Sage Journals: Discover world-class research

Abstract

Informed by social support theory, this study examined gender differences of revocation and relapse of 526 probationers diagnosed with a substance use disorder, mental health disorder, or co-occurring disorder who were court-ordered for treatment. Within 18 months of starting probation, 51% of men and 56% of women were revoked. Although not statistically significant, the relapse rate was 34% for women and 26% for men. Significant predictors of revocation were similar across gender with a lack of family support being the strongest predictor for both groups, though notably stronger for women. Family support was a significant predictor of relapse for women but not for men. Inpatient treatment of at least 180 days had the greatest effect in reducing relapse and revocation for women. Treatment dosage of at least 90 days had modest effects in reducing men’s revocation, but the treatment effect was less pronounced for men’s relapse rates.

Keywords

probation relapse recidivism social support substance use disorder mental health disorder gender differences

Probation supervision remains the most common sanction in the criminal justice system to address offender substance use and mental health needs. In 2021, nearly 3 million adults were supervised on probation in the United States, which represents more than 3.5 times as many people released annually from prison to community supervision (Carson & Kluckow, 2023; Kaeble, 2023). A considerable number of men and women on probation have either a substance use disorder (SUD), suffer from a serious mental illness (SMI), or simultaneously experience both substance use and mental health disorders, also known as a co-occurring disorder (COD). Research suggests that the presence of one or more of these disorders increases the likelihood of revocation and relapse during supervision (Gu et al., 2023; Wolff et al., 2014). Although some studies have examined within-group differences of offenders with SUD, SMI, and COD (Houser et al., 2019), even fewer studies have examined gender differences within these three groups.

Notably, scholars have taken two different theoretical approaches to examine the gender differences of SUD and SMI offenders. First, the gender pathways approach supports the idea that women face different challenges and risks to relapse and recidivism than men (Daly, 1992; Jeffries et al., 2019; Salisbury & Van Voorhis, 2009). The second theoretical framework is social support (Albrecht & Adelman, 1987; Cullen, 1994) that emphasizes the importance of relationships and open communication, which is particularly important during supervision for offenders with SMI (Epperson et al., 2017). Still, the research remains unclear the extent to which resource differences for women on probation such as less employment experience and lower education levels, affect women’s ability to achieve positive behavioral outcomes (Curcio & Pattavina, 2018; O’Brien, 2002; Peters et al., 1997), and also the effect that social support has on revocation and relapse.

The current study examined the key risk factors with men and women on specialized probation who had a SUD, a SMI diagnosis, or a COD to determine whether gender differences exist with respect to relapse and revocation, and second, the influence of social support on probationers. In the next section, we review the literature on the prevalence of SMI and SUD found among probationers in the United States. We then examine what is currently known about specialized probation and gender differences relative to SUD and SMI as it affects relapse and recidivism. Finally, we introduce the gender pathways to crime perspective and social support theory as potential explanations for our data.

Prevalence of Substance Use and Mental Health Disorders on Probation

Drug and alcohol dependency/use is a common problem for justice-involved individuals. Nearly half (47.5%) of all individuals who had been on probation supervision in the last year met the criteria for SUD, according to national statistics (Substance Abuse and Mental Health Services Administration, 2021, Table 8.39b). Other research has estimated the SUD rate to be higher; between 60% and 80% for those on community supervision (Galvin et al., 2022; Reichert & Gleicher, 2019).

The prevalence of SMI was 8% for men and 22% for women on probation, which is twice the percent found in the general population (Substance Abuse and Mental Health Services Administration, 2021, Table 6.16b). Thus, it is not surprising to find individuals on community supervision who have a COD, where both mental health and substance abuse coexist as a major problem. For example, between 52% and 74% of justice-involved individuals with an SMI also had co-occurring SUD problems (Hartwell, 2004; Modestin & Wuermle, 2005). Research found that individuals with COD are more likely to recidivate than individuals with only one of these risk factors (Hartwell, 2004; McReynolds et al., 2010; Messina et al., 2004). Initial examination of this issue found that risk factors other than SMI played a more direct role in recidivism (Skeem et al., 2014). However, others have argued continued alcohol or substance abuse may be a significant risk factor (Castillo & Alarid, 2011) suggesting the inclusion of relapse alongside revocation as an outcome variable should be further explored.

Specialized Probation for Substance Use and Mental Health Disorders

Specialized probation caseloads became the preferred method to supervise clients with mental health and substance use needs. Probation officers were trained to build rapport and use problem-solving strategies over formal revocation found in traditional probation (Skeem et al., 2006; Van Deinse et al., 2021). Studies comparing specialized probation caseloads to traditional supervision found that clients on specialized probation completed more treatment sessions and were more likely to adhere to medication regimens. However, specialized probationers committed more technical violations than clients on traditional probation (Manchak et al., 2014). This difference may be due in part to the clients’ exposure to cognitive-behavioral treatment that can be effective in changing attitudes and behaviors (Landernberg & Lipsey, 2005). Two independent studies found specialized probation clients had fewer revocations and rearrests for new crimes than similar probationers on traditional supervision (Skeem et al., 2017; Wolff et al., 2014). More recently, Van Deinse and colleagues (2022) found the opposite conclusion in a randomized control trial where clients on specialized probation were more likely to commit a new crime than those on standard probation despite having similar rates of technical violations.

Previous SUD and/or SMI research with justice-involved, community-supervised individuals suffer from two main limitations: First, most studies lack treatment dosage data to measure the number of days in treatment. Second, studies typically include mental health or substance use, and rarely attempt to understand their independent effects in the same model. The closest known study (Houser et al., 2019) compared four groups of Pennsylvania parolees diagnosed with SUD, SMI, COD, and no known problems. The researchers found that the SMI and COD parolees were more likely to commit a new crime over the SUD and regular parolee groups. Houser et al. (2019) also found that the probability of being revoked for a new crime or technical violations was higher for men. However, the type of treatment and dosage were not available. It is clear that research has not yet fully explored important gender differences that may exist within specialized probation groups.

Gender Comparisons of Relapse and Recidivism

Most justice-involved, gender comparisons of relapse found that women at various stages in the system were less likely than men to relapse (Gil-Rivas et al., 1996; Grella et al., 2005; Inciardi et al., 2004), or no gender differences existed (Nielsen et al., 1996). While most relapse research concentrated on prison therapeutic communities, success in the community for women relied on their active participation and completion of treatment (Fiorentine et al., 1997; Green et al., 2004). Relapse was indeed a primary predictor of arrest for a new crime (Kopak et al., 2016), with men being more likely than women to be rearrested (Olson et al., 2003). In non-criminal justice populations, women generally relapsed less often than men, but the reasons for these gender differences were inconclusive (Walitzer & Dearing, 2005).

Gender Pathways to Crime

The gender pathways to crime perspective argues that women offenders engage in crime for different reasons than men (Daly, 1992). Although women’s relationships can be potential motivators for or inhibitors against criminal behavior (Wright et al., 2013), many crime trajectories begin with early trauma and victimization experiences, and develop into substance use, and/or mental health problems as coping mechanisms (Gehring, 2016; Wright et al., 2012). For example, justice-involved women are more likely than men to abuse alcohol and drugs as a coping mechanism for childhood abuse, trauma, and family dysfunction (Karlsson & Zielinski, 2020; Langan & Pelissier, 2001; Leban & Gibson, 2020; Messina et al., 2007; Pelissier & Jones, 2005; Spjeldnes & Goodkind, 2009). This also holds true for women who have never been in trouble with the law (Heffner et al., 2011). In general then, Green and colleagues (2004) found mental health severity predicts continued substance use (Green et al., 2004). Along these same lines, the gender pathways perspective would advocate correctional programming that is gender-responsive, or more likely to address these gender differences (Holtfreter & Morash, 2003).

