Low Reincarceration Rate Associated with Yoga and Meditation
Pashupati Steven Landau, MD, FAAFP, ABHM, E-RYT
1
Jagat Bandhu John Gross, PhD2
1. University of North Carolina, Chapel Hill, Schools of Medicine and Nursing
2. Department of Economics, Duke University, North Carolina
Abstract:  Yoga was taught to male inmates at Wake Correctional Center in Raleigh, NC. A five-year study of 190 inmates found that those who were taughtYoga were significantly less likely to be reincarcerated upon release. Subjects were divided into two groups, those who attended at least one, but less than four classes, and those who attended more than four classes. Differences in reincarceration rates between these two groups during a two-year post-release period were striking. Of
those who attended more than four classes, 8.5% were reincarcerated, while 25.2% of those who attended fewer than four classes were reincarcerated during this same period. "is difference was found to be statistically significant at the 0.025 level.

Keywords: Yoga, meditation, prison, recidivism, reincarceration

Introduction
!ere is growing evidence to support the teaching of spiritual practices, such as Yoga and meditation, in prisons. A Minnesota-based consensus panel observed that spiritual approaches may be quite valuable in substance-abuse treatment for offenders, and recommended providing a time and suitable place for individual meditation, reflection, or prayer.
1 A 2002 study at Seattle’s North Rehabilitation Facility found that the recidivism rate for inmates who took a 10-day Vipassana meditation retreat was 56%, a 25% improvement over recidivism rates in the general inmate population.

2  A follow-up study by the Addictive Behaviors Research Center at the University of Washington found that drug use, drug and alcohol-related consequences, and self-reported levels of depression and hostility were significantly lower among those who took the course, compared to those who did not.

3 In January of 2002, the first author (Landau) began teaching weekly classes of Yoga postures, meditation, and philosophy to male inmates at Wake Correctional Center (WCC) in Raleigh, NC. WCC is a minimum-security prison with dorms instead of individual cells, and many of the inmates spend much of their time on work release. Some of them are allowed to visit their families on weekends. Almost all of the prisoners at WCC are felons, having been convicted of burglary, sex crimes, assaults, drug dealing, DWI, fraud, and the like.!is study follows the outcomes of prisoner release over a five-year period. We expected that regular participation in a Yoga program would help inmates stay out of prison once they were
released. !is study tests the hypothesis that inmates who took four or more Yoga classes would have lower reincarceration rates after release than inmates who took fewer than four classes. Acknowledgements: To Paulette Stiles, for helping out with data acquisition and analysis. To the Wake Correctional staff and administration, for being supportive and tolerant. To Stephan Kiefer, David Edwards, and Michelle Beck
at the NC DOC Research and Planning Department, for doing the detailed analysis of the first group of four-class graduates. To the inmates themselves for being wonderful in their own way and for demonstrating what is possible. To the  volunteers and acharyas who helped teach and cook. To Rakesh Rickey Moore, for helping research the bibliography. 44 INTERNATIONAL JOURNAL OF YOGA THERAPY  –  No. 18 (2008)

