INTRODUCTION
The research literature attributes great importance to integrating released prisoners into the labour market. Engaging in legitimate work in the labour market allows released prisoners to make a respectable living and may neutralize or at least reduce their motivation to take part in criminal activity (Weisburd et al. Reference Weisburd, Hasisi, Shoham, Haviv and Zelig2017). Scholars further note that being employed in the labour market enhances released prisoners’ self-esteem, provides them with a positive social network, a sense of belonging to the community, and conventional lifestyle, going as far as to serve as informal social control for ex-offenders (Hagan and Dinovitzer Reference Hagan and Dinovitzer1999; Peled-Laskov, Shoham, and Cojocaru Reference Peled-Laskov, Shoham and Cojocaru2019; Shoham and Peled-Laskov Reference Shoham and Peled-Laskov2022; Uggen Reference Uggen1999, Reference Uggen2000; Visher, Winterfield, and Coggeshall Reference Visher, Winterfield and Coggeshall2005; Western, Kling, and Weiman Reference Western, Kling and Weiman2001; Wilson, Gallagher, and MacKenzie Reference Wilson, Gallagher and MacKenzie2000). According to rehabilitation approaches that focus on desistance, genuine rehabilitation is a multiphase process leading to transformation in lawbreakers’ lives and self-perception, with desistance being a process of change performed by lawbreakers themselves, at times with the help of experts acting on behalf of the law enforcement system (Ward and Maruna Reference Ward and Maruna2007). Integrating into the employment world plays a key role in this multiphase process of change in perception and conduct that affects various aspects of former prisoners’ lives (Berg and Huebner Reference Berg and Huebner2011; Shoham and Peled-Laskov Reference Shoham and Peled-Laskov2020; Uggen and Staff Reference Uggen and Staff2001). To increase released prisoners’ employability and their chances of earning a legitimate living following their release from prison, correctional facilities in many Western countries offer professional training programmes of various kinds, such as educational teaching, vocational training, prison industries and employment services (Bouffard, MacKenzie, and Hickman Reference Bouffard, MacKenzie and Hickman2000; Davis et al. Reference Davis, Bozick, Steele, Saunders and Miles.2013; Lawrence et al. Reference Lawrence, Mears, Dubin and Travis2002). The types of training offered to prison inmates vary according to participants’ objectives and interests, prison decision-makers budget considerations, the professions in demand by society’s labour market, and the availability of facilitators or teaching staff. Although the characteristics of these various types of training programmes differ, their main goal is to provide prisoners with a positive routine, knowledge, tools, working skills and habits that would help them integrate into a labour market profession following their release from prison. Furthermore, these programmes aim to instill a sense of commitment, responsibility and motivation in prisoners to integrate into the labour market.
Most of the available literature that centres on examining the effectiveness of training programmes makes reference primarily to recidivism as an outcome measure (Maltz Reference Maltz1984; Shoham et al. Reference Shoham, Zelig, Hesisi, Weisburd and Haviv2017; Walk et al. Reference Walk, Haviv, Hasisi and Weisburd2021), although it is often an indirect variable, while the direct variable that should, perhaps, be examined in such studies pertains to former prisoners’ post-release employment (Walk et al. Reference Walk, Haviv, Hasisi and Weisburd2021). Moreover, for the most part, research literature focuses less on the effects of training programmes that emphasize prisoners’ process of entry into the labour market and all of its associated challenges (Shoham et al. Reference Shoham, Zelig, Hesisi, Weisburd and Haviv2017).
The present study seeks to fill this gap by examining the effectiveness of an employment training programme offered to prisoners by the Israel Prison Service (IPS), which focuses on helping participants re-enter the labour market. The programme aims to help prisoners write suitable résumés, prepare them for job interviews, and introduce them to personal budget management, business administration, etc. The uniqueness of this study lies in its ability to examine the programme’s effectiveness by measuring recidivism and other outcome measures, such as income, tax payment, need for welfare services that are considered positive progress, and referral to out-of-home frameworks.
Barriers and Difficulties in the Reintegration of Released Prisoners into the Labour Market
Significant concerns include barriers and difficulties reintegrating released prisoners into the labour market. Despite the acknowledged importance of helping prisoners find employment upon release (Newton et al. Reference Newton, Day, Giles, Wodak, Graffam and Baldry2018; Sampson and Laub Reference Sampson, Laub and Thornberry1997; Uggen Reference Uggen1999; Uggen and Staff Reference Uggen and Staff2001; Visher et al. Reference Visher, Winterfield and Coggeshall2005; Wilson Reference Wilson1997; Wilson et al. Reference Wilson, Gallagher and MacKenzie2000), numerous studies have highlighted a multitude of obstacles that hinder their successful reintegration into the workforce (Holzer, Raphael, and Stoll Reference Holzer, Raphael and Stoll2003; MacKenzie Reference MacKenzie2008; Newton et al. Reference Newton, Day, Giles, Wodak, Graffam and Baldry2018; Shoham and Peled-Laskov Reference Shoham and Peled-Laskov2020). These challenges extend beyond legal restrictions, such as prohibiting individuals with criminal convictions from assuming government or public positions in many countries (Bloom Reference Bloom2006; Holzer et al. Reference Holzer, Raphael and Stoll2003; Pager Reference Pager2003). Additionally, the personal and personality traits of former prisoners further complicate their employment prospects. Incarcerated individuals often exhibit low cognitive capabilities, limited education and literacy, and inadequate reading and writing skills (Bloom Reference Bloom2006; Davis et al. Reference Davis, Steele, Bozick, Williams, Turner, Miles, Saunders and Steinberg.2014; Lawrence et al. Reference Lawrence, Mears, Dubin and Travis2002; Saylor and Gaes Reference Saylor and Gaes1997; Silver and Nedelec Reference Silver and Nedelec2018). In the labour market context, released prisoners typically lack a substantial occupational history, marketable skills and professional competencies (Cullen and Gendreau Reference Cullen and Gendreau2000; Davis et al. Reference Davis, Steele, Bozick, Williams, Turner, Miles, Saunders and Steinberg.2014; Gaes et al. Reference Gaes, Flanagan, Motiuk and Stewart.1999; Lawrence et al. Reference Lawrence, Mears, Dubin and Travis2002; Petersilia Reference Petersilia2003; Visher et al. Reference Visher, Winterfield and Coggeshall2005; Western et al. Reference Western, Kling and Weiman2001). Their work history often includes a pattern of unstable employment (National Institute of Justice 2014; Petersilia Reference Petersilia2003; Visher et al. Reference Visher, Winterfield and Coggeshall2005; Western et al. Reference Western, Kling and Weiman2001), low self-esteem and negative expectations regarding their employability (Saylor and Gaes Reference Saylor and Gaes1997). Consequently, released prisoners frequently lack the expertise and knowledge to successfully integrate into labour markets (Visher et al. Reference Visher, Winterfield and Coggeshall2005).
