Who Are the Compliant Correctional Clients? New Evidence on Protective Factors among Federal Supervised Releases

Abstract

Protective factors facilitate success on community supervision, but relatively little is known about correctional clients who are highly compliant particularly in the federal system. Drawing on a near population of federal clients on supervised release in the Midwestern United States, the current study examined variables associated with compliant supervision status. One day on supervision contributed to a 1% reduction in the logged odds of supervision compliance. Clients with no drug history had 793% increased odds, clients with sustained remission had 620% increased odds, and clients with early remission had 458% increased odds of compliant supervision status relative to clients actively using drugs. Among the federal Post Conviction Risk Assessment (PCRA) indices, only PCRA Criminal History was significant as clients with less extensive criminal history were more likely to be compliant supervision clients. A one-unit change in PCRA Criminal History status was associated with 25% reduced odds of supervision compliance. Total conditions were inversely associated with compliant supervision status with each additional condition associated with a 19% reduced likelihood of compliant supervision status. None of the demographic variables was significantly associated with compliant supervision status. Implications of the findings for the protective factor paradigm in corrections are discussed.

Keywords

supervision protective factors corrections substance use federal offenders

Introduction

Risk assessment is a ubiquitous feature of the criminal justice system employed at virtually every stage of the correctional process from pretrial assessment and supervision, to intake and classification in jail and prison, to supervision on probation, parole, or federal supervised release. Overwhelmingly, criminal justice practitioners and criminologists have focused the bulk of their attention on clients that are higher risk and thus pose greater challenges for treatment and supervision purposes (cf., Baglivio et al., 2018; Feeley & Simon, 1992; Garland, 2012; Petersilia, 1999; Singh et al., 2011; Wang & Diamond, 1999). Although it is prudent to focus on more severe offenders for correctional needs and to facilitate public safety, this focus obscures the reality that a non-trivial amount of offenders or correctional clients function well on supervision, are compliant, and in some cases, are less resource intensive and therefore easier to supervise.

The protective factor paradigm effectively frames consideration of correctional clients that fare well on supervision. Common in prevention and juvenile delinquency research (e.g., Farrington et al., 2016; Li et al., 2019; Lösel, & Farrington, 2012; Ttofi et al., 2016), protective factors are characteristics that insulate, buffer, or protect offenders from offending, substance use, or recidivism. In many respects, protective factors represent behaviors that are consistent with immersion in conventional activities, such as employment, education, church, engagement with prosocial family members and friends, and others help to transition former offenders to non-offender status (see, Petersilia, 2007). Variously referred to as strengths, assets, or resilience factors, protective factors span multiple domains and include a mix of static and dynamic factors the latter of which are often a target for correctional change (De Vries Robbé & Willis, 2017; Garland et al., 2011; Klepfisz et al., 2017; Maruna, 2001; Shepherd et al., 2018). To the extent that correctional clients exhibit protective factors, correctional officers seek to encourage, cultivate, and enhance those protective factors to facilitate the desistance process during the transition from prison to community.¹

Protective Factors and Correctional Clients

Numerous correctional studies examined protective factors that facilitate success on supervision or studied risk factors for recidivism and noncompliance in which case one can infer protective factors. For example, younger age is commonly a risk factor for noncompliance and older age is commonly negatively associated with noncompliance. In other words, risk factors and protective factors are obverse. Irrespective of the approach, correctional researchers identified several protective factors based on diverse data sources and analytical techniques. In a comparative study of parolees, probationers, and shock incarceration clients, MacKenzie et al. (1992) reported several correlates of positive adjustment to supervision including older age, higher intelligence, less criminal history, shorter criminal career, employment, and involvement in positive social activities. Schram et al. (2006) investigated 546 female parolees and found that employment, residential stability, and involvement in substance abuse treatment were associated with successful functioning on supervision.

In a study of more than 3,000 felony probation clients, Harris (2011) identified clients whose first criminal offense did not occur until adulthood and thus had no prior criminal history. These offenders were highly successful on supervision and easier to supervise because they maintained employment, abstained from drug use, and conscientiously followed conditions of supervision (e.g., paid supervision fees, never missed appointments). Evans et al. (2012) studied more than 25,000 parolees and probationers and examined multiple positive and adverse supervision outcomes. They found several predictors of successful completion of supervision including older age, white race/ethnicity, greater educational attainment, employment, more social support, and prior history of drug treatment. Drawing on data from nearly 30,000 offenders released from state prison, Ostermann (2013, 2015) found several factors that insulated offenders from recidivism. These included older age, female gender, white ethnicity, less criminal history, and less deprivation in social background (also see, Ostermann et al., 2013).²

Grattet and Lin (2016) analyzed data from the California Parole Study and examined five forms of misconduct including absconding, technical violations, drug use or possession, (non-sexual) violent reoffending, and sexual reoffending. Several protective factors emerged from their analyses. Little to no history of imprisonment, better mental health, female gender, and older age (age 45+) were protective for all five parole outcomes suggesting that parolees who exhibit these characteristics are likely easier to supervise and more compliant. Recently, Powers et al. (2018) examined a sample of more than 30,000 parolees and found several protective factors that were associated with reduced likelihood of absconding. These included older age, white race/ethnicity, greater educational attainment, better psychosocial functioning, and less criminal history. These characteristics typified individuals that fared well on parole in terms of their compliance and behavioral functioning.

