Can Yoga Overcome Criminality? The Impact of Yoga on Recidivism in Israeli Prisons

Abstract

In recent years, yoga practices have been integrated into formal prison rehabilitation programs of the Israel Prison Service (IPS), as part of the informal education system, giving rise to innovative criminological theories such as positive criminology that emphasize the development of offenders’ strengths by facilitating rehabilitation and reintegration processes. The purpose of the present study was to examine the correlation between yoga practice and recidivism among released prisoners who participated in yoga programs during their incarceration in comparison with a matched control group of those who did not participate in yoga programs over a follow-up period of 5 years. To examine the effectiveness of the program, propensity score matching was used to compile the comparison group from among all convicted prisoners who were released from the Israeli prisons. Study results indicate that yoga may affect recidivism, supported by a finding of lower recidivism rates among released prisoners who had practiced yoga during their incarceration, compared with a matched control group. However, further study is needed including randomized controlled trials (RCTs). In light of these positive results, we recommend policy-makers consider expanding alternative practices such as yoga into prisons, in recognition of their contribution to the rehabilitation process through the development of personal and social strengths.

Keywords

yoga prison prisoners rehabilitation positive criminology

Introduction

Returning to incarceration is an enduring problem in many countries. For example, in the United States, 45% of released prisoners between the years 2005 and 2010 were reincarcerated within 5 years after their release (Fazel & Wolf, 2015; Pew Center on the States, 2011). Similarly, data of the Israel Prison Service (IPS) reveal that 41.3% of prisoners in the country return to prison within 5 years of their release (Berman & Walk, 2015). These high recidivism rates indicate the existence of a recurring problematic pattern that entails high social and economic costs (MacKenzie, 2006; Zara & Farrington, 2015).

As part of an attempt to address this recidivism issue, together with the growing emphasis on prisoner rights, prisons in the Western world offer a variety of rehabilitation programs aimed at assisting people who are incarcerated and ex-prisoners with recovery and reintegration into society through the acquisition of certain life skills (Andrews et al., 1990, 2011; Enosh et al., 2013; Lipsey & Cullen, 2007; MacKenzie, 2006; McGuire, 2003; Shoham et al., 2017, 2018). The objectives of such programs include, inter alia, changing prisoners’ state of mind, reducing risk factors in their lives, and developing prosocial skills, which, in turn, are designed to reduce and stop antisocial behaviors (Ross, 1980; Ross & Hilborn, 2008).

In addition, in recent years, the criminological discourse has brought into focus various rehabilitative theories and approaches from positive perspectives such as desistance from crime (Maruna, 2001; Sampson & Laub, 1993), reintegrative shaming, and restorative justice (Braithwaite, 2001). These approaches were recently tied together under the definition of “positive criminology” (Ronel & Elisha, 2011), placing the emphasis on the personal and social strengths, increasing the degree to which individuals can embrace the prosocial lifestyle and avoid criminality.

Yoga practice, which can be viewed as one aspect in the above positive criminology approach, is now offered in some prisons and considered as an informal rehabilitative alternative or addition (Karup & Hall, 2016; Muirhead & Fortune, 2016). It should be emphasized, however, that there are many different types of yoga practiced, meaning that the practice is not always consistent. Yet, very few controlled studies have been conducted to examine the impact of yoga practicing in prison and the reduction of recidivism, some of whom suffer methodological flows, as detailed bellow.

The present study aims to fill this void and assess the effectiveness of yoga programs using a quasi-experimental design and a wide sample, by taking into consideration a wide range of variables to better understand the influence of yoga programs on recidivism, and by extending the monitoring period for released prisoners to 5 years, which allows for the comparison with other studies in the field. The purpose of the present study was to examine the correlation between yoga practice and recidivism among prisoners released from the IPS who participated in yoga during incarceration as compared with a matched control group of non–yoga-participating released prisoners over a follow-up period of 5 years after discharge from prison. The hypothesis of the present study is that practicing yoga has positive ongoing effects on prisoners’ lives, which assist in their rehabilitation program and thereby reduce recidivism rates.

Rehabilitation Programs From the Positive Criminology Perspective

Positive criminology focuses on enhancing offenders’ personal and social strengths and the factors that promote successful rehabilitation and reintegration into the community (Ronel & Elisha, 2011), underscoring the importance of holistic rehabilitation in addressing the various dimensions of life. Holistic rehabilitation deals with improving and developing offenders’ physical, emotional, and mental health as well as their interpersonal relationships in an effort to assist them with refraining from engaging in further criminal activity (Ronel et al., 2013). Positive criminology also emphasizes the positive effects of informal rehabilitative programs that underscore the development of personal and interpersonal strengths such as sports and yoga (Barrett, 2017; Guse & Hudson, 2014; Kewley, 2017; Ronel & Segev, 2014; Ward et al., 2007).

Yoga can be viewed as a positive informal practice combining several aspects such as postures, breathing techniques, relaxation techniques, and meditation (Bilderbeck et al., 2013; Chiesa & Serretti, 2009) that aim to provide individuals with the opportunity to connect their bodies and souls, reflect upon their lives, and be introspective (Norman, 2015). Yoga is viewed as a positive tool for personal and human growth as well as a holistic path that facilitates practitioners’ self-awareness and understanding (Karup & Hall, 2016).

Yoga From the Positive Criminology Perspective

Interest in the practice of yoga has developed rapidly in recent years. As of 2015, 13.2% of the U.S. population engages in yoga (Cramer et al., 2016), a substantial increase when compared with the only 6% reported to be practicing yoga in 2007 (Barnes et al., 2008). These findings attest to yoga’s growing popularity and the rise in overall awareness of this practice in the Western world as well as in the literature (Bilderbeck et al., 2013; Norman, 2015).

Moreover, from the perspective of positive criminology, yoga has had several rehabilitative advantages manifest, inter alia, in the development of interpersonal integration, understanding and interpreting daily occurrences from a spiritual outlook, and embracing humane and prosocial views and norms, all of which may assist offenders through their rehabilitation process (Ronel & Ben Yair, 2018).

