Correctional Boot Camps: A Tough Intermediate Sanction - Chapter 18. Multisite Study of Correctional Boot Camps MENU TITLE: Multisite Study of Boot Camps Series: NIJ Report Published: February 1996 11 pages 24,537 bytes Multisite Study of Correctional Boot Camps by Doris Layton MacKenzie, Ph.D., and Claire Souryal Doris Layton MacKenzie is an Associate Professor in the Department of Criminology and Criminal Justice at the University of Maryland. As a Visiting Scientist at the National Institute of Justice, she provided expertise to Federal, State, and local jurisdictions on shock incarceration, intermediate sanctions, and other aspects of correctional policy and practice. Claire Souryal is a graduate student in the Department of Criminology and Criminal Justice, University of Maryland. She worked on a multisite evaluation of eight correctional boot camp programs. The proliferation of boot camp programs nationwide appears to have been largely unguided by comprehensive and methodologically rigorous evaluations. This chapter provides a brief overview of boot camp evaluation literature, beginning with a review of early boot camp research and concluding with selected findings of a recently completed NIJ-sponsored evaluation of eight State-level programs. The findings concern attitudinal change during incarceration, adjustment during community supervision, and offender recidivism. The study found that the in-prison phase of boot camps had a positive impact on the attitudes of boot camp participants, in contrast to a sample of regular prison inmates, whose attitudes deteriorated during their incarceration. However, boot camp graduates did not adjust more positively to community supervision following boot camp than did comparison samples of boot camp failures, prison releasees, and probationers. Recidivism findings suggest that common components of boot camps, such as military-style discipline, physical training, and hard labor, did not reduce recidivism. The authors conclude that future evaluations should focus on the efficacy of specific boot camp program components, such as academic training, and on the relative effectiveness of different types of community supervision provided to boot camp releasees. Since the inception of the first military-style correctional boot camp (or shock incarceration) programs in 1983 in Georgia and Oklahoma, boot camp programs have become a common intermediate sanction in States and counties across the country. In one decade, 36 State correctional systems and the Federal Bureau of Prisons have implemented more than 47 boot camp programs for adult offenders. Boot camp programs have also been developed for juvenile offenders in six States and for use in local jails. The passage of the Violent Crime Control and Law Enforcement Act of 1994, which allocated $25 million for the development of boot camp programs, virtually ensures their continued growth and development. What accounts for the tremendous popularity of boot camp programs? Why have correctional agencies so readily embraced them as viable intermediate sanctions? Certainly, it is not due to the favorable results of empirical research. Until recently, comprehensive and methodologically rigorous evaluations of boot camp programs have been noticeably lacking. Rather, the rapid growth of boot camp programs across the Nation is attributable to other factors, including the politics of crime control (i.e., politicians clamoring to appear "tough" on crime), the compelling images of boot camps produced in the mass media and especially on television, the appeal of military-style discipline as a correctional tool with young offenders, the punitive nature of the program in comparison to other intermediate sanctions, and the promise of meeting a wide range of goals for both the system and individuals (MacKenzie and Parent, 1992). Evaluation research appears to have played a very small role, if any, in the proliferation of boot camp programs nationwide. Indeed, until recently, comprehensive evaluations of boot camp programs were rarely conducted. This chapter begins with a review of early boot camp research and concludes with an overview of principal findings about boot camps' effects on individuals--one aspect of a recently completed NIJ-sponsored evaluation of eight State-level programs. Early Boot Camp Evaluations This Nation's first correctional boot camp programs were developed in Georgia and Oklahoma in 1983. Other States swiftly followed suit and by the end of the decade, 14 had developed their own programs. The development of the earlier programs seems to have been largely unguided by evaluation research. In November 1987, for example, then-Senator Lloyd Bentsen of Texas requested a study of State boot camp programs from the U.S. General Accounting Office (GAO). The resulting report, released in September 1988, concluded that "available data are not sufficient to determine if boot camps reduce prison overcrowding, costs, or recidivism" (GAO, 1988). Until recently, boot camp evaluations were limited in scope and methodology and typically conducted by the State correctional agencies that run them. The results of these in-house studies were not widely disseminated, nor were they very thorough. Examinations of recidivism, for example, often simply assessed the return-to-prison rates of boot camp graduates without selecting adequate comparison samples or controlling for the amount of time graduates spent on the street after release. Moreover, some of the earlier evaluations compared the return-to-prison rates of boot camp graduates with those of the total population of prison releasees in a State. Data used for this comparison unfairly favored boot camp graduates because boot camp programs specifically targeted nonviolent offenders who did not have extensive criminal histories. Clearly, the selection of adequate comparison samples is essential to conducting thorough and unbiased evaluations. Another major problem