Social Support Theory

Social support theory emphasizes the importance of relationships and open communication with the goal of reducing uncertainty, reassuring individuals, and increasing personal control in stressful situations (Albrecht & Adelman, 1987). In criminology, social support is premised on the idea that individuals are driven by a need to connect with others informally and formally in the community, through social networks, and through intimate relationships that help them to be successful (Cullen, 1994). Social support is a useful theoretical framework to explain both relapse (Latkin et al., 1999; McMahon, 2001) and criminal behavior (Cullen, 1994). Authors collectively agree that there are different types of social support. For example, instrumental social support is known as providing monetary assistance, a service or tangible material goods to someone else. Informational social support is providing advice or guidance to another person, while expressive social support focuses on affirming one’s self-worth and/or allowing another person to vent or express themselves emotionally.

Sources of social support originate from networks such as family, friends, therapy groups, and work colleagues (Albrecht & Adelman, 1987). For justice-involved individuals, additional support can emanate from supervision officers and social service organizations (Liu et al., 2023). Social support from family and friends has mixed results because the quality and consistency of these support sources can encourage or aid in conformity (Berg & Huebner, 2011; Cullen, 1994; Spjeldnes et al., 2012; Taylor, 2015). Dysfunctional, coercive, or stigmatizing social networks promote alcohol and drug use for women (Connors et al., 1998; Dobkin et al., 2002). This is consistent with why women drug users report more family dysfunction than men (Knight et al., 2001). Conversely, the absence of family more often predicts continued drug use for men (Green et al., 2004; Walitzer & Dearing, 2005).

Participation in group treatment can be a powerful source of social support, but there is a dearth of studies assessing treatment effects by gender. Through reciprocal storytelling and listening to others with similar problems, group sessions allowed participants to feel heard and to help others in their own recovery process (Arntson & Droge, 1987). Research has consistently shown that completing at least 3 months of substance abuse treatment is the minimum to make a difference on community supervision (Houser et al., 2012; Inciardi et al., 2004; Latessa et al., 2020; Taxman & Belenko, 2012). However, studies were mixed on whether inpatient or outpatient treatment would be more effective with a justice-involved group with mental health and/or SUDs. The only studies we found examined jail and prison-based custodial treatment rather than a community-based inpatient program One such study found that a jail-based treatment program with aftercare showed no significant differences in recidivism rates compared with just aftercare for both groups of individuals with COD (Chandler & Spicer, 2006). In a different study of in-custody treatment, incarcerated men with COD were randomly assigned to either a modified therapeutic community for both substance abuse and mental health or to a mental health treatment program. The modified therapeutic community group had lower reincarceration rates than the mental health treatment alone group. The difference was even lower with the therapeutic community group that opted to continue with aftercare following prison release (Sacks et al., 2004).

Studies assessing the effect of drug treatment with individuals on probation are inconsistent, and most lack a gender comparison. Successful completion of substance abuse treatment reduced recidivism for clients on community supervision (English & Mande, 1991; Hartman et al., 1994; Hiller et al., 2006), especially when compared with no treatment at all (Warner & Kramer, 2009). Other studies found no appreciable relapse difference for drug court versus regular probation (Deschenes et al., 1995) or group therapy plus Cocaine Anonymous versus Cocaine Anonymous alone (Wagoner & Piazza, 1993).

The Present Study

In the current study, clients on probation were court-ordered to complete an inpatient and/or outpatient treatment program determined by the severity of their problems. The following treatment modalities were available: (a) an inpatient SUD treatment program followed by aftercare, (b) an inpatient SMI/COD treatment facility followed by aftercare, and (c) a weekly outpatient only. The outpatient treatment location was also the aftercare provider for all three groups that offered sessions specific to each individual’s primary need.

The current study adds to the existing research in two ways. First, we examine whether gender differences exist with respect to relapse and revocation for clients on probation who had a SUD, a SMI diagnosis, or a COD. This analysis is guided by two overarching research questions. To what extent do predictors of revocation and relapse overlap across men and women probationers associated with this sample? For those variables that overlap, to what extent does the magnitude of their effect differ by gender? We hypothesize that women on probation will be more positively influenced by family support than men on probation. We also predict that involvement in treatment will act as a protective factor to reduce relapse and revocation for both men and women, but that the effect will be greater for women. Third, we examine the extent to which inpatient treatment dosage at 90 and 180 days or greater has an effect on revocation and relapse. Consistent with social support theory, we hypothesize that longer exposure to inpatient treatment will lead to more significant reductions in relapse and revocation compared with the no treatment group.

Method

Participants

All cases originated from a single county-level probation department located in a large urban metropolitan area in a southern state. Data were obtained for 526 adult men and women who were assessed and diagnosed for Axis 1 mental health problems (bipolar, schizophrenia, anxiety) and/or SUD. Of these cases, 428 clients were court-ordered to receive treatment: inpatient treatment only, inpatient followed by outpatient, or outpatient treatment only. The “no treatment” group (n = 98) was comprised of traditional probation clients who were diagnosed with a mental health or SUD but were court-ordered directly to traditional probation without treatment. The “no treatment” group would have qualified for treatment, but treatment facilities were not available in the jurisdiction at the time.

About 18% of the sample were identified as high risk, with 54% moderate and 28% low risk. Alternatively, about 25% were identified as high need, with 48% moderate and 27% low need. SUD only was the primary need for 39%, 14% had a serious mental health disorder only, and 47% had co-occurring SUD and SMI. The current offense was felony level 65% of the time, and 35% were convicted of a misdemeanor. Possession/sale of a controlled substance was the most common offense (37%) followed by property offenses (25%), violent crimes (19%), alcohol-related offenses (11%), and public order/nuisance offenses at 8.7%. About 26% of the sample possessed at least one prior felony and 69% had at least one prior misdemeanor. For individuals who had prior criminal histories, the average number was two felonies and three misdemeanors. Over 55% of the probationers did not possess a high school degree or equivalent, including a General Educational Development (GED). Eight out of ten probationers were single or divorced, 57% were Latina/Latino, and 16% were African American. The average age was about 32 with a standard deviation of 11 years. The gender composition was 41% women and 59% men.

Procedure

Data were collected through criminal history databases and probation case management files and de-identified prior to providing for research purposes. One of the authors conducted a site visit of both inpatient facilities and examined program curricula. Random assignment was not possible given that all clients had been sentenced with court-ordered conditions. All individuals in the sample had a primary diagnosis of a major mental health disorder, a severe SUD, or a co-occurring mental health and SUD. Outcomes were tracked only from an individual’s probation start date through 12 to 18 months of the probation discharge date. It was not possible to track outcomes beyond the probation discharge date.

Data Analytic Plan

The goal of this study was to determine if predictors of revocation and relapse differ, or differ in magnitude, by gender. As Cullen et al. (2021) have noted, reviews of past research have generally found considerable overlap among predictors of men’s and women’s criminality, but some factors have been shown to possess a larger effect by gender. We hypothesized that the social support indicators would be more likely to reduce relapse and revocation for women than men. To observe patterns in revocation and relapse by gender in this sample, frequency and descriptive data were compiled, bivariate analyses were conducted to identify initial differences between the two populations, and area under the curve (AUC) statistics were computed, bifurcated by gender for each dependent variable. SPSS was used to compute all statistics. In terms of initial bivariate analyses, phi coefficients (φ) and point-biserial correlations (r_pb) are performed. Effect sizes of .10 are considered small, .24 moderate, and .37 large (Rice & Harris, 2005). The AUC statistic provides a measure of magnitude in terms of a given independent variable’s impact on a binary outcome of interest (i.e., revocation, relapse). The AUC ranges from 0 to 1, with a value of .50 indicating the variable of interest performs no better than random chance. In terms of effect size, Rice and Harris’ (2005) standards are commonly referenced elsewhere (see also Ellis, 2010; Hamilton et al., 2020). They suggest an AUC value <.56 should be considered negligible, a value of .56 to .63 small, .64 to .70 moderate, and a value over .70 is considered a large direct effect.

Finally, the impact of inpatient treatment dosage on revocation and relapse is also explored. Prior literature has demonstrated a significant improvement in recidivism reduction with at least 3 months of treatment (Houser et al., 2012; Inciardi et al., 2004; Latessa et al., 2020; Taxman & Belenko, 2012). The number of days in treatment is categorized into multiple dichotomous variables in relation to probationers that received no inpatient treatment: less than 90 days, 90 days or more, and 180 days or more of treatment. Phi coefficients are produced for each combination in relation to revocation and relapse.

