Equal Justice

Abstract

Empirical analysis of the disproportionate application of carceral punishment has traditionally targeted race and class inequality while omitting noncitizens as a systematically disadvantaged population within the criminal justice system. Of the limited extant literature on this issue, nearly all have examined overall incarceration odds while failing to account for prison alternative eligibility, inaccurately measuring judicial discretion. Likewise, none have disaggregated noncitizens across nationality, an oversight that implicitly assuming that all noncitizens are equal recipients of discrimination, likely suppressing noncitizen disadvantage. Finally, these studies often fail to include contextual measures in their analyses. Using data from the U.S. Sentencing Commission (USSC)’s Monitoring of Federal Sentences from 1999 to 2013, this study examines case-, district-, and cross-level effects of citizenship status, documentation status, and nationality on incarceration odds, prison alternatives, and sentence length for federal drug offenders. The results of this study support the hypothesis that noncitizens receive more severe sentencing outcomes than U.S. citizens, Mexican noncitizens receive more severe outcomes compared to those from other countries, and undocumented noncitizens receive more punitive outcomes, though these findings vary across districts. However, counter to minority threat theory, noncitizen (offender) populations do not appear to influence incarceration outcomes for noncitizen offenders in the projected direction.

Keywords

sentencing courts/law inequality and crime other race and crime/justice quantitative methods legal issues

Immigration concerns have yet again reached volatile levels within societal discourse, infiltrating politics, the media, and even entertainment (Sowards & Pineda, 2013). This modern eruption of immigration controversy, while analogous to preceding ones regarding economic, cultural, and criminogenic themes, has surpassed its predecessors in sheer magnitude with larger turnouts at rallies and protests as well as increased media coverage (Sowards & Pineda, 2013). Considering the ubiquity of immigration discourse, related rhetoric risks influencing attitudes and responses toward noncitizens throughout various institutions, including the criminal justice system. Unfortunately, courtroom actors do not think within a sociological vacuum but rather are situated within the context of the social, historical, and cultural and are therefore in jeopardy of inadvertently or blatantly using carceral punishment as a means of reasserting formal control over noncitizens.

Although the resulting depictions of those entering the country are varied and often competing, the prevailing immigrant narrative is not a flattering one. Noncitizens are increasingly depicted as a danger to the economy, national security, and public safety (Citrin, Green, Muste, & Wong, 1997; Demleitner & Sands, 2002; Stacey, Carbone-López, & Rosenfeld, 2011). Among other stereotypes, noncitizens are often portrayed as inherently criminogenic, frequently associated with drug lords and mafia kingpins (Vaughn, Salas-Wright, DeLisi, & Maynard, 2014). This association persists in spite of findings that noncitizens appear to commit far fewer crimes compared to U.S. citizens, either out of fear of deportation or cultural armamentarium, a concept known as the immigrant paradox (Vaughn et al., 2014). In actuality, assimilation and acculturation seem to increase, rather than reduce, crime among foreign-born populations (Alvarez-Rivera, Nobles, & Lersch, 2014). These often inaccurate narratives of the criminal immigrant promote apprehension and fear while further increasing punitiveness toward this demographic.

The association between noncitizens and criminality is further perpetuated in common vernacular referring to undocumented noncitizens as “illegal aliens.” This terminology proliferates criminogenic associations while simultaneously “othering” this population. Furthermore, imagery of “the illegal immigrant” is implicitly, and often unabashedly, applied to all immigrants (Flores, 2003), neglecting to acknowledge the migration of numerous documented noncitizens. As of 2012, only 3.5% of the U.S. population consisted of unauthorized immigrants, representing only 26% of the foreign-born population (Passel & Cohn, 2014). Due to the copious application of this label, contemporary stereotypes of noncitizens fail to concede important distinctions within the broader social issue of immigration.

Further contributing to the homogenization of the noncitizen narrative is the extent to which immigrant has become synonymous with Hispanic, an often misleading association (Chan, 2013). Focusing solely on Hispanic migration to the United States distorts the unique nuances of population changes across various ethnic group (Chan, 2013; Morín, 2009). In 2010, more than 12% of the foreign-born population in the United States were from Europe, 53% were of Latin America and North American decent, 28% were Asian, and almost 5% arrived from other localities (Grieco et al., 2012). It is essential that subsequent empirical research examining crime and punishment among noncitizens incorporate ethnicity and country of origin.

Unfortunately, the effects of the current immigrant narrative are not relegated to innocuous discourse but rather permeate numerous institutions resulting in tangible consequences. Noncitizens are perceived as inherently criminogenic, assumed to be undocumented, and consolidated into one ethnicity. As a result, they suffer from housing and employment discrimination, elevated levels of poverty, linguistic profiling, and hypersegregation (Massey, 2007; Vaughn et al., 2014). The effects of citizenship status have not yet been thoroughly examined within the criminal justice system. This is not to imply that noncitizens have been entirely omitted from criminological analysis, but rather that the existing literature seems to suffer from a limitation similar in nature to that of the existing immigrant narrative: that immigrants are a singularity. In doing so, distinctions across ethnicity and documentation status have been neglected. This study endeavors to rectify this omission by deconstructing the immigrant narrative in the examination of the effects of noncitizenship, documentation status, and country of origin on decisions to incarcerate as well as sentence length outcomes.

Extant Literature

Sentencing inequality has been an escalating concern among criminological researchers, with copious studies exhuming evidence of gender, age, and racial disparities across a multitude of sentencing decisions including prosecutorial charging practices, the utilization of parole and community sanctions, jail and prison incarceration outcomes, sentence length, and departures from guidelines at the county, state, and federal levels (e.g., Blowers & Doerner, 2015; Doerner & Demuth, 2014; Feldmeyer, Warren, Siennick, & Neptune, 2015; Koons-Witt, Sevigny, Burrow, & Hester, 2014; Kutateladze, Andiloro, Johnson, & Spohn, 2014; Logue, 2009; Morgan & Smith, 2008; Starr & Rehavi, 2013; Stringer & Holland, 2016; Walker, Spohn, & DeLone, 2012). One particularly detrimental lapse in the extant literature on sentencing disparities, however, is the extensive evaluation of noncitizen disparities in sentencing outcomes.

According to the USSC Section 5H1.10, prosecutors and judges should not “use race and national origin as criteria in making sentencing decisions” (Demuth, 2002, p. 272). As such, systematic incarceration disparities between citizens and noncitizens would violate federal policy. While the disproportionate incarceration of noncitizens is by no means a new phenomenon, it is one that has begun to amass considerable attention and continues to inspire criminological analysis.

Noncitizen Federal Sentencing Outcomes

Of the relatively few studies exploring noncitizen disadvantage in the criminal justice system, most reach a similar conclusion. It seems as though noncitizens convicted at the federal level are disproportionately more likely than their U.S. counterparts to receive a prison sentence compared to alternative sanctions even when controlling for legally relevant variables (Albonetti, 1997; Demuth, 2002; Katzenelson, Conley, & Martin, 1996; Light, 2014; Wolfe, Pyrooz, & Spohn, 2011). This finding remains consistent across diverse methodologies and time frames, though the gap varies from as little as a 2% increased odds (Albonetti, 1997), to as much as 9.5 times greater odds (Wolfe et al., 2011). Conversely, noncitizens may be the recipients of leniency with regard to sentence length once incarcerated. Empirical research has found a noncitizen advantage in sentence length ranging from roughly 3% shorter sentences (Wu & DeLone, 2012) to as much as 22 fewer months (Katzenelson et al., 1996), though some variations in the literature are discernible (Albonetti, 1997; Demuth, 2002; Light, 2014).

Several explanations have been proffered for why the aforementioned variations in sentence outcomes findings have occurred. These conflicting results are likely impacted by methodological, regional, temporal, and sampling dissimilarities. For instance, the effects of citizenship status may vary according to sampling strategies across offense type, year of analysis, and district in which the case was tried. Furthermore, some studies examine data prior to the 2005 Booker ruling which authorized judges to step beyond sentencing guidelines, explaining some variation between pre-Booker and post-Booker data. Finally, each of the existing studies above examine overall incarceration odds even though many violations are associated with mandated prison sentences. According to the USSC’s annual report (2013), the gap between citizens and noncitizens is particularly large when limiting the analysis of citizenship effects to only those who have the option of receiving a prison alternative. In response to this assertion, this study compares outcomes across overall incarceration disparities, imprisonment odds only for those cases eligible for prison alternatives, and sentence length.

