Abstract

Utilizing 434 offenders who participated in a Lake City Access to Recovery program sponsored by the Substance Abuse and Mental Health Services Administration (SAMHSA), empirical research validates the use of the Addiction Severity Index (ASI-6) as a good tool in predicting offender reincarceration from one year of release (28%) through seven years of follow-up (59%). Furthermore, the study provides evidence of the conceptual foundation of Sampson and Laub›s life course theory that explains the long-term pathways most offenders face when returning to their communities.

Keywords:Life course theory, Laub and Sampson, Addiction Severity Index, ASI-6, SAMHSA, Access to Recovery, substance use, recidivism, reoffending, reincarceration

Keywords:SAMHSA: Substance Abuse and Mental Health Services Administration; ASI-6: Addiction Severity Index; ACLU: American Civil Liberties Union; ATR: Access to Recovery; VIF: Variance Inflation Factor

Introduction

America is synonymous with the War on Drugs, coined in 1971 by President Nixon and implemented by Ronald Reagan’s 1980s administration (Dholakia, 2021). This Get Tough on Crime initiative led to the incarceration binge we see today (Sawyer and Wagner, 2020), where jails and prison populations are bursting at their seams. A recent American Civil Liberties Union (ACLU) report suggests that 65% of Americans now feel it’s time to eliminate criminal penalties for drug possession and invest resources towards treatment and addiction services Dholakia [1]. [3] A 2011 National Drug Intelligence Center study estimated drug offence-related costs at $113 billion. This cost has likely increased since 2011, as the government spends 182 billion annually on the correctional system Dholakia [1].

In 2020, approximately 2.3 million Americans were housed in 1,833 state prisons, 110 federal prisons, US Territorial prisons, 1,772 juvenile facilities, 3,134 local jails, 218 immigration detention facilities, 80 Indian detention facilities while also inclusive of military facilities, civil commitment centers and state psychiatric hospitals Sawyer and Wagner [3]. Five years later, just less than two million people are incarcerated in 1,566 state prisons, 98 federal prisons, 1,277 juvenile correctional facilities, 3,116 local jails, 133 immigration detention facilities, and 80 Indian country jails, as well as in military prisons, civil commitment centers, state psychiatric hospitals, and prisons in the US territories Sawyer and Wagner [4]. So, progress has been made in reducing the number of people who are incarcerated and the facilities that house them. Despite the gains and progress of reducing the total population, the correctional system is still backlogged and overpopulated with suspects and offenders churning in and out of the system. As Sawyer and Wagner [3,4] report, while there are fewer jail facilities in 2025 than in 2020, jail admissions were more than seven million (despite the daily numbers of approximately 562,000 in local jails and 203,000 in federal jails and prisons). Nearly 7 of 10 (65%) of prisoners have an active substance abuse issue, which likely precipitates their criminal adventures (Center on Addiction) [5].

Tonry and Wilson [6] coined the term drug-crime nexus to appreciate the relationship between substance use and criminal activity; in that the relationship is so strong that crime can lead to substance use, or the opposite could exist Goldstein [7]. The Substance Abuse and Mental Health Services Administration (SAMSHA) has suggested that offenders with treatment and rehabilitation could create an anticipated 80% reduction in criminal activity (2006:4). As such, SAMHSA has emphasized the need to prioritize offenders who had issues with drug use to prevent future criminality and/or the opposite; that the reduction of criminal activity would reduce substance use. Notwithstanding the difficulty of substance use treatment, the human costs of arrest, booking, bail,civil/criminal trials (and tribulations) impact the suspects and offenders in the system (as well as victims and the criminal justice system personnel) who all view (in)justice, deterrence, and treatment (or the lack thereof, differently).

