Do Education and Employment Programs in Prison Affect Recidivism?

Do Education and Employment Programs in Prison Affect Recidivism?

Introduction

 Decreasing recidivism is critical to slowing the increasing trend in the prison population and its associated costs. Over the past two decades, expenditures on corrections have grown more rapidly than any other spending category of state budgets with the exception of health care. With 1.57 million inmates under State or Federal jurisdiction in 2006 and an average annual growth rate of 1.9% from 2000 to 2005, the criminal justice system expects continued increases in spending (Sabol, Couture, & Harrison, 2007).

 Recidivism is a significant issue when one considers that over 630,000 prisoners are released each year and that more than 95% of all state prisoners will eventually be released from prison. Langan and Levin’s (2002) study of 272,111 prisoners released in 15 states in 1994 showed 5% returning within six months for new offences and a cumulative 25.4% for a new offence within three years. Over half of the prisoners released from prison in 1990 in Iowa returned to prison within ten years. Offenders being returned to prison because of parole violation or re-arrest now account for a much larger proportion of all prison admissions. With ‘three strikes’ legislation, repeat offenders are also being incarcerated for longer periods.

 Having prisoners work and/or participate in education programs addresses two possible causes of incarceration and recidivism, lack of job skills and lack of education. In the US in 2001, 1.4 million people were incarcerated and the majority of these inmates participated in institutional prison jobs, such as laundry and kitchen duties. An additional 85,000 were employed in prison industry programs (Correctional Industries Association, 2001). In 1996, $1.2 billion was spent on inmate programs, which included both educational and non-educational programs. Of the state prisoners incarcerated in 1997, 550,000 inmates had participated in educational programs at some point during their incarceration. Twenty-six per cent of state prison inmates report completing their General Education Diploma (GED) while incarcerated (Harlow, 2003). In spite of the number of people participating in educational programs and the number of dollars allocated to them, little analysis has been carried out on their effectiveness. This paper will look at the question of whether or not educational and employment programs are effective in reducing recidivism rates.

Literature Review

 There has been much prior research done on the role of employment and education programs and their effects on recidivism rates across the country. The first scholarly article that provides theoretical background to this paper is titled Work as a Turning Point in the Life Course of Criminals: A Duration Model of age, employment, and recidivism written by Christopher Uggen (Uggen, 2000).

This article looks to answer the question if work and employment is a turning point for offenders. It looks at three questions about causality, age-dependence, and the timing of criminal recidivism: Does work cause a reduction in crime, or is the association spurious? Do work effects depend on the life-course stage of offenders, or are they uniform across age groups? And when are released offenders at greatest risk of recidivism?

Uggen shows that studies of work effects on crime are prominent in sociological research, and this research has several basic empirical findings. The first is the meaning of work and its implications for crime change at some point during the transition to adulthood (Bachman & Schulenberg, 1993). Second, a high incidence of short jobless spells rather than long-term unemployment characterizes the work histories of criminals (Cook, 1975). Third, even though evidence points to positive effects work has on adult offending, most experimental efforts to reduce crime through employment have had disappointingly small effects (Sampson & Laub, 1990). Uggen points out that crime reduction among older offenders has been observed in employment programs, adolescents benefit far less, raising the question if age has a relationship with crime.

Crime rates rise in the early teen years, peak during the mid to late teens, and decline thereafter. In the life-course camp, they argue that the causes of crime are variable over the life cycle. They feel work effects will vary with the age of the worker, and employment reinforces a non-criminal identity among older offenders, but not among younger ones. On the contrary, Hirschi and Gottfredson posit the “non-interaction hypothesis”, or the “invariance camp”, stating age does not interact with work or anything else in affecting crime (Hirschi & Gottfredson, 1983). They feel work and crime is spurious, and that work is not causally related to crime. The life course studies feels that it isn’t employment by itself that reduces crime, but the stability and commitment associated with work that do reduce crime. The invariance camp counter that stance by stating that people with good work habits would be less likely to commit crime in the first place (Sampson & Laub, 1990).

