Spirituality and Desistance From Substance Use Among Reentering Offenders

Abstract

Prior research has indicated an inverse relationship between religion and criminal behavior; however, few studies have specifically examined the effect of spirituality on the desistance process among a contemporary and diverse sample of reentering drug-involved offenders. A comprehensive understanding of how spirituality is related to desistance from substance use can lead to more effective and evidence-based preventive and rehabilitative interventions. Using data from a longitudinal study of 920 diverse offenders returning to the community after a period of incarceration, the current study examines three distinct forms of substance use (alcohol, marijuana, and cocaine) to gauge the effect that spirituality plays in the desistance process. The findings suggest a relatively high importance of spirituality in terms of preventing substance use during reentry, particularly concerning the use of both alcohol and cocaine.

Keywords

desistance spirituality substance use incarceration

Introduction

Researchers have long sought to examine and explain the factors related to the etiology of criminal behavior over a lifetime. Despite the varied explanations regarding the onset of crime, recent research has made a concerted effort to increase our empirical understanding of those factors associated with its cessation or desistance. While the majority of studies have focused on life events, such as marriage and employment to predict desistance, fewer studies have examined those intra-individual characteristics that have the potential to influence this process by creating shifts in an individual’s identity. More specifically, the role that religion and spirituality play in the cessation from criminal behavior and drug use has yet to be fully understood. Although rarely a central focus in prior life-course studies, preliminary research has supported the idea that religion and spirituality can initiate desistance and create continued behavioral change in regard to illegal activity, including substance use (Chu, 2007; Giordano, Longmore, Schroeder, & Seffrin, 2008; Maruna, 2001; Schroeder & Frana, 2009; Terry, 2003).

The theoretical framework used to study desistance has primarily relied on the work of Robert Sampson and John Laub (Laub & Sampson, 2003; Sampson & Laub, 1993), which identified an array of salient events (e.g., marriage or employment) that occur during the life course and can serve as potential turning points to move offenders away from crime and substance use. These structural conditions are thought to provide offenders with a “stake in conformity” and the informal social controls needed to promote and maintain a state of nonoffending. Despite Laub and Sampson (2003) not emphasizing the role of religion or spirituality in the desistance process among the men in their study, a small number of the Glueck respondents did in fact report having benefited from either religion or religiously involved self-help groups. As such, religion and spirituality can be considered a formal social institution that has the ability to increase one’s social capital and redirect individuals into a more conventional life. As individuals obtain strong bonds to religious institutions or through their own spirituality, the bond can serve as a turning point to deter them from returning to their prior criminal behavior, shorten their criminal careers, aid them in coping with the terms of their imprisonment, and spur on a change in their self-identity (see Koenig, 1995; Maruna, 2001; Maruna, Wilson, & Curran, 2006). The formal affiliation with a religious institution or an increase in spirituality through one’s own personal experiences could potentially facilitate the development of a new and more prosocial identity (e.g., from addict to nonaddict; Giordano et al., 2008; Johnson, 2009; Terry, 2003). The current study seeks to extend this research by investigating the process by which religion and spirituality influence desistance from continued substance use among a contemporary and diverse sample of reentering male and female offenders.

A problematic aspect of examining the influence religion has on desistance from drug use is that it is an extremely difficult concept to define (Clark, 1958). Before moving on, it may be helpful to identify the distinctions between the concepts of religion and spirituality. Religious participation and affiliation has been conceptualized as belonging to a religious organization that involves a system of worship or doctrine that is shared in a group setting (Fetzer Institute, 2003). More pertinent to the current study is the empirical conceptualization of spirituality. Spirituality is a multifaceted concept that is very personal and experiential and is “concerned with the transcendent, addressing ultimate questions about life’s meaning, with the assumption that there is more to life than what we see or fully understand” (Fetzer Institute, 2003, p. 2).¹ A systematic review of 40 studies by Johnson, Li, Larson, and McCullough (2000) pointed to the need to specify religion as a multidimensional construct that is more inclusive and theoretical in its measurement. The complex nature of religion or spirituality means that it cannot be represented using a single indicator. Many empirical studies examining the link between religion, spirituality, and crime have used a simple operational or unidimensional definition of religion such as regular attendance at religious services or perceived closeness to God (see, for example, Cochran & Akers, 1989; Johnson, Jang, Larson, & Li, 2001). The current study, however, uses a multifaceted approach in operationalizing spirituality, using a variety of pragmatic and theoretical factors to gauge the relative importance of religion and spirituality in a person’s life.

Spirituality, Drug Use, and Desistance

Empirical support for religion, religious involvement, and spirituality as a protective factor concerning criminal behavior and substance use has been mixed. Recent research suggests that a negative relationship between religion and criminal behavior exists (Albrecht, Chadwick, & Acorn, 1977; Evans et al., 1996; Johnson et al., 2001; Powell, 1997), although the effects have been found to be modest. Baier and Wright’s (2001) meta-analysis of 60 studies examining religion and crime suggested that religious behavior (e.g., church attendance, prayer, etc.) and spiritual beliefs exerted only a moderate deterrent effect on criminal behavior, although they found that religion did operate as a social bond. The authors did emphasize that religion both deters criminal behavior and promotes prosocial behavior through the threat of supernatural sanctions and rewards (Baier & Wright, 2001).

