Extralegal Disparity in the Application of Intermediate Sanctions: An Analysis of U.S. District Courts

Abstract

The sentencing literature is replete with studies that have examined the influence of extralegal offender characteristics on two key sentence outcomes: the imprisonment and sentence length decisions. Yet the study of other outcomes, such as the application of intermediate sanctions, is rarely addressed. To date, no studies have been conducted in the federal courts to examine the potential influence of race/ethnicity, age, gender, and educational attainment on the decision to apply intermediate sanctions. Consequently, the present analysis employs U.S. Sentencing Commission data to examine direct and interactive effects of these extralegal characteristics on this understudied outcome. Findings indicate that extralegal effects may play an important role in the use of intermediate sanctions. The implications of this research are discussed in detail.

Keywords

sentencing minorities offenders

Introduction

In recent decades, the desire to eliminate unwarranted sentencing disparity has produced numerous reform efforts aimed to control judicial discretion (e.g., state and federal sentencing guidelines; Frase, 2000; Spohn, 2000; Zatz, 2000). This desire has also produced an extensive body of research aimed to clarify the nature and extent of such disparity, especially as it pertains to the influence of race, ethnicity, sex, and age (for recent reviews, see Baumer, 2013; Ulmer, 2012). Although this literature has produced an impressive knowledge base, perhaps one of its greatest shortcomings is the routine oversimplification of the sentencing decision itself. The majority of existing research simplifies the sentencing process into two decisions: (a) the imprisonment or “in/out” decision and (b) the sentence length decision. So commonplace is this practice that Baumer (2013) identified it as a key characteristic of the “modal approach” to the study of sentencing (p. 234). Researchers typically analyze one or both of these outcomes, neglecting the reality that sentencing options are more complex and varied. Gainey, Steen, and Engen (2005) recognized this shortcoming a decade ago, stating that “[a]lthough a huge literature has developed around the factors affecting these two sentencing decisions [in/out and sentence length], research has virtually ignored factors affecting the application of qualitatively different kinds of sentences like alternative or intermediate sanctions” (p. 489). This critique is no less relevant today. To be sure, only a handful of studies have examined unwarranted disparity in the use of intermediate sanctions (i.e., sentences that fall between probation and prison) at the state level and, to date, no studies have been conducted at the federal level. Thus, a considerable gap remains in our understanding of the sentencing process.

The need to investigate the use of intermediate sanctions is important for at least two additional reasons. First, many states and the federal government have been faced with the challenge of managing overcrowded prisons in the context of considerable budget constraints (Andrews & Bonta, 2010; Petersilia & Cullen, 2015). Intermediate sanctions are touted as less costly alternatives to incarcerating offenders, and this has sparked a renewed interest in expanding their use at both the state and federal levels (Frase, 2005; Johnson & Dipietro, 2012; U.S. Sentencing Commission [USSC], 2009). Given the current salience of intermediate sanctions in punishment discourse, it may be a timely opportunity to ask and answer questions about the fairness of their application and to identify where improvements may be needed. Currently, it is unknown whether (or to what extent) race, ethnicity, gender, and/or other extralegal factors influence the use of intermediate sanctions in the federal courts.

Second, researchers and policy makers have begun to recognize the collateral consequences of mass incarceration and the differential impact of imprisonment on African American communities. As Roberts (2004) put it, “African Americans experience a uniquely astronomical rate of imprisonment, and the social effects of imprisonment [the damaging of social networks, social norms, and social citizenship] are concentrated in their communities” (pp. 1272-1273). Moreover, the coercive mobility perspective (see Clear, Rose, Waring, & Scully, 2003; Rose & Clear, 1998) highlights the implications of such conditions. That is, concentrated incarceration forces residential turnover within communities, breaking down informal social controls and ultimately increasing crime rates. As intermediate sanctions offer a community-based alternative to incarceration, they have the potential to help alleviate this concern. Yet, if race plays a role in their use—that is, if minorities are less likely than nonminorities to receive intermediate sanctions—this potential may not be fully realized.

To help advance the research in this area, the current study draws on a large, representative sample of federal offenders who were sentenced within the 90 U.S. districts located in the United States. Specifically, this study examines the influence of key offender characteristics—race, ethnicity, age, sex, and educational attainment—on this understudied sentence outcome. In doing so, the present research advances the literature in three notable ways. First and foremost, it examines the application of intermediate sanctions in an important, but previously unstudied context—U.S. district courts. Second, this study examines the effect of educational attainment on the use of intermediate sanctions and thus provides a more complete assessment of key offender influences on this decision; prior research has been limited to the study of race, sex, and age, and the need to study additional offender characteristics (to help offset potential omitted variable bias) has been recently recognized in the literature (e.g., Johnson & Dipietro, 2012). Third, the current study extends prior research by providing a first look at the interactive effects of race/ethnicity, sex, and age in the context of intermediate sanctions. More specifically, studies of traditional sentencing outcomes indicate that young, Black and Hispanic, male offenders are often punished most severely (e.g., Spohn & Holleran, 2000; Steffensmeier, Ulmer, & Kramer, 1998), and this possibility is examined here in the context of intermediate sanctions.

Intermediate Sanctions and the Federal Courts

The use of intermediate sanctions—those whose severity lies somewhere between that of probation and prison—has been advocated by numerous scholars in the development of a more rational sentencing system. Several decades ago, Morris and Tonry (1991) succinctly described a key shortcoming with sentencing options. In their words, “[i]mprisonment is used excessively; probation is used even more excessively; between the two is a near vacuum of purposive and enforced punishments” (p. 3). The large gap in punishment severity found between prison and probation created a problematic scenario: Judges were, at times, forced to sentence offenders either too leniently (via probation) or too harshly (via prison). In other words, traditional probation and prison sentences were not always well suited to fit the offender and/or offense (Tonry & Lynch, 1996). This issue became a growing concern as the “just deserts” perspective, and its emphasis on proportional punishment, began to dominate punishment practices (Tonry & Lynch, 1996; Von Hirsch, 1976). Thus, to enhance the rationality of sentencing and to provide a smoother transition along the continuum of available punishments (and more recently to provide cost-effective alternatives to incarceration), all states have adopted at least some forms of community-based intermediate sanctions, such as intensive supervision probation, house arrest, or boot camps (Gainey et al., 2005; Tonry, 1998). Moreover, several states, such as Michigan, Pennsylvania, and Washington, have structured intermediate sanctions into their guidelines systems to normalize and regulate their use, at least to some extent (Frase, 2005; Tonry, 1998).

At the federal level, intermediate sanctions (often referred to as alternative sentences) are also built directly into the existing guidelines system. Federal sentencing depends primarily on an offender’s guideline-calculated offense level (ranging from 1 to 43) and criminal history category (ranging from 1 to 6). Once these two factors are calculated, the offender’s presumed sentence range is identified by a sentencing grid. The grid itself is divided into four specific zones (A through D), which further clarify the types of sanctions approved by the guidelines (see USSC, 2009). Offenders who fall into Zone A can be sentenced, at the discretion of the judge, to probation, an intermediate sanction, or a prison sentence up to 6 months in length. Moving up in severity, offenders in Zone B face the presumption of prison; however, judges maintain discretion to substitute (in part or in full) the specified prison time with a confinement-based intermediate sanction (e.g., home detention, community confinement). Offenders in Zone C also face the presumption of prison but are only required to serve half the minimum guideline sentence in prison. The remaining half can be substituted with a confinement-based intermediate sanction. Finally, in the absence of a departure, offenders in Zone D face the presumption of prison with no alternative sentences structured into the guideline grid.¹ Beyond these stipulations, intermediate sanctions are generally not applicable to offenders who face mandatory minimum prison sentences or to noncitizens who face potential deportation, despite the zone in which they fall (USSC, 2009).

The federal guidelines clearly define which cases are eligible for intermediate sanctions, but much like state guidelines systems (e.g., Pennsylvania, Washington), they do little to actually guide judges in deciding which offenders should receive them (or which specific factors should be considered). Engen, Gainey, Crutchfield, and Weis (2003) argued that when used in this fashion, intermediate sanctions open “windows of discretion” for court actors and may encourage the consideration of substantive concerns that are inconsistent with key purposes of sentencing guidelines (e.g., uniformity, elimination of unwarranted disparity). This point will be revisited more fully after a discussion of existing research that has examined disparities in the use of intermediate sanctions.

Existing Literature on Disparity and Intermediate Sanctions

Hundreds of empirical studies have been conducted to uncover the nature of sentencing disparities (particularly race-based disparities) in the context of the “in/out” and/or sentence length decisions (see Baumer, 2013), yet only a handful have examined disparity in the context of intermediate sanctions (e.g., Engen et al., 2003; Gainey et al., 2005; Johnson & Dipietro, 2012). Instead of focusing on disparity, intermediate sanctions research has concentrated largely not only on the issue of recidivism (Smith, Goggin, & Gendreau, 2002; Turner, Petersilia, & Deschenes, 1992; Ulmer, 2001) but also on perceptions of these sanctions by offenders, correctional staff, and even judges (e.g., May & Wood, 2005; Moore, May, & Wood, 2008; Wooldredge & Gordon, 1997). As a result, current knowledge of disparity linked to court actors’ use of intermediate sanctions is primarily limited to three studies conducted in two states: Pennsylvania and Washington.