The author states that the effect of employment is the strongest when recidivism is most likely. In order to determine how and when employment affects recidivism, Uggen uses work effects that correspond to three behavioral models: assignments, participation, and eligibility. Assignment effects are indicators of program effectiveness, and answers if the work experience permanently changes the track of the offender. It encompasses the overall impact of being in the program, regardless of the level of participation. The participation effects refer to the immediate impact of actually working in the job. Eligibility effects fall between assignment and participation effects. These look at whether or not being in the program and having a job available to the offenders is looked at as a safety net that reduces recidivism, but won’t have lasting effects after the program ends (Uggen, 2000).

In order to test the effects, Uggen analyzed data from a large-scale experimental employee program, the National Supported Work Demonstration Project. Over 3000 people with an arrest history were randomly assigned to the control or treatment groups. These participants were criminal offenders, hardcore drug users, and youth dropouts with a history of chronic unemployment. They had to have been incarcerated within the last 6 months. Both groups had many young African American males with less than a high school education and an extensive criminal history.

When testing the data, Uggen chooses age 26 as the cutting point for several reasons. The strongest effects were found for those aged 26 and older, those older than 26 were excluded from large-scale programs, the median age of US inmates is about 26, labor force participation rates peak between ages 25 and 34, and it assured a sample of about 1000 offenders in each age category.

Uggen uses event history analysis as his method. He feels it is better than cross-sectional or panel designs. He states most recidivism studies operate on a “duration clock” that models the time from prison release until parole violation, rearrest, or reconviction.

The results show that aged 26 and under were unaffected by treatment. The numbers show that after 2 years, 55 percent in the control group remained free while 54 percent in the treatment group remained free. Uggen states the difference is neither statistically nor practically significant. The older group, however, showed the experimental group in the work program with 30 percent recidivism rates as opposed to 40 percent in the control group. Uggen concludes that work appears to be a turning point for offenders over 26 years old. Offenders who are provided even marginal employment opportunities are less likely to reoffend. Older offenders given jobs are less likely to reoffend than those of comparable age who were not provided the same opportunities.

Some researchers feel that criminal propensity embeds itself in a person in early childhood and is almost impossible to change thereafter. This study provides clear evidence that a modest work experience later in life can change the behavior of older criminals.

 The second scholarly article that provides theoretical background to this paper is entitled College Education and Recidivism: Educating Criminals Is Meritorious by Dennis J. Stevens and Charles S. Ward. In this article, the authors posit that, when incarcerated, the general public feels less threatened because the inmates are no longer an immediate threat. However, once released, without education and employment programs, the inmates are returning unprepared, uneducated, and usually bitter, which could be a higher threat to the public, and increases recidivism rates (Stevens, The impact of time-served and regime on prisoners’ anticipation of crime, 1997)

 While the authors state that there is doubt on the effectiveness of prison treatment programs, they do show one point as clear: the average educational level achieved prior to incarceration by inmates is far below the average level achieved by free individuals, and many inmates are unemployed prior to imprisonment (Bureau of Justice Statistics, 1995). The authors believe that inmates who earn a quality college degree while incarcerated tend to lead law-abiding lives once released more often than inmates who have not earned a college degree while incarcerated (Stevens & Ward, College Education and Recidivism: Educating Criminals is Meritorious, 1997). The role of correctional education, according to the authors, is critical in relieving boredom of prison time, give inmates a better understanding of society, and alter behavior preventing costly reincarceration.

 There is controversy among researchers when dealing with correctional education and recidivism rates. Some argue that there is no conclusive evidence correlating correction education and reduced recidivism, and some go as far as stating that nothing can alter criminally violent behavior. Opponents argue that criminal tendencies learned on the outside cannot be “unlearned” on the inside. They also feel that offenders gave up their rights to education when they took away the rights of others in the crimes they committed. However, a total of 97 articles published between 1969 and 1993 examined the relationship between correctional education and recidivism levels. Of those 97 articles, 83 (85%) reported documented evident of recidivism control through correctional education, while only 14 (15%) reported a negative relationship (Ryan & Mauldin, 1994).

