Monday, July 23, 2012

A thousand small sanities

 NYPD's hotspot tactics & Compstat are controversial - but effective
[Photo: Ianqui Doodle/Flickr]
Readers of SafeGrowth know certain high crime properties are incubators for gangs and violence. That isn't destiny, it's reality. If SafeGrowth (and approaches like it) prove anything, it proves residents are not doomed to a life of mayhem. Environment can be changed and streets transformed.  

It also proves we can do it with coherent planning, mobilized neighborhoods and intelligent anticrime strategies like hotspot policing.

Case in point: crime declines in New York.

I recently read Frank Zimring in the New York Times:  "The 40% drop in crime that occurred across the U.S. from 1991 to 2000 largely remains an unsolved mystery. Even more puzzling then is the crime rate drop in New York City, which lasted twice long and was twice as large. This 80% drop in crime over nineteen years represents the largest crime decline on record."


Zimring studies New York's crime declines - A book worth reading.
A mystery?

I'm not big on mysteries that aren't. It's like watching a Hollywood flick and expecting some magical, non-formulaic finale. Not going to happen!

That 40% drop nation-wide followed a decades-long demographic metamorphosis that swept North America more than anywhere else since WW2. Since the 1990s crime-prone cohorts aged out of crime in record numbers. Those crime declines continue today.                                                          
Then New York built on that perfect demographic storm as NYPD added crime suppression tactics like proactive street stops and controversial (but clearly effective) quality-of-life enforcement.               
Intensive street stops increased the risk of getting caught with an illegal gun. That led to a 39% drop in gun toting criminals from 1993-1995. Is it really a mystery that kind of informal gun control cut violence?

FEWER PRISONERS?

It's what Greg Bergman calls A Thousand Small Sanities (another excellent read).

During the peak crime declines fewer arrestees went to prison. Why? Bergman describes the vast network of incarceration alternatives evolved in New York - drug courts, mental health courts and community courts providing meaningful community alternatives like drug treatment and restorative justice.

Says Bergman "there needs to be a continuum of non-incarcerative interventions for offenders with the most intensive options reserved for populations that are both high risk and high-need."

Hotspot policing, neighborhood justice courts, and targeted suppression. Anchor that with permanent SafeGrowth planning and neighborhood capacity building and voila -  a finale that makes sense.


7 Replies so far - Add your comment

Tim Hegarty said...

My wife and I just returned from New York. It was my third visit there in six years, and at no point during any visit did I ever feel anything but safe. It's tempting to chalk that up to the fact that there were cops on every block, but that's not the whole answer. Several years ago, I was with a group of eight, all of us police officers, in downtown Baltimore after an evening ball game. There were cops on every street there, too, but we did not feel safe.

There is much to learn from New York, as well as other cities that have become safer. The first step involves following Braga's advice by focusing police actions "on the places, the times, and the people who pose the highest risk to public safety rather than dilute their crime prevention potency by spreading them thinly across the urban landscape." The second step is understand that those "actions" should involve not only enforcement but also alternatives to incarceration, problem solving, and community engagement. These are the lessons of safe cities.

GSaville said...

Thanks Tim. Excellent points.

NY has lessons to offer, both good and bad. My prior blogs on cooking-the-books-with-compstat is one glaring embarrassment in NYPD.

There is also a limit to what Bragga suggests (though he is a stellar researcher and rarely misses much). It is this: any place/time hotspot strategy requires one main element: crime! Obvious? Yes, but it shows that this whole strategy is reactive. After all, there can be no hotspot strategy without a hotspot - meaning the crime has already happened.

That's why I add "permanent SafeGrowth planning" along with capacity building. As an urban planner first, criminologist second, I'm more interested in getting those neighborhoods right in the first place than creating tactics to go after-the-fact. Both are needed, of course...yet when we do one without the other we risk treating the symptom and not the illness.

For my money the ultimate point is creating sustainable, safe, vibrant, and healthy neighborhoods!

Tim Hegarty said...

Agree, but I think that we have to admit that the nature of most police work is reactive to one degree or another. Even the latest "predictive policing" software (getting a look at that tomorrow, btw), makes its calculations based upon crimes that have already happened. Being reactive isn't a bad thing per se; we fail when we react in ways that the past has shown to be ineffective in creating and sustaining safe and secure communities.

Joshua said...

"The 40% drop in crime that occurred across the U.S. from 1991 to 2000 largely remains an unsolved mystery."

Theories abound. My wife and I came across this one a few months ago:
http://www.freakonomics.com/2005/05/15/abortion-and-crime-who-should-you-believe/

GSaville said...

Thanks Josh. Yes I know Levitt and love his work. His chapter fits into the general demographic story I described in the blog. Makes sense to me.

Thanks also Tim. You make a good point.

I think we must return to the core strength of street cops and help them reconnect with the residents they serve in a more intense and personal manner. For me that means problem-solving partnerships and problem-oriented policing. If "predictive math" makes that happen - count me in. But, I am skeptical. I've seen these things come and go - mostly go.

I recall testing the PCAM predictive deployment software back in the 1980s. Even that early version suggested this may redirect precious funds away from the proactive policing that actually solves crimes.

I'm not a solve-the-city-with-math kind of guy. I know you agree that, along with PBL, problem-oriented policing that is an important thread of the fabric we call public policing. Those threads are now very frayed and, IMHO, predictive analysis might be the wrong direction.

My blogs on this are many:

http://safe-growth.blogspot.com/2011/01/solving-city-with-math.html

http://safe-growth.blogspot.com/2011/12/crime-prediction-and-precog-paradox.html

Long ago Goldstein suggested the primary determinant of police service delivery was the telephone. More than anything else, that determines the reactive nature of police work. He followed that by saying there is a "limited capacity of the police to deal with crime by simply reacting when a crime occurs."

Encouraging, participating, and planning for cops to help build problem-solving capacity in neighborhoods - that's a future that makes sense to me.

Tim Hegarty said...

I sat through a presentation yesterday involving an algorithm-based program that attempts to predict future crime. It is the one tested by LAPD. Rather than call it predictive policing, a more apt descriptive term would be probability policing. I won't go into a lot of detail, but I will say that it's something that should not be dismissed. It could very well be one piece of the problem-solving puzzle, but only one piece. It doesn't take the place of a human crime analyst. It doesn't eliminate the need for problem solving. And it doesn't reduce the importance of collaborating with others. It's not cheap, and it could very well tempt agencies to divert funds away from more effective crime reduction strategies, but like any other new idea, only time will tell.

GSaville said...

Thanks Tim for the update. Very interesting. I hope you are right.

I am familiar with probability modeling. Back in the 1990s my universitiy research center developed predictive crime pattern mapping using a new model we developed called tipping point threshold theory. It worked find but it was pretty expensive and data heavy. Ultimately I just could not justify diverting precious public dollars for problem-solving onto fancy (and costly) math.

But that's just me.