Methods
Yoga Participants
Participation in the ongoing Yoga class is voluntary; prisoners do not need special permission to attend, nor are they required to attend a minimum number of classes. At about 1:00 PM on Sundays, the Yoga class is announced over the camp loudspeaker. From four to eight inmates attend each session. About once a month, local volunteers prepare a delicious vegetarian meal that is pre-approved by the prison officials,
and we announce the meal over the loudspeaker. !is usually brings a substantially larger crowd, up to 15 at a time.Each inmate signs an attendance sheet in class that includes his Department of Corrections (DOC) OPUS number. After an inmate has attended four sessions, he gets a Certificate of Attendance encased in a blue-and-gold certifi-cate jacket. !ese inmates are welcome to continue attending classes, and receive additional certificates at four-week intervals. In this study, we considered inmates who attended
at least four classes to be part of the “Yoga” group.!e Department of Corrections performed an analysis
of the first 52 inmates who participated in at least four Yoga classes (based on participant attendance from the roll sheets distributed at each class). !e median age at which these inmates most recently entered prison was 31 years. !e median grade level these inmates reported completing was 12th grade.
All inmates were male. 21.1% of participants were married; the rest were single, divorced, separated, or widowed.
 Table
1 shows the ethnic and racial distribution of the participants.
Table 2 shows the religious affiliations of the participants.
Most (88.5%) of the participants were felons. Tables 3 and 4
show the criminal backgrounds of the participants.
!e Yoga Program
!e classes take place in a modular classroom building with movable desks and chairs. !e first hour covers the
philosophy of Yoga. Participants are invited to share relevant personal experiences and discuss the applicability of Yogic principles to personal problems. We discuss yogic moral principles, including yama and niyama, conduct codes, relationships, the essence of meditation and mantra, and anything else the inmates wish to discuss, including sex, drugs, physiology, family, medical concerns, and religious issues.
!e Ananda Marga approach to religious issues is to emphasize the similarities between them, and to show how the Yoga/Tantra approach provides a rational basis for the common teachings of most religions. With this approach, most inmates are able to accept and absorb the principles, even though they come from a variety of religious backgrounds. For the second hour, the inmates move the tables and
desks to clear the space for Yoga practice. Blankets are retrieved from a locked cabinet, and Yoga mats are also used at times. A portable CD/tape player plays soft music or kirtan (chanting). All sit on the floor and center themselves with Namaskar (“I salute the Divinity within you with all the divine charms of my mind, and all the love and cordiality of my heart”). !is allows the participants to embody a sense of mutual respect, one of the central aims of the program.

!e âsanas are aimed at calming rather than energizing, and are targeted to help the inmates feel good in the situation that they find themselves in at the time. Participants practice the three-part Yogic breath (expanding the abdomen, rib cage, and chest while lifting arms overhead). Standing poses may include tree (vrksâsana), dancer pose (natarâjâsana), chair (utkatâsana), and hand-to-foot pose (pâdâhastâsana). Seated
poses include stretches to prepare to sit in a meditation pose (e.g., padmâsana or sukhâsana), Yogic seal (Yoga mudrâ), cobra (bhujangâsana), worship of the supreme (diirgha prânam),
twists of varying levels of difficulty (e.g., matsyendrâsana), bellows pose (bhastrikâsana),  knee-to-chest pose (utks’epa mudrâ) head to one knee (janushirsâsana), sitting forward bend (paschimottânâsana), shoulder stand (sarvângâsana), fish pose (matTable 1.  Ethnic Distribution
Ethnic Group Participants Percentage
African-American 23 44.2%
European 16 30.8%
Hispanic 3 5.8%
Asian/Pacific Islander 2 3.8%
Unknown 8 15.4%
Tables 1-4: Demographic analysis on 52 of the prisoners who attended the class at least 4 times. LOW REINCARCERATION RATE ASSOCIATED WITH YOGA 45
Table 2.  Religious Affiliation
Religion Participants Percentage
Christianity 21 40.4%
Islam 5 9.6%
Moorish Science 4 7.7%
Buddhism 3 5.8%
Wicca 1 1.9%
Rastafarianism 1 1.9%
Other 4 7.7%
None 13 25.0%
Table 3.  Crime Categories
Crime Category Participants Percent
Crimes Against a Person 19 36.5%
Robbery 14 26.9%
Assault 2 3.9%
Sexual Assault 2 3.9%
Other Sexual Offense 1 1.8%
Property Crimes 11 21.2%
Larceny 3 5.7%
Burglary 2 3.9%
Fraud 2 3.9%
Forgery 2 3.9%
Breaking and Entering 1 1.9%
Burnings 1 1.9%
Public Order Crimes 22 42.3%
Drugs: Trafficking 8 15.4%
Drugs: Non Trafficking 6 11.5%
Driving While Impaired 4 7.7%
Other Traffic Violations 1 1.9%
Other Public Order 3 5.8%
syâsana) with various modifications, depending on participants’ flexibility and blood pressure, and other âsanas based on the needs of the class (both spoken by the inmates, and as perceived by the instructor). We do not teach advanced prânâyâma. Usually a light and humorous tone is maintained. !ere have
never been any behavior disturbances in the classes.46 INTERNATIONAL JOURNAL OF YOGA THERAPY  –  No. 18 (2008)