Measures

Revocation

Probation discharge data distinguished whether the termination was due to successful completion of probation (= 0), or a revocation for a technical violation or a new arrest (= 1). Revocation was one of two dependent variables in the study.

Relapse

Relapse was a second dependent variable in the study, confirmed by a positive urinalysis test indicating that the individual had used alcohol or an illicit drug (not prescribed by a physician) sometime during probation supervision. The relapse measure also included probationers who admitted to their probation officer that they used alcohol or illicit drugs (=1).

Risk Score

Risk score was defined as the probability or likelihood of re-offending that was numerically computed using a revised version of the Wisconsin risk assessment tool. The risk scores were based on an 11-item questionnaire that factored in items such as the number of prior adjudications, number of prior revocations, motivation to change, number of address changes, percent of time employed in the last 12 months, among other items. Risk scores ranged from 1 to 40, with 0 to 11 = low risk; 12 to 24 = moderate risk; and 25 to 40 defined as high risk. Cutoffs did not differ by gender, race, or ethnicity.

Needs Score

The needs score was a numerical indicator of each individual’s number and severity of intervention areas that may be related to criminal behavior and/or treatment areas that should be addressed during probation. Needs scores were computed using a revised version of the Wisconsin needs assessment tool, and based on a 12-item questionnaire that asked about academic/vocational skills, employment, companions, emotional stability, financial management, peer relationships, physical health, etc. Needs scores ranged from 0 to 43, with 0 to 14 = low needs; 15 to 29 = moderate needs, and 30 to 43 defined as high needs, with no distinctions by gender, race, or ethnicity.

Treatment Type

No matter the diagnosis (SMI, SUD, or COD), there were three types of treatment: outpatient only, inpatient only, and inpatient followed by outpatient. Dichotomously coded variables were created for each type. All inpatient and outpatient treatment programs were staffed by licensed counselors from the same agency. Programs were identical for women and men participants and did not appear to incorporate gender-responsive approaches. The inpatient SUD program offered Thinking for a Change cognitive-behavioral curriculum (Bush et al., 2016) and substance abuse counseling in a community-based facility. The inpatient SMI and COD individuals were housed together in a different facility than SUD clients and had different group therapy schedules. The SMI clients received mental health counseling and medication management. The COD clients received all four services: mental health counseling, medication management, Thinking for a Change, and substance abuse counseling.

Inpatient Treatment Days

The number of days of residential treatment was tracked and divided into several dichotomous variables including no treatment (= 0) compared with those with 90 or more days of treatment (= 1), those with less than 90 days compared (= 1) to those with 90 or more (= 0), no treatment (= 0) compared with those with 180 days or more (= 1), those with less than 90 (= 0) to those with 180 or more (= 1), and those with 90 to 179 days (= 0) compared with those with 180 or more days of inpatient treatment (= 1).

Serious SUD

Substance use severity was measured using the Texas Christian University (TCU) drug screen which defines the need for drug treatment based on a 9-item instrument. The TCU Drug screen has been shown to be valid and reliable for recommending the type of treatment (Knight et al., 2002). Outpatient treatment was recommended for clients who scored between 4 and 6 points, while residential treatment followed by outpatient treatment was recommended for individuals who scored 7 and 9 points. This variable is represented by a dichotomous variable indicating the presence of a moderate (= 0) or serious (= 1) drug problem.

Current Offense Level

Current offense was dichotomously coded by level of severity (misdemeanor = 0, felony = 1).

Nature of Current Offense

The primary type of behavior (violent, property, drugs, alcohol, and public order offenses) associated with the offense was also documented. These were dummy-coded for each type of offense.

Prior Criminal History

Prior criminal history was measured by the number of misdemeanor and felony convictions accrued prior to the current offense. Two dichotomous measures were also created as indicators of any prior felonies (= 1), or any prior misdemeanors (= 1) compared with those that had no such history.

Social Support

Social support variables included lack of family support (0 = support is present; 1 = support is absent) at the time of the initial probation assessment. A lack of family support was present if the client reported that immediate family members were not supportive, or the home environment was not functional, safe, and/or stable. Marital status was coded as 0 = married and 1 = single/divorced.

Other Variables

Years of completed education was coded into 0 = high school degree or GED equivalent; 1 = less than high school degree. Employment while on specialized probation (0 = no; 1 = yes) was measured only after release from inpatient and/or during outpatient treatment. Age was measured at the time the client was on probation. Race/ethnicity was dummy coded into two separate variables with if Latina/Latino = 1 and if African American = 1.

Results

Revocation and Relapse Outcomes

As reported in Table 1, about 54% of all probationers had their supervision revoked. This proportion was similar for women and men at 51% and 56%. Alternatively, 29% of all probationers experienced a drug relapse. Although this proportion varied slightly between women and men at 34% and 26%, the difference was not statistically significant.

Table 1:

Frequency and Descriptive Data of Study Variables

Variable	Overall (N = 526)		Women (n = 217)		Men (n = 309)		φ/r_pb
Variable	n	%	n	%	n	%	φ/r_pb
Revocation	285	54.2	111	51.2	174	56.3	.05
Technical violation(s)	162	30.8	75	34.6	87	28.2
New crime	115	21.9	32	14.7	83	26.9
Other (e.g., transfer)	8	1.5	4	1.8	4	1.3
Drug relapse	153	29.1	73	33.6	80	25.9	−.08
Primary need
Substance use disorder	204	38.8	70	32.3	134	43.4	.09*
Serious mental health disorder	73	13.9	38	17.5	35	11.3	−.11*
Co-occurring disorder	249	47.3	109	50.2	140	45.3	−.05
Risk level
High (25+)	93	17.7	38	17.5	55	17.8
Moderate (12–24)	286	54.4	126	58.1	160	51.8
Low (0–11)	147	27.9	53	24.4	94	30.4
Mean risk score (SD)	526	16.5 (7.8)	217	17.0 (7.8)	309	16.1 (7.9)	−.06
Needs level
High (30+)	129	24.5	73	33.6	56	18.1
Moderate (15–29)	253	48.1	106	48.8	147	47.6
Low (0–14)	144	27.4	38	17.5	106	34.3
Mean needs score (SD)	526	21.6 (10.0)	217	23.8 (9.2)	309	20.1 (10.2)	−.18***
Treatment type/dosage
No treatment	98	18.6	35	16.1	63	20.4	.05
Outpatient only	93	17.7	10	4.6	83	26.9	.29***
Inpatient + outpatient	115	21.9	62	28.6	53	17.2	−.14**
Inpatient only	220	41.8	110	50.7	110	35.6	−.15***
Inpatient treatment days (SD)	335	170.8 (109.3)	172	183.2 (112.2)	163	157.7 (104.8)	−.26***
Inpatient <90 days	51	9.7	23	10.6	28	9.1	.05
Inpatient ≥90 days	284	54.0	149	68.7	135	43.7	−.05
Inpatient ≥180 days	109	20.7	63	29.0	46	14.9	−.09
Current offense level							.11**
Felony	342	65.0	127	58.5	215	69.6
Misdemeanor	184	35.0	90	41.5	94	30.4
Nature of current offense
Violent	99	18.8	31	14.3	68	22.0	.10*
Property	129	24.5	60	27.6	69	22.3	−.06
Drugs	195	37.1	73	33.6	122	39.5	.06
DWI/alcohol	57	10.8	31	14.3	26	8.4	−.09*
Public order/nuisance	46	8.7	22	10.1	24	7.8	−.04
Prior felony(ies)	134	25.5	48	22.1	86	27.8	.07
M (SD)		1.6 (1.0)		1.6 (0.9)		1.6 (1.1)
Prior misdemeanor(s)	362	68.8	151	69.6	211	68.3	−.01
M (SD)		3.3 (2.9)		3.0 (2.6)		3.5 (3.0)
Education mean (SD)	526	10.9 (2.0)	217	10.8 (2.0)	309	11.0 (1.9)	.05
Less than high school degree	287	54.6	119	54.8	168	54.4	−.01
Employed while on probation	221	42.0	83	38.2	138	44.7	.06
Serious substance use disorder	205	39.0	91	41.9	114	36.9	−.05
Single/divorced	427	81.2	182	83.9	245	79.3	−.06
Lack of family support	230	43.7	104	47.9	126	40.8	−.07
Demographics
Latina/Latino	298	56.7	114	52.5	184	59.5	.07
African American	82	15.6	33	15.2	49	15.9	.01
M_age (SD)	526	31.7 (10.7)	217	33.5 (10.2)	309	30.4 (10.8)	−.15***

Note. Risk and needs scores are based on the Wisconsin risk/needs tool, cutoffs did not differ by gender, race, or ethnicity. The presence of a serious drug problem was based on the Texas Christian University (TCU) drug screen. Statistically significant bivariate results by gender were examined using phi coefficient (φ) or point-biserial correlation (r_pb). According to Rice and Harris (2005), effect sizes of .10 are considered weak, .24 moderate, and .37 large (see also Ellis, 2010; Hamilton et al., 2020). DWI = Driving While Intoxicated.

