Punishing the “Others”

Abstract

Despite ongoing debates on the continued legal significance of citizenship in a globalized world, international comparative tests of the salience of citizenship under the law have yet to be undertaken. This article combines data from US and German courts with interviews of judges from both countries to 1) estimate the punishment consequences of lacking state membership; 2) compare the sentencing gap across international contexts, and; 3) identify and explicate the mechanisms linking citizenship to punishment considerations. Findings show noncitizens receive increased punishment in US and German courts net of legal factors, but this effect is less pronounced in Germany. The interviews suggest that a variety of intervening mechanisms explain these results. Prominent among these is that judges in both countries resent the fact that noncitizens compound their immigrant status with criminal transgressions. However, German judges place greater emphasis on consistency and proportionality at sentencing, thus guarding against overly harsh and disparate punishments.

Résumé

Malgré les débats en cours sur la persistance de la signification juridique de la citoyenneté dans un monde globalisé, une approche comparée internationale de l’importance légale de la citoyenneté restait à entreprendre. Cet article combine les données issues des tribunaux américains et allemands avec des entretiens de juges des deux pays pour 1/ évaluer l’impact de l’absence de citoyenneté sur la peine prononcée ; 2/ comparer l’écart de peine en fonction des contextes ; 3/ identifier et expliquer les mécanismes reliant la citoyenneté aux considérations punitives. Les résultats montrent que les non-citoyens reçoivent une sanction accrue dans les tribunaux américains et allemands, abstraction faite des facteurs juridiques, avec un effet moins prononcé pour l’Allemagne. Les entretiens suggèrent une variété de mécanismes susceptibles d’expliquer ces résultats. En particulier le fait que les juges des deux pays craignent que les non-citoyens ne compliquent leur statut d’immigrant par des infractions pénales. Toutefois, les juges allemands accordent davantage d’importance à la cohérence et à la proportionnalité lors de la détermination de la peine, ce qui permet de se prémunir contre des sentences excessivement dures et disparates.

Zusammenfassung

Trotz der aktuellen Kontroverse über die Fortdauer der juristischen Bedeutung der Bürgerschaft in einer globalisierten Welt, steht eine vergleichende, internationale Studie über die gesetzliche Bedeutung der Staatsbürgerschaft noch aus. Dieser Beitrag kombiniert Daten amerikanischer und deutscher Gerichte mit Gesprächen, die mit Richtern beider Länder geführt worden sind, um 1. die Auswirkung der fehlenden Staatsbürgerschaft auf die ausgesprochene Strafe zu ermitteln, 2. den Strafunterschied je nach Kontext zu vergleichen, 3. die Mechanismen zwischen Staatsbürgerschaft und Strafmaßnahmen aufzudecken und zu erklären. Die Ergebnisse zeigen, dass Nicht-Staatsbürger von amerikanischen und deutschen Gerichte zu höheren Strafen verurteilt werden, von den juristischen Faktoren einmal abgesehen, wobei der Unterschied in Deutschland geringer ausfällt. Die Gespräche legen eine Vielfalt von Mechanismen nahe, die diese Ergebnisse erklären. Besonders die Tatsache, dass die Richter beider Länder befürchten, dass Nicht-Staatsbürger ihren Migrantenstatus aufgrund von Straftaten gefährden. Im Rahmen der Straffestlegung schenken die deutschen Richter jedoch der Kohärenz und der Verhältnismäßigkeit mehr Bedeutung, was vor besonders harten und disparaten Verurteilungen schützt.

There is considerable debate on the legal relevance of citizenship in an increasingly globalized world. One view holds that the diffusion of international norms of “personhood” [Soysal 1994] has resulted in a new “postnational” citizenship, making national membership “inventively irrelevant” [Soysal 1997] by weakening national sovereignty and eroding the differences between citizens and aliens under the law [Jacobson 1996]. Alternatively, some argue that citizenship remains an essential feature of modern life, highlighting the enduring power of the nation-state in creating laws that shape immigrant integration, including work, rights, and social benefits [Menjivar 2006; Goodman 2014]. Others outright reject the fact that the significance of citizenship has declined, arguing that the boundaries of national membership represent the central axes of exclusion in the modern world [Wimmer 2002].

However, research in this area has primarily focused on analyses of immigration policies [Joppke 1999], the historical development of nation-states [Brubaker 1992] or constitutional legal reviews [Rubio-Marín 2000]. As a result, we know comparatively little about the cross-national salience of citizenship under the law. This is a particularly notable omission as one of the most authoritative and symbolic forms of national sovereignty—the power to punish—has dramatically increased in recent decades, entangling disproportionate numbers of foreigners in coercive state controls. In the US, half of all federal offenders lack US citizenship [USSC 2010] and the population of noncitizens incarcerated in federal prisons increased more than five-fold between 1985 and 2010. Though the US is often an outlier in the context of punishment, the increase of incarcerated foreigners is symptomatic of a broader punitive trend across many Western societies. In the European Union, between 2000 and 2011, the foreign-born population rose from 5 to 7 percent [Eurostat 2011] while the proportion of incarcerated foreigners increased from 15 to 22 percent [Aebi and Delgrande 2011].

Despite this international trend, much of the empirical work on immigration and punishment has been concentrated on case studies of individual countries [Bosworth and Kaufman 2011; Light et al. 2014; Light 2015]. As a result, the literature lacks a systematic examination of the legal treatment of noncitizens across diverse national contexts and the mechanisms underlying this relationship remain unknown. Attempting to address these gaps, this article utilizes a comparison of sentencing outcomes in the United States and Germany—two advanced Western democracies with strong rule of law traditions but markedly different conceptions of citizenship, nationhood, and punishment regimes [Joppke 1999; Savelsberg 1994]—to examine the punishment consequences of lacking state membership. Combining quantitative analysis of court records and qualitative analysis of interview data with US and German judges, this study is the first to empirically evaluate the relationship between citizenship and punishment in an international comparative perspective.

Bridging two distinct substantive traditions—citizenship studies and sociology of punishment—I propose and test a model of citizenship as an instrument of stratification which connotes cultural “otherness” and influences judges’ perceptions of appropriate punishment responses to foreign criminals, resulting in more severe sentences for non-state members. In doing so, this study advances the sociological study of citizenship by examining the extent to which state membership is an axis of legal inequality beyond well-known patterns of racial/ethnic stratification. This inquiry is motivated by two central questions.

First, are noncitizens punished more harshly in the United States and Germany and, if so, does the legal treatment of foreign nationals differ between these countries. The strength of this comparison is twofold. To begin with, both countries are home to large numbers of diverse international migrant groups; they are also both Western and industrialized, with capitalist economies, democratic governments, and constitutional protections for citizens and noncitizens under the law [Savelsberg 1994]. ¹ Yet, they differ distinctly in their normative definitions of citizenship. The US represents the ideal typical case of jus soli (right of soil) citizenship, as the historical conception of the US as a “nation of immigrants” has negated a purely ethnic form of nationhood. In contrast, Germany is one of the only advanced nations with a long-standing tradition of citizenship by ethnic descent (jus sanguinis), having adhered to the maxim that Germany is “not a country of immigration” for much of the 20^th century [Brubaker 1992]. While recent policies have loosened the rules for obtaining citizenship, the cultural conceptions that define Germany as an ethno-cultural nation still factor prominently in society, drawing stark boundaries between Germans and non-Germans [Green 2005]. Comparing the US and Germany thus provides powerful insights into sociological debates on the contemporary relevance of citizenship by assessing the legal salience of national membership in countries with significantly different conceptions of nationhood.