Recent in-depth analyses conducted by social work and critical criminology scholars have introduced a more nuanced perspective regarding these challenges. They have shed light on the stigmatizing and exclusionary consequences associated with re-entry programmes (Halushka Reference Halushka2020; Hinton and Cook Reference Hinton and Cook2021; Miller Reference Simon2021; Simon Reference Thompson2010; Thompson Reference Thompson2012). These scholars argue that although re-entry programmes aim to assist formerly incarcerated individuals, they may unintentionally perpetuate societal stereotypes and inequalities. For example, Halushka (Reference Halushka2020) suggests that these programmes might inadvertently reinforce negative preconceptions, making it more difficult for ex-offenders to secure employment. Hinton and Cook (Reference Hinton and Cook2021) delve into systemic issues within re-entry programmes, revealing how they can marginalize specific individuals or communities, thereby impeding their access to vital support.
Furthermore, Miller (Reference Miller2021) underscores the broader societal implications of re-entry programmes, linking them to social justice and equity matters. Simon (Reference Thompson2010) critically assesses the policies and practices within the criminal justice system that make an impact on re-entry, advocating for more comprehensive and equitable approaches. Thompson (Reference Thompson2012) offers insights into the experiences of marginalized populations within the criminal justice system, shedding light on how re-entry programmes can either provide meaningful support or fall short in assisting these individuals.
In light of these critical perspectives, it is evident that addressing the multifaceted challenges that released prisoners face in their quest for labour market integration requires a more comprehensive and equitable approach. Re-entry programmes must consider not only the legal and educational barriers but also the societal attitudes and systemic inequalities that can hinder successful reintegration.
Despite the great importance attributed to prisoners’ integration into the labour market (Newton et al. Reference Newton, Day, Giles, Wodak, Graffam and Baldry2018; Sampson and Laub Reference Sampson, Laub and Thornberry1997; Uggen Reference Uggen1999; Uggen and Staff Reference Uggen and Staff2001; Visher et al. Reference Visher, Winterfield and Coggeshall2005; Weisburd et al. Reference Weisburd, Hasisi, Shoham, Haviv and Zelig2017; Wilson Reference Wilson1997; Wilson et al. Reference Wilson, Gallagher and MacKenzie2000), studies reveal a large number of barriers and restrictions, impeding the prisoner population’s integration into the labour market following their release (Holzer et al. Reference Holzer, Raphael and Stoll2003; MacKenzie Reference MacKenzie2006; Newton et al. Reference Newton, Day, Giles, Wodak, Graffam and Baldry2018; Shoham and Peled-Laskov Reference Shoham and Peled-Laskov2020). For instance, in many countries, offenders convicted of criminal offences have been prohibited by law from assuming government or public positions (Bloom Reference Bloom2006; Holzer et al. Reference Holzer, Raphael and Stoll2003; Pager Reference Pager2003).
Furthermore, prisoners have many personal and personality traits that diminish their chances of being hired. In this context, prisoners typically tend to have low cognitive capabilities, low levels of education and literacy, as well as lesser reading and writing knowledge (Bloom Reference Bloom2006; Davis et al. Reference Davis, Steele, Bozick, Williams, Turner, Miles, Saunders and Steinberg.2014; Lawrence et al. Reference Lawrence, Mears, Dubin and Travis2002; Saylor and Gaes Reference Saylor and Gaes1997; Silver and Nedelec Reference Silver and Nedelec2018). In the occupational context, released prisoners characteristically have sparse occupational history, few working skills, a lack of professional competencies (Cullen and Gendreau Reference Cullen and Gendreau2000; Davis et al. Reference Davis, Steele, Bozick, Williams, Turner, Miles, Saunders and Steinberg.2014; Gaes et al. Reference Gaes, Flanagan, Motiuk and Stewart.1999; Lawrence et al. Reference Lawrence, Mears, Dubin and Travis2002; Petersilia Reference Petersilia2003; Visher et al. Reference Visher, Winterfield and Coggeshall2005; Western et al. Reference Western, Kling and Weiman2001), a history of unstable employment (National Institute of Justice 2014; Petersilia Reference Petersilia2003; Visher et al. Reference Visher, Winterfield and Coggeshall2005; Western et al. Reference Western, Kling and Weiman2001), low perceived self-esteem, negative expectations about their ability to get hired, and so on (Saylor and Gaes Reference Saylor and Gaes1997). As a result, released prisoners lack the expertise and knowledge to integrate into the labour market (Visher et al. Reference Visher, Winterfield and Coggeshall2005).