Criminologists have paid less attention to potential protective factors among correctional clients in the federal criminal justice system. Based on data from more than 40,000 federal probationers or supervised releases, Walters and Cohen (2016) found several factors that protected against recidivism among male offenders including older age, less criminal history, white racial status, and prosocial as opposed to criminal thinking/cognitive approach. Among women, older age, less criminal history, and prosocial thinking were also associated with compliance. In their recent study of more than 1,000 defendants on federal pretrial supervision, DeLisi et al. (2019) reported that minimal risk clients were those who were significantly older at first arrest, had no history of arrest or violence perpetration, and had no substance use history. Indeed, they identified 65 federal clients that fared well on supervision and across a 5-year follow-up period, this cohort of clients with these protective features amassed a mere four arrests, all for low-level offenses. Overall, older age, female gender, little to no prior criminal history or revocation history, sobriety, and involvement in conventional social institutions, such as family, work, and prosocial activities are among the most consistent protective factors (Anderson et al., 1991; Hochstetler et al., 2016; Mayzer et al., 2004; Pearl, 1998; Spruit et al., 2017; Threadcraft-Walker et al., 2018; Trulson et al., 2011; Wodahl et al., 2011) among correctional clients.

Current Aim

There is a broader importance to understanding compliant offenders. In recent years there is an emphasis on reducing the size of the correctional population, and indeed, during the Covid 19 pandemic, many jails and prisons have released prisoners to an unprecedented degree. Almost always, the decision on who to release relates primarily to offenders who perpetrated the least serious offenses (read non-violent) and who appear to be the “best” risk for adequate behavioral functioning in the community. The more that is known about compliant correctional clients, the more scientifically informed those decisions will be. A variety of protective factors characterizes successful correctional clients. The current study extends this literature in three ways by (1) examining clients who are less resource intensive to supervise based on having just one incident of noncompliance during their supervision, (2) using new covariates including the federal post-conviction risk assessment (PCRA), and (3) using a quasi-population of correctional clients on federal supervised release. The central research question is which factors are associated with compliant supervision status.

Method

Participants

The current study used retrospective, archival data from the total population of 865 active correctional clients in a federal jurisdiction in the Midwestern United States (two clients had incomplete data thus the analytical sample is 863). All clients were on supervised release after serving a confinement sentence under the supervision of the Bureau of Prisons. The sample was 84% male, 16% female, 79.4% white, 20.6% African American, 92% non-Hispanic, 8% Hispanic, and the mean age was nearly 44 years. The most common instant/most recent conviction offenses were distribution of methamphetamine (35%), felon in possession of firearm (13%), bank fraud, money laundering, and/or identity theft (13%), distribution of cocaine base (12%), possession or manufacturing of child pornography (6.5%), distribution of marijuana (6%), use of firearm during a drug trafficking offense (4.5%), and distribution of cocaine (3.6%). The clients were diverse in terms of their criminal history and criminal justice system involvement. The federal criminal history rank employs a 6-point system where I = lowest risk and VI = highest risk in terms of criminal history. In these data, 35.4% were Criminal History Rank I, 13.5% were Criminal History Rank II, 18.7% were Criminal History Rank III, 12.5% were Criminal History Rank IV, 6.7% were Criminal History Rank V, and 13.2% were Criminal History Rank VI.

Procedures

We collected data in two ways. First, all data in the client’s Probation/Pretrial Services Automated Case Tracking System (PACTS) file were electronically extracted and converted to an Excel spreadsheet. PACTS is the case management platform used in all 94 federal districts to track federal defendants. This electronic extraction contained information on numerous variables including demographics, case information, conditions, criminal history indices, and other documents relevant to the client’s social and criminal history. Second, the senior author manually collected additional data from presentence reports (PSR), offender documents from the Bureau of Prisons, local, state, and national criminal histories, psychological and psychiatric reports, treatment reports, and other relevant documents located in PACTS.

During the PSR interview process, defendants self-reported their address and residency history and requests for criminal history were sent to all of those areas. In addition, defendants were questioned about juvenile placements and if the defendant lived in any other location than with their parents, such as foster care, group homes, juvenile homes, state facilities, and others. Based on this information, verification was sent by United States Probation to those facilities.³ Additional self-reported information on antisocial behavior was gleaned from official mental health and educational records. The senior author coded all data and entered it into Excel then transferred into Stata/IC 14.2 for analyses. The Administrative Office of the U.S. Courts and Chief District Judge in this federal jurisdiction provided research approval for the study.

Measures

Dependent variables

Compliant supervision status (0 = no, 1 = yes) is defined as clients who have only one incident of noncompliance that was resolved with a verbal admonishment from the supervising probation officer. This dependent variable is used in the logistic regression model. Compliant supervision status is derived from a count measure of noncompliance incidents while on supervision ( $\bar{x}$ = 2.76, SD = 4.0, range = 1–41).⁴

Criminal career parameters

For the difference of means t-tests, nine count measures were included to quantify the criminal career including age of arrest onset, juvenile police contacts, juvenile confinement, total arrest charges, total convictions, total prison sentences, total probation revocations, total parole revocations, and criminal career span measured in years.

Independent Variables

Days on Supervision is a count measure indicating the number of days the client has been on supervised release and serves as the exposure or time-at-risk variable.

