Friday, September 23, 2016

The Economics of Crime

On September 28, the Economics Department at Haverford College will hold its annual alumni forum. The topic this year is "The Economics of Crime and Incarceration." Our panelists will be
Eric Sterling (Haverford class of '73), Executive Director of the Criminal Justice Policy Foundation, and Mark Kleiman (class of '72), Director of the Crime and Justice Program at New York University’s Marron Institute of Urban Management. In anticipation of the event, especially for any Haverford students who might be reading my blog, I wanted to do a quick survey of the literature on the economics of crime and some of the major topics and themes in this literature.

Why are crime and incarceration economics topics? In other words, given that there is an entire field--criminology--devoted to the study of crime, why are economists studying it as well?  Gary Becker suggested in 1968 that "a useful theory of criminal behavior can dispense with special theories of anomie, psychological inadequacies, or inheritance of special traits and simply extend the economist's usual analysis of choice" (p. 170).  In other words, he believed that criminal behavior could be modeled as a rational response to incentives; that the private and social costs of crime, and the costs of apprehension and conviction, could be quantified; and that a socially "optimal" (likely non-zero) level of crime could be computed.

How does the criminal justice system affect the incentives for crime, and, in turn, criminal behavior? Causal effects are quite challenging to study empirically. For example, consider the question of whether a larger police force deters crime. Suppose the data shows a positive correlation between crime rates and size of police force. While it is possible that larger police forces cause more crime, it is also possible that causality runs in the reverse direction: cities with higher crime rates hire more police. Steven Levitt, whose "Freakonomics" fame came in part from his clever approaches to these types of questions, has looked for "instruments," or ways to identify exogenous variations in criminal justice policies.

It is also difficult to identify causal effects of incarceration on criminal recidivism and other outcomes. Prison sentences are not "randomly assigned." So if we see that people who spend longer in prison are more likely to commit a second crime, we can't say whether the extra time in prison had a causal influence on the recidivism. A recent working paper by Manudeep Bhuller, Gordon B. Dahl, Katrine V. L√łken, and Magne Mogstad exploits the random assignment of criminal cases in Norway to judges who differ in their stringency of sentencing. They find that imprisonment discourages further criminal behavior. This decline in recidivism is driven by people who were unemployed before incarceration, and who participated in programs in prison aimed at increasing employability. The authors conclude that "Contrary to the widely embraced 'nothing works' doctrine, these findings demonstrate that time spent in prison with a focus on rehabilitation can indeed be preventive." But since not all prison systems have a focus on rehabilitation, they add that "It is important to recognize that our results do not imply that prison is necessarily preventative in all settings. While this paper establishes an important proof of concept, evidence from other settings or populations would be useful to assess the generalizability of our findings."

Some dimensions of crime can be difficult to measure. Many crimes go unreported or undetected. Black market activity, by its very definition, is hidden. Economists have also tried to come up with ways to measure illegal production or trade. See, for example, this study of elephant poaching and ivory smuggling. Online black markets, and other types of crime and fraud committed online, are also the subject of a growing economics literature.

Network economics is also applicable to the study of crime, since it can help with understanding the formation and workings of criminal networks.

Studies of the economics of crime are nearly always controversial. In part, this is because criminal justice itself is so controversial, so whenever an economic study draws implications about criminal justice, it is sure to find some resistance. In addition, many people find Becker's description of crime as a purely rational response to incentives to be lacking. Recall, for example, the controversy surrounding Roland Fryer's recent working paper on racial differences in police use of force. I think part of what people were uncomfortable with was the incorporation of racial discrimination into the utility function, and part was the distinction he made between "statistical discrimination" and racial bias.

I anticipate an interesting discussion on Wednesday and will try to update the blog with my impressions following the forum.


  1. How to match low skilled workers (Crips, Bloods, American raised taxi drivers) with low skilled work? Pay $800 a week for today's $400 jobs (how much many paid before).

    Collective bargaining sets the price of labor by HOW MUCH the consumer will put up with (how we set the minimum wage) -- instead of HOW LITTLE the most willing (most desperate foreign born) worker is willing to put up with.

    "It's the American dream, dog: flush toilet down the hall, AM radio, electric light in every room." Let's call that $200 a week living. The Crips and the Bloods and my "gang" the American raised cab drivers would gladly stack shelves and drive cabs for that -- the best the average person ever had it. And the year is 1916.

    100 years later, American raised workers wont show up for $400. Instead, 100,000 out of my guesstimate 200,000 Chicago, gang age males are in street gangs.

    Quick federal answer (w/o Repubs): an NLRB finding of firing organizers should lead to mandated certification election.

    Today, even if you have to rehire organizers for life at double pay, that does nothing to establish bargaining balance to the labor market (monopsony v. monopoly).

    Quick (doable) progressive states answer (WA, OR, CA, NV, MN, IL, NY, MD): make union busting a felony. States can set higher min wages and states can reinforce (non)enforcement of the right to collectively bargain.

    Today, labor market muscling doesn't even have the protection of the "FBI warning" we get before we watch a recorded movie. If you actually get caught making a movie in the movie house you will actually do a couple of years fed hospitality (for what little econ damage you perpetrate).

    It's just a matter of getting the blockade out of people's way and letting them do what they want -- and then watching the first 2,000 people in the telephone directories remake economic and political landscapes.

    1. I might add that the social pathologies in the white rust belt areas are probably parallel to the pathologies in the black ghetto areas: all basically anchored in low skilled jobs that could be paying $800 only paying $400 -- all due to nationwide de-unionization.

      5% union density in private business is a strange pathological condition -- that nobody sees. All discussions on economic and/or political degeneration take place within this uniquely American de-unionized context without ever recognizing that the context is what should be being discussed! !!!

      How to make the scales drop from all (ignorant) American lives?

  2. "This decline in recidivism is driven by people who were unemployed before incarceration, and who participated in programs in prison aimed at increasing employability." I sometimes wonder why it's so hard for Americans to understand that we have an elaborate set of rules set up that make a person nearly unemployable if they have once been imprisoned. Then we don't understand why recidivism rates are so high. Of course the idea of giving people education or job training while they're in prison is obviously crazy -- that will just be an incentive for people to commit crimes. /sarc


Comments appreciated!