Ethnicity and Documentation Status

With immigration and ethnicity so thoroughly intertwined within criminological discourse, ethnicity is often used informally as a proxy for noncitizenship (Chan, 2013). Although noncitizens consist of a wide variety of distinct cultures and ethnicities migrating from diverse countries, a majority of immigrants are of Latino descent (Massey, 2007; Musto, 1999). However, while U.S. policy was cracking down on undocumented Mexican immigration during the 1990s, concerns over Canadian immigration failed to illicit governmental response (K. R. Johnson, 1996), suggesting the “immigration crisis” is actually one concerning ethnicity. In light of findings such as these, researchers are being urged to examine ethnic distinctions between citizens and noncitizens as sentencing outcomes may be an especially punitive response resulting from the intersectionality of their ethnic minority status and citizenship status (Demleitner & Sands, 2002; Wolfe et al., 2011).

Of noncitizens, those with Latino heritage appear to be exceptionally disadvantaged within the criminal justice system. In her 1997 study, Albonetti found that the effects of being a noncitizen only increases sentencing severity for Black and Hispanic defendants. In 2002, Demuth found that Black and Hispanic defendants are more frequently incarcerated compared to White defendants, regardless of citizenship status, though race effects were found to be more prominent for citizens than for noncitizens. Wu and DeLone (2012) also note important ethnic/racial distinctions, finding that noncitizen Hispanics receive longer sentences, while noncitizen Asians received shorter ones.

Among only a few studies that endeavor to disaggregate Hispanic heritage and to acknowledge the importance of documentation status on federal sentencing outcomes, Logue (2009) found that noncitizen Mexican defendants receive shorter sentences and are more likely to be convicted on marijuana and powder cocaine than non-Mexican Latino noncitizens. In their 2011 study, Wolfe and colleagues find that Latino citizens receive shorter sentences than White citizens, but Latino undocumented noncitizens receive longer sentences. The identification of ethnic variations in sentencing outcomes between Mexican and non-Mexican Latino noncitizen offenders, both of which are often subsumed under the broader label of Hispanic, demonstrates the importance of further disaggregating ethnicity. This dichotomy, however, excludes noncitizens of other origins including Asia, Africa, and Europe among others. Furthermore, subsequent studies have been slow to emulate this approach, continuing to incorporate race and ethnicity as mere controls. As such, this is the first known endeavor to disaggregate ethnicity by country of origin with the intent of promoting a more robust understanding of noncitizen disadvantage across incarcerative outcomes.

Theoretical Framework

The labeling of noncitizens as inherently criminogenic is likely rooted in competition over perceived finite resources and exacerbated by media and political narratives that invoke harmful narratives (see Chavez, 2013). During periods of economic hardship, the widening of employable residents via immigration is construed as a financial burden. These populations are further viewed as a drain on limited education, medical, and welfare resources (Citrin et al., 1997; Demleitner & Sands, 2002; Stacey et al., 2011). Thus, the identification of this population as dangerous legitimizes criminal justice intervention, allowing the cultural majority to reclaim control over access to economic, political, and cultural resources. This theoretical perspective, often referred to as minority threat theory, ultimately assumes that “intergroup competition” is the source of disproportionately applied policies (Berg, 2013). Later extended to noncitizens and individuals of Latino ancestry in Chavez’s (2013) Latino threat perspective, this theoretical framework is rooted in critical theory and has been used to explain racial, ethnic, and gender disparities among the application of criminal justice sanctions directed at the enforcement of social control as a result of competition over limited resources (Rockques & Paternoster, 2011; Stolzenberg, D’Alessio, & Eitle, 2004). Therefore, noncitizens are expected to receive more punitive sentencing outcomes compared to U.S. citizens in an effort to limit their growing economic, political, and cultural power. With current rhetoric insinuating that migrants from Mexico are inherently criminogenic and “illegal,” this population is the likely recipient of the most severe applications of formal control.

Threat levels, however, are not constant across time and place. Blalock (1967), traditionally credited with introducing minority threat theory, argues that as the population of a racial minority increases, so does the level of threat perceived by the majority. The majority then attempts to reduce the threat via formal control (Wang & Mears, 2009). Although researchers often test the theory with linear models, assuming that as the proportion of minorities increase so does the level of threat, group threat theory actually implies a curvilinear relationship in which threat increases along with the out-group population until a particular point (Feldmeyer & Ulmer, 2011). Therefore, as noncitizens move into traditionally racially homogenous communities, the economic, cultural, and political levels of threat can be expected to increase.

Contextual Variations in Federal Sentencing Outcomes

Despite theoretical implications of the importance of controlling for geographic and population variations across place, the paucity of analyses acknowledging these contextual variants presents yet another critical limitation in the existing literature. While the greater noncitizen narrative is propagated at what appears to be a national level, receptiveness to this message is likely to vary. Residents of jurisdictions with larger noncitizen populations are more likely to have encountered and interacted with this demographic, potentially resulting in the rejection of the noncitizen singularity. With substantial variations in noncitizen population identified across geographic region, differential demographic and population measures are expected to influence sentencing disparities across place. As alleged by Cano and Spohn (2012), “aggregating data across all district courts may distort the reality of decision-making in each district court” (p. 327). Despite the importance of examining case-level outcomes nested within jurisdictions, very few studies have incorporated multilevel modeling in the examination of noncitizen sentencing disparities.

The value of controlling for geographic variations in federal sentencing outcomes was emphasized in Albonetti’s (1997) article in which she observed that sentence lengths vary significantly across circuits, with 8 of the 11 circuits demonstrating more punitive outcomes than the District of Columbia. Conversely, Kautt’s (2002) analysis of district-level variations in federal sentence length outcomes from 1998 to 1999 did not find considerable diversity across districts, with only 7% of variance explained at the district level and 3% explained at the circuit. Feldmeyer and Ulmer (2011) found that of district-level predictors, caseload measures proffer more potential for explaining district-level variations in sentencing severity than do population measures. Of studies examining differential sentencing outcomes across federal districts, none had specifically explored the importance of citizenship status until Light’s (2014) analysis of variations in incarceration odds and sentence lengths across federal districts from 1992 to 2009. According to this study, districts with recent increases in noncitizens appear to demonstrate more hardening effects for incarceration decisions. Surprisingly, the reverse is true for sentence length; districts with increases in noncitizen populations are associated with shorter sentence lengths. While a vital contribution to the existing sentencing literature, this study omits the relevancy of country of origin and documentation status. While race and ethnicity are included as control variables, this strategy neglects the complexity of ethnicity by oversimplifying unique cultural distinctions into arbitrary racial terms.

These, albeit few, studies endeavoring to contextualize immigrant sentencing disparities at the federal level are an invaluable contribution to the current sentencing literature, shedding light on many of the intricate effects of individual-level characteristics as well as geographic variations on understanding federal sentencing disparities between U.S. citizens and noncitizens. While it is worth noting that a few studies have also examined sentencing outcomes using multilevel techniques to analyze state data (see Ulmer & Johnson, 2004; Wang & Mears, 2009), such empirical approaches are not necessarily comparable to federal courts, with differing offenses, caseloads, and priorities, to name a few. Despite the flourishing attentiveness directed toward noncitizen courtroom disadvantage, many crucial themes remain unexplored. While incarceration decisions appear to be a prominent contributor of disadvantage, this sentencing gap between citizens and noncitizens sentenced to prison appears even greater when alternatives to incarceration are an option (USSC annual report, 2013). As asserted by Feldmeyer and Ulmer (2011), roughly 82% of federal defendants receive some prison. Therefore, a better measure of discretionary incarceration decisions is incarcerative decisions between citizens and noncitizens for those who qualify for nonincarcerative sanctions. Additionally, noncitizen characteristics, including ethnic and documentation status, influence noncitizen sentencing outcomes. In fact, the effects of citizenship may partially drive Latino sentencing disparities. However, researchers have traditionally included race as a trichotomy comprising Hispanic, Black, and White as control variables. Doing so neglects the complexity of ethnicity by oversimplifying unique cultural distinctions into arbitrary racial terms. A more useful strategy for measuring ethnicity would be to include one’s country of origin in addition to race. Another important distinction within noncitizens is the documentation status. Following Wolfe, Pyrooz, and Spohn (2011), the inclusion of whether the individual is in the country legally or illegally in studies examining the impacts of citizenship status across time and place is imperative. This study incorporates each of these elements.

