Sociolegal Characteristics and Parole Infractions Among Israeli Released Prisoners During Electronic Monitoring

Abstract

The objective of this study is to examine the part played by sociolegal characteristics such as ethnic background, family status, or criminal past in the rate of infractions among ex-prisoners in Electronic Monitoring (EM) Programs. In addition, it focuses on the nature of the formal decisions made by community supervision agents regarding such infractions and their correlation with the sociolegal characteristics of the participants. The research population included all prisoners on license (i.e., prisoners who have been granted conditional early release) who took part in the EM project from mid-2007 until mid-2009 (24 months), altogether 155 participants. The data show no significant correlation between the number of infractions and the participant’s sociolegal background. In spite of the fact that the EM coordinators have extensive discretionary power, which is likely to lead to discrimination attributable to variables such as ethnicity, this research shows that the most efficacious variable for explaining formal responses is an objective one—the number of infractions.

Keywords

reintegration electronic monitoring Israel

Introduction

The new administrative penology (Ajzenstadt & Barak, 2008; Simon & Feely, 1995) promotes limitation and incapacitation as the primary objective for temporarily reducing crime rates (Blomberg, Bales, & Waid, 2008). As an alternative to incarceration, the electronic monitoring (henceforth referred to as EM) programs, currently utilized by many Western states, are enjoying extensive empirical pursuit (Gable & Gable, 2005).

The main advantages attributed to the EM programs are (a) the decrease in the costs of punishment; (b) the reduction of the tax burden, by making it possible for the criminal to work for his living; (c) the decline of overcrowding in prisons; (d) the enabling of the criminal to fulfill personal and civic obligations; and (e) the decrease in recidivism during the duration of the program, even though the findings regarding a decrease in recidivism afterwards is inconclusive (Albrecht, 2005; Bonta, Rooney, & Wallace-Capretta, 1999; Finn & Muirhead-Steves, 2002; Gable & Gable, 2005; National Audit Office (NAO), 2006; Padgett, Bales, & Blomberg, 2006).

Supervision by electronic means includes fastening onto the monitored individual’s leg an electronic ankle bracelet (a tag) containing a transmitter, which transmits signals to a receiver in his house, and this in turn uses the telephone to transmit alarms on violations of the terms of the supervision in real time to a command center monitoring the prisoner (the receiver identifies interruptions in transmission or other signals). This system makes it possible to custom design a supervision regime that is uniquely suited to each individual offender, in accordance with his own individual characteristics (social, legal, and psychological), his personality traits, the crime he committed, and his rehabilitation needs (Efodi, 2010).

A review of this research literature shows that discharged prisoners who have already served part of their prison terms are the primary target population for inclusion in a program of supervision. This population is considered more stable and more ready for rehabilitation (in the sense of reintegration in society), as compared with other populations of criminals, detainees, and others (Segal & Wagschal, 2004).

The EM program is part of current community supervision trends in Western countries and in Israel, which enables the reintegration of convicted offenders into their communities and enables them to carry on a normative lifestyle. They are required to abide by limitations and a strict supervisory regimen, which should minimize the risk of a return to a criminal lifestyle (for the integration of discharged prisoners in various normative settings, see Hollin, 2001; McGuire, 2002; Petersilia, 2004; Uggen, 2000; Zhang, Roberts, & Callanan, 2006). The studies show that there is a positive correlation between the rates of recidivism among discharged prisoners and variables such as age, family status, criminal past, unemployment, drug addiction, living outside the home as well as many other factors (Bonta et al., 1999; O’Tool, 2004). Hucklesby (2008) suggests that for some offenders, curfew restrictions reduce offending and contribute to desistance by addressing levels of social behavior in two ways: first, by decreasing levels of antisocial behavior through reduction of offender’s links to situations, people, places, and network correlated with their offending and second, by improving levels of prosocial behavior by means of encouraging offenders to connect or reconnect with positive influences linked with desistance such as family and employment. In addition, Hucklesby (2009b) claims that the reasons offenders comply or fail to comply with the EM program are complex and interrelated, and they encompass many different factors, including perceptions of procedural justice, individual motivation, and attachment to significant others.

EM: The Israeli Case

At the end of 2006, the Prisoners’ Rehabilitation Authority in Israel began to operate an EM program for discharged prisoners.

The prisoners who participate in this program are released after serving two thirds of their sentence (henceforth referred to as licensed prisoners), and the rest of their sentence is served outside of prison walls, while they are under supervision and wear EM bracelets.

Once a prisoner has finished serving two thirds of his sentence, the Parole Board reviews his case and hands down a decision on his eligibility to be an early-release licensed inmate. The Board considers the rehabilitation process that the participant underwent during his incarceration, as well as the preparation for a rehabilitation program supervised by the Prisoners’ Rehabilitation Authority (henceforth referred to as the Authority) after his release. The Authority is responsible for rehabilitating prisoners and their family members, and bears the main responsibility for constructing and implementing supervised rehabilitation programs for the period of the prisoners’ early release.

During this time, they are required to report to the Parole Board for follow-ups and the monitoring of their progress in the treatment program. The Electronic Monitoring Rehabilitation Program includes a tight element of supervision, which is the electronic bracelet; psychosocial treatment; and constant employment. Such employment is made possible through periods of time in the day, referred to as “Windows,” allowing the monitored prisoners to leave the place of supervision to take part in the workforce and participate in various rehabilitation programs. For the most part, the curfew starts at 20:00 p.m., yet, in some of the cases it starts even later (22:00 p.m.), this according to the decision of the parole Board. In some of the cases, not in all of them, the curfew at home includes weekends as well. In Israel, EM is operating only by landlines and not by a cell phone system.

These windows are approved by the Parole Board, and vary from time to time according to the supervisors’ recommendations and to changing needs. The Electronic Monitoring Rehabilitation Program in Israel represents a holistic model, which includes four main elements for rehabilitation—treatment, employment, monitoring and reports (Efodi, 2010).

Treatment

The treatment of prisoners released into the EM program is carried out by municipality social workers located in the same area of the prisoner. These workers are the Prisoners’ Rehabilitation Authority’s community representatives; however, they do not directly work for the Authority. They are municipality employees. The treatment of a prisoner in the community includes treating addictions, group and individual therapy, family therapy, and so on. In addition, the monitored individuals are subject to urine tests to verify that he is clean from drugs.

Employment

Every monitored prisoner (with a few exceptions) is required to have a regular full-time job, this is where the Prisoners’ Rehabilitation Authority’s employment coordinators help by finding them jobs and helping them adjust to their new place of employment.

