The Drug Court Team and Discretionary Decision-Making: A Mixed-Methods Examination of Sanctioning Responses to Continued Participant Alcohol and Other Drug Use

Abstract

Drug courts attempt to gain participant compliance and alcohol and other drug (AOD) use abstinence through a strategy of moderate and progressive sanctioning, but its discretionary application possesses the capacity for disparity across participants and behaviors. The purpose of this study was to examine the drug court team’s (DCT) discretionary use of sanctions in response to continued participant AOD use. A mixed-methods approach was used for analyzing agency data (n = 1,032) and interviews of five members of the DCT. Data were collected from an adult felony drug court over a 6-year period (2008–2013) and use to answer the following research question: “What participant characteristics and program performance measures affected sanctioning outcomes?” We found that offender attributes did play a role in the sanctioning decision, but program performance measures were stronger predictors of sanction type.

Keywords

drug court punishment rehabilitation deterrence

Introduction

The drug court model employs two complementary components—substance abuse treatment and intensive supervision—to bring about client alcohol and other drug (AOD) abstinence and reduce recidivism among offenders (National Association of Drug Court Professionals [NADCP], 2013). The treatment component of this model focuses on rehabilitation that closely aligns with participant needs (Birgden, 2004). The supervision component relies on surveillance and punishment to obtain participant compliance and desistance from AOD use (Burns & Peyrot, 2003; Taxman, 2002). Drug courts use incremental and progressive punishments to deter future violations rather than program revocation for noncompliance (NADCP, 2013), which is often viewed as a strength and an integral component of the drug court model (Zweig et al., 2012). The punitive measures doled out by drug court team (DCT) have been tied to program outcomes and reduction in participant recidivism (Brown et al., 2010; Zweig et al., 2012); however, drug courts possess a capacity to apply sanctions differently across participants (Zweig et al., 2012).

Frequent drug screenings and immediate and moderate sanctioning are the primary tools used by drug courts to coerce AOD abstinence (Carey et al., 2012). Drug testing is seen as an opportunity to continuously assess a participant’s level of care (Chandler et al., 2009); however, sanctions are also levied to respond to the noncompliant behavior (NADCP, 2013). NADCP (2013) recognize that a reasonable amount of discretion should be applied to address individual needs (Carey et al., 2012; Zweig et al., 2012), despite critics advocating complete equity in sanctioning outcomes (Taxman et al., 1999). The discretionary use of punishment and disparate sanctions may engender feelings of unfairness, in which participants may question the legitimacy of the program (Goldkamp, 2002; Marlowe, 2008; Taxman, 2002). These perceptions may be amplified in instances of positive drug screens as prior research has found the harshest sanctions are reserved for those who continue their AOD use (Guastaferro & Daigle, 2012). Despite the breadth of drug court research, no study has examined the discretionary responses to positive drug screens and individual-level characteristics that may influence the sanctioning decision.

The purpose of this study was to examine a DCT’s discretionary use of sanctions to continued participant AOD use. We focused on responses to positive AOD screens due to its centrality in the drug court model. There has been only one other study that has quantitatively examined predictors of sanctioning responses (Guastaferro & Daigle, 2012) and found that positive drug tests were sanctioned more severely than other program violations. In the current study, a mixed-methods approach was used. We quantitatively analyzed 1,032 sanctioning events, which stemmed from positive AOD screens, as well as offender characteristics and program performance measures on the sanctions levied in positive AOD screens. Qualitatively, we interviewed the entire DCT (with the exception of two judges), focusing on how sanctioning decisions were made. Our primary research question was, “What participant characteristics and program performance measures affected sanctioning outcomes?” We focused on the relationship between offender characteristics, type of substance used, previous violations, duration in the program, and the type of sanction(s) rendered. Through qualitative interviews, we concentrated on the decision-making processes of the DCT when administering sanctions.

The current research contributes to the existing literature on drug court sanctioning in a few distinct ways. To our knowledge, there are no studies that used multivariate analysis to isolate the effects participant characteristics and program performance measures have on drug court sanctioning decisions. Second, we isolated the substance detected to explore any potential bias that may exist based on the type of drug used. Finally, we used a mixed-methods approach to understand the decision-making process of DCT and the context in which sanctions are rendered.

Review of the Literature

Drug courts employ contingency management techniques through sanctions to bring about participant AOD abstinence (Taxman, 2002). Fundamental to this strategy are principles of specific deterrence, where punishments should be delivered in a swift and certain manner with relatively mild to moderate punishments (Kleiman, 2001; Marlowe, 2008; Taxman et al., 1999). Conventional supervision programs are often reliant on unpredictable revocation hearings and subsequent incarceration as a response to noncompliant behaviors (Kleiman, 2001). Yet, supervision revocation fails to account for the impulsivity and risk-taking that leads to relapse in addicted persons (Kleiman, 2001; Rosenthal, 2002).

Previous drug court evaluations have outlined the application (Arabia et al., 2008; Guastaferro & Daigle, 2012; Hawken & Kleiman, 2009; Lindquist et al., 2006), perceptions (Goldkamp, 2002; Wolfer, 2006), and effects (Brown et al., 2010) of sanctions. Programs often collaboratively employ a variety of possible sanctions, which include nonincarceration sanctions (i.e. community service, written assignments, time in the jury box) and short periods of confinement (Arabia et al., 2008; Burns & Peyrot, 2003; Guastaferro & Daigle, 2012; Hawken & Kleiman, 2009). A punitive strategy to induce abstinence may appear antitherapeutic, but the use of sanctions is considered a key component of the drug court model (Carey et al., 2012). Punishment is used to interrupt relapse, teach accountability, and motivate participants (Burns & Peyrot, 2003; Lindquist et al., 2006).

Failure to administer sanctions swiftly may decrease the efficacy of punishments (Carey et al., 2012). Light sanctions may create an environment in which participants become conditioned to sanctions (Marlowe, 2008). Conversely, overly severe punishments may have a ceiling effect or participants may become resistant to the program. The NADCP suggest progressive, or graduated sanctions, as a best practice strategy (NADCP, 2013). This protocol calls for relatively light sanctions for initial program infractions while increasing the magnitude for subsequent noncompliant behavior. In addition, research suggests jail sanctions should be used sparingly, and if used, confinement should not exceed 5 days (Carey et al., 2012). Overall, research has found that the use of graduated sanctions has lessened participant AOD use and reduced recidivism within a drug-involved offender population (Hawken & Kleiman, 2009; Zweig et al., 2012). Thus far, there has been little debate regarding the utility of coerced rehabilitation, but the discretionary manner in which it is imposed appears to be unsettled (Marlowe, 2008; Taxman, 2002; Zweig et al., 2012).