Given released prisoners’ difficulties when attempting to integrate into the labour market, scholars (Visher et al. Reference Visher, Winterfield and Coggeshall2005), as well as prison policymakers and decision-makers (Newton et al. Reference Newton, Day, Giles, Wodak, Graffam and Baldry2018), have raised the need for effective intervention that would increase prisoners’ chances of integrating into the labour market. Moreover, indeed, over the past several decades, many correctional and training programmes have been designed, among other things, to provide inmates with the skills, competencies and knowledge required to integrate into the labour market following their release from prison, as well as persist at their workplace (Bloom Reference Bloom2006; Bouffard et al. Reference Bouffard, MacKenzie and Hickman2000; Lawrence et al. Reference Lawrence, Mears, Dubin and Travis2002; Visher et al. Reference Visher, Winterfield and Coggeshall2005). It is currently common to view the period of incarceration as an opportunity for the law enforcement system to instill working skills and competencies in prisoners while providing them with an education and preparing them for workplace placement following their release (Visher et al. Reference Visher, Winterfield and Coggeshall2005), thereby lowering the risk of these released prisoners’ return to active engagement in crime in the future (Bouffard et al. Reference Bouffard, MacKenzie and Hickman2000; Bushway and Reuter Reference Bushway, Reuter, Farrington, MacKenzie, Sherman and Welsh2002; Davis et al. Reference Davis, Steele, Bozick, Williams, Turner, Miles, Saunders and Steinberg.2014; Newton et al. Reference Newton, Day, Giles, Wodak, Graffam and Baldry2018; Visher et al. Reference Visher, Winterfield and Coggeshall2005). Visher et al. (Reference Visher, Winterfield and Coggeshall2005) argue that “A good job not only provides the means for basic survival, but also is a key element in rebuilding self-esteem, attachment to a conventional lifestyle, and a sense of belonging in the community.”
Employment and the integration into it provide released prisoners with the opportunity to become a productive member of the community while giving them occupational experience and allowing them to earn a living and provide for their families. Being employed also enables routine and the establishment of positive social ties (Laub and Sampson Reference Laub and Sampson2001; Maruna Reference Maruna2001; Sampson and Laub Reference Sampson, Laub and Thornberry1997). Integration into the labour market could serve as a promising starting point in ex-offenders’ process of re-entrance to society since it is regarded as significant “routine activity” for most released prisoners who seek to find a legitimate source of income (Bucklen and Zajac Reference Bucklen and Zajac2009; Bushway and Apel Reference Bushway and Apel2012; LeBel et al. Reference LeBel, Burnett, Maruna and Bushway2008). The result is a former prisoner who successfully integrates into the labour market and can thus play a positive social role with underlying pro-social values. Moreover, for an ex-offender, work stability symbolizes the transition toward a conventional, crime-free life.
Prison-Based Vocational Training Programmes
Prisons in many Western countries offer various kinds of vocational training and employment programmes (Bouffard et al. Reference Bouffard, MacKenzie and Hickman2000), such as educational instruction, vocational training, prison industry and employment services (Lawrence et al. Reference Lawrence, Mears, Dubin and Travis2002). The training programmes offered by prisons differ from one another in numerous aspects. For instance, some prisons allow inmates to choose whether or not to participate in vocational training programmes, while others force prisoners to partake in them (Davis et al. Reference Davis, Bozick, Steele, Saunders and Miles.2013). In addition to prison-based training programmes, some vocational training is offered as part of work release programmes, in which prisoners work in and are supervised by institutions and organizations outside prison walls (Bouffard et al. Reference Bouffard, MacKenzie and Hickman2000; Glaze and Parks Reference Glaze and Parks2012; National Institute of Justice 2014). The cooperation between correctional facilities and external organizations provides prisoners with valuable work experience, improving their employment options following their release (Bouffard et al. Reference Bouffard, MacKenzie and Hickman2000). Such programmes are designed to help prisoners get hired by workplaces following their release and provide them with sequential therapy – including social, mental and medical services – in order to facilitate their coping with the various difficulties associated with reintegration into the community (Lawrence et al. Reference Lawrence, Mears, Dubin and Travis2002). There are some differences between the various vocational training programmes’ characteristics, yet their main aim is to provide prisoners with the positive routine, knowledge, tools, skills and work habits that would help prisoners integrate into a profession in the labour market following their release (Bouffard et al. Reference Bouffard, MacKenzie and Hickman2000; Lawrence et al. Reference Lawrence, Mears, Dubin and Travis2002; National Institute of Justice 2014; Visher et al. Reference Visher, Winterfield and Coggeshall2005). These programmes also aim to instill in prisoners a sense of commitment, responsibility and motivation to integrate into the labour market (Gaes et al. Reference Gaes, Flanagan, Motiuk and Stewart.1999). To this end, prison-based vocational training programmes present prisoners with the opportunity to gain experience, work habits and valuable knowledge in various professions, such as computers (Gaes et al. Reference Gaes, Flanagan, Motiuk and Stewart.1999), hairdressing, car mechanics, building maintenance, electricity, painting, plumbing, food services, cooking, landscaping, custodial maintenance, upholstery, welding, heating, air-conditioning (National Institute of Justice 2014), carpentry, electronic services, art and printing, construction, graphics, and so on (Lawrence et al. Reference Lawrence, Mears, Dubin and Travis2002).