Religion and spirituality have similarly been found to have a deterrent effect on substance use (e.g., Brownfield & Sorenson, 1991; Hawkins, Jenson, Catalano, & Lishner, 1988; Jang & Johnson, 2001; Stewart & Bolland, 2002). Johnson, Tompkins, and Webb (2006) reviewed 151 studies that examined the relationship between religion and drug and alcohol use. The authors concluded that the majority of these studies demonstrated that a greater participation in religious activities was associated with a decreased likelihood to use or abuse drugs or alcohol, and that these findings held true regardless of the population under study (i.e., children, adolescents, and adult populations). Other studies point to spirituality and religious involvement as being a contributing factor for positive outcomes among those participating in drug treatment (Avants, Warburton, & Margolin, 2001; Barrett, Simpson, & Lehman, 1988; Galanter, Dermatis, Bunt, Williams, Trujillo & Steinke 2007), and as an important determinant in lower recidivism rates among boot camp graduates (Benda, Toombs, & Peacock, 2003) and released incarcerated offenders (Johnson, Larson, & Pitts, 1997; Young, Gartner, O’Connor, Larson, & Wright, 1995). While it has been difficult to differentiate between the influence of religious and spiritual components of treatment programs from their secular elements, empirical evidence has suggested that spirituality can facilitate the prevention and treatment of substance use disorders (Sterling et al., 2007). For instance, research has found religion to be more likely to predict lower levels of substance use and abuse (Kendler, Gardner, & Prescott, 1997; Koenig, George, Meador, Blazer, & Ford, 1994), with spirituality being a stronger indicator of posttreatment abstinence and desistance (Avants et al., 2001).

Recent research has specifically pointed to the subjective and often individual importance placed on spirituality in aiding the desistance process among addicts participating in various 12-step programs. Galanter and colleagues (2007) found that a sense of spiritual direction and growth was viewed among those in recovery from addiction as more important in their lives than either professional drug treatment or employment. Flynn, Joe, Broome, Simpson, and Brown (2003) found that patients who indicated religion or spirituality as a source of recovery support were roughly twice as likely as those did not to desist from heroin and cocaine at 5 years. Similarly, in a study of long-term abstinence from drug addiction, Best, Safeena, Day, Rajashree, and Loaring (2008) found that for both achieving and maintaining abstinence, spiritual factors and support groups associated with 12-step programs were mentioned much more frequently by respondents compared with formal addiction treatment. The 12-step program attempts to diagnose the nature of addiction in the individual as a spiritual disease or illness and provides a clear set of directions to aid in the recovery process (Marsh, 2011). The central belief is that addiction is a disease of the spirit, where the substance is used as a substitute for an individual’s connection with a “higher power.” Once an addict has lost control over one’s substance-using behaviors, it is believed that all efforts to control or desist from any substance-using behavior under his or her own volition are doomed to fail. The “spiritual awakening” that occurs from accepting and committing one’s life to a higher power fills this void in spirituality and seeks to restore the individual to a state of recovery (Lyons, Deane, & Kelly, 2010; McCoy, Hermos, Bokhour, & Frayne, 2004; Piderman, Schneekloth, Pankratz, Maloney, & Altchuler, 2007). Research has found that the experience of a spiritual awakening during participation in a 12-step program can increase posttreatment abstinence by four times (Kaskutas, Turk, Bond, & Weisner, 2003). Similarly, and of great importance for released drug-involved offenders, these programs can afford a positive and stable social network that can provide emotional support (e.g., Kendler et al., 1997; Maruna, 2001) and increased social control (see also Baier & Wright, 2001; Giordano et al., 2008; Hirschi & Stark, 1969; O’Connor, 2002; Pardini, Plante, Sherman, & Stump, 2000).

Despite the vast empirical support linking spirituality and religion to reduced levels of criminal behavior and substance use, there remain only a handful of studies that have examined the role that these factors play in the desistance process, with even fewer focusing on reentering criminal offenders. Using Waves V to VII of the National Youth Survey, Chu (2007) found that increased religious attendance was associated with desistance from marijuana use, while religious salience (perceived importance of religion) reduced the onset of drug use. Another study by Chu and Sung (2009) examined racial differences in desistance from substance abuse among drug treatment clients participating in the Drug Abuse Treatment Outcomes Study. Church attendance (frequency of past year attendance) was found to have a significant effect on Black clients’ desistance from drug use at the 1-year follow-up; however, no association was found for White clients. Similarly, a study by Ulmer, Desmond, Joon Jang, and Johnson (2010) utilized data from the National Longitudinal Study of Adolescent Health (Add Health) to examine the relationship between one’s religious involvement and patterns of marijuana use during an adolescent’s transition to adulthood. Using multiple indicators measuring the respondent’s religious involvement, the authors concluded that the main effect of religion was a protective factor that prevented adolescents from initiating into marijuana use. Adolescent religious involvement did not, however, significantly predict desistance from marijuana use.

Giordano and colleagues (2008) examined the effects of church attendance and spirituality (perceived closeness to God) on criminal desistance and found that neither religious involvement nor spirituality significantly predicted long-term desistance among a sample of previously institutionalized high-rate juvenile offenders who were followed over a 21-year period. Their qualitative narratives did, however, indicate that an offender’s spirituality and religious devotion were mentioned as a potential “hook” that initiated a sustained behavioral change away from crime. Finally, Schroeder and Frana (2009) used qualitative interviews to examine how men living in a halfway house use religion and spirituality as an emotion-coping mechanism in their efforts to desist from crime. Their analysis revealed that religion was used to cope with feelings of stress, anger, and anxiety, to provide a distraction from adverse life conditions, and to begin the process of transition towards a more conventional and focused life.