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

Bivariate Gender Comparisons

The sample was categorized according to three primary needs: SUD, SMI, and COD. SUD was more likely to be a primary need for men whereas SMI was more likely to be associated with women probationers, though each was associated with a significant but weak correlation. There were no significant gender differences for COD.

In terms of risk classification, a large proportion of individuals fell under the moderate category (48%–58%), with about 17% defined as high risk. Risk was not significantly different across gender. Note, however, the proportion of high need women was 34% compared with 18% of men with high needs. Women had more needs requiring treatment or intervention than men, so a higher percentage of women qualified for inpatient treatment and spent more days in treatment. Conversely, men participated in more outpatient treatment. The differences in number of inpatient days in treatment and prevalence of outpatient treatment were associated with significant and moderate effect sizes.

In terms of current offense, men were more likely than women to be on probation for felony and violent crimes. Women were more likely to be driving while intoxicated or on probation for alcohol-related offenses. The associations between gender and current offense, while statistically significant, were weak. There were no significant differences observed in terms of prior criminal history, consistent with the similar risk levels reported previously. The social support variables were not significantly different by gender. In terms of demographics, men and women were similar according to race and ethnicity but differed significantly in terms of age. On average, women were about 3 years older than the men, represented by a small effect size.

Bivariate Prediction of Revocation and Relapse by Gender

Table 2 reports the AUC results by gender for each independent variable in relation to revocation and relapse. The aim of these analyses is to identify statistically significant predictors, assess the extent to which these predictors are similar or different across men and women probationers, and determine to what extent the magnitude of a given variable is more prominent for one group over the other.

Table 2:

AUC Results for Revocation and Relapse by Gender

Independent variables	Dependent variables
	Women(n = 217)				Men(n = 309)
	Revocation		Relapse		Revocation		Relapse
	AUC	[95% CI]	AUC	[95% CI]	AUC	[95% CI]	AUC	[95% CI]
Assessment
Risk score	.76***	[.70, .83]	.63**	[.56, .71]	.74***	[.69, .80]	.63**	[.56, .70]
Needs score	.69***	[.62, .76]	.62*	[.54, .70]	.69***	[.64, .75]	.65***	[.58, .72]
Primary need
Substance use disorder	.57	[.49, .65]	.61**	[.54, .69]	.53	[.47, .60]	.55	[.48, .62]
Serious substance use	.62**	[.54, .69]	.64**	[.56, .72]	.59**	[.53, .65]	.55	[.47, .62]
Mental health disorder	.40**	[.32, .47]	.38**	[.30, .46]	.52	[.45, .58]	.49	[.42, .56]
Co-occurring disorder	.47	[.32, .54]	.49	[.41, .58]	.55	[.48, .61]	.54	[.47, .61]
Treatment type/dosage
No treatment	.53	[.45, .61]	.44	[.36, .52]	.53	[.47, .59]	.44	[.37, .51]
Outpatient only	.48	[.40, .56]	.50	[.42, .58]	.40**	[.33, .46]	.50	[.43, .58]
Inpatient + outpatient	.33***	[.26, .40]	.41*	[.33, .49]	.41**	[.34, .47]	.50	[.43, .58]
Inpatient only	.66***	[.59, .74]	.66***	[.58, .73]	.67***	[.60, .73]	.56	[.48, .63]
Days of inpatient treatment	.34***	[.26, .41]	.47	[.39, .55]	.49	[.42, .56]	.52	[.45, .59]
Nature of current offense
Felony level	.65***	[.57, .72]	.63**	[.55, .70]	.59**	[.53, .66]	.50	[.43, .58]
Violent	.49	[.42, .57]	.44	[.37, .52]	.57*	[.51, .63]	.54	[.46, .61]
Property	.45	[.37, .52]	.47	[.39, .55]	.49	[.43, .56]	.52	[.44, .59]
Drug	.59*	[.51, .67]	.62**	[.54, .70]	.49	[.42, .55]	.50	[.43, .58]
DWI/alcohol	.44	[.36, .51]	.48	[.40, .56]	.46	[.39, .52]	.45	[.42, .57]
Public order/nuisance	.54	[.46, .61]	.50	[.42, .58]	.49	[.43, .56]	.49	[.42, .56]
Criminal history
If any prior felonies	.55	[.47, .63]	.53	[.45, .61]	.62***	[.56, .68]	.50	[.42, .57]
If any prior misdemeanors	.58*	[.51, .66]	.55	[.47, .63]	.58*	[.52, .65]	.51	[.44, .59]
Social support factors
Single/divorced	.47	[.40, .55]	.48	[.40, .56]	.50	[.44, .57]	.52	[.45, .59]
Lack of family support	.75***	[.68, .81]	.68***	[.60, .75]	.67***	[.61, .73]	.57	[.50, .59]
Other variables
Less than high school degree	.59*	[.52, .67]	.60*	[.52, .68]	.65***	[.59, .71]	.56	[.48, .63]
Employed	.38**	[.30, .45]	.40*	[.32, .48]	.41**	[.35, .47]	.49	[.41, .56]
Latina/Latino	.52	[.44, .59]	.52	[.44, .60]	.57*	[.50, .63]	.55	[.47, .62]
African American	.49	[.42, .57]	.48	[.40, .56]	.49	[.42, .56]	.53	[.45, .60]
Age	.38**	[.31, .46]	.44	[.36, .52]	.37***	[.31, .44]	.42*	[.35, .49]

Note. Risk and needs scores are based on the Wisconsin risk/needs tool. According to Rice and Harris (2005), AUC values <.56 would be considered negligible, .56 to .63 small, .64 to .70 moderate, and over .70 large (see also Ellis, 2010; Hamilton et al., 2020). AUC = area under the curve; CI = confidence interval; DWI = Driving While Intoxicated.

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

As anticipated, a lack of family support was consistently one of the strongest predictors. For women, it possessed a large effect on revocation and a medium effect on relapse, similar in magnitude to the risk score. For men, family support possessed a medium effect on revocation but was not significant for relapse. For women, six other predictors possessed at least a moderate effect size in association with revocation; risk score, needs score, if inpatient and outpatient treatment, if inpatient only, days of inpatient treatment, and if current offense was felony level. In terms of treatment, inpatient only was associated with an increased likelihood of revocation while inpatient with outpatient treatment was associated with a decrease in revocation. Furthermore, the greater the number of days in treatment, the lower the likelihood of revocation. Second, if her current offense was felony level, women probationers had a higher likelihood of revocation than if her current offense was a misdemeanor.

For men, education and inpatient only were the only significant predictors with at least a moderate effect size. Men who possessed less than a high school degree were more likely to be revoked from probation. Likewise, men probationers who participated only in inpatient treatment were more likely to experience a revocation. Note, those who participated in inpatient with outpatient treatment experienced a reduction in revocation, but the effect size was small. Significant predictors for men and women overlapped on 11 variables of varying effect size. They differed on the following seven variables, with the first three for women: If the current offense was drug-related, if had a mental health disorder and days of inpatient treatment. The last four variables favored men: if any prior felonies, if the current offense was violent, if Latino, and if participated in outpatient treatment only. With the exception of days of inpatient treatment, all of these diverging predictors had small effect sizes.