Economic and Social Outcome Measures beyond Reduced Recidivism Rates
This study investigates the effectiveness of prison-based vocational training programmes, considering a diverse range of outcome measures. While the general recidivism rate (released prisoners’ rearrest or reincarceration) and post-release employment are commonly used indicators (Aos, Miller, and Drake Reference Aos, Miller and Drake2006; Roberts and Camasso Reference Roberts and Camasso1990; Seiter and Kadela Reference Seiter and Kadela2003; Torrey et al. Reference Torrey, Mueser, McHugo and Drake2014; Wilson et al. Reference Wilson, Gallagher and MacKenzie2000), we also explore broader measures encompassing academic achievements (Adams et al. Reference Adams, Bennett, Flanagan, Marquart, Cuvelier, Fritsch and Burton1994; Davis et al. Reference Davis, Bozick, Steele, Saunders and Miles.2013), prisoners’ behaviour while incarcerated (Adams et al. Reference Adams, Bennett, Flanagan, Marquart, Cuvelier, Fritsch and Burton1994; Davis et al. Reference Davis, Bozick, Steele, Saunders and Miles.2013; French and Gendreau Reference French and Gendreau2006), perceived self-esteem, life satisfaction, ability to afford housing (Torrey et al. Reference Torrey, Mueser, McHugo and Drake2014), post-release substance abuse, traffic offences, intelligence quotient (IQ) test scores and changes in marital status (Adams et al. Reference Adams, Bennett, Flanagan, Marquart, Cuvelier, Fritsch and Burton1994). By considering this wide array of indicators, our study aims to comprehensively assess the impact and effectiveness of these vocational training programmes. The benefit of prison-based employment and training programmes is directly linked to one other key outcome measure – post-release employment – which includes various aspects such as employment rates (Drake Reference Drake2003; Smith et al. Reference Smith, Bechtel, Patrick, Smith and Wilson-Gentry2006; Evans and Koenig Reference Evans and Koenig2011; Wilson et al. Reference Wilson, Gallagher and MacKenzie2000), time taken to find work following release (Drake Reference Drake2003; Smith et al. Reference Smith, Bechtel, Patrick, Smith and Wilson-Gentry2006), the ability to keep a job for an extended period and income (Drake Reference Drake2003; Smith et al. Reference Smith, Bechtel, Patrick, Smith and Wilson-Gentry2006). For example, a meta-analysis (Davis et al. Reference Davis, Bozick, Steele, Saunders and Miles.2013), based on the outcomes of eight prison-based vocational training programmes, found that the chances of released prisoners who had participated in the training programmes finding work were more than double that of released prisoners who did not participate in vocational training programmes.
However, it is worth noting that while previous research on this topic has received criticism for its methodological quality (Bloom Reference Bloom2006; Davis et al. Reference Davis, Bozick, Steele, Saunders and Miles.2013; Lawrence et al. Reference Lawrence, Mears, Dubin and Travis2002; Newton et al. Reference Newton, Day, Giles, Wodak, Graffam and Baldry2018), our study seeks to distinguish itself by addressing some key limitations that have been identified in the existing literature. One of the primary limitations of prior studies assessing vocational training programmes has been the issue of “selection bias” (Bloom Reference Bloom2006; Wilson et al. Reference Wilson, Gallagher, Coggeshal and MacKenzie1999, Reference Wilson, Gallagher and MacKenzie2000), stemming from potential differences in characteristics between the study and comparison groups (Kim and Clark Reference Kim and Clark2013; Wilson et al. Reference Wilson, Gallagher, Coggeshal and MacKenzie1999, Reference Wilson, Gallagher and MacKenzie2000). Additionally, our research takes a novel approach by considering the significance of aligning training programmes with prisoners’ criminogenic needs (Cullen and Gendreau Reference Cullen and Gendreau2000; Newton et al. Reference Newton, Day, Giles, Wodak, Graffam and Baldry2018), adapting them to meet labour market requirements (Hawley, Murphy, and Souto-Otero Reference Hawley, Murphy and Souto-Otero2013; Shoham and Peled-Laskov Reference Shoham and Peled-Laskov2022), providing formal graduation diplomas (Erisman and Contardo Reference Erisman and Contardo2005; Passarell Reference Passarell2013) and offering them near the time of release (Newton et al. Reference Newton, Day, Giles, Wodak, Graffam and Baldry2018). These unique aspects of our study distinguish it from prior research in the field.
The IPS-Based Training Programme
The IPS policymakers have recognized that prisoners should not only be given vocational training but also the kind of training that would prepare them for functioning well within the labour market and provide them with skills such as effective interpersonal communication, time management and problem-solving (Shoham et al. Reference Shoham, Zelig, Hesisi, Weisburd and Haviv2017). Moreover, toward the first decade of the twenty-first century, the IPS began to create courses on topics such as personal budget management, small business management, business entrepreneurship, mediation and computerized warehouse management. During them, prisoners are taught how to write their résumés, pass a job interview, and interact with their superiors or co-workers (Hasisi et al. Reference Hasisi, Weisburd, Shoham, Haviv, Zelig and Kovalsky2018).
There are designated courses in preparation for the labour market adapted to the courses offered by the Israeli Ministry of Economy on these topics. The IPS believes that vocational training alone does not suffice, as a person should also be taught how to adjust to the employment world and dynamic requirements while learning how to interact with superiors and co-workers. Participants may be dismissed from these courses if they fail to follow the IPS code of conduct, for instance, by skipping class unjustifiably, lacking discipline or motivation, and having an insufficiently low level of learning. Prisoners are dismissed by the vocational training officers after being called in for a talk. The prison warden is then notified of the dismissal, recorded alongside the reasons that led to it in the prisoner’s personal file. The courses are taught by external teachers via a tender issued by the IPS and take place in the IPS training centres in each criminal correctional facility. They last between 30 and 50 hours, allowing prisoners to begin and complete a course within one month and move directly on to the follow-up course. Courses open with a minimum number of 15 participants. Admission criteria are the ability to read and write, the approval of both intelligence and security officers, recommendations from both the social worker and wing commander, the prisoner’s request to participate in the course, and passing an admission interview.
As mentioned, the present paper will focus on this one kind of vocational training programme preparing prisoners for the employment world in Israeli prisons and is based on a quasi-experimental study employing a strong methodology that examines various outcome measures beyond the measure of recidivism and monitors ex-offenders for up to five years following their release from prison. We hypothesize that prisoners who participated in the vocational training programmes known as “preparation for the employment world” will be less likely to return to crime than similar, non-participant prisoners. The study further hypothesizes that prisoners who participated in employment world preparatory courses will exhibit more positive results on other outcome measures, such as employment, income, tax-paying and need for welfare services, compared to similar prisoners who did not participate in such programmes.
