Pretrial drug testing of arrestees and defendants released before trial is increasingly being implemented in state and local criminal justice agencies. The federal courts are also considering whether pretrial drug testing should be initiated in each district. But researchers and policymakers are currently engaged in a vigorous debate about the merits and drawbacks of pretrial drug-testing programs. Critics contend that the programs are difficult to implement, are too expensive, and have little real impact on criminal behavior and illegal drug use. Proponents argue that pretrial drug testing enables criminal justice officials to reliably detect drug use and improve release decisions, to effectively supervise drug-involved offenders before trial, and to encourage these offenders to seek treatment. This article reviews state and local experience with pretrial drug testing, discusses the evaluations that have been conducted, and offers some suggestions as to the role of pretrial drug testing in our nation's drug control policy.
Get full access to this article
View all access options for this article.
References
1.
1. U.S., Office of National Drug Control Policy, National Drug Control Strategy (Washington, DC: Government Printing Office, 1989), p. 26.
2.
2. John Carver, “Drugs and Crime: Controlling Use and Reducing Risk through Testing,”NIJ Reports, no. 199 (Sept.-Oct. 1986).
3.
Eric Wish and Bruce Johnson, “The Impact of Substance Abuse on Criminal Careers,” in Criminal Careers and “Career Criminals,” vol. 2, ed. A. Blumsteinet al. (Washington, DC: National Academy Press, 1986).
4.
Ko-lin Chin and Jeffrey Fagan, “Impact of Crack on Drug and Crime Involvement” (Paper delivered at the annual meeting of the American Society of Criminology, Baltimore, MD, 1990).
5.
5. Arrestees who are charged with minor crimes—for example, misdemeanors and traffic violations—are often released at a district police station. Since they are not brought to the central booking facility, they are not tested for drug use.
6.
Christy A. Visher and Richard L. Linster, “A Survival Model of Pretrial Failure,”Journal of Quantitative Criminology, 6:153-184 (1990).
7.
7. Visher and Linster, “Survival Model of Pretrial Failure,” p. 167, tab. 5.
8.
8. Toborget al., “Assessment of Pretrial Urine Testing,” p. 10.
9.
9. Ibid., pp. 13-14. Surprisingly, this finding was independent of whether those who showed up for testing were positive or negative for illegal drugs.
10.
10. Ibid., pp. 23-24. This type of evaluation is the strongest statistical test of whether regular drug testing during release might reduce pretrial misconduct. Such experiments, however, are difficult to carry out successfully in an actual criminal justice environment, and implementation problems usually occur.
11.
11. Christy A. Visher, “Using Drug Testing to Identify High-Risk Defendants on Release: A Study in the District of Columbia,”Journal of Criminal Justice, 18:321-332 (1990).
12.
Eric Wish , Mary Cuadrado, and Stephen Magura, “Drug Abuse as a Predictor of Pretrial Failure-to-Appear in Arrestees in Manhattan” (Final report submitted to the National Institute of Justice, U.S. Department of Justice, 1988).
13.
13. S. Belenko and I. Mara-Drita, “Drug Use and Pretrial Misconduct: The Utility of Prearraignment Drug Tests as a Predictor of Failure-to-Appear” (Manuscript, Criminal Justice Agency, New York, 1988). These results were based on a slightly different sample and different analytic methods from those of the analysis reported in the published study of the Manhattan data by Smith, Wish, and Jarjoura, “Drug Use and Pretrial Misconduct in New York City.” The Study by Belenko and MaraDrita analyzed only failure to appear and did not adjust for sample-selection effects or time at risk.
14.
14. John Goldkamp, Michael Gottfredson, and Doris Weiland, “Pretrial Drug Testing and Defendant Risk,”Journal of Criminal Law and Criminology, 81:585-652 (Fall 1990). The Dade County study appears to be seriously flawed, however. It is difficult to interpret the results because of the simultaneous inclusion of four measures of the drug test results—namely, positive for marijuana, positive for cocaine, positive for either, and positive for both—in the logit analysis. Ibid., p. 624 (tab. 7), p. 626 (tab. 9). These measures are likely to be highly correlated—one did not pass tolerance—which would cause unstable and unreliable coefficient estimates. These problems raise serious doubts about substantive conclusions based on this analysis.
15.
15. Comparisons of empirical studies carried out by different investigators in different jurisdictions are inherently difficult. These two studies had many important differences: they were carried out in different years—1984 in Manhattan and 1987 in Dade County—during a period when illegal drug use was experienceing rapid change; they measured variables differently, especially the drug test results; and they used dissimilar analytic techniques. The overall rearrest rate in Manhattan was 25 percent compared to 15 percent in Dade County; the FTA rates were 33 percent and 9 percent in Manhattan and Dade County, respectively—these differences are probably related, in part, to the length of the follow-up period in the two studies. Opiate use, a characteristic of serious drug use and criminal behavior, was practically nonexistent in Dade County, whereas 21 percent of the Manhattan sample tested positive for opiates. Finally, 56 percent tested positive for drugs in general in Manhattan compared to 80 percent in Dade County.
16.
16. Smith, Wish, and Jarjoura, “Drug Use and Pretrial Misconduct in New York City,” p. 119.
17.