Studies have shown that the inmate population typically exhibits many disorders (e.g., anxiety, depression, impulsiveness) along with risky behaviors at higher rates as compared with in the normative population (Auty et al., 2015; Muirhead & Fortune, 2016). Yoga addresses many of these problems and can, therefore, serve as both a complementary factor and possibly a positive alternative to the existing rehabilitation programs (Auty et al., 2015; Muirhead & Fortune, 2016; Ronel & Ben Yair, 2018; Wimberly & Xue, 2016).

Rehabilitating Prisoners Through Yoga

Some prison yoga programs are operated by the prison staff, whereas others are managed by community yoga organizations that run the program in the prison using volunteers or by prison-specific yoga organizations such as the Insight Prison Project (Norman, 2015; Silva & Hartney, 2012). Most prison yoga programs offer weekly sessions, during which time, prisoners learn a series of yoga postures and relaxing breathing techniques aiming to calm and ease their pressures as well as the pain associated with incarceration. During the meetings, inmates are encouraged to pay attention to their physical and mental states and needs and identify negative feelings, while acquiring effective coping tools (Auty et al., 2015; Muirhead & Fortune, 2016; Norman, 2015; Silva & Hartney, 2012; Wimberly & Xue, 2016). Some scholars suggest that practicing yoga may help prisoners to channel their acquired self-control into positive and rewarding directions as well postrelease (Silva & Hartney, 2012).

Yoga in prison is considered to be an informal rehabilitative alternative (Karup & Hall, 2016; Muirhead & Fortune, 2016), characterized as having a holistic approach designed to help one gain control over their emotions and behaviors (Auty et al., 2015; Muirhead & Fortune, 2016; Norman, 2015; Silva & Hartney, 2012; Wimberly & Xue, 2016). Studies on yoga indicate that it contributes to prisoners’ beings in various aspects, such as by reducing stress and anxiety (Bilderbeck et al., 2013, 2015), reducing depression (Duncombe et al., 2005; Harner et al., 2010), improving mental health, enhancing anger management capabilities, and reducing impulsiveness (Barrett, 2017; Bartels et al., 2019; Derezotes, 2000; Himelstein, 2011). Inmates engaging in yoga were able to gain greater self-restraint and self-awareness, which, in turn, enhanced their sense of self-worth and helped them control their emotions (Barrett, 2017; Bartels et al., 2019; Griera & Clot-Garrell, 2015; Karup & Hall, 2016; Rucker, 2015).

One study reported that, following their participation in a 10-week-long yoga program, prisoners from a U.K. correctional facility had adopted more positive attitudes, exhibited greater self-control, and reported lower levels of stress and psychological distress than they had experienced before the program (Bilderbeck et al., 2013). Another study evaluated a program called Free Inside, which is offered in Hawaii prisons and which combines yoga classes combined with meditation and Chinese martial arts (chi-gong). It was found that treated prisoners reported significantly less depression, a greater sense of hope, better physical and mental wellness, and a more positive perspective on life (Duncombe et al., 2005). In addition, a qualitative study conducted in Israel examined the overall feeling of participants in yoga programs. The researchers interviewed 22 treated prisoners who overall reported more positive relationships with the prison guards, a positive social atmosphere, and a positive view of their environment (Ronel et al., 2013).

With regard to recidivism as an index for the assessment of the effectiveness of yoga programs, Landau and Gross (2008) examined the effects of a yoga program in a correctional facility in the United States. In their study, participants were not randomly divided into groups; instead, the study group consisted of 52 prisoners who had taken yoga voluntarily, whereas the comparison group was selected out of a group of prisoners who were willing to take part in less than four sessions but eventually did not take part in the actual program. It was found that the rate of reincarceration among the comparison group was higher (25.2%) when compared with that in the treated group (8.5%). However, the researchers noted that the monitoring period among released prisoners in this study did not amount to 3 years, so it is, therefore, difficult to compare its outcomes accurately with recidivism rates commonly found in other studies (Landau & Gross, 2008).

From the above, it would appear that yoga practice has generally positive effects on prisoners in the various areas of mental health, physical health, and anger management. However, an analysis of the rate of recidivism as a measure of rehabilitative programs’ effectiveness has scarcely been adopted in this context. Moreover, most studies conducted on the topic of yoga suffer from weak methodologies, making it harder for them to provide strong conclusions as to the effectiveness of prison yoga programs (Auty et al., 2015; Muirhead & Fortune, 2016; Wimberly & Xue, 2016).

In quasi-experimental studies, as the present study draws on, unlike in randomized-control(led) trials, the process of participants’ selection might cause a bias in which only the best and most motivated patients enter the program (also known as the “creaming effect”; Haviv & Hasisi, 2019). The strength of this study lies in the fact that, beyond the many variables we controlled for, we also controlled for participation in other rehabilitation programs including informal and formal education, vocational programs, domestic violence programs (Hasisi et al., 2016), and the work–release program, which is considered the flagship program of the IPS and which has been found specifically to be effective in reducing recidivism (Elisha et al., 2017; Weisburd et al., 2017). Also, as an indicator of motivation, we controlled for the number of informal education programs that each prisoner participated in.

Method

Yoga Programs in the IPS

IPS’ yoga programs are currently being delivered using the framework of an informal education program. Participation in the program is based on personal choice and willingness on the part of the prisoner. Registration for the program is completed by IPS education officers. Any prisoner can participate in the program regardless of the length of imprisonment or the type of offense he or she committed. Also, there are no prior conditions for participating in the program other than positive behavior in prison and a lack of mental or physical limitations.