The Present Study

The current study builds on the existing research by examining the role that spirituality plays in the desistance process from a variety of illicit substances among a contemporary and diverse sample of reentering offenders. Specifically, we focus on whether measures of spirituality can significantly predict desistance above and beyond the traditional life-course explanations of desistance. Our study builds on prior efforts in a variety of ways. First, the studies of Chu (2007), Chu and Sung (2009), and Giordano and colleagues (2008) were all limited in terms of using single-item measures, whereas we use a multi-item construct to measure a respondent’s spirituality. Second, the different classifications of substance use make it difficult to compare results across studies. For instance, Ulmer and colleagues (2010) focus solely on the interplay between religious involvement and desistance from marijuana, while Chu (2007) focuses on desistance from both soft (marijuana) and hard (a construct of 11 different drugs) drugs. In addition, Chu and Sung use a measure asking if respondents have used any illicit drug in the past 12 months, not allowing for any specific interpretations regarding a particular drug or classification of drugs. The current study examines three distinct forms of substance use (alcohol, marijuana, and cocaine) to gauge the effect that spirituality plays in the desistance process.² Finally, prior research is limited in terms of assessing the relationship between spirituality and desistance among incarcerated offenders after release (see also Giordano et al., 2008; Schroeder & Frana, 2009). Given the fact that 730,000 offenders were released from state and federal prisons in 2009 (Sabol & West, 2010), and that faith-based programs and organizations are becoming more involved in the correctional treatment and reentry process (Johnson, 2008), it is vital that researchers, policymakers, and practitioners understand all facets that might aid offender reintegration and promote desistance among today’s diverse offender population.

Method

Sample and Data Collection

In 2001, the Urban Institute in Washington, DC, initiated Returning Home: Understanding the Challenges of Prisoner Reentry, a multistate, longitudinal study designed to explore the process of prisoner reentry, the challenges that returning prisoners face, and the pathways to successful reintegration. The data gathered through Returning Home provide a comprehensive look at individuals’ life circumstances prior to and during prison, and over the course of a year after the release from prison. From 2002 to 2005, the study targeted prisoners serving at least 1 year in state prison and returning to the city and county areas of Chicago, Cleveland, and Houston. Potential respondents were identified either through compulsory prerelease programs where prisoners were already convened (Illinois and Texas) or from lists of individuals who were within 60 days of release (Ohio). A member of the research team provided an overview of the study and details of informed consent to assembled groups of potential respondents. Individuals who agreed to participate (more than 80%) were given consent forms and asked to complete a self-administered survey.

The prerelease samples of prisoners in each state were generally representative of all prisoners being released to the study areas in the previous year in terms of race, sentence length, and time served. After each prisoner’s release, experienced interviewers conducted up to three personal interviews within 15 months, including interviews with those who were in jail or had been returned to prison. Individuals were paid US$25 for each interview in the community. The Returning Home data from Ohio and Texas³ combined include a total of 1,100 participants. To be included in the present study, at least one postrelease interview had to be completed due to the longitudinal nature of the analyses. After eliminating participants who did not complete at least one follow-up interview, the sample size for the present study became 920, which is an 86% retention rate overall.

Dependent Variable

In selecting the dependent variable, several forms of substance use included in the Returning Home questionnaires were considered. Because only former users could desist from substance use, the sample sizes for analyses of each type of substance user was determined by how many were using the respective substance prior to incarceration (as determined through self-report data on preincarceration substance use via the prerelease interview). Ultimately, only alcohol, marijuana, and cocaine included enough preincarceration users for subsequent analyses (i.e., these substances were used by 38%-85% of the full sample, compared with 10% or less for other substances).

To determine what typologies might exist between different types of desisters (e.g., nondesisters, quick desisters, etc.), trajectory modeling—also called group-based modeling of longitudinal data—was utilized. The TRAJ procedure was used via SAS 9.2 to perform a logit model to identify groups based on postrelease substance use. These models were based solely on whether the subject used the respective substance at each available time point. The model fit statistics are presented in Table 1. Used here are the Bayesian information criterion (BIC) scores from each model, with larger BIC scores indicating a better fit (Nagin, 2005). For alcohol and marijuana use, these models indicated that a two-group model best fits the data. Further inspection of the cases in each group indicates that the two groups represent (a) participants who did not report using the substance at all after release, and (b) participants who reported using the substance at least once after release. In other words, these groups represent desisters and nondesisters. For cocaine use, the model fit statistics actually indicated that the best fit was a one-group model. However, this is likely because of the small number of postrelease users (n = 65) and a larger sample size might at least recognize the distinction between users and nonusers. Therefore, the two-group model (desisters and nondesisters) will also be used for subsequent analyses of cocaine users.⁴ These groups are coded simply by whether participants used or did not use the associated substance after their release based on their actual reported behavior.

Table 1.

Trajectory Model Fit Estimates.

	Model fit (BIC)
Number of groups	Alcohol	Marijuana	Cocaine
1	−679.21	−323.16	−221.15
2	−648.05	−316.94	−226.46
3	−657.71	−325.58	−234.21

Note. BIC = Bayesian information criterion.

The study of desistance has been made difficult by conflicting definitional and operationalization issues across empirical studies (see also Kazemian, 2007). Desistance has been increasingly thought of as a gradual process towards a state of nonoffending instead of a static event (Bushway, Thornberry, & Krohn, 2003). However, given the difficulty of following offenders over long periods of the life course, desistance measures in most empirical studies focus on a discrete state of nonoffending as opposed to defining it as a gradual process involving reductions in the severity and frequency of offending. As such, these studies have defined desistance as individuals who have not committed (using self-report or official records) an offense for a certain time period that ranges from 1 to 11 years (Bushway, Piquero, Broidy, Cauffman, & Mazerolle, 2001; Warr, 1998). Given the data available for the current study, we use a discrete measure of desistance here.

The dependent variables for further analyses will be dichotomous indicators of group membership. For each substance, these variables will be recorded as nondesisters (0) and desisters (1). These data were obtained via self-report substance use questions (e.g., Did you use alcohol since the last interview?). Former users were most successful in desisting from cocaine (76%) and least successful in desisting from alcohol use (32%). Of course, this is likely because alcohol use is legal for individuals over age 21, and most of the sample (97%) meets this threshold. The only legal motive for desistance of alcohol use might be a condition of parole. The descriptive statistics for these and other variables are presented in Table 2.

Table 2.

Descriptive Statistics.

