http://informahealthcare.com/ada ISSN: 0095-2990 (print), 1097-9891 (electronic) Am J Drug Alcohol Abuse, 2014; 40(3): 200–205 ! 2014 Informa Healthcare USA, Inc. DOI: 10.3109/00952990.2013.858722
ORIGINAL ARTICLE
Twelve-month employment intervention outcomes for drug-involved offenders J. Matthew Webster, PhD1,2, Michele Staton-Tindall, PhD2,3, Megan F. Dickson, MA2,4, John F. Wilson, PhD1, and Carl G. Leukefeld, DSW1,2 1
Department of Behavioral Science, 2Center on Drug and Alcohol Research, 3College of Social Work, and 4Department of Sociology, University of Kentucky, Lexington, KY, USA Abstract
Keywords
Background: Employment has been identified as an important part of substance abuse treatment and is a predictor of treatment retention, treatment completion, and decreased relapse. Although employment interventions have been designed for substance abusers, few interventions have focused specifically on drug-involved offenders. Objectives: The purpose of this study was to examine employment outcomes for drug-involved offenders who received a tailored employment intervention. Methods: In a randomized controlled trial, baseline and follow-up data were collected from 500 drug-involved offenders who were enrolled in a drug court program. Participants were randomly assigned to drug court as usual (control group) or to the employment intervention in addition to drug court. Results: Intent-to-treat analyses found that the tailored intervention was associated only with more days of paid employment at follow-up (210.1 vs. 199.9 days). When focusing on those with greater employment assistance needs, a work trajectory analyses, which took into account participants’ pre-baseline employment pattern (negative or positive), revealed that intervention group participants had higher rates of employment (82.1% vs. 64.1%), more days paid for employment (188.9 vs. 157.0 days), and more employment income ($8623 vs. $6888) at follow-up than control group participants. Conclusion: The present study adds to the growing substance abuse and employment literature. It demonstrates the efficacy of an innovative employment intervention tailored for drug-involved offenders by showing positive changes in 12-month employment outcomes, most strongly for those who have not had recent employment success.
Drug-involved offenders, employment, intervention
Employment has been identified in the research literature as an important part of substance abuse treatment. It is a predictor of treatment success and has been shown to be a significant factor in substance treatment entry (1), treatment retention (2), and treatment completion (3). Employment has also been identified as a desirable outcome of substance abuse treatment (4) and has been associated with positive treatment outcomes. In particular, employment decreases the likelihood and severity of relapse (5) and lowers depression and other mental health issues (6). Employment is also associated with fewer rearrests (7) and more time spent crime-free among offenders (8). Reports also suggest that stable employment eases community reintegration and has been correlated to several factors
Address correspondence to J. Matthew Webster, PhD, Department of Behavioral Science, 120 Medical Behavioral Science Building, Lexington, KY 40536-0086, USA. Tel: +1 859 323 6100. Fax: +1 859 323-5350. E-mail: [email protected]
Received 4 March 2013 Revised 15 October 2013 Accepted 18 October 2013 Published online 9 January 2014
that improve treatment outcomes and successful community reentry. Specifically, former treatment clients with regular employment report healthier social and professional networks as well as decreased involvement in their former substanceusing networks (9). Stable employment also provides a steady income and enables legal economic independence. Time at work increases time involved in a constructive, valuable activity and reduces the free time that may have previously been devoted to unhealthy behaviors such as drug use. Although employment positively correlates with other indicators of treatment success, substance abusers frequently face barriers in finding and keeping a job. These individuals often have poor work habits, including regular absences, tardiness, sleeping on the job, using vulgar language in the workplace (10,11), and have difficulties staying organized, handling stress, meeting deadlines, and maintaining stamina during the day (12). Drug-involved offenders have the additional burden of poor employment history due to incarceration (8) and the stigma associated with a criminal record (13,14) and status as a drug abuser (15) to overcome when pursuing employment opportunities. As a result, the
20 14
Introduction
History
Twelve-month employment outcomes
DOI: 10.3109/00952990.2013.858722
most commonly available jobs for drug-involved offenders are poorly paid, entry-level service jobs. To help substance abusers overcome these obstacles, employment training programs and interventions have been incorporated into existing substance abuse treatment programs, but these employment services vary widely across programs with mixed results (16). Some studies have shown that employment interventions lead to positive changes among treatment clients. For example, Hall et al. (17) conducted an employment workshop (Job Seeker’s Workshop) for parolees and probationers with a history of drug use and concluded that the intervention significantly increased levels of employment. Other studies have found that employment intervention programs for substance abuse treatment clients have subtle, but positive, changes at treatment completion, including increases in the number of days paid for work (18) and in income (19). McIntosh et al. (9) concluded that clients who received employmentrelated help along with traditional treatment were more likely to obtain paid employment. Not all studies, however, have found clear support for employment interventions. For example, Hubbard et al. (20) reported that only those clients in long-term treatment showed improvement in employment status as compared to clients in methadone maintenance treatment. Another study examining long-term substance abuse treatment clients and methadone maintenance found that neither group experienced any significant improvement in employment (21). More recently, a multi-site randomized trial of the Job Seekers’ Workshop found no difference in 12-week or 24-week employment outcomes, with only modest employment gains overall (22). Few interventions have focused specifically on druginvolved offenders (16). This literature gap led to the development of an employment intervention tailored to drug-involved offenders, which is grounded in established employment interventions and life skills training (23). The purpose of the current study was to examine employment outcomes for drug-involved offenders who received a tailored employment intervention. It was predicted that those who received the employment intervention would have better overall employment outcomes across multiple employment measures including higher rates of employment, days worked, and income.
Methods Participants Five hundred participants were recruited from two Kentucky drug court sites, which were administered following the established national drug court key components (24). Study eligibility was based on criteria for entry into the drug court program including non-violent charges, a self-admitted drug problem, urine test consent, and an Addiction Severity Index classification as a drug abuser. During the recruitment period, 525 drug court participants were eligible for the study across the two drug courts. Of those eligible, 500 were interviewed (95%), 7 clients refused, and 18 clients were terminated within 30 days of program entry and before a scheduled baseline interview could be conducted. After completing a baseline interview, participants were randomized to one of
201
two conditions, and then completed a 12-month follow-up interview. Four subjects died between the baseline and followup interview, and therefore, were ineligible for follow-up. The follow-up interview rate was 96% with 477 of 496 eligible participants interviewed. Because the participating drug courts valued employment, the research team was given permission to conduct the study independent of the drug court program. Drug court staff and judges were not informed about each client’s decision to participate in the study, their study condition, or level of participation. At baseline, the majority of participants were male (65.4%), more than half were white (61.8%), and a few were married (17.7%). Participants had an average age of 30.5 years (SD ¼ 8.7), a mean of 11.8 years of education (SD ¼ 2.0), and several incarcerations (M ¼ 4.6, SD ¼ 7.6). About half of the participants (53.6%) were employed either full-time or part-time. Participants most frequently reported cocaine (36.0%) as the drug causing the most problems followed by marijuana (23.4%) and alcohol (11.2%). An attrition analysis found no significant demographic differences between baseline and follow-up samples. Procedure Drug court clients were recruited for study participation between March 2000 and December 2002. Study participation was voluntary and subjects interested in participating were interviewed face-to-face by a trained interviewer within 2 weeks of drug court program entry. At the time of interview, participants were given a description of the project, procedures, and assurance of confidentiality following IRBapproved informed consent procedures. The interview lasted approximately 2 hours, and all participants were paid for their participation. Following the baseline interview, each participant was randomized to one of two groups: the employment intervention group or the drug court as usual control group. Participants randomized to the employment intervention began a series of individual and group sessions with a study-hired employment specialist with experience in both employment and substance abuse counseling. The employment specialist coordinated scheduling of intervention sessions with the drug court office. Individual sessions incorporated behavioral contracting, motivational interviewing, and strengths-based case management with a focus on employment barriers and resolving issues that impede employment success. Group sessions also used motivational interviewing as well as structured stories and thoughtmapping to address employment needs. The 26-session intervention was organized and delivered in three phases: obtaining employment, maintaining employment, and upgrading employment. Specific topics included networking, preparing effective resumes, writing a cover letter, filling out job applications, addressing and overcoming a criminal record in the application and interview, and proper workplace behavior (see 23 for a full description of the employment intervention, and 4 for employment outcomes related to the upgrading phase of the intervention). Sixty-one percent of intervention participants completed half or more of the available sessions.