with many of the early evaluations was that researchers failed to control for the time offenders spent in the community following their release from boot camp. At a meeting of the American Correctional Association, a boot camp program official bragged about low recidivism rates for a program that had been operating for little more than 6 months. The program had released only one group of graduates, none of whom had been supervised in the community for more than 3 months. The recidivism rates for such a program would appear to be low, of course, if compared to those of prison releasees who had been supervised in the community for a year or more. In general, the longer of ex-offenders are in the community, the greater the percentage who fail. The great diversity among boot camp programs is also problematic from a research perspective. While some programs devote 3 or more hours per day to therapeutic activities such as drug treatment, academic education, or life skills training, others limit the schedule of daily activities to marching, military ceremony, and hard labor. Therefore, the results of any single evaluation have been difficult to generalize from one jurisdiction to another. The NIJ Multisite Evaluation In 1990, the National Institute of Justice (NIJ) funded an evaluation of boot camp program efficacy in eight States. This multisite study's overarching research question was whether boot camp programs successfully met their stated goals. As previously noted, prior NIJ-funded research examining the effectiveness of boot camp programs had focused on a single jurisdiction, a major limitation given the large differences among programs nationwide (see MacKenzie et al., 1993). The 1990 evaluation was designed to address this deficiency by selecting multiple sites for participation, which would significantly enhance the ability to generalize results. The eight States that participated in the study were Florida, Georgia, Illinois, Louisiana, New York, Oklahoma, South Carolina, and Texas. Sites in those States were selected because they incorporated the core elements of boot camp programs and because they varied on several dimensions hypothesized to influence the achievement of program goals. Core elements of boot camp programs were defined as follows: o Strict rules, discipline, and a military boot-camp-like atmosphere. o Mandatory participation in military drills and physical training. o Separation of program participants from other prison inmates. Differences among programs were examined in these areas: o Decisionmaking authority. o Program location. o Program components. o Supervision intensity upon release. The evaluation was also designed to provide a comprehensive assessment of each program (MacKenzie and Souryal, 1994). Rather than focusing on one measure of program success, such as recidivism reduction, the evaluation consisted of five major components: a process evaluation or qualitative description of the eight programs based on interviews with staff and inmates, official program materials, and observation; a study of inmate attitudinal change during incarceration; a study of offender recidivism; a study of positive adjustment during community supervision; and a study of prison bedspace savings. At the individual level, the first stage of the evaluation examined offenders' experiences during the in-prison phase of the programs. Its objective was to ascertain whether the boot camp experience engendered positive or negative changes in attitudes over the course of the programs. Next, after offenders had either completed the program and been released to the community or failed the program, their community performance was assessed in terms of both positive adjustment to community supervision and recidivism. Positive adjustment to community supervision included a variety of measures such as employment/educational status; recidivism was measured by arrests, revocations for new crimes, and revocations for technical violations. This chapter presents a brief overview of these findings. The multisite evaluation's results and methodology have been presented in detail elsewhere. (See Brame and MacKenzie, this volume; MacKenzie, 1994a; MacKenzie, 1994b; MacKenzie and Souryal, 1994; MacKenzie and Piquero, 1994; MacKenzie and Brame, 1995.) Also note that system-level analysis results have not been presented (see chapter 16 in this volume and MacKenzie and Piquero, 1994, for a discussion of the impact of boot camp programs on prison bedspace savings). The remainder of this chapter will focus on examination of the programs at the individual level. Overview of Evaluation Results Three sections of the evaluation assessed the impact of boot camp programs at the individual level. In each study the performance of boot camp participants (either during incarceration or during community supervision) was compared to the performance of demographically similar offenders who were legally eligible for boot camp programs in their States but had served time in a conventional prison. Samples of probationers were also examined in the recidivism and community adjustment studies. It should be emphasized here that comparison samples were selected to be as similar as possible to boot camp participants in each State. Attitudinal change during incarceration. Boot camp participants and comparison samples or prison inmates were asked to complete a self-report questionnaire two times during their terms of incarceration. The questionnaire was first administered at the beginning of offenders' incarceration periods and again approximately 90 days later (180 days later in New York). The questionnaire focused on two types of attitudes-- antisocial attitudes and attitudes toward the program (or prison). An antisocial attitudes scale was used to gauge opinions of police and authority, levels of maturity, and degree of social deviance (MacKenzie and Souryal, 1993b). The scale has been found to be related to recidivism and