The earliest of these studies was conducted by Engen et al. (2003) and examined the use of Washington’s structured sentence alternatives for felony offenders sentenced during the 1990-1992 fiscal years. In Washington, structured sentence alternatives (i.e., intermediate sanctions) were built into their guidelines system, allowing eligible offenders to receive sentences below the guidelines’ recommended range. Sentence alternatives consisted of partial confinement, community service, and/or treatment, and served as a replacement (or alternative) to standard incarceration, in part or in full. Findings from Engen et al.’s (2003) analysis indicated noteworthy disparities in the application of these sentence alternatives. Specifically, African American, Hispanic, male, and younger offenders were significantly less likely to receive sentence alternatives as compared with White, female, and older offenders, net of controls for several legal and case factors. Consequently, Engen’s et al.’s (2003) analysis revealed potentially concerning extralegal influences in what they refer to as discretionary decisions that are unlikely to face review or scrutiny by the courts.

In a second analysis of sentencing practice in Washington State, Gainey et al. (2005) analyzed the decision to employ intermediate sanctions for felony drug offenders sentenced during the 1996-1999 fiscal years. Their focus on drug offenders aimed to provide an examination of sentence alternatives where offender rehabilitation was likely to be of primary concern. The researchers presumed that judicial disagreement with standard guideline sentences (which were based largely on the principle of proportional punishment) would be more substantial in such cases, providing a unique opportunity to explore potential disparities arising from court actors’ discretionary decision making. Gainey et al.’s (2005) findings indicated that race, ethnicity, age, and sex each influenced the decision to administer intermediate sanctions, though their effects varied depending on the type of structured alternative sentence that was analyzed (i.e., the drug offender sentence alternative, work ethic camp, or first-time offender waiver). Overall, however, the results suggested a disadvantage (i.e., reduced likelihood of receiving intermediate sanctions in lieu of incarceration) for offenders who were male, older, and Hispanic. Black offenders, too, faced a disadvantage, but only in the context of the first-time offender waivers. In short, their analysis further confirmed earlier findings (i.e., Engen et al., 2003) that indicated the importance of extralegal factors in the use of intermediate sanctions.

Johnson and Dipietro (2012) provided the most recent analysis of disparity in the application of intermediate sanctions. Their study examined nearly all misdemeanor and felony offenders sentenced in Pennsylvania from 1998 to 2000. Like Washington, Pennsylvania also relied on sentencing guidelines that allowed for the use of intermediate sanctions (e.g., community-based drug and alcohol treatment, house arrest, electronic monitoring, intensive supervision, and boot camp). Whereas Engen et al. (2003) and Gainey et al. (2005) focused on the use of intermediate sanctions as a replacement for standard forms of incarceration, Johnson and Dipietro (2012) examined their use as a potential replacement for not only incarceration but also probation. To test for extralegal disparity in the use of intermediate sanctions, Johnson and Dipietro (2012) estimated multinomial hierarchical generalized linear models that included measures for typical legal/case factors, as well as characteristics of the judges and courts. Similar to the research conducted in Washington State, findings indicated significant extralegal influences. Compared with their counterparts, Black, Hispanic, male, and younger offenders were less likely to receive not only intermediate sanctions in lieu of prison and jail but also probation. These findings seemed to suggest that males, minorities, and younger offenders may be viewed as not only more dangerous and/or blameworthy (making them less suitable for intermediate sanctions relative to imprisonment) but also lacking the necessary offender agency involved in treatment-oriented sanctions (making them less suitable for intermediate sanctions relative to probation).

The application of intermediate sanctions at the federal level has yet to be studied in detail, but a recent report produced by the USSC provided a descriptive overview of their use during the 2007 fiscal year (USSC, 2009). Consistent with the structure of the federal sentencing guidelines (discussed earlier), the researchers reported that intermediate sanctions played the greatest role in the sentencing of eligible offenders who fell into Zones B and C of the guidelines grid, where approximately 50% of sentences involved an intermediate sanction. In Zone A, where standard probation was an option (unlike Zones B and C), the role of intermediate sanctions was less substantial; only 6.4% of offenders received them. Finally, in Zone D, a small portion of offenders also received intermediate sanctions (5.5%), but the majority received prison-only sentences as the guidelines recommended. The researchers also reported that intermediate sanctions were applied in greater proportion to females, Whites, older offenders, and offenders with higher levels of education, suggesting the possibility of important extralegal disparity in their use. The analysis, however, was limited to univariate descriptive statistics and thus the authors were unable to speak to the unique effects of offender characteristics.²

Sentencing Theory, Intermediate Sanctions, and Disparity

As mentioned previously, intermediate sanctions have the potential to open “windows of discretion” within the context of sentencing guidelines (Engen et al., 2003; Gainey et al., 2005). From a Weberian perspective (Weber, 1968), sentencing guidelines represent formal rational law, meaning that decisions are based on legally defined rules and regulations (see Savelsberg, 1992). The reliance on formal rational law, as embodied in guidelines, stands in stark contrast to the use of substantive law, which emphasizes nonlegal, individualized decision making—the kind of decision making that historically dominated sentencing prior to the advent of guideline-based systems (Savelsberg, 1992; Stith & Cabranes, 1998). Thus, guidelines represent a clear attempt to shift away from the inherent discretionary decision making of substantive law, toward nondiscretionary (or guided) decision making of formal rational law. Yet within guidelines systems, important pockets of discretion remain. As mentioned previously, the federal guidelines, along with well-established state guidelines (e.g., Pennsylvania, Washington), legally define which offenders are eligible for intermediate sanctions but do not legally define which of these offenders should receive them. Consequently, Johnson and Dipietro (2012) suggested that “intermediate punishments represent an important ‘gap’ in the ‘gapless system of rules’” (p. 818). When operating in the “gaps” of the system, how, then, do judges and court actors decide who should receive intermediate sanctions and who should not?

The answer to this question is at least partly rooted in Steffensmeier et al.’s (1998) focal concerns perspective of sentencing. This perspective posits that court actors make discretionary punishment decisions by assessing three focal concerns: (a) offender blameworthiness, (b) protection of the community (i.e., offender dangerousness), and (c) practical constraints and consequences of the sentence (e.g., offender’s ability to do prison time, amenability to treatment, family and employment status). Drawing on Albonetti’s (1991) earlier work, the focal concerns perspective suggests that court actors strive for a rational decision-making process, but in the face of organizational constraints (e.g., substantial caseloads and imperfect/incomplete information about offenders), decision making often becomes characterized by “bounded rationality.” That is, court actors are essentially forced to make critical sentencing decisions (e.g., whether or not to incarcerate) with too little information for judging the future risk of offenders. Steffensmeier et al. (1998) argued that these conditions promote the development of perceptual shorthands³ that can be readily used to aid sentencing decisions. Of particular importance, these shorthands are believed to be informed by not only crime seriousness and criminal history but also race, gender, and age attributions “relating these statuses to membership in social groups thought to be dangerous and crime prone” (Steffensmeier et al., 1998, p. 768). Moreover, Steffensmeier and Demuth (2000) further explained that this attributional decision-making process applies to not only concerns of dangerousness but also each of the focal concerns of sentencing (see also Johnson & Dipietro, 2012).

In short, this perspective suggests that discretionary punishment decisions will be influenced by a range of offense, offender, and practical concerns. Moreover, the use of “perceptual shorthands” may introduce unwarranted disparity in the application of punishment, and intermediate sanctions represent a “window of discretion” where this disparity may arise. For example, drawing on stereotypes about crime and criminality that routinely link race and dangerousness (e.g., Chiricos & Eschholz, 2002; Dixon & Maddox, 2005; Kennedy, 1997), numerous researchers have argued that minority, male, and younger offenders may face more punitive sanctions than their counterparts (e.g., Doerner & Demuth, 2010; Johnson & Betsinger, 2009; Spohn & Holleran, 2000). Steffensmeier and colleagues (1998) made this very argument in their research, claiming that “offenders who are young, male, and black are more likely to be forecast as dangerous or crime prone and viewed as less likely to be ‘harmed’ by incarceration” (p. 770). This argument was extended by Steffensmeier and Demuth (2000) who suggested that Hispanics, too, elicit similar attributions, due in part to being stereotyped as unintelligent, crime prone, and violent gang members (see also Spohn & Holleran, 2000). Thus, to the extent that intermediate sanctions are viewed as a less restrictive, more lenient form of punishment as compared with prison, it might be expected that certain offender groups will be deemed less fit for their use. This leads to the first hypothesis examined in the current study:

Hypothesis 1: Black, Hispanic, male, and younger offenders will be less likely to receive intermediate sanctions as compared with prison.