Total conditions is a count measure of the number of formal conditions of the client’s supervised release.⁵

Demographics

Dichotomous terms for sex (0 = female, 1 = male), race (0 = white, 1 = African American), Hispanic (0 = no, 1 = yes) and continuous measure of client’s current age were controlled.

Substance use history

Dichotomous terms for no drug history (0 = no, 1 = yes), sustained remission (0 = no, 1 = yes), and early remission (0 = no, 1 = yes) were used with active drug user the omitted referent in regression analyses.

Federal post conviction risk assessment

The Federal Post Conviction Risk Assessment (PCRA) consists of two sections. One section is completed by the officer—known as the Officer Assessment—and one section is completed by the offender—known as the Offender Self-Assessment. The Offender Assessment utilizes data in PACTS and the PSR and is a quantitatively-scored instrument that provides a consistent and valid method of predicting risk of re-arrest and reconviction. The Officer Assessment has seven domains with some scored and some unscored items.⁶ Prior research supports the reliability and validity of the PCRA (Cohen & Bechtel, 2017; Cohen & Whetzel, 2014; DeLisi et al., 2018; Harris et al., 2015; Johnson et al., 2011; Lowenkamp et al., 2013). Univariate statistics for PCRA Risk are ( $\bar{x}$ = .93, SD = .83, range = 0–3).

PCRA Criminal History ( $\bar{x}$ = 4.40, SD = 2.64, range = 0–9) contains 7-items that relate to the offender’s criminal history including whether they had a juvenile arrest, number of prior arrests, prior varied offending pattern, prior revocation/arrest while on supervision, prior institutional adjustment, and age at intake to supervision.

PCRA Education/Employment ( $\bar{x}$ = .71, SD = .83, range = 0–3) contains 5-items that relate to the offender’s work and school history including less than high school or has only GED, currently employed, multiple jobs in the past year, employed less than 50% over past 2 years, and good work assessment over past 12 months.

PCRA Drug/Alcohol ( $\bar{x}$ = .38, SD = .65, range = 0–2) contains 6-items relating to the offender’s drug and alcohol problems including drug use related to disruption at work, school or home, drug use in physically hazardous conditions, drug use led to legal problems, drug use continued despite social problems, current alcohol problem, and current drug problem.

PCRA Cognitions ( $\bar{x}$ = .08, SD = .27, range = 0–1) contains 2-items relating to antisocial cognition including whether the offender harbors antisocial attitudes and values and the offender’s attitude toward supervision and change.

PCRA Social Networks ( $\bar{x}$ = 1.00, SD = .74, range = 0–3) contains 6-items relating to the offender’s social network including single, divorced, or separated, lives with spouse and/or children, lacks family support, unstable family situation, associates with negative peers or no friends, and lacks positive prosocial support. Descriptive statistics for all study variables appear in Table 1.

Table 1.

Descriptive Statistics.

Variable	Mean	SD	Range
Days on supervision	683.36	540.99	6–3,304
Current age	43.71	11.45	21–81
PCRA risk	.93	.83	0–3
PCRA cognition	.08	.27	0–1
PCRA criminal history	4.4	2.64	0–9
PCRA drug/alcohol	.38	.65	0–2
PCRA education/employment	.71	.83	0–3
PCRA social network	.99	.74	0–3
Total conditions	5.98	2.89	0–18
Total arrest charges	14.31	14.75	1–97
Total convictions	8.51	8.26	1–67
Total prison sentences	2.33	2.06	1–16
Total probation revocations	.67	1.20	0–12
Total parole revocations	.13	.64	0–8
Age arrest onset	23.57	12.46	6–78
Juvenile police contacts	2.09	4.93	0–59
Juvenile confinements	.26	.75	0–7
Criminal career span	20.12	11.67	0–68
Compliant supervision	65.9% yes	34.1% no
No drug history	13.6% yes	86.4% no
Sustained remission	56.2% yes	43.8% no
Early remission	22.1% yes	77.9% no
Sex	15.7% female	84.3% male
Race	79.4% white	20.6% black
Hispanic	91.8% yes	8.2% no

Data Analysis

We used descriptive statistics (Table 1), difference of means t-tests (Table 2), and binary logistic regression (Table 3). In the binary logistic regression model, we specified odds ratios given their intuitive value. We performed the logistic regression model with bootstrapped standard errors with 500 replications to increase confidence in the estimates with corresponding 95% confidence intervals. To examine the sensitivity of the logistic regression model, we performed a supplemental analysis to include only clients at the 50^th percentile of above for days on supervision to reduce the potential confounding effects of short duration on supervision (e.g., less than 1 month).

Table 2.

Criminal Career Difference of Means by Supervision Status.

Variable	Compliant supervision	Non-compliant supervision	t	p
Total arrest charges	12.33	18.12	5.56	<.001
Total convictions	7.41	10.64	5.53	<.001
Total prison sentences	2.19	2.60	2.78	<.01
Total probation revocations	.60	.80	2.31	<.05
Total parole revocations	.13	.13	–0.02	ns
Age arrest onset	25.41	19.98	–6.19	<.001
Juvenile police contacts	1.55	3.14	4.55	<.001
Juvenile confinements	.21	.37	3.00	<.01
Criminal career span	19.04	22.22	3.82	<.001
PCRA risk	.76	1.22	7.92	<.001
PCRA cognition	.06	.12	3.28	<.001
PCRA criminal history	3.66	5.55	10.59	<.001
PCRA drug/alcohol	.31	.53	4.82	<.001
PCRA education/employment	.66	.80	2.34	<.05
PCRA social network	.88	1.22	6.54	<.001
Total conditions	5.49	6.93	7.11	<.001

Table 3.