Sanctioning Discretion

The use of judicial discretion is well documented in the literature, particularly as it relates to punishment (Albonetti, 1991; Steffensmeier et al., 1998). Studies have demonstrated that judicial decisions are often influenced by both legal (criminal charge, criminal history) and extralegal factors (gender, age, and race). Punishments are likely to vary across offense type and severity; however, sentencing disparity becomes problematic when legally irrelevant characteristics are found to influence decisions (Rodriguez et al., 2006; Warren et al., 2012). In drug courts, the discretionary use and sanctioning disparities may be more salient when participants feel a sense of unfairness and resistance toward program goals (Goldkamp, 2002). Some scholars argue sanctions should be devoid of discretion and affixed to a sanctioning schedule (Marlowe, 2008; Taxman, 2002; Taxman et al., 1999). This would provide participants a sense of control over their supervision and generate feelings of program legitimacy (Marlowe, 2008; Taxman, 2002). Although some programs have adopted this strategy to limit judicial discretion (Guastaferro & Daigle, 2012; Marchand et al., 2006), research has found that drug court professionals prefer to respond to noncompliance based on individual circumstances (Lindquist et al., 2006; Satal, 1998). Moreover, prior studies have found programs with rigid sanctioning templates are less successful at reducing recidivism than programs that employ some discretion (Zweig et al., 2012).

Drug court professionals often negatively view standardized protocols that possess a lack of flexibility because it limits their ability to tailor sanctions to individuals (Lindquist et al., 2006). They argue that the efficacy of any punishment is dependent upon a participants’ personal history, circumstances, and perceived severity of sanction-type. Satal (1998) reported that many judges recognize that a fixed schedule of sanctions would be most fair, but they do not want to be constrained in sanctioning decisions. Moreover, differences in punishment are not eliminated with fixed sanctioning schedules since existing standardized protocols distinguish between violation-types (noncompliance infractions and AOD relapse) (Arabia et al., 2008; Guastaferro & Daigle, 2012). Regardless, positive AOD screens are frequently met with harsher punishments or more cumbersome treatment responses (Arabia et al., 2008; Guastaferro & Daigle, 2012; Lindquist et al., 2006). Although this differentiation is considered legitimate, sanctioning discretion leaves room for disparity to exist within similar violations, particularly responses to relapses.

To our knowledge, no study has examined the offender and incident-level characteristics of sanctioning responses to program drug use violations. Beyond extralegal offender attributes, several incident-level characteristics may influence a DCT’s response to continued AOD use, such as the number of previous violations, past sanctions imposed, timing of infractions, and substance(s) being abused (Guastaferro & Daigle, 2012). Drug court scholars make no mention of differentiating punishments based on the substance(s) detected during AOD screens (Marlowe, 2008). Moreover, we have no knowledge of drug courts that account for specific substances in their standardized sanctioning protocols (Guastaferro & Daigle, 2012; Marchand et al., 2006). This does not suggest that considering substance use would be illegitimate or extralegal in sanctioning decisions. In fact, ignoring this factor may fail to recognize the realities of differentiated drug use. Legal criminal codes and prior research provide some support in the differentiation between drugs (Hartley et al., 2007; Spohn & Sample, 2013) and past studies have found that program outcomes and recidivism have been affected by participants’ drugs of choice (Gottfredson et al., 2008; Hickert et al., 2009). Hickert and colleagues (2009) found participants who report cocaine and other stimulants as their drug of choice were less likely to successfully complete drug court. In addition, research examining recidivism rates of drug treatment participants has found those who use alcohol and cocaine during their participation may be more likely recidivate post-participation (Gottfredson et al., 2008).

In sum, existing drug court research has provided a great deal of information regarding the application and effects of drug court sanctions. Yet, there is a paucity of research related to the discretionary use of sanctioning in drug courts and the possible disparity of imposed punishments based on incident-level characteristics. We sought to fill this gap in the literature by examining a DCT response to violations of a program’s AOD use policy. We employed a mixed methodology to determine whether sanctioning differences existed based on incident-level characteristics, such as previous violations, timing of violations, and drugs detected during screenings. Quantitatively, we conducted a multivariate analysis to determine the existence of sanction disparity based on certain violations and offender characteristics. Qualitatively, interviews with a DCT were conducted to provide context to potential sanctioning disparities. Combined we sought to answer our primary research question, “What participant characteristics and program performance measures affected sanctioning outcomes?”

Research Design

We conducted a convergent mixed-methods design, in which we utilized both quantitative and qualitative data to answer our research questions. In a convergent design, both sets of data are combined to provide multiple perspectives of a phenomenon (Creswell, 2014). Within a mixed-methods design, the quantitative analyses tend to produce relationships between variables, whereas qualitative analyses provide a deeper understanding on various roles and perspectives.

Data for the current study derive from an adult felony drug court serving a metropolitan area in the Midwestern United States. The program was established in 1997 and moved solely to a post-plea program in 2008. The supervision element of the program requires participants to submit random AOD screenings approximately 2 to 3 times a week, attend regular judicial status hearings (once a week for new clients), and report to case managers as directed. In addition, participants attend varying levels of substance abuse treatment, which is informed by a chemical dependency assessment. The program maintains a wide range of treatment options, ranging from standard outpatient to long-term residential treatment. The program does not comply by a rigid schedule of sanctions, but administers punishments on a case-by-case basis. Participants who yield a positive AOD screen can expect a response ranging from a fine, short periods of incarceration, to an increased level of care. Finally, the program is delineated into three phases and, on average, it takes participants 12 to 36 months to complete the program.

Quantitative Data Collection

This research examined discretionary sanctioning decisions by a DCT in response to the detection of continued AOD use by their participants. Data were collected from all participants (N = 824) who entered the program from 2008 to 2013. From this population, 560 individuals tested positive for continued AOD use at least once during the program. The unit of analysis was recoded from participants to individual sanctioning events. Specifically, events were coded as each sanctioning hearing subsequent to a participants’ positive AOD screen. This yielded a final sample of 1,032 incidents. Offender demographics and program performance information were gathered from a web-based management information system that included AOD monitoring results and sanctioning decisions.

Dependent Variable

Sanction type was the dependent variable and included four categories—fines, nonincarceration, jail, and the Day Reporting Center (DRC). All positive urine analyses (UA) resulted in an automatic US$10 fine. At no point did a fine exceed the US$10 amount, nor were they ever discretionarily administered. Due to the lack of decision-making on the part of the DCT, fines were viewed as a non-sanction and served as the reference category. Nearly half of all events resulted in only a fine (n = 490; 47%; see Table 1). The category of nonincarceration included sanctions beyond US$10 fine but did not include a period of incarceration. These sanctions included increased AA, the Offender Work Programs (8 hr of unpaid work with the county jail work detail), community service work (range from 8 to 40 hr), and chemical dependency education. This group of sanctions could not be analyzed separately due to cell population (events per predictive variables) considerations. Nonincarceration responses were the least utilized response to continued AOD use, making up 11% of all sanctioning events (n = 109). Jail sanctions were administered in 18% (n = 179) of events and ranged from 1 to 10 days.