Intermediate sanctions can take numerous forms, but as discussed previously, they were designed, in part, to provide a middle ground between probation and prison (Morris & Tonry, 1991; Tonry, 1996). Thus, while intermediate sanctions are less restrictive than prison, they typically take on a more restrictive role when compared with standard forms of probation. This is particularly true in the context of federal sentencing; within the federal courts, nearly three quarters of offenders sentenced to intermediate sanctions receive home confinement for the purpose of restricting their mobility in the community (USSC, 2009). This suggests that when deciding between probation and intermediate sanctions, assessments within the federal courts will depend largely on perceptions of who requires greater control.⁴ This leads to the second study hypothesis:

Hypothesis 2: Black, Hispanic, male, and younger offenders will be more likely to receive intermediate sanctions as compared with probation.

Ulmer and Kramer’s (1996) work suggests that perceptual shorthands are likely influenced by additional offender statuses, including their educational attainment. Not only are offenders with greater educational attainment a lower risk to the community (Bellair & Kowalski, 2011; Harer, 1995), but they may also be perceived as having better employment prospects (Everett, Rogers, Hummer, & Krueger, 2011; Snyder & Dillow, 2011). This may be important to court actors because intermediate sanctions, such as home confinement, have the ability to preserve community ties, including those surrounding the offender’s work experience (Hofer & Meierhoefer, 1987; Tonry & Lynch, 1996). Thus, better educated offenders may be viewed as more suitable for intermediate sanctions as compared with prison. However, when compared with probation, the education/lower risk link may lead court actors to draw less on the greater restrictive capacity of intermediate sanctions for better educated offenders. This leads to the following two hypotheses:

Hypothesis 3: Offenders with greater levels of education will be more likely to receive intermediate sanctions as compared with prison.

Hypothesis 4: Offenders with greater levels of education will be less likely to receive intermediate sanctions as compared with probation.

Steffensmeier et al.’s (1998) specification of focal concerns theory also drew special attention to the interactive nature of specific offender characteristics. Drawing on routine perceptions of criminality, the theorists argued that the dangerous, crime-prone stereotype was not uniformly applied to Black offenders but instead targeted young Black males more directly. In addition, qualitative interviews conducted by these researchers indicated that “[j]udges also may see young black men as doing time more easily and interpret their demeanor as indicating less remorse and more recalcitrance than offenders of other race-age-gender categories” (p. 787).

Later researchers have explained that perceptions of criminality take on similar meanings for Hispanic offenders, where again, it is specifically young males who absorb the brunt of criminal stereotypes (Rios, 2006; Spohn & Holleran, 2000). This leads to the final two study hypotheses⁵:

Hypothesis 5: Among race, sex, and age combinations, young, Black and Hispanic, male offenders will be least likely to receive intermediate sanctions as compared with prison.

Hypothesis 6: Among race, sex, and age combinations, young, Black and Hispanic, male offenders will be most likely to receive intermediate sanctions as compared with probation.

Similar to Johnson and Dipietro (2012), we use the frameworks described above to guide our hypotheses for the present study but caution the reader that testing these hypotheses more thoroughly will require data richer than that available for the federal courts (and many state courts). Perhaps most importantly, the present study lacks information regarding defendants’ employment status, which may influence the use of intermediate sanctions. The current analysis, however, is able to partially address this issue by including a measure of educational attainment, which serves as a proxy for employment status (as alluded to above). This represents a small advance over existing peer-reviewed research, which has been unable to account for either employment status or educational attainment (e.g., Engen et al., 2003; Gainey et al., 2005; Johnson & Dipietro, 2012). Even so, other potentially important factors remain unaddressed, both by the prior literature as well as the current study (e.g., marital status, number of dependents,⁶ drug use history), and will need to be addressed by future research.

Method

Data and Sample

To test the proposed hypotheses, the current study relies on data from the USSC for fiscal years 2000 through 2012. These data provide detailed information on all cases that fall under the jurisdiction of the federal sentencing guidelines and are widely considered useful for the study of extralegal influences on sentencing (Doerner & Demuth, 2010; Franklin, 2013; Johnson & Betsinger, 2009). Consistent with prior research (Franklin, 2015; Johnson & Betsinger, 2009; Ulmer, Light, & Kramer, 2011a), the sample selected for analysis excludes offenders whose cases were heard within U.S. territories, focusing only on district courts located within the United States. Noncitizens (who face potential deportation) and offenders who received mandatory minimum prison sentences were also excluded as they generally do not qualify for intermediate sanctions (USSC, 2009).⁷ Given the theoretical focus of the current study, the analysis was limited to White, Black, and Hispanic offenders.⁸ This produced a sample of 411,781 cases of which 14,960 were excluded as a result of missing data.⁹ Thus, 396,821 offenders were included in the final sample. This sample was then divided into pre-Booker (n = 157,088) and post-Booker (n = 239,733) time periods for the purpose of analysis (this choice is discussed in the “Analytic Strategy” section below).

Dependent Measure

The current study examines the use of intermediate sanctions relative to both incarceration and probation. Consequently, the dependent variable is a multinomial measure that indicates whether the offender received a probation sentence, an intermediate sanction, or a prison sentence. As Johnson and Dipietro (2012) pointed out, intermediate sanctions are often packaged along with probation or prison/jail sentences, and can be challenging to operationalize from a research standpoint. This is relevant to federal sentencing where intermediate sanctions are administered in conjunction with probation or in conjunction with a prison sentence via supervised release.¹⁰ As a result, the outcome measure for the present analysis can be operationalized in different ways. Following the lead of Johnson and Dipietro (2012), the dependent measure was first operationalized according to the most serious sentence type, yielding the following three categories: (a) prison with or without intermediate sanctions, (b) intermediate sanctions with probation, and (c) probation only. Second, the measure was altered such that the intermediate sanctions category included any sentence that involved an alternative punishment. Our analyses indicated similar findings across these two specifications, as did Johnson and Dipietro’s (2012) analysis of sentencing in Pennsylvania. Thus, only models based on the second specification are reported.

Independent Measures

Consistent with the primary focus of the study, a series of extralegal offender characteristics are included in the analysis. Race/ethnicity was measured with a set of three dummy variables (0 = no, 1 = yes) identifying whether the offender was White, Black, or Hispanic. Age was measured in years and identified how old the offender was at the time of sentencing. Male identified the sex of the offender (0 = female, 1 = male). Offender’s educational attainment was measured with a set of three dummy variables (0 = no, 1 = yes) that identified the offender’s highest level of education as less than high school, high school graduate, or some college or higher.

Numerous legal and case-processing measures are also included in the analysis to help isolate the unique influence of the above offender characteristics. Consistent with prior federal sentencing research (e.g., Franklin, 2013; Johnson & Betsinger, 2009), offense seriousness and criminal history are captured by a measure of the presumed sentence, which indicates the minimum guideline-recommended sentence in months. As previously described, guideline-recommended sentences are not only primarily based on the offender’s criminal history score (ranging from 1 to 6) and offense severity score (ranging from 1 to 43) but also take into consideration guideline-specified adjustments to the recommended sentence (e.g., acceptance of responsibility).¹¹ A separate measure for the 6-point criminal history score was also included in the analysis as it has been demonstrated to influence sentence outcomes above and beyond its indirect effect that operates through the presumed sentence measure (Franklin, 2013; Johnson & Betsinger, 2009; Ulmer et al., 2011a). Offense type is measured by a set of seven dummy variables (0 = no, 1 = yes) that identify whether the most serious conviction was for a drug, violence, fraud, immigration, property, firearm, or other offense. Multiple counts is a dichotomous measure that indicates whether or not the offender was convicted of multiple criminal charges (coded 1) as opposed to a single criminal charge (coded 0). In line with past research (e.g., Ulmer et al., 2011a) sentence-departure status was measured through a set of four dummy variables (0 = no, 1 = yes) that identified if the offender was sentenced within range, or granted an upward departure,¹² downward departure, or substantial assistance departure from the recommended guideline sentence. Detained is a dichotomous measure that identifies whether or not the offender was in custody prior to sentencing (coded 1) or remained in the community (coded 0). Finally, trial is a dichotomous measure that identifies whether the case was adjudicated via a jury or bench trial (coded 1) as opposed to a plea agreement (coded 0).

Prior research has demonstrated variation in sentencing practices across the U.S. district courts included in the analysis (e.g., Kautt, 2002; Ulmer et al., 2011a). Thus, we control for a series of dummy variables that identify each of the 90 districts in the sample. This approach takes into consideration unobserved characteristics of the court communities (including variation in the type and availability of intermediate sanctions) and also addresses the nonindependence of offenders sentenced within the same district (see Johnson & Betsinger, 2009). This method has been adopted by numerous federal sentencing researchers (e.g., Doerner & Demuth, 2010; Franklin, 2013; Johnson & Betsinger, 2009). Finally, the current study also controls for the fiscal year of sentencing to address potential changes in sentencing practice over time. To conserve space, coefficients for the district and year dummy measures will not be presented in tabular form (results available upon request).