For the relapse outcome, a lack of family support was the strongest predictor followed by two other variables with moderate effect sizes; the presence of a serious SUD and if probationers only received inpatient treatment. Each was associated with an increase in revocation. That said, there were nine significant variables with weak effects including risk score, needs score, presence of SUD, presence of mental health disorder, participation in inpatient and outpatient treatment, if current offense was felony level, if current offense drug-related, if less than a high school degree, and if employed. For men only, the needs score had a moderate association with relapse. The only other significant variables, albeit with weak effects, were risk score and age.

Bivariate Gender Comparison of Treatment Dosage on Revocation and Relapse

Additional bivariate analyses were examined in relation to inpatient treatment dosage and outcomes using the phi coefficient (not reported in Table 1 or 2). The following “treatment dosage hypotheses” guide these analyses for both men and women.

Hypothesis 1 (H1): Probationers who participated in 90 or more days of inpatient treatment would experience a significant reduction in revocation/relapse compared with probationers who did not participate in inpatient treatment.

Hypothesis 2 (H2): Probationers who participated in 90 or more days of inpatient treatment would experience a significant reduction in revocation/relapse compared with probationers who participated in less than 90 days of inpatient treatment.

Hypothesis 3 (H3): Probationers who participated in 180 or more days of inpatient treatment would experience a significant reduction in revocation/relapse compared with probationers who did not participate in inpatient treatment.

Hypothesis 4 (H4): Probationers who participate in 180 or more days of inpatient treatment would experience a significant reduction in revocation/relapse compared with probationers who participated in less than 90 days of inpatient treatment.

Hypothesis 5 (H5): Probationers who participated in 180 or more days of inpatient treatment would experience a significant reduction in revocation/relapse compared with probationers who participated in 90 or more, but less than 180, days of inpatient treatment.

For men and women on probation, there were no statistical differences on likelihood of revocation between individuals who did not attend treatment and those with 90 or greater days of treatment. Revocation rates were similar at 53% (no treatment) and 48% for women as well as 49% (no treatment) and 58% for men. There was a statistically significant increase in the likelihood of relapse for women with 90 or more days of inpatient treatment compared with those without treatment (φ = .15, p = .034, n = 194). Specifically, the relapse rate was 20% (no treatment) and 37%. This association did not hold true for men probationers with relapse rates of 21% and 25%. When comparing 90 or greater days of treatment to less than 90s days (excluding those without treatment), no statistically significant results were observed for women probationers. Alternatively, men with less than 90 days of inpatient treatment experienced a significant increase in their likelihood of revocation (φ = .25, p = .002, n = 163) and relapse (φ = .23, p = .003, n = 163) compared with men receiving 90 days or more of treatment. Revocation rates were 90% (<90) and 58% as well as 54% (<90) and 25% for relapse.

In terms of 180 or more days of inpatient treatment compared with no treatment, there was a moderately strong and significant correlation in favor of recidivism reduction for women (φ = −.36, p < .001, n = 108) and men (φ = −.29, p < .001, n = 192), but no such association for relapse with either group. The revocation rate was 53% (no treatment) and 19% for women and 49% (no treatment) and 15% for men. That said, strong significant correlations were observed when comparing 180 or more days with less than 90 days of treatment (excluding those with no inpatient treatment). For women, there was a large effect for reducing revocation (φ = −.48, p < .001, n = 86) but modest effect in terms of relapse prevention (φ = −.23, p = .034, n = 86). For men, the effect sizes were quite large for both revocation (φ = −.73, p < .001, n = 74) and relapse reduction (φ = −.41, p < .001, n = 74). Finally, when comparing individuals who completed 90 to 179 days of inpatient treatment to those with more than 180 days, the potential for revocation and relapse reduction remains, though the effect is reduced slightly for men. For women probationers, the impact on revocation didn’t change (φ = −.48, p < .001, n = 172) though the effect on relapse improved slightly (φ = −.31, p < .001, n = 172). The effect remained large for men’s revocation (φ = .62, p < .001, n = 163) and modest for relapse (φ = −.20, p < .010, n = 163).

To summarize, we reject the first treatment dosage hypothesis. In no circumstance did inpatient treatment of 90 or more days produce a significant reduction in revocation or relapse compared with individuals who did not participate in treatment. In fact, an increase in relapse was observed for women in 90 or more days of treatment. Treatment dosage H2 was supported for men, but we failed to reject the null hypothesis for women probationers. Ninety or more days of inpatient treatment produced a significant reduction in revocation and relapse for men compared with individuals who completed less than 90 days. H3 was partially supported. Both groups experienced a significant reduction in revocation associated with 180 or more days of treatment compared with those within the sample that did not participate in inpatient treatment. This, however, did not hold true for relapse. H4 and H5 were supported. Both men and women probationers experienced a significant reduction in revocation and relapse associated with 180 or more days of inpatient treatment compared with individuals with less than 180 days of treatment.

Discussion

This study examined gender differences related to key risk factors associated with relapse and recidivism with individuals on specialized probation who had one of the following primary needs: a SUD, a SMI diagnosis, or a COD. Gender pathways to crime and social support theory were used as foundations to explain potential gender differences. We found that family support was consistently the strongest predictor of probation revocation with this sample of justice-involved individuals. The findings lend credence to social support being important for justice-involved offenders (Cullen, 1994; Spjeldnes et al., 2012; Taylor, 2015). That said, the magnitude of these variables, while similar in pattern, differed as predicted by the gender pathways perspective. Specifically, while a lack of family support was the strongest predictor for revocation, the magnitude of that association was notably stronger for women, as noted by gender pathways scholars (Wright et al., 2013). However, with relapse these patterns diverge. Consistent with prior literature, family support/dysfunction and educational attainment remained significant for women (Connors et al., 1998; Dobkin et al., 2002), but such correlations did not persist for men. Women’s identity and self-worth may be linked to the quality of their relationships, and these relationships can be supportive or promote substance use to cope with relationship problems (Salisbury & Van Voorhis, 2009).

This study reinforced that probationers with an SMI were less likely to be revoked than those with COD, a finding consistent in other studies (Hartwell, 2004; McReynolds et al., 2010; Messina et al., 2004). Furthermore, individuals with SUD were more likely to relapse than those with COD, with no significant differences found for those solely with a mental health disorder compared with probationers with CODs. This suggests that individuals with SUD have a more serious problem with drugs or alcohol with more frequent use than individuals with COD who may use substances to self-medicate.

The treatment dosage results in the current study are exploratory. We found that those who exclusively participated in outpatient treatment were less likely to be revoked than those participating in inpatient and outpatient treatment. However, this is likely a proxy of higher risk and needs with individuals with higher needs qualifying for more treatment. The number of days in treatment was associated with a significant reduction in revocation, but not relapse. That said, the proportion of the sample experiencing relapse was a little more than half of those revoked. This is nonetheless unexpected. Bivariate analyses demonstrated that treatment dosage matters. While previous literature found that a minimum of 90 days of substance abuse treatment made a positive difference on probation (Houser et al., 2012; Inciardi et al., 2004; Latessa et al., 2020; Taxman & Belenko, 2012), completing more than 180 days was necessary to reduce revocation for the higher need COD women on probation.

Another view is that the treatment dosage results in the current study may be a reflection of the lack of gender-responsive programming in the curricula. Although the pathway to criminality can be distinctly unique between men and women, the needs for correctional programming should vary as well (Holtfreter & Morash, 2003). Indeed, with childhood abuse and trauma shown to affect women’s life course trajectories in connection with mental health and substance use issues (Karlsson & Zielinski, 2020; Langan & Pelissier, 2001; Leban & Gibson, 2020; Messina et al., 2007; Pelissier & Jones, 2005; Spjeldnes & Goodkind, 2009), responding to those needs with gender-responsive curricula is important to decrease both relapse and revocation.