Method

This study is designed to examine whether noncitizens are more likely to be incarcerated and receive longer sentence lengths than their U.S. citizen counterparts. Of particular interest, noncitizen disparities in incarceration odds are examined for those eligible for nonincarcerative sanctions, more effectively measuring discretionary discrimination. It also challenges traditional measures of ethnicity by incorporating noncitizen country of origin into the analysis, clarifying whether noncitizens of Mexican decent are more disadvantaged than those of other origins while controlling for documentation status. Finally, district-level variations in sentencing severity as well as cross-level interactions for noncitizens are examined.

Hypotheses

Hypothesis 1: In light of previous research on sentencing inequality, this analysis examines whether noncitizens receive more punitive sentencing outcomes compared to U.S. citizens. Noncitizens are expected to demonstrated higher incarceration odds (Hypothesis 1A and receive longer sentence lengths (Hypothesis 1B) while controlling for legally relevant variables.

Hypothesis 2: Furthermore, this study examines variations in sentencing outcomes within noncitizen populations. Specifically, this study examines whether noncitizens of Mexican origins receive for punitive sentencing than noncitizens of other nationalities (Hypothesis 2A) and whether undocumented noncitizens receive more punitive sentences than documented noncitizens (Hypothesis 2B).

Hypothesis 3: Finally, geographic variations in sentencing are explored. First, the direct effects of district-level predictors on sentencing outcomes are explored (Hypothesis 3A). Second, cross-level interactions are examined to discern whether district-level variables specifically influence sentencing outcomes for noncitizen offenders (Hypothesis 3B).

Data

According to Demuth (2002), a noncitizen-drug nexus exists in which assumptions of criminality among immigrants are especially robust for substance violations, indicating that noncitizen disproportionality is higher for drug violations compared to all other federal offenses, with the exception of immigration offenses. Furthermore, Demuth (2002) argues that drug crimes are associated with the largest race and ethnicity gap. Due to the unique intersectionality of these variables, it is particularly important to examine noncitizen disparities for federal drug violations. Therefore, this study analyzes sentencing outcomes for federal drug offenders using the Monitoring of Federal (MoF) Criminal Sentences data from 1999 to 2013. Compiled by the USSC, these data include all federal cases within the United States that were reported between October 1, 1998, and September 30, 2013, for a total of 1,081,779 federal cases over 15 years.

Level 1 data

The Level 1 data for these studies include all drug-related offenses reported to the USSC, a total of 388,229 cases (35.9% of total violations). However, due to the scope of this study, certain groups will be omitted for various reasons. Following Doerner and Demuth (2014), juveniles (ages 16 and 17 years) will be excluded because these cases are accompanied by unique circumstances that are likely to differ in substantial ways from adult cases. Only 88 cases from 1999 to 2013 included juvenile defendants, representing less than 0.1% of reported cases. Furthermore, cases in which the citizenship of the defendant is unknown will also be omitted. Once again, this population is relatively small, comprising of roughly 2% of cases. Finally, those who were extradited (1,191 or 0.1%) will have been removed from this study.

Level 2 data

Due to the frequency in which courtroom actors interact within the context of court processes, the court can function as a unique social world (Ulmer, 1997). As such, the courtroom environment may exert a greater influence on decision makers interpretations of threat than populations external to the courts. Examining variations in case outcomes across district courts is of further importance as these courtroom actors are at risk of being influenced by the state and local media and political rhetoric of the geographic location in which they live (Chavez, 2013). Therefore, the Level 2 data for this study incorporates district-level courtroom and offender measures by aggregating case-level measures from the MoF data to the district level for each of the districts. A total of 94 districts are present in the United States with each state housing between 1 and 4 districts. However, state and territories are likely to differ in various ways. Following Kautt (2002) and Ulmer, Eisenstein, and Johnson (2010), the Northern Mariana Islands, Guam, Puerto Rico, the Virgin Islands, and Washington DC are omitted. Thus, this study examines 89 of the 94 districts.

Variables

Dependent measures

This study examines case-level, district-level, and cross-level interactions for two dependent variables measuring sentencing severity: incarceration odds among eligible offenders and sentence length. In light of findings suggesting that noncitizens are disproportionately more likely to be incarcerated compared to U.S. citizens (Albonetti, 1997; Demuth, 2002; Wolfe et al., 2011), the study first examines disparities in incarceration odds. Incarceration outcomes are traditionally defined dichotomously as whether the offender was sentenced to prison (using the USSC’s PRISDUM measure). However, because the majority of federal drug cases result in a prison sentence (95%), the inclusion of incarcerative sentences for all offenders as a dependent variable presents a somewhat limited measure of discriminatory sentencing outcomes, potentially masking the extent and manifestation of inequality.

With so few cases receiving prison alternatives, a better measure of discretionary carceral decisions would include decisions to incarcerate among those eligible for prison alternatives. Although this measure is conveniently made available by the USSC (titled INOUTin the MoF Sentences data), it is infrequently examined in the sentencing literature with most studies utilizing the PRISDUM variable referenced above. According to the USSC’s annual report (2013), the gap between citizens and noncitizens is particularly large when examining citizenship effects for those who have the option of prison alternatives. Therefore, this study examines prison alternatives those offenders who are not required to serve a prison term using the USSC’s INOUT measure. As such, defendants who would receive a prison sentence regardless due to mandatory sentencing laws are excluded from this variable.

Furthermore, in order to address inconsistent findings regarding the effects of citizenship status on sentence length outcomes, this study will also examine legal and extralegal effects on this variable. This continuous measures ranges from 1 month to 2,880 months. Following Ulmer et al. (2010), this variable has been truncated at 470 months with longer sentences identified as life sentences; the average time sentenced is about 74 months. Due to the skewness of this variable, the natural log is used.

Theoretical variables

Several key theoretical variables are included as well. In order to examine variations between citizens and noncitizens, a dummy variable for noncitizen is included. Furthermore, documentation status is included with documented noncitizens omitted as the referent. Due to research suggesting that some nationalities are considered to be a “model minority” (Chan, 2013), while others are perceived as especially dangerous (Vaughn et al., 2014), the country of origin for noncitizens is included in the analysis. However, this variable is composed of 199 countries and was therefore recoded according to the United Nations Statistics Division’s (2013) definitions. However, several nationalities lacked a sufficient sample size, failing to demonstrate variability at Level 2 and thus misspecifying the statistical model. Therefore, several countries were subsumed under a miscellaneous nationality category. This final measure includes a dummy variable for each of the following: Africa, Asia, the Caribbean, Latin America, Mexico, and other countries (see Appendix A) with Mexico omitted as the reference category.

Control variables

Several legal variables are considered in this study. Considering judicial pressures to offer a specific sentence according to sentencing guidelines, it is necessary to control for the presumptive sentence as a legal control. This variable provides a measure of offense severity and criminal history. Without controlling for the presumptive sentence, the effects of extralegal characteristics on sentencing outcomes will likely be grossly exaggerated (see Engen & Gainey, 2000), therefore the minimum guideline sentence, coded in months, is included as a control variable. Although the severity of the crime and the offender’s criminal history are included in the calculation of the guidelines, it is possible that measures of crime severity and criminal history may influence sentencing outcomes beyond the presumptive sentence and are therefore included. The number of counts, which is operationalized numerically, is included as well as criminal history which is measured dichotomously with 1 indicating the presence of a criminal history. With so few cases consisting of more than one count, the variable was recoded dichotomously with multiple counts included in the analysis and single count omitted as the reference category. The case disposition is included as a dummy variable identifying whether the case went to trial. Although the sample includes only federal drug offenders, the drug type is also included, consisting of cocaine, crack, heroin, marijuana (referent), methamphetamine, and other.

Finally, in response to previously explored extralegal variables found to be related to sentence severity, this study controls for gender, education, age, and race. Female is included with males acting as a reference point. Offenders without high school degrees and those with some college are compared to offenders with a completed high school degree. Age is included as a continuous variable. Finally race, which was not found to demonstrate collinearity with nationality, is included. A dummy variable is added for Hispanic and Black, with White omitted as the reference category. Finally, due to the aggregation of multiple years of data, a dummy variable is included for each year with 1999 omitted as the referent. Due to space limitations, this control variable is omitted from the table. However, supplemental tables with these measures are available.

Table 1.

Case-Level Descriptive Statistics.