Monitoring

Monitoring is done by the EM coordinators, who work for the Prisoners’ Rehabilitation Authority. They are in charge of implementing the rehabilitation program in its entirety. The coordinators are in constant contact with all the elements relevant to the prisoner’s rehabilitation (for instance, their social worker and employment coordinator). Also, the monitored prisoners are called in for follow-up meetings or warnings, accordingly.

Reports

The supervisor’s function is to report to the Parole Board, which then decides, according to recommendations submitted, whether to administer leniencies or other changes to the structure of the rehabilitation program. The Parole Board has the sovereignty to decide to revoke a prisoner’s license, if he has not complied with the terms of his rehabilitation program. The duration of The Electronic monitoring Rehabilitation Program is from the day of his licensed release and up until the end of his sentence, in any case, not longer than 2 years from the day of his release (Shoham, Yehosha, Efodi, & Diamant, 2010).

The coordinators of EM at the Authority, who supervise the rehabilitation program, have broad discretionary powers regarding the number and nature of the violations which may lead to a variety of sanctions ranging between a simple warning to the prisoner to a formal report submitted to the Parole board which may result in a violation and reinstatement of the remainder of the prison term (Shoham et al., 2010).

Israel’s EM program is intended for prisoners who have already served two thirds of their prison terms, but their release can only be an option contingent on increased supervision. Three groups are included under this requirement: (a) weak prisoners who can be easily influenced and need close supervision during the critical period after discharge to prevent their coming into contact with criminals, and who would be unsuitable for early release without the EM; (b) prisoners convicted of serious offenses but underwent treatment programs during their incarceration; and (c) prisoners who were previously in a program of treatment-oriented supervision, and this would be the opportunity to toughen the conditions of their supervision (as this program is experimental, the criteria for participation in the program is based solely on the Authority’s internal protocols for EM [Diamant, 2006], describing guidelines for the selection of the prisoner population for this program). The Ministry of Public Security and the Attorney General made a decision in principle that the EM program in Israel is not intended for sex offenders, domestic violence, or drug trafficking because of the possibility of committing these crimes from the place of the monitoring—that is, their home—and also because of the high degree of danger they pose to their victims and to society in general (Ministry of Internal Security, 2006).

EM Programs: Selection Criteria

While EM programs as an alternative to imprisonment or arrest are being implemented in many countries throughout the world (Hucklesby, 2009a), selection criteria for the suitable population differs from one to another locality and from one individual to another. However, great weight is usually given to variables such as criminal past or the threat that the individual possesses (Gable & Gable, 2005). In Sweden, for instance, Marklund and Holmberg (2009) find that prior involvement in criminal activity has a significant effect on criminal behavior during the EM program. In New Zealand, EM matches for offenders who had not committed serious offenses and who were sentenced to, or serving, varying lengths of imprisonment (Gibbs & King, 2003).The nature of the EM program’s integration and its place in the array of sanctions in a given country is determined in accordance with that country’s judicial and enforcement systems (for an extensive description of the various programs, see Segal & Wagschal, 2004). While in all countries the program applies primarily to prisoners on conditional release and detainees awaiting completion of proceedings or on probation, there is great variation in the criteria for the classification of the population considered suitable for EM (Shoham et al., 2010). Crowe (2002), for example, reviews the different criteria for matching offenders to EM programs in a sampling of states in the United States. Among the criteria supporting the inclusion of individuals in the program are the absence of a significant criminal past, the readiness and motivation to be subjected to the monitoring regimen, medical considerations, whether the person is the primary caregiver for other family members, and a reasonable expectation of social support, along with the absence of prior or current sex or violence-related offenses. Criteria that would exclude him from the program include inappropriate behavior in prison, failed previous attempts during correction, drug addiction, the existence of a mental illness limiting the criminal’s ability to control his behavior, and so on. At the same time, these criteria serve only as a guideline: The decision regarding the assignment of any criminal to the EM program is based on the merits of each case (Crowe, 2002). For example, a long-term research in the United States shows that EM is often used to supervise grave offenders (Padgett et al., 2006).

The primary guiding consideration in the various programs described in Crowe’s (2002) study is the extent of the threat posed to society by the particular criminal. Other studies (Downing, 2006; Haverkamp, Mayer & Levy, 2004; Renzema & Mayo-Wilson, 2005) describe additional criteria, including the existence of a permanent residence and an active telephone in the home of the individual being monitored, the agreement of the adult members of the household to have an EM program operating in their home, the satisfactory physical and mental health of the individual, being employed for at least 20 hr a week while he is being monitored, and a period of supervision that does not exceed 1 year.

Extralegal Variables and Disparities in the Criminal Justice Systems

The “Retribution model” known also as the “Just Model” (Hudson, 2003; Shoham, 2009; Von Hirsch, 1986), which evolved in many Western countries as a reaction against subjective punishment, and the excesses of needs-based rehabilitative sentencing brought in its wake different ways of structuring the decisions made in the course of the criminal proceeding (Ben David, Fishman, & Tsadok, 2006; Garland, 2002; Stith & Cabanes, 1998; Ulmer & Cramer, 1996). The structuring of the discretion underlying various judicial decisions was intended to minimize the existing disparities in criminal justice and to increase the fundamental uniformity, consistency, and equality of these decisions (for a description of studies examining the interrelationships between such extralegal variables as race and ethnicity, and judicial decisions, see Brennan, 2008; Hawkins, 1995; Lahav, 2002; Steffenmeier & Demith, 2000, 2006; Walker, Spohn, & Delone, 2000).

One extralegal characteristic that has been found empirically to be related to disparities in judicial decisions in Israel was the ethnicity of the person convicted. Most of the studies conducted thus far in Israel indicate disparities in the legal system, nearly always to the disadvantage of the Arabs. For example, the findings of the comprehensive research done by Rattner and Fishman (2004) show that even when there is statistical control for the legal variables related to the nature of the crime, a criminal past, and the age of the accused, there are significant differences between Jews and Arabs in the probability of conviction and of receiving a “prison sentence.” Broadening discretion of the criminal proceeding from the beginning to the end could interfere with uniformity of punishment. It could even cause discrimination and inequality in the operations of the law enforcement and correctional system, of which the Prisoners’ Rehabilitation Authority is a part (see, for example, First & Agmon-Gonen, 2007; Lahav, 2002; Rattner & Fishman, 2004; Shoham & Abulafia, 2010).

The Israeli EM program is a pilot project; therefore, this study sets out to identify the sociolegal characteristics of the licensed prisoners selected for this program. The second aim is to examine the relationship between those characteristics, such as ethnicity or family status, and the level of infractions, and between those characteristics and the formal response toward these infractions.