Analytic Strategy

The analysis proceeds in two main stages. First, using the main sample of offenders described above, multinomial logistic regression models are estimated to determine the direct effects of extralegal offender characteristics on the application of intermediate sanctions relative to prison and probation. In doing so, Hypotheses 1 through 4 are addressed. These models are then reestimated with the inclusion of race–sex–age dummies¹³ to address Hypotheses 5 and 6 regarding the potential interactive nature of key extralegal factors. The second stage of the analysis largely replicates the first, but in a targeted subsample of offenders where intermediate sanctions play a more routine role in the sentencing process. As discussed earlier, this occurs in Zones B and C of the guidelines grid, where intermediate sanctions were specifically designed to serve as alternatives to incarceration. In addition, this stage of the analysis focuses on offenders who were sentenced within the guidelines’ recommended range to further emphasize cases where intermediate sanctions are most routinized. Here, the aim is to determine whether the extent and nature of disparity in these “routine” cases are similar to that of the full sample. Doing so will better elucidate the generalizability of study findings. As probation is not an option for Zones B and C offenders, this portion of the analysis examines the use of intermediate sanctions relative to prison only and relies on logistic regression modeling.

Given the lack of research examining intermediate sanctions in the federal courts, the current study aims to provide both a current and past look at their use. In doing so, the analyses described above are carried out in data from two time periods. The decision reached in United States v. Booker (2005) is widely recognized as the most fundamental change to the federal sentencing guidelines since their inception, reducing them from mandatory to advisory in nature (see Hofer, 2007; Ulmer et al., 2011a, 2011b). Thus, a current assessment of intermediate sanctions was carried out by examining data collected post-Booker (January 2005 through the 2012 fiscal year), and a past assessment by examining data collected pre-Booker (fiscal year 2000 to January 2005). Organizing the data in this manner also allows for an examination of intermediate sanctions in two different contexts—one where judicial discretion is more tightly controlled, and one where it is less tightly controlled—offering a broader view of their use in the federal courts over time.

Results

Descriptive Statistics

Table 1 reports descriptive statistics for the pre- and post-Booker main analytic samples. A brief examination of the dependent measure indicates that in the pre-Booker sample, approximately 14% of offenders received an intermediate sanction, while a slightly smaller proportion received probation (12.05%), and the majority received prison (73.9%). This trend is largely replicated in the post-Booker sample, where 12.3% of offenders received intermediate sanctions, 10.7% received probation, and 77.0% received prison. An examination of offender characteristics reveals that White offenders made up the greatest proportion of the samples, followed by Black and Hispanic offenders. In addition, offenders are largely male, have an average age in their mid-30s, and possess educational experience that is distributed fairly evenly across the three measured levels of schooling. Finally, offenders faced average presumed sentences of 48.0 and 55.5 months (pre- and post-Booker, respectively), and have moderate criminal history scores (below 3 on a 6-point scale).

Table 1.

Descriptive Statistics for Pre- and Post-Booker Samples.

		Pre-Booker(n = 157,088)		Post-Booker(n = 239,733)
	M	SD	M	SD
Multinomial outcome measure
Prison	73.94	—	76.98	—
Intermediate sanction	14.02	—	12.31	—
Probation	12.05	—	10.70	—
Extralegal factors
White (ref.)	0.49	0.50	0.47	0.50
Black	0.33	0.47	0.34	0.47
Hispanic	0.18	0.38	0.19	0.39
Age	35.32	11.56	36.59	11.94
Male	0.81	0.40	0.81	0.40
Less than HS graduate (ref.)	0.33	0.47	0.31	0.46
HS graduate	0.37	0.48	0.39	0.49
Some college or greater	0.29	0.45	0.29	0.45
Legal and case-processing factors
Presumed sentence	48.04	62.39	55.49	66.60
Criminal history score	2.32	1.70	2.46	1.79
Multiple counts	0.19	0.39	0.20	0.40
Drug offense (ref.)	0.37	0.48	0.36	0.48
Violent offense	0.07	0.26	0.07	0.26
Fraud offense	0.27	0.44	0.25	0.43
Immigration offense	0.04	0.19	0.05	0.22
Property offense	0.06	0.24	0.04	0.19
Firearm offense	0.14	0.34	0.17	0.38
Other offense	0.06	0.23	0.07	0.25
Within range (ref.)	0.63	0.48	0.50	0.50
Upward depart.	0.01	0.09	0.02	0.13
Substantial assist. depart.	0.26	0.44	0.27	0.45
Other downward depart.	0.11	0.31	0.21	0.41
Detained	0.49	0.50	0.56	0.50
Trial	0.03	0.18	0.03	0.18

Note. ref. = reference; HS = high school; assist. = assistance; depart. = departure.

Direct Effects—Full Analytic Samples

Turning attention to the main multivariate analyses, Table 2 reports the direct effects of extralegal factors on the decision to use intermediate sanctions relative to prison and probation. Beginning with the pre-Booker analysis (Model 1), it can be seen that, relative to White offenders, Hispanic offenders are somewhat less likely to receive intermediate sanctions as compared with prison. Black offenders, however, are treated no differently than White offenders. When examining the use of intermediate sanctions relative to probation, the odds of receiving an intermediate sanction are approximately 20% greater for Hispanic offenders relative to White offenders. Black offenders, however, are somewhat less likely to receive intermediate sanctions. Beyond race/ethnicity, male offenders are considerably less likely than female offenders to receive intermediate sanctions relative to prison, but equally likely to receive them relative to probation. Regarding age, younger offenders are less likely to receive intermediate sanctions relative to prison, though age has no bearing on the decision to use intermediate sanctions relative to probation. Taken together, these findings offer inconsistent/minimal support for the first two study hypotheses, which predicted that minorities, males, and younger offenders would be less likely to receive intermediate sanctions relative to prison, but more likely to receive intermediate sanctions relative to probation.

Table 2.

Multinomial Logistic Regression Estimates of Direct Effects on the Intermediate Sanction Decision.

	Model 1: Pre-Booker(n = 157,088)						Model 2: Post-Booker(n = 239,733)
	Intermediate sanction vs. prison			Intermediate sanction vs. probation			Intermediate sanction vs. prison			Intermediate sanction vs. probation
	b	SE	Exp(b)	b	SE	Exp(b)	b	SE	Exp(b)	b	SE	Exp(b)
Extralegal factors
White (ref.)	—	—	—	—	—	—	—	—	—	—	—	—
Black	−0.028	.023	0.972	−.071	.027	0.931**	−0.073	.019	0.930**	.022	.023	1.022
Hispanic	−0.063	.030	0.939*	.174	.037	1.190**	−0.111	.024	0.895**	.260	.031	1.297**
Age	0.007	.001	1.007**	.001	.001	1.001	0.003	.001	1.003**	−.003	.001	0.997**
Male	−0.402	.021	0.669**	.016	.023	1.016	−0.403	.017	0.668**	.049	.020	1.050*
Less than HS graduate (ref.)	—	—		—	—		—	—		—	—
HS graduate	0.094	.023	1.099**	−.121	.029	0.886**	0.061	.019	1.063**	−.092	.025	0.912**
Some college or greater	0.104	.024	1.110**	−.072	.029	0.931*	0.091	.020	1.095**	−.052	.026	0.949**
Legal and case-processing factors
Presumed sentence	−0.063	.001	0.939**	.074	.001	1.077**	−0.051	.001	0.950**	.062	.001	1.064**
Criminal history score	−0.262	.009	0.770**	.351	.015	1.420**	−0.221	.007	0.802**	.305	.012	1.357**
Multiple counts	−0.360	.026	0.698**	.189	.034	1.208**	−0.203	.021	0.816**	.221	.028	1.247**
Drug offense (ref.)	—	—		—	—		—	—		—	—
Violent offense	−0.438	.056	0.645**	−.115	.078	0.891	−0.125	.044	0.882**	−.131	.066	0.877*
Fraud offense	0.354	.028	1.425**	.440	.034	1.553**	0.194	.022	1.214**	.120	.029	1.127**
Immigration offense	0.026	.049	1.026	.640	.065	1.896**	0.213	.034	1.237**	.564	.049	1.758**
Property offense	0.770	.039	2.160**	.130	.042	1.139**	0.703	.037	2.020**	−.156	.039	0.856**
Firearm offense	0.079	.038	1.082*	.179	.051	1.196**	0.058	.029	1.060*	.029	.040	1.029
Other offense	0.578	.040	1.782**	−.018	.044	0.982	0.446	.031	1.562**	−.270	.035	0.763**
Within range (ref.)	—	—		—	—		—	—		—	—
Substantial assist. depart.	1.416	.026	4.121**	−.899	.030	0.407**	1.324	.021	3.758**	−.513	.027	0.599**
Other downward depart.	1.435	.028	4.200**	−.135	.034	0.874**	1.205	.019	3.337**	−.061	.024	0.941*
Detained	−1.122	.023	0.326**	.186	.032	1.204**	−0.982	.018	0.375**	.247	.025	1.280**
Trial	−0.569	.064	0.566**	.167	.095	1.182	−0.534	.051	0.586**	.296	.076	1.344**

Note. These models also include controls for district and year, but their coefficients are not presented due to space constraints. ref. = reference; HS = high school; assist. = assistance; depart. = departure.