Case Level Variables	Mean	Range	SD	Total
Dependent variables
Prison incarceration (total)	0.95	1	0.22	376,985
Prison alternatives (eligible)	0.45	1	0.50	23,022
Logged length	3.60	8.81	1.21	353,692
Citizenship
Noncitizens	0.30	1	0.45	369,154
Documented	0.42	1	0.28	98,619
Undocumented	0.58	1	0.37	98,619
Country of origin
Africa	0.012	1	0.109	98,619
Asia	0.016	1	0.126	98,619
Caribbean	0.121	1	0.326	98,619
Latin America	0.105	1	0.306	98,619
Mexico	0.698	1	0.459	98,619
Other	0.041	1	0.199	98,619
Legal variables
Logged presumptive	3.54	9.21	1.50	335,581
Number of counts	1.35	432	1.86	377,814
Criminal history	0.74	1	0.43	366,803
Trial	0.04	1	0.19	377,547
Drug offense
Cocaine	0.23	1	0.42	377,861
Crack	0.20	1	0.39	377,861
Heroin	0.07	1	0.26	377,861
Marijuana	0.28	1	0.44	377,861
Meth	0.18	1	0.38	377,861
Other	0.05	1	0.21	377,861
Extralegal variables
Female	0.13	1	0.33	374,358
Education
No HS degree	0.49	1	0.50	377,861
Completed HS degree	0.33	1	0.47	377,861
Some/completed college	0.18	1	0.37	377,861
Race
White	0.26	1	0.43	377,861
Black	0.28	1	0.45	377,861
Hispanic	0.43	1	0.49	377,861
Other	0.03	1	0.16	377,861
Age	33.30	80	9.94	375,799

Note. HS = high school.

District variables

Several caseload district-level variables are included in this study including the trial caseload, the guideline minimum, multiple offenses, and drug caseloads. The trial caseload is measured by aggregating the microlevel dichotomous trial variable to the district level. It is likely that districts with greater trial caseloads may rely more heavily on heuristic decision-making techniques in an effort to reach quick sentencing decisions and reduce the case backlog (Kautt, 2002). Furthermore, following Light (2014), the average offense severity score is included for each district. This variable is calculated by aggregating the minimum presumptive sentence length for each district. The case-level variable for whether multiple charges were brought forward was also aggregated to the district level as an additional measure of district-level caseload severity. The drug-crime caseload is measured as the percent of specific drug offenses brought forward per district. It is possible that districts serving a higher portion of certain drug offenders may be inclined to demonstrate more punitive sentencing decisions for these offenses. With so few level cases present at Level 2 (89 districts), it is not feasible to include a measure for all drug violations. Therefore, the two drugs most frequently associated with foreign populations are included cocaine and marijuana.

Several theoretical Level 2 variables are also included in the study. According to group threat theory, districts with larger populations of a minority group can be expected to demonstrate more severe outcomes toward that population in an attempt to increase social control. In order to examine whether increases in noncitizen, and noncitizen proxies, are related to sentencing inequality, the noncitizen, Hispanic, Mexico, and undocumented populations are included. Each of the variables is aggregated to the district level using case-level dichotomous measures.

Finally, a few additional district-level controls were added. It may be the case that greater numbers of White offenders in a given jurisdiction will lower sentencing severity for noncitizens, many of whom are non-White, because attributions of the criminal immigrant may be less frequent in these locations. In light of the frequency in which age is found to be related to sentencing outcomes at the individual level, this variable could very well demonstrate aggregate-level effects and is therefore included. Furthermore, previous research has found a relationship between citizenship status and education with noncitizens having disproportionately less educational attainment compared to their citizen counterparts. Therefore, it may be the case that districts with greater numbers of offenders without high school degrees rely more heavily on stereotypes within these districts. Therefore, district-level measures are included to assess the effects of White offender populations, age average, and education on sentencing outcomes (Table 2).

Table 2.

District-Level Descriptive Statistics.

District Level Variables	Mean	Range	SD	Total
Courtroom variables
% Prison	0.94	0.24	0.04	89
Logged length average	4.03	1.66	0.36	89
Logged presumptive average	4.20	1.45	0.31	89
% Number of counts	0.23	0.60	0.13	89
% Trial	0.05	0.09	0.02	89
% Cocaine	0.22	0.45	0.10	89
% Crack	0.25	0.55	0.14	89
% Heroin	0.06	0.36	0.07	89
% Marijuana	0.16	0.63	0.12	89
% Meth	0.23	0.75	0.21	89
% Other drugs	0.06	0.23	0.04	89
Demography variables
% Noncitizens	0.21	0.67	0.15	89
% Documented	0.06	0.27	0.05	89
% Undocumented	0.12	0.44	0.09	89
% Female	0.13	0.18	0.04	89
% No HS degree	0.43	0.36	0.07	89
% Completed HS degree	0.36	0.36	0.07	89
% Some/completed college	0.18	0.17	0.04	89
% White	0.31	0.65	0.14	89
% Black	0.34	0.70	0.20	89
% Hispanic	0.29	0.80	0.19	89
% Other race	0.04	0.59	0.08	89
Age average	33.65	6.16	1.07	89
% Asia origin	0.03	0.01	0.01	89
% Africa origin	0.05	0.00	0.01	89
% Caribbean origin	0.25	0.03	0.05	89
% Latin America origin	0.20	0.02	0.03	89
% Mexico origin	0.64	0.13	0.13	89
% Other origins	0.13	0.01	0.02	89

Analytical Strategy

In order to analyze the effects of citizenship status on sentencing outcomes, several models will be examined. Model 1 for each dependent variable introduces the theoretical measures; Model 2 includes all case-level measures. District-level measures are added in Model 3; the final model incorporates all statistically significant case-, district-, and cross-level predictors. Variables found to be significant at (p < .05) were left random (see Raudenbush, Bryk, Cheong, Congdon, & du Toit, 2004) and have been bolded to demarcate the random effects. Grand mean centering is applied.

Goodness-of-fit statistics are calculated for several of the models. Level 1 R² values are calculated for each linear model using Kreft and de Leeuw’s (1998) equation: R² = (unrestricted error − restricted error)/unrestricted error. Unfortunately, the sigma σ value is not provided in the Bernoulli models; thus, no Level 1 variance is available for the dichotomous-dependent variables. Likewise, the Level 2 model fit is calculated for the multilevel models. The results of the analysis are provided in Tables 3 and 4. Due to space limitations the following variables, though included in the analysis, are omitted from the tables: offender education, whether the case went to trial, offender criminal history score, case-level guideline minimum, whether the offender committed multiple offenses, year of the offense, district-level guideline minimum averages, district-level multiple offenses averages, and district-level age averages.

Table 3.

Models Predicting Prison Alternatives (Eligible).

Case and District Level Variables	Model 1	Model 2	Model 3	Model 4
	Theoretical Measures	Full Level 1	Level 1 and Level 2	Final Model With Interactions
Level 2 predictors
Noncitizen $\bar{x}$	—	—	0.22** (0.56)	0.20*** (0.49)
Hispanic $\bar{x}$	—	—	1.87^† (0.34)	1.99* (0.33)
Mexico $\bar{x}$	—	—	3.21* (0.58)	—
Undocumented $\bar{x}$	—	—	5.30 (1.09)	15.89*** (0.80)
Trial $\bar{x}$	—	—	0.01* (2.13)	0.03* (1.76)
Cocaine $\bar{x}$	—	—	2.32^† (0.44)	—
Marijuana $\bar{x}$	—	—	0.25*** (0.32)	0.27*** (0.22)
White $\bar{x}$	—	—	0.72 (0.32)	0.84 (0.27)
No high school $\bar{x}$	—	—	0.88 (0.69)	—
Noncitizen cross—level interactions	0.64*** (0.12)	0.35*** (0.09)	0.65** (0.16)	0.65 (0.27)
Noncitizen $\bar{x}$	—	—	—	—
Hispanic $\bar{x}$	—	—	—	—
Mexico $\bar{x}$	—	—	—	—
Undocumented $\bar{x}$	—	—	—	—
Trial $\bar{x}$	—	—	—	—
Cocaine $\bar{x}$	—	—	—	—
Marijuana $\bar{x}$	—	—	—	0.12*** (0.42)
White $\bar{x}$	—	—	—	4.39* (0.58)
No high school $\bar{x}$	—	—	—	—
Level 1 predictors
Hispanic	0.66*** (0.08)	0.84** (0.06)	0.76*** (0.06)	0.75*** (0.06)
Undocumented	0.59*** (0.11)	1.00 (0.14)	0.67** (0.13)	0.67** (0.13)
Age	—	0.99 (0.01)	0.99 (0.01)	0.99 (0.01)
Asia	—	1.34 (0.25)	0.72 (0.28)	0.67 (0.29)
Africa	—	1.66 (0.54)	0.60 (0.37)	0.54 (0.36)
Caribbean	—	1.38* (0.16)	0.84 (0.20)	0.77 (0.20)
Latin America	—	1.96*** (0.19)	1.29 (0.21)	1.20 (0.21)
Other countries	—	1.35 (0.19)	0.87 (0.21)	0.80 (0.21)
Black	—	0.71*** (0.06)	0.74*** (0.07)	0.73*** (0.07)
Female	—	1.53*** (0.05)	1.53*** (0.05)	1.53*** (0.05)
Cocaine	—	0.97 (0.07)	0.94 (0.08)	0.94 (0.07)
Crack	—	0.81** (0.08)	0.80** (0.08)	0.80** (0.08)
Heroin	—	0.80* (0.09)	0.76* (0.12)	0.75* (0.12)
Meth	—	0.77** (0.09)	0.74** (0.10)	0.73*** (0.09)
Other drugs	—	1.31** (0.08)	1.28** (0.09)	1.29** (0.09)
Reliability	0.646	0.780	0.693	0.674
n (Level 1)	23,022	23,022	15,823	15,823
n (Level 2)	—	—	89	89
R² (Level 2)	—	—	0.07	0.16
χ²	405.805***	252.324***	295.742***	262.510***