The current study is the first stage of a more comprehensive, three-stage study, designed to evaluate the efficiency of the EM project in Israel during the program and afterwards, as well as the level of satisfaction among all parties involved in this project.

Method

The Sample

The participating population included all the licensed prisoners (i.e., prisoners who have been granted conditional early release) who took part in the EM project from mid-2007 until mid-2009 (24 months). All of them were males, above the age of 21 who were under the supervision of the Authority for a period of at least 6 months—a total of 155 licensed prisoners. The mean age of the participants in the EM project was 34 (SD = 10.37), with an average of 10 years of schooling. Jewish participants comprised 68% of the participants; 32% were Muslim.

Unmarried men accounted for 44% of the participants, 37% were married, and 19% were divorced, separated, or widowed. Most were manual laborers (63.2%). Most of the monitored individuals (40%) were in the North, with a similar number of monitored individuals (30%) in the central and in the southern districts.

Table 1 shows that approximately 50% of the monitored individuals had been released from maximum security prisons, 33% from medium security prisons, and 16% from minimum security prisons. The average length of the most recent prison sentence was 53.66 months; however, the standard deviation was quite high (49.51). (The average length of prison sentences for men in Israel is 28 months).

Table 1.

Sample of Incensed Prisoners in Electronic Monitoring, 2007-2009, in Percentages (N = 155).

Characteristics of monitored individuals	%
Demographic characteristics
Nationality
Jewish	67.7
Arab	32.3
Country of origin
Native Israelis	76.1
Immigrants	23.9
Family status
Single	43.9
Married	37.4
Divorced/separated/widowed	18.7
Vocation
Laborer	63.2
Has had vocational training	6.5
Has no vocation	25.8
Unemployed	4.5
Area of residence
Center of Israel	29.0
North	40.0
South	31.0
Military service
Full	9.0
Partial	21.9
Did not serve (includes Arabs)	56.2
No information	12.9
Legal characteristics
Type of prison
Maximum security	49.0
Medium security	32.9
Minimum security	16.2
No information	1.9
Total number of incarcerations
One	51.6
Two	21.9
Three or more	26.5
Reason for present incarceration
Murder	17.4
Assault and battery	21.3
Property	29.0
Fraud	24.5
Drugs	26.5
Drug abuse
Did not use drugs	57.4
Use drugs	39.4
No information	3.2
Infractions
No infractions	60.0
One infraction	16.8
Two or more	23.2
Criminal offense during electronic monitoring
No criminal offense	89.7
One or more	10.3
Warning
No warning	69.7
One warning	12.9
Two or more	17.4
Reports
No reports	77.4
One report	14.2
Two or more	8.4

The average age of these monitored individuals at the time of their first arrest was 19 (SD = 6.02). For half of those under supervision (51%), this was their first incarceration.

The questionnaires filled out by the supervisors show that slightly less than half (40%) of the monitored ex-prisoners were drug abusers, the average age of the beginning of drug use was 18.5 (SD = 6.10). Table 1 also shows that the reasons for the recent arrest are many (as some of the prisoners committed a number of offenses, the total of these figures exceeds 100%).

The data showed that for 60% of those being monitored, there were no recorded infractions, 16.8% violated the terms once, and 23.2% committed two or more infractions. Only 10% of those being monitored were involved in criminal offenses during the EM period.

About two thirds of those being monitored (70%) received no warnings at all, 13% received one warning, and 17% were given two or more warnings. The supervisors issued reports on 23% of the participants to the Board which has the authority to revoke the license or to redefine its terms.

Tools

The researchers adjusted Weisburd, Shoham, Barak, Manspeizer, and Gideon’s (2005) Discharged Prisoners Characteristics Questionnaire to the specific characteristics of the EM program. The questionnaire was validated in two phases: In the first phase, the questionnaire was distributed among Authority personnel, who were asked to add other relevant questions or delete any that were irrelevant. At the end of this phase, a five-section questionnaire was prepared to adhere with the EM program:

The sociodemographic characteristics of the licensee;

The legal characteristics of the licensed prisoner;

Variables which we defined as either impeding or supporting rehabilitation. These variables were divided into two categories: personal and environmental. Personal variables included variables such as use of alcohol and/or drugs, unreliability at work, and an unwillingness to cooperate. Environmental variables included a lack of familial support, criminality in the family, and a negative or criminal environment.

A description of the prisoner’s behavior during the monitoring period.

The responses of the supervisors to infractions throughout the research period.

At the end of the questionnaire, the EM coordinators were asked to give their opinion in an open-ended question regarding each specific prisoner in the course of the EM, and how motivated he appeared to be to make a change in his former lifestyle.

In the second phase, the questionnaire was distributed among the EM district coordinators in the three primary districts of the Authority (North, South, and Central). The coordinators were asked to fill out the details of the questionnaire, based on the personal file of each of the licensed prisoners who was electronically monitored in the course of 2007 to 2009. A total of 155 questionnaires were filled out. The data collected from the files were analyzed by means of several parametric and nonparametric tests, whose purpose was to find a relationship between the sociolegal characteristics of the prisoner on license and his behavior during EM, and between these characteristics and the responses of the monitoring coordinators to infractions.

Results

To examine the relationship between infractions and various variables, we constructed a dichotomous variable: the total absence of infractions or one infraction during the entire period of monitoring, and two or more infractions during that period.

This was due to the fact that the finding indicated a very large statistical distribution in number of infractions, and the similar implication of one infraction or none, in the eyes of the supervisors.

Table 3 shows a significant correlation between the number of infractions and the prisoner’s relationship with his nuclear family, χ²(2) = 10.52, p < .01, φ = 0.30. Only a few licensees who have the contact defined by the Authority supervisors as being intimate and supportive (7%) violated the terms of their monitoring. In contrast, among those whose relationships with their nuclear families were partial and unstable, 35% committed infractions. As expected, a significant correlation also appeared between the prisoner’s ability to persevere on the job and his involvement in infractions, χ²(1) = 8.17, p < .05, φ= 0.23.

Table 2.

Frequencies in Percentages and the Correlation Between Infractions and Demographic, and Legal Variables (N = 155).