Table 1.

Sample Characteristics (N = 1,032).

Variables	Frequency	%	M	SD
Dependent variable: sanction
Sanction
Fines	490	47.48	—	—
Nonincarceration	109	10.56
Jail	179	17.34
DRC	254	24.61
Independent variables: offender characteristics
Age of offender (years)	—	—	30.19	9.25
18–24	402	38.95	—	—
25–32	308	29.84	—	—
≥33	322	31.20	—	—
Sex
Male	698	67.64	—	—
Female	334	32.36	—	—
Race
White	628	60.85	—	—
Non-White	404	39.15	—	—
Treatment phase at hearing
Phase 1	928	89.92	—	—
Phases 2 and 3	104	10.08	—	—
Employment at hearing
Unemployed	537	52.39	—	—
Employed	488	47.61	—	—
Misdemeanor criminal history	—	—	2.14	2.16
Felony criminal history
Yes	913	88.47	—	—
No	119	11.53	—	—
Independent variables: program violations
Alcohol positive	199	19.28	—	—
THC positive	372	36.05	—	—
Hazardous drug positive	461	44.67	—	—
Number of positives at one screen	—	—	1.17	0.469
Multiple positives
One	898	87.02	—	—
More than one	134	12.98	—	—
Number of previous positives	—	—	0.832	1.16
Number of previous general violations	—	—	0.300	0.745
Number of U/A technical violations	—	—	0.839	1.86
Time in program at positive screen	—	—	72.59	105.64
0–2 weeks	418	40.50	—	—
2–4 weeks	144	13.95	—	—
30–90 days	224	21.71	—	—
91 and above	246	23.84	—	—

Note. DRC = Day Reporting Center; THC = tetrahydrocannabinol.

The fourth sanction category was referral to the DRC. The DRC is a jail-based intervention program where individuals are incarcerated for an indeterminate amount of time. Participants are considered county jail inmates but are housed separate from the general population. While in DRC, individuals participate in substance abuse education program, job-skills training, and adult basic education to earn their general equivalency diploma (GED). Individuals may earn the privilege of leaving the facility to look for employment or work at their current job. Generally, participants remain in this program until bed space is identified for them in a residential treatment facility or halfway house. A referral to DRC was made in 25% (n = 254) of all sanctioning events.¹

Independent Variables

Offender characteristics

We included both offender characteristics and program violation measures as independent variables (Table 1). The average participant age was 30.19 years (SD = 9.22), with over two thirds being male (n = 698, 68%). Due to a lack of heterogeneity in our sample, race was coded dichotomously (White = 0, non-White = 1). In nearly 61% of all events, the participants sanctioned were White (n = 628, 60.85%) and 52% (n = 537) were unemployed (0 = employed, 1 = unemployed) at the time of the sanctioning event. Unlike some drug courts, this program admitted offenders with previous criminal convictions. On average, participants had been previously convicted of two (SD = 2.16) misdemeanors, while 11% (n = 119) had at least one prior felony conviction (yes = 1, no = 0).

Positive drug type

Positive drug type was measured as a categorical variable with three attributes—alcohol, tetrahydrocannabinol (THC; marijuana), and hazardous drugs. Particular attention was paid when coding THC positives. According to the DCT, once a participant tests positive for marijuana use, they are given approximately 4 weeks to provide a negative result and are not sanctioned during this interim period beyond a fine. In these instances, we coded the first positive THC screen as a violation but did not count subsequent positive screens that occurred during the 4-week interim period. Unfortunately, we were unable to isolate the direct effect of cocaine, heroin, methamphetamine, MDMA (3,4-methylenedioxy-methamphetamine), and prescription medication (i.e. opiates, benzodiazepines) due to limited variance across drug types. We collapsed these drugs into one attribute—hazardous drugs. To be in possession of any hazardous drug is a felony, which distinguishes them from both alcohol and marijuana within the state’s legal code. Nearly 20% (n = 199) of the sanctioning events were responses to the detection of alcohol use (Table 1), approximately 36% (n = 372) was marijuana use (reference category), whereas 461 (45%) yielded hazardous drug use. In several instances, positive AOD screens detected the use of a hazardous drug and marijuana. Consistent with the hierarchy rule, these events were coded as hazardous drug positives.

Program performance measures

The current data included four variables that capture the participants overall compliance in the program up to the sanctioning event. Similar to multiple charges in a sentencing decision, we measured the number of distinct substances detected in one AOD screen. The sample averaged 1.170 detected substances per sanctioning event. Moreover, we included a dichotomous variable which measured the number of positive screens incurred during the period between status hearings (0 = one positive, 1 = multiple positives). Participants were required to attend drug court status hearings once a week or twice a month, potentially submitting to two to six AOD screens between hearings. It was not uncommon for participants to yield positive results in more than one screening. In nearly 13% (n = 134) of events, the DCT was responding to multiple positive screens yielded by a single participant (Table 1). We also included three measures to capture participants history of program noncompliance—the number of previous positive AOD screens, technical screening violations (i.e. no show for screen, diluted sample, tampering with sample, etc.), and general violations. At the time of the sanctioning event, participants averaged less than one previous positive AOD screen violation (M = 0.832, SD = 1.16), 0.839 (SD = 1.86) previous noncompliance violations, and 0.300 (SD = 0.745) general violations.

Timing of sanctioning event

Previous research has found that most sanctions are levied within the first couple of months of program participation (Guastaferro & Daigle, 2012; Marlowe, 2008). We included two variables to capture the timing of sanctioning events—treatment phase and timing of events. Drug court programs are typically delineated into phases, where the completion of each phase signifies the completion of specific program requirements. This particular drug court separates the program into three phases. Within the first phase, participants must complete their primary treatment episode. A positive AOD screen in the second and third phases indicates that the participant relapsed after their completion of primary treatment. The variable—phase at sanctioning event—was made binary, collapsing events in the second and third phases (Phase 1 = 0, Phases 2 and 3 = 1). The vast majority (n = 928, 90%) of positive screens occurred in Phase 1 (Table 1). The program specifies that it has no set time designation in which a participant should spend in each phase.