202
J. M. Webster et al.
Approximately 10 months after the baseline interview, attempts were made to contact participants in order to schedule a 12-month follow-up interview. Locating and tracking activities were conducted without court assistance and included regular mailings, telephone calls, birthday and holiday card mailings, home visits, and electronic database checks. The 12-month interview contained the same measures as the baseline interview but focused on the time that had elapsed since the baseline interview. Measures Demographic information and employment data were collected using the Addiction Severity Index (25). Additional employment information was collected with measures developed at the University of Massachusetts School of Medicine (26). The key 12-month employment outcomes measures included: typical work status since the baseline interview, days paid for legal employment in the past 30 days and since the baseline interview (past year), and income from legal employment in the past 30 days and since the baseline interview (past year). Work status was treated as a categorical variable with four levels: full-time, part-time, unemployed, and not in the workforce (e.g. not working and not looking for a job). Days paid for legal employment and income from legal employment were continuous measures. Analytic plan In order to identify potential covariates for subsequent analyses, a series of t-tests and chi-square analyses were conducted to test for basic demographic differences between the intervention and control group. Next, a series of chisquare and ANCOVA intent-to-treat analyses were conducted to examine the effect of the intervention on employment outcomes including work status, days paid for legal employment (past year and past 30 days), and income from legal employment (past year and past 30 days). An alpha level of p50.05 was used for all analyses. An examination of the variables measuring days paid for legal employment and income from legal employment found that each was non-normally distributed (skewness range ¼ 0.23 to 4.43; kurtosis range ¼ 1.74 to 33.84). The income variables were moderately skewed and both the income and the days paid for legal employment variables were kurtotic. To remedy these distributional problems, a square root transformation of these dependent measures resulted in the greatest minimization of the skewness (range ¼ 0.92 to 0.80) and kurtosis (range ¼ 1.69 to 1.87) statistics across the range of variables affected (27). The transformed variables were used in subsequent analysis. Because almost half (44%) of the sample was employed full-time at baseline, we also examined the effectiveness of the intervention for those participants who were in the most need of employment assistance. Recognizing that many druginvolved offenders have unstable employment, we operationalized this by examining participants’ work trajectories. A pre-baseline work trajectory was created by examining participants’ typical job status in the 6 months prior to entering drug court and their typical job status since entering drug court. Both of these variables were measured at baseline.
Am J Drug Alcohol Abuse, 2014; 40(3): 200–205
Participants who experienced a decrease in their employment status (e.g. full-time to part-time or part-time to unemployed) or stayed unemployed from one time point to the next were considered to be on a negative pre-baseline work trajectory. Those participants whose employment status did not decrease (e.g. stayed employed full-time or went from unemployed at baseline to part-time at follow-up) were collapsed into a positive pre-baseline work trajectory group. A second work trajectory, follow-up work trajectory, was created as an outcome variable. This variable was created similarly to the pre-baseline work trajectory variable but examined participants’ employment status at the 12-month follow-up interview relative to their employment status at the time of the baseline interview. Thirty participants did not fit clearly into one of the follow-up work trajectory groups (e.g. missing data) and were eliminated from subsequent analyses.
Results Significant differences in age between the intervention group and control group were found at baseline (Mintervention ¼ 29.7, SD ¼ 8.4 vs. Mcontrol ¼ 31.2, SD ¼ 9.0; t(498) ¼ 1.96, p ¼ 0.05, d ¼ 0.18) and at 12-month follow-up (Mintervention ¼ 30.7, SD ¼ 8.4 vs. Mcontrol ¼ 32.6, SD ¼ 9.0; t(475) ¼ 2.28, p ¼ 0.02, d ¼ 0.21). Likewise, differences between the two groups in the number of years of education significantly differed at both time points (Mintervention ¼ 11.5, SD ¼ 1.9 vs. Mcontrol ¼ 12.0, SD ¼ 2.0 at baseline and followup; t(498) ¼ 2.47, p ¼ 0.01, d ¼ 0.22) and t(475) ¼ 2.70, p ¼ 0.01, d ¼ 0.25, respectively). No other group differences for demographic information (gender, race, marital status, drug court site) were found; thus, only age and education were used as covariates. The intent-to-treat analysis found that the employment intervention had a positive effect on the number of days paid for legal employment (see Table 1). Participants in the intervention group reported more days paid for employment in the past year (M ¼ 210.1, SD ¼ 114.1) than those in the control group (M ¼ 199.9, SD ¼ 130.1, d ¼ 0.20; F(1, 464) ¼ 4.69, p ¼ 0.03). No differences were found, however, for other employment outcomes, including the percentage of participants moving from unemployed at baseline to employed at follow-up (approximately 30% for both groups). All analyses of key employment outcomes (work status, days paid for legal employment, and income from legal employment) were then repeated to take into account participants’ pre-baseline work trajectory. An additional employment outcome, follow-up work trajectory, was also examined. Work status A chi-square analysis was performed to test whether work status patterns differed across combinations of the experimental condition (control vs. intervention) and the prebaseline work trajectory (negative vs. positive). A marginal difference in work status across the groups was found, 2 (6, N ¼ 447) ¼ 12.30, p ¼ 0.056, Cramer’s V ¼ 0.12. The pattern of results showed that those with a negative prebaseline work trajectory in the control condition had a lower percentage of full-time employment versus other groups.
Twelve-month employment outcomes
DOI: 10.3109/00952990.2013.858722
203
Table 1. Intent-to-treat and work trajectory analyses of 12-month employment outcomes. Intent-to-treat analysis
Work trajectory analysis Negative trajectory
Control (n ¼ 233) Work status Working fulltime (%) Working part-time (%) Unemployed (%) Not in the workforce (%) Days paid for legal employment In the past yeara,b In the past 30 daysb Income from legal employment ($) In the past yearb In the past 30 daysb
Intervention (n ¼ 244)
Control (n ¼ 64)
Intervention (n ¼ 68)
Positive trajectory Control (n ¼ 160)
Intervention (n ¼ 155)
59.7 15.9 1.7 22.8
67.6 12.9 1.7 17.8
51.6 12.5 1.6 34.4
65.7 16.4 1.5 16.4
63.8 17.5 1.9 16.9
69.0 12.3 1.9 17.3
199.9 16.1
210.1 16.7
157.0 11.7
188.9 15.0
217.3 17.8
219.3 17.5
11 073.6 845.5
9863.9 809.9
6888.4 531.8
8623.1 635.0
12 779.6 975.4
10 407.3 881.0
a
Significant intent-to-treat effects, p 0.05; bSignificant work trajectory effects, p 0.05.