Demographic variables	0-1 infractions (%)	2+ infractions (%)	χ² (φ/Cramer’s V)
Nationality			0.31 (0.045)
Jewish	78.1	21.9
Arab	74	26
Country of origin			1.33 (0.09)
Native Israelis	74.6	25.4
Immigrants	83.8	16.2
Family status			2.72 (0.13)
Single	80.9	19.1
Married	77.6	22.4
Divorced/separated/ widowed	65.5	34.5
Vocation			0.29 (0.04)
Laborer	77.6	22.4
Has had vocational training	70	30
Has no vocation/unemployed	76.6	23.4
Area of residence			9.6** (0.25)
Center of Israel	60.9	39.1
North	85.5	14.5
South	80.9	19.1
Military service			1.54 (0.11)
Full	64.3	35.7	Full
Partial	76.5	23.5	partial
Did not serve (includes Arabs)	79.3	20.7	did not serve (includes Arabs)
Legal variables	0-1 infractions (%)	2+ infractions (%)	χ² (φ)
Type of prison			0.76 (0.07)
Maximum security	73.7	26.3
Medium security	80.4	19.6
Minimum security	76	24
Reasons for present incarceration			0.13 (0.03)
Murder/manslaughter	74.1	25.9
Assault and battery	69.7	30.3	1.17 (0.09)
Property	66.7	33.3	3.63 (0.15)
Fraud	63.6	36.4	1.15 (0.09)
Drugs	82.9	17.1	1.18 (0.09)
Total number of incarcerations			1.22 (0.09)
One	80	20
Two	70.6	29.4
Three or more	75.6	24.4

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

Another significant correlation was found between the number of infractions and the regularity with which the participant reported to the supervision coordinator of the Authority, χ²(1) = 5.77, p < .01, φ = 0.20. The percentage of infractions among participants reporting to the coordinator once every 2 months (38%) was double than of those who reported once every 2 weeks or once a month (18%).

There were no significant correlations between infractions and family status, ethnicity, security level of the prison facility from which the prisoner was released, or the nature of the most recent crime for which the monitored individual was imprisoned, with the exception of property crimes (see Tables 3 and 4).

Table 3.

Frequencies in Percentages and the Correlation Between Infractions and Rehabilitation-Oriented Variables (N = 155).

Impeding or supporting variables	0-1 infractions (%)	2+ infractions (%)	χ² (φ)
Residence at time of supervision			2.23 (0.12)
With family of origin	82.1	17.9
With family of procreation	70.8	29.2
Rented apartment/other	75	25
Relationship with family of origin			3.86 (0.16)
Close and supportive	86	14
Reasonable	80.9	19.1
Partial	70.5	29.5
Frequency of visits to coordinator			5.77** (0.2)
Biweekly or monthly	81.4	18.6
Bimonthly	62.2	37.8
Assistance by other community agent			3.08 (0.14)
Present	74.6	25.4
Absent	91.3	8.7
Ability to hold down a job			8.17** (0.23)
Able	83.7	16.3
Unable	62.7	37.3
Criminality in family of origin			3.13 (0.14)
Absent	85	15
Present	73	27
relationship with nuclear family			10.5** (0.3)
Close and supportive	93	7
Reasonable	81.8	18.2
Partial	65.1	34.9

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

Table 4.

Averages (and Standard Deviations) of Parametric Variables and Their Relationship to Infractions of Terms of Supervision (N = 155).

Parametric variables	0-1 infractions	2+ infractions	t	df	Cohen’s d
Age of monitored individual	33.11 (10.05)	36.81 (11.13)	1.80*	146.0	0.4
Number of children	1.17 (1.63)	1.62 (1.70)	1.7	148.0	0.3
Years of schooling	10.18 (1.94)	9.64 (2.02)	1.3	134.0	0.3
Age at first incarceration	19.14 (5.88)	19.74 (6.64)	0.4	109.0	0.1
Length of last incarceration in months	Median (and interquartile range)	Median (and interquartile range)
	36 (32)	48 (38)

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

A significant difference was found in the distribution of the period of the last case of incarceration between the prisoners with zero or one infraction and the prisoner group with two or more infractions (Mann–Whitney U =1,484; n₁ = 112; n₂ = 35; p = .03). There was also a significant difference in average age of the prisoner at time of monitoring between those who did not violate the terms of their supervision or violated it at least once, and those who violated them two or more times, t(146) = 1.80, p < .5, Cohen’s d = 0.35.

Table 5 shows the results of a multivariate analysis for involvement in violations of the monitoring terms among all the participants.

Table 5.

Logistic Regression Results-Likelihood of Infractions (n = 123).

	B	SE	Odds ratio [95% CI]
Work-related problems (reference category = no problems)	1.7**	0.62	5.42 [1.6, 18.3]
Family criminal background (reference category = no criminal background)	1.5*	0.66	4.45 [1.22, 16.2]
Area of residence (reference category = center of country)
North	−2.1**	0.75	0.12 [0.03, 0.52]
South	−1.7*	0.71	0.19 [0.05, 0.77]
Age	0.06	0.04	1.06 [0.98, 1.2]
Total number of incarcerations	−0.54*	0.26	0.59 [0.35, 0.98]
Monitoring period	0.003	0.06	1.003 [0.9, 1.12]
Nationality (reference category =Arab Muslims)	−0.38	0.65	0.69 [0.19, 2.4]
Family status (reference category = single)
Married	0.063	0.87	1.07 [0.19, 5.8]
Divorced/widower	0.66	0.90	1.93 [0.33, 11.2]
Residency (reference category = origin family
Nuclear family	0.89	0.83	2.4 [0.48, 12.4]
Rented	0.74	0.87	2.1 [0.38, 11.5]
Drug addiction (reference category = no)	−0.52	0.63	0.60 [0.17, 2.04]
Last incarceration period	−0.004	0.006	0.99 [0.98, 1.008]

Note: Dependent variable = 1 for the prisoners who violate their monitoring terms at least twice, else 0; −2 log likelihood (−2LL) = 91.21; χ ² (14) = 27.3, p = .017; Nagelkerke R² = 32.2%; classification accuracy = 82.9%.

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

To identify what differentiates those who abide by the terms of their supervision from those who violate them, we included the following variables in a logistic regression model that examines the following variables: ethnic affiliation, residential area, family status, total number of incarcerations, length of most recent incarceration, drug addiction, employment problems, criminal family background, residency, and the monitoring period.

The results from Table 5 show that the odds of licensed prisoners experiencing employment difficulties were over 5.42 times more likely to violate their monitoring terms as compared with prisoners without employment difficulties.

The odds of a licensed prisoner from a family in which at least one other family member has a criminal background were over 4.45 times higher than those of prisoners with no criminal family background.

The results in Table 5 also show that prisoners with more prison sentences had a lower chance of violating their monitoring terms at least twice.