This contrast also informs our understanding of comparative punishment severity. Though anti-immigrant sentiment is manifest in many Western countries, the comparative study of criminal punishment has identified several key features of the US and German legal systems that are likely to promote more moderate and non-disparate sentencing in Germany [Whitman 2005]. These include institutional differences in the production of knowledge about crime and punishment that stress welfare-oriented rationales [Savelsberg 1994], insulation of the judiciary from political influences [Albrecht 2013], and differences in the legal education, appointment, and socialization of judges that promote consistency in sentencing decisions [Hörnle 2013]. While this research exhibits impressive theoretical insight, empirical evidence remains largely preliminary as tests of comparative sentencing disparities are lacking [Ulmer 2012]. By comparing the punishment gap between citizens and noncitizens across US and German courts, this study is a notable exception.

Second, what are the mechanisms that explain any differential legal treatment of noncitizens? While I draw from detailed court records to examine citizenship disparity, this approach alone cannot capture the interpretive processes that yield unequal outcomes, and research has yet to examine the legal treatment of noncitizens using both statistical and qualitative methods. This article casts light on this understudied area of research by using interview data to explicate the direct and indirect mechanisms linking citizenship to punishment considerations among US and German judges.

In sum, the cross-national study of state membership and punishment directly engages ongoing citizenship debates by testing whether legal rights correspond to legal reality for noncitizens adjudicated in US and German courts. It also informs broader concerns germane to the sociology of state social control and immigrant incorporation. In reference to state social control, this research draws from the Weberian insight that the exercise of violence and punishment is one of the defining features of the modern nation-state. As such, the courts are an appropriate—indeed perhaps ideal—venue for assessing whether “citizen” and “noncitizen” remain consequential statuses in contemporary society. This is especially the case given that postnational scholarship has identified national and supranational courts—more than other institutions—as responsible for eroding the distinctions between citizens and noncitizens by determining societal benefits through residency [Joppke 2010; Jacobson 1996]. However, as noncitizens continue to fill the prisons of many Western countries and research increasingly documents the deleterious effects of incarceration, the time has come for a thorough empirical analysis of how noncitizens actually fare under the law.

Links between national membership and criminal punishment

Inequality under the law has been a longstanding enterprise in sociology, producing a voluminous literature on state social control and legal decision-making. The bulk of this research has focused on racial cleavages and, to a lesser extent, ethnic and class inequalities under the law. While different theories stress diverse explanations for disparity in the criminal justice system, a unifying theme across relevant perspectives is that broader patterns of stratification are often reflected in the legal system and that the application of legal sanctions may be influenced by factors beyond legally prescribed elements, including offender attributes. This latter point is consistent with general theories of legal decision-making, which suggest that criminal justice actors rely on multiple external cues to make attributions of several judicial focal concerns, including defendant blameworthiness, dangerousness, and the practical implications of their decisions [Steffensmeier et al. 1998]. Accordingly, judges may form their conceptions of appropriate punishments partially through negative attributes associated with certain offender characteristics such as racial and ethnic stereotypes, or perhaps views on national membership and immigrant criminality.

Much of the previous research investigating citizenship in criminal courts has suggested that national membership likely indirectly affects punishment through the consideration of culpability and dangerousness [Wolfe et al. 2011; Wu and DeLone 2012]. However, this body of research has relied exclusively on court records, thus precluding a full examination of the mechanisms linking national membership to punishment. This is a notable gap given that there are strong theoretical reasons for thinking that citizenship may lead to disparate legal treatment—at least partially—for reasons other than those linking factors such as race and ethnicity to increased punitiveness.

Citizenship and punishment: direct links

Both classic and contemporary theoretical perspectives on legal decision-making suggest criminal justice actors consider the safety of the community when determining legal sanctioning. While this idea has been an underlying theme within punishment research, limited scholarly attention has been paid to how judges define the “community,” and there has been little consideration of whether different offenders may shift judges’ perceptions as to which “community” is in need of protection. The guiding assumption is that judges consider the community in a jurisdictional sense, and a significant amount of research has examined geographic variation in the application of laws based on this insight [Baumer and Martin 2013]. In this article I broaden the sociological lens by conceptualizing the “community” beyond jurisdictional boundaries to include the political, symbolic and normative boundaries of national membership.

Research on the role and function of citizenship suggests national membership can be understood as a mechanism of stratification, one inherently implicated in defining members and non-members to the community of the nation-state [Brubaker 1992]. As such, a fundamental feature of the modern nation-state involves determining community membership through citizenship. In this vein, Wimmer [2002] argues that the contours of national membership center around the “imagined community” of the nation—a political community grounded on common origin and historical experience—and is determined through cultural compromise which can be characterized as a consensus over the validity of collective norms, social classifications, and world-view patterns that separate the homogenous domestic realm of the nation-state from the heterogeneous external one.

Within this framework, citizenship is necessarily discriminatory—a status which by its nature signifies “otherness” and, unlike race or ethnicity, citizenship is formal and institutional and is far more immune to charges of “prejudice” [Joppke 2010]. By defining the terms of legal, political and social exclusion, the nation-state “must discriminate between citizens and foreigners” [Brubaker 1992: 47]. As a result, citizenship raises normative questions about identity, belonging, and state obligations. These normative assessments suggest the line between citizens and noncitizens may be a salient consideration in the legal system because courts are charged with protecting and enforcing state interests, which are in turn framed and shaped by the cultural conceptions of nationhood [Brubaker 1992]. Given the cultural, legal, political, and normative boundaries that citizenship defines, it is likely that criminal justice actors use national membership not as a “proxy” for attributing negative qualities, but as an explicit status that is deserving of increased punishment. Sayad’s [2004] discussion of “double punishment” captures this idea well:

The fact of being an immigrant delinquent constitutes, as a general rule, something of an aggravating circumstance […] Before we can even speak of racism or xenophobia, the notion of “double punishment” is therefore present within any judgment passed on the immigrant. It is rooted in state thought […] All immigrant behavior, and especially deviant behavior, has repercussions on the phenomenon of immigration itself, and leads to greater disapproval, greater disqualification, and greater stigmatization. [Some] do not conceal or hide their satisfaction at seeing two different modalities of crime and the two punishments that sanction them overlapping and aggravating one another—in their view, this is only fair, and basically, something that is quite right and that should be the rule […] Breaking the law means more than the infraction in question: it is an error of a different order, a lack of politeness [Sayad 2004: 282-285].

Based on this perspective, state institutions and judicial actors will be more punitive against the criminality of non-state members for reasons directly linked to their citizenship status. If this is indeed the case, it suggests that some of the mechanisms linking citizenship to punishment may be distinct from other statuses stressed in previous research (e.g. race, ethnicity, class). Thus, the focal concerns’ emphasis on the indirect influences of offender attributes may be somewhat incomplete for a full understanding of the processes linking issues of national membership to punishment, though still applicable in important respects.