A second chi-square analysis was conducted in which work status was dichotomized into working vs. not in the workforce and resulted in a significant effect for condition, 2 (3, N ¼ 447) ¼ 9.67, p ¼ 0.022, ’ ¼ 0.15. Participants in the control group with a negative trajectory had a significantly lower rate of employment than other groups (64.1% vs. 80.8%, 81.3%, and 82.1%). Days paid for legal employment The days paid for legal employment and the income from legal employment were analyzed in a series of 2 (Intervention Group: intervention vs. control) 2 (Pre-baseline Work Trajectory: negative vs. positive) ANCOVA tests controlling for age and education as in the intent-to-treat analysis. A significant main effect for pre-baseline work trajectory was found for the number of days paid in a legal job in the past year, F(1, 437) ¼ 10.82, p ¼ 0.001, d ¼ 0.31, such that those with a positive pre-baseline work trajectory had more days paid for legal employment than those with a negative prebaseline work trajectory (see Table 1). In addition, a main effect of intervention group emerged, F(1, 437) ¼ 6.85, p ¼ 0.009, d ¼ 0.25. Similar to the intent-to-treat analysis, those randomly assigned to the intervention group had more days of paid legal employment in the past year than those in the control group. A significant main effect of pre-baseline work trajectory also was found for the number of days paid for legal employment in the past 30 days, F(1, 400) ¼ 6.57, p ¼ 0.011, d ¼ 0.26. Participants with a positive pre-baseline work trajectory worked more days in the past 30 days than those with negative pre-baseline work trajectories. The intervention group, however, did not have a significant effect on the number of days paid for legal work in the past 30 days. Income from legal employment The amount of income from legal employment significantly varied as a function of pre-baseline work trajectory (F(1, 435) ¼ 11.77, p ¼ 0.001, d ¼ 0.33). Participants who had a negative pre-baseline work trajectory earned significantly less money for legal employment in the past year than
participants with a positive pre-baseline work trajectory (see Table 1). This main effect was qualified by a significant pre-baseline work trajectory by intervention group interaction, F (1, 435) ¼ 5.14, p ¼ 0.024, d ¼ 0.22. The interaction effect was characterized by a more pronounced difference between the intervention and the control group in the negative pre-baseline work trajectory than in the positive pre-baseline work trajectory group. This effect was not detected in the analysis of income from legal employment in the past 30 days, but the same main effect of pre-baseline work trajectory emerged, F (1, 403) ¼ 7.91, p ¼ 0.005, d ¼ 0.28. Change in work trajectory Logistic regression analyses using age and education as covariates were used to test whether the employment intervention was related to follow-up work trajectory. This analysis was performed separately for those who had a negative pre-baseline work trajectory and those who had a positive pre-baseline work trajectory. For those who had a positive pre-baseline work trajectory, the intervention was unrelated to their follow-up work trajectory status (OR ¼ 1.22, CI(95%) ¼ 0.58–2.57). However, for those who had a negative pre-baseline work trajectory, the intervention was strongly related to their follow-up work trajectory status (OR ¼ 6.83, CI(95%) ¼ 1.38–33.72). Specifically, participants who had a negative pre-baseline work trajectory in the employment intervention group had almost seven times the odds of no longer being on a negative work trajectory at 12-month follow-up than those in the control group.
Discussion The purpose of this study was to examine employment outcomes for drug-involved offenders who received a tailored employment intervention. The intent-to-treat analyses found that the intervention was successful in increasing the number of days worked in the past year, but no other employment outcomes. However, subsequent analyses focusing on work trajectories showed that those who were on a negative prebaseline work trajectory were most helped by the intervention. Specifically, participants with a negative pre-baseline
204
J. M. Webster et al.
work trajectory, who were randomly assigned to the tailored employment intervention, were more likely to be in the workforce, earn more income from legal employment, and more likely be on a positive work trajectory at 12-month follow-up than those with a negative pre-baseline work trajectory and assigned to the control group. These results support the efficacy of the intervention for those who were in greatest need of employment assistance. The present research adds to the growing substance abuse and employment literature in several ways. First, this study demonstrates that an innovative employment intervention tailored for drug-involved offenders can positively affect 12-month employment outcomes, most strongly for those without recent employment success. The substance abuse research literature is replete with references to the importance of employment in the recovery process, yet the literature on strategies to enhance employment is equivocal. For example, recent examinations of employment interventions for substance abusers have been inconsistent with some finding success (e.g. 28) and others reported no intervention effects on key employment outcomes (e.g. 22). These inconsistent findings are further complicated by differences in study design and employment measures. The present study incorporates a number of strengths that make the interpretation of the results clearer. Specifically, this study used random assignment, had high follow-up rates (96%), used multiple employment measures, and tested an intervention that was tailored to drug-involved offenders. Employment is an important dynamic process for many drug abusers, shifting from job to job and from periods of employment and unemployment. The current research used a work trajectory approach to capture this movement across employment statuses over time. Previous studies have used work or employment trajectories as a predictor or outcome variable (29,30), and some have applied concepts similar to the negative and positive work trajectories used in the current study (e.g. 31). However, only a few studies have examined work trajectories among drug abusers (32), and to our knowledge, none have used this approach specifically with drug-involved offenders. While the intent-to-treat analysis provided inconsistent support for the efficacy of the intervention, the work trajectory analyses highlighted the intervention’s positive impact on several employment outcome measures including increased employment, days paid for employment, and employment income. These findings suggest that it may be important to look at employment patterns and work trajectories in addition to current employment status when targeting individuals for employment interventions. Findings also highlight the idea that employment interventions for drug-involved offenders may be most appropriately targeted to those who have the least employment success, particularly if intervention resources are limited. Although the intervention included information relevant to employed participants (i.e. maintaining and upgrading employment), participants in the negative pre-baseline work trajectory most benefitted from the intervention. Furthermore, future studies may examine how much employment intervention participation is needed to enhance favorable outcomes. Along with participation data, measures of intervention engagement can be critical to understanding overall
Am J Drug Alcohol Abuse, 2014; 40(3): 200–205
intervention effectiveness and provides another direction for future research. In addition to the importance of employment patterns, this study shows the utility of using multiple measures of employment to examine outcomes. In the present research, outcome measures including employment status, days worked, days paid, and income were used with inconsistent results. In fact, some of these employment measures may be more or less mutable within a 12-month time frame depending on the individual and nature of the job. Other substance abuse research has shown that employment is a multidimensional construct and these dimensions differentially correlate with drug use and criminality (15). Future studies that examine employment as a key outcome, therefore, should carefully select the type and number of employment measures to ensure that more than one facet of this construct is measured. Furthermore, additional research should more closely examine how changes in these different employment outcomes affect other important variables such as drug use. Finally, the study provides an example of how a criminal justice setting – drug courts in this case – can be a platform for randomized controlled studies, particularly when judges and staff ‘‘buy in’’ to the research. Drug court judges involved in this study strongly believed that employment is important for drug court clients and allowed our research team to randomly assign participants to the intervention or control group. They were kept apprised of study progress, interim findings from aggregated data, and given opportunities to provide feedback to the research team during the course of the study. There is mounting evidence that drug courts are effective in reducing recidivism (33), and interventions focusing on areas that relate to treatment retention and outcomes, such as employment, may further reduce drug use and criminality. Drug courts may be able to enhance employment by investing more resources into integrating job training with drug court programming and by developing ties with community employers to hire clients during or upon completion of drug court. This study has limitations. First, the generalizability of findings is unclear since the data are not recent. The initial intent-to-treat analyses were performed when the study was completed. However, a continued interest in the employment of drug-involved offenders led to further data analysis using a work trajectory approach several years later. In addition, generalizability is limited since data were collected from two Kentucky drug courts. Although the administration of these drug courts remains the same since data collection, results of the intervention could vary for other samples of drug courts clients or groups of offenders at different time points, in other geographic areas, and under different economic conditions. In fact, both the historic and current economic climate of Kentucky, which is often associated with rural poverty, may have attenuated the overall effects of the intervention. Furthermore, several employment variable comparisons were made, which could increase the likelihood of a Type I error, although the pattern of significant differences in the work trajectory analyses were consistent across the days worked and the income variables. Finally, this study relies on self-reported employment data, which could be limited by recall, truthfulness, or demand characteristics associated with
Twelve-month employment outcomes
DOI: 10.3109/00952990.2013.858722
assignment to the intervention or control condition. For example, intervention participants could have over-reported their employment; however, it seems that this type of bias would have appeared in the intent-to-treat analyses if it were present. Despite these limitations, the current study provides evidence that a tailored employment intervention for druginvolved offenders can produce positive outcomes. Future research should examine employment interventions for druginvolved offenders that include both the development and enhancement of employment skills along with job placement, which might include cottage industries and other small businesses.
Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. This study was supported by Grant R01 DA013076 from the National Institute on Drug Abuse; Carl G. Leukefeld, Principal Investigator; and by the staff and resources of the Center on Drug and Alcohol Research at the University of Kentucky. Opinions expressed are those of the authors and do not represent the position of the National Institute on Drug Abuse.
13. 14. 15. 16. 17. 18.
19.
20.
21.
References 1. Corsi KF, Kwiatkowski CF, Booth RE. Treatment entry and predictors among opiate-using injection drug users. Am J Drug Alcohol Abuse 2007;33:121–127. 2. Platt JJ. Vocational rehabilitation of drug abusers. Psycholog Bull 1995;117:416–433. 3. Ginexi EM, Foss MA, Scott CK. Transitions from treatment to work: employment patterns following publicly funded substance abuse treatment. J Drug Issues 2003;33:497–518. 4. Leukefeld CG, Webster JM, Staton-Tindall M, Duvall J. Employment and work outcomes among drug court clients: 12-month outcomes. Subst Use Misuse 2007;42:1109–1126. 5. Evans JL, Hahn JA, Lum PJ, Stein ES, Page K. Predictors of injection drug use cessation and relapse in a prospective cohort of young injection users in San Francisco, CA (UFO Study). Drug Alcohol Depend 2009;101:152–157. 6. Wolkstein E, Bausch R, Weber G. Work as a critical component of recovery. Paper presented at New York University, 2000. Retrieved September 30, 2011 from http.//www.med.wright.edu/ 7. Evans E, Huang D, Hser Y. High-risk offenders participating in court-supervised substance abuse treatment: characteristics, treatment received, and factors associated with recidivism. J Behavioral Health Serv Res 2011;38:510–525. 8. Tripodi SJ, Kim JS, Bender K. In employment associated with reduced recidivism? The complex relationship between employment and crime. Int J Offender Therapy Compar Criminol 2010;54: 706–720. 9. McIntosh J, Bloor M, Robertson M. Drug treatment and the achievement of paid employment. Addiction Res Theory 2008;16: 37–45. 10. Carpenedo CM, Needham M, Knealing TW, Kolodner K, Fingerhood M, Wong CJ, Silverman K. Professional demeanor of chronically unemployed cocaine-dependent methadone patients in a therapeutic workplace. Subst Use Misuse 2007;42:1141–1159. 11. Wong CJ, Silverman K. Establishing and maintaining job skills and professional behaviors in chronically unemployed drug abusers. Subst Use Misuse 2007;42:1127–1140. 12. Batiste LC. Accommodation and compliance series: employees with drug addiction. Accommodation and Compliance Series,
22.
23. 24. 25. 26. 27. 28. 29. 30.
31.
32. 33.
205
Job Accommodation Network, 2005. Retrieved from http:// www.jan.wvu.edu/media/drugadd.html Harris PM, Keller KS. Ex-offenders need not apply: the criminal background check in hiring decisions. J Contemp Criminal Justice 2005;21:6–30. Holzer HJ, Raphael S, Stoll MA. Employment barriers facing exoffenders. Center for the Study of Urban Povery Working Paper Series, 2003. Webster JM, Staton-Tindall M, Duvall JL, Garrity TF, Leukefeld CG. Measuring employment among substance-using offenders. Subst Use Misuse 2007;42:1187–1205. Magura S, Staines GL, Blankertz L, Madison EM. The effectiveness of vocational services for substance users in treatment. Subst Use Misuse 2004;39:2165–2213. Hall SM, Loeb P, LeVois M, Cooper J. Increasing employment in ex-heroin addicts II: criminal justice sample. Behavior Ther 1981; 12:443–452. Lidz V, Sorrentino DM, Robison L, Bunce S. Learning from disappointing outcomes: an evaluation of prevocational interventions for methadone maintenance patients. Subst Use Misuse 2004; 39:2287–2308. Back SE, Sonne SC, Killeen T, Dansky BS, Brady KT. Comparative profiles of women with PTSD and comorbid cocaine or alcohol dependence. Am J Drug Alcohol Abuse 2003;29: 169–189. Hubbard RL, Craddock SG, Flynn PM, Anderson J, Etheridge RM. Overview of 1-year follow-up outcomes in the drug abuse treatment outcome study (DATOS). Psychol Addict Behaviors 1997;11: 261–278. Schildhaus S, Gerstein D, Brittingham A, Cerbone F, Dugoni B. Services research outcomes study: overview of drug treatment population and outcomes. Subst Use Misuse 2000;35:1849–1878. Svikis DS, Keyser-Marcus L, Stitzer M, Rieckmann T, Safford L, Loeb P, Allen T, et al. Randomized multi-site trial of the Job Seekers’ Workshop in patients with substance use disorders. Drug Alcohol Depend 2012;120:55–64. Leukefeld C, McDonald HS, Staton M, Mateyoke-Scrivner A, Webster JM, Logan TK, Garrity T. An employment intervention for drug-abusing offenders. Federal Probation 2003;67:27–31. National Association of Drug Court Professionals. Defining drug courts: the key components. NCJ 205621, 1997. McLellan AT, Luborsky L, O’Brien CP, Woody GE. An improved evaluation instrument for substance abuse patients: the Addiction Severity Index. J Nerv Mental Dis 1980;168:26–33. Barrow S, Hellman F, Lovell AM, Plapinger JD, Robinson CR, Struening EL. Personal history form. New York: New York State Psychiatric Institute; 1985. Tabachnick BG, Fidell LS. Using multivariate statistics. Boston: Allyn and Bacon; 2001. Walls RT, Batiste LC, Moore LC, Loy B. Vocational rehabilitation and job accommodations for individuals with substance abuse disorders. J Rehab 2009;75:35–44. Becker D, Whitley R, Bailey EL, Drake RE. Long-term employment trajectories among participants with severe mental illness in supported employment. Psychiatric Serv 2007;58:922–928. Meyer JD, Mutambudzi M. Construction of life-course occupational trajectories: evidence for work as a mediator or racial disparities in hypertension. J Occup Envir Med 2012;54: 1201–1207. Virtanen P, Vahtera J, Broms U, Sillanma¨ki L, Kivima¨ki M, Koskenvuo M. Employment trajectory as determinant of change in health-related lifestyle: the prospective HeSSup study. Eur J Public Health 2008;18:504–508. Huang DY, Evans E, Hara M, Weiss RE, Hser YI. Employment trajectories: exploring gender differences and impacts of drug use. J Vocat Behav 2011;79:277–289. Mitchell O, Wilson DB, Eggers A, MacKenzie DL. Assessing the effectiveness of drug courts on recidivism: a meta-analytic review of traditional and non-traditional drug courts. J Crim Justice 2012; 40:60–71.
