., at [424]. 100 The word “adequate” is, of course, an evaluative one. Its meaning can range from “barely reaching an acceptable standard,” to “quite sufficient,” to “[f]ully satisfying what is

COMMENTARY

the facts presented above, people will express different views on whether the provisions for judicial oversight are adequate or not. It is hoped that this Act, in its attempt to safeguard democracy, will allow people to continue to do so.

Appendix A – the libellous paragraphs in Oversea-Chinese Banking Corp Ltd v Wright Norman [1994] SGHC 174; [1994] 3 SLR(R) 410, with the statement considered emphasised in bold “A case of how not to go about attracting top executives was demonstrated in the front page article which appeared in your November 17 issue regarding OCBC’s reported plans to hire four, named, senior banking officials. Professional executive search firms would recoil in horror at such publicity. Your reporter has met his professional duty to ‘find and report the news’, but there is a prima facie case of rank amateurism or carelessness at OCBC to have allowed such a breach of confidentiality to happen. I doubt that a professional executive search firm would have been involved. It shows clearly how difficult it is for an organization to maintain confidentiality when it conducts its own search. The leakage is potentially damaging to those named, at the very least embarrassing, and the effect on the morale of their present staff and, indeed, on that of the staff at OCBC, must be unfortunate.

required.” On the facts presented above, people will express different views on whether the provisions for judicial oversight are adequate or not. It is hoped that this Act, in its attempt to safeguard democracy, will allow people to continue to do so.

247

COMMENTARY

A Call for Judicial Activism in response to current Anti-Terror Legislation. U.Sudharshanraj Naidu, University of Bristol The threat of terrorism is the primary national security concern for most if not all states. The notion of terrorism entails with it, ‘the fear of the unknown’ which accounts for the rather drastic measures adopted by states.1 This mainly includes the pre-emptive detention of terror suspects. This essay will begin by exploring the modern threat of terrorism vis-a-vis the need for scrutiny, namely judicial scrutiny, of preemptive measures adopted in response to terrorism. Following this, an analysis of the law in Singapore and the UK in relation to the powers circumscribed by the respective counter-terrorism legislations. The level of judicial review of the same will also be tackled. Finally, the essay concludes with the call for legal reform in Singapore considering the vital role of the judiciary with the rationale of preventing arbitrary decisions by the executive. The modern threat of terrorism Terrorism, post-9/11, has become a borderless and unpredictable threat to the national security of states.2 The indiscriminate malice of terrorist acts feeds on the fears of states and prompts an interventionist approach to be adopted.3 Furthermore, there is an instinctive response to empower and strengthen the government with regards to these terror threats.4 This empowerment of the executive often leads to

248

1 Imran Awan, “The erosion of civil liberties: precharge detention and counter-terror laws” (2011) Pol. J., 84(3), 272-284, at p276. 2 Alan M. Dershowitz, Why Terrorism Works: Understanding the Threat, Responding to the Challenge (Yale University Press 2003), p35. 3 ibid. 4 Eunice Chua, “Reactions to Indefinite Preventive Detention: An Analysis of how the Singapore, United Kingdom and American Judiciary Give Voice to the Law in the Face of (Counter) Terrorism” (2007) 25 Sing. L. Rev. 3, 3-23, at p4.

the infringement of civil liberties. The alarming need for potential safeguards is instrumental considering the ‘Singapore Terrorism Threat Assessment Report 2019’ published by the Ministry of Home Affairs that detailed the increasing prevalence of radicalised individuals.5 Much emphasis was placed on the radicalisation of foreign workers persisting in particular.6 This was evidenced in the report by the 27 Bangladeshi workers and 5 Indonesian domestic workers that have been detained under the Internal Security Act 19887 between 2016 and 2019 due to affiliations with terrorist organisations.8 The protection awarded to these foreigners in terms of legal representation remains elusive and illustrates the evident lack of protection awarded to minorities. The majoritarian rule of elected governments often fails to consider the interests of minorities that may be marginalised and lacks safeguards against arbitrary decisionmaking.9 Hence, the courts have an obligation to exercise a counter-majoritarian role in response that ensures the protection of minorities guided by the substantive principles of the rule of law.10

5 <https://www.mha.gov.sg/newsroom/others/ news/singapore-terrorism-threat-assessment-report-2019> accessed 20 January 2019. 6 Hariz Baharudin, “Singapore still faces strong threat from terrorism: MHA”, the Straits Times (23 January 2019). 7 Internal Security Act (Cap. 143, 1985 Rev. Ed. Sing.) [ISA 1988]. 8 Danson Cheong, “Shanmugam: Terror threat real, Singapore must stay alert” the Straits Times (5 May 2016); Feliz Solomon, “Domestic Helpers in Singapore Have Been Investigated for Suspected Islamic Radicalism” Time (19 December 2016). 9 Stephen Sedley, “Judicial Politics” (2012) 34(4) London Review of Books 15. 10 Stephen Sedley, The Making and Remaking of the British Constitution (London, 1997), p. 25.

COMMENTARY

The law on Preventive Detention Singapore The power to order detention is found under S8 of the ISA 1988,11 which gives the Minister the authority to preventively detain a person “acting in any manner prejudicial to the security of Singapore… or to the maintenance of public order or essential services therein” on the prerequisite “satisfaction” of the President.12 The period of detention can also be indefinite, “for as long as the grounds for detention continue to exist”.13 This authority vested on the executive was challenged in Chng Suan Tze v Minister of Home Affairs.14 The Court of Appeal in Chng’s case, in obiter, advocated for an objective test when reviewing the exercise of executive discretion.15 This objective test applied by the courts, permits them to examine whether the decision maker’s satisfaction for discharging a detention order was based on objective facts.16 Furthermore, the subjective test granting the executive a high margin of appreciation to decide the applicability to detain a person was denied by the courts. This was because “all power has legal limits and the rule of law demands that the courts should be able to examine the exercise of discretionary power.”17 Yet, this was legislatively overruled by amendments to the ISA 1988 restricting interpretation and judicial review reflected in S8A and S8B respectively.18 The powers to review the detention under judicial review are now subject to strict limitations – only under procedural grounds. This legislative intervention had been adopted by the courts and was affirmed by the decision in Teo Soh Lung v Minister of Home Affairs.19 As the law 11 Internal Security Act (Cap. 143, 1985 Rev. Ed. Sing.) [ISA 1988]. 12 ibid, s8. 13 ibid. 14 Chng Suan Tze v. Minister of Home Affairs [1988] S.L.R. 132 [Chng]. 15 ibid at [83]. 16 Chua (n 4), p12. 17 Chng (n 14) at [86]. 18 ISA 1988, s8A & s8B. 19 Teo Soh Lung v Minsiter for Home Affairs [1989] 1 S.L.R.(R.) 461, H.C [Teo].