Yoga classes are taught by professionally certified yoga instructors who are employees of an external company (Ankori¹). The program involves 12 group sessions with 10 to 12 participants in each group. Classes are held once a week, during evening hours, in the prison wing or at the education center of the prison, with each session lasting 90 min. However, this is an ongoing program that can be entered into or exited at any point in time during its course. This feature may be considered a limitation; however, it should be noted that participants are encouraged to practice the techniques they have learned on a daily basis, meaning that it is a prolonged program whose effects continue even after participation in formal classes has ended.

As for the content of the lessons, they involve both practical and theoretical parts. The theoretical section includes discussions of yoga principles (e.g., body relaxation, inner peace, intention, connection, breathing, expansion), the meaning and advantages of yoga practicing (e.g., to reduce tensions, anxiety, and depression while increasing the level of calm), and the meaning of the body–soul connection. The practical part is based on physical techniques aiming to achieve an inner silence space, calm, and balance through disengaging from external noises, releasing tension, and relaxation. Although meditation is practiced in a separate course, all participants begin the class with a few minutes of silence. Teachers come from different streams of yoga and each teacher has the freedom to choose the appropriate technique. However, most of them practice Vijnana, Ashtanga, Iyengar, or Vinyasa.

Study Design

To examine the effectiveness of prison yoga programs, this study employed a retrospective quasi-experimental design that examined the rate of recidivism among program participants and a comparison group over 5 years of follow-up after release from prison.

Data

The present study employed a secondary database provided directly by the IPS that contained data on 57,764 prisoners in Israel who were released from prison between the years 2004 and 2012, including prisoners’ demographic characteristics and information on their criminal records and incarcerations, including programs in which they had participated. Of this database, 800 prisoners who had participated in a yoga program during their incarceration were selected. To assess the yoga program, we filtered from the database 343 prisoners who had too many missing values in the variables included in the analysis, leaving us with 728 prisoners in the treatment group and 56,693 prisoners in the control group. We then compared these inmates to a similarly sized comparison group by using propensity score matching (PSM).

Variables

The outcomes variable examined in the present study was recidivism within 5 years after release from prison, whereas the independent variable was participation in a prison yoga program. To match the comparison group with the study group, more prisoner background variables were introduced into the model such as incarceration length, date of release, age when first incarcerated, number of prior incarcerations, current age, nationality, marital status, number of years of education, alcohol and drug addictions, violent according to law or the IPS, and the number of informal education courses participated in. Also, we controlled for participation in the following programs: 3 to 5 years of formal education, 6 to 8 years of formal education, 10 years of formal education, 12 years of formal education, vocational program participation, domestic violence program participation, and work–release program participation.

Study Method

To examine the effectiveness of the program and facilitate an appropriate comparison between the two groups of inmates, PSM was used to build the comparison group. This method is designed to minimize differences between study and comparison group members, with the exception of the fact that the former received therapy (i.e., in this case, they participated in a prison yoga program) and the latter did not (Apel & Sweeten, 2010).

The first step was to use a statistical model, based on the prisoners’ wide range of background variables, to calculate scores between 0 and 1 point(s) denoting their propensity to enter treatment. According to this method, when a prisoner’s score was closer to 0 points, their chances of entering treatment were low, whereas a score closer to 1 point meant that their chances of entering the program were higher.

Having calculated the propensity scores, each prisoner from the study group was matched with a single prisoner closest to them in score but differing from them, in that, the matched prisoner had not participated in the program under evaluation (nearest neighbor matching). In other words, to establish the comparison group, every participant in the yoga program was matched with a prisoner who had not participated in the yoga program but who displayed the closest propensity score to theirs. This matching was performed using the best matching method at a caliper of 0.0001, meaning that a successful match between a yoga participant and a non–yoga-participating prisoner was only made if the difference between the two prisoners’ propensity scores did not exceed 0.01% (Hasisi et al., 2016).

Once the matching was completed, and the two groups—study and comparison—were established, a test was conducted to ensure that the propensity-score approach had indeed yielded balanced samples and that no consistent differences were found between the groups with regard to background variables. This test was based on a series of chi-square tests for nominal variables and t tests for interval variables conducted as part of the PSM to ensure that no differences remained consistent between the study and comparison groups with regard to background variables following the matching process (Table 1).

Table 1.

A Comparison Between the Prisoners Who Participated in the Yoga Program and the Comparison Group, Before and After the Matching.

	Before matching (n = 57,421)		Bias %	After matching (n = 1,182)		Bias %
Variables	Yoga (n = 728)	Comparison group (n = 56,693)	Bias %	Yoga (n = 591)	Comparison group (n = 591)	Bias %
Sociodemographic variables
New immigrant	17.31%	18.28%	2.6	17.94%	16.07%	4.9
Married	32.42%	32.87%	1.0	32.83%	32.99%	0.4
Nationality (Jewish)	67.58%***	57.89%***	20.1	66.49%	64.81%	3.5
Age	34.10	33.57	4.7	33.93	33.12	7.2
Number of years of education	9.15***	8.19***	22.3	9.05	9.29	5.7
Criminal background
Number of prior incarcerations	1.99*	2.21*	7.9	1.96	1.99	1.0
Incarceration characteristics
Age when first incarcerated	28.16	27.54	5.9	28.18	27.48	6.7
Violence by law	16.62%	18.68%	5.4	17.09%	15.74%	3.6
Violence by IPS	4.94%	4.53%	2.0	4.57%	4.74%	0.8
Incarceration length (in months)	32.50***	13.54***	72.3	30.32	28.25	7.9
Sexual offense	15.38%***	4.32%***	37.7	11.84%	11.84%	0.0
Against human life offense	4.54%***	2.30%***	12.3	4.74%	3.21%	8.4
Violent offense	45.19%	44.55%	1.3	47.04%	43.82%	6.5
Property offense	38.60%**	33.41%**	10.8	38.41%	40.93%	5.3
Drug offense	23.63%	22.68%	2.2	23.01%	26.40%	8.0
Date of release (year)	2008.1	2008.1	0.3	2008.1	2008.1	0.8
Courses and treatments during incarceration
Number of informal education courses	1.80***	0.35***	216.7	1.57	1.57	0.8
3–5 years of formal education program	12.36%***	4.22%***	29.8	10.66%	12.52%	6.8
6–8 years of formal education program	16.07%***	4.25%***	39.9	14.21%	16.24%	6.8
10 years of formal education program	2.75%***	0.67%***	16.1	2.03%	1.52%	3.9
12 years of formal education program	6.18%***	0.76%***	29.9	4.06%	3.21%	4.7
Vocational programs	19.78%***	4.16%***	49.6	16.41%	12.86%	11.3
Domestic violence program	1.65%***	0.47%***	11.6	1.86%	1.69%	1.7
Work–release program	6.18%***	1.11%***	27.3	5.75%	3.89%	10.0
Prisoner characteristics
Alcohol addiction	5.08%	5.16%	0.4	5.41%	4.57%	3.8
Drug addiction	19.78%	21.65%	4.6	19.63%	21.10%	5.8