After a 30- to 45-minute âsana practice, participants give themselves a total body self-massage in the Ananda Marga style, and then practice corpse pose (savâsana) with a guided relaxation. After relaxation, the instructor introduces  mantra and meditation, and usually a  kirtan using the mantra  “Baba Nam Kevalam” (“Beloved’s Name Only”). !e instructor plays the guitar, and the inmates enjoy playing percussion instruments, including African drum, Spanish castanets, Latino ebony sticks, metal triangle, maracas, and egg-shakers. At times, kirtan is combined with Lalita Marmika, a movement meditation (the
inmates are not comfortable hearing it called a “dance”).

Many are shy about this in the beginning, and enjoy it immensely later on. After  kirtan, all sit for meditation, often introduced with the “Sam’gacchadvam’” mantra from the Rg Veda. !e meaning of practicing this mantra is explained: In a proper society, we all move together and do things together, even though in prison and in school we are usually encouraged and required to do things independently. !is is followed
by ten minutes of silent meditation, in which participants are encouraged to imagine that they are surrounded by the
Divine. !ey may focus on the religious figure or prayer of their choice, or use the mantra “Baba Nam Kevalam,” breathing in with “Baba Nam” and breathing out with “Kevalam.” At the end of meditation, there may be an offering of “the colors of the mind” to the Creator, so that whatever the participants are attached to and troubled with gets offered back to the Source of All. Single-page handouts are provided, including summaries of the âsanas, song sheets, and descriptions of the eightlimbed path of Ashtanga Yoga, the yamas and niyamas, and the 15 shiilas (interpersonal behavior guidelines). Participants are encouraged to read copies of We’re All Doing Time by Bo Lozoff, generously donated by the Human Kindness
Foundation. We also provide some meditation books on spiritual practices. Every few months, a visiting meditation teacher or sannyasin from Ananda Marga comes to visit. !ese teachers give inspiring talks, and give personal initiation into mantra meditation. Monthly vegetarian meals are provided to the inmates, prepared by local Ananda Marga folks, and the class size doubles or triples at
these times.

Practical Considerations for Teaching and Conducting Research in Prison. Prison is not a place where people bend the rules, and our program must adhere to many rules. Before the study began, the Department of Corrections reviewed the protocol of the study and required the first author to sign a confidentiality and ethics agreement to protect the rights of the inmates. It takes several weeks to clear the background
checks of any guest meditation instructor. Providing things to inmates like pratiks (symbolic Ananda Marga pendants) may cause cancellation of the program. Keys are signed in and out, and roll is carefully kept. Prisoners are not allowed to do Yoga  âsanas in their dorms, so an arrangement has been made that allows them to borrow Yoga mats from the library at certain times during the week and practice their exercises outside in the yard. !is is done only occasionally, and mostly in warm weather. Often they do meditation or
savâsana in their beds. Guards walk in and out of the classes at intervals to check on the situation and count the number of attendees. !e prison guards are generally very supportive of the Yoga activities, and often express their desire to join in. !ey have sometimes taken instruction separately when they are off duty.

Table 4.  Sentence Lengths
Sentence Length Category Participants Percentage
6 months or less  3  5.8%
> 6 months - 1 year  8 15.4%
> 1 - 2 years  9 17.3%
> 2 - 5 years  8 15.4%
> 5 - 10 years 20 38.4%
> 10 years - life  3  5.8%
Life  1  1.9%LOW REINCARCERATION RATE ASSOCIATED WITH ANANDA MARGA YOGA 47


Analysis
Selecting a Comparison Group!e authors attempted to use the demographics and histories of the Yoga group to select a comparison program group (for example, a Bible study group). However, the
Department of Research and Planning could not identify a program group similar enough to the Yoga group to do a comparison study of reincarceration rates. !e researchers did not think that it would be appropriate to make direct comparisons to the general WCC population. Instead, the author identified a comparison group (N = 131) from the list of inmates who had attended fewer than four classes. Often these were inmates who came primarily for the vegetarian buffet, which was announced over the loudspeaker,
but chose not to regularly participate in the Yoga classes.