*p < .05. **p < .01. ***p < .001. ^†Approaching significance at p < .10.

Bolded values are used to demarcate which variables are free to vary.

Table 4.

Models Predicting Logged Sentence Length.

Case and District Level Variables	Model 1	Model 2	Model 3	Model 4
	Theoretical Measures	Full Level 1	Level 1 and Level 2	Final Model With Interactions
Level 2 predictors
Noncitizen $\bar{x}$	—	—	−0.24* (0.12)	−0.22* (0.09)
Hispanic $\bar{x}$	—	—	−0.04 (0.07)	—
Mexico $\bar{x}$	—	—	0.02 (0.11)	0.13^† (0.08)
Undocumented $\bar{x}$	—	—	0.12 (0.23)	−0.08 (0.21)
Trial $\bar{x}$	—	—	0.39 (0.41)	—
Cocaine $\bar{x}$	—	—	−0.02 (0.08)	—
Marijuana $\bar{x}$	—	—	0.03 (0.06)	—
White $\bar{x}$	—	—	−0.12^† (0.07)	−.09^† (0.05)
No high school $\bar{x}$	—	—	0.16 (0.14)	—
Noncitizen cross-level interactions	−0.05* (0.02)	0.07*** (0.01)	0.07*** (0.01)	0.11*** (0.02)
Noncitizen $\bar{x}$	—	—	—	—
Hispanic $\bar{x}$	—	—	—	—
Mexico $\bar{x}$	—	—	—	−0.33*** (0.04)
Undocumented $\bar{x}$	—	—	—	0.49*** (0.06)
Trial $\bar{x}$	—	—	—	—
Cocaine $\bar{x}$	—	—	—	—
Marijuana $\bar{x}$	—	—	—	—
White $\bar{x}$	—	—	—	—
No high school $\bar{x}$	—	—	—	—
Level 1 predictors
Hispanic	0.06** (0.02)	0.11*** (0.01)	0.11*** (0.01)	0.11*** (0.01)
Undocumented	0.05* (0.02)	0.02* (0.01)	0.02* (0.01)	0.02* (0.01)
Age	—	0.01 (0.01)	0.01 (0.01)	0.01 (0.01)
Asia	—	−0.01 (.02)	−0.02 (0.02)	−0.02 (0.02)
Africa	—	−.02 (.02)	−.02 (.02)	−.03 (.02)
Caribbean	—	−0.02** (0.01)	−0.03*** (0.01)	−0.03*** (0.01)
Latin America	—	−0.01 (0.01)	−0.01 (0.01)	−0.01 (0.01)
Other countries	—	−0.01 (0.02)	−0.01 (0.02)	−0.00 (0.02)
Black	—	0.12*** (0.01)	0.12*** (0.01)	0.12*** (0.01)
Female	—	−0.31*** (0.02)	−0.31*** (0.02)	−0.31*** (0.02)
Cocaine	—	0.08*** (0.01)	0.08*** (0.01)	0.08*** (0.01)
Crack	—	0.15*** (0.01)	0.15*** (0.01)	0.15*** (0.01)
Heroin	—	0.12*** (0.01)	0.12*** (0.01)	0.12*** (0.01)
Meth	—	0.13*** (0.01)	0.12*** (0.01)	0.12*** (0.01)
Other drugs	—	−0.01 (0.01)	−.02 (0.01)	−.02 (0.01)
Reliability	0.991	0.971	0.968	0.967
n (Level 1)	353,692	353,692	323,772	323,772
n (Level 2)	—	—	89	89
R² (Level 1)	0.02	0.64	0.65	0.65
R² (Level 2)	—	—	.80	.80
χ²	41,441.604***	7,850.978***	7,499.689***	7,439.006***

*p < .05. **p < .01. ***p < .001. ^†Approaching significance at p < .10.

Bolded values are used to demarcate which variables are free to vary.

Results

Exploratory Diagnostics

Before conducting the multilevel analyses, several assumptions were tested. All continuous variables were tested for skewness. Age, criminal history, and offense severity met the criteria for normality with values ranging between −1 and 1. The sentence length and guideline minimum were severely skewed and were transformed using the log of the variables. The number of counts were also skewed; however, few cases incorporated more than one count. Therefore, the number of offenses was transformed into a dummy variable that consists of those with only one count and those cases with multiple counts. Next, each variable was examined for multicollinearity. Of the variables, only offense severity and guideline minimum were found to be collinear with a variance inflation factor value over 4. The collinearity between severity and sentencing guidelines was expected considering that sentencing guidelines take into account the severity of the offense as well as criminal history; offense severity was, therefore, dropped from the models. Surprisingly, criminal history was not found to be collinear with the guideline minimum and is included in the model.

Prison Alternatives for Eligible Offenders

Noncitizens are significantly less likely to receive a nonprison sentence compared to their U.S. counterparts. When examining only theoretical variables, noncitizens are roughly 36% less likely to receive a prison alternative. Contrary to suggestions that sentencing disparities are a product of differential offending patterns, the introduction of legal variables not only fails to account for noncitizen punitiveness but rather appears to exacerbate this gap. When controlling for legally relevant case measures, the odds of receiving a prison alternative drop to 65% for noncitizen drug offenders, demonstrating a notable suppression effect (see MacKinnon, Krull, & Lockwood, 2000).

Of noncitizens, those without documentation are 41% less likely to receive a prison alternative than documented noncitizens; however, this finding is no longer statistically significant once legal variables are included. Further distinctions within noncitizens are also discernable among various nationalities. Noncitizen drug offenders from the Caribbean and Latin America are more likely to receive a prison alternative than those from Mexico, though these findings disappear once district-level findings are introduced. Individuals from Asia, Africa, and other countries do not differ significantly from those of Mexican descent with regard to prison alternative odds. These findings indicate that noncitizens appear somewhat consistently disadvantaged with fewer variations across nationality.

Sentencing outcomes were not found to be identical across districts, often varying according to offender populations, thus supporting the necessity of multilevel approaches to understanding federal sentencing disparities. Larger noncitizen caseloads are affiliated with lower odds of receiving a prison alternative. However, upon examining cross-level interactions, the relationship does not appear to specifically impact noncitizens outcomes. Conversely, greater Hispanic populations and undocumented populations appear to increase prison alternative odds, but yet again demonstrate no cross-level interactions. Contrary to expectations, larger White offender populations are related to an increase in prison alternative odds for noncitizens. Finally, increases in marijuana caseloads severely decrease prison alternatives specifically for noncitizens. This may be due to historical associations of this drug as originating in Latino/a communities. Therefore, of district-level threat indicators, none were found to be significantly related to noncitizen disadvantage in a manner consistent with minority threat (Blalock, 1967).