currently stands, the passive approach of the Singaporean Court towards judicial review of executive decisions is afforded by a subjective test instead of an objective test. This has awarded an indiscriminate amount of discretion to the executive with regards to detention under the ISA 1988. The court’s positivist approach through adoption of the legislative reform could perhaps be attributed to the Singaporean judiciary’s abidance of the doctrine of separation of powers.20 The basis of separation of powers relies on a presumption of constitutionality.21 The courts presume state officials decide policies in the best interests of the state and act in accordance of legality.22 However, non-justiciability on grounds of a lack of institutional capacity by the courts is arguably overstated.23 This is especially so when the protection of rights at stake are for minorities devoid of safeguards by the political process, due to the disproportionate response to the threat of terrorism. For despite judges not having an elected mandate, “a democracy is more than a state in which power resides in the hands of a majority of elected representatives: it is a state in which individuals and minorities have an assurance of certain basic protections from the majoritarian interest.”24 Hence, the courts should adopt a wider formulation of review that extends beyond the presumption of constitutionality to principles of legality.25 As Thio purports, a strong commitment to the principle of legality is important to ensure that the Executive does not “effectively become a power unto itself.”26 Hence, this would ensure 20 Jothie Rajah, The Authoritarian Rule of Law: Legislation, Discourse and Legitimacy in Singapore (Rev ed., Cambridge University Press 2012), p212. 21 Thio Li-ann, A Treatise on Singapore Constitutional Law (Academy Publishing, 2012) a p539. 22 ibid. 23 Trevor Allan, ‘Human Rights and Judicial Review: A critique of ‘Due Deference’’, C.L.J. 2006, 65(3), p671-695, at p673. 24 Stephen Sedley, The Making and Remaking of the British Constitution (London,1997), p. 25. 25 Jaclyn L Neo, ‘All Power has Legal Limits’, 29 SAcLJ 2017, p667- 689, p687. 26 Thio Li-ann, “The Constitutional Framework of Powers” in The Singapore Legal System (Kevin Tan ed) (NUS Press, 1999) at

249

COMMENTARY

the response to the very real threat of terrorism that thrives on forwarding the majoritarian agenda is not at the expense of the civil and human rights of minorities. UK In contrast, the judiciary in England has played a slightly more active role in holding the Executive accountable for the wide powers entailed by counter-terrorism legislation. The UK’s preventive detention laws were conferred under S21 and S23 of the AntiTerrorism, Crime and Security Act 2001.27 This had awarded the executive wide powers to detain persons, particularly non-citizens, on grounds of a “reasonable belief… and reasonable suspicion…”28 This led to the infringement of the civil liberties of various non-nationals who were detained without any supervision of the exercise of the powers entailed by the legislation. However, the House of Lords in the Belmarsh case,29 empowered by the HRA 1998,30 declared the provisions of the ATCSA 2001 incompatible with the European Convention of Human Rights (EHCR), under S4 of the HRA 1998.31 The HOL’s decision rested mainly on the concept of ‘proportionality’. Under the doctrine of proportionality any measure or action restricting the rights under the ECHR should be suitable and necessary to achieve an envisaged legitimate aim. Additionally, the “means to achieve the aim”32 and the intrusiveness of the means was analysed. It was deemed to be a disproportionate response to the exigencies of the situation in the Belmarsh case for it unfairly discriminated against nonnational terror suspects. The issuing of a declaration of incompatibility

250

p94. 27 Anti-Terrorism, Crime and Security Act 2001 (U.K.), 2001, c.24 [‘ATCSA 2001’]. 28 ibid. s21. 29 A and ors, v, Secretary of State for the Home Department [2005] 3 All. E.R. 169 (H.L.) [Belmarsh]. 30 Human Rights Act 1998 c.24 [HRA 1998]. 31 HRA 1998, s4. 32 Belmarsh (n 29), at 55 (per Bingham LJ).

conferred upon the judiciary by the adoption of the HRA 1998 is not binding on the legislature but it had the effect of placing considerable political pressure on the legislature to make amendments on the ATCSA 2001. Also, the judgement did maintain deference in the executive’s authority to decide whether there is a “public emergency threatening the life of the nation” which still awards the executive copious amounts of discretionary power.33 However, the UK courts had been able to ensure the protection of minority interests by prompting legislative review of counter-terrorism laws with regards to the preventive detention of terror suspects through creating a “democratic dialogue” between the courts and other branches of the state.34 This ensures there are safeguards to question the executive’s decisions and places pressure on the branches of the state preventing executive tyranny while adhering to the legislative supremacy within the UK. The role of the judiciary The presence of apparent safeguards in the context of the UK distinguishes it from Singapore with regards to the response to counter-terrorism measures. Though, from a comparative analysis of the executive’s response to the threat of terrorism, the UK adopts a lower threshold requirement for the detention of individuals by merely requiring a “reasonable belief or suspicion” whereas the Singaporean executive held the requirement of detention being a “necessary” response.35 However, the criticism for Singapore is the lack of safeguards present in response to these pre-emptive measures by the executive. This is especially so considering the judiciary’s commitment to ensure compliance with the law, a fair hearing process and rationality in decision-making through judicial review.36 Thus, it is suggested 33 ibid. 34 Stephen Gardbaum, ‘Reassessing the new commonwealth model of constitutionalism’ (2010) 8 IJCL, 167-206, at p179. 35 Chua (n 4), p8. 36 Chan Sek Keong, ‘Judicial Review – From Angst to Empathy’ (2010) 22 SAcLJ, 469-486, p474.