Copyright of American Journal of Drug & Alcohol Abuse is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
ORIGINAL ARTICLE
Twelve-month employment intervention outcomes for drug-involved offenders J. Matthew Webster, PhD1,2, Michele Staton-Tindall, PhD2,3, Megan F. Dickson, MA2,4, John F. Wilson, PhD1, and Carl G. Leukefeld, DSW1,2 1
Department of Behavioral Science, 2Center on Drug and Alcohol Research, 3College of Social Work, and 4Department of Sociology, University of Kentucky, Lexington, KY, USA Abstract
Keywords
Background: Employment has been identified as an important part of substance abuse treatment and is a predictor of treatment retention, treatment completion, and decreased relapse. Although employment interventions have been designed for substance abusers, few interventions have focused specifically on drug-involved offenders. Objectives: The purpose of this study was to examine employment outcomes for drug-involved offenders who received a tailored employment intervention. Methods: In a randomized controlled trial, baseline and follow-up data were collected from 500 drug-involved offenders who were enrolled in a drug court program. Participants were randomly assigned to drug court as usual (control group) or to the employment intervention in addition to drug court. Results: Intent-to-treat analyses found that the tailored intervention was associated only with more days of paid employment at follow-up (210.1 vs. 199.9 days). When focusing on those with greater employment assistance needs, a work trajectory analyses, which took into account participants’ pre-baseline employment pattern (negative or positive), revealed that intervention group participants had higher rates of employment (82.1% vs. 64.1%), more days paid for employment (188.9 vs. 157.0 days), and more employment income ($8623 vs. $6888) at follow-up than control group participants. Conclusion: The present study adds to the growing substance abuse and employment literature. It demonstrates the efficacy of an innovative employment intervention tailored for drug-involved offenders by showing positive changes in 12-month employment outcomes, most strongly for those who have not had recent employment success.
Drug-involved offenders, employment, intervention
Employment has been identified in the research literature as an important part of substance abuse treatment. It is a predictor of treatment success and has been shown to be a significant factor in substance treatment entry (1), treatment retention (2), and treatment completion (3). Employment has also been identified as a desirable outcome of substance abuse treatment (4) and has been associated with positive treatment outcomes. In particular, employment decreases the likelihood and severity of relapse (5) and lowers depression and other mental health issues (6). Employment is also associated with fewer rearrests (7) and more time spent crime-free among offenders (8). Reports also suggest that stable employment eases community reintegration and has been correlated to several factors
Address correspondence to J. Matthew Webster, PhD, Department of Behavioral Science, 120 Medical Behavioral Science Building, Lexington, KY 40536-0086, USA. Tel: +1 859 323 6100. Fax: +1 859 323-5350. E-mail: [email protected]
Received 4 March 2013 Revised 15 October 2013 Accepted 18 October 2013 Published online 9 January 2014
that improve treatment outcomes and successful community reentry. Specifically, former treatment clients with regular employment report healthier social and professional networks as well as decreased involvement in their former substanceusing networks (9). Stable employment also provides a steady income and enables legal economic independence. Time at work increases time involved in a constructive, valuable activity and reduces the free time that may have previously been devoted to unhealthy behaviors such as drug use. Although employment positively correlates with other indicators of treatment success, substance abusers frequently face barriers in finding and keeping a job. These individuals often have poor work habits, including regular absences, tardiness, sleeping on the job, using vulgar language in the workplace (10,11), and have difficulties staying organized, handling stress, meeting deadlines, and maintaining stamina during the day (12). Drug-involved offenders have the additional burden of poor employment history due to incarceration (8) and the stigma associated with a criminal record (13,14) and status as a drug abuser (15) to overcome when pursuing employment opportunities. As a result, the
20 14
Introduction
History
Twelve-month employment outcomes
DOI: 10.3109/00952990.2013.858722
most commonly available jobs for drug-involved offenders are poorly paid, entry-level service jobs. To help substance abusers overcome these obstacles, employment training programs and interventions have been incorporated into existing substance abuse treatment programs, but these employment services vary widely across programs with mixed results (16). Some studies have shown that employment interventions lead to positive changes among treatment clients. For example, Hall et al. (17) conducted an employment workshop (Job Seeker’s Workshop) for parolees and probationers with a history of drug use and concluded that the intervention significantly increased levels of employment. Other studies have found that employment intervention programs for substance abuse treatment clients have subtle, but positive, changes at treatment completion, including increases in the number of days paid for work (18) and in income (19). McIntosh et al. (9) concluded that clients who received employmentrelated help along with traditional treatment were more likely to obtain paid employment. Not all studies, however, have found clear support for employment interventions. For example, Hubbard et al. (20) reported that only those clients in long-term treatment showed improvement in employment status as compared to clients in methadone maintenance treatment. Another study examining long-term substance abuse treatment clients and methadone maintenance found that neither group experienced any significant improvement in employment (21). More recently, a multi-site randomized trial of the Job Seekers’ Workshop found no difference in 12-week or 24-week employment outcomes, with only modest employment gains overall (22). Few interventions have focused specifically on druginvolved offenders (16). This literature gap led to the development of an employment intervention tailored to drug-involved offenders, which is grounded in established employment interventions and life skills training (23). The purpose of the current study was to examine employment outcomes for drug-involved offenders who received a tailored employment intervention. It was predicted that those who received the employment intervention would have better overall employment outcomes across multiple employment measures including higher rates of employment, days worked, and income.
Methods Participants Five hundred participants were recruited from two Kentucky drug court sites, which were administered following the established national drug court key components (24). Study eligibility was based on criteria for entry into the drug court program including non-violent charges, a self-admitted drug problem, urine test consent, and an Addiction Severity Index classification as a drug abuser. During the recruitment period, 525 drug court participants were eligible for the study across the two drug courts. Of those eligible, 500 were interviewed (95%), 7 clients refused, and 18 clients were terminated within 30 days of program entry and before a scheduled baseline interview could be conducted. After completing a baseline interview, participants were randomized to one of
201
two conditions, and then completed a 12-month follow-up interview. Four subjects died between the baseline and followup interview, and therefore, were ineligible for follow-up. The follow-up interview rate was 96% with 477 of 496 eligible participants interviewed. Because the participating drug courts valued employment, the research team was given permission to conduct the study independent of the drug court program. Drug court staff and judges were not informed about each client’s decision to participate in the study, their study condition, or level of participation. At baseline, the majority of participants were male (65.4%), more than half were white (61.8%), and a few were married (17.7%). Participants had an average age of 30.5 years (SD ¼ 8.7), a mean of 11.8 years of education (SD ¼ 2.0), and several incarcerations (M ¼ 4.6, SD ¼ 7.6). About half of the participants (53.6%) were employed either full-time or part-time. Participants most frequently reported cocaine (36.0%) as the drug causing the most problems followed by marijuana (23.4%) and alcohol (11.2%). An attrition analysis found no significant demographic differences between baseline and follow-up samples. Procedure Drug court clients were recruited for study participation between March 2000 and December 2002. Study participation was voluntary and subjects interested in participating were interviewed face-to-face by a trained interviewer within 2 weeks of drug court program entry. At the time of interview, participants were given a description of the project, procedures, and assurance of confidentiality following IRBapproved informed consent procedures. The interview lasted approximately 2 hours, and all participants were paid for their participation. Following the baseline interview, each participant was randomized to one of two groups: the employment intervention group or the drug court as usual control group. Participants randomized to the employment intervention began a series of individual and group sessions with a study-hired employment specialist with experience in both employment and substance abuse counseling. The employment specialist coordinated scheduling of intervention sessions with the drug court office. Individual sessions incorporated behavioral contracting, motivational interviewing, and strengths-based case management with a focus on employment barriers and resolving issues that impede employment success. Group sessions also used motivational interviewing as well as structured stories and thoughtmapping to address employment needs. The 26-session intervention was organized and delivered in three phases: obtaining employment, maintaining employment, and upgrading employment. Specific topics included networking, preparing effective resumes, writing a cover letter, filling out job applications, addressing and overcoming a criminal record in the application and interview, and proper workplace behavior (see 23 for a full description of the employment intervention, and 4 for employment outcomes related to the upgrading phase of the intervention). Sixty-one percent of intervention participants completed half or more of the available sessions.
