Drug courts constitute a clear paradigm of an integrated public health-public safety strategy that has shown promise for reducing drug use and recidivism among probationers and pretrial defendants. Drug courts are separate criminal court dockets that provide judicially supervised treatment and case-management services for drug offenders in lieu of prosecution or incarceration.

The core components of a drug court typically include:

Common examples of negative sanctions include:

Common examples of rewards include:

Counseling requirements may also appropriately be decreased when the client complies well with treatment or increased if he or she has poor attendance or participation or other problems. Clients who satisfactorily complete the program may have their current criminal charges dropped or may be sentenced to time served in the drug court program. Defendants are generally required to plead guilty or "no contest" as a precondition of entry into drug court. Therefore, termination from the program for noncompliance ordinarily results in a criminal drug conviction and sentencing to supervised probation or incarceration.

The evidence is clear that drug courts can increase clients' exposure to treatment. Reviews of nearly 100 drug-court evaluations concluded that an average of 60 percent of drug court clients completed a year or more of treatment, and roughly 50 percent graduated from the program (Belenko, 1998, 1999, 2001). This compares favorably to typical retention rates in community-based drug treatment programs where, as noted, more than 70 percent of clients on probation and parole drop out of drug treatment or attend irregularly within 3 months and 90 percent drop out in less than 1 year.

Promising, although less definitive, is the evidence with regard to the effects of drug courts on drug use and crime. Two experimental studies have compared outcomes between participants randomly assigned to either drug court or a comparable probationary condition. In one study, the Maricopa County (Arizona) Drug Court was found to have had no impact on rearrest rates 12 months after admission to drug court (Deschenes et al., 1995). However, a significant "delayed effect" was detected at 36 months, at which time 33 percent of the drug court participants had been rearrested, compared to 47 percent of subjects in various probationary tracks (Turner et al., 1999).

Similarly, in a randomized study of the Baltimore City Drug Treatment Court, 48 percent of drug court clients and 64 percent of adjudication-as-usual control subjects were rearrested within 1 year of admission (Gottfredson and Exum, 2002). At 2 years postadmission, 66 percent of the Baltimore drug court participants and 81 percent of the controls had been rearrested for some offense, and 41 percent of the drug court participants and 54 percent of the controls had been rearrested for a drug-related offense (Gottfredson et al., 2003).

Nearly 100 quasi-experimental evaluations have compared outcomes between drug court participants and those of nonrandomized comparison groups. In the majority of those evaluations, drug court clients achieved significantly greater reductions-differences of approximately 20 to 30 percentage points during treatment and 10 to 20 percentage points

than did individuals on standard probation or intensive probation (Belenko, 1998, 1999, 2001). The magnitudes of the posttreatment effects are comparable to the 15 percentage-point reduction in recidivism obtained in the two experimental studies reviewed above.

It is important to note, however, that many drug court evaluations have used systematically biased comparison samples, such as offenders who refused, were deemed ineligible for, or dropped out of the interventions. This may have led to an overestimation of positive outcomes for drug court clients in some studies because the comparison subjects are likely to have had more severe criminal histories or lower motivation for drug abuse treatment from the outset. Further, most of the studies evaluated outcomes only during the course of drug court or up to 1 year postdischarge, and hardly any studies have assessed substance-use outcomes after discharge. Thus, it remains largely unknown how drug court clients generally fare after the criminal justice supervision ends.

These limitations in the extant research on drug courts led the congressional General Accounting Office (GAO) to conclude there are insufficient data available to gauge the effectiveness of federally funded drug court programs in this country (GAO, 2002). In response to the GAO report, the National Institute of Justice (NIJ) released a request for proposals for long-term client-impact evaluations of up to 10 drug courts that will include