Finally, we included a categorical variable with four attributes capturing the timing of the sanctioning event (0–14 days, 15–29 days, 30–90 days, 91 days and above). Forty percent of positive AOD screens occurred during the first 2 weeks of program participation (reference), whereas 14% (n = 144) occurred between 15 and 29 days. Nearly 22% (n = 224) of all sanctioning events occurred within the second and third months of participation, whereas 246 (24%) events took place after the three-month period.

Quantitative Analytical Strategy

We examined the direct effects of continued participant drug use on the DCT’s sanctioning decision through a multinomial probit model (see Table 2). The multinomial regression model is appropriate due to the existence of a four-category dependent variable (Long & Freese, 2006) and the use of a probit regression is advised within the sentencing literature (Bushway et al., 2007; Koons-Witt et al., 2014).² The current analysis is similar to the initial in/out decision in two-part sentencing analyses. As Bushway and colleagues (2007) explained, probit models assume normality, rather than log normality as its functional form, thus a more appropriate test than logistic regression. Results from probit models cannot be converted into odds and the coefficients are not directly interpretable. Subsequently, the effects event characteristics have on the sanctioning decisions are reported in average marginal effects, rather than log odds. Finally, potential collinearity issues were examined through a Pearson’s correlation matrix and variance inflation factors (VIF). The highest VIF was 1.96 with an average of 1.29.

Table 2.

Predictors of Sanction-Type (Average Marginal Effects).

Independent Variables	Fines	Nonincarceration	Jail	DRC
Female	−0.0369887 (0.0276)	0.0211776 (0.0214)	−0.0363383 (0.0239)	0.0521493* (0.0265)
Non-White	0.0150451 (0.0275)	−0.0060135 (0.0205)	0.0506579* (0.0248)	−0.0596894* (0.0259)
Age (18–24: reference)
25–32	−0.0576519 (0.0310)	0.0155185 (0.0224)	0.0036373 (0.0271)	0.0457707 (0.0308)
≥33	−0.0014711 (0.0337)	0.0402307 (0.0251)	0.0068236 (0.0295)	−0.0455832 (0.0298)
Unemployed	−0.088969 (0.0265)	−0.0158464 (0.0200)	−0.0356759 (0.0235)	0.0704191* (0.0253)
Misdemeanor history	−0.0125723* (0.0062)	0.0008401 (0.0046)	0.00443 (0.0053)	0.0073022 (0.0058)
Felony history	0.0230593 (0.0397)	−0.0492135* (0.0238)	0.066623 (0.0380351)	−0.0404688 (0.0350)
Positive drug (THC: reference)
Alcohol	−0.2116731** (0.0408)	−0.015331 (0.0260)	0.103079* (0.0346)	0.1239252** (0.0346)
Hazardous	−0.2242223** (0.0360)	0.0153715 (0.0265)	0.0306508 (0.0291)	0.1782** (0.0320)
Multiple positive screens	−0.1881678** (0.0399)	−0.0293318 (0.0270)	0.021874 (0.0359)	0.1956222** (0.0416)
Number of positives at one screen	0.0288268 (0.0294)	−0.0454451 (0.0278)	−0.0048793 (0.0269)	0.0214875 (0.0258)
Number of previous positive screens	−0.0738854** (0.0140)	0.0135477 (0.0090)	0.0050252 (0.0102)	0.0553126** (0.01110)
Number of U/A tech violations	−0.003364 (0.0097)	−0.0152248* (0.0076)	−0.0091469 (0.0073)	0.0277257** (0.0075)
Number of general violations	−0.0188314 (0.0191)	0.0255012* (0.0121)	−0.0082501 (0.0144)	0.0015804 (0.0158)
Phases 2 and 3	−0.0384603 (0.0573)	0.1117648* (0.0545)	0.0611369 (0.0446)	−0.1344415** (0.0319422)
Time in program at positive screen (1–14 days: reference)
15–30 days	−0.2492352** (0.0470)	0.0366567 (0.0307)	0.1166957** (0.0363)	0.0958828* (0.0390)
31–90 days	−0.34018** (0.0442)	0.1206253** (0.0330)	0.1301023** (0.0323)	0.0894524* (0.0365)
91 and above	−0.3536909** (0.0579)	0.0121918 (0.0331)	0.2041691** (0.0479)	0.13733* (0.0479)

Note. Standard errors are in parentheses. DRC = Day Reporting Center.

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

Qualitative Data Collection

We conducted face-to-face, semi-structured interviews with five members of the DCT from January to April 2011. This included the entire DCT, with the exception of the two judges. During the interviews, we began by asking, “How long have you been with drug court?” “Please describe your professional background before joining the drug court team.” We followed background questions with inquiries regarding types of sanctions, responses to positive drug tests, and decision-making processes in regard to sanctioning. Interview questions included, but were not limited to, “What is the standard sanctioning response when a person tests positive for drug use?” “Do you consider the severity of drugs when considering how a client will do in the program?” “Are there particular drugs a client may use or abuse that you consider when recommending sanctions?” These interviews were conducted using informal conversational techniques (Kvale & Brinkmann, 2009), in which participants were free to discuss their thoughts about what affected their sanctioning decisions. This allowed us to understand what they found most important in their decision-making process. Interviews lasted between 1 and 3 hr. All interviews were recorded for accuracy and transcribed verbatim.

Qualitative Analytical Strategy

The qualitative analysis was conducted in three phases. Based on the open-ended nature of the interviews, we began the analytic process by reading the interviews holistically to understand (a) whether the DCT used a sanctioning protocol in their decisions, (b) the extent to which length of time and previous violations affected sanctioning decisions, and (c) whether the type of substance detected influenced sanctioning outcomes. Through this initial read, two investigators created a list of codes based on patterns in the data, which included five to 10 shorthand codes often referred to as “lean coding” (Creswell & Poth, 2017). In addition, they came to an agreement on the lean codes, which was calculated between 85% and 97% agreement. After the initial coding agreement, both investigators discussed all inconsistencies and came to a 100% agreement on the coding scheme. Once all the interviews were initially read and a lean list of codes were created, we re-read the interviews line-by-line and the number of codes increased. In the third phase, we conducted cross-case comparisons of our codes to identify potential themes and subthemes that emerged in our data.