Note. IPS = Israel Prison Service.

p < .05. **p < .01. ***p < .001.

In addition, the standardized difference, also known as bias, was also calculated per research by Rosenbaum and Rubin (1985) for each variable to reflect the rate of the difference between the study and comparison groups with reference to the root of the variances between them (Formula 1). When this index drops below 20%, it is considered a sufficient cause to rule that the groups are balanced with regard to the particular background variable examined (Table 1; Caliendo & Kopeinig, 2008; Rosenbaum & Rubin, 1985).

Formula 1: Standardized difference

Bias = \frac{100 ({\bar{x}}_{t} - {\bar{x}}_{c})}{\sqrt{\frac{(s_{t}^{2} + s_{c}^{2})}{2}}}

Finally, having completed the establishment of the study and comparison groups and after ensuring that they are balanced with regard to their background variables, the effectiveness of the yoga program was assessed using recidivism up to 5 years after release as an outcome index. Using a chi-square test, the recidivism rates among prisoners who participated in the yoga program were compared with those found among the comparison group members in two channels: arrests and reincarcerations. In cases where statistically significant differences were found, Cohen’s d was calculated to examine the effect size (Formula 2).

Formula 2: Cohen’s d

Cohen s d = \frac{{\bar{X}}_{t} - {\bar{X}}_{c}}{\sqrt{\frac{S D_{t}^{2} + S D_{c}^{2}}{2}}}

Results

Matching the Comparison (Nonyoga) Group to the Yoga Participants Group

The group of prisoners who participated in the prison yoga program involved a total of 726 prisoners prior to matching, with an average propensity score of 0.131 (SD = 0.151; minimum = 0.006, maximum = 0.849). Separately, the comparison group consisted, prior to matching, of 56,693 prisoners who had not participated in the prison yoga program, and their average propensity score was 0.011 (SD = 0.037; minimum = 0.000, maximum = 0.854). After matching, the yoga program participant group consisted of 591 prisoners with an average propensity score of 0.077 (SD = 0.074; minimum = 0.006, maximum = 0.437) and the comparison group consisted of 591 prisoners with an average propensity score of 0.077 (SD = 0.074, minimum = 0.006, maximum = 0.437).

Table 1 shows that, prior to PSM group matching, there were significant differences between the two groups in 15 variables: nationality, number of years of education, number of prior incarcerations, incarceration length, sexual offense, “against human life” offense, property offense, number of informal education courses, 3 to 5 years of formal education, 6 to 8 years of formal education, 10 years of formal education, 12 years of formal education, vocational program participation, domestic violence program participation, and work-release program participation.

Rearrests Among Yoga Program Participants

Table 2 and Figure 1 present rearrest rates among the prisoners who participated in the yoga program as compared with among the prisoners who did not take part in the yoga program during the incarceration period. No significant differences between the two were found.

Table 2.

Rearrest Rate Following Release Among Prisoners Who Participated/Did Not Participate in the Yoga Program.

	Within 1 year	Within 2 years	Within 3 years	Within 4 years	Within 5 years
Comparison group	(N = 591)	(N = 591)	(N = 575)	(N = 496)	(N = 432)
Comparison group	24.03%	38.41%	44.87%	51.21%	55.56%
Study group	(N = 591)	(N = 591)	(N = 583)	(N = 531)	(N = 479)
Study group	23.69%	37.56%	44.94%	49.15%	51.36%
Chi-square	0.02	0.09	0.00	0.43	1.61
Cohen’s d	0.008	0.017	0.001	0.041	0.084

p < .05. **p < .01. ***p < .001.

Figure 1.

Rearrest rate following release among prisoners who participated/did not participate in the yoga program.

Reincarcerations Among Yoga Program Participants

Table 3 and Figure 2 present recidivism rates observed among the prisoners who participated in the yoga program in comparison with those who did not during their respective incarceration periods. It is evident that, throughout the monitoring period, the rates of recidivism among the study group are significantly lower as compared with those in the comparison group. Of note, the risk of recidivism within 1 year of release from prison among the yoga program participants group was 29.8% smaller than that among the comparison group, whereas, within 2 years of release, the risk of recidivism among the yoga program participants group was 20.4% smaller than that among the comparison group. Within 3 years of their release from prison, the risk of recidivism was 17.8% smaller among yoga participants than among the comparison group prisoners, whereas, in the fourth year, it was 16.7% smaller and, in the fifth year, it was 17.0% smaller. It is important to note that the effect sizes over the same period of time remain small throughout.

Table 3.

Reincarceration Rate Following Release Among Prisoners Who Participated/Did Not Participate in the Yoga Program.

	Within 1 year	Within 2 years	Within 3 years	Within 4 years	Within 5 years
Comparison group	(N = 591)	(N = 591)	(N = 575)	(N = 496)	(N = 432)
Comparison group	15.91%	26.57%	31.30%	37.10%	40.74%
Study group	(N = 591)	(N = 591)	(N = 583)	(N = 531)	(N = 479)
Study group	11.17%	21.15%	25.73%	30.89%	33.82%
Chi-square	5.67*	4.77*	4.42*	4.42*	4.66*
Cohen’s d	0.14	0.13	0.12	0.13	0.14

p < .05. **p < .01. ***p < .001.