	M	SD	Minimum	Maximum
DVs
Alcohol desistance	32%	—	—	—
Marijuana desistance	67%	—	—	—
Cocaine desistance	76%	—	—	—
IVs
Work release	06%	—	—	—
Spirituality	.00	1.00	−3.19	1.01
Religious affiliation	42%	—	—	—
Family attachment	.00	1.00	−4.02	1.37
Peer deviance	.00	1.00	−1.89	2.61
Bad neighborhood	.00	1.00	−1.82	2.88
Employment	35%	—	—	—
Demographics/controls
Female	23%	—	—	—
Black	65%	—	—	—
White	22%	—	—	—
Hispanic	09%	—	—	—
Other race	04%	—	—	—
Texas	58%	—	—	—
Age	36.35	9.85	18.00	73.00
Missed interviews	.98	.74	.00	2.00

Note. DV = dependent variable; IV = independent variable.

Independent Variables

Variables used as predictors in the models can generally be understood in three categories: characteristics when leaving prison, characteristics after reentry, and demographic or control variables. A spirituality scale⁵ was constructed using several indicators from the prerelease survey. These items included the following: How often do you pray or meditate; how often do you read the Bible or other religious literature (not at all, less than once a month, once a month, a few times a month, once a week, a few times a week, daily); I find strength in my religion or spirituality; I feel guided by God in the midst of daily activities; my faith helps me know right from wrong; and my spiritual beliefs help define the goals I set for myself (strongly disagree, disagree, agree, strongly agree). The variables scaled together well with a reliability alpha of .82 and were combined using factor extraction that retained 4.13 eigenvalues of variance. In addition, for characteristics when leaving prison, one variable was used. A simple dichotomous indicator was used measuring whether the individual participated in a work-release program while in prison. Though this was not asked at baseline, the respondent was asked the question retrospectively at the first survey (e.g., if a participant missed the first scheduled follow-up, he or she would have been asked the question as part of the next survey).

The variables used for postrelease characteristics largely measure the social environment to which the respondent returned, including religious affiliation, family attachment, deviant peers, neighborhood characteristics, and employment status. The measures used for the present study are from the first available postrelease time point for each respondent. If, for example, the respondent completed the first follow-up, only that time point was used. If a respondent missed the first follow-up, but participated in the second, only that time point was used. This method allows for the earliest possible measurement for each case.

Religious affiliation was measured using a simple yes-or-no question that asked whether the participant belonged to a “church, synagogue, mosque, or other religious organization.” Employment was also measured with a simple yes-or-no question: “Are you currently employed?” Using Likert-like statements relating to family attachment and support, a factor analysis indicated that 20 statements loaded well together on a single construct. These items included the following questions: Do you feel close to your family; do you have someone in your family to talk to about yourself or your problems; and do you have someone in your family who would provide support for dealing with a substance abuse problem.⁶ These items scaled together very reliably (α = .97) and retained 12.69 eigenvalues when the factor was extracted. Given that these statements each respond to positive family characteristics, this factor is conceptualized as positive family attachment.

In terms of potentially negative influences, constructs for both peers and neighborhood were used. For peers, the following questions were used: How many of your friends have ever been in prison; how many of your friends do you think have ever committed a theft; how many of your friends do you think have ever assaulted someone; how many of your friends do you think have ever used illegal drugs; and how many of your friends do you think have ever sold illegal drugs (none, some, most, all). These items scaled together moderately well (α = .84) and retained 2.97 eigenvalues after factor extraction. For the neighborhood indicator, the following statements were used: It is hard to stay out of trouble in your neighborhood; you are nervous about seeing certain people in your neighborhood; drug selling is a major problem in your neighborhood; living in this neighborhood makes it hard for you to stay out of prison. These items scaled together moderately well (α = .78) and retained 2.39 eigenvalues after factor extraction.

For demographic indicators and other controls, several dichotomous indicators were used. These included female, Black, Hispanic, other race (non-Hispanic White being the reference category for all race/ethnicity variables), and Texas (Ohio as the reference category). In addition, age, as measured in years at the time of prerelease interview, was also used. Finally, as a methodological control, a count for number of missed follow-ups was used. Because of the nature of the dependent variables, someone who participated in more follow-ups had more chances to report using a substance. It was unfeasible due to the limited sample size to only include those who participated in all follow-ups, especially because Texas participants were only asked to participate in two (and therefore had a minimum number of one missed follow-up). Thus, this control was used to model missed opportunities to report substance use. Although interpreting its direct effects was substantively meaningless, incorporating it as a control was important for those reasons.

Analytical Strategy

Due to the dichotomous nature of the dependent variables, logistic regression was used for the analyses. The sample sizes varied by the substance being modeled, as only those participants using the substance prior to prison were theoretically able to desist. The sample sizes were 491 for alcohol, 307 for marijuana, and 270 for cocaine. The samples were further reduced through listwise deletion due to missing responses in some cases. Because most cases did not have any missing data, the majority of cases (97%−98% depending on the substance) were retained for analyses. Due to the relatively small sample sizes, the analyses reported significance using both traditional notations (i.e., .05 and .01) and a marginal significance level (.10) to provide information on relationships that border on true significance. Rather than suggesting that this is evidence of significance, it should be considered more of a note that additional research could provide such evidence.

Desistance from each substance was modeled in two phases. First, a basic model was presented that included only spirituality and religious affiliation.⁷ This indicated whether there was any statistical relationship between these concepts and desistance from substance use. Second, the full model was analyzed incorporating all theoretical and control variables, which indicated whether religious indicators remained significant after controlling for other factors.

Results

The results from the models predicting desistance from alcohol use are presented in Table 3. First, Model 1a presents the abridged model. Higher levels of spirituality seem to indicate a greater chance of desisting after release. Specifically, for each standard deviation difference in spirituality, the odds of desisting increased by a factor of 1.36, which is a significant effect. Religious affiliation after release, however, appeared to have no discernable effect on alcohol desistance.

Table 3.

Logistic Regression Results Predicting Desistance From Alcohol (n = 481).