Citizenship and punishment: indirect links

By conceptualizing citizenship as a mechanism of stratification and an outcome of cultural compromise, two aspects of national membership are particularly likely to influence considerations of offender dangerousness and criminal risk—social marginalization and cultural dissimilarity. It is in these regards that the effect of citizenship at punishment parallels other offender statuses that are prone to stereotyping and disparate treatment (e.g. race).

A common thread within sociolegal studies is that those who are socially marginal are more likely to receive harsh treatment in the criminal justice system [Black 1976]. Several reasons have been put forth to explain these disparities, including lack of resources to impede the imposition of legal sanctions or lack of meaningful social bonds to demonstrate a low criminal risk, both of which are highly applicable to noncitizens. Across Western societies, noncitizens occupy a uniquely disadvantaged niche as their formal exclusion exacerbates other social inequalities [Massey 2007]. As Koopmans and Statham [1999: 662] note, “the different symbolic labels that nations attribute to migrants directly influence the distribution of material resources to them, and their potential for mobilizing challenges and participating within […] society.” Consistent with the literature on inequality and punishment, it follows that noncitizens will receive harsher treatment at sentencing due to their lack of societal integration and the inability to resist negative labels.

A second important aspect of citizenship for sentencing is its connotation with cultural distinctiveness. General theories of legal decision-making––including the focal concerns and Black’s theory of law—suggest that those perceived to be culturally dissimilar are more likely to be viewed as “threatening” or “unpredictable” [Steffensmeier and Demuth 2000]. As the organizing criterion for distinguishing cultural “insiders” from “outsiders,” it follows that those who are not members to the nation-state will be viewed as culturally dissimilar, and thus more likely to be viewed as criminal risks. This is especially the case in the context of perceived links between immigration and crime in public, political, and criminal justice discourse.

Despite considerable evidence to the contrary [Lee and Martinez 2009], large portions of the public in the US and Europe associate immigration with increased crime, particularly drug offending. In 2010, nearly 90 percent of Americans agreed that “protecting our borders is important to prevent drug trafficking and organized crime” [Ipsos 2010]. In Germany, 65 percent of respondents feared an “increase in drug trafficking and international organized crime” from further European Union integration. According to Bauman [2002], immigration, security and crime have become interwoven in much political discourse across Europe, where contemporary immigrants now poignantly represent today’s “figures of fear.”

Perhaps nowhere is the trend of the increased criminalization of immigration better captured than in national efforts to combat drug crimes. Since 2000 the European Union has developed multiple initiatives to curb drug crimes linked to immigration such as the Border Management Programme in Central Asia and the EU Drug Strategy [2007-2012]. In a similar vein, in recent decades the US has considerably revised immigration laws to include relatively minor drug crimes as deportable offenses. Because noncitizens face eventual deportation at the conclusion of their sentences, this may funnel them away from the non-incarceration punishments (e.g. probation) that are available to US citizens.

Taken together, the lack of resources and the possibility of deportation, combined with attributions that associate noncitizens with criminal dangerousness, are likely to increase sentence severity for noncitizens in US and German courts. However, comparative research provides ample reasoning to anticipate differences in this punishment gap across dissimilar contexts.

Differences in citizenship severity

The rich theoretical literature on comparative criminal punishment highlights several institutional, cultural, and normative aspects of the German legal system that guard against harsh and disparate sentencing. First, the institutional arrangement of the judiciary in Germany better insulates legal decision-making from political influences and shifts in public opinion than in the US [Savelsberg 1994]. Due to the fact that German judges are appointed as civil servants with tenured positions based on tests that operate in a fairly neutral and meritocratic fashion, they are not dependent on public approval or political parties. Though US federal judges are appointed rather than elected, research suggests that the political appointment and confirmation processes have become increasingly partisan [Epstein et al. 2013], and “politically motivated appointments of judges could be as distorting for moderate, equal sentencing as politically charged elections” [Hörnle 2013: 207]. Thus, to the extent that sentencing decisions are better insulated in Germany from anti-immigrant sentiment in public and political discourse, it is likely that there will be a lesser punishment gap between citizens and noncitizens.

Second, the legal culture and professional socialization among German judges places substantial emphasis on a corporatist self-understanding and internal consistency. Albrecht [2013] suggests this conservative inertia results from judges learning sentencing patterns almost exclusively from information on past sentences combined with judicial policies that minimize opportunities to diverge from established sentencing patterns, such as an appellate review system which views sentencing as an application of law, rather than a discretionary decision. In line with the aims of the Civil Law system, this corporatist spirit within the judiciary discourages individual expressions of personal and political preferences in sentencing decisions [Hörnle 2013], arguably lessening the influence of non-legal factors. In contrast, perhaps due to its history rooted in the Common Law tradition [Scott 2007], the US is better characterized by a “judges-as-individuals” model which allows for greater individual expression, as US judges retain considerable discretion and standards of appellate review remain highly deferential [Epstein et al. 2013].

A third distinction concerns the fundamentally different normative assessments of the appropriateness of severe punishments in the United States and Germany. In his historical analysis of the growing punishment divide between the US and Europe, Whitman [2003] suggests that a key factor explaining divergent punitive responses is the emphasis on degradation through punishment in the US, whereas the French and German systems are more willing to treat criminals with respect. Savelsberg [1994] attributes these penal differences to the institutional conditions that shape public, political, and academic knowledge about crime and punishment, and how this knowledge influences legal decision-making. Hörnle [2013], on the other hand, highlights the role of legal education in instilling a commitment to proportionality in sentencing and a deep skepticism of deterrence through harsh sanctioning in Germany. Regardless of the source, what is evident is that German judges are likely to view much of the severe sanctioning in the US as grossly disproportionate to the harm caused by the prohibited conduct. This emphasis on proportionality is likely to mitigate the influence of national membership at sentencing.

Summary

Overall, my review of the theoretical literature suggests that non-state members will be punished more severely than their citizen counterparts through a variety of direct and indirect pathways. Citizenship will be directly linked to punishment as the criminal transgressions of those who lack state membership will be viewed by judges as particularly deserving of harsher penalties, and indirectly related to sentencing through perceptions of lacking societal integration, the practical issues surrounding eventual deportation, and negative attributes associated with immigrants’ perceived cultural dissimilarity. This general framework suggests that the boundaries of citizenship confer legal disadvantages to those who lie outside the community of national membership, even in countries with distinct political, legal, and cultural definitions of nationhood. However, key institutional and legal culture differences in criminal sentencing suggest that there will be a lesser punishment gap between citizens and noncitizens in Germany.

Data, methods and logic of analysis

Statistical analysis

I address the above research questions using individual-level data from US and German courts in 2009 to 2010. The US data comes from the US Sentencing Commission’s (ussc) research files and the German data comes from the Strafverfolgungsstatistik (Prosecution Statistics) provided by the German Federal Statistical Office. Both datasets are rich sources of information on sentencing and provide detailed case and offender information, making available an extensive set of controls to assess the punishment consequences of citizenship in both countries.

In the ussc data, the unit of analysis is each sentenced case and the universe is all offenders sentenced in US federal courts. ² The unit of analysis in the Strafverfolgungsstatistik data is each prosecution and the universe is all defendants in German courts. Because only those who are convicted are at risk of punishment, I removed all cases that resulted in an acquittal.