Of the nearly two million people incarcerated in 2025, almost everyone, approximately 98%, of people admitted to correctional facilities will be released and return to their communities at some point Petersilia [8]; Sabol [9]. Therefore, lessening the cycling or churning effect is integral to stabilizing the criminal justice system. Then consider, as Wilson [10] and Walker [11] report, whether there is any significant merit to imprisoning offenders to deter their previous criminal activity. Incarceration won’t fix the underlying causes of crime. Additionally, attempting to provide treatment and rehabilitation to people who are involuntarily admitted to their custodial facilities wouldn’t likely be anything more than reactive (versus an offender’s need for readiness to change). So, we are left with a multitude of questions. Can offenders adjust and adapt to these knifing off points of custody (according to Laub and Sampson [12]) to reintegrate back to their families and communities? Who is at risk, and should fewer be incarcerated post-release? What programs or services do offenders require to best desist from crime (because general deterrence and incarceration don’t appear to work)? And furthermore, is there an empirical approach that may validate these processes? This article is meant to shed some light on these questions.

Recidivism

Offender reoffending reflects the effect or impacts of the stimulus of one or more criminal justice interventions to enable an offender to desist from criminal activity actively. These numbers do not, however, tell the story of the affect on physical health, mental health, substance use, and other structural factors (such as poverty, unemployment) on the offender, victims, their families, and the communities at large. While 98% of offenders will be released from custody Petersilia [5], most offenders will fail in desistance from criminal activity.

National Bureau of Justice Statistics studies consistently report that as many as seven out of ten offenders will be re-arrested within three years of release, while three to four offenders will be reconvicted or reincarcerated within three years of release from custody (Langan and Levin, 2002; DuRose et al, 2014; Antenangeli and DuRose [13]. The question was always whether these threeyear follow-ups on recidivism were simply a short-term issue of criminal offending (perhaps considered casual or variable offending) or whether offenders were more chronic offenders where they desist from crime due to merely aging out of crime Ray and Jones [14] or life’s turning points Laub and Sampson [12]. Thankfully, more recent research in the last ten years has shed more empirical evidence on whether incarceration/ reincarceration and/ or other factors are reducing an offender’s long-term criminogenic outcomes.

A 2021 report on ten-year recidivism rates from prisoners across twenty-four States suggests that crime continues but progressively diminishes over time Antenangeli and DuRose [13]. The Bureau of Justice reported that while annual percentages of re-arrest, re-conviction, and reincarceration fell over time, 82% of offenders were re-arrested over ten years, and 75% of identified drug offenders were re-arrested for non-drug offences 10 years later Antenangeli and DuRose [13]. In terms of re-conviction, approximately seven in ten offenders (69%) released in 2008 were convicted of a new arrest in 2018 Antenangeli and DuRose [13]. Like other studies on recidivism, this Bureau of Justice Statistics was unable to access all 24 State databases to determine reincarceration (because most correctional systems operate across a variety of digital software and platforms). Of the 18 States that were able to report, the authors report that 61% of prisoners released in 2008 were reincarcerated in 2018. According to the Bureau of Justice Statistics (2021), within the first year, 31% of prisoners returned to custody, which continuously increased to 49% (by year three), 55% (by year five), 58% (by year seven) to 61% (by year ten). These statistics highlight the lack of long-term offender desistance from crime and reincarceration. As a result, the churning continues for most offenders once they are released from custody.

The churning of offenders in and out of criminal justice facilities (whether it be as a result of a misdemeanor or felony offense) significantly impacts system resources, never mind the offender, victims, families, and/or communities who are directly affected. Therefore, identifying needs-assessment tools to profile offenders once released could assist in reducing recidivism in the short and/or long term. While statistics offer insight into the prevalence of the problem of reoffending, we often require more insight into why offenders may or may not desist from crime Ray and Jones [14]. This study explores the merits of the Addiction Severity Index (ASI-6) in predicting offender reincarceration to reduce the overload to the criminal justice system, while inevitably assisting offenders with their short and/or long-term needs and assessments.