Sentence Length

Noncitizens initially appear to benefit from leniency when examining sentence length outcomes in Model 1 of Table 4. However, once legal and control variables are introduced, sentence lengths are longer for noncitizens than for their U.S. citizen counterparts. Undocumented noncitizens receive longer sentences than documented offenders, though the gap is somewhat narrowed once legal and extralegal variables are added. Once again, unique patterns emerge across nationality, though not necessarily as anticipated. Only Caribbean offenders receive significantly shorter sentences than Mexican offenders. Furthermore, only the miscellaneous category of nationalities varies across district, demonstrating a surprising degree of consistency.

Although threat indicators do not explain noncitizen disadvantage regarding prison alternatives, they appear slightly more applicable for sentence length disparities. Although undocumented offender populations demonstrate no direct effect on sentence length, this district measure was found to increase sentence length for noncitizen offenders in a manner consistent with minority threat theory. Conversely, larger caseloads of offenders from Mexico are related to increased sentence lengths, but severely decrease lengths for noncitizens. Larger noncitizen offenders are related to shorter sentence lengths overall but are not significantly related to noncitizen outcomes. Finally, while White caseloads are correlated with lower sentence lengths at a level approaching significance, this measure does not appear to specifically influence noncitizen outcomes.

Discussion and Conclusion

The results of this study contribute to the emerging criminological literature across multiple dimensions. First, it elucidates some of the conflicting empirical findings on noncitizen sentencing disadvantage across incarceration odds and sentence lengths while simultaneously expanding this limited literature with the introduction of differential odds of prison alternative outcomes. Like previous research (Albonetti, 1997; Demuth, 2002; Light, 2014; Wolfe et al., 2011), the results of this study indicate that noncitizens are more likely to be incarcerated than their U.S. counterparts when controlling for legally relevant variables. Furthermore, this disparity is unlikely to be a product of differential offending patterns. The greater incarceration odds for noncitizens who are eligible for prison alternatives while controlling for legal factors, an outcome that has been largely omitted from the sentencing literature, is demonstrative of discretionary disadvantage. Finally, this study attempts to shed light on the conflicting findings on sentence length outcomes for noncitizens. Although noncitizens receive shorter sentences than U.S. citizens (see Katzenelson et al., 1996; Wu & DeLone, 2012), this leniency disappears with the introduction of legal controls. Once legal controls for offense severity, criminal history, and the presumptive sentence are introduced, noncitizens receive significantly longer sentences as indicated in the works of Albonetti (1997) and Light (2014). Therefore, the results of this study support Hypothesis 1 in its entirety, demonstrating that noncitizens are disadvantaged across both sentencing outcome measures.

Yet another unique contribution of this study is the inclusion of country of origin as a more reliable indicator of ethnicity. As Chan (2013) observed, failure to distinguish within noncitizen variations introduces a layer of invisibility for several overlooked populations. However, the hypothesized differential outcomes across nationality was not found to be as prominent as anticipated, thus demonstrating little support for Hypothesis 2A. Noncitizens of Caribbean descent receive significantly more lenient sentencing outcomes compared to those from Mexico, congruent with immigration research citing Mexican immigrants as more at risk of punitiveness due to current rhetoric identifying them as dangerous (Vaughn et al., 2014). However, this finding was not significant for any of the other included nationalities. It is possible that ethnicity mediates the relationship between nationality and sentencing outcomes so that offenders identified as Hispanic are sentenced similarly regardless of nationality. Although the term Hispanic incorporates a large multitude of unique populations that do not necessarily share language, race, or culture (Hayes-Bautista & Chapa, 1987), this term has come to encapsulate these populations into a singularity resulting in the implicit homogenization of differing identities (Jones-Correa & Leal, 1996). Decision makers may be susceptible to this homogenization, failing to differentiate nationalities. This explanation fails to explain the significance of the Caribbean measure. Future research should further disentangle potential causal mechanisms related to the unique courtroom response to Caribbean drug offenders.

Although noncitizens of various nationalities may be uniformly disadvantaged in the courtroom, demonstrating evidence of the noncitizen singularity, sentencing outcomes vary significantly according to documentation status. In juxtaposition to country of origin, undocumented noncitizens received significantly more punitive sentencing outcomes across both sentencing decisions in nearly every model, as anticipated according to Hypothesis 2B. The variability in sentencing disparities among various noncitizen backgrounds demonstrates the importance of further disentangling noncitizen disadvantage.

Finally, the introduction of contextual measures further expands contemporary knowledge of noncitizen disadvantage. While differential offender populations across districts appear to be related to sentencing outcomes for noncitizen offenders, this relationship is not always exhibited in a manner consistent with traditional measures of minority threat theory. Noncitizen and Hispanic offender populations fail to demonstrate significant cross-level interactions effects on noncitizen sentencing severity. Of threat measures demonstrating cross-level interactions on noncitizen sentencing outcomes, increases in Mexican offender populations are related to reduced punitiveness for noncitizens rather than the hypothesized increase in severity, though this inverse relationship may be demonstrative of nonlinearity (see Feldmeyer & Ulmer, 2011). Future research should explore potential curvilinearity. Only undocumented populations increase incarceration odds and sentence lengths for noncitizens in a manner consistent with minority threat. Additionally, like Kautt (2002) and Feldmeyer and Ulmer (2011), caseload predictors were related to sentencing severity. In fact, caseloads predictors demonstrated the only cross-level interactions for prison alternatives. Marijuana caseloads decrease alternatives while White offender caseloads increase these odds. Punitiveness toward noncitizens in districts with higher marijuana caseloads may be evidence that immigrants are blamed for the introduction and popularity of this drug which has been historically associated with Latin American culture (Musto, 1999). Furthermore, increased caseloads are likely to increase the reliance on stereotypes in the effort to quicken the sentencing process, suggesting that the “assembly line justice” approach may exacerbate inequality (Lydgate, 2010). While district-level variables appear to influence sentencing outcomes for noncitizens, this relationship is not always manifested along the anticipated trajectory, providing mixed support for Hypothesis 3 while contemporaneously illustrating the importance of further contextualizing criminal justice outcomes across geographic context.

These findings are essential for the evaluation and elaboration of criminological theory. Minority threat theory argues that increases in the population of a particular out-group will be met with heightened punitiveness in an effort to control this population (Berg, 2013; Blalock, 1967; B. D. Johnson, Steward, Pickett, & Gertz, 2011; Wang & Mears, 2009); therefore, increases in noncitizen populations should increase sentencing severity only for noncitizens. Contrary to this assumption, no cross-level interactions were found for noncitizen populations on noncitizen case-level outcomes. However, this study finds that larger populations of undocumented offenders increase sentence lengths for noncitizens even when controlling for the defendant’s documentation status. This finding suggests that threats of undocumented immigrants disproportionately affect all noncitizens, a probable indicator of the immigrant singularity resulting from the noncitizen narrative. Therefore, theoretical elaborations of minority threat theory should consider the potential for the transmission of threat.

Furthermore, the consistent disadvantage of noncitizens across federal districts despite the history of reduced judicial discretion resulting from the implementation of sentencing guidelines indicates valuable policy implications. The controversy over criminal justice punishment disparities in conjunction with a dedication to ensuring sociological equality has led to several criminological policies designed to ensure justice and fairness in an attempt to reify the legitimacy of the criminal justice system. In response to these challenges, judicial discretion has been limited via mandatory drug minimums as well as the establishment of sentencing guidelines. Congressional mandatory drug minimums became a popular response to efforts at controlling substance abuse and enterprises during the 1980s, though the origins of mandatory drug policies can be traced much earlier (USSC, 2011). Such policies include the Anti-Drug Abuse Act of 1986 and the Omnibus Anti-Drug Act of 1988, both of which established minimum penalties for offenders convicted of certain drug offenses (Albonetti, 2011). In addition to these minimums, the Sentencing Reform Act of 1984 created the USSC, a commission tasked with creating federal sentencing guidelines to which federal judges were expected to strictly adhere. Although the purpose of the resulting guidelines was to reduce discrimination by identifying the appropriateness of a sentence in lieu of the offense severity and criminal history of the offender, judges maintained the authority to sentence within a given range and depart from the guidelines under certain conditions.