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

The findings regarding educational attainment, however, offer consistent support for the third and fourth hypotheses. Together, these hypotheses predicted that greater educational attainment would increase the odds of receiving an intermediate sanction relative to prison but decrease the odds relative to probation. This can be seen in Model 1, where the odds of receiving an intermediate sanction in lieu of prison increased by approximately 10% for high school graduates and 11% for those with some college or greater as compared with those with less than high school education. Furthermore, the odds of receiving an intermediate sanction in lieu of probation were reduced by approximately 11% for high school graduates and 7% for those with some college or greater.

Turning to the post-Booker analysis (Model 2) reveals a pattern of extralegal effects that provides stronger overall support for the first four study hypotheses. Consistent with the first hypothesis, Black and Hispanic offenders are less likely than their White counterparts to receive intermediate sanctions relative to prison (by 7.0% and 10.5%, respectively). Moreover, this same pattern applies to younger and male offenders who are also less likely to receive these alternative sentences as compared with older and female offenders. When the point of comparison is shifted to probation, Hispanic offenders (though not Black offenders) are significantly more likely (29.7%) than White offenders to receive these intermediate sanctions, providing support for the second hypothesis. This hypothesis is further supported by the finding that younger and male offenders are also more likely to receive intermediate sanctions. Finally, the third and fourth hypotheses regarding the effect of educational attainment are also supported. Consistent with Hypothesis 3, it can be seen that the odds of receiving an intermediate sanction in lieu of prison are greater for high school graduates (by 6.3%) and offenders with some college or greater (by 9.5%), as compared with offenders who did not complete high school. When comparing the use of intermediate sanctions with probation, these effects are reversed as predicted by Hypothesis 4. That is, offenders who completed high school and offenders with some college experience or greater were less likely to receive intermediate sanctions (by 8.8% and 5.1%, respectively).

Numerous legal and case-processing control measures were also included in the analyses, and their effects are largely consistent with expectations¹⁴ and similar across time periods. For example, offenders with higher presumed sentences, higher criminal history scores, and multiple counts of conviction were less likely to receive intermediate sanctions relative to prison, but more likely to receive them relative to probation. Likewise, offenders who were detained prior to sentencing and who opted for a trial were less likely to receive intermediates in lieu of prison, but more likely to receive them in lieu of probation. However, offenders who were granted downward departures (of any type) were considerably more likely to receive intermediate sanctions relative to prison, and less likely to receive them relative to probation. The offense type measures, however, were less consistent across time periods and, in some instances, their effects were similar when comparing intermediate sanctions with both prison and probation. The cause of these findings is unclear, but some offense types may be considered inappropriate for certain intermediate sanctions.

Interaction Effects—Full Analytic Samples

Table 3 reports the findings from the analyses that examined race–sex–age interactions in the main pre- and post-Booker samples. Only coefficients for the race–sex–age dummy measures are reported, but the models were estimated with the same set of covariates as presented earlier and their effects were virtually unchanged. As a reminder, the fifth and sixth study hypotheses predicted that among race–sex–age categories, young Black and Hispanic males would receive the most punitive/restrictive treatment, meaning these two groups would be the least likely to receive intermediate sanctions relative to prison, and the most likely to receive them relative to probation. Beginning with the pre-Booker time period, it can be seen that these hypotheses received partial support. Compared with young White male offenders, those who were young, Black, and male were the least likely to receive intermediate sanctions, relative to prison (odds = .826). Young Hispanic males are the next least likely group to receive intermediate sanctions (odds = .914), but the effect is only marginally significant (p = .071). However, when comparing the use of intermediate sanctions with probation, young Black males are treated similarly to young White males, and it is young Hispanic males who receive the most punitive/restrictive treatment. Their odds of receiving an intermediate sanction, relative to probation, are approximately 1.5 times greater than young White males.

Table 3.

Multinomial Logistic Regression Estimates of Race/Ethnicity, Sex, and Age Interactions.

Race/ethnicity–sex–age interaction dummies	Model 1: Pre-Booker(n = 157,088)						Model 2: Post-Booker(n = 239,733)
	Intermediate sanction vs. prison			Intermediate sanction vs. probation			Intermediate sanction vs. prison			Intermediate sanction vs. probation
	b	SE	Exp(b)	b	SE	Exp(b)	b	SE	Exp(b)	b	SE	Exp(b)
Young White male (ref.)	—	—	—	—	—	—	—	—	—	—	—	—
Young Black male	−.191	.045	0.826**	−.017	.056	0.983	−.220	.039	0.803**	.115	.053	1.122*
Young Hispanic male	−.090	.050	0.914	.424	.067	1.528**	−.118	.041	0.889**	.414	.058	1.513**
Older White male	.132	.033	1.141**	.090	.039	1.094*	.026	.029	1.026	−.053	.037	0.948
Older Black male	.080	.040	1.083*	.017	.050	1.017	−.102	.034	0.903**	.023	.044	1.023
Older Hispanic male	.033	.049	1.034	.214	.062	1.239**	−.209	.040	0.811**	.096	.053	1.101
Young White female	.406	.053	1.501**	.061	.058	1.063	.326	.045	1.385**	−.064	.053	0.938
Young Black female	.388	.054	1.474**	−.063	.057	0.939	.410	.051	1.507**	−.061	.057	0.941
Young Hispanic female	.411	.066	1.508**	.197	.077	1.218*	.393	.051	1.481**	.349	.066	1.418**
Older White female	.388	.041	1.474**	.107	.046	1.113*	.278	.035	1.320**	−.052	.042	0.949
Older Black female	.560	.048	1.751**	.024	.053	1.024	.403	.041	1.496**	−.087	.047	0.917
Older Hispanic female	.365	.065	1.441**	.082	.075	1.085	.262	.048	1.300**	.186	.061	1.204**

Note. These models include the full spectrum of measures from the initial models. Only coefficients for the relevant interaction dummy variables are reported as the remaining measures did not substantively differ in their effects presented in Table 2. ref. = reference.

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

The fifth and sixth hypotheses receive similar partial support in the post-Booker sample where, relative to prison, young Black males (but not young Hispanic males) are again the least likely to receive an intermediate sanction. Likewise, it is young Hispanic males (but not young Black males) who are the most likely to receive an intermediate sanction, relative to probation. Though not directly tied to the proposed hypotheses, the interaction analyses also reveal that certain categories of female offenders face increased odds of receiving intermediate sanctions not only relative to prison (as might be expected) but also relative to probation (which may be less expected). We return to this finding at a later point.

Direct Effects—Zones B and C Subsamples

Table 4 reports the findings from the analyses restricted to cases from Zones B and C of the sentencing guidelines where intermediate sanctions play a routine role as imprisonment alternatives. Thus, the focus here is only on the use of intermediate sanctions relative to prison. In the pre-Booker analysis, it can be seen that race/ethnicity does not exhibit direct effects on the decision to use intermediate sanctions as an alternative to prison. Other extralegal factors do, however, play a role in this decision, providing mixed support for the first hypothesis and consistent support for the third hypothesis. Specifically, males and younger offenders are less likely than their counterparts to receive intermediate sanctions (supporting Hypothesis 1), while offenders with greater educational attainment are more likely to receive these sentences (supporting Hypothesis 3). These findings are replicated in the post-Booker sample with two exceptions. First, age does not play a role in the use of intermediate sanctions as alternatives to prison. Second, the effect of race is significant, indicating that Black offenders have 13.5% lower odds of receiving intermediate sanctions as compared with White offenders.

Table 4.

Logistic Regression Estimates of Direct Effects: Zones B and C Subsamples.