To test our hypothesis that regularly attending Yoga classes reduces reincarceration rates, we compared the reincarceration rates of the Yoga group and the group of inmates who attended fewer than four classes.
Determining Reincarceration Rates Data on these participants’ release from prison and subsequent reincarceration was obtained from the DOC Public Information website : http://webapps6.doc.state.nc.us/apps/
offender/search1).
 Analyses of this type are complicated by the fact that some individuals who are reincarcerated are not
ultimately convicted, some individuals who commit crimes are not caught, and some individuals who are convicted are not guilty. Further, difficulties arise because some reincarcerations are based on parole violations (violations of curfew, lack of reporting on time, driving without permission, etc.) and not on actual crimes committed. However, the authors view reincarceration rates as an important practical outcome
for both the inmates and Department of Corrections. !is is not to be confused with the rate of recidivism, which usually means reincarceration within a 3-year period after release. Because we followed inmates as they were released within a set study period, and not for three years following each inmate’s release, it is not possible to accurately extrapolate 3-year recidivism rates from our data. !erefore, our results
should not be directly compared with other published reports of recidivism.

Results
A total of 190 inmates attended at least one Yoga class between January 2002 and January 2007. Of these, 52 completed at least four classes (the Yoga group). Of these, 47 were released from prison. Four (8.5%) were reincarcerated for crimes including larceny from person, drug possession, assault on a female, breaking and entering, and “habitual felon.” !ose reincarcerated committed their offenses a median of 7.5 months (ranging between 2 and 18 months) after release, and had multiple prior incarcerations for various offenses. !e remaining 43 who were released have stayed out of prison for a median of 18 months, with a range of 1 to 58 months. Eleven have been out in society for over three years, and four for over four years.
Of the total of 190 inmates, 131 attended the class less than four times (the comparison group). Of these, 111 were released and 28 were reincarcerated (25.2%) within a median of 12 months, and a range of 6 to 31 months.

New crimes included DWI, drug use, breaking and entering, firearms possession, speeding, and eluding arrest. Only one was for technical probation violation. !e median time post-release for those not re-incarcerated was 15 months, ranging between 4 and 60 months.We used a chi-square test to compare the actual observed frequencies of reincarceration with the frequencies of reincarceration that we would expect if there were no difference in reincarceration rates between the two groups. Under the null hypothesis (no difference between groups), we would expect 9-10 participants of the Yoga group to have
been reincarcerated, and 22-23 members of the comparison group to have been reincarcerated. We found that the actual observed frequencies were significantly different than these expected values, with a chi-square statistic (1, N = 190) of 5.68,  p < 0.025. !is suggests that regular attendance of Yoga classes was associated with lower reincarceration rates.

Discussion
!e results show a strikingly low rate (8.5%) of reincarceration in the group that attended four or more classes. !e comparison group that attended less than four classes showed a much higher rate (25.2%), comparable to that seen in the general population of inmates released from North Carolina prisons. In a North Carolina governmentsponsored study of all inmates released between 1998 and 1999,  4
 22% of those released were re-arrested within one year, 33% within two years, and 41% within three years. A Federal study of prisoners released from facilities in 15 different states in 1992 showed that 67.5% of released prisoners were re-arrested for a new crime within three years of release, with approximately two-thirds occurring within the first year.5 Although these figures cannot be directly compared to ours, due to differences in populations and in the 48 INTERNATIONAL JOURNAL OF YOGA THERAPY  –  No. 18 (2008)methods of collecting reincarceration data, the general trend is instructive.
Limitations of the Study Because the program was set up originally as a service
project and not a research project, the comparison group was not actually a control group, and participation was voluntary, not randomly assigned. It is possible that the lower
reincarceration rate is due to the self-selection of a motivated and committed group of inmates. !e authors were unable to fully explore any selection bias in the two groups because confidentiality issues prohibited access to data that could have allowed a thorough analysis of pre-participation recidivism risk. However, comparison of reincarceration rates to other reports of similar populations suggests that
the reincarceration rate of the Yoga group is indeed lower than is typical.
It is useful to keep in mind that more than 50% of participants in the Yoga group were non-white, and most were single. Both of these demographics predict a high reincarceration rate. !e low reincarceration rate in our study was seen even among felons with long prior prison sentences, who would have also been predicted to be at high risk for reincarceration.