Limitations

First, data were obtained from one large urban county probation department and may not be generalizable to other jurisdictions. Second, predictors in other studies relevant to relapse were not available in this study, such as past trauma, treatment motivation, and the degree to which spouse/significant other used drugs or alcohol (Connors et al., 1998; Dobkin et al., 2002). Additional measures of social support that may have been important such as probation officer support (Epperson et al., 2017; Skeem et al., 2009), peer support, and support from children were also not available (Jacoby & Kozie-Peak, 1997; McMahon, 2001). Model diagnostics from a separate set of logistic regression models, not reported, demonstrate this is particularly an issue for relapse. Though not all available variables could be included in these models due to multicollinearity issues, McFadden’s pseudo R² was found to range from .096 to .224 for relapse compared with .358 to .390 for revocation. Though all models were significant, these diagnostics confirm the presence of specification error (i.e., not all relevant variables were accounted for). Note, results obtained from logistic regression were similar to the results presented in Table 2 using the AUC, family support remains consistently one of the strongest predictors.

Furthermore, as displayed in Table 1, while women and men were similar on a variety of key measures, there were some significant differences in terms of SUD, SMI, needs scores, and treatment type. Specifically, a larger proportion of women were classified as high need (34%) compared with men (18%), but this could be a function of the risk/needs assessment tool that was used. Alternatively, the inverse was true, a larger proportion of men in the sample were classified as low need (34%) compared with women (18%). Random assignment was not possible given the data had been collected after judicial sentencing. Propensity score methods were attempted but no suitable matching could be established to create a comparison group in terms of treatment/no treatment or across gender (Bai & Clark, 2019). Finally, this study made no attempt to formally evaluate any of the specialized probation programs.

Conclusion

Three key conclusions can be drawn from this research. First, our results support the gender-responsive strategies approach espoused by Bloom and colleagues (2003) and others (see also Spjeldnes et al., 2014). Gender-specific programming should be tailored to the unique reality and needs of probationers. Second, for probationers with serious SUDs, the conventional 90-day minimum for treatment effectiveness will likely be insufficient. Agencies working with these populations should target a minimum of 180 days of treatment. Finally, results from this study also demonstrated that continued outpatient treatment is critical to success. This is consistent with the health care system’s emphasis on continuity of care (see Gulliford et al., 2006), which coincides with many correctional reentry initiatives (e.g., Woods et al., 2013), including the National Institute of Corrections’ Transition from Prison to Community and Transition from Jail to Community Initiatives.

Footnotes

Authors’ note:

The authors and publication of this research received no external funding support. The authors report no conflicts of interest. The authors wish to thank the Bexar County Community Supervision and Corrections Department for data access,and to Shonna Webster and Kayli Hoffer for data collection and coding assistance.

ORCID iD

Leanne Fiftal Alarid

Leanne Fiftal Alarid,PhD,is a professor of criminal justice and security studies at The University of Texas at El Paso. She has published more than 60 journal articles and eight books in the research areas of corrections,mental health,and gender and crime. In 2023,she was selected for the Peter P. Lejins Research Award through the American Correctional Association,and in 2024 for the John Howard Award through the Corrections Section of the Academy of Criminal Justice Sciences. She is currently evaluating a jail-based reentry program in El Paso Texas funded by the Bureau of Justice Assistance.

Adam K. Matz,PhD,is an associate professor of criminal justice at the University of North Dakota. He has more than two dozen scholarly publications and numerous professional reports. His research interests include evidence-based program and practice identification,risk assessment validation,and recidivism. His current projects include evaluating an intimate partner violence program at the Grand Forks Domestic Violence Court,and a workload evaluation with Westat for the Washington State Department of Corrections,Community Corrections Division.

References

Albrecht

T. L.

Adelman

M. B.

(1987). Communicating social support. Sage.

Arntson

Droge

(1987). Social support in self-help groups: The role of communication in enabling perceptions of control. In Albrecht

T. L.

Adelman

M. B.

(Eds.), Communicating social support (pp. 148–171). Sage.

Bai

Clark

M. H.

(2019). Propensity score methods and applications. Sage.

Berg

Huebner

(2011). Reentry and ties that bind: An examination of social ties, employment and recidivism. Justice Quarterly, 28(2), 382–410. https://doi.org/10.1080/07418825.2010.498383

Bloom

Owen

Covington

(2003). Gender-responsive strategies: Research, practice, and guiding principles for women offenders. U.S. Department of Justice, National Institute of Corrections. https://s3.amazonaws.com/static.nicic.gov/Library/018017.pdf

Bush

Glick

Taymans

(2016). Thinking for a change 4.0 (Accession No. 032650). U.S. Department of Justice, National Institute of Corrections. https://nicic.gov/resources/nic-library/web-items/thinking-change-40

Carson

Kluckow

(2023). Correctional populations in the United States, 2021 (NCJ 305542). U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics. https://bjs.ojp.gov/library/publications/correctional-populations-united-states-2021-statistical-tables

Castillo

E. D.

Alarid

L. F.

(2011). Factors associated with recidivism among offenders with mental illness. International Journal of Offenders Therapy and Comparative Criminology, 55(1), 98–117. https://doi.org/10.1177/0306624X09359502

Chandler

D. W.

Spicer

(2006). Integrated treatment for jail recidivists with co-occurring psychiatric and substance use disorders. Community Mental Health Journal, 42(4), 405–425. https://doi.org/10.1007/s10597-006-9055-6

10.

Connors

G. J.

Maisto

S. A.

Zywiak

W. H.

(1998). Male and female alcoholics’ attributions regarding the onset and termination of relapses and the maintenance of abstinence. Journal of Substance Abuse, 10(1), 27–42. https://doi.org/10.1016/S0899-3289(99)80138-2

11.

Cullen

F. T.

(1994). Social support as an organizing concept for criminology: Presidential address to the Academy of Criminal Justice Sciences. Justice Quarterly, 11(4), 527–559. https://doi.org/10.1080/07418829400092421

12.

Cullen

F. T.

Agnew

Wilcox

(2021). Criminological theory: Past to present (7th ed.). Oxford University Press.

13.

Curcio

Pattavina

(2018). Still paying for the past: Examining gender differences in employment among individuals with a criminal record. Women and Criminal Justice, 28(5), 375–396. https://doi.org/10.1080/08974454.2018.1441773

14.

Daly

. (1992). Women’s pathways to felony court: Feminist theories of lawbreaking and problems of representation. Southern California Review of Law and Women’s Studies, 2(1), 11–52. http://hdl.handle.net/10072/386914

15.

Deschenes

E. P.

Turner

Greenwood

P. W.

(1995). Drug court or probation? An experimental evaluation of Maricopa County’s drug court. Justice System Journal, 18(1), 55–73. https://doi.org/10.1080/23277556.1995.10871222

16.

Dobkin

P. L.

De Civita

Paraherakis

Gill

(2002). The role of functional social support in treatment retention and outcomes among outpatient adult substance abusers. Addiction, 97(3), 347–356. https://doi.org/10.1046/j.1360-0443.2002.00083.x

17.

Ellis

P. D.

(2010). The essential guide to effect sizes: Statistical power, meta-analysis, and the interpretation of research results. Cambridge University Press.

18.

English

Mande

(1991). Empirical support for intervention strategies in community corrections. Journal of Contemporary Criminal Justice, 7(2), 95–106. https://doi.org/10.1177/104398629100700203

19.

Epperson

M. W.

Thompson

J. G.

Lurigio

A. J.

Kim

(2017). Unpacking the relationship between probationers with serious mental illness and probation officers: A mixed-methods examination. Journal of Offender Rehabilitation, 56(3), 188–216. http://doi.org/10.1080/10509674.2017.1290005

20.

Fiorentine

Anglin

M. D.

Gil-Rivas

Taylor

(1997). Drug treatment: Explaining the gender paradox. Substance Use and Misuse, 32(6), 653–678. https://doi.org/10.3109/10826089709039369

21.

Galvin

M. A.

Davidson

K. M.

Kleiman

(2022). Substance involvement and probation outcomes: Evidence from a cohort study. Journal of Drug Issues, 52(3), 329–348. https://doi.org/10.1177/00220426211062560

22.

Gehring

K. S

. (2016). A direct test of pathways theory. Feminist Criminology, 13(2), 115–137. https://doi.org/10.1177/1557085116646195

23.