Quantitative Results

We employed a multinomial probit model to answer our primary research question—what participant characteristics and program performance measures affected sanctioning outcomes? The model was significant, which included a sample size of 1,024 after listwise deletion, χ²(1,010) = 425.97, p < .001 (Table 2). We found that offender demographics played a minimal role in the DCT’s sanctioning decisions (Table 2). Females marginally had a greater probability than males of being sanctioned to the DRC than receiving a fine (AME = 0.052). The probability of being sanctioned to jail rather than incurring a fine was 5 percentage points higher for non-Whites than for Caucasian participants. Yet, the probability of being sanctioned to the DRC over a fine was nearly 6 percentage points lower for these same individuals. The probability of an unemployed client being sanctioned to the DRC rather than a fine was seven points greater than those who were employed. In regard to criminal history measures, the probability of participants receiving only a fine for a positive AOD screen slightly decreased (AME = −0.013) for each previous misdemeanor conviction. The existence of a prior felony conviction had little impact on the DCT’s response to continued substance use. Participants who had previously been convicted of a felony were less likely (AME = −0.049) to receive a nonincarceration sanction than a fine, compared with those with no felony history.

The DCT’s sanctioning response to detected alcohol use or other drugs was a primary focus of the analysis. Relative to marijuana use, clients who tested positive for alcohol or a hazardous drug were significantly less likely to incur only a fine as a response to their continued use. Moreover, participants with a positive alcohol screen experienced a probability of a jail sanction, rather than a fine, 10 percentage points greater than those who tested positive for marijuana. These individuals were also more likely to be sanctioned to DRC (AME = 0.124) over incurring merely a fine. We found a pronounced difference in the DCT’s use of DRC sanctions in response to a hazardous drug positive. Participants who yielded a positive AOD screen for hazardous drugs experienced a probability of a DRC sanction over a fine 18 percentage points greater than those who tested positive for continued marijuana use.

Statistically significant results related to additional program performance measures were limited to fines and DRC sanctions. Clients who appeared for a status hearing with multiple positive screening events since their last judicial appearance had an increased probability (AME = 0.196) of receiving a DRC sanction relative to a fine. This result was not surprising, as clients with multiple positives were demonstrating a prolonged relapse that may indicate to the DCT that a more intensive intervention is necessary. Participants with an existing history of positive AOD screens were also more likely to incur a DRC sanction than having only a fine levied. For every additional previous positive screen, the probability of receiving only a fine decreased by 7%, where the probability of a DRC sanction increased by 6%. Participants who completed their primary treatment episode and other positive compliance measures were sanctioned differently than those who were in Phase 1 of the program. Clients who yielded a positive AOD screen in Phase 2 or 3, compared with those in Phase 1, were less likely (AME = −134) to be sanctioned with DRC than a fine. In lieu of either jail or DRC, clients in Phase 2 were more likely than those in Phase 1 to receive a nonincarceration sanction (AME = 0.112) than a fine.

Finally, we examined the timing of positive AOD screens as it related to particular sanctions. We found that clients who tested positive for continued AOD use were less likely to be sanctioned with only a fine as their time increased in the program. Similarly, we found that the longer participants were in the program, the more likely they were sanctioned to jail or DRC relative to a fine. Those who yielded a positive AOD screen during the 15- to 30-day period of participation possessed the probability of receiving a jail sanction, over a fine, was 12 percentage points greater than those who tested positive during the 1- to 14-day period. The probability of a jail sanction, over a fine, increased to 13 and 20 percentage points within the 31- to 90-day and more than 90-day periods, respectively. In addition, clients who tested positive for AOD use between 15 and 30 days were more likely (AME = 0.089) to be sanctioned with DRC, rather than a fine, than those in the first 2 weeks of participation. This difference remained relatively consistent over the 31- to 90-day period, but increased in probability to 14 percentage points greater after 90 days.

Qualitative Results

After reviewing the interview transcripts of the DCT, we found a number of prominent themes that influenced their decisions when recommending sanctions. We specifically focused on the DCT’s philosophy regarding the use of sanctions and how they were levied in instances of positive AOD screens. Moreover, we examined the extent to which sanctioning decisions were based on the substance used by clients to gain context to our quantitative findings.

The Overall Sanctioning Philosophy

Although this program utilized a sanctioning menu rather than a standardized schedule, all members of the DCT reported the program operated under a graduated sanctions philosophy. One member of the DCT conveyed an informal hierarchy of sanctions:

At the bottom [of the hierarchy] you may see self-help meetings and a person may be given more self-help meetings as a sanction . . . The next level is offender work program where you have to work for eight hours . . . The next one up you got is jail, which is utilized a lot. The next level is probably seven days in jail, or a just revoke their bond and have been transported back the next week. The next level is probably termination. We do not see taking a person into custody and day reporting center as a sanction. It is more of a therapeutic move.

Although the use of graduated sanctions was widely accepted among the DCT, each member indicated there were caveats rather than a strict adherence. The DCT recognized that sanctions should be individualized to participants and specific violations. One member of the DCT stated,

Everything is done on a case-by-case basis; it is not always black and white. If we got somebody where jail is like a second home to them and it really does not faze them, how was that a sanction? So, we will do something else with that person. It all depends on what they did, the consequence fits the crime . . .

All of the DCT conveyed that sanctions were doled out with the individual in mind. The DCT has a variety of sanctioning options to choose from, but their decisions often depended on the type of infraction, prior violation, and choice of substance use.

Prior History and Drug Court Sanctions

In drug courts, the participants’ prior behavior and noncompliance history play a significant role in sanctioning decisions. One member described the informal process on how previous violations are taken into account, the progressive nature of sanctions, how case managers make recommendations to drug court judges. One of the drug court members stated,

Well, we don’t go back in the file. Everybody knows their people well enough and it depends on what has happened. If the behavior is relatively minor, then everyone pretty much considers what the least punitive sanction. If it is a huge behavior problem, even if there wasn’t a problem before, we will go for jail to get their attention . . . [I]t depends on their case manager. Some of them are a little easier and some of them are not. Then it depends on the judge . . . So, a lot of times we try to discuss it amongst ourselves before we ever get to the judge. So, we can at least bring one opinion to the judge and the judge has the final word.

The DCT also agreed that sanctions are not always levied in a progressive manner. Participants may receive a relatively severe sanction as a response to a first violation. One member highlighted a specific incident: “this particular client came to my orientation intoxicated. I took them downstairs and he blew a point .088 . . . So his sanction, right off the bat, was jail for two days.” Participants’ prior behavior, such as criminal history or overall risk, may also factor into the initial sanctioning decision. As one member stated,

If you already have some guy that is borderline coming into this program and they got a horrible record, the chances are that I’ll be a little bit more quicker to respond to going right to jail or to the day reporting center, or to lock them up . . . If this is the first time, they’ve been to jail I may give them a day of work first.

The DCT were likely to use what they know about their participants’ backgrounds to make decisions about punishments. Based on expertise and experience, they acknowledged that some participants needed tougher immediate sanctions than others.

