Crime: Costs and Benefits

Reprinted from The Joy of Economics by R. Stonebraker

Who knows what evil lurks in the minds of men?                                                                         …the Shadow

Presidents commit crimes; priests and pastors commit crimes.  Why should you be any different?  Why should I?  I commit uncountable crimes; jaywalking is a particular favorite.

Economists treat crime as a rational act. Certainly some crime does result from unthinking, irrational behavior driven by blind rage or passion, but I make choices about jaywalking, and I am not alone.  Perceived costs and benefits can trigger criminal acts just as they trigger decisions to buy vanilla rather than rocky road ice cream.

The benefits of committing a crime include a variety of monetary and psychic pleasures, but there are costs as well.  Even in a world without laws, criminals incur opportunity costs for the time and energy devoted to their activity.  The rational criminal will choose to pursue crime as long as its marginal benefit (MB) covers its marginal cost (MC); that’s Q0 in the graph below.

Of course, the most important costs are external. The victims of crime pay a heavy price as does society at large.  Crime burdens us all with the cost of police, prisons, courts, security devices and, perhaps worst of all, with fear.  Because criminals largely ignore these costs to others, they produce inefficiently high amounts of crime.  Should I steal $100 from you?  If the MB of theft to me is $100 and I estimate my MC as $80; I’ll steal.  The theft creates $20 of net value for me.  My $20 gain is more than offset by your $100 loss, however that does not necessarily enter my calculations.  The thievery is not socially efficient, but my private gains outweigh my private costs.

Using the model

What should we do?  The obvious choice is to shift the perceived costs and benefits of crime.  That is precisely what the criminal justice system is designed to do.  By imposing penalties we raise the MC of crime to its suppliers and decrease the equilibrium quantity.

More interestingly, if this economic view is correct, then differences in crime across time or across individuals should be directly attributable to differences in the costs and benefits of crime.  After a steady rise during the 1950’s, 60’s, 70’s and 80’s criminal activity fell significantly in the 1990’s. This drop occurred in almost every category of crime in every part of the nation.  According to economist Steven Levitt, shifts in costs and benefits tell much of the story.1

Levitt argued that the strong economy of the 1990’s explains a small but significant part of the decline.  Gainful employment can be considered a substitute for crime. If we flood the market with Pepsi, we lower the demand for Coke.  Similarly, if we flood the market with jobs, we lower the demand for crime.  As the economy grew and the job opportunities multiplied, potential criminals shifted from illegal to legal means of support.

Levitt also suggested that crime rates fell because we stuffed huge numbers of Americans into jails.  The incarcerations impact crime both by removing potential criminals from the streets and by acting as a potential deterrent to others.  By 2011 the U.S. led the world in terms of the percent of its population behind bars.  We imprisoned 716 per 100,000 residents; a rate more than five times higher than the median rate across the world.  As a comparison, the rate in Canada was 114 prisoners per 100,000 and rates in Western Europe ranged from 60 (Finland) to 149 (Spain).2

Levitt identified a third and far more controversial factor for decreased crime rates: legalized abortion.  Citing evidence that unwanted children are more likely to commit crimes, he argues that legalized abortion has cut the number of unwanted children.  The historical timing is what we should expect. Roe vs. Wade, the landmark case that led to increased abortion, was settled in 1973.  If abortion cuts crime, the effect should begin to appear about 18 years later; this is exactly what happened.  Moreover, differences in abortion rates across states correlate with differences in crime rates in subsequent years. Levitt does not tout abortion as a particularly efficient weapon against crime; he clearly prefers additional police officers for that. He merely notes that increased abortion accounts for a significant chunk of the slowdown in criminal activity.

While all of these potentially matter, subsequent research finds that other factors must also be at play.  For example, if job opportunities matter most, crime rates should have reversed when the unemployment rate soared after 2007.  But they did not; they continued to fall.  Increased incarceration might have mattered in the U.S., but it cannot explain why crime rates also fell in almost every developed nation, including countries such as Canada and the Netherlands where prison populations were falling.  And, similarly, changes in U.S. abortion laws cannot account for crime drops elsewhere in the world.

Changes in MC of crime seem to have been more critical.  First,  there were significant increases in the number of police officers on the streets.  While this was expensive, the benefits seem to have exceeded the costs by a wide margin. Levitt’s research fingers increased police presence as being the more cost effective approach.  He estimates that dollars spent on police protection create five times more bang for the buck than dollars spent on imprisonment.4  Second, not only has police presence increased, law enforcement tactics also have become more efficient.  Computerization now allows police to identify “hot spots” where their presence is likely to have larger impacts and technologies such as DNA testing and cell phone tracing have enabled them to build evidence more effectively.  Private anti-crime efforts also have increased dramatically.  Gated communities, home security systems and private security guards have burgeoned, and robbers that typically used stolen cars for their getaways now are thwarted by car alarms and circuits immune to hot-wiring.5  Pillaging and plundering simply is more difficult and costly than in the past.       _________________________________


  1. For an overview of his work, see Levitt, Steven D., “Understanding Why Crime Fell in the 1990’s,” Journal of Economic Perspectives, Winter 2004, volume 18, number 1, pp. 163-190.
  2. Data are collected by the International Centre for Prison Studies. See
  3. Levitt, op. cit., pp. 181-183. Not all researchers agree.  For an opposing view see Joyce, Ted, “A Simple Test of Abortion and Crime”, Review of Economics and Statistics, volume 91, number 1, February 2009, pp. 112-124.
  4. Ibid., p. 179.
  5. “Where have all the burglars gone?”, The Economist, July 20, 2013, pp. 21-23.


Testing Yourself

To test your understanding of the major concepts in this reading, try answering the following:

  1. Describe the benefits and costs of crime; explain why in a world without laws individuals will create more crime than is socially efficient.
  2. Explain why Levitt thought that strong economic growth, incarceration and abortion were factors in why U.S. crime rates fell in the 1990’s.  Also explain why these cannot be the entire explanation.
  3. Explain how the MC of crime has changed in recent decades and why this probably has been the most critical factor in declining rates of criminal activity.