Addiction Severity Index (ASI-6)

People live complex lives, and the new buzzword of clinical therapy is taking a holistic approach to treatment. This is undoubtedly a worthwhile venture; however, it doesn’t often allow for more objective measurement of the priority of needs and wants of those suffering from the complexities of physical and mental health with corresponding substance use and criminal activities (and/or desistence). Losing one’s relationships (or putting them on hold) and losing freedom of movement into custody and confinement have both short- and long-term effects on the offender, victims, their families, and the communities they are located in. While incarceration was always meant as a method to protect victims and society from dangerous criminal offenders, we must continue to ask if we are truly isolating those who are hazardous versus identifying alternatives (while also protecting our communities). While these ideological approaches are discussed in academic circles, this study focuses on the priorities and needs of the offender to determine if the ASI-6 could be a clinical tool to prevent prisoners from committing more crime and reduce the impact on potential victims, families, and communities.

The Addiction Severity Index (hereinafter known as the ASI-6) was first introduced in the United States as a clinical tool to assess the incidence and prevalence of substance use and co-occurring impairments by McLellan et al (1980). The ASI-6 was meant to assess the immediate, short, and long-term needs of patients/ subjects Hendriks [15]; McLellan [16]; Leonhard [17]. Makela [18] outlined the validated reliability of the ASI-6 within various clinical environments to assist people in different inpatient, outpatient and residential clinical settings and programs. The popularity of the ASI-6 among clinicians allowed for the evolution and modification of language and query-based items that focus on the depth of a respondent’s cognitive processing, especially immediate gratification Kokkevi and Hartgers [19]. The use of the ASI-6 has been adapted to predict antisocial behavior Cacciola [20], psychopathology Cacciola [21], and compulsive gambling Lesieur and Blume [22]. However, before this study, few studies studied the Addiction Severity Index’s predictive power on recidivism Seredycz [23].

The ASI-6 utilized in this study examines 118 closed and open-ended questions across seven aggregate domains McLellan [24]. Clinicians (or in the case of this study, intake screeners) are meant to assess participants in seven core domains: one’s (i) medical condition, (ii) employment and support, (iii) incidences and prevalence of alcohol use and (iv) drug use, (v) legal status, (vi) family relations, and (vii) psychological/psychiatric status. Participants self-report their behaviors across different time intervals (from last occurrence, monthly, six months, and lifetime prevalence) to assess the frequency and duration of problematic behavior McLellan [24]. The ASI-6 is a well-designed tool that takes 45-60 minutes to administer, ensuring respondents are still actively responding to the questions. The 15-minute scoring also allows for speed in clinical assessment to determine how to best proceed with each person/ offender. Within each of the seven domain areas, a composite score is derived from the questions that were asked by the interviewer Rosen [25]. The composite scores are considered measures of problem severity, with higher scores indicating greater problem severity Butler [26]; Grissom and Bragg [27]. Each of these seven domains should predict the circumstances that offenders face in their barriers to reintegration Petersilia [7].

Methodology

This study utilizes data tracking 434 offenders within a federally funded Access to Recovery (ATR) Lake City site. Access To Recovery (ATR) was a Substance Abuse and Mental Health Services Administration (SAMSHA) [28] voucher program that allowed offenders with substance abuse to self-select clinical treatment and recovery support services. Limiting participants to those with substance use issues is a limitation of the study. However, these parameters reinforce that these offenders are likely at more risk than offenders without substance use issues Antenangeli and DuRose [13]. Inclusion criteria for participation in the Access to Recovery (ATR) Lake City program were: i) adult aged 18 or older; ii) residency; iii) a need for alcohol and/or drug treatment; and iv) informed consent for treatment and research purposes Seredycz [23]. Consent for treatment and research purposes ensured that offenders would attain a collaborative/ holistic approach of formal support (80-plus treatment providers, case managers, probation/ parole agents) and informal supports (like families and children, should the offender wish). Overall, 434 offenders met the criteria over the first year of operation and were used for this analysis. The total number of possible jail and prison released offenders who tried to access the program was 456 that year; however, records and documentation of 22 offenders who participated in the program were ineligible and removed from the study.