	Model 1a			Model 1b
	b	SE	OR	b	SE	OR
Spirituality	.31	.12	1.36**	.27	.13	1.31**
Religious affiliation	−.04	.22	.96	−.09	.23	.91
Work release	—	—	—	−.37	.45	.69
Family attachment	—	—	—	−.01	.11	.99
Deviant peers	—	—	—	−.08	.11	.92
Bad neighborhood	—	—	—	−.23	.12	.80**
Employment	—	—	—	−.02	.23	.98
Female	—	—	—	.04	.27	1.04
Black	—	—	—	.23	.28	1.26
Hispanic	—	—	—	−.06	.40	.94
Other race	—	—	—	.01	.62	1.01
Age	—	—	—	.00	.01	1.00
Texas	.24	.25	1.27	.21	.28	1.23
Missed follow-ups	.56	.16	1.74***	.59	.17	1.80***
Intercept	−1.44	.21	.24	−1.74	.54	0.18
Nagelkerke R²	.09			.11

Note. OR = odds ratio.

p < .10. **p < .05. ***p < .01.

The results from the full model are presented as Model 1b.⁸ Notably, spirituality retained much of its effect despite the additional controls in the model, with a change in the odds by a factor of 1.31 for each standard deviation difference in spirituality. The only other significant effect in the model is the effect of being in a bad neighborhood. Specifically, for each standard deviation increase in the negative characteristics of the neighborhood, the odds of desisting from alcohol use declined by a factor of .80. No other variables in the model are significant, with the exception of the methodological control for the number of missed follow-ups.

The next set of models is presented in Table 4. First, Model 2a presents the abridged model results, in which no predictors are significant using the traditional .05 alpha level. However, religious affiliation is significant under the less stringent .10 alpha level. Given the limited sample sizes, it can be reasonably understood that such a relationship might be significant under traditional standards if a larger sample were available. This is, of course, just speculation, so caution is advised for variables significant under this standard. This effect suggests that those affiliated with a church or other religious group were more likely to desist marijuana use after release and had their odds increased by a factor of 1.60.

Table 4.

Logistic Regression Results Predicting Desistance From Marijuana (n = 300).

	Model 2a			Model 2b
	b	SE	OR	b	SE	OR
Spirituality	.16	.14	1.17	.29	.16	1.34*
Religious affiliation	.47	.27	1.60*	.39	.30	1.48
Work release	—	—	—	−.11	.55	.90
Family attachment	—	—	—	.24	.14	1.27*
Deviant peers	—	—	—	−.13	.15	.88
Bad neighborhood	—	—	—	−.22	.14	.80
Employment	—	—	—	−.28	.31	.76
Female	—	—	—	−1.17	.37	.31****
Black	—	—	—	−.63	.35	.53*
Hispanic	—	—	—	.07	.61	1.07
Other race	—	—	—	1.29	1.12	3.65
Age	—	—	—	.04	.01	1.04***
Texas	−.05	.34	.95	.33	.41	1.38
Missed follow-ups	.09	.22	1.09	.23	.24	1.26
Intercept	.48	.23	1.61	−.28	.66	.75
Nagelkerke R²	.03			.18

Note. OR = odds ratio.

p < .10. **p < .05. ***p < .01.

The results for the full model are presented as Model 2b.⁹ Among the theoretical variables, there were again no predictors significant with a .05 alpha level. However, both spirituality and family attachment were significant under looser standards,¹⁰ with higher levels of spirituality and higher levels of family attachment corresponding to greater chances of desisting from marijuana. In contrast to the abridged model, religious affiliation was weakened to nonsignificance after adding the other variables into the model. Several demographic predictors were also significant, with female and Black participants less likely to desist and older participants more likely to succeed in desisting from marijuana use.

The final sets of models, those predicting cocaine use, are presented in Table 5. In the abridged model, spirituality is a significant predictor of cocaine desistance. Specifically, each standard deviation increase in spirituality equated to a 1.46 factor increase in the odds of desisting. Religious affiliation did not have a significant effect.

Table 5.

Logistic Regression Results Predicting Desistance From Cocaine (n = 267).

	Model 3a			Model 3b
	b	SE	OR	b	SE	OR
Spirituality	.38	.15	1.46**	.39	.17	1.47**
Religious affiliation	.45	.32	1.57	.46	.34	1.58
Work release	—	—	—	−.62	.52	.54
Family attachment	—	—	—	.15	.16	1.16
Deviant peers	—	—	—	−.05	.17	.95
Bad neighborhood	—	—	—	−.31	.17	.74*
Employment	—	—	—	.33	.36	1.39
Female	—	—	—	−.28	.44	.75
Black	—	—	—	−.01	.42	.99
Hispanic	—	—	—	.14	.58	1.15
Other race	—	—	—	.97	1.15	2.63
Age	—	—	—	−.01	.02	.99
Texas	−.58	.42	.56	−.64	.47	.53
Missed follow-ups	.36	.28	1.44	.42	.30	1.53
Intercept	1.04	.29	2.84	1.31	.89	3.70
Nagelkerke R²	.07			.13

Note. OR = odds ratio.

p < .10. **p < .05. ***p < .01.

The full model is presented as Model 3b, and indicates that spirituality retained its effect despite the additional controls in the model. Of the other variables in the model, only the neighborhood environment variable had any significant effect. Even it, however, did not reach the traditional significance level. Nevertheless, these findings suggest that individuals reentering into a bad neighborhood had lower odds of desisting cocaine use by a factor of .74 for each standard deviation change in the neighborhood.