COMMENTARY

that more can be done by the judiciary in terms of judicial review regarding detainees under the ISA 1988 to reaffirm their role as “guardians of constitutional propriety”.37 The appropriate response would be one where the judiciary ensure that both formal democracy (in terms of separation of powers) and substantive democracy (in terms of protection of civil liberties) are given equal importance in judicial decision-making.38 In countering terrorism, the judiciary’s role remains relevant and important – “amid the clash of arms, the laws are not silent”39 – it is up to the judge to take up arms in the defence of the law. This would avoid the situation of anointing the executive with absolute power. However, considering legislative limitations posed by S8 ISA 1988, the courts may need to adopt an interventionist approach on constitutional grounds relying on the presence of a written constitution. The clear and unambiguous language that prevents judicial review on an objective test that had been adopted in Teo should be corrected for it fails to achieve the pre-requisite element of substantive democracy. The Singaporean judiciary are greatly empowered and have legitimate standing to void legislation.40 This is found in Article 4 of the constitution which declares “This Constitution is the supreme law of the Republic of Singapore and any law enacted by the Legislature after the commencement of this Constitution which is inconsistent with this Constitution shall, to the extent of the inconsistency, be void”.41 The breaches imposed on civil liberties through the preventive detention laws under the ISA 1988 could be challenged by affording a more 37 Jeffrey Jowell, “Parliamentary Sovereignty under the new constitutional hypothesis” P.L. 2006, Oct, 562-580, at p 572. 38 Aahron Barak, “Foreword: A judge on Judging: The Role of a Supreme Court in a Democracy” (2002) 116 Harv. L. Rev. 19, at p1207. 39 Liversidge v Anderson [1942] A.C. 206 (H.L.) at p.244 (Atkin LJ). 40 Chua (n 4), p20. 41 Constitution of the Republic of Singapore (1999 Rev. Ed.), art. 151.

liberal interpretation towards Article 151 of the constitution, which provides for restrictions on preventive detention. As detailed, Article 151 requires detainees to be informed of the grounds of detention and there needs to be fair representation to hear one’s case.42 If these procedures are not met, the court could deem such restrictions on judicial review to be unconstitutional. Hence, the judiciary can afford to utilise this constitutional article more liberally insofar as to prevent a breach of civil liberties. Furthermore, having overcome the possible restrictions imposed by the statute through constitutional review, judicial review on the principle of legality can be adopted by the courts as a possible extension to the available grounds of review. This extension of review would also not be contrary to the current trajectory of review adopted by the courts. This is evidenced in Yong Vui Kong v Attorney General,43 where then Chief Justice Chan Sek Keong affirmed that the principles enunciated in Chng were fully alive.44 Additionally, one such case which affirmed the principle of legality is Tan Seet Eng v AG.45 Tan is significant as it mirrors Chng in that the impugned power also entailed preventive detention (albeit under a different statute).46 Tan involved a detention order under the Criminal Law (Temporary Provisions) Act 200047 (CLTPA). S30 of the CLTPA 2000 empowers the Minister for Home Affairs to preventively detain a person who has been associated with activities of a criminal nature if the Minister considers it “necessary in the interests of public safety, peace and good order”. The Court of Appeal in ordering the detainee’s release, in obiter, asserted its authority to enquire into areas of ministerial discretion through an objective

42 ibid. 43 Yong Vui Kong v Attorney General [2011] 2 SLR 1189 [Yong]. 44 ibid, at [79] 45 Tan Seet Eng v AG [2016] 1 SLR 779 [Tan]. 46 Neo (n 25), p681. 47 Criminal Law (Temporary Provisions) Act 2000, Cap 67, Rev Ed. [CLTPA 2000].

251

COMMENTARY

enquiry.48 This supports the judiciary’s adoption of responsibility as a bulwark against executive tyranny.49 Hence, having hinted the possible adoption of review guided by the principles of legality in preventing arbitrary decisions by the Executive, the courts can act as a safeguard against pre-emptive detention cases under the ISA 1988, once having deemed S8 ISA 1988 unconstitutional.

values, comes not from terrorism but from laws such as these. That is the true measure of what terrorism may achieve. It is for Parliament to decide whether to give the terrorists such a victory.”51

Conclusion In conclusion, mounting fears posed by terrorism are apparent in Singapore with ever increasing calls by the state to stay vigilant. This is expounded by the introduction of ‘SGSECURE’, an online platform that empowers citizens to be wary of terror threats and to report them immediately, which steer the possible response to counter-terrorism in a positive direction.50 However, the draconian response adopted by state actors in response through the preventive detention of suspected terrorists requires an appropriate safeguard that ensures the protection of civil liberties. The judiciary in their capacity as a ‘countermajoritarian’ entity need to stand to prevent executive tyranny. This issue is can be resolved by a two-fold solution. Firstly, declaring the restraints imposed by S8 ISA 1988 as unconstitutional allows judicial review claims to be brought to court by applicants subject to preventive detention under the ISA 1988. Then, extending such grounds for judicial review thereby allows for an objective test to apply. This prevents arbitrary decision-making. At the same time, there is a need for some reflection by the executive encompassed in the powerful dissent by Lord Hoffman in the Belmarsh case that “The real threat to the life of the nation, in the sense of a people living in accordance with its traditional laws and political

252

48 CLTPA 2000, s30(1)(a). 49 Christopher Forsyth and Mark Elliot, ‘The Legitimacy of Judicial Review’, P.L. 2003, Sum, p286-307, p293. 50 <www.sgsecure.sg/> accessed 18 March 2019.

Belmarsh (n 29), at [97] (per Hoffman LJ).

COMMENTARY

Rage Against the Machines: AI or Nay? Annabelle Lee, University of Warwick Introduction Early last year, we welcomed a new addition to the family – our first ever Google Home Mini, a smart speaker with a built-in virtual artificial intelligence (hereinafter AI) assistant.1 While the “OK, Google(s)” which were responded to by an inanimate voice were a source of fresh amusement, a gradual moment of epiphany dawned upon me – that we were one step closer to living in the then seemingly distant world of I, Robot.2 Automation technology is dominating our lives in the twenty-first century of an information society and digital age, and we are without a doubt embracing it. My Google Home Mini plays music; runs the household smoothly; tells me whether I should bring along an umbrella before I leave the house. Although there are a great number of definitions and takes of tech industry players on what AI is, they all share a prevalent idea: that its main objective is to build systems that perform tasks normally done by intelligent beings.3 Even though the idea of a machine functioning as a human may seem daunting, AI has supported our daily lives tremendously by increasing our efficiency in completing the most mundane tasks. What then, does the future hold for AI in our professional lives?