Within forty-eight hours of a prisoner’s release (and acceptance into the ATR program), offenders would be referred to an intake specialist where the ASI-6 would be utilized Seredycz [23]. The ASI-6 and its seven composite scores were used to assist and prioritize offender needs throughout their rehabilitation and treatment program, which could last as long as eighteen to twenty-four months. Within the intake process, offenders self-select from a list of over 80 (religious and non-religious) treatment providers that they felt could best serve them, while being guided by the ASI-6 scoring parameters. Reincarceration was the predicted offender outcome and required a triangulation of data sources.

The first source was the review of social services/ clinician case managers’ case notes. While these case managers and clinicians may not be privy to all criminal justice system information, dates and times would articulate the length of lapses in treatment, supervision, meetings, and/or incarceration. The second validation measure was the State’s parole database system, which included all case notes and contact information for offenders under community supervision after release from custody. This documentation offered data to validate the discretion associated with additional custody because of technical violations and/or arrest for new/ pending charges. The third validation measure was using a State correctional database system which could identify an offender based on their identification number, name, and date of birth. These three methods were extremely reliable in measuring an offender’s potential reincarceration in and out of the State’s jurisdiction (if they abscond). Utilizing a one-group one-shot treatment design involved exposing a group of offenders to treatment, which was further followed by annual interval incarceration measures over a ten-year follow-up. The limitation of this method is that it is not an experimental design. This was because a control group could not be accessed for a comparison study. While there was no control group, the study did assess all offenders in the program (rather than a random sample). While limitations should be considered, it would still be regarded as reliable and valid for the program participants Sherman [29]; Farabee [30].

ASI-6 initial analysis

As explained previously, the ASI-6 was used to assess offender’s histories, frequencies, and consequences based on seven domains: (i) drug and (ii) alcohol use, (iii) medical, (iv) employment, (v) legal, (vi) social and family, and (vii) psychological functioning. Table 1 shows the profile of the 434 offenders sampled and reported areas of need within the AS1-6. Table 2 offers more insight into the levels of severity as scored by clinicians to determine whether the risk is none to severe. Respectfully, each of the Tables is ranked based on the most prevalent areas of need and severity of risk.

The application of both Tables explores the complexities and co-occurring factors associated with how an offender perceives themselves (Table 1) versus the reality of the barriers, situations, and/or circumstances they face as scored within the risk-need assessment. For instance, while drug use was considered the most significant area of need as reported by offenders (82%), it ranked fourth (of seven) in terms of severity of risk factors, where only 4.2% of offenders scored in the high or severe categories. While it is encouraging that offenders felt it was an area of concern, the severity of high or severe risk was a lesser concern as compared to one’s medical/ physical health issues (22%) and mental health/ psychological issues (18%). This could hypothetically explain an offender’s expected or anticipated assistance within health care facilities (or lack thereof); further accentuating how institutions and their care (or lack thereof) are tipping points for offenders according to life course theory Sampson and Laub [31]. Perhaps the medical and psychological services they require are unaffordable, unattainable, or have not fixed the issues (based on previous situations). This reiterates employment’s emphasis and primary risk factor in sustaining the necessary social bonds to find stability versus the immediate gratification of criminal enterprise Laub and Sampson [12]. Again, while an offender’s legal situation (particularly the monitoring and supervision of their activities by probation/ parole officers) is an area they perceive to be important (77%), it was considered a relatively low risk factor within the ASI-6 scoring (when we combine high and severe risk at 16%). Again, offenders felt it was a very pressing need when abiding by the terms and conditions of probation/ parole would likely reduce their level of concern. These Tables highlight the need for riskneed assessment. They are unaware of why empirical clinical tools should be more widely available as offenders’ previous experiences or perceptions may be clouding their judgment on the severity of certain risk factors.