METHOD
Our study used a comprehensive dataset comprising 57,783 Israeli residents released from IPS facilities between 2004 and 2012. This dataset, obtained from the IPS data system known as “Tzohar”, provided a wealth of information on various aspects. Regarding sociodemographic attributes, we collected data on factors such as new immigrant status, marital status, nationality (Jewish), age and years of schooling. Additionally, the dataset included details about prisoners’ criminal backgrounds, encompassing variables like the number of prior incarcerations, age at the first incarceration, classification as violent by law and by IPS standards, sentence duration in months, involvement in sexually based offences, offences against human life, violent offences, property offences and drug-related offences. We also documented the year of release and indicators of alcohol and drug abuse. Regarding data access, it is important to clarify that we obtained this dataset as part of a research grant awarded to our team from the Hebrew University of Jerusalem, led by the authors, in collaboration with a team at Ashkelon College, led by Professor Efrat Shoham. The grant, totalling 1,000,000 New Israeli Shekels (approximately 275,000 US dollars), was specifically allocated for evaluating various rehabilitation programmes within the IPS. The dataset was made available to our research team for this study, and we adhered to ethical and legal guidelines for its utilization.
Of the 57,783 prisoners included in the data file, 2,519 participated in some vocational training programme; of them, 758 participated in “employment world” training (small business management, preparation for the employment world, business entrepreneurship, personal budget management and business budget management).
The comparison group for the training programme was comprised of the 57,783 prisoners contained in the data file. To form the comparison group for the “employment world” training programme, 1,760 prisoners who had participated in another training programme were filtered out of the data file to maintain the clean “employment world” programme-only effect. Once omitted, the comparison group consisted of 55,264 prisoners.
Propensity score matching (PSM) matched the comparison group to the prisoners who participated in the “employment world” programme. This method was chosen to minimize participant selection bias and ensure that members of the comparison group would be as similar to those of the study group as possible, except that they were not themselves part of the participants’ group (Jordan Reference Jordan2012). The first phase of calculating the propensity score was selecting the variables to calculate the probability of participating in the programme. These variables included prisoners’ sociodemographics, alcohol abuse, incarceration characteristics, criminal history, profile and characteristics while in prison. Propensity scores were calculated using a statistical model that predicts propensity based on subjects’ background characteristics. Each prisoner received a value between 0 and 1, with 0 expressing zero probability and 1 expressing a certain probability of participating in the programme.
Once each prisoner’s probability of being included in a training programme was calculated, every prisoner who had indeed participated in the programme was “matched” with another prisoner whose score was the most similar to his (his “twin”) out of the group of non-participant prisoners. Since the groups of prisoners who participated in the training programmes were sufficiently large, the “best match” method was used, whereby each study participant was matched with a single subject from the non-participant group (the single match approach). For each subject in the group of vocational training programme participants, a prisoner with the closest propensity score was selected from the comparison group. A 0.01 calliper was used for the selection, meaning that the match was only made if the comparison group subject’s score was within a range of 1% of the participation probability of the subject in the participant’s group. Once the matched sample had been completed, a test was carried out to ensure that the propensity score approach resulted in balanced samples and that no consistent differences could be found between the groups in the variables used for the selection process.
Matching the Comparison Group to the Employment World Programme Participants
The group of prisoners who participated in the “employment world” programme consisted, prior to matching, of 758 prisoners with an average “matching” score of 0.04 (standard deviation [SD] = 0.005, minimum = 0.00, maximum = 0.59). The comparison group of prisoners who did not participate in the programme consisted, prior to matching, of 55,264 prisoners with an average propensity score of 0.01 (SD = 0.02, minimum = 0.00, maximum = 0.88). Following matching, the training group consisted of 757 prisoners with an average propensity score of 0.04 (SD = 0.05, minimum = 0.00, maximum = 0.59), and the comparison group too consisted of 757 prisoners with an average propensity score of 0.04 (SD = 0.05, minimum = 0.00, maximum = 0.59).
Table 1 compares the prisoners who participated in the “employment world” programme and the comparison group before and after matching. The significance of the differences between them was tested using the χ 2 test for nominal variables and the t test for continuous variables.
Comparison of Group of Prisoners Who Participated in the “Employment World” Programme and Comparison Group, Before and After, Based on Propensity Scores Matching
* p < 0.05, ** p < 0.01, *** p < 0.001.
Differences existed between the treatment and control groups in 10 variables: sentence duration; against human life offence; violence by law; property offence; drug-related offence; sexually based offence; nationality; number of years of schooling; drug abuse; and alcohol abuse. These disparities encompassed various dimensions, including sociodemographic attributes like marital status and nationality, where the control group had a higher percentage of married individuals and a greater representation of Jewish nationality.
Additionally, differences were observed in criminal background and profile, with the control group showing a slightly higher average number of prior incarcerations. Notable variations in incarceration characteristics included differences in the percentage of individuals incarcerated for violent offences by law and average sentence duration. Disparities in prisoner characteristics extended to variations in the reported prevalence of alcohol abuse and drug abuse. However, the rigorous application of propensity score matching effectively mitigated these differences, resulting in a more balanced and comparable comparison between the treatment and control groups. To comprehensively evaluate the effectiveness of the IPS “employment world” training programme, our analysis began by comparing general recidivism risk among programme participants and non-participants. Subsequently, we merged the dataset containing the treatment and matched comparison groups with the Israel Central Bureau of Statistics data. This integration was achieved by matching individual prisoners using their unique Israeli IDs, allowing us to track each participant precisely after release.
We employed specific statistical tests to assess the significance of our findings and discern differences between the treatment and matched comparison groups across various outcome measures. Specifically, we utilized χ 2 tests to examine reincarceration and rearrest rates and the rates of prisoners registered with welfare services. In parallel, we conducted t tests to appropriately analyse the number of months employed yearly, annual income and annual tax payments. Cohen’s d was also calculated to provide further insight into the effect size associated with these measures. This meticulous and comprehensive statistical approach empowered us to conduct a comprehensive analysis of diverse dimensions beyond recidivism, reinforcing the robustness of our analysis and enhancing the clarity of the interpretation of our results.