AI technology offers the possibility of automating tasks traditionally done by lawyers. This will increase the productivity and costeffectiveness of legal firms, allowing lawyers to shift their focus on higher value-added work such as advising clients, negotiating deals, and appearing in court.4 For example, Slaughter and May in the United Kingdom (UK) adopted Luminance, a document analysis software used to sort and rank documents in its data room for due diligence of its merger and acquisition transactions. This led to halving the time of document organisation done by its employees.5 Likewise, WongPartnership became the frontrunner in Singapore’s legal industry to adopt AI technology with Luminance.6 Notwithstanding these benefits, the novelty of AI technology may also impose a burden on both lawyers and clients alike to adjust and adapt to in the face of time and resource constraints. As such, the Singapore Academy of Law has rolled out initiatives such as the Future Law Innovation Programme (FLIP) – to assist industry players in streamlining integration of new technology into their processes, through enabling the exchange of ideas between legal and tech sectors.7 Denton’s NextLaw Labs and Allen & Overy’s Fuse in the UK serve similar functions.8 In addition, AI technology brings

1 ‘Google Home Mini’ (Google LLC) <https://store. google.com/us/product/google_home_mini?hl=en-US> accessed 29 March 2019 2 (2004). A science-fiction action film which depicted the future set in 2035 where humankind was heavily reliant on robots in their daily lives. Through a murder investigation, a police detective discovered human-like traits and emotions in the robots. The robots and its central intelligent system turned out to be the offenders, who were planning to enslave and control humanity. 3 Bernard Marr, ‘The Key Definitions of Artificial Intelligence (AI) That Explain Its Importance’ (Forbes, 14 February 2018) <https://www.forbes.com/sites/ bernardmarr/2018/02/14/the-key-definitions-of-artificialintelligence-ai-that-explain-its-importance/#8be57874f5d8> accessed 29 March 2019

4 ‘How Law Firms Can Benefit from Artificial Intelligence’ (SwissCognitive, 17 November 2018) <https:// swisscognitive.ch/2018/11/17/how-law-firms-can-benefitfrom-artificial-intelligence/> accessed 29 March 2019 5 Andrew Burgess, The Executive Guide to Artificial Intelligence: How to Identify and Implement Applications for AI in your Organisation (Palgrave Macmillan, 2018) 82 6 Jason Thomas, ‘AI and big data in the legal profession’ (The Asean Post, 14 March 2019) <https://theaseanpost.com/ article/ai-and-big-data-legal-profession> accessed 29 March 2019 7 Peck Gek Tay, ‘Legaltech and the law: How tech will change the practice’ Business Times (Singapore, 4 March 2019) <https://www.singaporelawwatch.sg/Headlines/legaltechand-the-law-how-tech-will-change-the-practice> accessed 29 March 2019 8 Alex Heshmaty, ‘Legal tech in 2018: threats and opportunities’ (The Law Society, 13 June 2018) <https://www.

253

COMMENTARY

about substantive benefits that better reflect the current digital age, allowing legal services to be more relevant to society. For example, in the area of public legal education, a group of University of Cambridge students created LawBot, a chat system using AI Markup Language. This tool was created with the aim of providing free and easy-to-understand legal advice to the general public on criminal offences, assisting them in deciding which legal actions to pursue. They hope to increase access to justice by compelling more people to seek qualified legal advice, and have since globally expanded to USA, Canada, Singapore, Hong Kong, Australia and New Zealand.9 However, associated challenges can be identified.

laid out the key principle that a work cannot be capable of copyright protection without an author as his identity is essential in determining who is conferred the legal right, and that only a natural person can qualify as one.12 In the UK, on the other hand, section 9(3) and 178 of the Copyright Designs and Patents Act (CDPA) 1988 defines a “computer-generated” protected work as one without a human author, and recognises the author of these works to be a person “by whom the arrangements necessary for the creation of the work are undertaken”.13 The UK’s approach thus poses a lower threshold in order for AI-generated works to be copyrightprotected, allowing for programmers to have legal ownership rights over the work produced rather than have no rightful owner at all.

1. Title and Accountability With any great innovation comes title and responsibility. Given its lack of human touch, AI technology poses a challenge in the field of intellectual property law with regards to the recognition of ownership rights. Consequently, litigating the intellectual property protection of AI-generated works becomes complex. As parties to the World Trade Organisation (WTO), both Singapore and the UK have implemented copyright laws compliant with the minimum standards set out in the Trade Related Aspects of Intellectual Property Rights (TRIPS) Agreement. However, UK legislation appears superior in being better equipped to protect ownership rights of AI works, perhaps due to its roots as an early adopter of the technology. Under section 7A of Singapore’s Copyright Act 1987, AI technology can constitute a “computer programme”, which is capable of being protected as a piece of literary work.10 Still, case law has shown copyright does not subsist without a human author. The Court of Appeal’s decision in Asia Pacific Publishing Pte Ltd v Pioneers & Leaders (Publishers) Pte Ltd11

254

lawsociety.org.uk/news/blog/legal-tech-2018-threats-andopportunities/> accessed 29 March 2019 9 The Law Society, ‘Horizon Scanning: Artificial Intelligence and the Legal Profession’ (2018) 1, 7. 10 Copyright Act 1987, s7A. 11 [2011] SGCA 37.