Figure 2.

Reincarceration rate following release among prisoners who participated/did not participate in the yoga program.

Treatment Effect Sensitivity Test

When using PSM, one would ideally take into consideration every variable that may affect the outcomes variable; however, because it is impossible to introduce all these variables into the present analysis and identify every factor that may affect prisoners’ decision to enter into treatment, this study is vulnerable to implicit bias (Rosenbaum, 2002). A sensitivity test was, therefore, conducted in an attempt to assess the degree of bias inherent in the present study. Due to the outcomes variable’s binary nature, the most suitable sensitivity test in this case was deemed to be the Mantel–Heanszel bounds test (MHbounds; Aakvik, 2001). During this test, when γ exceeds 1, such may indicate the existence of implicit variables affecting the outcomes of the test. Importantly, although the test examines an analysis’ sensitivity to the existence of implicit variables that were not introduced into it, it does not determine that such variables do indeed exist.

The outcomes of MHbounds tests may be interpreted in one of two ways. The first is Qmh+, which assumes a treatment’s overestimation, that is, a state in which treated participants have a greater chance of achieving higher values for the outcomes variable, whereas, the second option, Qmh−, assumes a treatment’s underestimation, that is, a state in which treated participants have a greater chance of achieving lower values in the outcomes variable (Caliendo & Kopeinig, 2008; Weisburd et al., 2017). Because respondents in the present study were expected to achieve lower values in the recidivism rate, a Qmh− interpretation was employed.

In the present study, the γ index test remained significant up to a value of 1.1 to 1.15 (Table 4), indicating the study’s vulnerability to bias only if an implicit variable that has not been taken into account presents a 10% to 15% effect on the likeliness of entering into treatment. The results of the sensitivity test in the present study correlate with the findings of both other social science studies in general and criminology in particular, though are lower than those of medical studies (Weisburd et al., 2017). Therefore, in any event, the results of the present study must be considered with caution.

Table 4.

MHbounds Sensitivity Test Results.

	Gamma (Qmh−)	p critical
Reincarceration
Recidivism 1 year	1.1	.04
Recidivism 1 year	1.15	.07
Recidivism 2 years	1.05	.04
Recidivism 2 years	1.1	.08
Recidivism 3 years	1.05	.04
Recidivism 3 years	1.1	.09
Recidivism 4 years	1.05	.04
Recidivism 4 years	1.1	.09
Recidivism 5 years	1.05	.04
Recidivism 5 years	1.1	.08
Rearrest
Recidivism 1 year	1.0	.47
Recidivism 2 years	1.0	.40
Recidivism 3 years	1.0	.51
Recidivism 4 years	1.0	.27
Recidivism 5 years	1.0	.11

Discussion

In recent years, various rehabilitation programs have been introduced into prisons worldwide with the aim of reducing recidivism rates (Enosh et al., 2013; Gideon et al., 2010; Hasisi et al., 2016; Lipsey & Cullen, 2007; MacKenzie, 2006; McGuire, 2003; Ronel et al., 2013). These include informal practices such as sports and yoga (Barrett, 2017; Guse & Hudson, 2014; Kewley, 2017; Ronel & Segev, 2014; Ward et al., 2007). Currently, yoga courses are offered as part of prisons’ rehabilitation programs worldwide, including in the IPS.

Yoga is perceived as a holistic tool for achieving a greater body–soul connection by practicing physical postures and breathing techniques that connect practitioners to their inner silence and assist them with achieving balance and internal control through reflection, emotions recognition, and anger management. Thus, yoga contributes to development on both personal and spiritual levels. Improving the connection between the body and soul along with performing breathing and relaxation techniques are therapeutic tools that assist in rehabilitation and healing processes (Auty et al., 2015; Chong et al., 2011; Hartfiel et al., 2011; Karup & Hall, 2016; Kirkwood et al., 2005; Li & Goldsmith, 2012) while also serving as mechanisms of social control (Norman, 2015).

The present study was designed to examine the correlation between yoga practice and recidivism reduction among prisoners released from the IPS who practiced yoga during their incarceration in comparison with a matched control group of non–yoga-participating prisoners over a follow-up period of 5 years after discharge from prison. Study findings support the hypothesis that practicing yoga during incarceration has a positive effect on reducing recidivism rates (e.g., Barrett, 2017; Bartels et al., 2019; Griera & Clot-Garrell, 2015; Karup & Hall, 2016; Rucker, 2015).

In this study, it was found that released prisoners who had practiced yoga while incarcerated demonstrated various lower rates of reincarceration throughout the 5-year follow-up period. For example, in the first year after release, the study group’s risk of reincarceration (11.17%) was significantly lower than of the comparison group (15.91%). Similar differences were found between the study and comparison groups’ risks of reincarceration 2 years after release (21.15% vs. 26.57%, respectively), 3 years after release (25.73% vs. 31.30%, respectively), 4 years after release (30.89% vs. 37.10%, respectively), and 5 years after release (33.82% vs. 40.74%, respectively). Conversely, an examination of the rearrest outcomes index did not show any significant differences between the study and comparison groups’ risks of rearrest.

These positive findings are in line with the positive criminology approach that emphasizes the importance of informal support factors such as spirituality, which yoga is included in. Yoga allows for practitioners to acquire personal and social strengths such as greater self-awareness, emotional regulation, self-control, and social acceptance. It is possible that these specific skills (in addition to other tools gained through other rehabilitation programs in prison) increased prisoners’ ability to cope with life stressors after their release from prison. As positive criminology argues, prison rehabilitation programs should be holistic and include positive elements, as they enhance rehabilitation processes (e.g., hope, faith, love, trust, compassion, happiness, acceptance, humanity; Ronel, 2015; Ronel & Toren, 2012). Similarly, yoga contains positive rehabilitative components that assist offenders in identifying and revealing their positive potential, while recognizing human goodness (Rucker, 2015).

