« Chemo Aftermath--Stuntz | Main | Father Richard John Neuhaus--Skeel »

Policing and Stimulus Packages--Stuntz

Obama and his underlings have emphasized, rightly, that federal spending designed to pump up the economy should do more than that: spending should rebuild needed infrastructure, invest in cleaner energy, and the like, so that money spent now would yield economic returns years later.

There is one kind of spending that would do just that: federal aid for local police. The number of urban police officers per unit population held steady through the 1970s and 1980s, while urban violence steadily rose. In the 1990s, that number rose 17%, and violent crime fell sharply. In this decade, nearly half of the gains of the 1990s have been wiped out during this decade--and that was true before the collapse of the credit markets this past fall and the broader recession that is now taking hold. Urban violence is rising again. If the federal government doesn't subsidize police spending, we will see more cuts in local police budgets, and probably more crime.

Police spending has another large benefit: over time, it reduces prison spending. Take a look at this graph--the blue curve is annual change in the number of urban police officers per unit population with a one-year delay; the green curve measures annual change in the number of prison inmates per unit population. The inverse relationship between the two curves is hard to miss:

Imprisonment and Urban Policing Rate Graph

Less crime, fewer prisoners, and more cops is a rare policy trifecta. This is an opportunity not to be missed.

TrackBack

TrackBack URL for this entry:
https://apps.law.upenn.edu/weblog/mt-tb.cgi/617

Comments ( 3 )

This is really interesting - and hopefully instructive. What's the source of the graph? Is it published anywhere?

I would also be interested in the source of this graph, is it available somewhere?

Like Anna and Joseph in the 2 previous comments, I would like to know the source of the graph and the data behind it.

You say "The inverse relationship between the two curves is hard to miss." I disagree. Although there are about 14 years that show a clear inverse relationship, it looks to me like there at least 6 that don't, and the changes hardly have what I would call a consistent relationship to one another.

Even if the relationship were clear and consistent, I would not draw the same conclusions that you draw from it. The data would suggest merely a correlation, not causation.

Additionally, is an increase in the U.S. imprisonment rate really the goal? Increased imprisonment does not mean reduced crime.

In short: the graph does not support your analysis, the effect (assuming it even exists) is causative rather than correlative, and your end result may or may not have anything to do with a bona fide public good.