The area of residence also affects the likelihood of infractions. The odds of prisoners living in the North and South were 88% and 81% (respectively) less likely to violate their conditions than those living in the center.

In addition, the logistic regression analysis shows that ethnicity, family status, drug addiction, area of residence during the monitoring period, age, monitoring period, and last incarceration period do not significantly influence the odds for infractions.

Formal Responses Toward Parole Infractions

From interview with the EM coordinator on the topic of notifications and warnings (Interview with Mrs. Rotem Efodi, the Authority’s Northern District Supervisor, May 2, 2010), it emerges that not every infraction is necessarily reported to the Board. There are several stages that precede reporting to the Board and their deliberation on the revocation of the licensed release—the prisoner must be summoned and given a warning, and the process must include his caregiver’s recommendation to the Board to let him off with a warning. Whereas the warning can remain a matter that is between the supervisor and the monitored individual, a report obliges the Board to make a recommendation as to toughening the terms of the supervision, lengthening the period of the licensed release, or revoking it altogether. In their interviews, while validating the questionnaire, the supervision coordinators indicated that there were no guidelines for submitting the reports. Also, the decision to report to the Committee is a product of the totality of the prisoner’s ability to function and the supervisor’s discretion (Shoham et al., 2010). According to supervisors at the Prisoners Rehabilitation Authority,

Due to many differences between prisoners, as well as the multiple variables examined, they are required to examine each case individually . . . therefore it is very hard to stick to structured protocol, especially when they themselves haven’t been properly solidified. As professionals, we use our discretion to decide what is worthy of being reported to the Parole Board. (Interview with Efodi, May 2, 2010)

To examine the decision whether to warn the supervised individual, we constructed a dichotomous measure of warnings (similar to the measure of infractions), comprising two values: zero to one warning, and two or more warnings. Although we found a significant correlation, χ²(1) = 32.34, p < .001, φ= 0.72, between the measures of infractions and warnings, a graphic depiction of the relationship shows that the supervisor has broad discretionary powers on issuing a warning to the monitored individual. In all, 28% of the prisoners who violated the terms of their supervision two or more times received only one warning or none at all. There was a significant correlation, χ²(1) = 7.39, p < .01, φ = 0.35, between the measure of infractions and the decision whether or not to report an infraction to the Board. For 30% of the prisoners who violated the monitoring terms two or more times, the supervision coordinators decided not to report the infraction to the Parole Board.

The relationship between a decision to warn the prisoner and the decision to report him to the Board was significant, χ²(1) = 20.44, p < .001, φ = 0.57: In nearly all the cases (86%) of two or more warnings, the Board received a report that included a recommendation to hold a deliberation regarding the revocation of the prisoner’s license or toughening its terms. Tables 6 and 7 show only the significant correlations that were found between formal responses to infractions (warnings or reports to the Board) and sociolegal characteristics of the monitored individuals. In the rest of the variables we examined (appearing also in Tables 2, 3, and 4), there were no significant relationships with the measure of warnings or with the decision to report infractions to the Parole Board.

Table 6.

Significant Correlations Between Demographic, Legal, and Rehabilitation-Oriented Variables and Formal Responses (Warnings) to Infractions (n = 62).

Characteristics	0-1 warnings (%)	2+ warnings (%)	χ² (φ/Cramer’s V)
Nationality			6.96** (0.34)
Jewish	66.7	33.3
Arab	29.4	70.6
Criminality in family of origin			4.77* (0.30)
Absent	71	29
Present	41.7	58.3
Violation of curfew			4.8* (0.28)
Absent	68.6	31.4
Present	40.7	59.3

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

Table 7.

Significant Correlations Between Demographic, Legal, and Rehabilitation-Oriented Variables and Formal Responses (Reports) to Infractions (n = 62).

	0 reports	1+ reports
Characteristics	%	%	χ² (φ/Cramer’s V)
Nationality			4.42*(0.27)
Jewish	53.3	46.7
Arab	23.5	76.5
Family status			5.43* (0.3)
Single	60	40
Married	45	55
Divorced/separated/widowed	23.5	76.5
Personal impeding factors			6.74** (0.34)
Absent	68.4	31.6
Present	32.5	67.5
Quality of relationship with nuclear family			4.91* (0.34)
Close and supportive	66.7	33.3
Normal	50	50
Partial	26.1	73.9
Violation of curfew			4.65* (0.27)
Absent	57.1	42.9
Present	29.6	70.4
Prognosis			5.66* (0.30)
Positive	53.8	46.2
Negative	30	70
Future not mentioned	63.2	36.8

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

There were two additional significant variables on which we performed a t test. A significant correlation was found between age at first incarceration and number of warnings issued to the prisoner, t(21.96) = 2.16, p = .042, Cohen’s d = 0.71—participants first arrested at a later age committed more infractions. Another significant relationship was evident between the number of years of schooling of the monitored individual and the number of warnings that he received, t(49) = 2.27, p = .028, Cohen’s d = 0.64—participants with more years of schooling received fewer warnings.

Prompted by studies performed in Israel dealing with a lack of penal uniformity based on ethnic affiliation (for example, see Rattner & Fishman, 2004), we set out to examine whether this tendency also underlies the formal decisions that are made throughout the monitoring process. The preliminary analysis showed a significant correlation between the monitored individual’s ethnicity (Jewish as opposed to Muslim Arabs) and the average number of warnings, t(60) = 2.06, p = .043, Cohen’s d = 0.48. The average number of warnings given to Muslim Arab participants (M = 3.24, SD = 4.5) was far greater than that given to the Jewish ones (M = 1.56, SD = 1.99). In view of the significant relationship between the monitored individual’s ethnicity and the number of warnings, we examined each of the groups separately—Jewish and Muslim Arabs. In each of the groups, there was a significant positive correlation between number of infractions and number of warnings; however, the correlation for the Arabs was stronger. To examine whether there is a significant tendency for the supervisors to issue more warnings to the Muslim Arab offenders, we performed a one-way ANCOVA, comparing ethnicity and warnings, with the number of infractions serving as a mediating variable. This analysis revealed that although there was a significant relationship between number of infractions and number of warnings, F(1, 58) = 18.069, p = .000, there was no interaction between ethnicity and number of infractions, F(1, 58) = 0.165, p = .686. It seems from this analysis that controlling for the number of infractions, there was no significant difference between Jewish and Muslim Arabs, F(1, 58) = 0.148, p = .702, in the number of warnings that the prisoner received from the supervision coordinator.