How Yoga Helps
Although we did not measure any variables that could account for the lower observed reincarceration rate, we speculate that our low reincarceration rate may be related to the following aspects of the Yoga program:
1. !e esteem and self-worth generated in the inmates by the care and attention received during the class.
2. !e wisdom imparted during the philosophy classes, including instruction in the eight-limbed path of Yoga.
3. !e sharing and introspection done during the philosophy classes.
4. !e feelings of peacefulness and self-control generated
during the Yoga/meditation classes, and continued outside the classes.
5. !e camaraderie generated during the classes, and the
emphasis on interdependent behavior that is fostered
during the class.
6. !e blue-and-gold certificates of attendance, which recognized the participants’ achievements.
Ananda Marga has been described by its founder
as a “Man-Making Mission.” The entire thrust of our program has been to transform individuals and empower them to make proper choices. By giving them the direct experience of bliss, they can observe the world and participate in it from an entirely different and positive perspective. Although obtaining a GED and training in vocational skills may be of real practical benefit to inmates, it is the authors’ belief that prevention of reincarceration also depends on developing spiritual values and morality.

Conclusions
Yoga and meditation can safely and effectively be taught in prison to a varied population, irrespective of religion or race. Inmates voluntarily participating in four or more Ananda Marga Yoga/meditation classes, along with proper philosophy/sharing sessions, are found to have a lower than expected reincarceration rate, in spite of significant risk factors for recidivism.

References 
1. AHCPR Archived reports, Put Prevention Into Practice and Minnesota
Health Technology Advisory Committee   SAMHSA/CSAT Treatment
Improvement Protocols    TIP 44. Substance Abuse Treatment for Adults
in the Criminal Justice System.
Available at http://www.ncbi.nlm.nih.gov/books/bv.fcgi?highlight=medit
ation&rid=hstat5.section.80442#80553. Accessed January 29, 2008.
2. Murphy, LD.  !e NRF Vipassana recidivism study: final report.
Unpublished manuscript, 2002. Data cited in: Parks GA, Marlatt GA,
Bowen SW, et al. !e University of Washington Vipassana Meditation
Research Project at the North Rehabilitation Facility.  American Jails
Magazine, July/August 2003. Available at: www.prison.dhamma.org/amjarticle.pdf. Accessed January 29, 2008.
3. Bowen S, Witkiewitz K, Dillworth TM, et al. Mindfulness meditation
and substance use in an incarcerated population. Psychology of Addictive
Behaviors. 2006;20(3):343-7.
4. Dawes D, Ebron R, Freeman L, Hevener G, Jones K, Katzenelson
S. Correctional program evaluation: Offenders placed on probation or
released from Prison in fiscal year 1998-99.
Available at:
http://www.nccourts.org/SearchResult.asp?Keyword=recidivism&Usin
g=1&What=1&StdContent=true&Forms=true&PageSize=10.   Article
#10. Accessed on January 29, 2008.
5. Federal Bureau of Justice Statistics Recidivism Study – June 2002.
Summary found at:
http://www.cor.state.pa.us/stats/lib/stats/BJS%20Recidivism%20Study.
pdf. Accessed January 29, 2008. (no author listed). Full article at http://
www.ojp.usdoj.gov/bjs/abstract/rpr94.htm . Accessed April 10, 2008.
Direct correspondence to Pashupati Steven Landau, MD, at
707 Wilkins Street, Smithfield, NC 27577. Tel: (919) 209-
9930. Email: pashupati@bellsouth.net.