 The authors send letters to every Chief Adult Correctional Education Administrator in each of the fifty states, and each North Carolina public administrator who might have some information about inmate education. All degree granting institutions participating in NC inmate education were also asked to submit data about inmates who obtained degrees from their institutions. Also, criminal records for a ten-year period of nondegree-inmates and degree-inmates were examined in NC. 320 inmates earned 373 post secondary degrees in NC prisons over a 10-year period from 1981 to 1991. 60 of these inmates were released in 1991 and traced for a three-year period. Of these 60 NC released inmates, 5% were returned to prison for criminally violent offenses in the 36 months following their release. Offenders who earned four-year degrees were not reincarnated during the three-year period after their release, and all but one of these individuals found employment related to their degree. On the other hand, 40% of the general population in NC was reincarcerated within three years of their release.

 The third scholarly article that give theoretical background to this paper is entitled Studies of Correctional Education Programs by Barbara Wade. She states that the purpose of her study is to examine educational program evaluations and the validity of research focused on various prison education programs (Wade, 2007). The premise of correctional education is threefold. First, as inmates gain knowledge and skills, they should be qualified for employment upon their release into the community; second, education in prison should serve as a mechanism that enables inmates to learn to think about responsibility; and last, this combination should make it less likely that they will return to prison (Messemer & Valentine, 2004)

 Wade looks at 13 different research articles to make her review. Most of the studies conducted in these articles were quantitative. She states that it may be difficult to measure recidivism because of the many definitions and characteristics used in research. For example, if the purpose of the research is to assess whether an educational program has an effect on the amount of time offenders remain out of prison, then recidivism is defined as reincarceration (Brewster & Sharp, 2002). Other studies define recidivism vaguely as an inmate being returned to custody, returned to a specific county facility, or returned to prison.

 Wade highlights the work done by Messemer and Valentine. In their study, they measured the inmates’ learning gains through the use of pre- and posttests. The results indicated that inmates who participated made significant learning gains in reading, math, and language (Messemer & Valentine, 2004). The majority of participants in these studies had failed academically prior to their incarceration. They showed that students who were previously considered at risk could become successful in the classroom. This indicates that perhaps the reason they failed was because most offenders went to schools in communities infested with crime and poverty, and for the first time in prison were achieving academic success because the educational programs were structured to meet their needs (Wade, 2007).

 Wade concludes that her literature review supports the proposition that inmates who participate in various types of education programs are less likely to revert back to criminal behavior upon release.

Current Study/Hypothesis

 Based on the literature review, I feel there is substantial information pointing to the positive relationship between education and employment programs and a reduced recidivism rate.

Data

 There are several ways of collecting data for this study. A broad way is to follow the lead of Stevens and Ward (1997) by sending letters to each Chief Adult Correctional Education Administrator in each of the fifty states, asking for their data at their respective facilities. In that study, it was found that only 8 states had conducted studies on the correctional education and recidivism question. Ten years later, it would be interesting to see if the data has changed for those same 8 states, and also to see if more states have become involved in measuring this data.

 A second way to collect data will be more focused on one particular area of the country. We can collect data from a number of prisoners released from the Ohio prison system, for example, in 1992. Prisoners must have served at least one year in order to guarantee the have been “in the system” long enough to benefit from educational programs. One group will not have participated in any educational program and will be the control group. The second group will have participated in an education program. We can follow these prisoners for a period of 5 to 10 years and measure and compare the level of recidivism. The challenge will be finding the inmates once they have been released, as some inmates will not want to be found or participate. It can also be costly in trying to locate these inmates, especially at the 10 year mark, where there might not be an easy trail to find the men. Also, we must take into consideration any inmates who have died during the trial and account for them.