All cases involving traffic violations are removed, as are those involving juveniles because in both countries there are different sets of laws and procedures for juvenile offenders. As a practical matter, this analysis removes offenders charged with unlawfully entering or remaining in the United States because US citizens are not at risk of being sentenced under this provision. For this same reason, I eliminate all asylum violations in the Strafverfolgungsstatistik.

• • Dependent variables

I assess sentencing as a two-stage process: (1) the decision to incarcerate, and (2) the length of imprisonment. In both the US and German data, incarceration is measured as a dichotomy for whether or not an offender was sentenced to prison (1 = yes). ³ For the ussc data, the sentence length measure is the logged number of months of incarceration. The Strafverfolgungsstatistik only releases information on the sentence range. There are 10 different sentencing categories, ranging from less than 6 months to life imprisonment. Because the sentence categories are not uniform and a “life” sentence has a qualitatively different meaning in Germany (usually resulting in approximately 15 years), a one unit increase in the regression models lacks a clear substantive interpretation. For this reason, I place greater emphasis on the direction of the citizenship results in the German sentence length analysis. However, to increase the comparability of the analyses, in the Online Supplement (Appendix A) I draw from an alternative German data source where the sentence length variable mirrors the US analysis. Coding and descriptive statistics for all variables are shown in Online Appendix B.

• • Focal predictor variables

Citizenship status is the focal independent variable (coded 1 for non-US/German citizens). For the US I further code for legal and undocumented immigrants. In both data sources I also classify noncitizens on the basis of country or region of origin. In the US data I use indicators for Mexican, European Union countries, Other European countries, Central or South American, Middle Eastern, and African/Asian. In the German data I use indicators for European Union countries, Other European, Turkish, Middle Eastern, those from the Americas, and African/Asian.

• • Legal, case, and offender characteristics

Previous research demonstrates that offense severity and criminal history are salient determinants of sentencing. In the US data, I capture crime severity by including the recommended punishment set forth by the sentencing guidelines. This measure accounts for the complexity of the guidelines calculation (e.g. sentencing adjustments, mandatory minimums) and represents the minimum length of incarceration recommended by the guidelines (logged). Consistent with prior research, I include a control for the offender’s criminal history score. In the Strafverfolgungsstatistik I measure offense severity using the statutory prescribed sentencing range for each offense. This measure is consistent with research conducted in the United States in non-guideline jurisdictions [Baumer and Martin 2013]. This variable ranges from a 6-month statutory maximum (coded 1) to a possible life sentence (coded 17). I also include two measures of the defendant’s criminal history: the number of prior convictions (ranging 0 to 9) and the most severe penalty from a prior conviction (ranging from no criminal history to incarceration exceeding 2 years).

In the US data I capture crime type differences with measures for drugs, immigration, ⁴ violent, firearms, sex offenses, fraud, and other offenses (property offenses as reference). In the German analysis I include indicators for fraud, public, drug, sex offenses, violent, immigration, and other offenses (property/theft as reference). Given the identification between foreigners and drug crimes, I pay particular attention to the relationship between citizenship and drug offenses throughout the analysis.

In both data sources I include a control for whether there were multiple convictions (1 = yes) and a dichotomous variable for the year of sentencing to account for any policy and legal changes between 2009 and 2010. In the ussc data I further control for whether the individual was convicted at trial.

Finally, I account for two important offender characteristics in both sources: the offender’s age and sex (male = 1). In the US data I further control for race/ethnicity with measures for black, Hispanic, and other race (white as reference) ⁵ and highest level of education. Several victim characteristics are also taken into account. The US guidelines explicitly incorporate victim enhancements into the recommended sentence for harm to “vulnerable victims” (e.g., minors). Similarly, I include a control for the number of victims under the age of 14 (ranging from 0 to 9) in the German analysis.

Analytical strategy

In both data sources I include measures for each district to capture all stable, unmeasured characteristics of court communities. Standard errors are also adjusted to account for unmeasured interdependence of cases within the same district.

Treating imprisonment and sentence length as distinct stages raises the possibility of selection bias because only a subset of offenders receives incarceration. Because most US offenders are incarcerated this form of selection is not problematic ⁶ . However, due to the more select use of imprisonment in Germany, I model these decisions simultaneously using Tobit regression alongside models of incarceration and length separately. Tobit estimation accounts for selection into the incarcerated population by assuming that the outcome measure would be normally distributed if it were fully observed. It thus accounts for both the probability that the outcome exceeds zero (i.e. that the offender was incarcerated) and the mean value of the outcome variable, adjusted by covariates.

I investigate the central question of this study by estimating the direct effect of citizenship at sentencing in US and German courts. Formally, these models are expressed as:

(1) $${Z_i} = \ln \left( {{\pi \over {1 - \pi }}} \right) = {\beta _0} + {\beta _1}(noncitizen) + {\beta _j}{X_{ij}} + {\alpha _k}$$

(2) $${Y_i} = (\hat \mu ) = {\beta _0} + {\beta _1}(noncitizen) + {\beta _j}{X_{ij}} + {\alpha _k} + {\varepsilon _i}$$

where Z _i represents the log odds of incarceration for individual i in a logistic regression, Y _i represents the logged number of months of imprisonment $(\hat \mu )$ for individual i in an ols regression for the US data, and the severity score of imprisonment in the German data; β ₁ estimates the focal independent variable—citizenship status—, X _ij represents a vector of j control variables, and α _k is a block of k – 1 federal districts (in the US) or Landgerichtsbezirk (in Germany) in the analysis. The Tobit estimators are presented as:

(3) $$Y_i^* = {\beta _0} + {\beta _1}(noncitizen) + {\beta _j}{X_{ij}} + {\alpha _k} + {\varepsilon _i}$$

where Y* is a latent variable that is observed for all values greater than 0 and censored otherwise. The observed Y is defined as:

(4) $${Y_i} = \left\{ {\matrix{ {{Y^*}\,if\,{Y^*} > 0} \cr {0\, \it if\,{Y^*} \;\le\; 0} \cr} } \right\}$$

Combined, equation 3 corresponds to the linear regression model of sentence length among those incarcerated, while equation 4 corresponds to the relevant probabilities that an individual received incarceration.

While these analyses can demonstrate the relationship between citizenship and sentencing, one shortcoming is the inability to identify the pathways through which this relationship operates. To provide a fuller and more nuanced explanation of the interpretive grounds for punishing non-state members, I draw on semi-structured interviews with judges in major cities within the United States and Germany.

Qualitative analysis

The interviews were conducted between May and September 2012 and consisted of four US Federal Court judges in one “East Coast District” and four judges in a “West Coast District” court, as well as five Landgerichte ⁷ (District Court) judges in one “West German District” (see Online Appendix C for descriptions of participant recruitment and interview methodology). The scope of the findings from these interviews merits attention. While the research sites for these interviews were chosen strategically, the findings are not meant to be general statements about the views of all judges in each country, or even all judges within these chosen districts. Rather, within this intentionally narrow domain, my goal is to use these interviews to capture the diversity and intricacy of pathways that may operate to explain the relationship between citizenship and punishment. Given this focus, the important consideration is the social position of the judges to offer strategic knowledge about the sentencing process generally, and the punishment of foreigners specifically. To this end, this analysis is well-suited.