	Model 1: Pre-Booker(n = 17,894)			Model 2: Post-Booker(n = 19,004)
	Intermediate sanction vs. prison			Intermediate sanction vs. prison
	b	SE	Exp(b)	b	SE	Exp(b)
Extralegal factors
White (ref.)	—	—	—	—	—	—
Black	−0.082	.047	0.921	−0.145	.045	0.865**
Hispanic	−0.062	.066	0.940	−0.064	.060	0.938
Age	0.008	.002	1.008**	0.002	.039	1.002
Male	−0.498	.043	0.608**	−0.443	.039	0.642**
Less than high school (ref.)	—	—	—	—	—	—
High school graduate	0.190	.047	1.209**	0.090	.043	1.094*
Some college or greater	0.197	.050	1.217**	0.108	.048	1.114*
Legal and case-processing factors
Presumed sentence	−0.099	.007	0.905**	−0.110	.007	0.896**
Criminal history score	−0.571	.018	0.565**	−0.533	.017	0.587**
Multiple counts	−0.301	.056	0.740**	−0.106	.054	0.899*
Drug offense (ref.)	—	—	—	—	—	—
Violent offense	−0.480	.143	0.619**	−0.435	.138	0.647**
Fraud offense	0.078	.065	1.082	−0.019	.059	0.981
Immigration offense	−0.570	.097	0.565**	−0.426	.074	0.653**
Property offense	0.159	.078	1.173*	0.162	.078	1.176*
Firearm offense	0.056	.097	1.058	0.017	.081	1.017
Other offense	−0.051	.081	0.951	−0.095	.070	0.909
Detained	−1.504	.043	0.222**	−0.809	.127	0.445**
Trial	−0.668	.124	0.513**	−1.392	.038	0.249**

Note. These models also include controls for district and year, but their coefficients are not presented due to space constraints. ref. = reference.

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

Interaction Effects—Zones B and C Subsamples

Race–sex–age interactions for the restricted subsamples are reported in Table 5. Consistent with the fifth study hypothesis, young Black male offenders are the least likely of all race–sex–age groupings to receive intermediate sanctions as an alternative to imprisonment. More specifically, the odds of receiving an intermediate sanction are 26.7% and 28.0% lower (pre- and post-Booker, respectively) for young Black males as compared with young White males. Inconsistent with the fifth hypothesis, however, young Hispanic males were treated no differently than young White males during either the pre- or post-Booker periods. Overall, the direct and interactive influence of extralegal factors in the targeted subsamples indicate the presence of important disparities in the routine application of intermediate sanctions, and these disparities appear to be largely similar to that reported in the full sample.

Table 5.

Logistic Regression Estimates of Race/Ethnicity, Sex, and Age Interactions: Zones B and C Subsamples.

Race/ethnicity–sex–age interactions	Model 1: Pre-Booker(n = 17,894)			Model 2: Post-Booker(n = 19,004)
	Intermediate sanction vs. prison			Intermediate sanction vs. prison
	b	SE	Exp (b)	B	SE	Exp (b)
Young White male (ref.)	—	—	—	—	—	—
Young Black male	−.310	.088	0.733**	−.329	.090	0.720**
Young Hispanic male	−.104	.103	0.901	.045	.095	1.046
Older White male	.106	.069	1.112	.049	.073	1.050
Older Black male	−.004	.081	0.996	−.128	.081	0.880
Older Hispanic male	−.008	.102	0.992	−.258	.095	0.773**
Young White female	.364	.114	1.439**	.312	.116	1.366**
Young Black female	.444	.086	1.560**	.364	.121	1.439**
Young Hispanic female	.352	.141	1.422*	.647	.117	1.910**
Older White female	.470	.086	1.599**	.325	.087	1.384**
Older Black female	.607	.101	1.835**	.368	.098	1.445**
Older Hispanic female	.473	.147	1.605**	.258	.113	1.294*

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

Discussion

The present study set out with a series of primary aims: (a) to expand the sentencing literature by examining extralegal disparity in the context of a rarely studied sentencing outcome—the application of intermediate sanctions; (b) to do so in the context of U.S. district courts, which represent a new and important jurisdiction of study; (c) to incorporate previously unexamined predictors of the intermediate sanction decision, particularly offenders’ educational attainment; (d) to consider the interactive influence of key extralegal factors on the decision to apply intermediate sanctions; and (e) to paint a broad image of their use in the federal courts by investigating multiple time periods, representing not only current but also past practice. These aims were addressed by analyzing USSC data spanning the years 2000 through 2012, and the findings have several implications as discussed below.

First, in support of the focal concerns perspective, the pattern of findings offers partial or full support for each of the six study hypotheses (summarized in Table 6), suggesting that extralegal factors often play a measurable role in the application of intermediate sanctions. In all, 32 of the 48 tested extralegal effects favored the study hypotheses. In a general sense, these findings are consistent with state-level research conducted in Pennsylvania (Johnson & Dipietro, 2012) and Washington (Engen et al., 2003; Gainey et al., 2005), where extralegal factors were also found to influence the use of intermediate sanctions. Not only at the federal level, but also in many state-level systems, intermediate sanctions are structured in a way that potentially invites the consideration of individual offender characteristics (encouraging substantive rationality). This invitation stems from the absence of clearly established factors to guide the use of intermediate sanctions, thereby leaving room for considerable judicial discretion (Engen et al., 2003; Frase, 2000, 2005). In light of the present findings, the USSC may consider further developing the guidelines to systematically outline the process by which various intermediate sanctions should be meted out in relation to punishments such as prison or probation (encouraging formal rationality). This certainly poses considerable challenges, such as standardizing the purposes/goals of intermediate sanctions, managing resource availability, and identifying relevant offender characteristics (see Frase, 2005), but the potential for reduced disparity may be desirable.

Table 6.

Summary of Support for Proposed Hypotheses.

Proposed hypotheses	Full sample		Zones B and C subsamples
Proposed hypotheses	Pre-Booker	Post-Booker	Pre-Booker	Post-Booker
1. Black, Hispanic, male, and younger offenders will be less likely to receive intermediate sanctions relative to prison.	−/+	+	−/+	−/+
a. Black offenders	−	+	−	+
b. Hispanic offenders	+	+	−	−
c. Male offenders	+	+	+	+
d. Younger offenders	+	+	+	−
2. Black, Hispanic, male, and younger offenders will be more likely to receive intermediate sanctions relative to probation.	−/+	−/+	—	—
a. Black offenders	−	−	—	—
b. Hispanic offenders	+	+	—	—
c. Male offenders	−	+	—	—
d. Younger offenders	−	+	—	—
3. Offenders with greater levels of education will be more likely to receive intermediate sanctions relative to prison.	+	+	+	+
a. High school graduate	+	+	+	+
b. Some college or greater	+	+	+	+
4. Offenders with greater levels of education will be less likely to receive intermediate sanctions relative to probation.	+	+	—	—
a. High school graduate	+	+	—	—
b. Some college or greater	+	+	—	—
5. Young, Black and Hispanic, male offenders will be least likely to receive intermediate sanctions relative to prison	−/+	−/+	−/+	−/+
a. Young Black male offenders	+	+	+	+
b. Young Hispanic male offenders	−	−	−	−
6. Young, Black and Hispanic, male offenders will be most likely to receive intermediate sanctions relative to probation	−/+	−/+	—	—
a. Young Black male offenders	−	−	—	—
b. Young Hispanic male offenders	+	+	—	—

Note. “+” = supported; “−” = not supported; “−/+” = partially supported; “—” = not applicable.

Second, the current study provides a first look at the role of offenders’ educational attainment in the decision to use intermediate sanctions. Overall, the findings demonstrate that offenders with higher levels of education (i.e., high school graduate, some college or greater) are more likely to receive intermediate sanctions in lieu of prison. When compared with the less restrictive probation sentence, offenders with higher levels of education are less likely to receive intermediate sanctions. This pattern is consistent with focal concerns theory and the notion that court actors may view better educated offenders as lower risk (i.e., less dangerous) to the community (Bellair & Kowalski, 2011; Harer, 1995), providing them with less restrictive sentences where possible (i.e., intermediate sanctions instead of prison, and probation instead of intermediate sanctions). But perhaps more importantly, the finding that better educated offenders are favored for intermediate sanctions over prison is also consistent with the notion that court actors may seek to protect offender/community bonds (a practical constraint/consequence of sentencing), especially those fostered by greater educational attainment (i.e., stable/quality employment).

Third, much like prior sentencing research (e.g., Doerner & Demuth, 2010; Johnson & Betsinger, 2009; Steffensmeier et al., 1998), we find that race, sex, and age interact to influence punishment outcomes, though not always in expected ways. As predicted by the focal concerns framework, young Black males were the least likely to receive intermediate sanctions relative to prison. Young Hispanic males, however, were treated more similarly to young White males, despite the fact that both offender groups are arguably stereotyped as crime prone and dangerous (Rios, 2006; Spohn & Holleran, 2000). Adding further complication, when contrasting intermediate sanctions with probation, it is not young Black males, but rather young Hispanic males, who face the greatest risk of intermediate sanctions. Consequently, the interaction of race, sex, and age appears to be important in the context of intermediate sanctions but cannot be easily explained through stereotyping alone. Other interesting patterns of interaction also emerged, such as the channeling of Hispanic females away from prison and probation and into intermediate sanctions. Again, patterns such as these demand an explanation that extends beyond perceptions of dangerousness/risk, and likely include court actors’ considerations of practical constraints and consequences. Perhaps the available treatment/service-oriented intermediate sanctions are viewed as more appropriate or useful for offenders with greater familial responsibility. In the sample analyzed here, this is likely to be Hispanic females, who are primarily of Mexican origin and whose fertility rates stand out as being among the highest in the nation (see Cherlin, 2010). Unfortunately, making further sense of such patterns will require data richer than that collected by the USSC (marital status and number of dependents have not been collected by the USSC since 2003).