Logistic Regression Model for Compliant Supervision.

Variable	Odds ratio	Bootstrap standard error	z	95%CI
Days on supervision	.99	.01	−5.54***	.99/.99
Sex	.93	.25	−0.27	.55/1.57
Race	.77	.19	−1.08	.48/1.24
Hispanic	.81	.28	−0.61	.41/1.59
Current age	1.01	.01	0.65	.99/1.02
No drug history	8.93	5.65	3.46***	2.58/30.88
Sustained remission	7.20	3.61	3.94***	2.69/19.23
Early remission	5.58	2.96	3.24***	1.97/15.81
PCRA risk	1.08	.34	0.25	.59/1.98
PCRA cognition	.67	.23	−1.16	.35/1.32
PCRA criminal history	.75	.06	−3.49***	.64/.88
PCRA drug/alcohol	.76	.14	−1.53	.53/1.08
PCRA education/employment	1.10	.18	0.59	.80/1.53
PCRA social network	.74	.14	−1.59	.51/1.07
Total conditions	.81	.03	−5.33***	.76/.88
Wald χ²	175.97***
Pseudo R²	.221

Note. ^***p < .001.

Findings

Criminal Career and PCRA Risk Difference of Means by Supervision Status

As shown in Table 2, current supervision status revealed significant criminal career differences. Clients currently characterized as compliant on supervision had later starting, shorter, less extensive, and less severe criminal careers compared to clients that were not currently compliant to supervise. This is seen for total arrest charges (t = 5.56, p < .001), total convictions (t = 5.53, p < .001), total prison sentences (t = 2.78, p < .01), total probation revocations (t = 2.31, p < .05), age of arrest onset (t = –6.19, p < .001), juvenile police contacts (t = 4.55, p < .001), juvenile confinements (t = 3.00, p < .01), and criminal career span (t = 3.82, p < .001). The only criminal career parameter that was not significant was total parole revocations (t = –0.02, ns). In addition, compliant clients had uniformly lower PCRA risk assessments for total risk (t = 7.92, p < .001), cognition (t = 3.28, p < .001), criminal history (t = 10.59, p < .001), drug/alcohol (t = 4.82, p < .001), education/employment (t = 2.34, p < .05), and social network (t = 6.54, p < .001) in addition to fewer total conditions (t = 7.11, p < .001).

Logistic Regression Model for Compliant Supervision

As shown in Table 3, several factors were associated with compliant supervision status. Days on supervision was inversely associated with compliant supervision (OR = .99, z = –5.54, p < .001), the shorter length of time an offender was on supervised release, the more likely he or she had few problems. Each day on supervision reduced the likelihood of compliant supervision status by 1% lower odds. None of the demographic variables was significantly associated with compliant supervision status. In contrast, drug history was robustly associated with compliant supervision status. Clients with no drug history had 793% increased odds (OR = 8.93, z = 3.46, p < .001), clients with sustained remission had 620% increased odds (OR = 7.20, z = 3.94, p < .001), and clients with early remission had 458% increased odds (OR = 5.58, z = 3.24, p < .001) of compliant supervision status. Among the PCRA indices, only PCRA Criminal History was significant as clients with less extensive criminal history (OR = .75, z = –3.49, p < .001) were more likely to be compliant supervision clients. A one-unit change in PCRA Criminal History status was associated with 25% reduced odds of compliant supervision status. Total conditions were inversely associated with compliant supervision status (OR = .81, z = –5.33, p < .001) with each additional condition associated with a 19% reduced likelihood of compliant supervision status.

Sensitivity Analysis for Logistic Regression Model

Given the significant association for days on supervision, we tested the sensitivity of the model by limiting the analysis to clients at the 50th percentile or above of days on supervision, which was 534 days. This removed clients that had been on supervision for potentially too brief a duration to accumulate noncompliance incidents. The results were substantively the same as those in Table 2 regarding current drug status and total conditions. No drug history, sustained remission, and early remission were positively associated with compliant supervision status and total conditions were inversely associated with compliant supervision status. Two other effects attenuated. The PCRA Criminal History variable was no longer significant but trended toward significance (OR = .83, z = –1.77, p < .077) and the effect for race (indicating white clients) also trended toward significance (OR = .57, z = –1.85, p < .064).

Discussion

For several decades, correctional practitioners have employed scientific principles in the reliable and valid identification, supervision, and treatment of correctional clients (Andrews & Bonta, 2010; Duriez et al., 2018; Hanson et al., 2009; Smith et al., 2009). Implicit in this paradigm is the realization that some offenders are low-risk and will likely require fewer resources for supervision. Moreover, even some clients with extensive criminal histories who are high-risk in their assessment will nevertheless fare well on supervision. As such, identifying the factors associated with compliant supervision status is important for both research and correctional practice. Several discussion points are worth noting.