Members of the DCT also addressed the relationship between specific violation-types and resulting sanctions. One of the DCT members suggested there existed two sanctioning tracks, one for noncompliance and one for positive AOD screens. One member stated, “There are two different systems. If you are saying you need more support such as more AA meetings or another sponsor . . . that is on this side at the house (continued AOD use).” One of our participants described the positive AOD track:

[A]s very progressive. If they keep using and say I just can’t stop using . . . I still handle it the same way and say your honor we need to protect this person because he needs to be protected from himself and his friends and family and whoever he’s going out with . . . We need to put him some place where he can’t get this stuff for a while, so he knows what it feels like to be clean and sober . . . Usually put them in county for seven days or however many days it takes to get them in the day reporting center.

One member of the DCT suggested that the use of jail can be more of a therapeutic response rather than a punitive one:

If they told us they used again and then they used again . . . well this is getting ridiculous. They can’t use because this is drug court. Then I say that they need to find out what it means to not be around it, so it’s not a sanction or punishment, but sort of timeout. It’s a “go rest for a little” while you clean your body out, eat some good food for a change, and we will see you and a week . . .

Jail time is used as a means for punishment, but importantly, it is another way the DCT help drug court participants get sober and clean. Before resorting to kicking participants out of the program, the DCT seems to use all available options to help their participants.

A Client’s Drug of Use

The DCT conveyed that the type of drug clients used did not factor into their sanctioning decisions. When asked about differential treatment in sanctioning outcomes based on the drug of use, one of the members of the drug team stated,

Everything is the same. We treat alcohol the same as heroin, heroin the same as marijuana, etc. We treat them all and think of them all as totally the same. Another member of the team stated, “we don’t go . . . well it was just alcohol as opposed to, oh my gosh, it was meth. A drug is a drug is a drug and you’re not supposed to be using.

DCT members were clear as they did not focus on the client’s choice of drugs. Their concern was centered on the dangerousness of various drugs and getting their clients the help that they needed. They were more concerned about the client’s health and severity of drug abuse than the choice of drug. One member expressed,

As far as like to danger to self or others, I may take that into consideration . . . If I have somebody that has a history of chronic alcoholism and they even have some medical issues because of it, like liver or hepatitis, then I am going to look at that. If I have someone that is smoking marijuana, but they seem to be smoking this stuff that is always laced with something, then I am going to look it that a little more . . . Are they putting themselves or their children in more danger because of their using, if so, then we may have them go to DRC because of that . . . So it’s not always so much the drug itself, but the danger to themselves.

One DCT member suggested the drug of choice made no difference in the sanctioning decisions, but later admitted that some substances may be more dangerous than others:

I think we see more danger with certain drugs. People who are using needles to shoot heroin are probably at greater risk than a pot smoker. In terms of the big picture, I don’t think we see it differently [when making sanctioning decisions]. There are certain things or certain drugs that are more harmful, medically. People who are using OxyContin are not only at great risk using them, but when they withdraw from it it’s also dangerous. Alcohol is the same.

The DCT members recognized that implicit bias was difficult to eliminate from the decision-making process; therefore, they participated in training and tried to make a conscious effort to reduce bias when making decisions. One DCT member stated,

So do the case managers, at times, show bias, sure they do. They are human beings. I don’t know if we can take that out of any human involved process . . . Hopefully if we have done our homework and we’ve written things down in our case management notes and we are not doing that.

Overall, our qualitative results indicated that sanctioning decisions were largely based on the client’s personal history and addiction needs. DCT members emphasized the importance of providing individualized treatment and focusing on their client’s health care when taking drug of choice into consideration. These qualitative results are somewhat consistent with the quantitative results, with the exception of the influence on drug of choice on sanctioning decisions.

Discussion and Conclusion

The current study sought to fill a gap in the drug court literature by examining a DCT’s discretionary sanctioning decisions in response to detected AOD use. Specifically, we sought answers to our primary research question: “What participant characteristics and program performance measures affected sanctioning outcomes?” We examined the relationship between offender characteristics, type of substance used, previous violations, duration in the program, and the type of sanction(s) rendered. We also concentrated on the decision-making processes of the DCT when doling out sanctions through qualitative interviews.

Criminological literature has consistently found biases in punitive decision-making based on offender characteristics (Hartley et al., 2007; Steffensmeier et al., 1998). Scholars have suggested that court actors are often left with limited offender information, thus base their decisions on stereotypes (Albonetti, 1991; Steffensmeier et al., 1998). A DCT, however, has frequent contact with participants over several months, providing them greater insight to individuals and their circumstances. This unique situation should reduce their reliance on stereotypes and eliminate bias based on offender characteristics (Gibbs, 2020; Hoffman, 2000). The current analysis, however, found moderate bias based on participant age, race, and employment status. Females were more likely sanctioned to the DRC than males upon the detection of continued AOD use, which could be a result of a paternalistic effect (Chesney-Lind, 1986). This potential explanation suggests that the DCT more often perceived that female participants were leading unstable lives and needed “protection” and increased care relative to their male participants.

The disparity in sanctions between Whites and non-Whites was more complicated. Non-White participants were modestly more likely to be sanctioned to jail over a fine, but less likely than White participants to be sanctioned to DRC rather than fined. In terms of liberty lost, the indefinite confinement period of a DRC sanction is more cumbersome than the typical jail sanction. Unemployed compared with employed participants were more likely to be sanctioned to DRC rather than incurring only a fine. DCT members indicated in their interviews that DRC was often used to stabilize participants’ lives. Unemployment can be one factor indicating life instability. In addition, an indefinite period of confinement may jeopardize an employed participants’ job status, where the DCT may be cognizant an individual’s need to maintain employment.

In our interviews, we inquired whether the substance used by clients influenced the type of sanction levied. The consensus across all interviewed DCT members was that the detected substance did not matter in their sanctioning decisions; yet, all of them offered caveats to this proclamation. One drug court professional admitted that, at times, case managers showed bias in their responses to specific drug use. Through the multivariate analysis, we found sanction disparities based on the substance detected. A positive screen for hazardous drugs, compared with marijuana, possessed a greater probability that participants would be referred to DRC relative to being fined. Furthermore, the continued use of alcohol, relative to marijuana, was more likely to be met with a jail or a DRC sanction over a fine. The differential sanctioning based on detected substance use may be a product of the DCT’s perception of immediate dangers associated with a particular drug. Subsequently, the differential sanctioning may not necessarily be a direct result of the substance being used, but the perceived dangerousness the substance represents in participants’ individual life circumstances. As a whole, these case managers believed that the use of a hazardous drug places participants and other individuals in their lives in greater harm than marijuana use. As a result, the recommendations to incarcerate these individuals and/or refer them to DRC were to remove them from danger. Although these explanations may suffice for continued hazardous drug use, the results related to detected alcohol use were more curious.

