Head Strong: A tiny group of city residents commits most of violent crimes

In understanding the murder of Officer John Pawlowski, here's one crime statistic that's significant: 0.14 percent.

That's the percentage of city residents that makes up the tiny band of marauders who are committing most of the crimes burdening the city. And they are just slightly more than one-tenth of 1 percent.

They are young, male and African American. Broken homes. Lengthy rap sheets. Escalating violent practices. Rasheed Scruggs fits that bill. If Central Casting went looking for someone likely to murder a cop, Scruggs would have gotten a call back.

Consider: A Daily News piece about his family featured an interview with his mother and reference to five sisters, but made no mention of a father figure.

His rap sheet includes nine arrests for a litany of offenses. He spent five years in jail after pleading guilty to armed robbery in 1997. He served an additional six months in 2004 after violating his parole. Before he allegedly murdered Pawlowski, Scruggs was due in court that week as a result of a September charge of car theft.

He's the latest in the line of recidivist cop killers to strike in the City of Brotherly Love.

The temptation might be to cast the entire city as irretrievably crime-ridden. Or to criticize judges for imposing lenient sentences. Or urge the city and state to build more prisons. Or even to lay the blame on young African American men, who we know make up disproportionate numbers of murderers and their victims.

All of which is completely unfair.

Consider this: In July 2001, the city's Repeat Offenders Unit identified more than 2,000 career criminals who'd been arrested at least 25 times each - a collection of all-star rap sheets that accounted for as much as 60 percent of the city's crimes.

I'm willing to bet there's a comparable number of ultra-recidivists today. Which would mean 0.14 percent of the city's 1,449,634 residents is committing more than half of its crimes.

In other words, we're dealing with a relative handful of career criminals committing a disproportionate amount of the misdeeds.

That's not to say that every parolee is a likely murderer. Or even that each of the city's 2,000 uber-criminals is going to pull the trigger someday. But seven Philadelphia officers have been killed in the line of duty since 2006, and something must be done to quell what has become the most dangerous city in America for cops.

Mayor Nutter's aides will spend the next several weeks sorting through the federal stimulus package in anticipation of receiving money to be spent on hiring more police officers. Those resources should be directed toward predicting which of the ex-offenders among us are likely to commit a homicide or kill a cop, and targeting them before they have the chance to act. A study completed last year concluded that such predictions are more accurate given recently developed statistical procedures.

In "Forecasting Murder Within a Population of Probationers and Parolees: A High Stakes Application of Statistical Learning" (which then-City Councilman Nutter helped fund), University of Pennsylvania criminology professor Richard Berk and Penn's Jerry Lee Center of Criminology considered a set of 66,000 cases under the jurisdiction of the Philadelphia Adult Probation and Parole Department. They determined that certain factors can help forecast whether an individual is more likely to commit a homicide (or attempted homicide) while they're under parole supervision.

In particular, both an individual's age and the age at which that person first enters the adult court system are the most statistically meaningful factors, according to the analysis. The number of prior firearms convictions is another important predictor. Race improves the ability to forecast violent offenders, but to a lesser degree.

Parolees and those on probation who represent a confluence of those factors are almost eight times more likely to be charged with murder or attempted murder within two years of coming in contact with APPD, the study concludes. The point? "If useful forecasts can be made, scarce resources can be concentrated on the 'power few' subset of offenders who cause the greatest harm and public concern about crime."

Exactly. In other words, the city should think less in terms of monitoring police districts, neighborhoods, or street corners and more in terms of identifying and targeting the actual people - ex-offenders, parolees, those on probation - likely to commit a homicide.

We know it's a relatively small number of career criminals to follow. And the more we do, the less that officers like John Pawlowski will have to encounter guys like Scruggs on the street.