The Deterrent Effects of California's Proposition 8: Weighing the Evidence

Whether, and the extent to which, stiffer sanctions deter crime is an integral question in fashioning corrections policy. To the extent that the behavior of the criminally inclined is completely insensitive to variation in punishment severity, crime-control policy should focus on incapacitation. Sentences should be long enough to ensure that offenders are removed from society during the period of their lives when they are most criminally active, but not so long as to be incarcerating individuals at a high cost to society who no longer pose a threat. On the other hand, if potential offenders take into account the severity of punishment when deciding whether to commit a felony (presumably, moderating their behavior in the face of stiffer sanctions), harsher sentencing may reduce overall offending, in part obviating the need to incapacitate those on the margin between offending and not offending. In the presence of such deterrence effects, optimal sentence lengths may extend beyond incapacitating offenders during the criminally active portions of their lives, as stiffer sentences will preclude some potential offenders from entering the system in the first place. In practice, creating an effective and efficient sentencing regime requires precise knowledge of the importance of the deterrent effects of sentence length. Such knowledge, however, is hard to come by as empirical research on deterrence has faced several fundamental challenges that have been difficult to overcome. First, the cross-sectional as well as longitudinal policy variation in sentence regimes often used to study deterrence is likely to be endogenously related to crime rates, a factor that could bias statistical inference in either direction. For example, if states with more severe crime problems enhance sentences as a result, cross-sectional analysis of the relationship between crime and sentencing severity is likely to reveal a positive relationship between sentence severity and overall crime. Alternatively, longitudinal analysis of crime rates within states may erroneously conclude that sentence enhancements reduce crime if states introduce harsher sentences in response to transitory increases in crime rates.