In response to the many criticisms levied at these policies, including accusations that federal guidelines and mandatory drug sentences are excessively harsh (Bowman, 2005; Cassell, 2003; Steiker, 2013) and promote an overreliance on long-term incarceration as a response to offending (Bowman, 2005), efforts have been directed at restructuring federal guidelines, increasing judicial discretion, and addressing controversial mandatory minimum policies with varying degrees of success. Several Supreme Court decisions, including United States v. Booker (2005), United States v. Fanfan (2005), and Gall v. United States (2007), have addressed the need for a modicum degree of judicial discretion by reducing guideline adherence to advisory. Unfortunately, this approach has failed to alleviate sentencing inequality (Albonetti, 2011) and may even have exacerbated this social problem (Ulmer, Light, & Kramer, 2011). According to Albonetti (2011), “…18 years of following the step-by-step path to determine sentence outcomes probably has been routinized” (p. 1153) and did nothing to address the controversial mandatory minimum policies. This is not to imply that mandatory minimum policies have remained unchanged; efforts at drug policy reform include the introduction of a “safety valve” provision established by the Sentencing Improvement Act of 1993 and the reevaluation of disparate cocaine minimums in the Fair Sentencing Act of 2010 (USSC, 2011). Efforts at addressing the appropriateness of federal sentencing outcomes are not static but rather continue to evolve. In 2014, the USSC amended the existing federal drug guidelines, reducing offense levels by two (USSC, 2014). Efforts at restructuring the existing guidelines, like the 2014 drug guidelines amendment, may hold the key to improving social perceptions of fairness and legitimacy in sentencing outcomes.

While this study is an important contribution to the development of the criminological literature examining noncitizen disadvantage, a few limitations are worth noting. In order to conserve degrees of freedom at Level 2, this study does not include quadratic terms in order to assess nonlinear relationships, as advocated by some when testing minority threat theory. However, this study was not designed with the intent of theory testing but rather incorporates theory as a methodological guide. Furthermore, previous research has indicated that nonlinear relationships fail to explain ethnic disparities in incarceration outcomes (see Wang & Mears, 2009), thus this omission does not appear to be particularly detrimental to the study. Furthermore, examining relational patterns between sentencing severity and population change inherently relies on assumptions of threat. In order to confirm this chain of causal reasoning, future studies should incorporate attitudinal measures. Likewise, it is possible that relying on court level data as a population measure is somewhat tautological; attitudes of threat may influence arrests, inadvertently inflating noncitizen offender populations rather than actual population growth within a community, further complicating causal inferences. However, the utilization of offender population data for the Level 2 analysis can be justified according to the social world framework which suggests that the courtroom provides an environment comparable to a community (Ulmer, 1997). Therefore, courtroom actors are likely to be influenced by these surroundings, not necessarily those of the community. Future research should examine differential arrest and charging patterns among noncitizen populations. Furthermore, the observed significant leniency toward Caribbean noncitizen offenders compared to Mexican noncitizens is not readily explained by minority threat theory, especially considering other nationalities frequently failed to achieve statistical significance. As the introduction of nationality is somewhat exploratory in nature, future research should further elucidate the causal mechanisms behind this unique finding as well as further enhance policies designed to address inequality.

The purpose of this study is not to express a scathing indictment of the criminal justice system, but rather to illustrate the severe and damaging consequences of the current immigrant narrative. These messages demonstrate lasting impacts on the noncitizen population, proliferating xenophic, nativistic, and isolationist ideologies at the risk of marginalizing and stigmatizing noncitizens. As such, the most valuable policy suggestion for the resolution of this social justice issue is to advance pluralism through a changed immigrant narrative that highlights the importance of diversity and contributes to a more holistic and accepting society.

Supplemental Material

Supplemental Material, CJR_Supplemental_Tables_(2) - Equal Justice: Examining the Effects of Citizenship, Documentation Status, and Country of Origin on Carceral Punishment Across Federal Districts

Supplemental Material, CJR_Supplemental_Tables_(2) for Equal Justice: Examining the Effects of Citizenship, Documentation Status, and Country of Origin on Carceral Punishment Across Federal Districts by Melanie M. Holland in Criminal Justice Review

Footnotes

Appendix A

Table A1.

Country of Origin Recode.

Africa	Algeria, Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Cape Verde, Central African Republic, Chad, Comoros, Congo, Cote d’Ivoire, Djibouti, Egypt, Eritrea, Ethiopia, Gabon, Gambia, Ghana, Guinea, Kenya, Lesotho, Liberia, Libya, Madagascar, Malawi, Mali, Mauritius, Mauritania, Morocco, Mozambique, Namibia, Niger, Nigeria, Republic of Tanzania, Rwanda, Sao Tome and Principe, Senegal, Seychelles, Sierra Leone, Somalia, South Africa, South Sudan, Swaziland, Togo, Tunisia, Uganda, Zambia, Zimbabwe
Asia	Afghanistan, Armenia, Azerbaijan, Bahrain, Bangladesh, Bhutan, Brunei Darussalam, Cambodia, China, Cyprus, Georgia, Hong Kong, India, Indonesia, Iran, Iraq, Israel, Japan, Jordan, Kazakhstan, Korea, Kyrgyzstan, Kuwait, Laos, Lebanon, Malaysia, Maldives, Mongolia, Myanmar, Nepal, Oman, Pakistan, Philippines, Qatar, Saudi Arabia, Singapore, Sri Lanka, Syria, Taiwan, Tajikistan, Thailand, Turkey, Turkmenistan, United Arab Emirates, Vietnam, Yemen, and Uzbekistan
Caribbean	Antigua and Barbuda, Bahamas, Barbados, Cuba, Dominica, Dominican Republic, Grenada, Haiti, Jamaica, Montserrat, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, and Trinidad and Tobago
Latin America	Argentina, Belize, Bermuda, Bolivia, Brazil, Chile, Colombia, Costa Rica, Ecuador, El Salvador, Guatemala, Guyana, Honduras, Nicaragua, Panama, Paraguay, Peru, Suriname, Uruguay, and Venezuela
Mexico	Mexico
Others	Albania, Andorra, Australia, Austria, Belarus, Belgium, Bosnia and Herzegovina, Bulgaria, Czech Republic, Canada, Croatia, Denmark, Estonia, France, Fiji, Finland, Germany, Gilbert Islands, Greece, Guam, Hungary, Iceland, Ireland, Italy, Kosovo, Latvia, Liechtenstein, Lithuania, Luxembourg, Macedonia, Malta, Marshall Islands, Micronesia, Monaco, Moldova, Montenegro, Netherlands, New Zealand, Northern Mariana Islands, Norway, Other, Palau, Papua New Guinea, Poland, Portugal, Romania, Russia, Samoa, San Marino, Serbia, Slovakia, Slovenia, Solomon Islands, Spain, Sweden, Switzerland, Tonga, Ukraine, Great Britain/UK, United States, Vanuatu (208), and Yugoslavia

Declaration of Conflicting Interests

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

Funding

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

Supplemental Material

Supplemental material for this article is available online.

References

Albonetti

C. A.

(1997). Sentencing under the federal sentencing guidelines: Effects of defendant characteristics, guilty pleas, and departures on sentence outcomes for drug offenses, 1991–1992. Law and Society Review, 31, 789–822.

Albonetti

C. A.

(2011). Judicial discretion in federal sentencing. Criminology & Public Policy, 10, 1151–1155.

Alvarez-Rivera

L. L.

Nobles

M. R.

Lersch

K. M.

(2014). Latino immigrant acculturation and crime. American Journal of Criminal Justice, 39, 315–330.

Berg

J. A.

(2013). Opposition to pro-immigrant public policy: Symbolic racism and group threat. Sociological Inquiry, 83, 1–31.

Blalock

H. M.

Jr . (1967). Toward a theory of minority group relations. New York, NY: John Wiley.

Blowers

A. N.

Doerner

J. K.

(2015). Sentencing outcomes of the older prison population: An exploration of the age leniency argument. Journal of Crime and Justice, 38, 58–76.

Bowman

F. O.

III (2005). The failure of the federal sentencing guidelines: A structural analysis. Columbia Law Review, 105, 1315.

Cano

M. V.

Spohn

(2012). Circumventing the penalty for offenders facing mandatory minimums: Revisiting the dynamics of “sympathetic” and “salvageable” offenders.” Criminal Justice and Behavior, 39, 308–332.

Cassell

(2003). Too severe: A defense of the federal sentencing guidelines (and a critique of federal mandatory minimums). Stanford Law Review, 56, 1017.

10.

Chan

(2013). An invisibility cloak: The model minority myth and unauthorized Asian immigrants, UC Irvine Law Review, 3, 1281.

11.