RESULTS
As mentioned in the Method section, we will begin by comparing the test and control groups’ recidivism rates (general rearrests and reincarcerations) and then examine the differences between the two groups in the other outcome measures previously mentioned. The rates of recidivism among participants in the “employment world” programmes were examined using two measures: rearrests and reincarcerations. While monitored, the number of prisoners in each group – the training group and comparison group – was identical throughout the monitoring period: after one year, two years, three years, four years and five years. Naturally, the rate of recidivism cannot diminish over the years. In the file used in the present study, no recidivism data exist for the later years of prisoners released in later years since the file was closed before the monitoring period had terminated. Therefore, the follow-up each year was conducted only for the group of prisoners for which we had recidivism data, and therefore, the groups of prisoners became smaller as the years progressed. Under these circumstances, whereby the recidivism among a different-sized group was measured each year, a decline in recidivism may emerge as the years progress. The decline in the number of training group and comparison group participants also affects statistical power and the possibility of reaching statistical significance in the later years of the monitoring period.
It is further noted that the test group includes all prisoners who participated in the programme, including ex-offenders who were re-incarcerated during the monitoring period.
Reincarcerations: “Employment World” Training Programme Participants
Table 2 shows the reincarceration rates among prisoners who participated in the “employment world” programme versus those of the comparison group. The findings reveal no significant differences in the reincarceration rate between the participant and non-participant groups throughout the monitoring period.
Rate of Reincarcerations Following Release among Prisoners Who Participated/Did Not Participate in the “Employment World” Training Programme
Rearrests: “Employment World” Training Programme Participants
Table 3 shows the rearrest rates among prisoners who participated in the “employment world” programme versus those of the comparison group. The findings reveal consistently lower recidivism rates among members of the training group compared to members of the non-participant group throughout the monitoring period. A statistically significant difference was only found between the two groups in the third year of monitoring, with a small effect. That year, the risk of rearrest was 11.30% lower among the training group prisoners than in the comparison group.
Rate of Rearrests Following Release among Prisoners who Participated/Did Not Participate in the “Employment World” Training Programme
* p < 0.05.
As the findings of the first part show, the data on the effectiveness of the “employment world” programme are not encouraging where recidivism rates are concerned. This fact heightens the need to examine additional outcome measures that could attest to the programme’s effectiveness.
Employment Months: “Employment World” Training Programme Participants versus the Control Group
Table 4 shows the number of months of employment yearly among the prisoners who participated in the training programmes versus those of the comparison group. For the first four years following release, the average number of employment months among the prisoners who participated in the “employment world” programmes was significantly higher than the control group. The average number of employment months at the end of the first year following release among “employment world” programme participants was 44.37% higher than that of the control group. Two years after release, the average number of employment months among “employment world” programme participants was 33.61% higher than that of the control group. Three years after their release from prison, the average number of employment months among “employment world” programme participants was 40.88% higher than that of the control group, and four years after release, it was 98.77% higher than that of the control group. The effect size was small to medium throughout the first four years of the monitoring period.
Number of Employment Months Yearly among Prisoners Who Participated/Did Not Participate in the “Employment World” Training Programme
*** p < 0.001.
Annual Income: “Employment World” Programme Participants versus the Control Group
Table 5 shows the average annual income of prisoners participating in the “employment world” training programme versus the control group. Throughout the monitoring period, the average annual income among the prisoners who participated in the “employment world” programme was significantly higher than that of the control group. The average annual income at the end of the first year following release among “employment world” programme participants was 52.64% higher than that of the control group. Two years after release, the average annual income among programme participants was 51.12% higher than that of the control group. Three years after their release from prison, the average annual income among programme participants was 54.62% higher than that of the control group. Four years after release, it was 107.18% higher; five years post-release, it was 45.73% higher. The effect size was small to medium throughout the monitoring period.
Annual Income (New Israeli Shekels) among Prisoners Who Participated/Did Not Participate in the “Employment World” Training Programme
* p < 0.05, *** p < 0.001.
Annual Tax Payment: “Employment World” Programme Participants versus the Control Group
Table 6 shows the average annual tax payment among prisoners who participated in the “employment world” training programme versus the control group. During the first two years following their release, the average annual tax payment among the prisoners who participated in the “employment world” programme was significantly higher than that of the control group. The average annual tax payment at the end of the first year following release among “employment world” programme participants was 108.36% higher than that of the control group, and two years after release, it was 80.94% higher. The effect size was small throughout the first two years post-release.
Annual Tax Payment (New Israeli Shekels) among Prisoners Who Participated/Did Not Participate in the “Employment World” Training Programme
* p < 0.05.
Rate of Prisoners Registered with the Welfare Services: “Employment World” Programme Participants versus the Control Group
Table 7 shows the rate of prisoners registered with the welfare services among the group of prisoners who participated in the “employment world” training programme versus the control group. It seems that, during the first three years of monitoring, the rate of prisoners registered with the welfare services among the prisoners who participated in the “employment world” programme was significantly higher compared to that of the control group. The rate of prisoners registered with the welfare services at the end of the first year following release among “employment world” programme participants was 13.98% higher than that of the control group. Two years after release, the rate of prisoners registered with the welfare services among programme participants was 16.97% higher than that of the control group, and three years into the monitoring period, it was 19.54% higher. The effect size was small in those years.
The Rate of Prisoners Registered with the Welfare Services among Prisoners Who Participated/Did Not Participate in the “Employment World” Training Programme
* p < 0.05.
It is noteworthy that, while examining other “employment world” programme outcome measures, a test was conducted to examine the differences between “employment world” programme participants and the control group in the measure of prisoners placed in out-of-home frameworks; however, throughout the monitoring period, no statistically significant differences were found.
