Introduction What is the optimal level of incarceration?

This website is an interactive extension of The Imprisoner’s Dilemma, which presents a cost-benefit analysis of three policy changes that impact incarceration. The paper draws on a large amount of recent empirical literature, on general deterrence, specific deterrence, incapacitation, costs of crime, value of freedom and prison costs. In order to produce the cost-benefit calculation, a large number of parameter values need to be chosen. This website allows the user to modify parameter values and calculate their influence on the net benefit. To get started, choose one of the policies below. I look forward to discussing your reactions in the policy-oriented and model-oriented comments sections.

David S. Abrams
Assistant Professor of Law, Business, and Public Policy
University of Pennsylvania Law School

Proposed Policy Changes change sentence lengths, reclassify crimes, release prisoners

Policy 1: Change Sentence Lengths

The first policy change is an increase or decrease of all sentences by a certain percentage, for example a proposal to increase all criminal sentences in a state by 10%. This policy change is not a realistic one, and its purpose is as a simple point of comparison with the other more realistic policy changes. The analysis of this proposed change can give insight into more realistic proposals where sentences may change according to offense type or offender criminal history. The complication of adding variation to each of the components of the cost-benefit calculation may then be invoked in examining the other policy changes.

Policy 2: Reclassify Crimes

The second type of policy change is one that reclassifies some crimes so that they either gain or lose eligibility for incarceration. An example of this type of change is an increase in the minimum threshold that delineates felony larceny from $500 to $1,000 (in a jurisdiction where misdemeanors may not be punished by incarceration). This decreased penalty for thefts of property valued between $500 and $1,000 leads to a lower prison population. Another common policy change of this type deals with “technical violations” – infractions that lead to a revocation of parole, but which are not otherwise crimes. Some states have recently changed rules so that failure of a single drug test while on parole does not trigger automatic incarceration. An additional type of policy change that falls within this category, but would lead to a larger prison population, is the addition of new substances to the category of illegal drugs.

Policy 3: Release Prisoners

The third type of policy change is a one-time release of prisoners. Unlike the previous two categories, this change only works in one direction: it decreases the incarcerated population. This has been a policy change that has been used in the past, albeit reluctantly, as a response to prison overcrowding. For example, Alabama released 1100 prisoners in 1981 in response to a court order.

The impact of a prisoner release will vary substantially depending on which prisoners are chosen for release. Some of the most common groups include those with minimal time remaining on their sentences, elderly or medically infirm prisoners, or those convicted only of relatively minor crimes. Inmates with minimal time remaining on their sentences are more likely to have been convicted of relatively minor crimes which may decrease the harm or likelihood of recidivism. However, they are also likelier to be younger than prisoner releases based on age. This is sometimes the preferred policy because of the well-known relationship between criminality and age: crime is a young man’s game. In all-known cases of prisoner releases, those released are highly non-representative of the criminal population. Prisoners chosen to be released early are chosen because of their lower likelihood of recidivism and lower likelihood of committing particularly costly crimes upon release.