Each research site is home to a large and diverse set of immigrant groups and few cities in either country process more noncitizens in their criminal courts. In both US districts, noncitizens account for over 30 percent of the overall docket and there are over 50 different nationalities represented. In the “West German District,” noncitizens accounted for over 20 percent of all of those punished, with over 100 different nationalities. In addition, in these districts, noncitizen groups did not strongly overlap with one particular race or ethnicity, thus providing ideal locations for discussing issues of national membership across diverse racial/ethnic groups.

While prior work has examined noncitizens using aggregate court data, a necessary limitation of this approach is the inability to capture the interpretive processes that generate sociolegal inequalities. As Blumer noted long ago, “the variable relation is a single relation, necessarily bare of the complex of things that sustain it […] When caught in its actual social character, it turns out to be an intricate and inner-moving complex” [1956: 685, 688]. The analytical strategy adopted here mitigates such oversimplification. By combining statistical analyses of court records with qualitative interviews of US and German judges, this analysis provides a considerable degree of depth to our understanding of legal inequality by making the realization of legal disparities visible in real contexts.

Establishing the punishment gap

US courts

Models 1 and 2 of Table 1 report the effects of citizenship and legal status on incarceration net of all offender, case, legal, and district controls. Models 3 and 4 report the results for the length of incarceration. Before turning to the focal variables, I examine the estimates of the control variables. Across all models, the results are consistent with theoretical expectations and prior sentencing research. Racial and ethnic minorities, men, those with low education, more severe criminal histories and longer prescribed sentences are all more likely to receive harsher penalties.

Table 1 Incarceration and sentence length models in U.S. federal courts, 2009-2010

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

Note: Standard Errors are clustered on judicial district. All models include controls for district.

Next I consider the effect of citizenship status. Consistent with theoretical expectations, non-US citizens pay a significant penalty at criminal sentencing. Starting with model 1, compared to US citizens, noncitizen offenders are five times more likely to be incarcerated. Model 2 further examines the effect of citizenship status by assessing punishment differences by legal status. The results show that both legal status groups are at greater risk of incarceration, particularly undocumented immigrants, who are over ten times more likely to be imprisoned than US citizens.

The sentence length results in models 3 and 4 show a remarkably similar pattern. Non-state members receive sentences that are approximately 8 percent longer than similarly situated citizens. On average, this corresponds to an additional 6 months of incarceration. ⁸ The pattern is also consistent across legal statuses—relative to their citizen counterparts, both groups of noncitizens are considerably disadvantaged at sentencing, with undocumented immigrants receiving the longest prison terms.

The results in Table 1 suggest that citizenship status is a consequential sentencing criterion that leads to harsher penalties. It is worth noting that the effects of national membership are on par with or exceed statuses that have dominated sociological discourse on legal inequality (e.g. race and ethnicity). Moreover, these results are not dependent on any particular offense type and are also consistent across victimless crimes. For example, the interaction effect between noncitizen status and drug offenses in the incarceration model (results not shown) is significant and positive (b = 0.81; p < 0.01), suggesting that the citizenship penalty is even more pronounced for noncitizen drug offenders. This aligns with the increasing punitiveness towards immigration in the US “War on Drugs.”

While the results are in line with theoretical predictions, several questions remain regarding the scope of the citizenship findings that speak to the salience of state membership. Most important among them: is the citizenship penalty observed for all groups lacking state membership? The next series of models addresses this question.

Table 2 reports incarceration and sentence length models using six different noncitizen categories based on nationality or region (note that both models include all controls reported in Table 1). The pattern of results in Table 2 is clear across both the incarceration (model 1) and sentence length (model 2) decisions—all noncitizen groups receive more severe punishments compared to US citizens and, with the exception of the Other European group in model 2, all effects are significant at the p < 0.05 level or higher.

Table 2 Incarceration and sentence length models for different non-citizen groups in U.S. federal courts, 2009-2010

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

Note: Standard Errors are clustered on judicial district. All models include controls for district.

Overall, the results in Table 2 show that the citizenship penalty displayed in the main models is not reserved for one particular group, but rather is applied to all groups who fall outside the national community. In short, the pattern of enhanced punitiveness suggests that it is not just about being from Mexico or from the Middle East that leads to increased punishment; it is the fact of not being a US citizen that yields disparate treatment. Together with the findings in Table 1, these results suggest that citizenship is a salient measure of legal stratification in the application of US law, even after accounting for a multitude of legal and extra-legal factors. The next section examines whether this pattern holds in Germany—a country with a markedly different conception of national membership.

German courts

Perhaps more so than the US, Germany as a case study is exceptionally well positioned to inform our understanding of the legal consequences of citizenship in the wake of ongoing globalization. As one of the original members of the European Union, Germany is considerably limited by the many protocols, conventions, and guidelines governing the legal rights and punishment of non-state members. ⁹ In addition, Germany’s legal decisions, especially those that involve foreign nationals, are subject to review by the Court of Justice of the European Union. Against this backdrop, Germany may offer an even more stringent test of citizenship as a measure of sociolegal inequality.

Similar to the US analysis, Table 3 shows the effects of multiple case and offender characteristics on sentencing. Despite markedly different legal systems, the results for the control variables are highly consistent with prior research in US courts—those convicted of more severe offenses, males, and younger offenders all receive harsher penalties. Most important for this study is the consistency in the relationship between citizenship and punishment. Net of legal controls, non-German citizens have 30 percent higher odds of receiving a prison sentence. In addition, among those who are incarcerated, noncitizens receive significantly longer prison terms. The citizenship effect is even more pronounced once the incarceration and length decisions are combined in the Tobit model (model 3). As in the US analysis, I find that the citizenship penalty is particularly acute in drug cases (results not shown). When I interact citizenship by drug offenses in the Tobit model, the result is positive and significant (b = 1.95; p < 0.001). Taken together, the pattern of results is highly consistent with those reported in the US analysis.

Table 3 Incarceration and sentence length models in german courts, 2009-2010

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

Note : Standard Errors are clustered on judicial district. All models include controls for district.

Yet, it is important to note the sizeable differences in the citizenship effect between US and German courts, particularly in the incarceration decision. Consistent with theoretical expectations, the sentencing penalty for noncitizens in the US is greater than in Germany (b = 1.62 compared to b = 0.27). This could be a reflection of the cases that come before the US federal courts which, on average, tend to handle more serious offenses compared to state courts. This appears to be part of the explanation. As I demonstrate in Online Appendix A, when I compare US federal courts to German Landgerichte courts—which handle only serious violations where the potential penalty is greater than three years of prison—the difference in the punishment gap lessens, but still remains significantly larger in the US for both the incarceration and sentence length decisions (incarceration: b = 1.62 v. b = 0.63; length: b = 0.08 v. b = 0.03). ¹⁰ While one must be cautious against drawing overly strict comparisons across samples, both the main and supplementary analyses offer suggestive evidence that the effect of national membership is lower in the German courts.

The next stage investigates the punishment of different noncitizen groups in Germany. This examination is especially important for the German case because, unlike the US, Germany is a member of the European Union which clearly demarcates member from non-member states. Thus, Germany’s EU membership raises the possibility that the postnational position could still be correct but in a more limited way. To the extent that EU citizenship has replaced national membership as the important legal status, we should observe no difference between Germans and other EU citizens, but may still observe sentencing disadvantages for those from outside the EU. The results reported in Table 4 address this possibility.