What is present in both Tables is the importance of attaining assistance for substance use and unstable informal social supports. This too is highlighted by Laub and Sampson [12] as potential turning points in one’s life where an offender’s pathways/ trajectories are directly impacted by the drug-crime nexus and/ or co-occurring factors Bunting [32]; Visher and Travis [33]. Employment was the most significant reported offender’s area of need (98%) and was also the highest-severe risk factor (93%) based on ASI-6 scoring. Substantial evidence suggests that gainful employment leads to success in desisting from crime. A recent study by Bunting [32] reports that unemployed individuals have a 127% increase in odds of reincarceration. Consistent with the findings of Bunting et al. and other earlier studies Sherman [29]; Petersilia [8] would suggest that life course theory (as previously explained by Sampson and Laub [31] would explain how integral developing social capital (money, housing, insurance, stability) is to reintegrate into one’s community and prosperity. Employment was also found to be the most significant and severe barrier to an offender’s criteria. This would explain why (gainful and potentially meaningful long-term) employment has and will continue to be the most significant factor in reducing recidivism.

Reincarceration initial analysis

As explained, reincarceration data were triangulated across three validated data sources, including case managers, probation/ parole officer case notes, and official correctional sources. Table 3 presents the findings of the first seven years of statistics on the offender’s likelihood to return to custody once released.

There were changes and unavoidable intervals based on the triangulation of data sources and the life course theory assumptions. In year four, the data was too unreliable to report due to revisions of system software platforms and access. In years four to five, two Lake City ATR program participants were removed from the analysis. In years five to six, an additional two participants were removed. In year seven, 13 participants were removed from the analysis (due to mortality, alias removal/ name change, absconding from jurisdictions).

As viewed in Table 3, the reincarceration of offenders remained relatively high each year, with the most significant one-year increases between their first year of release (22.8%) and their last year, seventh year of recorded follow-up (20.1%), to 73.8%. This would suggest, similar to the US Bureau of Statistics (hereinafter known as BJS) 10-year follow-up report, offenders continue to offend far past a short-term transition period from custody despite the years progressing. However, Lake City ATR program’s offenders appeared to be more successful at avoiding short term custody than the offenders within 18 States reported by Antenangeli and DuRose [13]. ATR Lake City offenders were less likely (23%) than BJS offenders to be reincarcerated (31%) after the first year. This also occurred during years two (28% v. 43%) and three (33% v. 49%). This was also true for years five (41% v. 55%) and six (54% v. 57%). While the US BJS is not a perfect comparison, aggregate data might suggest that the Lake City Access to Recovery (ATR) offered the treatment and ancillary services that may have improved an offender’s success in reincarceration (versus a national comparison group). However, whatever gains the Lake City ATR program participants had been lost as reincarceration statistics were significantly higher in year 7 (74%) than the US BJS (58%). The statistics reported in this study and presented by Antenangeli and DuRose (2021) would suggest that most offenders will have continued contact with the criminal justice system. The author would speculate that the initial custody for a large majority [74%] of Lake City ATR offenders was not simply a turning point in their lives but a knifing off point according to Laub and Sampson [12].

The predictive power of the ASI-6 on reincarceration

Using the ASI-6 as a predictive model for reincarceration, ensuring there was no likelihood of multi-collinearity for each of the four logistic regression models was essential. Therefore, this study employed the removal of any case outliers with over a 2.0 residual tolerance and/or a Variance Inflation Factor (VIF) above 4. As such, 13, 13, 11, and 18 were removed from each of the four logistic multivariate models. Table 4 below illustrates the predictive power of seven domains on offender reincarceration after the first year of an offender’s release. This model was statistically significant (.001 with a confidence level of 95% with p < .05 being significantly different than zero. The seven domains of the ASI-6 within the model explained 28% of future offender reincarceration. This number is based on the average of two variances utilized for the analysis. The Nagelkerke R Square was reported at .340, and the Cox and Snell R Square was .224. The regression reported a Chi-square of 109.881 and a model-2 Log likelihood of 356.221.

