The Hamilton County Drug Court: Outcome Evaluation Findings
Final Report
Introduction | Methods | Profile | Treatment Considerations | In-Program Behavior | Subsequent Criminal Behavior | Subsequent Criminal Behavior Among Graduates | Conclusions and Implications | Report Authors
The main purpose of an outcome evaluation is to determine the impact of the intervention, in this case the drug court, on behavior. The commonly used measure of behavior is recidivism. Hence, the purpose of this section is to address the recidivism rates between the two groups and identify factors associated with outcome. A number of research questions are examined. What are the rearrest rates among the experimental and comparison groups? What is the rate of time to elapse between initial arrest to rearrest among the two groups? What are the determinants or factors associated with recidivism?
A statistically significant difference exists with regard to rearrest for a new offense. Figure 14 presents the data with regard to rearrest. The data indicate that offenders in the treatment group were less likely to be rearrested than those in the comparison group. Specifically, table 15 reveals that 29 percent of the experimental group was rearrested in comparison to 39 percent of the comparison group.
Table 15 also indicates that both groups were likely to be arrested on a drug related offense. Specifically, of those arrested, 47 percent of the experimental group and 44 percent of the comparison group were arrested on a drug charge. In addition, 22 percent of the experimental group and 26 percent of the comparison group were arrested for a theft. Although not significant, a slightly higher number of the treatment participants were likely to be arrested for a felony charge and have their charge result in a conviction.
In contrast, table 16 reveals that significantly more individuals in the control groups were likely to be arrested multiple times during the follow-up period. Specifically, Figure 15 reveals that 30 percent of the experimental group was arrested multiple times during the follow up period in contrast to 47 percent of the comparison group. Finally, table 16 illustrates that comparison group participants, on average, were rearrested 1.87 times.
Evaluations that explore only the number of arrests of two or more groups may overlook important treatment effects. Specifically, participation in treatment may only delay onset of reoffending. To explore this possibility, we compare the rate at which drug court participants and the comparison group participants were rearrested. Table 16 indicates that both the groups are similar with regard to the average number of days to elapse between initial arrest and subsequent rearrest. Specifically, table 16 illustrates that the typical ADAPT participant was likely to remain arrest free for the first 205 days. In comparison, the typical comparison group participant was likely to remain arrest free for 218 days. It appears, then, that participation in the treatment program does not represent a period of delay. In fact, those participating in treatment are significantly less likely to be rearrested regardless of the follow-up period in comparison to the control group.
This section is designed to isolate factors associated with recidivism. The analysis will allow the court to discern what type of clients benefit from services and for whom additional services are needed. Logistic regression was used to determine which factors were associated with recidivism. The characteristics included in the model were group membership (i.e. experimental or control group), race, gender, marriage, education level, prior record, employment, age, and time at risk. Time at risk was included as a control for the different lengths of time clients were followed.
Analysis of the model revealed that five variables predicted who is likely to be rearrested: group membership, race, prior record, age, and time at risk. Specifically, the data indicate that comparison group participants were more likely to be rearrested. The data also indicate that African Americans were more likely to be rearrested. In addition, it was found that individuals with a prior record are more likely to continue engaging in criminal behavior. The data also indicate that the younger the individual, the more likely he or she may be rearrested. Finally, the longer an individual remained at risk, that is the longer the time to elapse between initial contact with the court to the end of the follow-up period, the more likely a rearrest will occur.
In order to examine the results in more detail, the five variables predictive of rearrest were translated into log-odds ratios to simple odds. Failure rates for each of the significant factors were estimated from the odds. Figure 16 and Figure 17 present the estimated probabilities and delineate the percentage with which each factor has in predicting arrest. Those individuals who are a member of the control group, African-American, have a prior record, are between the ages of 18-22, and remain at risk longer have a 63% probability of being arrested for a new offense.
The purpose of this section was to address the recidivism rates of the two groups and identify factors associated with recidivism.
What are the rearrest rates among the experimental and comparison groups?
* With regards to rearrest, the experimental or ADAPT group was significantly less likely to be rearrested in comparison to the control group that did not receive treatment.
* Of those rearrested, the typical offender in both groups was charged with a drug related offense.
* Slightly more experimental participants were rearrested for a felony in comparison to control group participants
* Significantly more experimental group participants were convicted of the offense in comparison to control group cases.
* The comparison group participants were more likely to be rearrested multiple times.
What is the length of time to elapse between initial arrest to rearrest between the two groups?
* The length of time to elapse was similar for both groups. Data indicate that treatment participation did not simply delay subsequent criminal behavior.
What are the determinants or factors associated with recidivism?
* Five variables predicted rearrest: group membership, race, prior record, age, and time at risk.
• 65 South Front Street Columbus, Ohio 43215-3431 •