short-term change. The scale for program attitudes measured the degree to which offenders expected their period of incarceration to motivate them to change in a positive manner, such as growing more mature or learning self-discipline. Findings for this section were remarkably consistent across sites despite differences in program content, such as in rehabilitative programming. Inmates in both samples developed fewer antisocial attitudes during incarceration. On program attitudes, however, results differed over time. Boot camp participants' attitudes became more positive as their programs progressed, while prison inmates' attitudes became less positive or stayed the same. Thus, contrary to the expectations of some critics, the boot camp experience did not result in the development of more negative program or antisocial attitudes. In short, regardless of variations in structure and programming, the boot camp programs appeared to have a positive impact on inmate attitudes. Compared to samples of offenders incarcerated in conventional prisons, boot camp inmates believed that their boot camp experience had changed them in a positive way. Prison inmates, on the other hand, did not view prison as a positive experience. The fact that the program attitude changes were so consistent across sites led us to conclude that these attitude changes were likely a result of the boot camp atmosphere and not a product of additional treatment or therapy available in some camps. Thus, one potential benefit of the programs is that participants think of boot camps as a positive experience, a change in attitude which could help produce behavioral changes after graduates are released to the community. Positive adjustment to community supervision. Performance during community supervision was assessed in five States using a positive community adjustment scale. Positive adjustment was measured in terms of success in pursuing employment, education, residential and financial stability, and treatment opportunities. During 1 year of supervision, probation or parole officers were asked to complete a 10-item index at 3-month intervals. Community adjustment was considered to be a stronger measure of success after release than simple arrest or revocation rates because boot camp programs are believed to increase the capability of offenders to adjust successfully to the day-to-day requirements of community living. (See chapter 17 of this volume for a fuller description of this portion of the evaluation.) The results of this section of the study were also quite consistent across sites. With the exception of one jurisdiction, which did not stand out from the other programs in rehabilitative programming, boot camp graduates did not adjust more positively to community supervision than comparison samples of boot camp failures, prison releasees, and probationers. Positive community adjustment was related to demographic and offense-related characteristics, criminal history, and supervision intensity, but not to whether offenders had completed their programs. Similar to the study's examination of attitude change, then, results were fairly consistent across sites despite programmatic differences. This section's results, however, provided rather convincing evidence that the in-prison phase of boot camp programs had little, if any, effect on positive community adjustment. Offender recidivism. Recidivism levels for boot camp graduates and comparison samples were examined using the following outcome measures: arrests, probation revocations for new crimes, and revocations for technical violations. The followup period was either 1 or 2 years of community supervision. It should be noted that because revocations for new crimes could not be distinguished from revocations for technical violations in two States, a general revocation measure was constructed. The results of the study were not quite as consistent across sites due to variations in the measures of recidivism available, the length of the followup period, and differences in supervision intensity. In some States, for example, boot camp graduates underwent intensive supervision while comparison samples did not. Nevertheless, taken as a whole, the results indicated that the in-prison phase of the boot camp programs had a negligible impact on recidivism. By and large, boot camp graduates appeared to perform at the same level as probationers and prison releasees who had served longer terms in prison. In five States--Florida, Georgia, Oklahoma, South Carolina, and Texas--there were no differences in recidivism that could be attributed to the in-prison phase of the program. But in three States--Illinois, Louisiana, and New York--evidence emerged indicating that boot camp graduates may have had lower recidivism rates on particular recidivism measures. Graduates in Illinois and Louisiana, for example, were less likely to have their supervision status revoked as a result of a new crime revocation than were comparison samples. Similarly, graduates of New York's program were less likely to be returned to prison as a result of a technical violation, although they did not differ from the comparison sample on any other measure of recidivism. Graduates in Illinois, however, were more likely to have their supervision status revoked as a result of a technical violation. What do the Illinois, Louisiana, and New York programs have in common that might explain the similarity in results? These programs stand out as the only boot camps in the study that have developed an intensive supervision phase for boot camp graduates. It is therefore very likely that differences in recidivism rates in these States were due to the type of community supervision provided to graduates, not in-prison programming. In other words, the effect of the in-prison phase on participants was likely confounded with the level of supervision intensity after their release. These programs also devoted a considerable amount of time