The present study findings strengthen those of previous studies conducted on the effects of the inclusion of yoga practice in prisons. For example, Pham (2013) found that prisoners who practiced yoga while incarcerated reported higher levels of spirituality and mental stability as compared with prisoners who did not practice yoga. An experimental study conducted in the United Kingdom found that prisoners who continued to practice yoga in their cells after the program’s official time was over reported significantly lower levels of stress and negative perceptions (Bilderbeck et al., 2015). Moreover, Norman (2015) found that prisoners in his study regarded yoga as a tool that could assist them with reducing violent behavior as well as optimally channeling their negative feelings such as anger and frustration (Norman, 2015).

In addition to the above findings on the advantages of yoga practice during incarceration, the present study indicates that yoga has an ongoing positive effect reaching far beyond the prison walls, as the recidivism rates remained lower over the 5 years following release among released prisoners who participated in prison yoga programs than their non–yoga-participating counterparts. This can be explained by the fact that this is an enduring program, regardless of consistently participating in formal classes, as the independent practice of the techniques they acquired, along with other restorative components learned during incarceration, can strengthen their rehabilitative process in the community. These findings are supported by previous studies that also observed positive effects on prisoners who participated in a yoga program (Duncombe et al., 2005; Landau & Gross, 2008; Ronel et al., 2013). However, although previous studies suffered from weak methodologies including small samples, unbalanced groups, and short follow-up periods (Auty et al., 2015; Muirhead & Fortune, 2016; Wimberly & Xue, 2016), the present study aims to overcome these methodological problems. This can be explained by the fact that this is an enduring program, regardless of consistently participating in formal classes, as the independent practice of the techniques they acquired with their therapeutic ingredients can strengthen their rehabilitative process in the community.

Another issue that requires explanation is why this study found a significant decrease in reincarceration rates rather but no reduction in rearrests. In this context, it is worth noting that released inmates are often supervised by intelligence officers and monitored by law enforcement agencies after release and might be viewed as an “immediate suspects” when crime occurs. This could lead to an increase of arrests among released prisoners. But these arrests do not necessary end up with incarcerations. Hence, we view reincarceration as a much more accurate indicator of recidivism than rearrests.

Of note, our findings indicating a decline in recidivism rates have small effect sizes. Whenever recidivism is measured and monitored in a certain group, the number of prisoners in the study group and comparison group is expected to remain identical throughout the 5-year follow-up period. However, the database file used in the present study does not contain recidivism data of prisoners released in later years, as the file was not further updated beyond 2015. Therefore, recidivism measurements each year were conducted only among the group of prisoners, for whom recidivism data were found, making the groups examined increasingly smaller over the 5 years of follow-up. This decline in the number of both study and comparison group members throughout the follow-up period affects the statistical strength of this investigation as well as the possibility of achieving statistical significance in later follow-up years.

Limitations

Despite the positive results found in the present study, it is not without limitations. Although it offers significant progress and a strong methodology, it is not experimental—a fact that must be taken into consideration. In this context, although the present study has controlled a variety of characteristics related to the prisoner and his imprisonment, it is still subjected, like any quasi-experimental studies, to nonobservable biases. Therefore, the authors encourage future researchers to use randomized controlled trials when examining rehabilitation programs in general and yoga programs specifically.

Also, the study is retrospective, utilizing secondary data retrieved in advance; therefore, for these reasons, it is subject to limitations and inflexibility when presenting and encoding the variables. For example, as a result of inconsistent system coding, it is not possible to know from the data the exact number of yoga classes to which each prisoner attended in prison. However, we know for sure that the treatment group includes only prisoners who participated in yoga classes, whereas in the control group, no prisoner participated. Hence, the positive effect found can be attributed to the yoga program.

With regard to the sensitivity analysis conducted in the present study, another limitation is that there is no consensus in the literature regarding a certain threshold value that the MHbounds sensitivity test should meet. The results of the sensitivity test conducted in the study show that it is necessary to treat the findings with caution.

An additional limitation of the present study is its inability to control prisoners’ motivation and character. In an attempt to counter this limitation, an effort was made during participant selection to choose those who were most similar to members of the study group. We also added variables that might reflect levels of inmates’ motivation (e.g., prior participation in education programs) to the matching process. Nevertheless, in the absence of the random allocation of respondents to study and comparison groups, it is hard to rule out variables that might have biased the results. However, our analysis showed that both the effect sizes and the sensitivity tests indicated that the results are relatively stable. Yet, it is difficult to exclude the bias of motivation or other variables; therefore, the findings should be treated with caution.

Also, in the “Method” section, we noted that any prisoner could participate in the yoga program, regardless of their imprisonment length or offense type. However, a prisoner who wishes to participate in the program should behave positively and should be able to engage both mentally and physically. In this context, one limitation in the present study is that the data used are lacking a clear indication of the prisoner’s behavior during imprisonment. In this respect, however, it is conjectured that the relevant variables controlled for in the matching process, including “Violence by IPS” (which is an internal profile of the prison service relating to violence), participation in additional programs, addiction, and the type of offense may improve the chances of getting balanced groups.

In addition, the data file we received from the IPS did not specify in which prison did the prisoner undergo the program. In this context, it is worthy to note that the number of prisons in which the program operates is limited; however, the facilities, security level, and conditions of the prisons are much alike.

Conclusion

The contribution of the present study is an enrichment and expansion of the current body of knowledge on prisoner rehabilitation, both in light of the positive criminology approach and, more specifically, on the role of yoga programs. Yoga provides new insights into positive and spiritual rehabilitation approaches that holistically meet prisoners’ needs during their rehabilitation process and can help in the reentry process. Our findings suggest that yoga programs have a positive effect on reducing recidivism. However, follow-up experimental studies are needed to understand the specific effects of yoga on prisoners, and also to examine whether participation in yoga programs can have a similar effect on different prisoners based on the type of offense, motivation, and other characteristics.