Table 4 Incarceration and sentence length models for different non-citizen groups for all offenders and drug offenders only in german courts, 2009-2010

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

Note: Standard Errors are clustered on judicial district. Models include all variables shown in Table 3 as well as controls for district.

Models 1-3 report the findings for all offense types, while models 4-6 specifically target the effects in drug cases. Though the picture of noncitizen disadvantage from the results in models 1-3 is not as clear as the US analysis, it shows very little support for the postnational position. Particularly important is the finding that offenders from other EU countries receive more severe sanctions compared to their German counterparts. According to model 1, EU members are 28 percent more likely to be imprisoned, and models 2 and 3 suggest they also receive significantly longer prison terms.

Turning to the other groups, with the exception for Turkish noncitizens, all noncitizen groups in model 1 have higher odds of incarceration than German citizens, though the effect for noncitizens from the Americas is only marginally significant (p < 0.055). Focusing on the overall effects in model 3, the picture presented is one of disadvantage for nearly all noncitizen groups and for the overwhelming majority of noncitizens in German courts. Turkish noncitizens, while comprising a sizeable group in the German courts, still only make up 23 percent of the noncitizen population. Thus, for the other five noncitizen groups that comprise over three-quarters of convicted foreigners, the overall story is one of increased punishment compared to similarly situated Germans.

Turning to the results for the different noncitizen groups in drug cases (models 4-6), the pattern is unambiguous—all those who lack German citizenship, including Turkish noncitizens, receive harsher punishment. Across all three drug models the effects are positive and significant (though the Americas effect remains marginally significant), and the effect sizes are considerably more pronounced.

Taken together, the results in Table 4 align with theoretical expectations and the empirical results presented for the US. The punitive consequences of lacking German citizenship are not reserved only for certain groups, and there appears to be no sentencing advantage to being from another EU country. Particularly in drug cases, the findings in Table 4 demonstrate that the citizenship penalty applies to all groups who fall outside the boundaries of German national membership. As with the US, it appears that the key distinction is not who you are, but who you are not. While the EU and other supra-national institutions have no doubt enhanced the legal rights of non-state members, there remains a considerable gap in the treatment of non-German citizens, one that disproportionally incarcerates non-state members and imprisons them for longer periods.

Given the markedly different approaches to crime, punishment, and citizenship in the US and Germany, the consistency of the findings suggests that the boundaries of national membership result in legal disadvantages at sentencing for non-state members regardless of how these boundaries are drawn, albeit with varying levels of severity. I attempt to uncover the mechanisms that might explain these findings through interviews with judges in the United States and Germany.

Explaining the punishment gap

The interviews revealed several themes that shed light on the mechanisms that help explain the statistical results. Prominent among these were expressions of resentment linked to the criminality of non-state members.

Resentment

Is there an annoyance because some of these criminals are basically biting the hand that feeds them? Yes [US Judge].

The criminality of non-state members engendered responses that explicitly linked citizenship status and punishment. Specifically, many of the judges resented that those who were not members of their society, especially those who entered unlawfully, would compound their status by committing crimes. This was often framed by judges in terms of violating their countries’ hospitality. As one German judge told me: “When I give reasons for the judgment I would certainly say that in some way it is objectionable to abuse the right of hospitality in order to take advantage of the benefits and then to commit crimes in addition […] if it was the case I would punish it more severely.”

Not only did these ideas resonate with other judges in both the United States and Germany, but they were often expressed as legitimate considerations. As one German judge put it:

These are people who enjoy the right of hospitality, take advantage of the benefits and don’t contribute anything towards the country that they receive it from but instead they go beyond that—they abuse the right of hospitality by committing crimes. I can relate to that and I understand that as a basis for punishment. I consider it right.

The legitimacy of this consideration was stated even more pointedly by a US judge:

I certainly can understand that as a view and, you know, there is something to be said for that; you come to my country, my birthplace, the land that I love and I’ve sworn to uphold and protect, and what do you do here? You commit crimes, and some of these crimes are absolutely devastating to American citizens. So I don’t think there’s anything wrong with that sentiment.

This quotation illustrates how resentment can be expressed as a legitimate punishment criterion and also highlights the salience of citizenship as a significant division between judges and the offender (e.g. “you come to my country…”). A German judge drew the line between herself as part of the German citizenry and noncitizen offenders even more explicitly when asked whether foreign criminals bred resentment: “That’s true. I’m quite sure about that. It would be strange if it wasn’t the case… We [judges] are also a part of society.”

Another way in which this sentiment affected specific sentencing considerations was through the consideration of mitigating circumstances. Most of the judges interviewed discussed family background as an important issue they consider at sentencing, particularly poor upbringing as a mitigating factor. As one US judge put it, “In mitigation are some of the horrible, horrible examples of just the most abused and neglected childhoods that these people have had to endure… I have to admit that I really do take that into consideration when it comes to sentencing.” However, this same judge viewed troubled backgrounds differently when it came to the punishment of foreigners, as illustrated in the following discussion:

Invariably, the Armenians and some of the Russian immigrants will tell me a story of their childhood and how oppressive it was living under some of those totalitarian governments and how they wanted to come to the United States for a much better life. Well, and I’m thinking to myself, then why are you screwing the country that’s saving you, which has rescued you…? Is there an annoyance because some of these criminals are basically biting the hand that feeds them? Yes. Does that translate into an enhanced sentence? Probably not. What it does do, it will offset, perhaps, some other mitigating factors, okay? Then I’ll consider it a wash, probably, okay? […] I do, I feel it, and it really does upset me, and I’ve said as much on the record, particularly when the plea is the terrible conditions that this person was living in… and how they’ve come to the promised land. I’m going, “Well, damn it. Why are you crapping all over the promised land?” And once again, what happens is that if you have raised some legitimately mitigating factors, it gets washed out.

Several aspects of this quote stand out. First, the annoyance that immigrants would “bite the hand that feeds them” cancels out other mitigating factors, even those that the judge concedes would otherwise be “legitimate” factors in cases involving US citizens. Second, consistent with the statistical findings, the ire over the crimes committed by noncitizens is especially targeted at undocumented immigrants. Finally, the mechanisms discussed by this judge, as well as many of those discussed above, are inextricably linked to issues of national membership. In this regard, these specific ties to citizenship differ in several respects from many of the offender statuses identified in previous literature, such as race, ethnicity, or class. Taken together, the interview data suggests resentment is a particularly salient consideration in the sentencing of non-state members.

Social marginalization and societal integration

Foreigners don’t have a permanent residence here or any social bonds so it would be harder for them to receive suspension of imprisonment compared to Germans [German Judge].

Judges in both the US and Germany raised concerns linked to foreigners’ lack of societal integration. These considerations manifested in several ways, including (1) the fear of absconding, (2) through “secondary effects” such as the foreigners’ inability to obtain gainful employment, and (3) language proficiency.

• • Absconding

Without permanent residence or employment, many judges considered noncitizens as posing a particularly high flight risk. This affected whether judges would consider suspending the sentence in favor of a non-incarceration sentence. As one German judge explained:

When somebody goes to Turkey, he can prove his identity there and stay in the country because he also has the Turkish citizenship […] of course there are less chances for the suspension of imprisonment because he has an option to disappear and stay somewhere else legally.