Unexpectedly, detected alcohol use was sanctioned more severely than marijuana. This result was surprising considering alcohol is a legal substance and few participants reported it as their primary drug of choice. Case managers conceded that they did not believe most participants were abusing alcohol, but merely consuming one to two beers in the evening. Interview responses suggested that the DCT believed that alcohol use can, and frequently does, lead to harder drug use. It is for this reason, perhaps, case managers were proactively attempting to interrupt a possible relapse with a hazardous drug. This explanation may also support the significant finding that alcohol consumption increased the likelihood of jail sanctions than the use of hazardous drugs.³ The DCT may believe that meeting a positive alcohol screen with a short period of incarceration is a mechanism to disrupt a potential relapse. It may also remind participants that any substance use is a program violation, where the detected use of hazardous drugs is cause to increase one’s level of care. Subsequently, case managers were less likely to respond to hazardous drug use with a jail sanction, believing a DRC referral or a more immediate increase in a participant’s level of treatment to be more appropriate. Regardless, this drug court did not take alcohol use lightly and detection was met with punitive measures.

We examined several measures capturing participants’ program performance history. These measures included the number of substances detected in one screen, the number of previous positive screens, the number of prior screening technical violations, and the total number of previous general violations. We believed these variables were potential indicators of progressive sanctioning. Consistent with a graduated sanctions protocol, the number of previous AOD positives and technical violations did impact the use of a therapeutic response. As the number of these violations increased, the likelihood of the DCT responding with a referral to the DRC, rather than only a fine, also increased. These two variables indicated a continued pattern of substance abuse; thus, it should be expected that as the number of these violations increase so the probability of being referred to DRC. Further evidence of progressive sanctioning was found in the impact of the timing of the detected use or the duration one had spent in the program. The DCT responded more harshly to a positive AOD screen the longer the participant had been in the program.

Our qualitative findings in regard to timing are consistent with the quantitative results. All of the drug court members expressed jail time and DRC was utilized to help drug court clients get sober and clean, especially when clients were unable to control their addiction. Jail time was used to protect clients from themselves, especially those who used more dangerous drugs. Our findings indicated that sanctions were doled out in an incremental manner. Scholars have suggested that AOD-involved offenders are better deterred through progressively moderate sanctioning (Marlowe, 2008). The use of relatively severe sanctions early in program or for first infractions may be detrimental to the success of participants. The inconsistent use of sanctions may create feelings of helplessness and illegitimacy toward the program (Marlowe, 2008).

Although our research significantly contributes to the existing literature, we must note certain limitations to the study. First, the quantitative data analyzed were collected by agency professionals and not necessarily for this study (Maxfield & Babbie, 2014). In the quantitative model, we were unable to include a risk/needs measure. We recognize the value of such a measure, but the assessment records kept by the agency contained an insurmountable number of missing data. Despite this limitation, our analysis did include measures of offender risk, which included two criminal history variables—the number of prior misdemeanors and prior felonies. In addition, we included an employment measure that partially captures offenders’ risks and needs.

Second, the interviews were not conducted for the purposes of this study, nor were they intended to answer our specific research questions. These interviews were part of an earlier evaluation of an alcohol screening instrument. For this evaluation, it was unnecessary to interview the judges overseeing the drug court program. We recognize the limitation of our qualitative data due to the omitted opinions and perspectives of the drug court judges; however, this only excluded two DCT members. It is also important to note that the DCT indicated judges agreed with the case managers’ sanctioning recommendations in most instances. In fact, one interviewee suggested that judges adhered to 98% of their recommendations. We believe this lessens the adverse impact of the omitted interviews on our findings.

Third, we cannot guarantee our marijuana coding scheme completely aligned with the drug court professionals. As reported above, the DCT recognized an approximate 30-day detection period for THC and claimed not to consider THC-positives as violations during this 4-week grace period. It is possible the DCT unintentionally provided some participants an extra week or two of a grace period. Finally, our findings are limited regarding its generalizability. Our sample consisted of 6 years of data but was limited to one drug court located in the Midwest of the United States. Although our data derived from a single drug court, we believe our findings are useful for drug courts that serve similar populations.

Overall, our analysis indicates that this drug court makes sanctioning decisions based on individual participant circumstances. This drug court was less constrained and maintained a great deal of discretion in sanctioning recommendations than those with rigid sanctioning templates. Although there are advantages to individualizing punitive responses, bias of legally relevant and irrelevant characteristics exists. Research has shown that a medium amount of discretion within a drug court’s sanctioning protocol reduces a greater amount of reoffending than those programs that use a strict schedule (Zweig et al., 2012). Further research should examine the simultaneous effect drug court sanctioning has on program outcome and recidivism. Prior research has demonstrated that drug courts can reduce recidivism (Wilson et al., 2006), but no research has attempted to parse out the individual effects of sanctioning, program outcomes, and the subsequent adjudication (for program failures) has on future criminality.

We believe the current study contributes to the existing literature by filling a gap in drug court sanctioning. Few studies have undertaken this examination quantitatively (Guastaferro & Daigle, 2012), and to our knowledge, no previous research has used multivariate analysis. Second, we reported sanctioning patterns as responses to positive AOD screens, which allowed us to isolate a major component of the drug court model, responses to continued AOD use, and detect differential sanctioning based on the substance used. Finally, our mixed-methods approach demonstrated the importance of both quantitative and qualitative methodologies in understanding how and in what manner sanctions were doled out. The interview responses provided extensive context and potential explanations to our quantitative findings. We were able to glean insight into the decision-making processes of case managers and their sanctioning decisions.

Footnotes

Declaration of Conflicting Interests

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

Funding

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

ORCID iD

Benjamin R. Gibbs

Notes

Author Biographies

Benjamin R. Gibbs, PhD, is an assistant professor in the Department of Criminal Justice and Criminology at Ball State University. His current research includes sentencing, institutional and community corrections, and sex offenders.

Tusty ten Bensel, PhD, is the graduate coordinator and an associate professor in the Department of Criminal Justice at the University of Arkansas at Little Rock. She specializes in research on sexual violence and victimization, women and crime, identity changes, and reentry.

Madison K. Doyle is a graduate student at the University of Arkansas at Little Rock. Her current research focuses on community corrections and has presented original works at annual conferences for the American Society of Criminology and the Academy of Criminal Justice Sciences.

William Wakefield, PhD, is a professor in the School of Criminology and Criminal Justice at the University of Nebraska Omaha. His research concerns comparative cross-cultural criminal justice and agency research evaluation, focusing on drug court process and outcome measures.

References

Albonetti

(1991). An integration of theories to explain judicial discretion. Social Problems, 38, 247–266.