Footnotes

Authors’ Note

Noam Haviv is also affiliated with The Hebrew University, Israel.

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) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Noam Haviv

Ety Elisha

Notes

References

Aakvik

(2001). Bounding a matching estimator: The case of a Norwegian training program. Oxford Bulletin of Economics and Statistics, 63(1), 115–143.

Andrews

D. A.

Bonta

Hoge

R. D.

(1990). Classification for effective rehabilitation: Rediscovering psychology. Criminal Justice and Behavior, 17(1), 19–52.

Andrews

D. A.

Bonta

Wormith

J. S.

(2011). The risk-need-responsivity (RNR) model: Does adding the good lives model contribute to effective crime prevention? Criminal Justice and Behavior, 38, 735–755.

Apel

R. J.

Sweeten

(2010). Propensity score matching in criminology and criminal justice. In Piquero

A. R.

Weisburd

(Eds.), Handbook of quantitative criminology (pp. 543–562). Springer.

Auty

K. M.

Cope

Liebling

(2015). A systematic review and meta-analysis of yoga and mindfulness meditation in prison effects on psychological well-being and behavioural functioning. International Journal of Offender Therapy and Comparative Criminology, 61(6), 689–710. https://doi.org/10.1177/0306624X15602514

Barnes

P. M.

Bloom

Nahin

R. L.

(2008). Complementary and alternative medicine use among adults and children: United States, 2007. National Health Statistics Reports, 12, 1–23.

Barrett

C. J.

(2017). Mindfulness and rehabilitation: Teaching yoga and meditation to young men in an alternative to incarceration program. International Journal of Offender Therapy and Comparative Criminology, 61(15), 1719–1738.

Bartels

Oxman

L. N.

Hopkins

(2019). “I would just feel really relaxed and at peace”: Findings from a pilot prison yoga program in Australia. International Journal of Offender Therapy and Comparative Criminology, 63, 2531–2549. https://doi.org/10.1177/0306624X19854869

Berman

Walk

(2015). The recidivism of Israeli criminal prisoners released in 2008. Israel Prison Service Research Unit. (In Hebrew)

10.

Bilderbeck

A. C.

Brazil

I. A.

Farias

(2015). Preliminary evidence that yoga practice progressively improves mood and decreases stress in a sample of UK prisoners. Evidence-Based Complementary and Alternative Medicine, 2015, Article 819183.

11.

Bilderbeck

A. C.

Farias

Brazil

I. A.

Jakobowitz

Wikholm

(2013). Participation in a 10-week course of yoga improves behavioural control and decreases psychological distress in a prison population. Journal of Psychiatric Research, 47(10), 1438–1445.

12.

Braithwaite

(2001). Restorative justice and a new criminal law of substance abuse. Youth & Society, 33(2), 227–248.

13.

Caliendo

Kopeinig

(2008). Some practical guidance for the implementation of propensity score matching. Journal of Economic Surveys, 22(1), 31–72.

14.

Chiesa

Serretti

(2009). Mindfulness-based stress reduction for stress management in healthy people: A review and meta-analysis. The Journal of Alternative and Complementary Medicine, 15(5), 593–600.

15.

Chong

C. S.

Tsunaka

Chan

E. P.

(2011). Effects of yoga on stress management in healthy adults: A systematic review. Alternative Therapies in Health and Medicine, 17(1), 32–38.

16.

Cramer

Ward

Steel

Lauche

Dobos

Zhang

(2016). Prevalence, patterns, and predictors of yoga use: Results of a US nationally representative survey. American Journal of Preventive Medicine, 50(2), 230–235.

17.

Derezotes

(2000). Evaluation of yoga and meditation trainings with adolescent sex offenders. Child and Adolescent Social Work Journal, 17(2), 97–113.

18.

Duncombe

Komorosky

Wong-Kim

Turner

(2005). Free inside: A program to help inmates cope with life in prison at Maui Community Correctional Center. Californian Journal of Health Promotion, 3(4), 48–58.

19.

Elisha

Shoham

Hasisi

Weisburd

(2017). For prisoners, “work works”: Qualitative findings from an Israeli program. The Prison Journal, 97(3), 342–363.

20.

Enosh

Buchbinder

Smith

L. N.

Shafir

(2013). From the “greenhouse” to reality: Challenges faced by graduates of in-prison batterers intervention program. Journal of Interpersonal Violence, 28(5), 910–937.

21.

Fazel

Wolf

(2015). A systematic review of criminal recidivism rates worldwide: Current difficulties and recommendations for best practice. PLOS ONE, 10(6). https://doi.org/10.1371/journal.pone.0130390

22.

Gideon

Shoham

Weisburd

D. L.

(2010). Changing prison into a therapeutic milieu: Evidence from the Israeli national rehabilitation center for prisoners. The Prison Journal, 90(2), 179–202.

23.

Griera

Clot-Garrell

(2015). Doing yoga behind bars: A sociological study of the growth of holistic spirituality in penitentiary institutions. In I. Becci & O. Roy (Eds.), Religious diversity in European prisons (pp. 141–157). Springer.

24.

Guse

Hudson

(2014). Psychological strengths and posttraumatic growth in the successful reintegration of South African ex-offenders. International Journal of Offender Therapy and Comparative Criminology, 58(12), 1449–1465.

25.

Harner

Hanlon

A. L.

Garfinkel

(2010). Effect of Iyengar yoga on mental health of incarcerated women: A feasibility study. Nursing Research, 59(6), 389–399.

26.

Hartfiel

Havenhand

Khalsa

S. B.

Clarke

Krayer

(2011). The effectiveness of yoga for the improvement of well-being and resilience to stress in the workplace. Scandinavian Journal of Work, Environment & Health, 37(1), 70–76.

