Ann Dryden Witte

Estimating the Economic Model of Crime With Individual Data

The paper develops a variant of the economic model of crime. The model developed is estimated using information on the post-release activities of a sample of men released from the North Carolina prison system. Both the expected certainty and severity of punishment are found to deter criminal activity in a number of instances and a 1 percent increase in certainty is generally found to have a greater effect than a similar increase in severity. Certainty of punishment is found to have a greater effect for relatively minor offenders and severity for persons offenders. While higher legal wages are found to deter, their effect is quite weak.

Predicting criminal recidivism using 'split population' survival time models

Abstract In this paper we develop a survival time model in which the probability of eventual failure is less than one and in which both the probability of eventual failure and the timing of failure depend (separately) on individual characteristics. We apply this model to data on the timing of return to prison for a sample of prison releases, and we use it to make predictions of whether or not individuals return to prison. Our predictions are more accurate then previous predictions of criminal recidivism. The model we develop has potential applications in economics; for example, it could be used to model the probability of default and the timing of default on loans.

Social Policy Experimentation A Position Paper

We review the argumentsfor and against randomized field experiments design to address important questions of social policy. Based on this review, we make a number of recommendations about how the use of randomized field experiments might be fostered.

Is Rehabilitation Dead

This paper examines rising crime rates, findings on the relative effectiveness of deterrence and rehabilitation, and the rise of a civil rights movement in correction, all of which have led to greater emphasis on deterrence and incapacitation and less emphasis on rehabilitation in correctional philosophy and practice. The conclusiveness of the findings that nothing rehabilitates and deterrence works is questioned. A more careful reading of existing evidence leads to no decisive conclusion on the relative effectiveness of these two philosophies of correction. Much of the failure of rehabilitative programs to date stems from programs limited in duration and quality and evaluated for their ability to alter lifestyles dramatically; many that have been shown to be failures are the result of an inadequate tailoring to offender problems. The great increase in economic crimes points to the need to improve the economic opportunities of offenders rather than altering personality. The civil rights movement in correction alerts us to the need for curbs on certain types of rehabilitative programs, but it should not force us to abandon all attempts at rehabilitation. Adherence to a strict deterrence philosophy because of its economic and humanistic cost is questioned. Finally, a more careful application of rehabilitative programs and their continued use are called for as one approach to the crime problem.

Experimental Assessment of the Effect of Vocational Training On Youthful Property Offenders

In this article, we report results that suggest that carefully integrated and implemented vocational training and reentry programs for youthful property offenders can reduce the rate at which such individuals are arrested after release. This result is important since most evaluations of programs for such offenders show no significant effects. The question has been Why have programs rarely been shown to have significant effects on the behavior of offenders? Our results suggest that the major reasons may be that programs evaluated to date have been weak and implementation has been poor. Even with substantial backing from correctional management only 16% of the experimental group participated in all aspects of the Vocational Delivery System (VDS). Members of the experimental group were most likely to participate in early aspects of the VDS (e.g., a three-week evaluation of vocational interests and aptitudes) than in later elements (e.g., work with the Employment Security Commission to find a job). Even with relatively weak implementation, the experimental group subjects were significantly less likely to be arrested than were control group subjects.

Socially optimal and equilibrium distributions of office activity: Models with exogenous and endogenous contacts

Abstract Two models of the location of office activities within a city are developed. In the first model, the pattern of interoffice contacts is specified exogenously as has been traditional in the literature. In the second model the contact patterns as well as the distribution of firms are endogenous. For both models, the market equilibrium distribution of office activities is not socially optimal. Two sources for this market failure are identified in these models: (1) agglomeration economies and (2) transactional externalities arising from firm interdependence and mobility.

Evaluating Correctional Programs Models of Criminal Recidivism and an Illustration of Their Use

In this article, we develop a model of the length of time until recidivism, and estimate it on a sample of releasees from the North Carolina prison system. Evidence of the models predictive accuracy is presented. The model is then used to evaluate a correctional program by comparing the actual time until recidivism of program participants to that predicted by the model. The program appears to be effective only in the very short run.

The Proportional Hazards Model

In this chapter and Chapter 7, we will consider models of the length of time until recidivism that contain individual characteristics as explanatory variables. The models of Chapter 7 will be parametric models in the sense that they will assume a particular distribution for the survival times; for example, we will estimate a model based on the exponential distribution, in which the mean of the distribution depends on individual characteristics. In this chapter, on the other hand, we will consider the proportional hazards model of Cox (1972). This model allows one to estimate the effects of individual characteristics on survival times without having to assume a particular form for the distribution function (or the density or hazard). Instead, the distribution of the survival times is estimated in a nonparametric way. This model has been used a good deal recently in the study of recidivism. For example, see Barton and Turnbull (1981), Rhodes and Matsuba (1985), Sherman and Berk (1984), and Witte et al. (1982).

Provision of Child Care: Cost Functions for Profit-Making and Not-for-Profit Day Care Centers

In this paper, we explore and compare the operation of profit making (PMOs) and not-for-profit (NPOs) day care centers using data from a random sample of Massachusetts day care centers. We find many differences in the characteristics of PMOs and NPOs. Most interestingly, NPOs pay significantly higher wages than PMOs although the staff of NPOs has aproximately the same education and less experience than the staff of PMOs. NPOs also have significantly more staff per child than PMOs. We estimate separate cost functions for PMOs and NPOs and find the parameters of these functions to be insignificant different imiplying that the higher observed costs for NPOs result from different input (e.g., paying higher wages) and output (e.g., more interaction and larger size) choices and not from differences in methods of operation. Finally, we find that current monetary and inkind subsidies have no significant effect on the costs of day care centers.

Tax Compliance: An Investigation Using Individual Taxpayer Compliance Measurement Program (TCMP) Data

In this paper, we analyze the tax compliance behavior of U.S. taxpayers by using a 1979 data set that combines information from a random sample of individual tax returns, each of which has been thoroughly audited, IRS administrative records, and sociodemographic data from the Census. We find evidence that both audits and tax code provisions affect compliance. However, the effects are significant for only low-and high-income groups. Interestingly, work on the underground economy also suggests that it is these two groups that are least compliant. Our results for audits suggest that the “ripple” or general deterrent effect of audits may be substantially larger than the direct revenue yield of audits for high-income taxpayers. Our results for allowable subtractions from income imply that the 1986 Tax Reform Act changes that lowered allowable subtractions may have procompliance effects.