Interestingly, while weak social bonds within the host country worked to the disadvantage of foreign offenders, weak social bonds to their country of origin also sometimes led to more punitive sentences. This concern overlapped with the fear of absconding but for a different reason—judges understood that many noncitizen offenders immigrated as children and had very little connection to their country of origin. For some, this meant that offenders who feared deportation would have an incentive to flee the terms of their sentence. This concern was acute in the US where most federal noncitizen offenders faced deportation after their sentence. ¹¹ The following exchange with a US judge captures this dynamic:

There are so, so, so many people that are here illegally… I mean they’ve been here since they were two years old… They know nothing other than the United States… And so if you have them in your clutches, you better hang onto them because if you’re talking about sending them back to Chile… They’re in the wind.

Consistent with this view, the eventual deportation of noncitizens meant that some judges felt that they had limited sentencing options. When asked why there was a gap in the incarceration rates between citizens and noncitizens in federal courts one judge offered the following explanation:

This is custodial. That’s all it is. We’re just gonna warehouse you. Maintain custody of you until [Immigration and Customs Enforcement] gets you. Maybe that’s all there is… We’re not going to give you a noncustodial sentence, for example, even if you might give someone else probation… Because, what, we’re gonna put you on five years’ probation? Oh, by the way, sometime during the course of that probation, ice is gonna come pick you up. Yeah, good luck with that.

Issues surrounding deportation also influenced judges’ consideration of mitigating circumstances, particularly those linked to families within the host country. Especially in the United States where noncitizens are likely to be deported, judges put less emphasis on familial ties at sentencing because regardless of whether the offender is incarcerated, eventually these ties will be severed. This was described by a US judge: “it [family ties] doesn’t become less of a mitigating factor. It’s simply—if it might otherwise be a factor, it just drops out.” In other words, what might otherwise be a consideration for an offender who would be released back into the United States simply “drops out” for offenders who face deportation.

• • Recidivism

Lack of social bonds also affected the judges’ evaluations of an offender’s ability to reintegrate. For foreigners, especially those with few ties to society, judges routinely saw little chance for successful reintegration. As a US judge described, “because of their status, they don’t have a snowball’s chance in hell of going straight because they can’t get a job.” Similarly, a German judge informed me that “Germans of course are the most promising candidates of the rehabilitation facilities because they are easier to manage and have better possibilities to be integrated, in terms of therapy through occupational integration.” In these instances, most of the judges saw their primary consideration as being an evaluation of whether the offender would recidivate, and only considered their citizenship status as indirectly affecting that consideration.

• • Language proficiency

At times, considerations about the offenders’ likelihood of reintegration were intimately linked with language proficiency. This was particularly the case in Germany, where judges and defendants interact throughout the trial. ¹² For example, in discussing why Germans may receive favorable sentences, one German judge opined:

Because of the fact that the natives grow up here and get a native command of the language, on principle they will naturally have a more favorable prognosis, options for job placement on the job market or something like that.

One of the specific ways in which lacking native language proficiency disadvantaged foreign offenders was through their inability to effectively express remorse. For example, one German judge explained that offenders can “score points” when they can “convincingly express that he or she… is contrite and sorry. It has a lot to with the language and self-presentation.” However, foreign offenders are likely to be at a disadvantage because “on average the self-presentation is probably not as good if somebody isn’t speaking their own language.”

Cultural practices

Now one of the things I’ve noticed from a lot of the Hispanic illegal aliens that I find here… they don’t seems to have a great deal of problem in beating up their women, and that bothers me… It just seems to be a cultural thing I keep encountering [US Judge].

Judges in both countries discussed the role of cultural differences in the handling of foreign offenders. Such considerations affected their sentencing decisions in a number of ways. One such way was through the consideration of the defendant’s criminal history where judges attributed meaning to defendants’ behavior consistent with criminal stereotypes. In several instances, these attributions were linked in specific ways to citizenship. As the above quotation illustrates, the cultural practice of partner abuse was not perceived as a problem among minorities or even Hispanics, but rather was reserved only for Hispanic illegal aliens. This clearly played a role in deciding appropriate punishment for these offenders, as the judge explained: “So if this is your way of life, I’m probably gonna hurt you…” In Germany too, some judges thought that foreigners were more prone to violence: “based on my experience, the readiness to use a weapon is higher among foreigners… than among Germans.” It bears mentioning that in both countries descriptive evidence suggests that noncitizens were less likely to be sentenced for violent offenses and had significantly less prior criminal involvement than citizens.

Because some criminal practices were perceived as rooted in “culture,” this meant that one of the goals of punishment was to send a message to the broader immigrant community. While deterrence is one of the primary goals of sentencing for all offenders, these symbolic messages were aimed at specific immigrant groups. The following exchange illustrates:

— Interviewer: And so do you see other kinds of cultural practices that you may take into account?

— Judge: I make note of. For example, Russian immigrants, Armenians, look at the US Treasury as their personal piggy bank… It looks like half the community… And we [judges] also look at in terms of communities… I really do think about the deterrent effect that the message needs to be sent... It won’t be tolerated.

When asked about the practice of sending messages to certain immigrant groups, another federal judge viewed this as a legitimate sentencing goal but acknowledged that this may be seen as problematic: “So I can see the people who consider that important, that may be something that people are interpreting as cultural distinctions. And, you know, depending on what side you’re looking at, you can describe it in a way that is perfectly justified or in a way that seems unjustified.”

Differences in disparity

While the quantitative and qualitative data converge on the general finding that noncitizens receive harsher treatment, recall that this punishment gap appears to be lower in Germany. Theoretical work on comparative criminal punishment points to a number of features that could help explain this difference, some of which were evident throughout the interviews.

• • Views on punishment

Compared to judges in the US, German judges expressed much more lenient views on sentencing. In this vein, there appeared to be consensus that prison should be used sparingly. As one German judge told me, “I think it’s appropriate that we impose monetary penalties quite frequently and rarely give prison sentences.” Part of this was linked to the professional socialization of judges. As one judge put it: “As a rule the way we go about it is by saying: ok, if somebody didn’t get any suspended sentence until now and this is his first prison sentence that can be suspended, we can give him a chance and impose a suspended sentence. I think that this is how we normally do it here and I think that I have internalized this concept quite well. I can hardly imagine refusing somebody a chance for probation.” In stark contrast, a US judge told me, “I’ve never granted probation to any illegal entry, even those that fall in Zone A or B (referring to those zones in the sentencing guidelines that are eligible for probation)… I don’t know why not. I’ve never even thought about giving them probation, ever, not once.”

This punishment divide was even more evident when judges in Germany discussed penalties in the United States. Referencing the proliferation of “three-strikes” laws in the US, a German judge told me: “Regarding the American legal system… that someone is out after three times, I find it shocking… if it was the case here I can tell you that 90% would stay in prison for life.”

• • Individual expressions

While German judges frequently described sentencing as an application of established practices, judges in the US were more likely to discuss their individual sentencing preferences. One theme that came up repeatedly was their general antipathy towards the aim of the US sentencing guidelines of curbing their discretion: “so they get the sentencing guidelines, and it puts everybody in a straightjacket, and you’ve got to go in the grid and out and around and up up up—all of that stuff… You have to nod at [the guidelines] is what we say. Hello, guidelines. And then tell them go fuck themselves.” In a similar vein, another US judge told me, “I think the guidelines are to be considered and then thrown in the trash.” Another judge simply told me, “I hate the sentencing commission.” In other instances judges described how they disagreed with the practices set forth in the guidelines: “I’ve only had this conversation with one other judge, and we were each in agreement that we oughta just give all of the illegal entry folks six months, regardless of the guidelines.”

