Local Penn professor talks crime

by Tom Beck
Posted 5/11/23

Just last month, a Pew Research Center poll of Philadelphia residents confirmed what a lot of city residents already believed: reducing the city’s crime rate is a “top priority.”

This item is available in full to subscribers.

Please log in to continue

Log in

Local Penn professor talks crime


Just last month, a Pew Research Center poll of Philadelphia residents confirmed what a lot of city residents already believed: reducing the city’s crime rate is a “top priority” for the Philadelphia electorate. A whopping 89% of Philadelphia residents, the poll showed, considered reducing crime a top priority, the highest polling policy issue of the seven polled. Education ranked a somewhat distant second place, with 75% of Philadelphians considering it a top priority. Jobs and the economy rounded out the top three at 65%. 

And it’s a hot topic within the 8th District too. When the Local asked readers to submit questions for last month’s 8th councilmanic debate between incumbent City Councilmember Cindy Bass and her challenger, Seth Anderson-Oberman, more public safety-related questions were submitted than for any other issue. 

Luckily for us, our neighborhoods have their own resident crime expert: University of Pennsylvania Emeritus Professor of Criminology and Statistics Richard Berk. 

A Mt. Airy resident, Berk is most known for creating an algorithm that helps judges and parole boards determine which people in the criminal justice system are most likely to commit additional crimes, and which are not. It’s designed to assist judges and parole boards in deciding whether prison is the right choice for an individual – and Berk argues it can help avoid unnecessary imprisonment and also save taxpayers money.

Berk’s algorithm factors in such things as an offender’s prior record, how they were when first arrested, and how far back in time they committed their crimes. It has seen its fair share of criticism - in large part for its propensity, detractors say, to reduce human beings to mere numbers.

“When someone is committing crimes right now and you arrest them, it's a pretty good bet that they'll commit those crimes again,” Berk said. “If the crimes they committed were over 20 years ago, then no, maybe not.”

Berk acknowledges the shortcomings but said he thinks human bias can be worse. 

“It's going to make mistakes, but humans make the same mistakes and more,” Berk said. “The idea is to improve current practice. Nobody is trying to produce a perfect anything.”

With residents in the 8th District raising concerns over specific issues related to crime – whether or not to reinstate stop-and-frisk policies, Philadelphia District Attorney Larry Krasner’s performance and the movement to defund the police – we picked up the phone to consult with him on the current state of crime in the city, both in our own neighborhoods and beyond. 

The interview was lightly edited for clarity and length.

Is the algorithm you created any more relevant now than it's been in recent years?

Yes. Here's the thing, I don't know if you’ve watched any of the recent mayoral debates, but if you did, every one of them said we've got to do something about crime and get the shooters off the street. It's obviously a serious problem. 

The problem with the current criminal justice system is that most of these folks – or at least a large fraction – already have had contact with the criminal justice system, and for one reason or another they're still on the street, or they're back on the street. 

The current way in which we screen individuals [for incarceration] is very blunt. What we're proposing with these algorithms is to be more surgical. It does not mean we incarcerate more people, it means incarcerating fewer people, but doing it smartly. 

The problem with so many of the proposals for homicide prevention is that they're not smart. Stop-and-frisk, for example.

These algorithms can give you anything you want, in a sense. You can reduce crime, and you can reduce the prison population, and you can get increased fairness. Take these decisions partly out of the hands of current decision-makers, because the algorithm can be more surgical.

Cherelle Parker is unique among the city’s mayoral candidates in that she’s open to the idea of implementing “stop and frisk.” Where do you stand?

This is complicated because stop-and-frisk means different things in different cities. The policies are different. In New York City, there's no evidence that stop-and-frisk reduced crime. Yes, crime went down, but it went down all over the country during that period. Stop-and-frisk doesn't have anything to do with that. 

My complaint about stop-and-frisk is that it's very blunt. There are other objections you might have also, including civil liberties. But in terms of finding the bad guys, stop-and-frisk is very blunt. And there's no evidence that it really works. 

In the debate, someone said ‘Yeah, I'm for stop-and-frisk, but I'm for fair stop-and-frisk that protects civil liberties.' Great, tell me how you do that. It's not clear what that even means. 

Homicides are increasing in the city and some blame the progressive policies of District Attorney Larry Krasner. What do you say?

That's a hard question. The answer is nobody knows why homicides are up. There are about a half dozen different explanations floating around. It's gone up all over the country, not just here in Philadelphia. It's gone down in some places too, and it's gone up in some rural areas. 

There's no single story that accounts for this, and there certainly is no smoking gun for Larry Krasner. Maybe his policies are involved, we just don't know. 

While homicides are going up, the rate is not nearly as high as it was in the 1990s.

Is it true to say crime, in general, is going up?

That gets really tricky because it depends on what kind of crime you're talking about. A lot of very serious crimes are not reported. Rape, for example, is underreported terribly, so we don't really know. 

It turns out you've got all kinds of weird things going on. For example, in some cities, the number of homicides has gone up but the number of shootings has not. How could that be? Now it gets complicated. When someone is shot and medical attention is required immediately, a lot of other factors come into play – how quickly they get to a trauma center, the quality of that trauma center, and whether they’re shot with a high-caliber or low-caliber weapon. If you're shot during rush hour, for instance, your chances of surviving are much worse than if you're shot at 3 in the morning because you can get to the emergency room much more quickly. 

I think the only thing you can be pretty confident about is the enormous, easy availability of firearms – handguns in particular – but also things like AR-15s. That's clearly playing a role because those are particularly lethal.

I think a lot of people feel the city is in a much worse place than it's been crime-wise in decades past. How true is that?

During the crack epidemic in the early 90s, it was much worse. Let me just emphasize that when I say much worse, it depends on where you live in the city. I live in Mt. Airy. How many homicides were there last year? One or two? That's not like 10 a week, which is what happens in certain areas of the city. 

In fact, there was recent data on downtown Philadelphia being very safe, and it is. 

What’s your stance on the movement to defund the police?

If there's one thing that we know will reduce crime in the short run it's better law enforcement. I would never for a moment support any ideas to defund the police. We need more police, we need better police and we need oversight. 

We know more police equals less crime. That's been demonstrated. But we should be doing long-term fixes too. Better schools and better health will, over a period of a generation or two, improve the situation dramatically. We should be doing both. 

Unfortunately, the political time horizon is about two or three years, so the long-term stuff gets lost and the short-term stuff gets mixed up with all kinds of silly politics.

There's no doubt that very often, some police officers around the country – including in Philadelphia – are behaving in unprofessional and dangerous ways. But most are not. Most are good people trying to do a very difficult job. 

What's the evidence for that? 

Studies at major police departments around the country have identified police officers who disproportionately arrest people of color, and shoot with unnecessary force – all kinds of stuff. It’s a tiny fraction. You can call them bad apples. But that's a structural problem.  In a police department that's run effectively with oversight, you find those bad apples quickly and you either sanction them or fire them. 

Every occupation has some bad apples. Some college professors are bad apples.  I'm sure some newspaper reporters are as well. The question is whether they're caught and proper remedies are introduced.