The research also indicated a significant difference between the average number of warnings received by the prisoners with a criminal family background and those without such a background, t(53) = 1.78, p < .05, Cohen’s d = 0.46—the average number of warnings among the former (M = 2.92, SD = 4.04, n = 24) was significantly greater than the average for prisoners from normative families (M = 1.45, SD = 1.91, n = 31).

To eliminate the possibility that this difference might be related to the different levels of infractions between the two groups, we performed a one-way ANCOVA, controlling the number of infractions. It appears that, similar to the analysis comparing ethnic affiliation and warnings, there is no significant difference in the number of warnings between participants from a criminal family background and those who come from a normative background—controlling for the number of infractions, F(1, 52) = 0.696, p < .05. Here also, apparently the difference between the two groups is not related to the background from which they come, but rather to the number of infractions they commit.

Characteristics of the Monitored Individual and Reports to the Parole Board

Indeed, it seems that not every infraction leads to a report to the Parole Board. Minor infractions such as coming home a few minutes late or stepping out of the permitted zone will not always lead to a formal response. This section deals with whether there was a relationship between extralegal characteristics and the supervision coordinator’s decision to report an infraction to the Board, who has the authority to revoke the monitored individual from the monitoring program and return him to prison.

There was a significant difference in the average number of reports regarding the participant’s family status, F(2, 59) = 3.56, p < .05: single (M = 0.52, SD = 0.77, n = 25), married (M = 0.95, SD = 1.28, n = 20), and divorced/widowed/separated (M = 1.41, SD = 1.18, n = 17). The highest average number of reports appeared in the divorced/widowed/separated group. In a follow-up Bonferroni analysis, there was a significant difference between single and divorced/widowed/separated participants (p < .05). In this case too, it was necessary to determine whether the source of the significant differences was the participant’s family status, or perhaps each of the three groups differed significantly from each other in the number of infractions.

To this end, we performed another one-way ANCOVA, controlling for number of violations. This analysis revealed that although a significant relationship exists between the number of violations and the number of reports, F(1, 58) = 34.23, p < .001, η² = 0.37), controlling for average number of infractions, a significant difference in the number of reports between married, single, and divorced/widowed/separated prisoners remains, F(1, 58) = 4.18, p < .05, η² = 0.13. An examination of the relationship between ethnicity and average number of reports to the Parole Board revealed that the latter number was greater for Muslim Arab prisoners (M = 1.24, SD = 1.25, n = 17), compared with Jews (M = 0.78, SD = 1.04, n = 45). However, an analysis of the results showed that there was no significant difference in the average number of reports between Jewish and Muslim Arabs, t(60) = 1.46, p > .1, Cohen’s d = 0.4.

Discussion

The EM program for licensed prisoners in Israel, like various other Western countries, places great emphasis on the participation of the monitored individuals in employment and treatment programs (Albrecht, 2005; Gable & Gable, 2005; Gibbs & King, 2003; Padgett et al., 2006; Renzema & Mayo-Wilson, 2005). At the same time, in contrast to some of these countries, the Israeli program is intended primarily for a population of discharged prisoners that would not be suitable for early release, without the demanding supervision terms of this program.

The initial aims of this research were to examine the nature of licensed prisoner’s sociolegal characteristics and the level of their infractions during the EM program.

Even though the licensed prisoners have a relatively high level of involvement in criminal offenses prior to participation in the monitoring program, accompanied by a variety of behavior problems during their imprisonment and relatively long prison terms, the rate of their criminal activities during the program was relatively very low.

These low recidivism percentages (10%) while the prisoners were in the supervision program may be explained according to our findings, primarily by a positive, supportive relationship with the nuclear family and by stable ongoing employment (for the major impact of employment on successful reintegration, see Hucklesby, 2008; Travis & Petersilia, 2004; Zhang et al., 2006).

The low infringement level among the EM program participants can be attributed to a number of core variables. First, the participation in psychosocial treatment, apart from EM, enables treating the prisoner’s main issues, which may prevent a relapse into a criminal lifestyle (such as drug addictions). Second, forcing the released prisoner to be integrated into the normative workforce gives the prisoner a structured daily agenda and the ability to make an honest living, and finally, the coordination between all the elements of treatment, and the parole committee’s obligation to report, constitute a threat to the supervised and therefore acts as a catalyst for abiding by the rules and the treatment. Taxman, Byrne, and Tharnner (2002) raise the question of how to utilize the leverage of the criminal justice system in a manner that supports—rather than subvert—treatment goals. Even though our research population are not only drug abusing offenders, our finding supports the basic assumption of the HIDTA (High Intensity Drug Trafficking Areas) model (Taxman, 2002), designed to implement a seamless system to target hard-core offenders, who are both difficult clients for treatment providers and difficult offenders for community supervision agents. The seamless system is tailored to their own sociolegal environment, which includes system reforms consistent with the core components of the HIDTA model (Taxman & Bouffard, 2000). The positive outcomes of the Israeli EM project, which redefine the relationship bet-ween criminal justice and the treatment system, can be explained by all those key components—continuum of care, supervision, urinalysis testing, and compliance measures and graduated sanctions.

At this stage, we do not have a good theoretical explanation for the geographical differences found in the level of infractions. One possible explanation might be the level of permissiveness and the level of temptations in the central district which leads to higher level of infractions.

The second aim of this research was to examine whether variables, such as ethnicity or family status, are related to the nature of the formal response to monitoring violations. Although the penology literature examining judicial decisions indicates that judges give great consideration to variables such as normative background versus criminal past in making their decisions (see, for example, First & Agmon-Gonen, 2007; Shoham & Abulafia, 2010), in the present study, variables such as number of previous imprisonments, or length of most recent sentence, do not provide a significant explanation for the level of infractions. The results show that personal or situational variable such as an inability to cope with the demands of employment or having other family members involved in crime are better predictors for the risk of infractions. Most of the studies conducted in Israel regarding the lack of uniformity in punishment have focused primarily on variables such as ethnic affiliation and gender (Bogush & Don Yichye, 1999; Rattner & Fishman, 2004). Because the EM population are solely composed of males (a kind of discrimination by itself), we chose to address the question whether the ethnicity of a monitored individual bore any relationship to decisions made about him in cases of infractions. The comprehensive study by Rattner and Fishman (2004), for example, indicates that there are significant differences between Jews and Arabs in the probability of conviction and of receiving a prison sentence. The present study also indicates, prima facie, a greater tendency to issue a warning and a report of infractions among Arab prisoners on license, compared with the Jewish prisoners in the program.