DISCUSSION
The prevailing approach in evaluating the effectiveness of rehabilitation programmes has traditionally centred on comparing recidivism rates between programme participants and non-participants. While recidivism remains the primary metric for assessing the efficacy of rehabilitation programmes (Bales et al. Reference Bales, Clark, Scaggs, Ensley, Coltharp, Singer and Blomberg2016; Ellison et al. Reference Ellison, Szifris, Horan and Fox2017; Gibbons Reference Gibbons1999; Sechrest, White, and Brown Reference Sechrest, White and Brown1979), recent studies have shifted their focus towards exploring the impact of these programmes on a broader range of outcome measures beyond recidivism (Cook et al. Reference Cook, Kang, Braga, Ludwig and O’Brien2015; Creese Reference Creese2016; Hunter and Boyce Reference Hunter and Boyce2009; Skardhamar and Telle Reference Skardhamar and Telle2012).
This study distinguishes itself by examining the programme through the lens of recidivism and incorporating additional economic and social outcome measures. These measures include employment, income, tax contributions, acknowledgement of the need for welfare services, and engagement with out-of-home frameworks among released prisoners who participated in an employment training programme designed to prepare offenders for re-entry into society. To mitigate the inherent selection bias when forming a control group, this study employed PSM to minimize bias in subject selection, ensuring comparability between programme participants and non-participants (Jordan Reference Jordan2012). The findings of this study reveal that while programmes aimed at preparing inmates for employment did not emerge as strong drivers of significant change in reported recidivism rates when compared to a control group comprised of individuals who did not partake in such programmes, a notable difference becomes evident when examining additional outcome measures. Specifically, released prisoners who had participated in programmes focused on employment persevered and retained their jobs for nearly twice as long (44.7%) as their non-participant counterparts during the first 12 months following their release – a period recognized in the research literature as the time when former prisoners are most susceptible to recidivism (Shoham and Peled-Laskov Reference Shoham and Peled-Laskov2022). Intriguingly, this significant difference persisted over the subsequent four years, with the average number of employment months yearly among programme participants being nearly 100% larger (98.77%) than that of their non-participant peers. A similar pattern emerged when assessing average annual income.
These findings raise a compelling question: how do these programmes succeed in enhancing pro-social behaviours among released prisoners while falling short in reducing recidivism? The answer lies in the programme’s primary objective, which is to transform the daily activities of released prisoners and promote their integration into society through legal employment and income generation. The underlying assumption is that these positive activities will, in due course, influence criminal involvement and recidivism, thus resulting in an indirect and seemingly weak correlation between rehabilitation programmes and recidivism.
Indeed, the integration of employment serves as a foundational stepping stone for released prisoners re-entering society, offering a legitimate source of income and playing a vital role in shaping their daily routines and adherence to pro-social norms (Bucklen and Zajac Reference Bucklen and Zajac2009; Bushway and Apel Reference Bushway and Apel2012; LeBel et al. Reference LeBel, Burnett, Maruna and Bushway2008). Over the past two decades, the research literature has underscored the importance of assessing perseverance in the workplace (Duwe Reference Duwe2015) and examining average wages as an indicator of released prisoners’ reintegration into the community (Peled-Laskov et al. Reference Peled-Laskov, Shoham and Cojocaru2019). Many released prisoners carry numerous debts and fines (Pogrebin et al. Reference Pogrebin, West-Smith, Walker and Unnithan2014), making their integration into employment with appropriate pay a pivotal element in their rehabilitation process – a potential turning point in their criminal trajectories (Duwe Reference Duwe2015; Skardhamar and Telle Reference Skardhamar and Telle2012). It is important to acknowledge that the programme under study goes beyond mere vocational training. It places significant emphasis on equipping participants with practical skills and attitudes necessary not only to secure employment but to thrive in a workplace environment. This holistic approach includes imparting effective interpersonal skills, emphasizing qualities such as persistence and punctuality, and providing valuable lessons on financial management and personal responsibility.
The significance of this nuanced approach becomes evident when attempting to explain the observed outcomes in our study. By focusing on more than just vocational skills, the programme equips released prisoners with broader tools needed to excel in the workforce. This multifaceted approach may, in part, elucidate the positive occupational outcomes observed among programme participants.
Furthermore, considering the unique sociocultural context of Israel, influenced by factors such as Judaism, the concept of repentance (Haviv et al. Reference Haviv, Shoham, Hasisi, Weisburd and Toren-Rozanski2020) and legal restrictions on inquiring about criminal history during hiring (Peled et al. 2019), provides insights into the complexities surrounding our findings. These factors probably contribute to positive employment outcomes and the reduction of stigma associated with a criminal record, further reinforcing the programme’s success. In this context, our study’s findings reveal that, at the end of the first year following release, prisoners who had participated in the employment programmes earned annual wages 52% higher than those in the comparison group. This substantial difference persisted throughout the five-year monitoring period. The decision to focus on the earnings of former prisoners, rather than just their employment status, aligns with existing research indicating that integration into legitimate work bolsters released prisoners’ self-esteem, enabling them to earn a living and support their families with dignity (Drake Reference Drake2003; Smith et al. Reference Smith, Bechtel, Patrick, Smith and Wilson-Gentry2006). Adequate pay can contribute to workplace stability and a positive routine, thereby reinforcing the adoption of pro-social values. It is worth noting that the gap between the two groups could be attributed to the larger number of hours worked by released prisoners who participated in the employment training programme, as opposed to securing higher-paying jobs with greater potential for reducing recidivism (Ramakers et al. Reference Ramakers, Nieuwbeerta, Van Wilsem and Dirkzwager2016; Uggen Reference Uggen1999).