Discussion

This study builds on the existing literature by examining the effects of spirituality on desistance from alcohol, marijuana, and cocaine among a contemporary and diverse sample of reentering offenders. The present findings suggest the relatively high importance of spirituality in terms of preventing substance use during reentry. For both alcohol and cocaine, spirituality was a significant predictor of desistance. And although the neighborhood environment was also a factor in desistance, no other theoretical variables were significant in either model, further highlighting the theoretical and practical importance of spirituality. Despite this consistent effect, religious affiliation was not significant in any of the alcohol or cocaine models, suggesting an important difference between simply belonging to a religious group, and the influence of one’s spirituality in refraining from drug use.

The findings from the marijuana model differ substantially from the other models. In addition to demographic variables being significant in those models, which was not observed for the other substances, the relationship between spirituality and marijuana use was mixed and statistically weaker than that in the models for alcohol and cocaine. Moreover, religious affiliation showed some significance in the abridged model, but the effect weakened to nonsignificance after adding controls. Overall, while the findings for alcohol and cocaine are somewhat compatible, marijuana seems to suggest a different process in the factors that affect desistance.

Religion and spirituality have the potential to act as a mechanism of external social control over the individual, as well as a facilitator in the development of a new prosocial identity (Maruna, 2001; Terry, 2003). While simply having a religious affiliation had no significant effect on desistance from substance use for reentering offenders in this study, this was not the case for spirituality. Individual spirituality is unique in that it can provide reentering offenders with both psychological and emotional comfort during a time that can be both stressful and chaotic. Becoming spiritually centered can provide a sense of clarity for these offenders in actively choosing to forgive their prior transgressions, to hope for a better future through the use of religious guidance, and to fill the void left by substance use with a higher power. Similarly, a strong sense of spirituality can serve as a guide through and a distraction from the tumultuous life circumstances that await many offenders on release, such as dealing with issues relating to substance use, unemployment, reconnecting with family and peers, and finding adequate housing. Finally, these spiritual transformations may mark an important turning point in the lives of these men and women, facilitating a shift in one’s identify from an offender to an ex-offender, which can serve as a catalyst in the desistance process. Given the many difficulties awaiting these men and women when returning home, it is well understood that many offenders will succumb to the immense pressures that come with a conventional lifestyle, making desistance extremely challenging. Yet, despite these difficulties, spirituality can at the very least be the foundation from which they start to rebuild their lives, creating the potential to stimulate long-term behavioral change away from crime and substance use.

In addition, it has been noted that the present-day landscape of desistance is likely influenced by aggregate declines in marriage, a decreased stability and security within social relationships, lack of legitimate and sustainable opportunities for employment, and the marginalizing effects of heavy drug use (see also Booth, Crouter, & Shanahan, 1999; Giordano, Cernkovich, & Rudolph, 2002; Giordano et al., 2008; Maruna, 2001). As the relative importance of these traditionally stable areas of employment and social attachments continue to weaken in terms of predicting desistance, some offenders may be progressively influenced by rehabilitative programming that stresses positive identity transformation and change that can be accepted and reinforced by others familiar with their situation (e.g., other addicts, counselors, or peers and/or relatives). In What Works in Corrections, MacKenzie (2006) concluded that programs focused on individual change had greater effects than those focused on practical improvements, such as employment training. If individual spirituality aids in an identity transformation, policymakers should not minimize the role of religion and spirituality when considering potential rehabilitative and treatment programs and initiatives. The continued use and effectiveness of faith-based initiatives in drug treatment and reentry programs should be considered, as these programs may provide both external social control and internal spiritual guidance that may initiate a transformative change in an offender’s identity from an addict to an ex-addict.

Like any research, this study is not without limitations. While this study focused primarily on the spirituality of reentering offenders, other measures of religion and religious involvement were limited in the data. Research has suggested that religion may affect drug use differently based on religious denomination (Bahr, Hawks, & Wang, 1993) and church attendance (Chu & Sung, 2009); however, these measures were unavailable in our data. Future research should attempt to measure the effect of complex measures of both religion and spirituality on the desistance process. This may clarify the importance of religion as a means of social control or as an agent of identity change. In addition, this research was limited by a short follow-up of reentering offenders. Studying the relationship of religion and desistance among offenders over an extended follow-up period would be desirable (see also Giordano et al., 2008), would allow desistance to be viewed as a “process” instead of a single event, and may clarify the relationships between criminal behavior, religion, and the well-established correlates of desistance (i.e., employment and marriage). Another important limitation to note is that these data are unable to account for causality and self-selection bias. In short, those who are more spiritual may be biased towards desisting for reasons other than spirituality itself. Thus, though the present study supports spirituality as a factor in desistance, these results cannot prove such causality. Future research would also benefit from studying the independent variables longitudinally, which was beyond the scope of the present study.

This study provides a useful extension to the literature examining the relationship between spirituality and substance use. Religious involvement and spirituality have long been recognized as a significant source of social capital among drug-involved offenders, providing enhanced physical and mental well-being and increased social control (Baier & Wright, 2001; Ellison, 1991). Released offenders face multiple challenges when they return home to their families and communities. While maintaining meaningful employment and healthy social attachments are recognized as the primary means of promoting desistance, a strong sense of spirituality seems to further emphasize the promotion of a meaningful identity change that will only encourage successful offender reintegration. The findings from this study provide insight into the utility of incorporating spirituality and religion-based programs into or as an alternative to traditional substance abuse treatment for incarcerated and reentering offenders. As current treatment approaches continue to focus on traditional external structural factors to encourage desistance, those factors that promote internal motivation and change should not be ignored.

Footnotes

Acknowledgements

This paper stems from a larger research effort at The Urban Institute’s Justice Policy Center under the direction of Dr. Christy Visher and Dr. Nancy La Vigne. The authors would like to thank the individuals who shared their experiences, opinions, and expectations with us. Without their willingness to share the details of their lives, this study would not have been possible.

Authors’ Note

This paper stems from a larger research effort at the Urban Institute’s Justice Policy Center. The data used in this analysis were collected by Research Support Services directed by Alisu Schoua-Glusberg (Ohio) and NuStats directed by Robert Santos (Texas).