Chavez

(2013). The Latino threat: Constructing immigrants, citizens, and the nation. Stanford, CA: Stanford University Press.

12.

Citrin

Green

D. P.

Muste

Wong

(1997). Public opinion toward immigration reform: The role of economic motivations. The Journal of Politics, 59, 858–881.

13.

Demleitner

N. V.

Sands

J. M.

(2002). Non-citizen offenders and immigration crimes: New challenges in the Federal system. Federal Sentencing Reporter, 14, 247–254.

14.

Demuth

(2002). The effect of citizenship status on sentencing outcomes in drug cases. Federal Sentencing Reporter, 14, 271.

15.

Doerner

J. K.

Demuth

(2014). Gender and sentencing in the federal courts are women treated more leniently? Criminal Justice Policy Review, 25, 242–269.

16.

Engen

R. L.

Gainey

R. R.

(2000). Modeling the effects of legally relevant and extralegal factors under sentencing guidelines: The rules have changed. Criminology, 38, 1207–1230.

17.

Feldmeyer

Ulmer

J. T.

(2011). Racial/ethnic threat and federal sentencing. Journal of Research in Crime & Delinquency, 48, 238–270.

18.

Feldmeyer

Warren

P. Y.

Siennick

S. E.

Neptune

(2015). Racial, ethnic, and immigrant threat: Is there a new criminal threat on state sentencing? Journal of Research in Crime and Delinquency, 52, 62–92.

19.

Flores

L. A.

(2003). Constructing rhetorical borders: Peons, illegal aliens, and competing narratives of immigration. Critical Studies in Media Communication, 20, 362–387.

20.

Grieco

E. M.

Trevelyan

Larsen

Acosta

Y. D.

Gambino

De La Cruz

… Walters

(2012). The size, place of birth, and geographic distribution of the foreign-born population in the United States: 1960 to 2010. Population Division Working Paper, 96. Retrieved from https://www.census.gov/content/dam/Census/library/working-papers/2012/demo/POP-twps0096.pdf

21.

Hayes-Bautista

D. E.

Chapa

(1987). Latino terminology: Conceptual bases for standardized terminology. American Journal of Public Health, 77, 61–68.

22.

Johnson

B. D.

Steward

Pickett

Gertz

(2011). Ethnic threat and social control: Examining public support for judicial use of ethnicity in punishment. Criminiology, 49, 401–441.

23.

Johnson

K. R.

(1996). ‘Aliens’ and the U.S. immigration laws: The social and legal construction of nonpersons. Inter-American Law Review, 28, 263–292.

24.

Jones-Correa

Leal

D. L.

(1996). Becoming “Hispanic”: Secondary panethnic identification among Latin American-origin populations in the United States. Hispanic Journal of Behavioral Sciences, 18, 214–254.

25.

Katzenelson

Conley

Martin

(1996). Non-US citizen defendants in the federal court system. Federal Sentencing Reporter, 9, 259–263.

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.

Kreft

G. G.

de Leeuw

(1998). Introducing multilevel modeling. London, England: Sage.

28.

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.

29.

Kutateladze

B. L.

Andiloro

N. R.

Johnson

B. D.

Spohn

C. C.

(2014). Cumulative disadvantage: Examining racial and ethnic disparity in prosecution and sentencing. Criminology, 52, 514–551.

30.

Light

M. T.

(2014). The new face of legal inequality: Noncitizens and the long-term trends in sentencing disparities across U.S. District Courts, 1992–2009. Law & Society Review, 48, 447–478.

31.

Logue

M. A

. (2009). ‘The price of being Mexican’: Sentencing disparities between noncitizen Mexican and non-Mexican Latinos in the Federal courts. Hispanic Journal of Behavioral Sciences, 31, 423–445.

32.

Lydgate

J. J.

(2010). Assembly-line justice: A review of operation streamline. California Law Review, 98, 481–544.

33.

MacKinnon

D. P.

Krull

J. L.

Lockwood

C. M.

(2000). Equivalence of the mediation, confounding, and suppression effect. Prevention Science, 1, 173–181.

34.

Massey

D. S.

(2007). Categorically unequal: The American stratification system. New York, NY: Russel Sage Foundation.

35.

Morín

J. L

. (2009). Latino/as and US Prisons. In Oboler

, Behind bars (pp. 17–38). New York, NY: Palgrave Macmillan US.

36.

Morgan

K. D.

Smith

(2008). The impact of race on parole decision-making. Justice Quarterly, 25, 411–435.

37.

Musto

D. F.

(1999). The American disease: Origins of narcotic control. New York, NY: Oxford University Press.

38.

Passel

J. S.

Cohn

(2014). Unauthorized immigrant totals rise in 7 States, fall in 14. Pew Research Center. Retrieved from http://www.pewhispanic.org/2014/11/18/unauthorized-immigrant-totals-rise-in-7-states-fall-in-14/

39.

Raudenbush

Bryk

Cheong

Y. F.

Congdon

Du Toit

(2004). HLM 6. Lincolnwood, IL: Scientific Software International.

40.

Rockques

Paternoster

(2011). Understanding the antecedents of the “school-to-jail” link: The relationship between race and school discipline. Journal of Criminal Law & Criminology, 101, 633–665.

41.

Sowards

S. K.

Pineda

R. D.

(2013). Immigrant narratives and popular culture in the United States: Border spectacle, unmotivated sympathies, and individualized responsibilities. Western Journal of Communication, 77, 72–91.

42.

Stacey

Carbone-López

Rosenfeld

(2011). Demographic change and ethnically motivated crime: the impact of immigration on anti-Hispanic hate crime in the United States. Journal of Contemporary Criminal Justice, 27, 278–298.

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.

44.

Steiker

C. S.

(2013). Lessons from two failures: Sentencing for cocaine and child pornography under the federal sentencing guidelines in the United States. Law & Contemporary Problems, 76, 27.

45.

Stolzenberg

D’Alessio

S. J.

Eitle

(2004). A multilevel test of racial threat theory. Criminology, 42, 673–698.

46.

Stringer

R. J.

Holland

M. M.

(2016). It’s not all black and white: A propensity score matched, multilevel examination of racial drug sentencing disparities. Journal of Ethnicity in Criminal Justice, 14, 327–347.

47.

Ulmer

(1997). Social worlds of sentencing: Court communities under sentencing guidelines. Albany: State University of New York Press.

48.

Ulmer

J. T.

Eisenstein

Johnson

B. D.

(2010). Trial penalties in federal sentencing: Extra-guidelines factors and district variation. Justice Quarterly, 27, 560–592.

49.

Ulmer

J. T.

Johnson

(2004). Sentencing in context: A multilevel analysis. Criminology, 42, 137–178.

50.

Ulmer

J. T.

Light

M. T.

Kramer

J. H.

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

51.

United Nations Statistics Division. (2013). Composition of macro geographical (continental) regions, geographical sub-regions, and selected economic and other groupings. Retrieved from http://unstats.un.org/unsd/methods/m49/m49regin.htm

52.

U.S. Sentencing Commission. (2013). U.S. Sentencing Commission 2013 Annual Report. Retrieved from https://www.ussc.gov/about/annual-report/archive/annual-report-2013

53.

U.S. Sentencing Commission. (2014). U.S. Sentencing Commission 2014 Drug Guidelines Amendment Retroactivity Data Report. Retrieved from https://www.ussc.gov/sites/default/files/pdf/research-and-publications/retroactivity-analyses/drug-guidelines-amendment/20161025-Drug-Retro-Analysis.pdf

54.

Vaughn

M. G.

Salas-Wright

C. P.

DeLisi

Maynard

B. R.

(2014). The immigrant paradox: Immigrants are less antisocial than native-born Americans. Social Psychiatry and Psychiatric Epidemiology, 49, 1129–1137.

55.

Wang

Mears

D. P.

(2009). A multilevel test of minority threat effects on sentencing. Journal of Quantitative Criminology, 26, 191–215.

56.

Walker

Spohn

DeLone

(2012). The color of justice: Race, ethnicity, and crime in America. Belmont, CA: Cengage Learning.

57.

Wolfe

S. E.

Pyrooz

D. C.

Spohn

C. C.

(2011). Unraveling the effect of offender citizenship status on federal sentencing outcomes. Social Science Research, 40, 349–362.

58.

DeLone

M. A.

(2012). Revisiting the normal crime and liberation hypotheses: Citizenship status and unwarranted disparity. Criminal Justice Review, 37, 214–238.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.08 MB