per day (3 hours or more) to rehabilitative activities, had high dropout rates, required voluntary participation, and selected participants from a pool of prison-bound offenders. (Note, however, that the characteristics listed above are not exclusive to these programs.) Although this study was not able to untangle the particular effects of each program component on recidivism, results clearly show that the core elements of boot camp programs--military-style discipline, hard labor, and physical training--by themselves did not reduce offender recidivism. At this time it is unclear which component or components are critical to reducing recidivism, but it is likely that some mixture of rehabilitation and intensive followup supervision plays an important role. Summary of Results Paradoxically, the results for each individual-level component of the evaluation were remarkably similar across sites despite differences among the State programs. While the results of the attitudinal change survey revealed that the in-prison phase had a clear and consistent impact on the attitudes of boot camp participants across sites, this change did not appear to affect offenders' performance during community supervision. In fact, there is very little evidence that the in-prison phase had much of an impact on community supervision performance. The results of the positive adjustment to the community supervision phase of the study would seem to be a case in point. With the exception of Florida, boot camp graduates did not adjust more positively to the community supervision than did comparison samples of boot camp failures, prison releasees, and probationers. Although results from the recidivism study were not as consistent as results from the other two studies, they strongly suggest that in-prison components common to each program, such as military-style discipline, physical training, and hard labor, did not reduce recidivism. Conclusion Unquestionably, more research is needed to indicate what can be accomplished with boot camp programs. Many times people ask, "Is this a correctional fad?" Yes, today boot camps are a fad, and they will fade away as fads inevitably do unless we explore indepth which aspects of this sanction are worth keeping and which are not. As jurisdictions around the country consider implementing their own programs, the many issues raised in this study need to be explored. For example, as there does appear to be some positive attitudinal change associated with boot camp participation, it is important to explore whether this atmosphere facilitates other positive changes associated with participants' success in the community after release. Do boot camp inmates learn more in academic classes than they would in a traditional prison during the same time period? Research to answer these important questions could easily be accomplished using a pretest/posttest design. Academic achievement could be measured first upon entry to the boot camp and again 90 days later. Performance in boot camps could be compared with the performance of similar prison inmates over the same time period. Studies could also be conducted examining such factors as moral development and life skills. Moreover, more information is needed about the adjustment and activities of boot camp graduates in the community. There is evidence that intensive supervision helps boot camp graduates when they return to the community; it is important to know exactly what elements of this supervision are helpful. Does such supervision have to be preceded by incarceration in a boot camp or would other prisoners do just as well if they were provided with the same type of supervision and services? These are the kinds of questions we need to answer to improve our correctional system over the long term and move beyond following the latest correctional fad. What is clear from the multisite study is that boot camp programs are not one-dimensional--a boot camp is not a boot camp is not a boot camp. We must identify the elements of boot camps that work and reject those that do not. This type of exhaustive evaluation will not be possible, however, until enough time, money, and effort are devoted to independent and objective study of the programs. References MacKenzie, D.L. "Boot Camps: A National Assessment." Overcrowded Times: Solving the Prison Problem, 5, no. 4 (1994a). MacKenzie, D.L. "Results of a Multisite Study of Boot Camp Prisons." Federal Probation, 58, no. 2 (1994b):60-66. MacKenzie, D.L., and R. Brame. "Shock Incarceration and Positive Adjustment During Community Supervision." Journal of Quantitative Criminology, 11, no. 2 (1995):111-142. MacKenzie, D.L., R. Brame, D. McDowall, and C. Souryal. "Boot Camp Prisons and Recidivism in Eight States." Criminology, 33, no. 3 (1995):401-430. MacKenzie, D.L., and D. Parent. "Boot Camp Prisons for Young Offenders." In Smart Sentencing: The Emergence of Intermediate Sanctions, J.M. Byrne, A.J. Lurigio, and J. Petersilia (eds.). Newbury Park, California: Sage Publications, 1992:103-119. MacKenzie, D.L., and A. Piquero. "The Impact of Shock Incarceration Programs on Prison Crowding." Crime and Delinquency, 40, no. 2 (1994):222-249. MacKenzie, D.L., J.W. Shaw, and V.B. Gowdy. An Evaluation of Shock Incarceration in Louisiana. Research in Brief. Washington, D.C.: U.S. Department of Justice, National Institute of Justice, 1993. MacKenzie, D.L., and C. Souryal. Multi-Site Study of Shock Incarceration: Process Evaluation. Final Reports I-IV and Executive Summary to the U.S. Department of Justice, National Institute of Justice. Washington, D.C., 1994. MacKenzie, D.L., and C. Souryal. "Inmate Attitude Change During Incarceration: A Comparison of Boot Camp With Traditional Prison." Justice Quarterly, 12(2) 1995. U.S. General Accounting Office. "Prison Boot Camps: Too Early To Measure Effectiveness." Briefing Report to the Honorable Lloyd Bentsen, U.S. Senate. Washington, D.C.: U.S. General Accounting Office, 1988.