First, compared to clients that are actively using drugs while on supervision, those in early or sustained remission or who abstained from drugs throughout life were dramatically more likely to be compliant correctional clients in terms of supervision. The effect sizes were large and revealed a gradient of time or latency from drug use from early remission (458%) to sustained remission (620%) to no drug history (793%). Consistent with research that shows drug use is a risk factor for noncompliance, violation, and revocation of correctional sentence (Bennett et al., 2008; Salas-Wright et al., 2016), avoidance of drug use is commensurately linked with compliant supervision status. We suspect that practitioners are well aware of the critical role of drug use and by extension drug abstention and compliance on supervision. Prolonged sobriety can represent a variety of circumstances that bode well for a correctional client’s ability to navigate supervision, including involvement in drug treatment during BOP confinement, severance of former friendship networks where drug use was pivotal, and willingness and desire to lead a drug-free, crime-free life. Indeed, in our various practitioner roles, clients have stated that cessation of drug use was “the switch” that helped to change them from offender to non-offender (or at least set into motion the desistance processes therein). This comports with Opsal’s (2011) qualitative study of women parolees who shed their formerly deviant self-identity by renouncing or disassociating themselves from their prior drug use and drug lifestyle.

It is also important to recognize that substance use and abuse is itself a fundamental marker of correctional status. To illustrate, Vaughn et al. (2012) analyzed data on approximately 38,000 participants selected from the 2009 National Survey on Drug Use and Health. Compared to those in the general population, persons or probation or parole had significantly higher past year binge drinking, cigarette, marijuana, cocaine, hallucinogens, inhalants, prescription opiates, tranquilizers, crack cocaine, heroin, and methamphetamine use. In addition, correctional clients were more likely to have early onset substance use, meet diagnostic criteria for four types of substance dependence, to drive while intoxicated, and to sell illegal drugs. In this regard, substance use is an element of a broader antisocial behavioral pattern and as the current findings indicate, once that connection to drug use is broken—easier, compliant conduct ensues.

Second, criminal history is crucial for understanding which clients are less resource intensive to manage on supervision. In the logistic regression model, a one-unit interval change in PCRA Criminal History resulted in a 25% decreased odds likelihood of compliant supervision. Thus, high-risk clients are 75% less likely (reduced odds) than low-risk clients to be compliant. Put another way, the lowest risk clients are 75% more likely (higher odds) than their high-risk peers to be compliant. That finding provides additional support for the PCRA and its classification accuracy regarding various offender types. The significance of PCRA Criminal History also makes intuitive sense given the profound role of criminal history for understanding offender risk (Andrews et al., 2006; Bonta, 2002; Caudy et al., 2013; DeLisi, 2016; Whitten et al., 2017).⁷

It is also important to recognize that current compliant supervision status is not a transitory phase in the criminal career, but likely represents a broader tendency toward compliance—even among offenders with at times lengthy criminal records. The t-tests make clear that compliant supervision offenders have less extensive criminal careers and less noncompliance with the justice system compared to clients that are more difficult to manage currently. The relatively advanced age of these offenders is also illustrative. The mean age is approximately 44 years that is considered old in criminological terms (DeLisi & Piquero, 2011). Thus, the inexorable processes of aging among offenders with significant but not extreme criminal careers likely facilitates their compliance.

Third, an inverse association occurred for total conditions and compliant supervision. This likely reflects multiple issues. First, clients that receive fewer conditions exhibit fewer criminogenic risks and needs and thus have fewer problems to address during their supervision. Thus, it is reasonable that a correctional client with fewer entanglements would be easier to supervise. Second, fewer conditions also means there are fewer opportunities for the client to be noncompliant, which is supportive of research showing that more intensive community supervision with more conditions is associated with increased violations and revocation (cf., Cohen, 2019; Petersilia & Turner, 1993; Zettler & Medina, 2019) among community correctional clients.

Fourth, although correctional practitioners commonly use risk assessment tools, there is variation among correctional staff in terms of the degree to which they adhere to actuarial tools, to the degree that they “believe” in the outcomes produced by the actuarial tools, and to the degree that actuarial tools (compared to their own professional judgment, experiences, and training) best identify offender risk levels (Hochstetler et al., 2017; Miller & Maloney, 2013; Shook & Sarri, 2007; Viglione et al., 2015). Most of the subscales of the PCRA, a respected risk assessment instrument (Cohen & Bechtel, 2017; Cohen & Whetzel, 2014; DeLisi et al., 2018), were not significantly associated with compliant supervision status. That points to the challenges of trying to identify the positive cases of clients that perform well. Indeed, the more “intuitive” measures of current substance use status fared well in this regard.

Finally, despite the strengths of a near population, comprehensive data collection that minimized shared methods variance, and the use of multiple control variables to guard against confounding effects, the current study had limitations as well. For instance, despite the salience of risk indices to assessing the correctional population, there are critical qualitative issues that are crucial for understanding offender change and we were not able to capture these. Our practitioner experiences suggest that some correctional clients—even ones with substantial criminal histories and polysubstance abuse—finally reach a point of fatigue with drug use, offending, and criminal justice system involvement, and it is this maturational point that facilitates their compliance. Qualitative data (e.g., Gideon, 2009) have shown the importance of the correctional client-officer relationship and that a nuanced blend of enforcement, care, supervision, and support helps offenders desist from crime. In addition, although the multivariate models contained multiple validated risk indices pertaining to the PCRA, we lacked a measure of self-control that has shown to be importantly related to an array of offending and justice system outcomes (Beaver et al., 2009; DeLisi & Berg, 2006; Vazsonyi et al., 2017).