As mentioned earlier, this study aimed to explore measures associated with pro-social values and the often-perceived indirect variables related to recidivism. According to desistance-oriented rehabilitation approaches, genuine rehabilitation involves a transformation in the lives and self-perception of lawbreakers, with crime desistance representing a gradual process of change and integration into the community. Fair compensation and positive self-perceptions are pivotal in this process (Raphael Reference Raphael2007; Shoham and Peled-Laskov Reference Shoham and Peled-Laskov2020). Tax-paying can also be viewed as an expression of pro-social values and normative lifestyles embraced by former prisoners. Our study demonstrates that, in the first two years following release, the rate of taxpayers among programme participants was significantly higher than among non-participants.
A similar trend emerged when examining registration with welfare services. During the first three years following their release from prison, a significantly higher proportion of released prisoners who participated in employment programmes were registered with welfare services to receive various benefits. While there is a tendency to view the need for assistance from welfare services as a sign of weakness, among the deprived population of released prisoners, seeking help upon re-entering society should be considered a positive step. Welfare personnel consulted by the authors have explained that it reflects the former prisoners’ willingness to adhere to the accepted rules of societal engagement and is in line with the principles of the crime desistance theory (Maruna Reference Maruna, McNeill, Raynor and Trotter2010). This theory posits that work integration and tax-paying signify the embrace of positive pro-social values and a willingness to cooperate with state systems and adhere to social norms – an integral part of the crime desistance process (Shoham and Peled-Laskov Reference Shoham and Peled-Laskov2022). This study sheds light on the effectiveness of a specific prison employment programme within the IPS. However, several limitations warrant consideration. One limitation is the inclusion of participants who returned to prison in the analysis of occupational outcomes, potentially introducing bias. To isolate the programme’s pure impact on occupational outcomes, the analysis should have included only participants who completed the programme without returning to prison.
Notwithstanding this limitation, the positive impact on occupational measures, the statistically significant improvements in income, months of employment yearly, tax payment rate and welfare registration rate compared to the control group all suggest that the programme is effective in improving occupational outcomes, contrary to what might be expected given the lack of improvement in recidivism. Since this programme was not primarily designed to reduce recidivism, further investigation is needed to understand the complex and multidimensional correlation between employment and criminality.
Another limitation is the omission of post-release supervision from the control variables, which may have influenced the results. However, due to the nature of PSM analysis, variables that the programme can influence cannot be included.
In summary, while this study provides valuable evidence of the effectiveness of a prison-based employment programme, the highlighted limitations underscore the need for further research to confirm and expand upon these findings. The study emphasizes that assessing the effectiveness of various programmes designed to reintegrate former prisoners into the community requires a broader perspective encompassing a wider range of outcome measures beyond formal measures of rearrest and reincarceration rates. Much remains to be explored in the intricate and lengthy process of reintegrating former prisoners into society, necessitating additional quantitative and qualitative studies to elucidate the significance of different variables in this multifaceted re-entry and crime desistance process. As demonstrated in this study, leveraging robust statistical matching methods and adopting the perspective of released prisoners themselves could contribute to developing more comprehensive and effective intervention programmes within the corrections system.
CONCLUSION AND FUTURE DIRECTIONS
In conclusion, this study has offered valuable insights into the effectiveness of a prison-based employment programme, highlighting its significant impact on employment, income, tax compliance, welfare services utilization and overall pro-social habits among released prisoners. While the findings indicate that the programme may be ineffective in reducing recidivism, it has shown promise in enhancing various aspects of the lives of those re-entering society. As emphasized in the literature review, the next critical step is recognizing the detrimental effects of stigmatization, exclusion and systemic inequalities often accompanying re-entry initiatives. As evidenced by the work of scholars in social work and critical criminology, including Halushka (Reference Halushka2020), Hinton and Cook (Reference Hinton and Cook2021), Miller (Reference Miller2021), Simon (Reference Simon2010) and Thompson (Reference Thompson2012), these aspects cannot be overlooked. Therefore, policymakers must take proactive measures to improve the overall impact of re-entry programmes while addressing these systemic challenges.
Moving forward, policymakers should consider the following steps to enhance the effectiveness of re-entry programmes:
-
1. Comprehensive assessment: Conduct a thorough assessment of existing re-entry programmes to identify areas where improvements can be made. This assessment should include input from formerly incarcerated individuals and experts in the field.
-
2. Anti-stigmatization initiatives: Develop strategies to counteract stigmatization and negative stereotypes faced by released prisoners. Public awareness campaigns, anti-discrimination policies and employer education initiatives can be instrumental in challenging societal biases.
-
3. Equity in access: Ensure that re-entry programmes are designed with a focus on equity. Address systemic barriers disproportionately affecting marginalized communities and promote fair access to support services for all individuals.
-
4. Social justice frameworks: Frame re-entry programmes within a broader social justice and equity context. Policies should actively reduce societal disparities and promote fair opportunities, aligning with broader social justice goals.
-
5. Holistic approaches: Promote comprehensive re-entry approaches that address not only legal and educational barriers but also broader societal and systemic factors. This includes providing support for mental health, substance abuse treatment, housing and family reunification.
-
6. Research and evaluation: Invest in ongoing research and evaluation of re-entry programmes to measure their impact on various outcome measures, including employment, income and recidivism. This research can inform evidence-based practices and policy adjustments.
Acknowledgements
We are grateful to the Research Unit at the Israeli Prison Service for their invaluable assistance in data creation and overall study development.
Efrat Shoham serves as the Chair of the Criminology Department and leads the Shaam Institution for the reintegration of formerly incarcerated individuals at Ashkelon Academic College. With a prolific academic career, she has authored 11 books and contributed to over 60 articles spanning topics such as prisoners’ rehabilitation, cultural criminology and social control within segregated communities, among others.
Noam Haviv an assistant professor at John Jay College’s Department of Law, Police Science, and Criminal Justice Administration, specializes in prisoners’ rehabilitation and programme evaluation. His research explores reducing recidivism and successful reintegration. With a strong background in statistical analysis, he has published extensively on topics such as risk factors, programme effectiveness and the role of ethnicity in juvenile justice.