Despite the prevalence of substance use as an identified risk by offenders, alcohol and drug use were the only two domains within the ASI-6 that were not found to be significant. It did not appear that the offender’s risk severity of alcohol or drug use impacted their immediate or short-term reincarceration. Employment (2.47) was identified as the most significant factor in reincarceration. Therefore, the higher an offender scored on the employment domain, the higher the likelihood of reincarceration within the first year of release from custody. Medical (2.15) and legal (2.14) domains were also found to be the second and third most significant domains that impacted future custody. A prior criminal record and prior incarceration were determinants of future custody, further explaining the churning and cyclical nature of offending Sherman [29]; Maxwell et al, 1998). Perhaps those with medical conditions found themselves with fewer options and would almost choose incarceration because custody provides the health care they may require. This has been cited in several studies (Travis et al, 2003). Finally, the remaining two domains of social stability and psychological assistance of the ASI-6 were prevalent in predicting reincarceration.

The second model in Table 5 suggests that the ASI-6 was a much better predictor of reincarceration within a three-year follow-up. Like the last, this model was found to be statistically significant (.001 with a confidence level of 95% with p < .05 being significantly different than zero. The variance explained within this model increased from the previous one-year follow-up. The first model predicted 28% of future offender reincarceration. This model predicts approximately 43% of reincarceration, a significant/ increase. This number is based on the average of two variances utilized for the analysis. The well-respected Nagelkerke R Square was reported at .496, and the Cox and Snell R Square was .357. The regression reported a Chi-square of 191.774 and a model- 2 Log likelihood of 361.181.

Table 5 reports that every domain, except drug use, was statistically significant. Offender’s self-reported employment and medical domains remained the best predictors of reincarceration over three years post-custody release. The legal, social, and psychological factors also remain statistically significant. The domain of alcohol use becomes a more important factor (and now statistically substantial) while the domain of drug use remains insignificant. Perhaps this may concern whether the offender has attained treatment or rehabilitation. Otherwise, this is a unique finding as substance use in year three is only a moderate factor in reincarceration.

The findings of the third model (Table 6) suggest that by the fifth year of follow-up, the ASI-6 has become an excellent tool for predicting long-term reincarceration. The model was statistically significant (.001 with a confidence level of 95% with p < .05 being significantly different than zero, and the variance explained in the model increases. The ASI-6 predicts 47% of the offenders will be reincarcerated. Although there was a slight increase from the three-year model, these findings indicate that the ASI-6 may be a better long-term predictive model than a short-term model. While it is certainly possible that the offender’s risks can change while completing programming and attaining assistance, it seems that an intake assessment tool such as the ASI is still a great predictor of future custody, not considering demographic variables (which certainly would have skewed the results). The Nagelkerke R Square was reported at .542, and the Cox and Snell R Square was .402. The regression reported a Chi-square of 223.086 and a model-2 Log likelihood of 382.919. All three of the logistic regression models had seven degrees of freedom. The severity of employment, psychological, medical, and substance use was still the most significant predictors of reincarceration.

The findings from Table 7 further substantiate the efficacy of the ASI-6 domains in predicting offender reincarceration seven years after release from custody. All domains are statistically significant, and it could be argued that these seven co-occurring issues offenders face are issues that persist upon their first year of release or earlier (before they were in custody). The model was statistically significant (.001 with a confidence level of 95% with p < .05 being significantly different than zero, and the variance explained in the model increased. The ASI-6 domains predicted 59% of the reincarceration of offenders. The Nagelkerke R Square was reported at .594, and the Cox and Snell R Square was .540. The regression reported a Chi-square of 251.516 and a model-2 Log likelihood of 302.704. This last model would suggest a shuffling of the domains in terms of the level of importance based on the reincarceration of offenders. Employment, medical care, and offender’s legal struggles become more acute. However, all domains remain significant while predicting nearly 60% of the reincarceration of offenders.