Gil-Rivas

Fiorentine

Anglin

M. D.

(1996). Sexual abuse, physical abuse, and posttraumatic stress disorder among women participating in outpatient drug abuse treatment. Journal of Psychoactive Drugs, 28(1), 95–102. https://doi.org/10.1080/02791072.1996.10471718

24.

Green

C. A.

Polen

M. R.

Lynch

F. L.

Dickinson

D. M.

Bennett

M. D.

(2004). Gender differences in outcomes in an HMO-based substance abuse treatment program. Journal of Addictive Diseases, 23(2), 47–70. https://doi.org/10.1300/J069v23n02_04

25.

Grella

C. E.

Stein

J. A.

Greenwell

(2005). Associations among childhood trauma, adolescent problem behaviors, and adverse adult outcomes in substance-abusing women offenders. Psychology of Addictive Behaviors, 19(1), 43–53. https://doi.org/10.1037/0893-164X.19.1.43

26.

Stone

Varghese

Braziel

Snyder

Zaller

(2023). Drivers of probation revocation in Pulaski County, Arkansas. Criminal Justice and Behavior, 50(9), 1405–1426. https://doi.org/10.1177/00938548231177701

27.

Gulliford

Naithani

Morgan

(2006). What is “continuity of care”? Journal of Health Services Research & Policy, 11(4), 248–250. https://doi.org/10.1258/135581906778476490

28.

Hamilton

Kowalski

M. A.

Schaefer

Kigerl

(2020). Recrafting youth risk assessment: Developing the modified positive achievement change tool for Iowa. Deviant Behavior, 41(10), 1268–1289. https://doi.org/10.1080/01639625.2019.1609302

29.

Hartman

D. J.

Friday

P. C.

Minor

K. I.

(1994). Residential probation: A seven-year follow up study of halfway house discharges. Journal of Criminal Justice, 22(6), 503–515. https://doi.org/10.1016/0047-2352(94)90092-2

30.

Hartwell

(2004). Triple stigma: Persons with mental illness and substance abuse problems in the criminal justice system. Criminal Justice Policy Review, 15(1), 84–99. https://doi.org/10.1177/0887403403255064

31.

Heffner

J. L.

Blom

T. J.

Anthenelli

R. M.

(2011). Gender differences in trauma history and symptoms as predictors of relapse to alcohol and drug use. The American Journal on Addictions, 20(4), 307–311. https://doi.org/10.1111/j.1521-0391.2011.00141.x

32.

Hiller

M. L.

Knight

Simpson

D. D.

(2006). Recidivism following mandated residential substance abuse treatment for felony probationers. The Prison Journal, 86(2), 230–241. https://doi.org/10.1177/0032885506287951

33.

Holtfreter

Morash

(2003). The needs of women offenders: Implications for correctional programming. Women and Criminal Justice, 14(2/3), 137–160. https://doi.org/10.1300/J012v14n02_07

34.

Houser

K. A.

Salvatore

Welsh

(2012). Individual-level predictors of community aftercare completion. The Prison Journal, 92(1), 106–124. https://doi.org/10.1177/0032885511429261

35.

Houser

K. A.

Saum

C. A.

Hiller

M. L.

(2019). Mental health, substance abuse, co-occurring disorders, and 3-year recidivism of felony parolees. Criminal Justice and Behavior, 46(9), 1237–1254. https://doi.org/10.1177/0093854819856924

36.

Inciardi

J. A.

Martin

S. S.

Butzin

C. A.

(2004). Five-year outcomes of therapeutic community treatment of drug-involved offenders after release from prison. Crime and Delinquency, 50(1), 88–107. https://doi.org/10.1177/0011128703258874

37.

Jacoby

J. E.

Kozie-Peak

(1997). The benefits of social support for mentally ill offenders: Prison-to-community transitions. Behavioral Sciences and the Law, 15(4), 483–501. https://doi.org/10.1002/(SICI)1099-0798(199723/09)

38.

Jeffries

Chuenurah

Wallis

(2019). Gendered pathways to prison in Thailand for drug offending? Exploring women’s and men’s narratives of offending and criminalization. Contemporary Drug Problems, 46(1), 78–104. https://doi.org/10.1177/0091450918818174

39.

Kaeble

(2023). Probation and parole in the United States, 2021 (NCJ 305589). U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics. https://bjs.ojp.gov/library/publications/probation-and-parole-united-states-2021

40.

Karlsson

M. E.

Zielinski

M. J.

(2020). Sexual victimization and mental illness prevalence rates among incarcerated women: A literature review. Trauma, Violence and Abuse, 21(2), 326–349. https://doi.org/10.1177/1524838018767933

41.

Knight

Logan

Simpson

(2001). Predictors of program completion for women in residential substance abuse treatment. American Journal of Drug and Alcohol Abuse, 27(1), 1–18. https://doi.org/10.1081/ada-100103116

42.

Knight

Simpson

D. D.

Morey

J. T.

(2002). Evaluation of the TCU drug screen (NCJ Document 196682). U.S. Department of Justice. http://www.ncjrs.gov/pdffiles1/nij/grants/196682.pdf

43.

Kopak

A. M.

Hoffmann

N. G.

Proctor

S. L.

(2016). Key risk factors for relapse and rearrest among substance use treatment patients involved in the criminal justice system. American Journal of Criminal Justice, 41(1), 14–30. https://doi.org/10.1007/s12103-015-9330-6

44.

Landernberg

Lipsey

(2005). The positive effects of cognitive-behavioral programs for offenders: A meta-analysis of factors associated with effective treatment. Journal of Experimental Criminology, 1(4), 451–476. https://doi.org/10.1007/s11292-005-3541-7

45.

Langan

N. P.

Pelissier

B. M. M.

(2001). Gender differences among prisoners in drug treatment. Journal of Substance Abuse, 13(3), 291–301. https://doi.org/10.1016/S0899-3289(01)00083-9

46.

Latessa

E. J.

Johnson

S. L.

Koetzle

(2020). What works (and doesn’t) in reducing recidivism (2nd ed.). Routledge.

47.

Latkin

C. A.

Knowlton

A. R.

Hoover

Mandell

(1999). Drug network characteristics as a predictor of cessation of drug use among adult injection drug users: A prospective study. American Journal of Drug and Alcohol Abuse, 25(3), 463–473. https://doi.org/10.1081/ada-100101873

48.

Leban

Gibson

C. L.

(2020). The role of gender in the relationship between adverse childhood experiences and delinquency and substance use in adolescence. Journal of Criminal Justice, 66, Article 101637. https://doi.org/10.1016/j.jcrimjus.2019.101637

49.

Liu

Becker

Mowen

T. J.

(2023). Social support during reentry: Family mentor, religious, parole officer, and social service roles. Criminal Justice and Behavior, 50(7), 1053–1070. https://doi.org/10.1177/00938548231166154

50.

Manchak

S. M.

Skeem

J. L.

Kennealy

P. J.

Louden

J. E.

(2014). High-fidelity specialty mental health probation improves officer practices, treatment access, and rule compliance. Law and Human Behavior, 38(5), 450–461. https://doi.org/10.1037/lhb0000076

51.

McMahon

R. C.

(2001). Personality, stress, and social support in cocaine relapse prediction. Journal of Substance Abuse Treatment, 21(2), 77–87. https://doi.org/10.1016/S0740-5472(01)00187-8

52.

McReynolds

L. S.

Schwalbe

C. S.

Wasserman

G. A.

(2010). The contribution of psychiatric disorder to juvenile recidivism. Criminal Justice and Behavior, 37(2), 204–216. https://doi.org/10.1177/0093854809354961

53.

Messina

Burdon

Hagopian

Prendergast

(2004). One year return to custody rates among co-disordered offenders. Behavioral Sciences and the Law, 22(4), 503–518. https://doi.org/10.1002/bsl.600

54.

Messina

Grella

Burdon

Pendergast

(2007). Childhood adverse and current trauma distress. Criminal Justice and Behavior, 34(11), 1385–1401. https://doi.org/10.1177/0093854807305150

55.

Modestin

Wuermle

(2005). Criminality in men with major mental disorder with and without comorbid substance abuse. Psychiatry and Clinical Neurosciences, 59(1), 25–29. https://doi.org/10.1111/j.1440-1819.2005.01327.x

56.

