|There is great enthusiasm for crime prevention as a supplement or alternative to ex post incarceration. Cost-effectiveness evaluations have indicated that prevention programs might compare favorably with incarceration in terms of crimes averted per program dollar invested and that they might pay for themselves with savings or "cost-offsets". Until recently, however, the evaluations have not addressed in any comprehensive way the implications of the uncertainties inherent in their conclusions because of data and methodological limitations. This paper examines that uncertainty, addressing specifically the following issues:
* Data supporting the analyses are limited. What happens if they are wrong? The specific benefit and cost estimates would, of course, hange, but would they change enough to alter the overall conclusions?
* What can be said about whether programs can be expanded to populations ess at risk without sacrificing the net savings they achieve?
* Because estimates are uncertain, there is some possibility that a pogram expected to have a favorable benefit-cost ratio will instead ail. If policymakers wish to hedge against such failures, what program nvestment strategy should they pursue?
(Return to Program Resources)