Arabia

P. L.

Fox

Caughie

Marlowe

D. B.

(2008). Sanctioning practices in an adult felony drug court. Drug Court Review, 61(1), 1–32.

Birgden

(2004). Therapeutic jurisprudence and responsivity: Finding the will and the way in offender rehabilitation. Psychology, Crime & Law, 103(3), 283–295.

Brown

R. T.

Allison

P. A.

Nieto

F. J.

(2010). Impact of jail sanctions during drug court participation upon substance abuse treatment completion. Addiction, 106, 135–142.

Burns

S. L.

Peyrot

(2003). Tough love: Nurturing and coercing responsibility and recovery in California drug courts. Social Problems, 50(3), 416–438.

Bushway

Johnson

B. D.

Slocum

L. A.

(2007). Is the magic still there? The use of the Heckman two-step correction for selection bias in criminology. Journal of Quantitative Criminology, 23, 151–178.

Carey

S. M.

Mackin

J. R.

Finigan

M. W.

(2012). What works? The ten key components of drug courts: Research-based best practices. Drug Court Review, 8, 6–42.

Chandler

R. K.

Fletcher

B. W.

Volkow

N. D.

(2009). Treating drug abuse and addiction in the criminal justice system: Improving public health and safety. Journal of the American Medical Association, 301(2), 183–190.

Chesney-Lind

(1986). “Women and crime”: The female offender. Signs, 12(1), 78–96.

10.

Creswell

J. W.

(2014). A concise introduction to mixed methods research. SAGE.

11.

Creswell

J. W.

Poth

C. N.

(2017). Qualitative inquiry and research design: Choosing among five approaches. SAGE.

12.

Gibbs

B. R.

(2020). Sentencing drug court failures: Judicial considerations with increased offender information. Criminal Justice Policy Review, 31(2), 237–261.

13.

Goldkamp

(2002). An honest chance: Perspectives on drug courts. Federal Sentencing Report, 14(6), 369–372.

14.

Gottfredson

D. C.

Kearley

B. W.

Bushway

S. D.

(2008). Substance use, drug treatment, and crime: An examination of intra-individual variation of drug court. Journal of Drug Issues, 38(2), 601–630.

15.

Guastaferro

W. P.

Daigle

L. E.

(2012). The use of graduated sanctions in a felony level drug court. Journal of Drug Issues, 42(4), 396–419.

16.

Hartley

R. D.

Madden

Spohn

C. C.

(2007). Prosecutorial discretion: An examination of substantial assistance departures in federal crack-cocaine and powder-cocaine cases. Justice Quarterly, 24(3), 382–407.

17.

Hawken

Kleiman

(2009). Managing drug involved probationers with swift and certain sanctions: Evaluating Hawaii’s HOPE. National Institute of Justice.

18.

Hickert

A. O.

Boyle

S. W.

Tollefson

D. R.

(2009). Factors that predict drug court completion and drop out: Findings from an evaluation of Salt Lake County’s Adult Felony Drug Court. Journal of Social Service Research, 35(2), 149–162.

19.

Hoffman

M. B.

(2000). The drug court scandal. North Carolina Law Review, 78(5), 1439–1534.

20.

Kleiman

(2001). Controlling drug use and crime with testing, sanctions, and treatment. In Heyman

P. B.

Brownsberger

W. N.

(Eds.), Drug addiction and policy: The struggle to control dependence (pp. 168–192). Harvard University Press.

21.

Koons-Witt

B. A.

Sevigny

E. L.

Burrow

J. D.

Hester

(2014). Gender and sentencing outcomes in South Carolina: Examining the interactions with race, age, and offense type. Criminal Justice Policy Review, 25, 299–324.

22.

Kvale

Brinkmann

(2009). Interviews: Learning the craft of qualitative research interviewing. SAGE.

23.

Lindquist

C. H.

Krebs

C. P.

Lattimore

P. K.

(2006). Sanctions and rewards in drug court programs: Implementation, perceived efficacy, and decision making. Journal of Drug Issues, 361(1), 119–146.

24.

Long

J. S.

Freese

(2006). Regression models for categorical dependent variables using stata. Stata Press.

25.

Marchand

Waller

Carey

S. M.

(2006). Barry county adult drug court outcome and cost evaluation. NPC Research.

26.

Marlowe

D. B.

(2008). Application of sanctions. Quality improvement for drug courts: Evidence based practices. National Drug Court Institute.

27.

Maxfield

M. G.

Babbie

E. R.

(2014). Research methods for criminal justice and criminology. Cengage Learning.

28.

National Association of Drug Court Professionals. (2013). Adult drug court best practice standards (Vol. I). National Association of Drug Court Professionals.

29.

Rodriguez

S. F.

Curry

T. R.

Lee

(2006). Gender differences in criminal sentencing: Do effects vary across violent, property, and drug offenses. Social Science Quarterly, 87, 318–339.

30.

Rosenthal

T. A

. (2002). Therapeutic jurisprudence and drug treatment courts: Integrating law and science. In Nolen

J. L.

De Gruyter

(Eds.), Found in drug courts in theory and in practice (pp. 145–171).

31.

Satal

S. L.

(1998). Observational study of courtroom dynamics in selected drug courts. National Drug Court Institute, 11(1), 56–87.

32.

Spohn

Sample

L. L.

(2013). The dangerous drug offender in federal court: Intersections of race, ethnicity, and culpability. Crime & Delinquency, 591(1), 3–31.

33.

Steffensmeier

Ulmer

Kramer

(1998). The interaction of race, gender, and age in criminal sentencing: The punishment cost of being young, black, and male. Criminology, 36(4), 763–797.

34.

Taxman

F. S.

(2002). Supervision- exploring the dimensions of effectiveness. Federal Probation, 66(2), 14–28.

35.

Taxman

F. S.

Soule

Gelb

(1999). Graduated sanctions: Stepping into accountable systems and offenders. Prison Journal, 79(2), 182–204.

36.

Warren

Chiricos

Bales

(2012). The imprisonment penalty for young Black and Hispanic males: A crime-specific analysis. Journal of Research in Crime & Delinquency, 49, 56–80.

37.

Wilson

D. B.

Mitchell

MacKenzie

D. L.

(2006). A systematic review of drug court effects on recidivism. Journal of Experimental Criminology, 2, 459–487.

38.

Wolfer

(2006). Graduates speak: A qualitative exploration of drug court graduates’ view of the strengths and weaknesses of the program. Contemporary Drug Problems, 33, 303–320.

39.

Zweig

J. M.

Lindquist

Downey

P. M.

Roman

Rossman

S. B.

(2012). Drug court policies and practices: How program implementation affects offender substance use and criminal behavior outcomes. Drug Court Review, 81(1), 43–79.