As described early on, the EM coordinators in Israel have extensive discretionary power, which is likely to lead to discrimination which is attributable to variables such as ethnicity. However, statistically controlling for the number of infractions, we did not find a significant correlation between ethnicity and the decision to warn the monitored individual or to report an infraction to the Parole Board. It is important to note that most of the legal variables examined, such as the criminal past of the licensed prisoner, the type of crime that his last imprisonment was for, sentence length, or number of previous imprisonments, were not significantly related to the decision whether to give the individual a warning on account of his infraction. Our study shows that the most effective variable for explaining the decision whether to toughen the terms of the monitoring or to request its cancellation is an objective one—the number of infractions.

Our findings also suggest that the decision whether to request more stringent terms of supervision or cancellation of participation in the program, or perhaps to grant the individual another chance and to let him off with a warning, was mainly related to the existence of impeding personal variables like unreliability on the job, drug addiction, his family status, and the quality of his relationship with his nuclear family. It appears that the existence of a problematic relationship with one’s family members is likely to upset his situation to such an extent that he might return to criminal behavior. Hence, it obligates the organization to consider alternative rehabilitation programs (e.g., hostels) that place the prisoner at a distance from his family members. Another issue that rose by our research findings, which is likely to enhance the success of the program, is the intensiveness of the supervision. The findings show that prisoners who visit the EM coordinator relatively infrequently commit twice as many infractions as those who visit often. It seems that the close supervision and the intensive relationship contribute to the maintenance of stability and boundaries in EM; accordingly, it is recommended that a schedule of frequent visits to the EM coordinator be set.

The great emphasis placed on treatment within the framework of the EM program has led to a relatively high degree of discretionary power and subjective consideration of each case on its own merits, thereby making the continued participation of prisoners on license possible, even after one or more infractions of the terms of supervision. Nevertheless, despite this high degree of discretionary power on the part of the Rehabilitation Authority supervisors, we found a relatively low involvement in recidivism among the released prisoners on one hand, and no formal discrimination among the licensed prisoners on the other hand. In spite of the fact that the most efficacious variable found in this pilot research was an objective one, the great discretion enjoyed by the Prisoners’ Rehabilitation Authority reinforces the need for further studies that will assist to improve the criteria selection process for the EM project in Israel and establish clear guidelines to achieve a better homogeneity of supervisors across all districts.

This study relates to all licensed prisoners supervised between the years 2007 and 2009. Their number was relatively small and it made it difficult to achieve significant statistical conclusions. In addition, because the EM project in Israel is in its initial stages, we found a lot of technical infractions that reflected malfunctions in the monitoring system that were removed from the database.

As mentioned above, this is the first part of a larger study; therefore we cannot conclude yet whether the EM program reduces recidivism in the long term. The second and third stages of the study will examine the recidivism for a period of up to 3 years from the end of the program and examine the contentment of all parties involved in the program.

Unfortunately, during the last year, the Ministry of Internal Security in Israel has temporarily stopped the EM program for licensed prisoners pending further examination.

Based on the positive outcomes of this study, we recommend to carry on with the EM program for the licensed prisoners in Israel and to recognize the important role of the EM supervisors in reducing nonnormative behaviors during the program.

Footnotes

Declaration of Conflicting Interests

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

Funding

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

References

Ajzenstadt

Barak

(2008). Terrorism and risk management: The Israeli case. Punishment & Society, 10, 355-374.

Albrecht

(2005). Electronic monitoring in Europe: A summary and assessment of recent developments in the legal framework and the implementation of electronic monitoring. Retrieved from http://www.mpicc.de/shared/data/pdf/albrecht.pdf

Ben David

Fishman

Tsadok

(2006). Havnayat shikul hada’at hashiputi ba’anisha [The structuring of judicial discretion in punishment]. Jerusalem, Israel: The Knesset Research and Information Center.

Blomberg

Bales

Waid

(2008). Punishment and culture. In Shoham

S. G.

Beck

Kett

(Eds.), The international handbook of penology and criminal justice (pp. 1-15). New York, NY: CRC Press.

Bogush

Don Yichye

(1999). Migdar Ve mishpat [Court decisions and gender bios]. Jerusalem, Israel: The Insinuation for Jerusalem Research.

Bonta

Rooney

Wallace-Capretta

(1999). Electronic monitoring in Canada. Ottawa, Ontario, Canada: Public Works and Government Services Canada.

Brennan

P. K.

(2008). Race/ethnicity and sentencing outcomes among drug offenders in North Carolina. Journal of Contemporary Criminal Justice, 24, 371-398.

Crowe

A. H.

(2002). Offender supervision with electronic technology: A user’s guide. Lexington, KY: American Probation and Parole Association.

Diamant

(2006). Kovetz nehalim—Tipul Hareshut Leshikum Ha’asir banoseh pikuach elektroni [Procedures—The Authority’s handling of the subject of electronic monitoring]. Jerusalem, Israel: Prisoners Rehabilitation Authority.

10.

Downing

(2006). Emergence of Global Positioning Satellite (GPS) Systems in correctional applications. Corrections Today, 68, 42-45.

11.

Efodi

(2010). Tochnit shikum befikuach elektroni [An electronic monitoring rehabilitation program]. Jerusalem, Israel: Prisoners Rehabilitation Authority.

12.

Finn

M. A.

Muirhead-Steves

(2002). The effectiveness of electronic monitoring with violent male parolees. Justice Quarterly, 19, 293-312.

13.

First

Agmon-Gonen

(2007). Hamechonit shava yoter?! Ha’anisha ba’aveirot alimot neged benot zug [Is the car worth more?! Punishment in domestic violence offenses]. In Barak-Erez

Yanisky-Ravid

Bitton

Pugatch

(Eds.), Studies in law, gender, and feminism (pp. 545-582). Kiryat Ono, Israel: Sarigim-Li-on, Nevo Press, and the Ono Academic College.

14.

Gable

R. K.

Gable

R. S.

(2005). Electronic monitoring: Positive intervention strategies. Federal Probation, 69, 1-7.

15.

Garland

(2002). The culture of control: Crime and social order in contemporary society. Oxford, UK: Oxford University Press.

16.

Gibbs

King

(2003). “The Electronic Ball and Chain? The Operation and Impact of Home Detention with Electronic Monitoring in New Zealand.” Australian and New Zealand Journal of Criminology, 36, 1-17.

17.

Haverkamp

Mayer

Levy

(2004). Electronic monitoring in Europe. European Journal of Crime, Criminal Law, and Criminal Justice, 12, 36-45.

18.

Hawkins

P. F.

(Ed.). (1995). Ethnicity, race and crime: Perspectives across time and place. Albany: State University of New York.