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded through the generous support of the John D. and Catherine T. MacArthur Foundation, Annie E. Casey Foundation, JEHT Foundation, Cleveland Foundation, Houston Endowment, Robert Wood Johnson Foundation, Rockefeller Foundation, George Gund Foundation, Smith Richardson Foundation, Ohio Department of Rehabilitation and Correction, Ohio Office of Criminal Justice Services, and the Urban Institute.

Notes

References

Albrecht

S. I.

Chadwick

B. A.

Acorn

D. S.

(1977). Religiosity and deviance: Of an attitude-behavior contingent consistency model. Journal for the Scientific Study of Religion, 16, 263-274.

Avants

S. K.

Warburton

L. A.

Margolin

(2001). Spiritual and religious support in recovery from addiction among HIV-positive injection drug users. Journal of Psychoactive Drugs, 33, 39-45.

Bahr

S. J.

Hawks

R. D.

Wang

(1993). Family and religious influences on adolescent substance abuse. Youth & Society, 24, 443-465.

Baier

C. J.

Wright

(2001). If you love me, keep my commandments: A meta-analysis of the effect of religion on crime. Journal of Research in Crime & Delinquency, 38, 3-21.

Barrett

M. E.

Simpson

Lehman

W. E.

(1988). Behavioral changes of adolescents in drug abuse intervention programs. Journal of Clinical Psychology, 44, 461-473.

Benda

B. B.

Toombs

N. J.

Peacock

(2003). Discriminators of types of recidivism among boot camp graduates in a five-year follow-up study. Journal of Criminal Justice, 31, 539-551.

Best

D. W.

Safeena

Day

Rajashree

Loaring

(2008). Breaking the habit: A retrospective analysis of desistance factors among formerly problematic heroin users. Drug and Alcohol Review, 27, 619-624.

Booth

Crouter

A. C.

Shanahan

M. J.

(Eds.). (1999). Transitions to adulthood in a changing economy: No work, no family, no future? Westport, CT: Praeger.

Brownfield

Sorenson

A. M.

(1991). Religion and drug use among adolescents: A social support conceptualization and interpretation. Deviant Behavior, 12, 259-276.

10.

Bushway

S. D.

Piquero

A. R.

Broidy

L. M.

Cauffman

Mazerolle

(2001). Anempirical framework for studying desistance as a process. Criminology, 39, 491-515.

11.

Bushway

S. D.

Thornberry

T. P.

Krohn

M. D.

(2003). Desistance as a developmental process: A comparison of static and dynamic approaches. Journal of Quantitative Criminology, 19, 129-153.

12.

Chu

D. C.

(2007). Religiosity and desistance from drug use. Criminal Justice and Behavior, 34, 661-679.

13.

Chu

D. C.

Sung

(2009). Racial differences in desistance from substance use: The impact of religious involvement on recovery. International Journal of Offender Therapy and Comparative Criminology, 53, 696-716.

14.

Clark

W. H.

(1958). The psychology of religion: An introduction to religious experience and behavior. New York, NY: MacMillan.

15.

Cochran

J. K.

Akers

R. L.

(1989). Beyond hellfire: An explanation of the variable effects of religiosity on adolescent marijuana and alcohol use. Journal of Research in Crime & Delinquency, 26, 198-225.

16.

Ellison

C. G.

(1991). Religious involvement and subjective well-being. Journal of Health and Social Behavior, 32, 80-99.

17.

Evans

T. D.

Cullen

F. T.

Burton

V. S.

Jr. Dunaway

R. G.

Payne

G. L.

Kethineni

S. R.

(1996). Religion, social bonds, and delinquency. Deviant Behavior, 17, 43-70.

18.

Fetzer Institute and National Institute on Aging Working Group. (2003). Multidimensional measurement of religiousness/spirituality for use in health research. Kalamazoo, MI: Fetzer Institute.

19.

Flynn

P. M.

Joe

G. W.

Broome

K. M.

Simpson

D. D.

Brown

B. S.

(2003). Recovery from opioid addiction in DATOS. Journal of Substance Abuse Treatment, 25, 177-186.

20.

Galanter

Dermatis

Bunt

Williams

Trujillo

Steinke

(2007). Assessment of spirituality and its relevance to addiction treatment. Journal of Substance Abuse Treatment, 33, 257-264.

21.

Giordano

P. C.

Cernkovich

S. A.

Rudolph

J. L.

(2002). Gender, crime, and desistance: Toward a theory of cognitive transformation. American Journal of Sociology, 104, 990-1064.

22.

Giordano

P. C.

Longmore

M. A.

Schroeder

R. D.

Seffrin

P. M.

(2008). A life course perspective on spirituality and desistance from crime. Criminology, 46, 99-131.

23.

Goldfarb

Galanter

McDowell

Lifshutz

Dermatis

(1996). Medical student and patient attitudes toward religion and spirituality in the recovery process. American Journal of Drug and Alcohol Abuse, 22, 549-561.

24.

Hawkins

J. D.

Jenson

J. M.

Catalano

R. F.

Lishner

D. M.

(1988). Delinquency and drug abuse: Implications for social services. Social Service Review, 62, 258-284.

25.

Hirschi

Stark

(1969). Hellfire and delinquency. Social Problem, 17, 202-213.

26.

Jang

S. J.

Johnson

B. R.

(2001). Neighborhood disorder, individual religiosity, and adolescent use of illicit drugs: A test of multilevel hypotheses. Criminology, 39, 109-143.

27.

Johnson

B. R.

(2008). The faith factor and prisoner reentry. Interdisciplinary Journal of Research on Religion, 4, 1-21.

28.

Johnson

B. R.

(2009). The role of religious institutions in responding to crime and delinquency. In Clarke

P. B.

(Ed.), The Oxford handbook of the sociology of religion (pp. 857–875). New York, NY: University of Oxford Press.

29.