Things becomes even trickier in more sensitive areas of society such as in the law of torts and criminal liability. The notion of accountability in the assistance provided by AI technology in our daily lives will be an area of contention as to who – the programmer or operator – will be liable for when something goes wrong. As a new area of technology, there still exists a grey area in whether someone can sue a non-living being without legal personality for negligence. Notably, in recent years, transportation systems have embodied AI technology with autonomous self-driving vehicles. The first pedestrian fatality caused by a self-driving Uber test car in Tempe, Arizona, in the United States was concluded with Uber reaching a settlement with the victim’s family.14 While there has not been any legal precedent on AI-generated negligence yet, Smith submits that litigating automated driving accidents representsv a shift from proving vehicular negligence of which driver and insurer normally hold primary liability to that of product liability, which holds the manufacturers providing the defective vehicles 12 ibid, para 41 and 67. 13 Copyright Designs and Patent Act 1988, s9(3) and s178. 14 Ian Bogost, ‘Can you sue a Robocar?’ (The Atlantic, 20 March 2018) <https://www.theatlantic.com/technology/ archive/2018/03/can-you-sue-a-robocar/556007/> accessed 30 March 2019

COMMENTARY

responsible if the harm was foreseeable.15 The UK currently appears to take a similar stand, with the AI user being less likely to be at fault than the manufacturer.16 On the other hand, prosecutors found no basis for criminal liability in Uber. Nevertheless, the safety backup driver who was present in the vehicle may be faced with charges of vehicle manslaughter for her involvement in the accident, as she could have prevented the crash had she been paying attention to the road.17 While the limitation of a civil lawsuit may possibly set precedent for litigating autonomous vehicle accidents in the future, technological infancy still remains a hurdle as there will come a time when there is no need for drivers behind the wheel to whom we can pin the blame on. As technology develops and AI systems become more robust and capable of openended learning, there may emerge more functions that a manufacturer cannot foresee, posing a challenge to traditional approaches of manufactural product liability.18 Having said that, both the UK and Singapore have shown the desire to be at the forefront of autonomous vehicle technology, showing initiatives to regulate specifically in these areas and rightly so. These include new legislative proposals for autonomous car insurance laws in the UK,19 and amending the Road Traffic Act of Singapore to include a regulatory framework for recognition of autonomous vehicle trials in addition to a 15 Bryant Smith, ‘Automated Driving and Product Liability’ (2017) 1 Michigan State Law Review 1, 45 (forthcoming) 16 Lee Gluyas and Stefanie Day, ‘Artificial Intelligence – Who is Liable when AI Fails to Perform?’ (CMS Cameron McKenna Nabarro) <https://cms.law/en/GBR/Publication/ Artificial-Intelligence-Who-is-liable-when-AI-fails-toperform> accessed 30 March 2019 17 Kristin Houser, ‘Prosecutors: Uber Isn’t Criminally Liable for Self-Driving Car Death’ (Futurism, 7 March 2019) <https://futurism.com/uber-not-criminally-liable-self-drivingcar-death> accessed 22 June 2019 18 Peter Asaro, ‘The Liability Problem for Autonomous Artificial Agents’ (Ethical and Moral Considerations in NonHuman Agents, 2016 AAAI Spring Symposium Series, 2016) 19 Matt Burgess, ‘WIRED’s need-to-know guide to driverless car testing in the UK’ (WIRED, 30 August 2018) <https://www.wired.co.uk/article/driverless-cars-uk-selfdriving-cars> accessed 30 March 2019

new legislative bill proposal in the latter half of 2018.20 2. Data Analysis and Bias Further, the growth of AI technology would mean the mass collection and production of data that is fed into these systems for them to operate. For instance, law enforcement authorities are increasingly using AI technology in the investigation of crime. Police of Kent in the UK were the first adopters of PredPol, a machine-learning algorithm that predicts potential crime hotspots with the aim of making allocation of crime control resources more effective.21 However, because the technology bases its predictions on old records in police databases, past human biases contained in those decisions are reinforced in the algorithm, exacerbating racial, gender, and geographical discrimination. The Royal Statistical Society, a Human Rights Data Analysis group, conducted a study which found that the software led to police unfairly targeting lower-income or racial minority-filled neighbourhoods.22 This pattern repeats itself in private sectors such as AI technology used in automating recruitment of employees like Amazon’s experimental hiring tool, which favoured male over female job applicants in the tech industry – and fortunately has been abandoned.23 The risk of bias from services using AI programming presents a potential infringement of its data subjects and their fundamental human right against 20 Kevin Kwang, ‘Singapore relooking road rules to allow for self-driving vehicles’ (Channel NewsAsia, 6 June 2018) <https://www.channelnewsasia.com/news/singapore/selfdriving-cars-road-rules-singapore-mot-10397586> accessed 31 March 2019 21 Daniel Cossins, ‘Discriminating algorithms: 5 times AI showed prejudice’ (NewScientist, 12 April 2018) <https:// www.newscientist.com/article/2166207-discriminatingalgorithms-5-times-ai-showed-prejudice/> accessed 31 March 2019 22 Kristian Lum and William Isaac, ‘To predict and serve?’ (2016) 13(5) Significance 14, 15 23 Jeffrey Dastin, ‘Amazon scraps secret AI recruiting tool that showed bias against women’ (Thomson Reuters, 10 October 2018) <https://www.reuters.com/article/ us-amazon-com-jobs-automation-insight/amazon-scrapssecret-ai-recruiting-tool-that-showed-bias-against-womenidUSKCN1MK08G> accessed 31 March 2019

255

COMMENTARY

discrimination. 3. Developments in Regulation: Ethics and Data Protection Even so, I believe that with sufficient measures in place, AI can be worked to its advantage. Early this year, a model AI governance framework pertaining to the ethical use of AI in private businesses was published by Singapore’s Infocomm Media Development Authority at the World Economic Forum meeting. Its key governing principles are that AI use remains explainable, fair and transparent to consumers, and solutions human-centric.24 Between the laissez-faire approach in the US and the precautionary stance in the EU and UK, Singapore seeks to strike a balance in its own response to the global discourse by designing an environment ripe for innovation but also risk-aware in dealing with ethical and governance issues that arise when deploying AI technologies. Namely, the Model Framework enforces a governance structure with clear roles and responsibilities assigned to individuals in the chain of processes in AI use, a persisting human role in decision-making processes, and responsible data management.25 Similarly, the UK government set up the Centre for Data Ethics and Innovation (CDEI) in 2017, an advisory body that promotes best practice of the ethical and innovative use of technology and address gaps in the regulatory landscape of data-enabled technologies.26 Although the recency of these initiatives has yet to see the fruits of its labour, the mere fact that authorities are taking a step in ensuring responsible AI use