Concluding statements

While this author would argue it is essential to recognize the importance of taking a holistic approach when assisting offenders, it also appears that more objective measurements like ASI-6 should be used to validate those same concerns or at least determine whether perception is reality. Tables 1 and 2 highlight the differences between offender perception and risk-needs assessment objective scoring that recognizes the severity of the area of need. Sampson and Laub [31] provide a potential theoretical foundation for the continuity of events for offenders once they are released from custody. The last (or most recent) time in custody was a knifing-off or turning point in most Lake City ATR offenders’ lives, where they continue to struggle long-term in attaining a different pathway from crime. While it appears that wraparound programs like SAMHSA’s Lake City ATR had a measurable difference in the short-term (as compared to the statistics and a national comparison offered by the US Bureau

of Statistics), long-term offenders appear to be facing the same or more struggles to desist from crime and reincarceration. This could be due to the conceptualization or operationalization of process and outcome measurements to the limitation and scope of the Lake City ATR data Sherman [29]. Either way, the data and analysis validate the pathway and trajectory of churning and cycling through the criminal justice system. More research is, of course, needed to reduce the cumulative disadvantage and continuity of people’s lives after being incarcerated.

The primary focus of this study was to evaluate the efficacy of the Addiction Severity Index (ASI-6) in predicting the reincarceration of 434 offenders across multiple intervals of time from the first year of release through seven years of follow-up. The multivariate models (not considering other demographic information) substantiate the efficacy of the models in predicting as much as 59% of reincarceration within seven years of release. The interval levels of the four models would suggest that the ASI-6 is a good clinical tool to diagnose offender severity while also impacting successful outcomes (within the first few years of release).

Using data indicators such as the ASI could benefit probation/ parole officers in assisting offenders in meeting appropriate needs (versus their perception) and the risk-severity measures to be aware of Hannah-Moffit [34]. This, too, highlights the need for correctional services to be mindful of the needs and issues probation/ parole officers face to assist them with better and more appropriate resources and programming Dholakia [1]. Circling back to where we began (in the introduction), citizens expect more treatment and assistance to curb an offender’s propensity to reoffend Dholakia [1]. While more research is needed into how we assist offenders Petersilia [8], we should continue to follow these validated and reliable empirical clinical tools to reduce the barriers offenders face returning to their communities.