Johnson

B. R.

Jang

S. J.

Larson

D. B.

(2001). Does adolescent religious commitment matter? A reexamination of the effects of religiosity on delinquency. Journal of Research in Crime & Delinquency, 38, 22-44.

30.

Johnson

B. R.

Larson

D. B.

Pitts

T. C.

(1997). Religious programs, institutional adjustment, and recidivism among former inmates in prison fellowship programs. Justice Quarterly, 14, 501-521.

31.

Johnson

B. R.

Larson

D. B.

McCullough

(2000). A systematic review of the religiosity and delinquency literature. Journal of Contemporary Criminal Justice, 16, 32-52.

32.

Johnson

Tompkins

R. B.

Webb

(2006). Objective hope assessing the effectiveness of faith-based organizations: A review of the literature (ISR Report). Waco, TX: Institute for Studies of Religion, Baylor University.

33.

Kaskutas

L. A.

Turk

Bond

Weisner

(2003). The role of religion, spirituality and alcoholics anonymous in sustained sobriety. Alcoholism Treatment Quarterly, 21, 1-6.

34.

Kazemian

(2007). Desistance from crime: Theoretical, empirical, methodological, and policy considerations. Journal of Contemporary Criminal Justice, 23, 5-27.

35.

Kendler

K. S.

Gardner

C. O.

Prescott

C. A.

(1997). Religion, psychopathology, and substance use and abuse: A multimeasure, generic-epidemiologic study. American Journal of Psychiatry, 154, 322-329.

36.

Koenig

H. G.

(1995). Religion and older men in prison. International Journal of Geriatric Psychology, 10, 219-230.

37.

Koenig

H. G.

George

L. K.

Meador

K. G.

Blazer

D. G.

Ford

S. M.

(1994). Religious practices and alcoholism in a southern adult population. Hospital & Community Psychiatry, 45, 225-231.

38.

Laub

J. H.

Sampson

R. J.

(2003). Shared beginnings, divergent lives: Delinquent boys to age 70. Cambridge, MA: Harvard University Press.

39.

Lyons

Deane

F. P.

Kelly

P. J.

(2010). Forgiveness and purpose in life as spiritual mechanisms of recovery from substance use disorders. Addiction Research & Theory, 18, 528-543.

40.

MacKenzie

(2006). What works in corrections: Reducing the criminal activities of offenders and delinquents. New York, NY: Cambridge University Press.

41.

Marsh

(2011). Narrating desistance: Identity change and the 12-step script. Irish Probation Journal, 8, 49-68.

42.

Maruna

(2001). Making good: How ex-offenders reform and reclaim their lives. Washington, DC: American Psychological Association.

43.

Maruna

Wilson

Curran

(2006). Why God is often found behind bars: Prison conversions and the crisis of self-narrative. Research in Human Development, 3, 161-184.

44.

McCoy

L. K.

Hermos

J. A.

Bokhour

B. G.

Frayne

S. M.

(2004). Conceptual bases of Christian, faith based substance abuse rehabilitation programs: Qualitative analysis of staff interviews. Substance Abuse, 25, 1-11.

45.

McDowell

Galanter

Goldfarb

Lifshutz

(1996). Spirituality and the treatment of the dually diagnosed: An investigation of patient and staff attitudes. Journal of Addictive Diseases, 15, 55-68.

46.

Nagin

D. S.

(2005). Group-based modeling of development. Cambridge, MA: Harvard University Press.

47.

O’Connor

T. P.

(2002). Introduction: Religion-offenders-rehabilitation: Questioning the relationship. Journal of Offender Rehabilitation, 35, 1-9.

48.

Pardini

D. A.

Plante

T. G.

Sherman

Stump

J. E.

(2000). Religious faith and spirituality in substance abuse recovery: Determining the mental health benefits. Journal of Substance Abuse Treatment, 19, 347-354.

49.

Piderman

K. M.

Schneekloth

T. D.

Pankratz

V. S.

Maloney

S. D.

Altchuler

S. I.

(2007). Spirituality in alcoholics during treatment. American Journal on Addictions, 16, 232-237.

50.

Powell

(1997). Correlates of violent and nonviolent behavior among vulnerable inner-city youths. Family & Community Health, 20, 38-47.

51.

Sabol

W. J.

West

(2010). Prisoners in 2009. Washington, DC: Bureau of Justice Statistics, U.S. Department of Justice.

52.

Sampson

R. J.

Laub

J. H.

(1993). Crime in the making: Pathways and turning points through life. Cambridge, MA: Harvard University Press.

53.

Schroeder

R. D.

Frana

J. F.

(2009). Spirituality and religion, emotional coping, and criminal desistance: A qualitative study of men undergoing change. Sociological Spectrum, 29, 718-741.

54.

Sterling

R. C.

Weinstein

Losardo

Raively

Hill

Petrone

Gottheil

(2007). A retrospective case control study of alcohol relapse and spiritual growth. American Journal on Addictions, 16, 56-61.

55.

Stewart

Bolland

J. M.

(2002). Parental style as possible mediator of the relationship between religiosity and substance use in African-American adolescents. Journal of Ethnicity in Substance Abuse, 1, 63-80.

56.

Terry

C. M.

(2003). The fellas: Overcoming prison and addiction. Belmont, CA: Wadsworth.

57.

Ulmer

J. T.

Desmond

S. A.

Joon Jang

Johnson

B. R.

(2010). Teenage religiosity and changes in marijuana use during the transition to adulthood. Interdisciplinary Journal of Research on Religion, 6, 1-19.

58.

Warr

(1998). Life-course transitions and desistance from crime. Criminology, 36, 183-216.

59.

Young

M. C.

Gartner

O’Connor

T. P.

Larson

D. B.

Wright

(1995). Long-term recidivism among federal inmates trained as volunteer prison ministers. Journal of Offender Rehabilitation, 22, 97-118.