Despite a growing consensus among scholars that substance abuse treatment is effective in reducing offending, strict eligibility rules have limited the impact of current models of therapeutic jurisprudence on public safety. This research effort was aimed at providing policy makers some guidance on whether expanding this model to more drug-involved offenders is cost-beneficial. A goal of this project was to determine the size of the drug-involved offender population that could be served effectively and efficiently by partnerships between courts and treatment.
Since data needed for providing evidence-based analysis of this issue were not readily available, micro-level data from three nationally representative sources were used to construct a 40,320 case synthetic dataset -- defined using population profiles rather than sampled observation--that was used to estimate the benefits of going to scale in treating drug-involved offenders. The principal investigators combined information from the NATIONAL SURVEY ON DRUG USE AND HEALTH, 2003 (ICPSR 4138) and the ARRESTEE DRUG ABUSE MONITORING (ADAM) PROGRAM IN THE UNITED STATES, 2003 (ICPSR 4020) to estimate the likelihood of drug addiction or dependence problems and develop nationally representative prevalence estimates. They used information in the DRUG ABUSE TREATMENT OUTCOME STUDY (DATOS), 1991-1994 (ICPSR 2258) to compute expected crime reducing benefits of treating various types of drug-involved offenders under four different treatment modalities. The four different treatment modalities were:
- Long-Term Residential (LTR) Treatment
- Short-Term Inpatient (STI) Treatment
- Outpatient Methadone Treatment (OMT)
- Outpatient Drug-Free (ODF) Treatment
The project computed expected crime reducing benefits that were conditional on treatment modality as well as arrestee attributes and risk of drug dependence or abuse. Moreover, the principal investigators obtained estimates of crime reducing benefits for all crimes as well as select sub-types. More specifically, in order to create the dataset, the principal investigators defined a detailed exhaustive set of client attributes (based on availability across the three data sources) as a base data set. The attributes that defined the synthetic database included all characteristics available for all the databases used in this study. A synthetic dataset of 40,320 client profiles with the attributes and categories that include age, race, gender, current offense, criminal history, drug treatment history, history of violence, current criminal justice status, co-occurring alcohol problem, and geographic location was created. The base synthetic dataset was a cross combination of all of these attributes. Each profile in the dataset was a list of attribute combinations that represented one row in the dataset. The next step was to determine how many individuals in the population were represented by each profile. For each profile, a value was estimated that was used for weighting the profile by its prevalence in the population of interest. Moreover, since the project was interested in assessing the prevalence of potential clients at risk of drug dependence or abuse, prevalence estimates were generated conditional on drug dependence or abuse. Independently of the profiles' prevalence, the project also generated estimates of the expected crime reducing benefits of treating potential clients under various treatment modalities. This information was obtained from DATOS, and the principal investigators interpolated all expected outcomes of treatment onto the profiles in the synthetic dataset.
The data are not from a sample. To construct the sample of 40,320 profiles, the principal investigators defined a potential client of a therapeutic jurisprudence program as any arrestee who was probably guilty of a crime and was also at risk of substance abuse. To determine the nature of the arrestee's substance use, they used the DSM-IV criteria to assess whether or not a drug involved offender was at risk of being dependent on drugs or of abusing drugs. The principal investigators also included a set of client attributes (e.g., age, race, gender) as traits to be included in the design of the synthetic data, e.g., as attributes along which dependency may vary. The analyses were performed on profiles that represent types of individuals, rather than on individual observations. Because information was combined from multiple data sources, it was not possible to observe information about the same individuals describing their risk, dependence, and responsiveness to treatment. Since it was possible to observe common attributes -- socio-demographic characteristics and information about substance abuse and criminal histories across datasets -- simulated (synthetic) profiles were created. These profiles, which are combinations of all attributes, were then used in the analysis in much the same way individual observations would be used had that information been available. From these multiple data sources, the principal investigators developed an exhaustive set of "profiles," where each profile was one possible combination of all of the potential client's attributes. Hence, the synthetic data used in this analysis included a profile for every client permutation, allowing the project to quantify the effect of drug treatment on every combination of client attributes and characteristics. Simulation models were used to estimate outcomes for each profile, conditional on receiving each of four possible treatment modalities. Data were used from studies linking client traits and characteristics to outcomes to identify expected outcomes for each of these profiles.
All individuals arrested in the United States in 2003.
DRUG ABUSE TREATMENT OUTCOME STUDY (DATOS), 1991-1994 (ICPSR 2258)
ARRESTEE DRUG ABUSE MONITORING (ADAM) PROGRAM IN THE UNITED STATES, 2003 (ICPSR 4020)
NATIONAL SURVEY ON DRUG USE AND HEALTH, 2003 (ICPSR 4138)
Variables include age, race, gender, offense, history of violence, history of treatment, co-occurring alcohol problem, criminal justice system status, geographic location, arrest history, and a total of 134 prevalence and treatment effect estimates and variances. More specifically, the prevalence and treatment effect variables include number of arrestees estimate and variance, arrestees at risk of dependence estimate and variance, and arrestees at risk of abuse estimate and variance. Furthermore, variables include estimates and variances for all crimes averted, drug crimes averted, frauds averted, burglaries averted, larcenies averted, robberies averted, aggravated assaults averted, and simple assaults averted for modalities 1-4 at risk of abuse and at risk of dependence.