202
J. M. Webster et al.
Approximately 10 months after the baseline interview, attempts were made to contact participants in order to schedule a 12-month follow-up interview. Locating and tracking activities were conducted without court assistance and included regular mailings, telephone calls, birthday and holiday card mailings, home visits, and electronic database checks. The 12-month interview contained the same measures as the baseline interview but focused on the time that had elapsed since the baseline interview. Measures Demographic information and employment data were collected using the Addiction Severity Index (25). Additional employment information was collected with measures developed at the University of Massachusetts School of Medicine (26). The key 12-month employment outcomes measures included: typical work status since the baseline interview, days paid for legal employment in the past 30 days and since the baseline interview (past year), and income from legal employment in the past 30 days and since the baseline interview (past year). Work status was treated as a categorical variable with four levels: full-time, part-time, unemployed, and not in the workforce (e.g. not working and not looking for a job). Days paid for legal employment and income from legal employment were continuous measures. Analytic plan In order to identify potential covariates for subsequent analyses, a series of t-tests and chi-square analyses were conducted to test for basic demographic differences between the intervention and control group. Next, a series of chisquare and ANCOVA intent-to-treat analyses were conducted to examine the effect of the intervention on employment outcomes including work status, days paid for legal employment (past year and past 30 days), and income from legal employment (past year and past 30 days). An alpha level of p50.05 was used for all analyses. An examination of the variables measuring days paid for legal employment and income from legal employment found that each was non-normally distributed (skewness range ¼ 0.23 to 4.43; kurtosis range ¼ 1.74 to 33.84). The income variables were moderately skewed and both the income and the days paid for legal employment variables were kurtotic. To remedy these distributional problems, a square root transformation of these dependent measures resulted in the greatest minimization of the skewness (range ¼ 0.92 to 0.80) and kurtosis (range ¼ 1.69 to 1.87) statistics across the range of variables affected (27). The transformed variables were used in subsequent analysis. Because almost half (44%) of the sample was employed full-time at baseline, we also examined the effectiveness of the intervention for those participants who were in the most need of employment assistance. Recognizing that many druginvolved offenders have unstable employment, we operationalized this by examining participants’ work trajectories. A pre-baseline work trajectory was created by examining participants’ typical job status in the 6 months prior to entering drug court and their typical job status since entering drug court. Both of these variables were measured at baseline.
Am J Drug Alcohol Abuse, 2014; 40(3): 200–205
Participants who experienced a decrease in their employment status (e.g. full-time to part-time or part-time to unemployed) or stayed unemployed from one time point to the next were considered to be on a negative pre-baseline work trajectory. Those participants whose employment status did not decrease (e.g. stayed employed full-time or went from unemployed at baseline to part-time at follow-up) were collapsed into a positive pre-baseline work trajectory group. A second work trajectory, follow-up work trajectory, was created as an outcome variable. This variable was created similarly to the pre-baseline work trajectory variable but examined participants’ employment status at the 12-month follow-up interview relative to their employment status at the time of the baseline interview. Thirty participants did not fit clearly into one of the follow-up work trajectory groups (e.g. missing data) and were eliminated from subsequent analyses.
Results Significant differences in age between the intervention group and control group were found at baseline (Mintervention ¼ 29.7, SD ¼ 8.4 vs. Mcontrol ¼ 31.2, SD ¼ 9.0; t(498) ¼ 1.96, p ¼ 0.05, d ¼ 0.18) and at 12-month follow-up (Mintervention ¼ 30.7, SD ¼ 8.4 vs. Mcontrol ¼ 32.6, SD ¼ 9.0; t(475) ¼ 2.28, p ¼ 0.02, d ¼ 0.21). Likewise, differences between the two groups in the number of years of education significantly differed at both time points (Mintervention ¼ 11.5, SD ¼ 1.9 vs. Mcontrol ¼ 12.0, SD ¼ 2.0 at baseline and followup; t(498) ¼ 2.47, p ¼ 0.01, d ¼ 0.22) and t(475) ¼ 2.70, p ¼ 0.01, d ¼ 0.25, respectively). No other group differences for demographic information (gender, race, marital status, drug court site) were found; thus, only age and education were used as covariates. The intent-to-treat analysis found that the employment intervention had a positive effect on the number of days paid for legal employment (see Table 1). Participants in the intervention group reported more days paid for employment in the past year (M ¼ 210.1, SD ¼ 114.1) than those in the control group (M ¼ 199.9, SD ¼ 130.1, d ¼ 0.20; F(1, 464) ¼ 4.69, p ¼ 0.03). No differences were found, however, for other employment outcomes, including the percentage of participants moving from unemployed at baseline to employed at follow-up (approximately 30% for both groups). All analyses of key employment outcomes (work status, days paid for legal employment, and income from legal employment) were then repeated to take into account participants’ pre-baseline work trajectory. An additional employment outcome, follow-up work trajectory, was also examined. Work status A chi-square analysis was performed to test whether work status patterns differed across combinations of the experimental condition (control vs. intervention) and the prebaseline work trajectory (negative vs. positive). A marginal difference in work status across the groups was found, 2 (6, N ¼ 447) ¼ 12.30, p ¼ 0.056, Cramer’s V ¼ 0.12. The pattern of results showed that those with a negative prebaseline work trajectory in the control condition had a lower percentage of full-time employment versus other groups.
Twelve-month employment outcomes
DOI: 10.3109/00952990.2013.858722
203
Table 1. Intent-to-treat and work trajectory analyses of 12-month employment outcomes. Intent-to-treat analysis
Work trajectory analysis Negative trajectory
Control (n ¼ 233) Work status Working fulltime (%) Working part-time (%) Unemployed (%) Not in the workforce (%) Days paid for legal employment In the past yeara,b In the past 30 daysb Income from legal employment ($) In the past yearb In the past 30 daysb
Intervention (n ¼ 244)
Control (n ¼ 64)
Intervention (n ¼ 68)
Positive trajectory Control (n ¼ 160)
Intervention (n ¼ 155)
59.7 15.9 1.7 22.8
67.6 12.9 1.7 17.8
51.6 12.5 1.6 34.4
65.7 16.4 1.5 16.4
63.8 17.5 1.9 16.9
69.0 12.3 1.9 17.3
199.9 16.1
210.1 16.7
157.0 11.7
188.9 15.0
217.3 17.8
219.3 17.5
11 073.6 845.5
9863.9 809.9
6888.4 531.8
8623.1 635.0
12 779.6 975.4
10 407.3 881.0
a
Significant intent-to-treat effects, p 0.05; bSignificant work trajectory effects, p 0.05.
