Evaluation Research

Recidivism

Recidivism would seem to be an obvious focus for evaluation studies of teen courts, but only a handful of evaluations have measured postprogram recidivism (see table 5). Most studies have relied on court records and official police data to detect recidivism. Few studies have attempted to collect personal data from teen court defendants. Only Swink's (1998) study of a teen court program in Onondaga County, NY, measured self-reported recidivism, and it relied on parents to report the illegal activities of their children.

Of the few studies that measured official recidivism, some found very low rates of reoffending among former youth court defendants. Several researchers found rates of postprogram recidivism that ranged from 3 to 8 percent within 6 to 12 months of appearance in teen court (Butler-Mejia, 1998; McNeece et al., 1996; SRA Associates, 1995). A few studies found recidivism rates in excess of 20 or 30 percent. One Texas study, for example, found that 24 percent of former youth court participants reoffended (Hissong, 1991). Minor and his colleagues found that nearly one-third (32 percent) of teen court youth reoffended within 1 year (Minor et al., 1999). It is not possible to say whether these higher rates are anomalies. Existing teen court evaluations are so different in scope and design that it is often impossible to compare the findings of one with another.

Most evaluations of teen court recidivism have employed relatively simple research designs. Even some of the best studies (Minor et al., 1999; LoGalbo, 1998; Swink, 1998; Wells, Minor, and Fox, 1998) have relied on data from a single group of teen court cases at a single point in time. Often, researchers have failed to use comparison groups or pre- and postmeasures. Thus, it is impossible to test the assumption that recidivism outcomes are due to teen court rather than to other factors (e.g., the type of youth selected for teen court may be unlikely to recidivate).

Only three published studies (Hissong, 1991; North Carolina Administrative Office of the Courts, 1995; Seyfrit, Reichel, and Stutts, 1987) have used reasonably appropriate comparison groups to measure the possible effects of teen courts on recidivism (see table 5). Hissong's evaluation of an Arlington, TX, teen court compared recidivism among teen court defendants with a group of non-teen-court participants matched on sex, race, age, and offense. The analysis suggested that teen court participants were significantly less likely to reoffend than the comparison group (24 percent versus 36 percent). Several important elements of the study, however, were poorly documented. The definition of recidivism used in the analysis (presumably rearrest) is unclear. The duration of the followup period is not described (subjects may have had different periods of risk), and there is a range of unexplored potential differences between the treatment group and the comparison group.

The North Carolina study used a comparison group that consisted of 97 cases diverted by police during a 6-month period prior to implementation of the teen court in Cumberland, NC. Researchers hypothesized that these youth would have been referred to teen court had the program been in existence. Teen court and comparison group cases were matched using several factors, including demographic characteristics and offense type, and researchers tracked the recidivism of both groups. The study failed to find statistically significant differences in the recidivism of the two groups. In fact, the analysis seemed to favor the comparison group. After 7 months, 20 percent of teen court participants had reoffended, compared with just 9 percent of the comparison group. The study also found little difference between the two groups in average time before a new offense (4.1 months for teen court offenders versus 4.6 months for the comparison group). Youth who successfully completed the teen court program were less likely to reoffend than were youth who began but failed to complete the program (11 percent compared with 42 percent), but this finding may reflect the greater tendency of low-risk youth to complete the program.

Seyfrit and her colleagues (1987) tracked recidivism outcomes for 52 youth referred to a Columbia County, GA, teen court during an 18-month period in the early 1980's. They also collected data for a comparison group of 50 youth matched on demographics and offenses. The study found little difference between the two groups. Although 12 percent of the comparison group recidivated during the followup period, the same was true for 10 percent of the teen court defendants. Like the North Carolina study, the Seyfrit study was unable to control statistically for different periods of opportunity to reoffend. The followup periods ranged from 6 to 18 months, which reduced the researchers' ability to infer any real differences in the recidivism of the two groups.

Other Outcomes

Several studies have suggested that teen courts may have effects on youth other than reduced recidivism. These potential benefits include client satisfaction with the teen court experience (Colydas and McLeod, 1997; McLeod, 1999; Reichel and Seyfrit, 1984; Swink, 1998; Wells, Minor, and Fox, 1998), enhanced perceptions of procedural justice (Butler-Mejia, 1998), improved attitudes toward authority (LoGalbo, 1998; Wells, Minor, and Fox, 1998), and greater knowledge of the legal system (LoGalbo, 1998; Wells, Minor, and Fox, 1998).¹

For example, McLeod's (1999) survey of former teen court participants found that at least 90 percent of youth referred to the Colonie (NY) Youth Court during 1997 and 1998 believed that the experience increased their understanding of the legal system, helped them improve their behavior, and helped them become more responsible. Nearly all survey respondents (95 percent) reported that going through teen court caused them to "make more thoughtful decisions." Nearly three in five (58 percent) reported better communication with their parents, and half (50 percent) reported improved grades in school. However, the study's very low response rate (24 percent of youth surveyed) raised the possibility that the youth responding to the followup survey may have been the most compliant and prosocial youth in the sample.

LoGalbo's (1998) evaluation of the Sarasota County, FL, teen court program also found that teen court positively affected defendant attitudes toward authority and understanding of the legal process. LoGalbo surveyed 111 youth immediately after their initial interview with teen court staff and again upon completion of the program. The survey asked participants about their knowledge of Florida laws and the justice system, their attitudes toward nine authority figures (e.g., police officer, judge, parent, teacher), their attitudes toward teen court and toward themselves, and their perception of the fairness of teen court procedures. The study found teen court participation was associated with increased self-esteem and positive attitudes toward select authority figures (e.g., judges). The analysis also suggested that recidivism was less likely among defendants with improved attitudes toward authority figures.

Strong client satisfaction was also reported by researchers in Kentucky. Exit interviews conducted by Wells and colleagues (1998) revealed high levels of satisfaction among 123 teen court participants, with 84 percent indicating that their sentences were fair. Several positive features of the teen court experience were cited by the Kentucky subjects, including "educational advantages" (37 percent) and the actual sentences youth received (21 percent). Teens also consistently indicated that the opportunity to serve as a teen court juror was an important, positive aspect of the teen court process.