Fourth, the current study found that patterns of extralegal disparity were more similar than different across the multiple contexts examined (see Table 6). For example, the pattern of disparity that emerged in the sample of Zones B and C offenders (where intermediate sanctions are a routinized part of sentencing) largely translated to that found in the full sample (which includes Zones A and D offenders where intermediate sanctions are much less routinized). This suggests that the nature of disparity is fairly generalizable and extends to the full spectrum of federal offenders for whom intermediate sanctions might be applied. When the emphasis is shifted to patterns of disparity across pre- and post-Booker time periods, similar conclusions can be drawn. That is, most of the extralegal effects operate similarly across time periods, though a few notable exceptions do occur. For example, there is no evidence during the pre-Booker period that race directly influences the use of intermediate sanctions, relative to prison; in contrast, during the post-Booker period, a small but statistically significant direct race effect does appear (particularly in the full sample). In addition, when examining the use of intermediate sanctions relative to probation, direct age and sex effects do not occur during the pre-Booker period, yet these measures do exhibit small but statistically significant effects during the post-Booker period. Thus, patterns of disparity are not identical across time periods, but they do share considerable similarity and seem to suggest that current sentencing practice, at least as it pertains to extralegal disparity in the use of intermediate sanctions, has not undergone substantial transformation. This interpretation is consistent with recent assessments of Booker and its effect on disparity when examining the traditional “in/out” and/or sentence length decisions (Starr & Rehavi, 2013; Ulmer et al., 2011a, 2011b).

Fifth, prior research has suggested that individual offender agency may play a central role in the decision to apply intermediate sanctions as they often take a strong rehabilitative focus (e.g., Gainey et al., 2005; Johnson & Dipietro, 2012). In short, it has been argued that certain offenders, such as males and minorities, may be perceived as less amenable to treatment and possibly less deserving of rehabilitative efforts. Johnson and Dipietro (2012) found support for this claim in their Pennsylvania-based analysis, noting that male and Black offenders were less likely to receive intermediate sanctions not only relative to prison and jail but also relative to probation. This pattern, however, does not appear to translate to the federal courts. Instead, it appears that males and minorities (e.g., Hispanics) are more likely to receive intermediate sanctions than probation (particularly during the post-Booker time period). While this may appear to be a contradictory finding, it was not unexpected in the context of the federal courts. As mentioned previously, intermediate sanctions are dominated by the use of home confinement in the federal system (USSC, 2009), and as such the role of offender agency is minimized. Instead, concerns appear to remain focused on perceptions of dangerousness, leading federal court actors to favor intermediate sanctions over probation for not only males and minorities but also younger offenders and those who fail to complete high school.

Sixth, young Black males were the least likely group among all race–sex–age categories to receive intermediate sanctions relative to prison. This finding is potentially troublesome. As discussed earlier, the collateral consequences of incarceration are especially consequential in African American communities, where significant numbers of young Black men are imprisoned (see Western, 2006). Not only does incarceration cause economic and social hardships within these communities, but it may also have serious negative effects on the children of offenders, including antisocial/aggressive behavior (e.g., Braman, 2004; Murray, Farrington, & Sekol, 2012). Such effects are concentrated among African American children who are estimated to be approximately 6.8 times more likely to experience parental incarceration than White children (see Foster & Hagan, 2009; Wildeman, 2009). Along these lines, researchers working within the coercive mobility framework have pointed out that concentrated incarceration forces residential turnover within communities, ultimately breaking down informal social controls and increasing (rather than decreasing) crime rates (see Clear et al., 2003; Rose & Clear, 1998). Community-based intermediate sanctions offer alternatives to incarceration, and thus represent a mechanism by which policy makers could attenuate some of these collateral consequences of incarceration. Yet, the present evidence suggests that the offenders in greatest need of alternatives to incarceration are, in fact, the least likely to receive them. At a minimum, court actors should be aware of this potentially damaging phenomenon, and this could be facilitated by the USSC.

Finally, the present study is not without limitations. Of primary concern, the analyses presented here (like nearly all studies of sentencing disparity) suffer from potential omitted variable bias. As mentioned previously, a number of additional measures would have been useful for understanding the application of intermediate sanctions, some of which might call into question the reported findings. For example, the young–Black–male disadvantage discussed above could be partially (or even fully) accounted for by unmeasured factors such as employment status; this measure has yet to be examined in the context of intermediate sanctions. We do think the inclusion of educational attainment as a proxy for employment status may help offset this concern, but it does not eliminate it. However, there is also reason to believe that this reported racial disadvantage is underestimated. Starr and Rehavi’s (2013) recent analysis of federal sentencing demonstrated evidence that Black males are disadvantaged prior to the final sentencing stage. In fact, their analysis suggested that Black males face greater charge severity as compared with similarly situated White males. As the current study is limited to the final sentence stage, such disparities that arise earlier in the process are unaccounted for, and this may have produced downward bias in the size of reported disparities. Starr and Rehavi (2013) also criticized federal sentencing studies that control for factors that filter out potential disparities arising from judicial fact-finding, including the judicial decision to depart from the guidelines. In a reanalysis of our main findings without the inclusion of departure measures, patterns of racial disadvantage remain similar, but do become more pronounced (results available upon request). In light of these concerns, we find little reason to suspect that meaningful disparities do not exist in the application of intermediate sanctions, even if the current data fall short of more precise estimations.

It should also be noted that the present analysis does not fully consider the diversity of intermediate sanctions employed by the federal courts. Similar to prior research (e.g., Johnson & Dipietro, 2012), all types of intermediate sanctions were examined together. This may be less problematic in the federal courts due to the heavy reliance on a single intermediate sanction (i.e., home confinement), but it would still be theoretically useful to, at minimum, examine confinement- and treatment-oriented sanctions separately. In addition, the current data are useful for examining patterns of disparity, yet they still require much speculation about the cause of such patterns. The interpretation of findings can become difficult, particularly when they do not line up with existing theory (as noted above). Thus, future researchers should conduct qualitative interviews with district court judges to further illuminate how they decide which offenders are best suited for intermediate sanctions. Such an approach could address numerous issues that are not captured in the USSC data, including the availability of resources, how federal courts communicate with these resources, the role of probation officers in identifying suitable offenders for intermediate sanctions, and the influence of court cultures on the normalization and use of alternative sentences. Doing so will go a long way to advance what we know about the use of intermediate sanctions and the nature of disparity.

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.

Notes

Author Biographies

Travis W. Franklin earned his PhD in criminal justice from Washington State University and is currently an associate professor in the Department of Criminal Justice and Criminology at Sam Houston State University. His current research focuses on understanding the nature of disparity in the prosecution and sentencing of offenders in state and federal courts. His recent work has appeared in Justice Quarterly, Crime & Delinquency, Criminal Justice and Behavior, and Journal of Criminal Justice.

Layne Dittmann is a PhD student in the Department of Criminal Justice and Criminology at Sam Houston State University. His current research focuses on private prisons, particularly ex-inmates’ perceptions and experiences of the quality of confinement in public and private prisons. He also focuses on disparity in sentencing, gangs in the media, media portrayals of private prisons, and inmate perceptions of family support and community support after release.

Tri Keah S. Henry is a PhD student in the Department of Criminal Justice and Criminology at Sam Houston State University. Her current research interests include disparities in the criminal justice system and procedural justice in policing.

References

Albonetti

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

Andrews

D. A.

Bonta

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

Baumer

E. P.

(2013). Reassessing and redirecting research on race and sentencing. Justice Quarterly, 30, 231-261.

Bellair

P. E.

Kowalski

B. R.

(2011). Low-skill employment opportunity and African American-White difference in recidivism. Journal of Research in Crime & Delinquency, 48, 176-208.

Braman

(2004). Doing time on the outside: Incarceration and family life in Urban America. Ann Arbor: University of Michigan Press.

Cherlin

A. J.

(2010). Demographic trends in the United States: A review of research in the 2000s. Journal of Marriage and Family, 72, 403-419.

Chiricos

Eschholz

(2002). The racial and ethnic typification of crime and the criminal typification of race and ethnicity in local television news. Journal of Research in Crime & Delinquency, 39, 400-420.

Clear

T. R.

Rose

D. R.

Waring

Scully

(2003). Coercive mobility and crime: A preliminary examination of concentrated incarceration and social disorganization. Justice Quarterly, 20, 33-64.

Dixon

T. L.

Maddox

K. B.

(2005). Skin tone, crime news, and social reality judgments: Priming the stereotype of the dark and dangerous black criminal. Journal of Applied Social Psychology, 35, 1555-1570.

10.

Doerner

Demuth

(2010). The independent and joint effects if race/ethnicity, gender, and age on sentencing outcomes in U.S. federal courts. Justice Quarterly, 27, 1-27.

11.