256

24 Personal Data Protection Commission Singapore, A Proposed Model Artificial Intelligence Governance Framework (Infocomm Media Development Authority, 2019) para 2.5 25 Info-communications Media Development Authority, ‘Singapore’s Governing Framework for Artificial Intelligence’ (The New York Times, 18 June 2019) <https://www.nytimes. com/paidpost/imda/singapores-governing-framework-forartificial-intelligence.html?smid=fb-share&fbclid=IwAR35K4 vwLJMHyjXqaOrij1eHEf0aaNXvlCjAEkTp7Jq0t6tGkmhzdG Ufylc> accessed 22 June 2019 26 ‘Centre for Data Ethics and Innovation to shape UK AI industry’ (Government Europa, 22 November 2018) <https://www.governmenteuropa.eu/centre-for-data-ethicsand-innovation/91232/> accessed 31 March 2019

is comforting in the face of the industry’s future challenges. Additionally, data protection laws help safeguard collection of data against misuse in data-enabled AI technologies. Singapore’s Personal Data Protection Act 2012 is one such legislation which is enforced by the Personal Data Protection Commission authority. It aims to regulate the flow of personal data among organisations, taking into account factors like the data subject’s consent, and the data processor’s duty to inform the data subject of the purpose of data collection, and reasonableness of personal data use.27 The Data Protection Act 2018 of the UK and Information Commissioner’s Office share comparable principles.28 The Act falls under the implementation of the European Union’s General Data Protection Regulation (GDPR), a comprehensive piece of legislation esteemed for its high standards of personal data protection. Further, the UK is party to the Charter of Fundamental Rights of the European Union, which has recognised the right to the protection of personal data in Article 8 as a fundamental right, in addition to the right to respect for private and family life in Article 7.29 Conversely, it remains unsettled as to the extent to which Singapore will prioritise the protection of its citizens’ personal data privacy over national interests, as illustrated in its lack of a right to privacy in its constitution, with only some laws regulating the use of personal data. Globally, the Republic has also failed to ratify the International Covenant on Civil and Political Rights, which includes the right against unlawful interference of privacy in Article 17.30 In comparison to the strict regime and outlook towards fundamental rights in the European Union, human rights protection evidently fares worse in Singapore at the expense of the data 27 Personal Data Protection Act 2012, s3 28 ‘Data Protection Act 2018’ (Information Commissioner’s Office) <https://ico.org.uk/for-organisations/data-protectionact-2018/> accessed 31 March 2019 29 Charter of Fundamental Rights of the European Union [2012] OJ C326/391 30 Privacy International, The Right to Privacy in Singapore (2015) paras 8-14

COMMENTARY

subjectâ&#x20AC;&#x2122;s privacy in AI use. 4. Conclusion As more interest is garnered towards AI and the beneficial value it has on society, more technological advancements will roll out as more research and development will be invested into it. Although the emergence of AI is being embraced in the legal profession to ease processes and tasks, it will not replace the craft of advocacy and ability to connect with real people. However, a host of related issues such as ownership, liability, and data discrimination will continue to persist as AI technology is still in its inception. With the advent of these times, it will now be an opportune time to utilise our resources and focus on developing frameworks to combat these potential issues, rather than fixate on the anxiety surrounding this transitory disruption brought about to the legal industry by AI technology.â&#x20AC;&#x192;

257

COMMENTARY: GUEST WRITER

Evaluating the constitutionality of Section 61, Protection from Online Falsehoods and Manipulation Act 2019 Suang Wijaya, Eugene Thuraisingam LLP Introduction On 25 June 2019, the Protection from Online Falsehoods and Manipulation Act 2019 (“POFMA”) was published on in the Government Gazette. This article focuses on Section 61 of the POFMA, which provides:The Minister may, by order in the Gazette, exempt any person or class of persons from any provision from this Act. What is the effect of Section 61? The Explanatory Statement does not explain the persons or classes of persons to whom Section 61 is meant to apply. The Explanatory Statement also does not explain the purposes that are intended to be achieved by Section 61. However, when we consider the other provisions of POFMA, we get a reasonably clear understanding of the nature of the Minister’s proposed powers under POFMA. The provisions of the POFMA can be divided into three broad categories:a. First, there are provisions conferring various powers on Ministers, and setting out the conditions for the exercise of those powers. For instance, Sections 11 to 13 confer on Ministers the power to issue “Part 3 Directions”, while Section 10 prescribes the conditions for the exercise of such power. b. Second, there are provisions prescribing the remedies are available to persons aggrieved by the exercise of powers by Ministers. There are also provisions setting out the procedures in order to 258

obtain such relief. Thus, Section 17 sets out the remedies and procedures available to persons who have been served with Part 3 Directions. c. Third, there are provisions which impose criminal liability (“offence-creating provisions”). The POFMA is littered with such provisions. These include Part 2, and Sections 15, 27, 28(3), 32(6), 33(4), 34(5), 36, 37(6), 38, 42, 43(3), 47(4), 50(3) and 54(4). Although on its terms the Minister has power to “exempt” persons from “any provision” of the proposed Act, Section 61 is unlikely to apply to the first category of provisions set out above. The ordinary meaning of “exempt” is to free a person from an obligation or liability imposed on other persons.1 If Section 61 were to apply to the first category of provisions, this would mean that a Minister can, by order in the Gazette, waive the conditions for the exercise of other Ministers’ or even his own exercise of their statutory powers. The outcome of such an interpretation would be manifestly absurd and arbitrary. I trust that the legislative drafters did not intend for Section 61 to have such an effect. However, Section 61 can conceivably apply to the second category of provisions. It is conceivable for the Minister to have the power to exempt persons from specific procedural requirements for the exercise of their rights to seek remedies against orders or directions made by Ministers against them. For instance, the Minister may exempt an aggrieved person from the requirement under Section 17(2) that, before exercising a right of appeal, there must first be an application to the Minister to vary or cancel a Part 3 Direction, and such application 1 https://en.oxforddictionaries.com/definition/ exempt

COMMENTARY: GUEST WRITER

must then be wholly or partially refused. Similarly, the drafters likely intended Section 61 to apply to the third category of provisions. I say this because Section 61 states that the Minister’s powers of exemption apply to “any provision”, and there are numerous offencecreating provisions in the POFMA. At this juncture, it is useful to elaborate on the legal effects of the Minister having Section 61 powers over offence-creating provisions. On its plain and ordinary meaning, Section 61 empowers the Minister to free a person from liability for an offence provided by the POFMA (a “POFMA offence”) where, but for the Minister’s decision, the person would be liable for such an offence. Section 61 can apply prospectively, in the sense that a person who has not committed a POFMA offence may, moving forward, lawfully do acts which would without the Minister’s intervention have amounted to such an offence. That is not all, however. The Minister can exercise the Section 61 power retrospectively, by ordering that a person who has already committed a POFMA offence, be exempt from liability for that offence. Taking things even further, the Minister may even exempt a person from liability for a past act, even where criminal proceedings are already afoot. In such a case, the Minister’s order exempting an accused person from criminal liability would mean that a court becomes bound to acquit the accused in question. It is my respectful view that, on its present terms, Section 61 is arguably unconstitutional. I take this view, even bearing in mind that the exercise of the Section 61 power will, in the usual way, be subject to judicial review. I elaborate.

