This week, we looked at the effect of prison on crime.
The rational choice model predicts that tougher sentencing laws should reduce crime by increasing the severity of punishment, and research suggests that the recent rise in the prison population was one of the important factors for the historic crime drop in the United States during the 1990s.
Tougher sentencing laws should reduce crime by preventing prisoners from committing new crimes while incarcerated (incapacitation effect) and deterring would-be criminals from carrying out the crime (deterrent effect). It is usually difficult to empirically quantify these two effects without using a randomized experiment, but some economists overcame this problem by using clever empirical strategies and found that the magnitudes of both incapacitation and deterrent effects were sizable. At the same time, researchers also worry about adverse effects of incarceration. Prison experience may adversely influence offenders’ future job prospects and aggravate their criminal risks. Children suffer from heavy emotional and financial losses while their parents are incarcerated.
This week, we also saw how we could use regression discontinuity analysis to study the effect of more severe punishment on criminal behavior. When the severity of punishment sharply rises at a threshold age, we can compare the offending rate between those who are just below and above the threshold age and credibly view the difference in offending rates as the causal effect of more severe punishment. (Researchers usually implement difference-in-differences and regression discontinuity analyses by running some form of linear regression we saw in Week 2, but describing exactly how to run these regression analyses would be out of scope of this course. Interested learners should refer to an econometrics textbook.)
Police and prison are two main crime-control policy tools used by governments, but as we saw from Weeks 1 and 2, the potential gains and costs from crime can be influenced by many other government policies. Next week, we will see how effective crime-control policies may involve more than just law enforcement, judiciary, and correctional systems. Crime rates may be influenced by the quantity and quality of public education, government regulations on alcohol and drug use, and public health and environmental policies. We will see existing research evidence on how these policies may “indirectly” affect crime rates.
See you next week.
© Songman Kang, Hanyang University