Doherty

(2013). Indeterminate sentencing returns: The invention of supervised release. New York University Law Review, 88, 958.

12.

Engen

R. L.

Gainey

R. R.

Crutchfield

R. D.

Weis

J. G.

(2003). Discretion and disparity under sentencing guidelines: The role of departures and structured sentencing alternatives. Criminology, 41, 99-130.

13.

Everett

B. G.

Rogers

R. G.

Hummer

R. A.

Krueger

P. M.

(2011). Trends in educational attainment by race/ethnicity, nativity, and sex in the United States, 1989–2005. Ethnic and Racial Studies, 34, 1543-1566.

14.

Foster

Hagan

(2009). The mass incarceration of parents in America: Issues of race/ethnicity, collateral damage to children, and prisoner reentry. The ANNALS of the American Academy of Political and Social Science, 623, 179-194.

15.

Franklin

T. W.

(2013). Sentencing native Americans in US federal courts: An examination of disparity. Justice Quarterly, 30, 310-339.

16.

Franklin

T. W.

(2015). Race and ethnicity effects in federal sentencing: A propensity score analysis. Justice Quarterly, 32, 653-679.

17.

Franklin

T. W.

(2017). Sentencing outcomes in US district courts: Can offenders’ educational attainment guard against prevalent criminal stereotypes? Crime & Delinquency, 63(2), 137–165.

18.

Frase

R. S.

(2000). Is guided discretion sufficient? Overview of state sentencing guidelines. St. Louis University Law Journal, 44, 425-430.

19.

Frase

R. S.

(2005). State sentencing guidelines: Diversity, consensus, and unresolved policy issues. Columbia Law Review, 105, 1190-1232.

20.

Gainey

R. R.

Steen

Engen

R. L.

(2005). Exercising options: An assessment of the use of alternative sanctions for drug offenders. Justice Quarterly, 22, 488-520.

21.

Harer

M. D.

(1995). Prison education program participation and recidivism: A test of the normalization hypothesis. Washington, DC: Federal Bureau of Prisons, Office of Research and Evaluation.

22.

Hofer

P. J.

(2007). United States v. Booker as a natural experiment: Using empirical research to inform the federal sentencing policy debate. Criminology & Public Policy, 6, 433-460.

23.

Hofer

P. J.

Meierhoefer

B. S.

(1987). Home confinement: An evolving sanction in the federal criminal justice system (Vol. 87, No. 4). Washington, DC: Federal Judicial Center.

24.

Johnson

B. D.

Betsinger

(2009). Punishing the “model minority”: Asian-American criminal sentencing outcomes in federal district courts. Criminology, 47, 1045-1090.

25.

Johnson

B. D.

Dipietro

S. M.

(2012). The power of diversion: Intermediate sanctions and sentencing disparity under presumptive guidelines. Criminology, 50, 811-850.

26.

Kautt

P. M.

(2002). Location, location, location: Interdistrict and intercircuit variation in sentencing outcomes for federal drug-trafficking offenses. Justice Quarterly, 19, 633-671.

27.

Kennedy

(1997). Race, crime, and law. New York: Pantheon Books.

28.

Kitchin

R. M.

(1994). Cognitive maps: What are they and why study them? Journal of Environmental Psychology, 14, 1-19.

29.

Matei

Ball-Rokeach

S. J.

Qiu

J. L.

(2001). Fear and misperception of Los Angeles urban space: A spatial-statistical study of communication-shaped mental maps. Communication Research, 28, 429-463.

30.

May

D. C.

Wood

P. B.

(2005). What influences offenders’ willingness to serve alternative sanctions? The Prison Journal, 85, 145-167.

31.

Moore

N. T.

May

D. C.

Wood

P. B.

(2008). Offenders, judges, and officers rate the relative severity of alternative sanctions compared to prison. Journal of Offender Rehabilitation, 46, 49-70.

32.

Morris

Tonry

M. H.

(1991). Between prison and probation: Intermediate punishments in a rational sentencing system. Oxford, UK: Oxford University Press.

33.

Murray

Farrington

D. P.

Sekol

(2012). Children’s antisocial behavior, mental health, drug use, and educational performance after parental incarceration: A systematic review and meta-analysis. Psychological Bulletin, 138, 175-210.

34.

Petersilia

Cullen

F. T.

(2015). Liberal but not stupid: Meeting the promise of downsizing prisons. Stanford Journal of Criminal Law and Policy, 2, 1-44.

35.

Rios

V. M.

(2006). The hyper-criminalization of Black and Latino male youth in the era of mass incarceration. Souls, 8, 40-54.

36.

Roberts

D. E.

(2004). The social and moral cost of mass incarceration in African American communities. Stanford Law Review, 56, 1271-1305.

37.

Rose

D. R.

Clear

T. R.

(1998). Incarceration, social capital, and crime: Implications for social disorganization theory. Criminology, 36, 441-480.

38.

Savelsberg

(1992). Law that does not fit society: Sentencing guidelines as a neoclassical reaction to the dilemmas of substantivized law. American Journal of Sociology, 97, 1346-1381.

39.

Smith

Goggin

Gendreau

(2002). The effects of prison sentences and intermediate sanctions on recidivism: General effects and individual differences (Rep. JS42-103/2002). Ottawa, Ontario: Solicitor General of Canada.

40.

Snyder

T. D.

Dillow

S. A.

(2011). Digest of Education Statistics 2010 (NCES 2011-015). Washington, DC: National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education.

41.

Spohn

C. C.

(2000). Thirty years of sentencing reform: The quest for a racially neutral sentencing process. In Horney

(Ed.), Policies, process, & decisions of the criminal justice system: Criminal Justice 2000 (pp. 427-501). Washington, DC: U.S. Department of Justice.

42.

Spohn

C. C.

Holleran

(2000). The imprisonment penalty paid by young, unemployed black and Hispanic male offenders. Criminology, 38, 281-306.

43.

Starr

S. B.

Rehavi

M. M.

(2013). Mandatory sentencing and racial disparity: Assessing the role of prosecutors and the effects of Booker. Yale Law Journal, 123, 2-80.

44.

Steffensmeier

Demuth

(2000). Ethnicity and sentencing outcomes in US federal courts: Who is punished more harshly? American Sociological Review, 65, 705-729.

45.

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, 763-798.

46.

Stith

Cabranes

J. A.

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

47.

Tonry

(1996). Sentencing matters. Oxford, UK: Oxford University Press.

48.

Tonry

(1998). Intermediate sanctions in sentencing guidelines. In Tonry

(Ed.), Crime and justice: A review of research (pp. 199-253). Chicago, IL: University of Chicago Press.

49.

Tonry

Lynch

(1996). Intermediate sanctions. In Tonry

(Ed.), Crime and justice: A review of research (pp. 99-144). Chicago, IL: University of Chicago Press.

50.

Turner

Petersilia

Deschenes

E. P.

(1992). Evaluating intensive supervision probation/parole (ISP) for drug offenders. Crime & Delinquency, 38, 539-556.

51.

Ulmer

J. T.

(2001). Intermediate sanctions: A comparative analysis of the probability and severity of recidivism. Sociological Inquiry, 71, 164-193.

52.

Ulmer

J. T.

(2012). Recent developments and new directions in sentencing research. Justice Quarterly, 29, 1-40.

53.

Ulmer

J. T.

Kramer

J. H.

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

54.

Ulmer

J. T.

Light

M. T.

Kramer

J. H.

(2011a). The “liberation” of federal judges’ discretion in the wake of the Booker/Fanfan decision: Is there increased disparity and divergence between courts? Justice Quarterly, 28, 799-837.

55.

Ulmer

J. T.

Light

M. T.

Kramer

J. H.

(2011b). Racial disparity in the wake of the Booker/Fanfan decision. Criminology & Public Policy, 10, 1077-1118.

56.

United States v. Booker, 543 U.S. 220 (2005).

57.

U.S. Sentencing Commission. (1996). Sentencing options under the guidelines. Washington, DC: Author.

58.

U.S. Sentencing Commission. (2009). Alternative sentencing in the federal criminal justice system. Washington, DC: Author.

59.

Von Hirsch

(1976). Doing justice: The choice of punishments. New York, NY: Hill & Wang.

60.

Weber

(1968). Economy and society. New York, NY: Bedminster Press.

61.

Western

(2006). Punishment and inequality in America. New York: Russell Sage Foundation.

62.

Wildeman

(2009). Parental imprisonment, the prison boom, and the concentration of childhood disadvantage. Demography, 46, 265-280.

63.

Wooldredge

Gordon

(1997). Predicting the estimated use of alternatives to incarceration. Journal of Quantitative Criminology, 13, 121-142.

64.

Zatz

(2000). Convergence of race, ethnicity, and class on court decision-making: Looking toward the 21st century. In Horney

(Ed.), Policies, process, & decisions of the criminal justice system: Criminal Justice 2000 (pp. 503-552). Washington, DC: U.S. Department of Justice.