19.

Hollin

(2001). To treat or not to treat? An historical perspective. In Hollin

(Ed.), The essential handbook of offender assessment and treatment (pp. 3-15). New York, NY: Wiley.

20.

Hucklesby

(2008). Vehicles of desistance? The impact of electronically monitored curfew orders. Criminology and Criminal Justice, 8, 51-71.

21.

Hucklesby

(2009a). Report of the 6th Electronic Monitoring Conference. Leeds, UK: Center for Criminal Justice Studies.

22.

Hucklesby

(2009b). Understanding offender’s compliance: A case study of electronically monitored curfew orders. Journal of Law and Society, 36, 248-271.

23.

Hudson

(2003). Understanding justice: An introduction to ideas, perspectives and controversies in modern penal theory. Buckingham, UK: Open University Press.

24.

Lahav

(2002). Yachas hama’arechet hamishpatit beyisrael likvutzot shonot ba’ochlusia [How the judicial system in Israel relates to various population groups]. Jerusalem, Israel: The Knesset Research and Information Center.

25.

Marklund

Holmberg

(2009). Effects of early release from prison using electronic tagging in Sweden. Journal of Experimental Criminology, 5, 41-61.

26.

McGuire

(2002). Criminal sanctions versus psychologically based interventions with offenders: A comparative empirical analysis. Psychology, Crime and Law, 8, 183-208.

27.

Ministry of Internal Security. (2006). The Minister of Internal Security’s policy toward the Israel Police in 2006. Jerusalem: Ministry of Internal Security.

28.

NAO. (2006). The electronic monitoring for adult offender. London, England: The Stationery Office.

29.

O’Tool

(2004). Study of factors affecting recidivism of offenders on electronic monitoring. Journal of Offender Monitoring, Winter/Spring, 5-21.

30.

Padgett

K. G.

Bales

W. D.

Blomberg

T. G.

(2006). Under surveillance: An empirical test of effectiveness and consequences of electronic monitoring. Criminology and Public Policy, 5, 61-91.

31.

Petersilia

(2004). What works in prisoner reentry: Reviewing and questioning the evidence. Federal Probation, 68, 4-8.

32.

Rattner

Fishman

(2004). Tsedek lakol? Yehudim ve’aravim bema’arechet hamishpat beyisrael [Justice for all? Jews and Arabs in the Israeli criminal justice system]. In Eden

Shadmi

Kim

(Eds.), Rodfei tsedek: Mechkarim bifeshia ve’achifat chok beyisrael [Seeking Justice: Research in Crime and Law Enforcement in Israel] (pp. 85-110). Tel Aviv, Israel: Cherikover and Beit Berl College.

33.

Renzema

Mayo-Wilson

(2005). Can electronic monitoring reduce crime for moderate to high-risk offenders? Journal of Experimental Criminology, 1, 215-237.

34.

Segal

Wagschal

(2004). Pikuach elektroni ba’olam [Electronic monitoring in the world]. Jerusalem, Israel: Ministry of Internal Security, Information Services Unit.

35.

Shoham

(2009). Hatzad ha’afel shel hashemesh [The dark side of the sun]. Beer Sheva, Israel: Ben-Gurion University of the Negev.

36.

Shoham

Abulafia

(2010). Hachlatot shiputiot be’aveirot alimot ben bnei zug bevatei mishpat shalom beyisrael [Judicial decisions in crimes of domestic violence in Magistrate’s Courts in Israel]. Megamot, 47, 103-129.

37.

Shoham

Yehosha

Efodi

Diamant

(2010). Shikum ba’azikim elektroni’im—Mechkar ha’aracha shel tochnit hapikuach ha’elektroni bekerev asirim berishi’on beyisrael [Rehabilitation through electronic shackling—An assessment study of the electronic monitoring program among prisoners released on license in Israel] (Report presented to the Authority). Jerusalem, Israel: Authority.

38.

Simon

Feely

M. M.

(1995). True crime: The new penology and public discourse on crime. In Blomberg

T. G.

Cohen

(Eds.), Punishment and social control (pp. 147-180). New York, NY: Aldine de Gruyter.

39.

Steffenmeier

Demith

(2000). Ethnicity and sentencing outcomes in U.S. federal courts. American Sociological Review, 65, 705-729.

40.

Steffenmeier

Demith

(2006). Does gender modify the effect of race/ethnicity on criminal sentencing? Journal of Quantitative Criminology, 22, 241-265.

41.

Stith

Cabanes

J. A.

(1998). Fear of judging: Sentencing guidelines in the federal courts. Chicago, IL: The University of Chicago Press.

42.

Taxman

F. S.

(2002). Supervision: Exploring the dimensions of effectiveness. Federal Probation, 66, 14-28.

43.

Taxman

F. S.

Bouffard

(2000). The importance of system in improving offender outcomes: Critical elements of treatment integrity. Justice Research and Policy, 2, 9-30.

44.

Taxman

F. S.

Byrne

J. M.

Tharnner

M. H.

(2002). Evaluating the implementation and Impact of a seamless system of care for substance abusing offenders—The HIDTA model (Final Report). Rockville, MD: National Criminal Justice Service.

45.

Travis

Petersilia

(2004). Reentry reconsidered: A new look at an old question. In Cole

Gretz

Bunger

(Eds.), The criminal justice system (pp. 415-434). Belmont, CA: Wadsworth.

46.

Uggen

(2000). Work as a turning point in the life course of criminals: A duration model of age, employment and recidivism. American Sociological Review, 67, 529-546.

47.

Ulmer

J. T.

Cramer

J. H.

(1996). Court communities under sentencing guidelines: Dilemmas of formal rationality and sentencing disparity. Criminology, 34, 383-408.

48.

Von Hirsch

(1986). Deservedness and dangerousness in the sentencing policy. Criminal Law Review, 33, 79-91.

49.

Walker

Spohn

Delone

(2000). Color of justice: Race, ethnicity and crime. Belmont, CA: Wadsworth.

50.

Weisburd

Shoham

Barak

Manspeizer

Gideon

(2005). Chazara la’avaryanut bekerev asirim meshuchrarim bogrei merkaz hashikum ha’artzi bekeleh haSharon 1994-1997 [Recidivism in discharged prisoners who have completed the program of the National Center for Rehabilitation at the Sharon Prison 1994-1997]. Jerusalem: Israeli Anti-Drug Authority and Authority.

51.

Zhang

S. X.

Roberts

Callanan

V. J.

(2006). Preventing parolees from returning to prison through community-based reintegration. Crime and Delinquency, 52, 551-571.