17. The statistical determination of whether specific information predicts individual criminal behavior is a two-step analytic task. Initially, a prediction instrument using the information is developed, and then a second analysis, with a different sample or with the same sample at a later point in time, is used to confirm the instrument's accuracy. The analyses reported here are only the first step in this process. For further discussion, see, for example, Don Gottfredson and Michael Tonry, eds., Prediction and Classification: Criminal Justice Decision Making (Chicago: University of Chicago Press, 1987), esp. pp. 21-52, 201-248.
18.
18. The federal court system also implemented a demonstration pretrial testing program in several sites without the monitoring component, but an evaluation of its effectiveness was not conducted. See U.S., administrative Office of the United States Courts, The Demonstration Program of Mandatory Drug Testing of Criminal Defendants (submitted to the U.S. Congress, 1991).
19.
Additional reports on the pretrial demonstration programs are available in seven issues of the Pretrial Reporter, the newsletter of the National Association of Pretrial Services Agencies (Oct., Dec. 1988; Feb., Apr., June, Aug. 1989; Feb. 1991).
20.
20. Goldkamp, Jones, and Gottfredson, “Measuring the Impact of Drug Testing at the Pretrial Release Stage: Pretrial Drug Testing,” pp. 81-96.
21.
21. In the Pima County experiment, 4 percent of the drug-monitored group was rearrested compared to 12 percent of the control group (p = .06). In the Maricopa County experiment, the drug-monitored group had a 29 percent FTA rate whereas the control group had a 39 percent FTA rate (p = .11). Sample sizes in both analyses were small, which reduced the likelihood of finding large statistically significant differences between the groups.
22.
22. In particular, the analyses of the release decisions in Prince George's County and Milwaukee (Goldkamp, Jones, and Gottfredson, “Measuring the Impact of Drug Testing at the Pretrial Release Stage: Experimental Findings,” p. 27, tabs. 2.7 and 2.8) suffer from the problems raised earlier about the Dade County study (Goldkamp, Gottfredson, and Weiland, “Pretrial Drug Testing and Defendant Risk,” n. 12). The simultaneous inclusion of multiple measures of drug test results—namely, positive for any drug, positive for cocaine, and number or positive test results—in the multivariate analysis confounds any substantive interpretation of the results.
23.
23. National Association of Pretrial Services Agencies, “Pretrial Drug Testing: A Review of Three Evaluations,”Pretrial Reporter, Feb. 1991, p. iv.
24.
24. Specifically, in Pima County, 29 percent of the defendants in the monitoring program had no positive tests during monitoring, in Prince George's County 36 percent had no positive tests, and in Milwaukee 32 percent had no positive tests. Similar data from Maricopa County and Washington, D.C., are not presented in available sources.
25.
25. Toborget al., Assessment of Pretrial Urine Testing, pp. 12-16.
26.
26. Data from Prince George's County and Milwaukee on positive test results and pretrial misconduct are difficult to interpret because of analytic problems. The analysis unfortunately combined defendants who had no positive tests with those who missed all tests, the latter being a group likely to comprise drug users. Thus the reported high FTA rates for those with no positive tests in Prince George's County and Milwaukee are not surprising since many of the defendants in that group had missed all tests. See Goldkamp, Jones, and Gottfredson, “Measuring the Impact of Drug Testing at the Pretrial Release Stage: Experimental Findings,” p. 92 (tab. 5.15), p. 114 (tab. 6.9).
27.
27. Mary A. Toborg and John P. Bellassai, “Assessment of Pretrial Urine Testing in the District of Columbia: The Views of Judicial Officers” (Final report, National Institute of Justice, U.S. Department of Justice, 1988), pp. 4, 10.
28.
28. Ibid., pp. 5-9.
29.
David Hoytet al. , “Drug Testing in the Workplace—Are Methods Legally Defensible?”Journal of the American Medical Association, 258: 504-509 (July 1987).
30.
30. See the review of drug-testing case law in U.S., Department of Justice, Bureau of Justice Assistance, American Probation and Parole Association's Drug Testing Guidelines and Practices for Adult Probation and Parole Agencies (Washington, DC: Bureau of Justice Assistance, 1991), pp. 87-108.
31.
Christy Visher , A Comparison of Urinalysis Technologies for Drug Testing in Criminal Justice (Washington, DC: Department of Justice, National Institute of Justice, 1991).
32.
32. National Association of Pretrial Services Agencies, “Pretrial Drug Testing in Six Jurisdictions: Legal Issues in Pretrial Urinalysis,”Pretrial Reporter, June 1989, pp. i-iv.
33.
Reggie B. Walton , Gary J. Peters, and J. Anthony Towns, “Pretrial Drug Testing—An Essential Component of the National Drug Control Strategy,”Brigham Young University Law Journal, in press.
34.
34. Berry v. District of Columbia, 833 F.2d 1031, 1034-36 (D.C. Cir. 1987).
35.
35. U.S., Department of Justice, Bureau of Justice Assistance, Estimating the Costs of Drug Testing for a Pretrial Services Program (Washington, DC: Bureau of Justice Assistance, 1989), p. 17.
36.
36. The Drug Use Forecasting program obtains voluntary and anonymous urine specimens and interviews from a sample of arrestees in 24 cities each quarter. For more information about the program, see Joyce Ann O'Neil and Virginia Baldau, Drug Use Forecasting: 1990 Annual Report (Washington, DC: Department of Justice, National Institute of Justice, 1991).
37.
37. Discussed in Walton, Peters, and Towns, “Pretrial Drug Testing.”