Court of Appeal’s judgment in Prabagaran a/l Srivijayan v PP and other matters2 (“Prabagaran”) stated at [56]:It is undisputed that the Constitution, based on the Westminster model of constitutional government, incorporates as part of its basic structure the principle of separation of powers: Mohammad Faizal bin Sabtu v PP [2012] 4 SLR 947 (“Faizal”) at [11]. By way of Arts 23(1), 38 and 93 of the Constitution, the executive, legislative and judicial powers of Singapore vest in their respective organs of state… [T]he judicial power of Singapore vests exclusively in the Supreme Court and “such subordinate courts as may be provided by any written law for the time being in force.” This entails that the legislative and executive branches may not interfere with the exercise of the judicial power by the judicial branch (see Faizal at [19]), and such acts of interference may be struck down as unconstitutional. The more difficult task, however, is ascertaining whether a particular Act of Parliament has in fact interfered with the judicial power. The starting point is to recognise the distinction between, on the one hand, a valid legislative rule of general application, and on the other hand, purported legislation that attempts to achieve a particular outcome for a particular person or group of persons in relation to particular sets of facts. The distinction is vividly described by Blackstone (Commentaries, Introduction, Section II):[I]t is a rule: not a transient order from a superior to or concerning a particular person; but something permanent, uniform and universal. Therefore, a particular act of the legislature to confiscate the goods of Titius, or to attaint him of high treason, does not enter into the idea of a municipal law: for the operation of this act is spent upon Titius only, and has no relation to the community in relation to the community in general; it is rather a sentence than a law. But an act to declare that the crime of which Titius is accused shall be deemed high treason: this

The separation of powers and the judicial power under Article 93 of the Constitution It is now firmly established that the principle of separation of powers is a bedrock constitutional principle. As Chao Hick Tin JA delivering the

[2017] 1 SLR 173.

259

COMMENTARY GUEST WRITER

has permanency, uniformity and universality, and therefore is properly a rule. The above passage was cited by the Privy Council in Liyanage v The Queen,3 “the leading modern authority for the proposition that under a written constitution based on the separation of powers, the legislature may not determine by statute the outcome of particular judicial proceedings” (Ferguson v Attorney General of Trinidad and Tobago4 (“Ferguson”) at [21]). Chan Sek Keong CJ in Faizal at [27] summarises the essential judicial task, which is not to be intruded upon by the legislative or executive organs of State, as follows:In essence, the judicial function is premised on the existence of a controversy either between a State and one or more of its subjects, or between two or more subjects of a state. The judicial function entails the courts making a finding on the facts as they stand, applying the relevant law to those facts and determining the rights and obligations of the parties concerned for the purposes of governing their relationship for the future. More specifically in the field of criminal law, “[o] ne of the exclusively judicial functions of government is the adjugment… of criminal guilt” (Nicholas v R5 at [16] per Brennan CJ). The indicia of impermissible legislative interference in the judicial power In the recent Privy Council decision of Ferguson at [22], Lord Sumption delivering the Board’s judgment observed that legislation may interfere in the judicial power either directly or indirectly. Statutes which directly interfere in the judicial power include bills of attainder (ancient statutory instruments which convicted specific named individuals of offences), legislation that determines innocence or guilt or the penalty to be imposed, or legislation which vests on the

260

3 4 5

[1967] AC 259 at 291. [2016] UKPC 2. (1998) 193 CLR 173.

executive the power to determine innocence or guilt or the penalty to be imposed. Indirect legislative interference in the judicial power occurs by “altering general rules of law in a manner which will in practice determine the outcome of particular proceedings or of particular issues in those proceedings, for example by changing the elements of an offence or a tort, or abrogating a special defence, or altering the rules of evidence or a relevant period of limitation, without any transitional provisions to ensure that current proceedings are unaffected” (Ferguson at [24]). Indirect interference is more common than direct interference, and “gives rise to more difficult problems” (Ferguson at [24]). The key to identifying such indirect interference, according to Lord Sumption, is to recognise that the legislation “should not simply affect the resolution of current litigation but should be ad hominem, ie targeted at identifiable persons or cases” (Feguson at [25]). Unfortunately, there is not a clear bright line between legislation which merely alters the law or affects current legislation, and legislation which is ad hominem. Some general insights can however be gleaned from the cases. First, “Each case must be decided in the light of its own facts and circumstances, including the true purpose of the legislation, the situation to which it was directed, the existence (where several enactments are impugned) of a common design, and the extent to which the legislation affects, by way of direction or restriction, the discretion or judgment of the judiciary in specific proceedings” (Liyanage at 290). In weighing all the facts and circumstances, the court considers the substance and not the form of the legislative scheme. “Parliament cannot evade a constitutional restriction by a colourable device” (Hinds v The Queen6 (“Hinds”) at 290). In Liyanage itself, a number of legislative provisions were under challenge. Had each provision been considered alone, there would be no constitutional objection. The provisions purported to provide for general rules of procedure and evidence. However, when the provisions are considered 6

[1967] 1 AC 259.

COMMENTARY: GUEST WRITER

cumulatively, the ad hominem nature of the legislative scheme becomes apparent. By:- (i) changing the rules of admission of evidence; (ii) changing the minimum sentence; (iii) providing specific answers to specific issues in the ongoing proceedings; (iv) making these changes retrospective by applying them to ongoing judicial proceedings; (v) disapplying the legislative scheme from future proceedings by the use of a sunset clause (Ferguson at [26]), the legislation was clearly directed towards securing the conviction and heavy punishment of a targeted group of individuals (and no others). The Privy Council had no difficulties finding that this “constituted a grave and deliberate incursion into the judicial sphere” (Liyanage at 290).