Personal expressions, however, were not limited to such disagreements. Other US judges described how their philosophy stands apart from that of their colleagues: “But I will tell you that if I’m not the toughest judge on sentencing, I’m up there in the top three.” It is important to note that there were no instances throughout the interviews where German judges discussed their disagreement with well-known sentencing practices or drew attention to their philosophy as distinct from other judges.

Taken together, the emphasis on proportionality and internal consistency among German judges helps explain why the court data suggests that national membership factors less in punishment decisions compared to the US.

Discussion

Despite controversy on whether citizenship is legally relevant in a world characterized by international migration, recognized human rights, and domestic legal protections for immigrants, our knowledge of the salience of citizenship under the law is limited, especially from an international perspective. With over 230 million international migrants globally and rapidly growing numbers of foreign prisoners, understanding the contemporary legal significance of state membership in criminal courts is paramount. As the first cross-national mixed methodological study of punishment and state membership, the findings from this research offer one answer as to why noncitizens are increasingly filling the prisons of the United States and Germany—they receive more severe punishment.

The interviews revealed several themes that help explicate this finding. In both countries, judges’ perceptions of social marginalization and cultural distinctiveness worked to the detriment of noncitizen offenders. In many respects, these themes align with insights from previous sociolegal research that stress the roles of inequality and criminal stereotyping in explaining racial/ethnic disparities under the law. A contribution of the current study has been to broaden the discourse on punishment and inequality by identifying the particular constellation of factors that judges consider when punishing non-state members. That different groups vary along dimensions of marginalization and dissimilarity might help explain why some judges discussed the criminality of certain immigrant groups over others (e.g. Armenian fraud offenders), and also why certain groups are more at risk for harsher treatment (e.g. foreign drug offenders).

The interviews also revealed a key influence that helps explain the salience of the citizenship effect—the resentment of immigrant criminality. Recall that the overwhelming majority of noncitizens in both countries received more severe sanctions compared to their citizen counterparts. While different groups no doubt vary in terms of their ties to the host country, financial means, and cultural distinctiveness, the sentiments of resentment apply equally well to all those who are not members of the state. In other words, any noncitizen that would compound their status by committing crimes, and thus violate the country’s hospitality, is deserving of increased punishment.

As members of the host society and also agents of the state, many judges felt that immigrants—as guests of the state—have an additional obligation to obey the law. And when they do not, judges resented their criminality and interpreted their crimes as qualitatively worse than those committed by citizens. Two factors make resentment fundamentally important for understanding the punishment consequences of state membership. First, resentment makes the link between citizenship and punishment direct. That is, the status of being a noncitizen itself breeds resentment for those who violate the law. Second, the form of resentment discussed by judges is exclusive to non-state members. While previous research—and indeed the current study—demonstrates that other offender attributes affect sentencing (e.g. race/ethnicity, sex), discussions of “hospitality” are necessarily tied to notions of state membership.

By examining the punishment costs associated with citizenship status this article not only informs on-going citizenship debates but also speaks to a number of significant themes germane to sociolegal inequality.

Revisiting the citizenship debates

By focusing almost exclusively on the policy realm, the citizenship literature has produced a sizeable gap between philosophy and practice, especially in regards to the handling of immigrants in the legal system. Looking beyond the policy making realm, this article asked whether noncitizens receive equal treatment in US and German courts despite having recognized equal legal rights. The series of empirical tests—and additional robustness checks—presented in this article suggest that lacking citizenship status results in more punitive judicial outcomes.

What are the implications of these findings for contemporary citizenship debates? At the most basic level, these results suggest that citizenship remains a significant construct under the law, and leave little doubt that citizenship is a salient mechanism of legal stratification. In this regard, the findings offer little support to the postnational argument that citizenship is legally “irrelevant.” The findings regarding the legal treatment of EU citizens in Germany are particularly informative here. If EU citizenship has now become the fundamental status of European membership, as some postnational scholars posit, it should be observed in comparisons between citizens of other EU countries. Yet noncitizens from EU countries are as equally disadvantaged in German courts as are other non-German citizens.

Taken together, the findings from this study align more with the nation-centered view of citizenship which holds that national models of citizenship continue to have important consequences for identities, forms of organization, and immigrant incorporation [Koopmans and Statham 1999]. As Calhoun [2007: 27] notes, despite the challenges of globalization to traditional notions of state sovereignty, nationalism and citizenship remain pervasive because they are widely used discursive formations that give shape to the modern world as ways of thinking and talking about the basic units of culture and belonging. This article investigated whether and to what extent citizenship was used by judges in US and German courts. The interview findings revealed the multiple and complex ways in which citizenship shaped judicial attitudes towards the criminality of foreigners and the appropriate punishment response to their transgressions. In this regard, the discourse surrounding national membership clearly demonstrates that citizenship is used to think, talk about, and act in the legal realm.

Citizenship and international punishment

This study contributes to the small yet increasingly important field of international criminological research. Despite near unanimity among criminologists that cross-national research is a major gap in the field [LaFree 2007], the comparative study of punishment is still in its relative infancy. This article adds to this nascent literature by investigating the punishment consequences of lacking state membership in the United States and Germany. The strength of this comparative approach is to test whether the relationship between citizenship and punishment holds across diverse international contexts. Because many countries have witnessed increases in both international migrants and foreign prisoners in recent decades, similar analyses with other countries or in other regions of the world would be a fruitful line of subsequent research.

However, future research in this area should be attentive to the institutional and normative environments in which legal decision-making operates. As the comparative citizenship findings shown here illustrate, the institutional setting of punishment can help to aggravate or mitigate the use of extra-legal factors at sentencing. Consistent with theoretical expectations, there was a lesser punishment gap between citizens and noncitizens in Germany, and the interview findings suggest this may be linked to more lenient normative assessments of imprisonment and a stronger emphasis on internal consistency among German judges, compared to their US counterparts. These results, however, likely only scratch the surface of the potential sources of punishment differences across international contexts, and future research amongst a larger pool of nations that explicitly examines factors such as the political insulation of the judiciary, Common vs Civil Law traditions, the bureaucratic structure of the legal profession, the production of knowledge about crime and punishment, and the legal education of judges would shed substantial light on the study of cross-national punishment severity.

Conclusion

Just as Thorsten Sellin [1935] investigated the gap between legal ideals and the legal treatment of African Americans in the US in order to raise questions regarding the substantive meaning of equality before the law, the current study examined this same gap for noncitizens on a global stage. In doing so this article used an explicitly sociological frame to ask what citizenship means for certain social actors and how this plays out in a specific institutional setting—the courts. Thus, aside from the contributions to citizenship studies and the sociology of punishment, this article takes a modest step towards a sociological study of citizenship, one interested in understanding the size, scope, and processes that produce the gap between abstract rights and rights in practice, between formal membership and substantive membership, and between the rule of law and the practice of law. With the prison playing an ever larger role in protecting the borders of the state, the results from this article suggest that, as international migration increases, the division between “insiders” and “outsiders” of the state may be a central axis of global legal inequality.