Program Outcomes: Completion Rates and Recidivism

Client completion rates for the different types of programs are shown in Table 12. The numbers, on the surface, appear to conflict with prior research that shows that completion rates generally run higher than those listed in table 12, and that formal programs usually are more successful than informal programs (Schneider and Schneider, 1984). The findings, however, are not comparable. First, the data in table 12 reflect the number of clients who completed restitution during 1990, divided by the number of clients in restitution programs during 1990. Many cases doubtless were still open at the time of the survey, and as they are closed, the successful completion rate will of course increase. Second, the distinction between "formal" and "informal" programs in the survey questionnaire was strict. Recall that only about 60 percent of the programs met the definition, and that 15 percent of those surveyed (about 83) lacked only one of the three components of the definition. A small shift in programs between the two categories could have a large effect on thepercentages of successful completions.

Similar cautions apply to Table 13 and Table 14, which show, respectively, the average amounts of restitution completed overall and the average amounts completed by type of client. To reiterate, some of the cases on which these numbers are based still were open at the time the data were collected, and the rates undoubtedly will improve when those cases are closed. Still, the numbers are impressive: as shown in table 13, the average program collected $80,850 in 1990, as compared with $138,715 ordered; this is a 1-year completion rate of about 58 percent. When multiplied by the number of programs identified in the survey and the average number of clients, these figures suggest that of approximately $72.3 million in restitution ordered during 1990, approximately $44.5 million was collected. Similarly, the community service completion rate (a much higher 88 percent) indicates that about 17.1 million hours of community work service was performed during 1990, along with 44,000 hours of direct victim service. Clearly, restitution has become a major industry.

Table 14 provides the average amounts of restitution, community service, and victim service ordered and completed by the types of clients accepted. While the completion rates are higher among juveniles than among adults, the amounts ordered are typically smaller. For example, the average referral to a program that accepts juvenile offenders paid only 85 percent of the financial restitution ordered, as compared with 47 percent by the average referral to adult-only programs, but the juvenile would have paid only about one-third as much as the adult. The figures are about the same for community service. Oddly, the average number of victim service hours completed in programs that accept juvenile offenders appears greater than the average number of hours ordered. While it is tempting to speculate that the average juvenile offender working directly for his or her victim voluntarily works extra time to make restitution, the numbers -- reflecting only portions of hours on the average -- are too small to draw any conclusions.

Programs contacted by the 1991 RESTTA survey were asked to report the number of their clients who recidivated (committed another offense) after either completing the program successfully or failing to complete the program. These figures were then divided by the total number of clients in each group (successful completions and failures) to calculate a recidivism rate for each category. An overall recidivism rate was calculated for all programs and for programs that accept juvenile offenders only by adding the numbers in each category of completion and dividing by the total referrals in 1990. The results, listed by type of referrals accepted (juveniles or adults), are presented in Table 15.

Across all types of programs, persons who completed their requirements successfully -- by meeting all of the restitution and/or community service obligations -- recidivated at a lower rate than those who failed to do so. The difference is particularly striking in juvenile-only programs, but a strong relationship between successful completion and recidivism also exists in programs that accept both adults and juveniles. Both the direction and the magnitude of this relationship areconsistent with findings in prior research (Schneider and Schneider, 1984). It should be noted that the n sizes (numbers of programs responding) are different for successful completion and program failure across all categories of programs. Generally, programs appear to keep better data (including reoffense information) on clients who successfully complete requirements than on those who fail. If this is the case, then the actual differences in recidivism rates between the successful and unsuccessful groups may be even greater.