Second, it is not enough to only show that the legislation indirectly affects the issue of liability or penalty in a case, or vests in a member of the Executive a power which, when exercised, indirectly affects the issue of liability or penalty in a case. The Legislature can legitimately provide for rules of general application that have the effect of restricting the range of decisions which the court can reach in a case (see Faizal). The Legislature can also validly prescribe conditions which circumscribe the range of decisions which the court can reach in a case, and then vest in a member of the Executive the power to determine whether those conditions are met (see Prabagaran). Applying the principles to Section 61

Ferguson is a useful contrast to Liyanage. In Ferguson, the legislation under challenge in essence repealed the limitation defence for criminal proceedings. Before the enactment of impugned legislation, this limitation defence had accrued in favour of the appellants in the Ferguson case. Some of the appellants had even filed applications for summary acquittal, on the basis of the accrued limitation defence. The impugned legislation was retrospective, such that:- (i) accused persons who had an accrued limitation defence were no longer able to rely on that defence; and (ii) those accused persons who had applied for summary acquittal necessarily had their applications dismissed. The Privy Council accepted that the legislation under challenge did impact ongoing proceeding. However, the Board then held that such the retrospectivity of the legislation was not enough to amount to direct or indirect interference in the judicial power. There was no direct interference because the legislation was a mere alteration to the general criminal law. There was no indirect interference because the appellants had failed to show that the legislation was colourable in the sense that in practice only a limited category of persons would be affected by the legislation. In Lord Sumption’s words, “The Amending Act not only looks like general legislation. It is general legislation” (Ferguson at [31]).

There can be no constitutional objection to the availability of Section 61 powers over the second category of provisions. Any exemption to this category of provisions would not amount to an executive pre-judgment of the ultimate issue of whether the Ministerial order or direction in question should be set aside. It would only affect the conduct of ongoing or future proceedings by persons or classes of persons. As I have explained, this is not enough to constitute direct or indirect interference in the judicial power. It is my respectful opinion, however, that Section 61 arguably cannot, in conformity with the principle of separation of powers, be applied to the offence-creating provisions of the POFMA (ie, the third category of provisions). First, I argue that Section 61 arguably constitutes a direct interference in the judicial power. By Section 61, the Legislature confers on the Minister the power to issue a bill of attainder in reverse form (ie, an executive judgment that a person or class of persons shall be innocent of criminal liability). It does not matter that Section 61 provides for an executive power to exempt from criminal liability, as opposed to an executive power to judge, determine, or direct the court to find that a person is not criminally liable. As I 261

COMMENTARY: GUEST WRITER

have explained, the courts look to the substance and not the form of a legislative scheme. There can be no meaningful difference between, on the one hand, an executive power to direct the court to adjudicate a controversy in a particular way, and on the other hand, an executive power to remove the controversy entirely for a specific person or class of persons (and not others).

would qualify a person or class of persons for exemption from criminal liability. The analysis may be different even if the Minister is vested with the exclusive power to determine if these general conditions are met. That is however not the case with Section 61.

The point can be illustrated this way. It cannot be disputed that legislation providing that a person shall be guilty of an existing offence is impermissible. What difference is there between such legislation, and legislation creating a completely new offence, and further providing that a specific person (and no one else) shall be guilty of that offence?

It may be noted that many such exemption clauses can be found in other Singapore legislation. For instance, under Section 29 of the Enlistment Act,7 it is provided that “[t]he proper authority may by notice exempt any person from all or any part of the liability of that person under this Act”. A full examination of whether all the various exemption clauses found in other legislation amount to a legislative interference in the judicial power is beyond the scope of this short article. I can however make a few general observations here. Firstly, it is not the case that an exemption clause necessarily amounts to an unconstitutional interference in the judicial power. As can be seen in this article, I accept that Section 61 of the POFMA would not impinge on the judicial power so far as the second category of provisions is concerned. Secondly, each exemption clause must be scrutinised in light with the entire legislative scheme and context, to ascertain whether the exemption clause:- (a) is a legitimate conferral of powers to make rules of general application in relation to a person’s or class of person’s rights and duties; or (b) is an illegitimate ad hominem legislation targeted at securing a particular outcome for specific persons (or classes of persons). Thirdly, if, however, an exemption clause is construed to be an unconstitutional interference in the judicial power, then the fact that similar exemption clauses can be found in many other legislation cannot save the provision from striking down under Article 4 of the Constitution.

Second, if I am wrong on the view that Section 61 constitutes direct interference, it is my view that Section 61 arguably amounts at least to an indirect interference with the judicial power As Lord Sumption explains in Ferguson, indirect interference occurs where the legislation in question putatively provides for mere alterations to the general law, but in practical effect is targeted towards a specific outcome for a limited and identifiable class of persons. The fact that Section 61 is framed in terms of exemption rather than determination may suggest that the Minister powers is given only powers of general subsidiary legislation. But the analysis of indirect interference does not stop there. What appears constitutionally objectionable is the fact that the Minister is, on Section 61’s own terms, given the power to determine that specific persons or classes of persons shall not be imposed with criminal liability for doing acts which would, if done by other persons, attract criminal liability. Underscoring the ad hominem nature of the powers sought to be conferred by Section 61 is the fact that Section 61 empowers the Minister to exempt persons after they have committed acts which would ordinarily attract criminal liability, and even after criminal proceedings are afoot. The analysis may be different if Section 61 prescribes general conditions which, if met, 262

Exemption clauses in other Singapore legislation

Conclusion There is no suggestion in this article that the 7

Cap 93, 2001 Rev Ed.

COMMENTARY: GUEST WRITER

proposed exercise of power under Section 61 shall be anything other than in good faith, with relevant considerations taken into account and irrelevant considerations not taken into account. No doubt, the proposed exercise of the Section 61 power shall be susceptible to judicial review in the normal way. However, as explained in Prabagaran and Faizal, it is a strict rule that any legislation that interferes with the judicial power shall be struck down for breach of separation of powers.

263

Thank You to our Sponsors: Gold Tier Baker McKenzie. Wong & Leow Clifford Chance LLP Rajah & Tann Singapore LLP WongPartnership LLP Silver Tier Allen & Gledhill LLP Allen & Overy LLP Dentons Rodyk & Davidson LLP Drew & Napier LLC Singapore Legal Service TSMP Law Corporation Bronze Tier Latham & Watkins LLP