Conclusion

Irrespective of these limitations, the current study provided some good news in a domain where skepticism and cynicism dominate the notion that serious criminal offenders can fare well on supervision. Following the lead of prevention and juvenile delinquency scholars, we implore penologists to also consider protective factors that facilitate supervision success and hasten the reentry transition. The more correctional officials can clearly identify the correlates of compliant offender status, the more informed the decisions about who to potentially release from custody during exigent circumstances, such as the Covid 19 pandemic, will be.

Footnotes

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Matt DeLisi receives consulting income and travel expenses in criminal and civil litigation relating to criminological and forensic assessment of criminal offenders, receives editorial remuneration from Elsevier, has received expert services income from the United States Department of Justice and the Administrative Office of the United States Courts, and receives royalty income from Cambridge University Press, John Wiley & Sons, Jones & Bartlett, Kendall/Hunt, McGraw-Hill, Palgrave Macmillan, Routledge, Sage, University of Texas Press, and Bridgepoint Education. No direct remuneration is associated with the current study. The views reflected in this study do not necessarily represent those of the Administrative Office of the U.S. Courts, Probation and Pretrial Services Office, or the Federal Judiciary.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Matt DeLisi

Notes

References

Anderson

D. B.

Schumacker

R. E.

Anderson

S. L.

(1991). Releasee characteristics and parole success. Journal of Offender Rehabilitation, 17(1–2), 133–145.

Andrews

D. A.

Bonta

(2010). Rehabilitating criminal justice policy and practice. Psychology, Public Policy, and Law, 16(1), 39–55.

Andrews

D. A.

Bonta

Wormith

J. S.

(2006). The recent past and near future of risk and/or need assessment. Crime & Delinquency, 52(1), 7–27.

Baglivio

M. T.

Wolff

K. T.

Howell

J. C.

Jackowski

Greenwald

M. A.

(2018). The search for the holy grail: Criminogenic needs matching, intervention dosage, and subsequent recidivism among serious juvenile offenders in residential placement. Journal of Criminal Justice, 55, 46–57.

Barklage

Miller

Bonham

Jr.

G. (2006). Probation conditions versus probation officer directives: Where the twain shall meet. Federal Probation, 70(3), 37–41.

Beaver

K. M.

DeLisi

Mears

D. P.

Stewart

(2009). Low self-control and contact with the criminal justice system in a nationally representative sample of males. Justice Quarterly, 26(4), 695–715.

Bennett

Holloway

Farrington

(2008). The statistical association between drug misuse and crime: A meta-analysis. Aggression and Violent Behavior, 13(2), 107–118.

Bonta

(2002). Offender risk assessment: Guidelines for selection and use. Criminal Justice and Behavior, 29(4), 355–379.

Caudy

M. S.

Durso

J. M.

Taxman

F. S.

(2013). How well do dynamic needs predict recidivism? Implications for risk assessment and risk reduction. Journal of Criminal Justice, 41(6), 458–466.

10.

Cohen

T. H.

(2019). The imposition of pretrial special conditions on released federal defendants. Paper presented at the annual meeting of the U.S. Pretrial Services and Probation Chief’s Research Group, Hoboken, N.J., April 24, 2019.

11.

Cohen

T. H.

Bechtel

(2017). Removal of non-scored items from the post-conviction risk assessment instrument: An evaluation of data-driven risk assessment research within the federal system. Federal Probation, 81(2), 37–51.

12.

Cohen

T. H.

Whetzel

(2014). The neglected R-Responsivity and the federal offender. Federal Probation, 78(2), 11–18.

13.

DeLisi

(2016). Career criminals and the antisocial life course. Child Development Perspectives, 10(1), 53–58.

14.

DeLisi

Berg

M. T.

(2006). Exploring theoretical linkages between self-control theory and criminal justice system processing. Journal of Criminal Justice, 34(2), 153–163.

15.

DeLisi

Elbert

M. J.

Drury

A. J.

(2018). Federal criminal careers: An empirical examination of the Post-Conviction Risk Assessment (PCRA). American Journal of Criminal Justice, 43, 792–809.

16.

DeLisi

Elbert

Tahja

(2019). Inverting risk assessment: Considering the lowest risk clients in the federal criminal justice system. Criminal Justice Policy Review, 30(7), 1043–1063.

17.

DeLisi

Piquero

A. R.

(2011). New frontiers in criminal careers research, 2000–2011: A state-of-the-art review. Journal of Criminal Justice, 39(4), 289–301.

18.

De Vries Robbé

Willis

G. M.

(2017). Assessment of protective factors in clinical practice. Aggression and Violent Behavior, 32, 55–63.

19.

Duriez

S. A.

Sullivan

Latessa

E. J.

Lovins

L. B.

(2018). The evolution of correctional program assessment in the age of evidence-based practices. Corrections: Policy, Practice and Research, 3(2), 119–136.

20.

Evans

Jaffe

Urada

Anglin

M. D.

(2012). Differential outcomes of court-supervised substance abuse treatment among California parolees and probationers. International Journal of Offender Therapy and Comparative Criminology, 56(4), 539–556.

21.

Farrington

D. P.

Ttofi

M. M.

Piquero

A. R.

(2016). Risk, promotive, and protective factors in youth offending: Results from the Cambridge study in delinquent development. Journal of Criminal Justice, 45, 63–70.

22.

Feeley

M. M.

Simon

(1992). The new penology: Notes on the emerging strategy of corrections and its implications. Criminology, 30(4), 449–474.

