« Choose My Midlife Crisis. Sponsorship Opportunities Available. | Main | Couple of teaching thoughts »

Wednesday, December 30, 2009

Crime Decline Conundrum

With aviation terrorism and a still lackluster employment market dominating year end headlines, the one piece of good news appears to be a fairly widespread decline in homicides in major cities. New York, as trumpeted in yesterday's NYtimes (read Al Baker's reporting) had a year with fewer homicides than any year since 1963 (essentially before the modern crime wave was evident). San Francisco also reported a record drop (read Jaxon Van Derbenken's article in the SFChron) to as low as the city has seen since 1961 (take that New York), and after a series of rather violent years in the middle of this decade. Chicago and LA have also reported declines this year. Providence, was one of the few cities reporting a homicide "spike," with the addition of two dead this week in a drug raid that also left three police officers wounded (read W. Zachary Malinowski's reporting in the Providence Journal). This is good news in a year with little of it.

The journalistic lead is that this is happening despite a severe recession (the man bites dog angle). Whatever the intuitive appeal to the notion that bad times generate crime, few criminologists believe it is a clean relationship. In many respects, times are always bad in those communities that experience the highest levels of crimes like homicide, aggravated assault, and robbery. This, not surprisingly, does not stop police chiefs and mayors from claiming credit (at least if they've been on the job for more than six months) whatever the hazard that their policies might be blamed when crime begins its inexorable return (like most gambles, it probably makes sense in the short term context of political survival). But even criminologists, this one included, are not immune from believing that, combined with the substantial crime declines of the 1990s, and the relative stability of crime through most of this decade, this end of decade crime decline could mark a longer term shift away from the pattern of high levels of gun violence concentrated in cities that has defined urban life for the much of the past forty years. What would drive such change? Here is a New Year's speculation list of the top three "positive" factors underlying declines in urban violence.

May they all continue in 2010!

1. Bottoming out of the de-industrialization of American cities that began in 1946 and continued through the 1980s. Even if new economic engines of prosperity have not exactly re-emerged in many cities, the process of losing existing assets has run its course.

2. Demographic diversification of urban neighborhoods through immigration and in-migration of suburbanites fleeing unsustainable lifestyles.

3. Better trained and motivated police forces.

cross-posted at:


Posted by Jonathan Simon on December 30, 2009 at 02:07 PM in Criminal Law, Jonathan Simon | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Crime Decline Conundrum:


Mark's point is well taken. The dictatorship of annual reporting periods creates all kinds of distortions (as our financial accounting wizards have long demonstrated). Pity the poor police chief in Providence whose year end numbers were blown out by a final violent week (of course it was a police raid that generated the two deaths which reminds us that crime is not fully exogenous to policing).

Posted by: Jonathan Simon | Dec 31, 2009 12:58:09 PM

This is clearly good news. But there are a couple of questions that I think are worth asking about the data (sadly, this is the one instinct I seem to have retained from my brief incarnation as a graduate student in sociology).

January 1 to January 1 borders are obviously unrelated to anything about the underlying phenomena the data describe. It is useful to compare apples to apples by measuring one January 1 to January 1 period against another, but it's worth remembering that if, say, murders spiked during a 2 month period that was April and May in 2007, and spiked again during a two month period that was December 08-January 09, 2007 would look like a worse year than either 2008 or 2009 if everything else was constant except for the timing of the spike.

That distortion could mask spikes or declines in the murder rate that might be traceable to particular factors. For example, it would be useful to know how the murder rate during a moving 10 day average relates, if at all, to the foreclosure rate during a moving 10 day averagethe previous 2 or 3 years.

I say all this without the slightest idea what the data actually show of course. It's easy to sit here and write about what the data might show; much harder to actually dig in and find out. My hat's off to scholars like Jonathan who do just that.

Posted by: Mark Edwards | Dec 30, 2009 7:51:38 PM

Larry I agree that New York is the best case scenario and I'm confident policing is a large part of that (although perhaps not for the reasons broken windows theory would require). San Francisco's new police chief George Gasgon is also claiming some credit for the use of concentrated policing in hot spots and targeting particularly high risk offenders, tactics bit different than NYPD but not unrelated. I would also point out however that New York since the 1990s has also been very big in both of my other two factors. It has enjoyed (at least until recently) a very strong post-industrial economy and it has experience strong immigration and in-migration.

Posted by: Jonathan Simon | Dec 30, 2009 3:16:43 PM

Could it be, perhaps, that New Yor outperforms all other major cities because its police force uses the most aggressive tactics to take control of the streetscape? The kind of tactics that most criminologists assured us cannot work because they do not address the "root causes" of crime?

Larry Rosenthal
Chapman University School of Law

Posted by: Larry Rosenthal | Dec 30, 2009 2:14:37 PM

The comments to this entry are closed.