Nielsen

A. L.

Scarpitti

F. R.

Inciardi

J. A.

(1996). Integrating the therapeutic community and work release for drug-involved offenders: The CREST program. Journal of Substance Abuse Treatment, 13(4), 349–358. https://doi.org/10.1016/s0740-5472(96)00112-2

57.

O’Brien

(2002). Reducing barriers to employment for women ex-offenders: Mapping the road to reintegration. Safer Foundation.

58.

Olson

D. E.

Lurigio

A. J.

Alderden

(2003). Men are from Mars, women are from Venus, but what role does gender play in probation recidivism? Justice Research and Policy, 5(2), 33–54. https://doi.org/10.3818/JRP.5.2.2003.33

59.

Pelissier

Jones

(2005). A review of gender differences among substance abusers. Crime and Delinquency, 51(3), 343–372. https://doi.org/10.1177/0011128704270218

60.

Peters

Strozier

Murrin

Kearns

(1997). Treatment of substance-abusing jail inmates: Examination of gender differences. Journal of Substance Abuse Treatment, 14(4), 339–349. https://doi.org/10.1016/s0740-5472(97)00003-2

61.

Reichert

Gleicher

(2019). Probation clients’ barriers to access and use of opioid use disorder medications. Health and Justice, 7(10), 1–11. https://doi.org/10.1186/s40352-019-0089-6

62.

Rice

M. E.

Harris

G. T.

(2005). Comparing effect sizes in follow-up studies: ROC area, Cohen’s d, and r. Law and Human Behavior, 29(5), 615–620. http://doi.org/10.1007/s10979-005-6832-7

63.

Sacks

J. Y.

McKendrick

Banks

Stommel

(2004). Modified TC for MICA offenders: Crime outcomes. Behavioral Sciences and the Law, 22(4), 477–501. https://doi.org/10.1002/bsl.599

64.

Salisbury

E. J.

Van Voorhis

(2009). Gendered pathways: A qualitative investigation of women probationers’ paths to incarceration. Criminal Justice and Behavior, 36(6), 541–566. https://doi.org/10.1177/0093854809334076

65.

Skeem

J. L.

Emke-Francis

Eno Louden

(2006). Probation, mental health, and mandated treatment: A national survey. Criminal Justice and Behavior, 33(2), 158–184. https://doi.org/10.1177/0093854805284420

66.

Skeem

J. L.

Eno Louden

Manchak

Vidal

Haddad

(2009). Social networks and social control of probationers with co-occurring mental and substance abuse problems. Law and Human Behavior, 33(2), 122–135. https://doi.org/10.1007/s10979-008-9140-1

67.

Skeem

J. L.

Manchak

Montoya

(2017). Comparing public safety outcomes for traditional probation vs specialty mental health probation. JAMA Psychiatry, 74(9), 942–948. https://doi.org/10.1001/jamapsychiatry.2017.1384

68.

Skeem

J. L.

Winter

Kennealy

P. J.

Eno Louden

Tatar

J. R.

(2014). Offenders with mental illness have criminogenic needs too: Toward recidivism reduction. Law and Human Behavior, 38(3), 212–224. https://doi.org/10.1037/lhb0000054

69.

Spjeldnes

Goodkind

(2009). Gender differences in offender reentry: A review of the literature. Journal of Offender Rehabilitation, 48(4), 314–335. https://doi.org/10.1080/10509670902850812

70.

Spjeldnes

Jung

Maguire

Yamatani

(2012). Positive family social support: Counteracting negative effects of mental illness and substance abuse to reduce jail ex-inmate recidivism rates. Journal of Human Behavior in the Social Environment, 22(2), 130–147. https://doi.org/10.1080/10911359.2012.646846

71.

Spjeldnes

Jung

Yamatani

(2014). Gender differences in jail populations: Factors to consider in reentry strategies. Journal of Offender Rehabilitation, 53(2), 75–94. https://doi.org/10.1080/10509674.2013.868387

72.

Substance Abuse and Mental Health Services Administration. (2021). National study on drug use and health (Tables 6.16B and 8.39B). Center for Behavioral Statistics and Quality. https://www.samhsa.gov/data/release/2021-national-survey-drug-use-and-health-nsduh-releases#detailed-tables

73.

Taxman

F. S.

Belenko

(2012). Implementing evidence-based practices in community corrections and addiction treatment. Springer.

74.

Taylor

C. J.

(2015). Gendered pathways to recidivism: Differential effects of family support by gender. Women and Criminal Justice, 25(3), 169–183. https://doi.org/10.1080/08974454.2014.989305

75.

Van Deinse

T. B.

Crable

E. L.

Dunn

Weis

Cuddeback

G. S

. (2021). Probation officers’ and supervisors’ perspectives on critical resources for implementing specialty mental health probation. Administration and Policy in Mental Health and Mental Health Services Research, 48(3), 408–419. https://doi.org/10.1007/s10488-020-01081-8

76.

Van Deinse

T. B.

Givens

Cowell

Ghezzi

Murray-Lichtman

Cuddeback

G. S

. (2022). A randomized trial of specialty mental health probation: Measuring implementation and effectiveness outcomes. Administration and Policy in Mental Health and Mental Health Services Research, 49(3), 415–428. https://doi.org/10.1007/s10488-021-01172-0

77.

Wagoner

J. L.

Piazza

N. J.

(1993). Group therapy for adult substance abusers on probation. Alcohol and Drug Rehabilitation, 19(3–4), 41–56. https://doi.org/10.1300/J076v19n03_02

78.

Walitzer

K. S.

Dearing

R. L.

(2005). Gender differences in alcohol and substance use relapse. Clinical Psychology Review, 26(2), 128–148. https://doi.org/10.1016/j.cpr.2005.11.003

79.

Warner

Kramer

(2009). Substance abuse treatment as an alternative to traditional sentencing for drug-dependent offenders. Criminal Justice and Behavior, 36(1), 89–105. https://doi.org/10.1177/0093854808326743

80.

Wolff

Epperson

Shi

Huening

Schumann

B. E.

Sullivan

I. R.

(2014). Mental health specialized probation caseloads: Are they effective? International Journal of Law and Psychiatry, 37(5), 464–472. https://doi.org/10.1016/j.ijlp.2014.02.019

81.

Woods

L. N.

Lanza

A. S.

Dyson

Gordon

D. M.

(2013). The role of prevention in promoting continuity of health care in prisoner reentry initiatives. American Journal of Public Health, 103(5), 830–838. https://doi.org/10.2105/AJPH.2012.300961

82.

Wright

E. M.

DeHart

D. D.

Koons-Witt

B. A.

Crittenden

C. A.

(2013). “Buffers” against crime? Exploring the roles and limitations of positive relationships among women in prison. Punishment and Society, 15(1), 71–95. https://doi.org/10.1177/1462474512466199

83.

Wright

E. M.

Van Voorhis

Salisbury

E. J.

Bauman

(2012). Gender-responsive lessons learned and policy implications for women in prison: A review. Criminal Justice and Behavior, 39(12), 1612–1632. https://doi.org/10.1177/0093854812451088

The Importance of Social Support for Individuals on Specialized Probation: A Gender Comparison of Relapse and Revocation

Abstract

Keywords

Prevalence of Substance Use and Mental Health Disorders on Probation

Specialized Probation for Substance Use and Mental Health Disorders

Gender Comparisons of Relapse and Recidivism

Gender Pathways to Crime

Social Support Theory

The Present Study

Method

Participants

Procedure

Data Analytic Plan

Measures

Revocation

Relapse

Risk Score

Needs Score

Treatment Type

Inpatient Treatment Days

Serious SUD

Current Offense Level

Nature of Current Offense

Prior Criminal History

Social Support

Other Variables

Results

Revocation and Relapse Outcomes

Bivariate Gender Comparisons

Bivariate Prediction of Revocation and Relapse by Gender

Bivariate Gender Comparison of Treatment Dosage on Revocation and Relapse

Discussion

Limitations

Conclusion

Footnotes

Authors’ note:

ORCID iD

References