References

1. Dholakia N (2021) Fifty years ago today, President Nixon declared the war on drugs. Vera Institute of Justice.
2. Sawyer W, Wagner P (2020) Mass incarceration: The whole pie 2020. Washington, DC: The Prison Policy Initiative.
3. National Drug Intelligence Center (2011) The Economic Impact of Illicit Drug Use on American Society. Washington D.C.: U. S Department of Justice.
4. Sawyer W, Wagner P (2025) Mass incarceration: The whole pie 2025. Washington, DC: The Prison Policy Initiative.
5. Center on Addiction and Substance Abuse (2010) Behind Bars II: Substance Abuse and America's Prison Population. New York, NY: Center on Addiction and Substance Abuse.
6. Tonry M, Wilson J (1990) Drugs and crime, in Chapter 13, Crime and justice: A review of research. Chicago, IL: University of Chicago Press.
7. Goldstein P (1985) The drugs violence nexus: A tripartite conceptual framework. Journal of Drug Issues 15: 493-506.
8. Petersilia J (2004) What works in prisoner reentry? Reviewing and questioning the evidence. Federal Probation 68(2): 4-8.
9. Sabol W (2007) Prison and jail inmates at mid-year 2006. US Department of Justice. Washington, DC: Bureau of Justice Statistics.
10. Wilson J (2000) Crime: Public policies for crime control, 2nd ed. Oakland, CA: ICS Press.
11. Walker S (2001) Sense and nonsense about crime and drugs: a policy guide. 5^th Wadsworth.
12. Laub J Sampson R (1993) Turning points in the life course: Why change matters to the study of crime, Criminology 31(3): 301-325.
13. Antenangeli L, DuRose R (2021) Special report: Recidivism of prisoners released in 24 States in 2018: A 10-year follow-up period (2008-2018). Washington, DC: US Bureau of Justice Statistics.
14. Ray J, Jones S (2023) Aging out of crime and personality development: A review of the research examining the impulsiveness of offending in middle and late adulthood. Psychological Research Behavioral Management 16: 1587-1596.
15. Hendriks V (1989) The Addiction Severity Index: Reliability and validity in a Dutch addict population. Journal of Substance Abuse Treatment 6(2): 133-141.
16. McLellan A (1992) The fifth edition of the Addiction Severity Index, Journal of Substance Abuse Treatment 9(3): 199-213.
17. Leonhard C (2000) The Addiction Severity Index: A field study of reliability and validity. Journal of Substance Abuse Treatment 18(2): 129-135.
18. Mäkelä K (2004) Studies of the reliability and validity of the Addiction Severity Index. Addiction 99(4): 398-410.
19. Kokkevi A Hartgers C (1995) Europe ASI: European adaptation of a multidimensional assessment instrument for drug and alcohol dependence. European Addiction Research 1(4): 208-210.
20. Cacciola J (1994) An examination of the diagnostic criteria for antisocial personality disorder in substance abusers. Journal of Nervous and Mental Disease 182(9): 517-523.
21. Cacciola J (2001) The relationship of psychiatric co-morbidity to treatment outcomes in methadone-maintained patients. Drug and Alcohol Dependence 61(3): 271-280.
22. Lesieur H Blume S (1991) Evaluation of patients treated for pathological gambling in a combined alcohol, substance abuse, and pathological gambling treatment unit using the Addiction Severity Index. British Journal of Addiction 86(8): 1017-1028.
23. Seredycz M (2010) Offender drug abuse and recidivism: An Access to Recovery (ATR) program. El Paso, TX: LFB Scholarly Publishing LLC.
24. McLellan T (2004) Commentaries on Mäkelä. Addiction 99(4): 411-413.
25. Rosen C (2000) Consistency of self-administered and interview-based Addiction Severity Index composite scores. Addiction 95(3): 419-425.
26. Butler R (2001) Initial validation of a computer-administered addiction severity index: ASI-MV. Psychology of Addictive Behaviors 15(1): 4-12.
27. Grissom G, Bragg A (1991) Addiction Severity Index: Experience in the field. International Journal of the Addictions 26(1): 55-64.
28. Substance Abuse and Mental Health Services Administration (2006) Access to Recovery implementation and toolkit - menus and recovery services technical assistance package. Washington, DC: Substance Abuse and Mental Health Services Administration.
29. Sherman L (1997) Preventing crime: What works, what doesn't, what's promising. College Park: University of Maryland Press.
30. Farabee D (1998) The effectiveness of coerced treatment for drug-abusing offenders. Federal Probation 62(1): 3-10.
31. Sampson R, Laub J (1990) Crime and deviance over the life course: The salience of adult social bonds. American Sociological Review 55(5): 609-627.
32. Bunting A (2018) Beyond the employment dichotomy: An examination of recidivism and days remaining in the community by post-release employment status. International Journal of Offender Therapy and Comparative Criminals 63: 712-733.
33. Visher C, Travis J (2011) Life on the outside: Returning home after incarceration. The Prison Journal 91: 102S-119S.
34. Hannah-Moffat K (2010) Negotiated risk: Actuarial assessment and discretion in probation. Canadian Journal of Law and Society 24(3).