A second chi-square analysis was conducted in which work status was dichotomized into working vs. not in the workforce and resulted in a significant effect for condition, 2 (3, N ¼ 447) ¼ 9.67, p ¼ 0.022, ’ ¼ 0.15. Participants in the control group with a negative trajectory had a significantly lower rate of employment than other groups (64.1% vs. 80.8%, 81.3%, and 82.1%). Days paid for legal employment The days paid for legal employment and the income from legal employment were analyzed in a series of 2 (Intervention Group: intervention vs. control) 2 (Pre-baseline Work Trajectory: negative vs. positive) ANCOVA tests controlling for age and education as in the intent-to-treat analysis. A significant main effect for pre-baseline work trajectory was found for the number of days paid in a legal job in the past year, F(1, 437) ¼ 10.82, p ¼ 0.001, d ¼ 0.31, such that those with a positive pre-baseline work trajectory had more days paid for legal employment than those with a negative prebaseline work trajectory (see Table 1). In addition, a main effect of intervention group emerged, F(1, 437) ¼ 6.85, p ¼ 0.009, d ¼ 0.25. Similar to the intent-to-treat analysis, those randomly assigned to the intervention group had more days of paid legal employment in the past year than those in the control group. A significant main effect of pre-baseline work trajectory also was found for the number of days paid for legal employment in the past 30 days, F(1, 400) ¼ 6.57, p ¼ 0.011, d ¼ 0.26. Participants with a positive pre-baseline work trajectory worked more days in the past 30 days than those with negative pre-baseline work trajectories. The intervention group, however, did not have a significant effect on the number of days paid for legal work in the past 30 days. Income from legal employment The amount of income from legal employment significantly varied as a function of pre-baseline work trajectory (F(1, 435) ¼ 11.77, p ¼ 0.001, d ¼ 0.33). Participants who had a negative pre-baseline work trajectory earned significantly less money for legal employment in the past year than
participants with a positive pre-baseline work trajectory (see Table 1). This main effect was qualified by a significant pre-baseline work trajectory by intervention group interaction, F (1, 435) ¼ 5.14, p ¼ 0.024, d ¼ 0.22. The interaction effect was characterized by a more pronounced difference between the intervention and the control group in the negative pre-baseline work trajectory than in the positive pre-baseline work trajectory group. This effect was not detected in the analysis of income from legal employment in the past 30 days, but the same main effect of pre-baseline work trajectory emerged, F (1, 403) ¼ 7.91, p ¼ 0.005, d ¼ 0.28. Change in work trajectory Logistic regression analyses using age and education as covariates were used to test whether the employment intervention was related to follow-up work trajectory. This analysis was performed separately for those who had a negative pre-baseline work trajectory and those who had a positive pre-baseline work trajectory. For those who had a positive pre-baseline work trajectory, the intervention was unrelated to their follow-up work trajectory status (OR ¼ 1.22, CI(95%) ¼ 0.58–2.57). However, for those who had a negative pre-baseline work trajectory, the intervention was strongly related to their follow-up work trajectory status (OR ¼ 6.83, CI(95%) ¼ 1.38–33.72). Specifically, participants who had a negative pre-baseline work trajectory in the employment intervention group had almost seven times the odds of no longer being on a negative work trajectory at 12-month follow-up than those in the control group.
Discussion The purpose of this study was to examine employment outcomes for drug-involved offenders who received a tailored employment intervention. The intent-to-treat analyses found that the intervention was successful in increasing the number of days worked in the past year, but no other employment outcomes. However, subsequent analyses focusing on work trajectories showed that those who were on a negative prebaseline work trajectory were most helped by the intervention. Specifically, participants with a negative pre-baseline
204
J. M. Webster et al.
work trajectory, who were randomly assigned to the tailored employment intervention, were more likely to be in the workforce, earn more income from legal employment, and more likely be on a positive work trajectory at 12-month follow-up than those with a negative pre-baseline work trajectory and assigned to the control group. These results support the efficacy of the intervention for those who were in greatest need of employment assistance. The present research adds to the growing substance abuse and employment literature in several ways. First, this study demonstrates that an innovative employment intervention tailored for drug-involved offenders can positively affect 12-month employment outcomes, most strongly for those without recent employment success. The substance abuse research literature is replete with references to the importance of employment in the recovery process, yet the literature on strategies to enhance employment is equivocal. For example, recent examinations of employment interventions for substance abusers have been inconsistent with some finding success (e.g. 28) and others reported no intervention effects on key employment outcomes (e.g. 22). These inconsistent findings are further complicated by differences in study design and employment measures. The present study incorporates a number of strengths that make the interpretation of the results clearer. Specifically, this study used random assignment, had high follow-up rates (96%), used multiple employment measures, and tested an intervention that was tailored to drug-involved offenders. Employment is an important dynamic process for many drug abusers, shifting from job to job and from periods of employment and unemployment. The current research used a work trajectory approach to capture this movement across employment statuses over time. Previous studies have used work or employment trajectories as a predictor or outcome variable (29,30), and some have applied concepts similar to the negative and positive work trajectories used in the current study (e.g. 31). However, only a few studies have examined work trajectories among drug abusers (32), and to our knowledge, none have used this approach specifically with drug-involved offenders. While the intent-to-treat analysis provided inconsistent support for the efficacy of the intervention, the work trajectory analyses highlighted the intervention’s positive impact on several employment outcome measures including increased employment, days paid for employment, and employment income. These findings suggest that it may be important to look at employment patterns and work trajectories in addition to current employment status when targeting individuals for employment interventions. Findings also highlight the idea that employment interventions for drug-involved offenders may be most appropriately targeted to those who have the least employment success, particularly if intervention resources are limited. Although the intervention included information relevant to employed participants (i.e. maintaining and upgrading employment), participants in the negative pre-baseline work trajectory most benefitted from the intervention. Furthermore, future studies may examine how much employment intervention participation is needed to enhance favorable outcomes. Along with participation data, measures of intervention engagement can be critical to understanding overall
Am J Drug Alcohol Abuse, 2014; 40(3): 200–205
intervention effectiveness and provides another direction for future research. In addition to the importance of employment patterns, this study shows the utility of using multiple measures of employment to examine outcomes. In the present research, outcome measures including employment status, days worked, days paid, and income were used with inconsistent results. In fact, some of these employment measures may be more or less mutable within a 12-month time frame depending on the individual and nature of the job. Other substance abuse research has shown that employment is a multidimensional construct and these dimensions differentially correlate with drug use and criminality (15). Future studies that examine employment as a key outcome, therefore, should carefully select the type and number of employment measures to ensure that more than one facet of this construct is measured. Furthermore, additional research should more closely examine how changes in these different employment outcomes affect other important variables such as drug use. Finally, the study provides an example of how a criminal justice setting – drug courts in this case – can be a platform for randomized controlled studies, particularly when judges and staff ‘‘buy in’’ to the research. Drug court judges involved in this study strongly believed that employment is important for drug court clients and allowed our research team to randomly assign participants to the intervention or control group. They were kept apprised of study progress, interim findings from aggregated data, and given opportunities to provide feedback to the research team during the course of the study. There is mounting evidence that drug courts are effective in reducing recidivism (33), and interventions focusing on areas that relate to treatment retention and outcomes, such as employment, may further reduce drug use and criminality. Drug courts may be able to enhance employment by investing more resources into integrating job training with drug court programming and by developing ties with community employers to hire clients during or upon completion of drug court. This study has limitations. First, the generalizability of findings is unclear since the data are not recent. The initial intent-to-treat analyses were performed when the study was completed. However, a continued interest in the employment of drug-involved offenders led to further data analysis using a work trajectory approach several years later. In addition, generalizability is limited since data were collected from two Kentucky drug courts. Although the administration of these drug courts remains the same since data collection, results of the intervention could vary for other samples of drug courts clients or groups of offenders at different time points, in other geographic areas, and under different economic conditions. In fact, both the historic and current economic climate of Kentucky, which is often associated with rural poverty, may have attenuated the overall effects of the intervention. Furthermore, several employment variable comparisons were made, which could increase the likelihood of a Type I error, although the pattern of significant differences in the work trajectory analyses were consistent across the days worked and the income variables. Finally, this study relies on self-reported employment data, which could be limited by recall, truthfulness, or demand characteristics associated with
Twelve-month employment outcomes
DOI: 10.3109/00952990.2013.858722
assignment to the intervention or control condition. For example, intervention participants could have over-reported their employment; however, it seems that this type of bias would have appeared in the intent-to-treat analyses if it were present. Despite these limitations, the current study provides evidence that a tailored employment intervention for druginvolved offenders can produce positive outcomes. Future research should examine employment interventions for druginvolved offenders that include both the development and enhancement of employment skills along with job placement, which might include cottage industries and other small businesses.
Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. This study was supported by Grant R01 DA013076 from the National Institute on Drug Abuse; Carl G. Leukefeld, Principal Investigator; and by the staff and resources of the Center on Drug and Alcohol Research at the University of Kentucky. Opinions expressed are those of the authors and do not represent the position of the National Institute on Drug Abuse.
13. 14. 15. 16. 17. 18.
19.
20.
21.
References 1. Corsi KF, Kwiatkowski CF, Booth RE. Treatment entry and predictors among opiate-using injection drug users. Am J Drug Alcohol Abuse 2007;33:121–127. 2. Platt JJ. Vocational rehabilitation of drug abusers. Psycholog Bull 1995;117:416–433. 3. Ginexi EM, Foss MA, Scott CK. Transitions from treatment to work: employment patterns following publicly funded substance abuse treatment. J Drug Issues 2003;33:497–518. 4. Leukefeld CG, Webster JM, Staton-Tindall M, Duvall J. Employment and work outcomes among drug court clients: 12-month outcomes. Subst Use Misuse 2007;42:1109–1126. 5. Evans JL, Hahn JA, Lum PJ, Stein ES, Page K. Predictors of injection drug use cessation and relapse in a prospective cohort of young injection users in San Francisco, CA (UFO Study). Drug Alcohol Depend 2009;101:152–157. 6. Wolkstein E, Bausch R, Weber G. Work as a critical component of recovery. Paper presented at New York University, 2000. Retrieved September 30, 2011 from http.//www.med.wright.edu/ 7. Evans E, Huang D, Hser Y. High-risk offenders participating in court-supervised substance abuse treatment: characteristics, treatment received, and factors associated with recidivism. J Behavioral Health Serv Res 2011;38:510–525. 8. Tripodi SJ, Kim JS, Bender K. In employment associated with reduced recidivism? The complex relationship between employment and crime. Int J Offender Therapy Compar Criminol 2010;54: 706–720. 9. McIntosh J, Bloor M, Robertson M. Drug treatment and the achievement of paid employment. Addiction Res Theory 2008;16: 37–45. 10. Carpenedo CM, Needham M, Knealing TW, Kolodner K, Fingerhood M, Wong CJ, Silverman K. Professional demeanor of chronically unemployed cocaine-dependent methadone patients in a therapeutic workplace. Subst Use Misuse 2007;42:1141–1159. 11. Wong CJ, Silverman K. Establishing and maintaining job skills and professional behaviors in chronically unemployed drug abusers. Subst Use Misuse 2007;42:1127–1140. 12. Batiste LC. Accommodation and compliance series: employees with drug addiction. Accommodation and Compliance Series,
22.
23. 24. 25. 26. 27. 28. 29. 30.
31.
32. 33.
205
Job Accommodation Network, 2005. Retrieved from http:// www.jan.wvu.edu/media/drugadd.html Harris PM, Keller KS. Ex-offenders need not apply: the criminal background check in hiring decisions. J Contemp Criminal Justice 2005;21:6–30. Holzer HJ, Raphael S, Stoll MA. Employment barriers facing exoffenders. Center for the Study of Urban Povery Working Paper Series, 2003. Webster JM, Staton-Tindall M, Duvall JL, Garrity TF, Leukefeld CG. Measuring employment among substance-using offenders. Subst Use Misuse 2007;42:1187–1205. Magura S, Staines GL, Blankertz L, Madison EM. The effectiveness of vocational services for substance users in treatment. Subst Use Misuse 2004;39:2165–2213. Hall SM, Loeb P, LeVois M, Cooper J. Increasing employment in ex-heroin addicts II: criminal justice sample. Behavior Ther 1981; 12:443–452. Lidz V, Sorrentino DM, Robison L, Bunce S. Learning from disappointing outcomes: an evaluation of prevocational interventions for methadone maintenance patients. Subst Use Misuse 2004; 39:2287–2308. Back SE, Sonne SC, Killeen T, Dansky BS, Brady KT. Comparative profiles of women with PTSD and comorbid cocaine or alcohol dependence. Am J Drug Alcohol Abuse 2003;29: 169–189. Hubbard RL, Craddock SG, Flynn PM, Anderson J, Etheridge RM. Overview of 1-year follow-up outcomes in the drug abuse treatment outcome study (DATOS). Psychol Addict Behaviors 1997;11: 261–278. Schildhaus S, Gerstein D, Brittingham A, Cerbone F, Dugoni B. Services research outcomes study: overview of drug treatment population and outcomes. Subst Use Misuse 2000;35:1849–1878. Svikis DS, Keyser-Marcus L, Stitzer M, Rieckmann T, Safford L, Loeb P, Allen T, et al. Randomized multi-site trial of the Job Seekers’ Workshop in patients with substance use disorders. Drug Alcohol Depend 2012;120:55–64. Leukefeld C, McDonald HS, Staton M, Mateyoke-Scrivner A, Webster JM, Logan TK, Garrity T. An employment intervention for drug-abusing offenders. Federal Probation 2003;67:27–31. National Association of Drug Court Professionals. Defining drug courts: the key components. NCJ 205621, 1997. McLellan AT, Luborsky L, O’Brien CP, Woody GE. An improved evaluation instrument for substance abuse patients: the Addiction Severity Index. J Nerv Mental Dis 1980;168:26–33. Barrow S, Hellman F, Lovell AM, Plapinger JD, Robinson CR, Struening EL. Personal history form. New York: New York State Psychiatric Institute; 1985. Tabachnick BG, Fidell LS. Using multivariate statistics. Boston: Allyn and Bacon; 2001. Walls RT, Batiste LC, Moore LC, Loy B. Vocational rehabilitation and job accommodations for individuals with substance abuse disorders. J Rehab 2009;75:35–44. Becker D, Whitley R, Bailey EL, Drake RE. Long-term employment trajectories among participants with severe mental illness in supported employment. Psychiatric Serv 2007;58:922–928. Meyer JD, Mutambudzi M. Construction of life-course occupational trajectories: evidence for work as a mediator or racial disparities in hypertension. J Occup Envir Med 2012;54: 1201–1207. Virtanen P, Vahtera J, Broms U, Sillanma¨ki L, Kivima¨ki M, Koskenvuo M. Employment trajectory as determinant of change in health-related lifestyle: the prospective HeSSup study. Eur J Public Health 2008;18:504–508. Huang DY, Evans E, Hara M, Weiss RE, Hser YI. Employment trajectories: exploring gender differences and impacts of drug use. J Vocat Behav 2011;79:277–289. Mitchell O, Wilson DB, Eggers A, MacKenzie DL. Assessing the effectiveness of drug courts on recidivism: a meta-analytic review of traditional and non-traditional drug courts. J Crim Justice 2012; 40:60–71.
Copyright of American Journal of Drug & Alcohol Abuse is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