Measurement

 When measuring the data, a few methods can be used, according to Wade (2007). These methods include t-tests, correlation coefficients, and descriptive statistics. When looking to see if educational programs even help increase inmates test scores, before being released, Messemer and Valentine (2004) used pre- and posttest paired t-tests. In their study, most inmates showed statistical significance regarding higher scores in math, language, and reading. This was found with an alpha level of .05.

 In 1998, a study was conducted using correlation coefficients (Koski, 1998). In this study, Koski stated his results were statistically significant with only an r value being reported. According to Wade (2007), a proper way to analyze the statistical significance of a correlation coefficient with only an r value is to convert the r value into a z value, and compare the observed z value to a critical z value at a specific alpha level.

 Other studies examined how recidivism rates vary with the completion of educational programs by using descriptive statistics. While this provides policy makers with nice visual graphics, it does not determine if there is a statistically significant relationship between educational programs and recidivism rates (Wade, 2007). It is important to note that if using descriptive statistics, it is wise to include other methods of data analysis as well to prove significance.

Analytic Strategy

 As mentioned in the measurement section, it is nice to use descriptive statistics to visually show reults, as that can be crucial in earning funding from lawmakers for educational programs in prisons. However, while that is visually pleasing, it does not show statistical significance. I would propose looking at paired t-tests to prove educational programing is helping inmates learn while in prison. Also, I would look at the data and compare the control group to the educated group, and longitudinally evaluate recidivism rates across a 10 year period.

Limitations and Directions for Future Research

 I believe the literature shows a positive relationship with educational and employment programs and recidivism rates. The studies clearly show some support that inmates who participate in various types of education programs are less likely to revert back to criminal behavior upon their release. Future researchers should also look at the learning gains the inmates achieve while in prison, as this is a foundation for measuring recidivism rates. However, even with education and employment programs, some offenders are unaffected and are rearrested. I feel future research should look to see what factors those inmates have in common, and if an educational and employment program was the right choice for them in the beginning, as they might have been more suited for vo-tech programing or something of that sort.

Works Cited

• Bachman, J. G., & Schulenberg, J. (1993). How part time work intensity relates to drug use. Developmental Psychology , 29, 220-235.
• Brewster, D., & Sharp, S. (2002). Educational programs and recidivism in Oklahoma: Another Look. The Prison Journal , 62 (2), 314-334.
• Bureau of Justice Statistics. (1995). Sourcebook of criminal justice statistics.
• Cook, P. (1975). The Correctional Carrot: Better Jobs for Parolees. Policy Analysis , 1, 11-54.
• Correctional Industries Association. (2001). CIA Directory .
• Harlow, C. (2003). Education and Correctional Populations. US Department of Justice .
• Hirschi, T., & Gottfredson, M. (1983). Age and the Explanation of Crime. American Journal of Sociology , 89, 552-584.
• Koski, D. (1998). Vocational Education in Prison. Journal of Offender Rehabilitation , 17 (4), 151-164.
• Langan, P., & Levin, D. (2002). Recidivism of Prisoners released in 1994. Federal Sentencing Reporter , 15 (1), 58-65.
• Messemer, J., & Valentine, T. (2004). The learning gains of male inmates participating in a basic skills program. Adult Basic Education , 14 (2), 67-89.
• Ryan, T., & Mauldin, B. (1994). Correctional education and recidivsm: An historical analysis.
• Sabol, W., Couture, H., & Harrison, P. (2007). Prisoners in 2006. Bureau Of Justice Statistics .
• Sampson, R. J., & Laub, J. H. (1990). Crime and Deviance over the Life Course. American Sociologial Review , 55, 609-627.
• Stevens, D. (1997). The impact of time-served and regime on prisoners’ anticipation of crime. The Howard Journal of Criminal Justice .
• Stevens, D., & Ward, C. (1997). College Education and Recidivism: Educating Criminals is Meritorious. Journal of Correctional Education , 48 (3).
• Uggen, C. (2000). Work as a turning point in the life course of criminals. American Sociological Review , 65 (4), 529-546.
• Wade, B. (2007). Studies of Correctional Education Programs. Adult Basic Education & Literacy Journal , 1 (1), 27-31.