23.

Garland

Wodahl

E. J.

Mayfield

(2011). Prisoner reentry in a small metropolitan community: Obstacles and policy recommendations. Criminal Justice Policy Review, 22(1), 90–110.

24.

Garland

(2012). The culture of control: Crime and social order in contemporary society. University of Chicago Press.

25.

Gideon

(2009). What shall I do now? Released offenders' expectations for supervision upon release. International Journal of Offender Therapy and Comparative Criminology, 53(1), 43–56.

26.

Grattet

Lin

(2016). Supervision intensity and parole outcomes: A competing risks approach to criminal and technical parole violations. Justice Quarterly, 33(4), 565–583.

27.

Grattet

Petersilia

Lin

Beckman

(2009). Parole violations and revocations in California: Analysis and suggestions for action. Federal Probation, 73(1), 2–11.

28.

Hanson

R. K.

Bourgon

Helmus

Hodgson

(2009). The principles of effective correctional treatment also apply to sexual offenders: A meta-analysis. Criminal Justice and Behavior, 36(9), 865–891.

29.

Harris

G. T.

Lowenkamp

C. T.

Hilton

N. Z.

(2015). Evidence for risk estimate precision: Implications for individual risk communication. Behavioral Sciences & the Law, 33(1), 111–127.

30.

Harris

P. M.

(2011). The first-time adult-onset offender: Findings from a community corrections cohort. International Journal of Offender Therapy and Comparative Criminology, 55(6), 949–981.

31.

Hochstetler

Peters

D. J.

Copes

(2017). Supervisory experience and data points: Implementation barriers, risk assessment and correctional professionals. Criminal Justice Studies, 30(3), 240–256.

32.

Hochstetler

Peters

D. J.

DeLisi

(2016). Classifying risk development and predicting parolee recidivism with growth mixture models. American Journal of Criminal Justice, 41(3), 602–620.

33.

Hoffman

P. B.

Beck

J. L.

(2005). Revocation by consent: The United States Parole Commission's expedited revocation procedure. Journal of Criminal Justice, 33(5), 451–462.

34.

Irizarry

May

D. C.

Davis

Wood

P. B.

(2016). Mass incarceration through a different lens: Race, subcontext, and perceptions of punitiveness of correctional alternatives when compared to prison. Race and Justice, 6(3), 236–256.

35.

Johnson

J. L.

Lowenkamp

C. T.

VanBenschoten

S. W.

Robinson

C. R.

(2011). The construction and validation of the federal Post Conviction Risk Assessment (PCRA). Federal Probation, 75, 16–29.

36.

Kennealy

P. J.

Skeem

J. L.

Manchak

S. M.

Eno Louden

(2012). Firm, fair, and caring: Officer-offender relationships protect against supervision failure. Law and Human Behavior, 36(6), 496–505.

37.

Klepfisz

Daffern

Day

(2017). Understanding protective factors for violent reoffending in adults. Aggression and Violent Behavior, 32, 80–87.

38.

Chu

C. M.

Zeng

Ruby

(2019). Risk and protective factors for probation success among youth offenders in Singapore. Youth Violence and Juvenile Justice, 17(2), 194–213. https://doi.org/10.1177/1541204018778887

39.

Lösel

Farrington

D. P.

(2012). Direct protective and buffering protective factors in the development of youth violence. American Journal of Preventive Medicine, 43(2), S8–S23.

40.

Lowenkamp

C. T.

Johnson

J. L.

Holsinger

A. M.

VanBenschoten

S. W.

Robinson

C. R.

(2013). The federal Post Conviction Risk Assessment (PCRA): A construction and validation study. Psychological Services, 10(1), 87–102.

41.

MacKenzie

D. L.

Shaw

J. W.

Souryal

(1992). Characteristics associated with successful adjustment to supervision: A comparison of parolees, probationers, shock participants, and shock dropouts. Criminal Justice and Behavior, 19(4), 437–454.

42.

Maruna

(2001). Making good: How ex-convicts reform and rebuild their lives. American Psychological Association.

43.

May

D. C.

Wood

P. B.

Mooney

J. L.

Minor

K. I.

(2005). Predicting offender-generated exchange rates: Implications for a theory of sentence severity. Crime & Delinquency, 51(3), 373–399.

44.

Mayzer

Gray

M. K.

Maxwell

S. R.

(2004). Probation absconders: A unique risk group? Journal of Criminal Justice, 32(2), 137–150.

45.

Miller

Maloney

(2013). Practitioner compliance with risk/needs assessment tools: A theoretical and empirical assessment. Criminal Justice and Behavior, 40(7), 716–736.

46.

Opsal

T. D.

(2011). Women disrupting a marginalized identity: Subverting the parolee identity through narrative. Journal of Contemporary Ethnography, 40(2), 135–167.

47.

Ostermann

(2013). Active supervision and its impact upon parolee recidivism rates. Crime & Delinquency, 59(4), 487–509.

48.

Ostermann

(2015). How do former inmates perform in the community? A survival analysis of rearrests, reconvictions, and technical parole violations. Crime & Delinquency, 61(2), 163–187.

49.

Ostermann

Miller

Matejkowski

(2013). Are “parole failures” parolees when they fail? Incorporating supervision terms and the timing of arrest events into parolee recidivism rates. Justice Research and Policy, 15(2), 1–20.

50.

