Scholars have published over 100 empirical studies in the last three decades testing for the deterrent effect of punishment on criminal activity. Most support the existence of a deterrent effect for both imprisonment risk and longer prison time.

• A National Academy of Sciences panel estimated a few years ago (1993) that a 50 percent increase in the probability of incarceration prevents twice as much violent crime as a 50 percent increase in the length of incarceration.

• More recently, using state data for 1940 and 1950, researchers found a 50 percent boost in the chance of imprisonment prevents three to five times as many murders as a 50 percent increase in prison time for murderers.

• A greater risk of execution (most states used the death penalty then) also deters murder.

A greater imprisonment risk discourages twice as many robberies as the same boost in prison time, while risk and time have about the same impact on aggravated assault.

Source: Isaac Ehrlich and Zhiquiang Liu, "Sensitivity Analyses of the Deterrence Hypothesis: Let's Keep the Econ in Econometrics," Journal of Law and Economics, April 1999.