27.

Hasisi

Shoham

Weisburd

Haviv

Zelig

(2016). The “care package,” prison domestic violence programs and recidivism: A quasi-experimental study. Journal of Experimental Criminology, 12(4), 563–586.

28.

Haviv

Hasisi

(2019). Prison Addiction Program and the Role of Integrative Treatment and Program Completion on Recidivism. International journal of offender therapy and comparative criminology, 63(15-16), 2741–2770.

29.

Himelstein

(2011). Mindfulness-based substance abuse treatment for incarcerated youth: A mixed method pilot study. International Journal of Transpersonal Studies, 30(1–2), 1–10.

30.

Karup

Hall

(2016). The meaning and effects of yoga in prison [Unpublished master’s thesis]. University of Cambridge.

31.

Kewley

(2017). Strength based approaches and protective factors from a criminological perspective. Aggression and Violent Behavior, 32, 11–18.

32.

Kirkwood

Rampes

Tuffrey

Richardson

Pilkington

(2005). Yoga for anxiety: A systematic review of the research evidence. British Journal of Sports Medicine, 39(12), 884–891.

33.

Landau

P. S.

Gross

J. B.

(2008). Low reincarceration rate associated with Ananda Marga yoga and meditation. International Journal of Yoga Therapy, 18(1), 43–48.

34.

A. W.

Goldsmith

C. A.

(2012). The effects of yoga on anxiety and stress. Alternative Medical Review, 17(1), 21–35.

35.

Lipsey

M. W.

Cullen

F. T.

(2007). The effectiveness of correctional rehabilitation: A review of systematic reviews. Annual Review of Law and Social Science, 3, 297–320.

36.

MacKenzie

D. L.

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

37.

Maruna

(2001). Making good: How ex-convicts reform and rebuild their lives. American Psychological Association.

38.

McGuire

(Ed.). (2003). Offender rehabilitation and treatment: Effective programmes and policies to reduce re-offending. John Wiley.

39.

Muirhead

Fortune

C. A.

(2016). Yoga in prisons: A review of the literature. Aggression and Violent Behavior, 28, 57–63.

40.

Norman

(2015). Prison yoga as a correctional alternative? Physical culture, rehabilitation, and social control in Canadian prisons. In Joseph

Crichlow

(Eds.), Alternative offender rehabilitation and social justice: Arts and physical engagement in criminal justice and community settings (pp. 78–98). Palgrave Macmillan.

41.

Pew Center on the States. (2011). State of recidivism: The revolving door of America’s prisons. Pew Charitable Trusts.

42.

Pham

K. H.

(2013). Outcomes of a recreation therapy yoga meditation intervention on prison inmates’ spiritual well-being [Unpublished master’s thesis]. San Jose University.

43.

Ronel

(2015). How can criminology (and victimology) become positive? In Ronel

Segev

(Eds.), Positive criminology (pp. 13–31). Routledge.

44.

Ronel

Ben Yair

(2018). Spiritual criminology: The case of Jewish criminology. International Journal of Offender Therapy and Comparative Criminology, 62(7), 2081–2102.

45.

Ronel

Elisha

(2011). A different perspective: Introducing positive criminology. International Journal of Offender Therapy and Comparative Criminology, 55(2), 305–325.

46.

Ronel

Frid

Timor

(2013). The practice of positive criminology: A vipassana course in prison. International Journal of Offender Therapy and Comparative Criminology, 57(2), 133–153.

47.

Ronel

Segev

(2014). Positive criminology in practice. International Journal of Offender Therapy and Comparative Criminology, 58(11), 1389–1407.

48.

Ronel

Toren

Y. A. T.

(2012). Positive victimology: An innovation or “more of the same”? Temida, 15(2), 171–180.

49.

Rosenbaum

P. R.

(2002). Sensitivity to hidden bias. In Rosenbaum

P. R.

(Ed.), Observational studies (pp. 105–170). Springer.

50.

Rosenbaum

P. R.

Rubin

D. B.

(1985). Constructing a control group using multivariate matched sampling methods that incorporate the propensity score. The American Statistician, 39(1), 33–38.

51.

Ross

R. R.

(1980). Socio-cognitive development in the offender: An external review of the UVIC program at Matsqui penitentiary. University of Ottawa Department of Criminology.

52.

Ross

R. R.

Hilborn

(2008). Rehabilitating rehabilitation: Neurocriminology for treatment of antisocial behavior. Cognitive Center of Canada.

53.

Rucker

(2015). Yoga and restorative justice in prison: An experience of “response-ability to harms.” Contemporary Justice Review, 8(1), 107–120.

54.

Sampson

R. J.

Laub

J. H.

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

55.

Shoham

Zelig

Hasisi

Weisburd

Haviv

(2018). “The whole is greater than the sum of the parts”: Prison staff perceptions of domestic violence rehabilitation programs. International Journal of Offender Therapy and Comparative Criminology, 62(11), 3298–3321.

56.

Shoham

Zelig

Hesisi

Weisburd

Haviv

(2017). The “black box” behind prison-based vocational training programs. European Scientific Journal, 13(12), 432–443.

57.

Silva

Hartney

(2012). Evaluation of the Insight Prison Project. National Council on Crime & Delinquency.

58.

Ward

Mann

R. E.

Gannon

T. A.

(2007). The good lives model of offender rehabilitation: Clinical implications. Aggression and Violent Behavior, 12(1), 87–107.

59.

Weisburd

Hasisi

Shoham

Aviv

Haviv

(2017). Reinforcing the impacts of work release on prisoner recidivism: The importance of integrative interventions. Journal of Experimental Criminology, 13(2), 241–264.

60.

Wimberly

A. S.

Xue

(2016). A systematic review of yoga interventions in the incarcerated setting. Journal of Sociology & Social Welfare, 43(4), 85–108.

61.

Zara

Farrington

D. P.

(2015). Criminal recidivism: Explanation, prediction and prevention. Routledge.