Pearl

N. R.

(1998). Use of community based social services to reduce recidivism in female parolees. Women & Criminal Justice, 10(1), 27–52.

51.

Petersilia

(1999). Parole and prisoner reentry in the United States. Crime and Justice, 26, 479–529.

52.

Petersilia

(2007). Employ behavioral contracting for earned discharge parole. Criminology & Public Policy, 6(4), 807–814.

53.

Petersilia

Turner

(1993). Intensive probation and parole. Crime and Justice, 17, 281–335.

54.

Powers

R. A.

Kaukinen

Khachatryan

(2018). Risk factors for absconding among adult parolees in Colorado. International Journal of Offender Therapy and Comparative Criminology, 62(14), 4622–4641.

55.

Rhine

E. E.

Petersilia

Reitz

K. R.

(2017). The future of parole release. Crime and Justice, 46(1), 279–338.

56.

Salas-Wright

C. P.

Vaughn

M. G.

Reingle Gonzalez

J. M.

(2016). Drug abuse and antisocial behavior: A biosocial life course approach. Palgrave Macmillan.

57.

Schram

P. J.

Koons-Witt

B. A.

Williams

F. P.

III McShane

M. D.

(2006). Supervision strategies and approaches for female parolees: Examining the link between unmet needs and parolee outcome. Crime & Delinquency, 52(3), 450–471.

58.

Seng

Lurigio

A. J.

(2005). Probation officers' views on supervising women probationers. Women & Criminal Justice, 16(1-2), 65–85.

59.

Shepherd

S. M.

Strand

Viljoen

J. L.

Daffern

(2018). Evaluating the utility of ‘strength’items when assessing the risk of young offenders. The Journal of Forensic Psychiatry & Psychology, 29(4), 597–616.

60.

Shook

J. J.

Sarri

R. C.

(2007). Structured decision making in juvenile justice: Judges’ and probation officers' perceptions and use. Children and Youth Services Review, 29(10), 1335–1351.

61.

Singh

J. P.

Grann

Fazel

(2011). A comparative study of violence risk assessment tools: A systematic review and metaregression analysis of 68 studies involving 25,980 participants. Clinical Psychology Review, 31(3), 499–513.

62.

Smith

Gendreau

Swartz

(2009). Validating the principles of effective intervention: A systematic review of the contributions of meta-analysis in the field of corrections. Victims and Offenders, 4(2), 148–169.

63.

Spruit

van der Put

Gubbels

Bindels

(2017). Age differences in the severity, impact and relative importance of dynamic risk factors for recidivism. Journal of Criminal Justice, 50, 69–77.

64.

Threadcraft-Walker

Threadcraft

M. M.

Henderson

Rembert

(2018). Gender, race/ethnicity and prediction: Risk in behavioral assessment. Journal of Criminal Justice, 54, 12–19.

65.

Ttofi

M. M.

Farrington

D. P.

Piquero

A. R.

Lösel

DeLisi

Murray

(2016). Intelligence as a protective factor against offending: A meta-analytic review of prospective longitudinal studies. Journal of Criminal Justice, 45, 4–18.

66.

Trulson

C. R.

Haerle

D. R.

DeLisi

Marquart

J. W.

(2011). Blended sentencing, early release, and recidivism of violent institutionalized delinquents. The Prison Journal, 91(3), 255–278.

67.

Vaughn

M. G.

DeLisi

Beaver

K. M.

Perron

B. E.

Abdon

(2012). Toward a criminal justice epidemiology: Behavioral and physical health of probationers and parolees in the United States. Journal of Criminal Justice, 40(3), 165–173.

68.

Vazsonyi

A. T.

Mikuška

Kelley

E. L.

(2017). It's time: A meta-analysis on the self-control-deviance link. Journal of Criminal Justice, 48, 48–63.

69.

Viglione

Rudes

D. S.

Taxman

F. S.

(2015). Misalignment in supervision: Implementing risk/needs assessment instruments in probation. Criminal Justice and Behavior, 42(3), 263–285.

70.

Walters

G. D.

Cohen

T. H.

(2016). Criminal thought process as a dynamic risk factor: Variable-and person-oriented approaches to recidivism prediction. Law and Human Behavior, 40(4), 411–419.

71.

Wang

E. W.

Diamond

P. M.

(1999). Empirically identifying factors related to violence risk in corrections. Behavioral Sciences & the Law, 17(3), 377–389.

72.

Whitten

McGee

T. R.

Homel

Farrington

D. P.

Ttofi

(2017). Disentangling operationalizations of persistent offending. Journal of Criminal Justice, 52, 22–33.

73.

Wodahl

E. J.

Garland

Culhane

S. E.

McCarty

W. P.

(2011). Utilizing behavioral interventions to improve supervision outcomes in community-based corrections. Criminal Justice and Behavior, 38(4), 386–405.

74.

Wodahl

E. J.

Ogle

Kadleck

Gerow

(2013). Offender perceptions of graduated sanctions. Crime & Delinquency, 59(8), 1185–1210.

75.

Wood

P. B.

May

D. C.

(2003). Racial differences in perceptions of the severity of sanctions: A comparison of prison with alternatives. Justice Quarterly, 20(3), 605–631.

76.

Zettler

H. R.

Medina

J. C.

(2020). Missing offender-probation contacts and its influence on rearrest. Criminal Justice Policy Review, 31(3), 374–394. https://doi.org/10.1177/0887403419828092