Universal Pension, Yes, Regressive Funding, No

Mariano Bosch, Carmen Pages, and Angel Melguizo at the Inter-American Development Bank (IDB) are proposing a new approach to expanding the coverage of pension systems in Latin America while helping create more and better jobs. Their ideas are spelled out in a new book "Better Pensions, Better Jobs: Towards Universal Coverage in Latin America and the Caribbean." The book is about Latin America but the problems discussed and proposed solutions are relevant for any middle-income country. I think the IDB's proposal is a great contribution to the debate on pension reform. Below I discuss some of the points they make that I agree with and those where I think other options could be considered.

Much to like about the IDB pension proposal...

Mariano recently posted a blog here on jobsknowledge.org that summarizes the book's main messages. In a nutshell, the proposed model for reforming pensions systems has two components: (i) an anti-poverty non-contributory pension that is financed out of general revenues; and (ii) a contributory system that can include contribution subsidies for middle- and "low-income" workers.

There are three main things that I like about the book and the model. The first is the connection the authors make between pension reform and the labor market. Indeed, an important objective of the recommended policy changes is to reduce taxes on labor (the tax-wedge) that we all agree can reduce formal employment. The second is the model makes a clear distinction between the redistributive function of the pension system (anti-poverty) and the insurance function (smoothing consumption/preserving standards of living). Thus, the authors recommend that workers' contributions should always be linked to benefits. And, finally, the third is that the issue of "funding" does not come out as an important element of the reform. All the recommendations can be implemented in the context of a pay-as-you-go system (see my posting on non-defined contributions) or funded defined contribution systems.

...but only if the model is NOT regressive

What I don't like is that the proposed model can be regressive — that it can redistribute income from low — to higher-income workers. Indeed, potentially large fiscal outlays would be used to subsidize the contributions of workers who enroll in the contributory system. They are, by definition, formal sector workers who are likely to have, on average, higher incomes. Clearly, the subsidies can provide incentives to workers (and employers) in the informal sector to enroll in the contributory system but the increase in the coverage rate is likely to be small (a few percentage points).

What is the rationale for those subsidies? If the function of the contributory system is consumption smoothing (on top of the social pension), it is unclear that subsidies should be used. If there are concerns that workers can’t afford the contributions needed to finance a given replacement rate (the ratio between the pension and pre-retirement earnings), then probably the social pension is too low or the targeted replacement rate is too high.

If, on the other hand, the objective of the subsidies is to reduce labor taxes it becomes important to identify what those taxes are. Implicit taxes used to finance minimum pension guarantees or to subsidize workers with short career histories should indeed be eliminated (given that the social pension is financed out of general revenues). However, workers' contributions that are linked to benefits shouldn't be considered a tax — unless the target level of benefits is well above what workers want, and in that case, the solution is to reduce these benefits.

The other tax is the pay-roll tax paid by employers. In theory, this "tax" can be passed on to workers through lower wages (in exchange for a higher pension). But it is a reality in developing countries that many employers, particularly small, low-productivity firms, might not be able to afford the tax. These, for the most part, are firms that also pay low wages and operate informally.

So in this case, the solution could be to reduce or eliminate pay-roll taxes for the firms (they would be targeted by productivity level, not size) but also the benefits that would have been financed from those taxes. Worker in these firms would be registered in the pension system (to be eligible for the social pension workers should be required to register/enroll) and they can contribute if they want. If there are concerns that their pension could be "too low," then the subsidies would be channeled through the "social pension." Basically, this pension would focus on the redistributive function of the pension system. Nick Barr in his recent JKP blog discusses this in the case of Chile. The social pension could take the form of deposits in the individual accounts. But again, the subsidies would go toward the poorest workers — whether wage employees in the informal sector or self-employed, and whether they contribute or not.

The JKP plans to continue this debate on pension reform through our blog. Also, a heads up for a new book that is coming out on the topic co-authored with Carmen and others: "Social Insurance, Informality, and Labour Markets: How to Protect Workers While Creating Good Jobs."

This post was first published on the Jobs Knowledge Platform.