Immigration detention is the fastest-growing, yet the least-examined, type of incarceration in the United States (Reference BernsteinBernstein 2008). In 2013, U.S. Immigration and Customs Enforcement (ICE) detained a total of over 440,500 immigrants pending completion of their immigration cases (Reference SimanskiSimanski 2014: 6). Because the primary purpose of such detention is to effect the removal of a noncitizen who allegedly has violated U.S. immigration laws, immigration detention is deemed to be strictly civil or administrative, not penal. Yet, most immigrant detention facilities were originally built and currently operate as jails and prisons that confine pre-trial and sentenced felons (Reference SchriroSchriro 2009: 4). This study examines the legal attitudes of immigrant detainees using original data on long-term immigrant detainees (defined in this study as noncitizens detained by ICE for a continuous period of six months or more) held in facilities across the Central District of California. More specifically, this study addresses the following two key questions. How do immigrant detainees perceive obligations to obey the law generally and U.S. immigration authorities in particular? What is the relationship between detainees’ procedural justice judgments and their perceived obligations to obey?
Addressing these questions is important for a number of reasons. Immigrant detainees constitute a rapidly growing segment of the noncitizen population in the United States due to the developments in U.S. immigration enforcement policy in the past few decades (Reference HernándezHernández 2014; Reference RyoRyo 2016). Many, if not most, detainees are racial/ethnic minorities of disadvantaged socio-economic background who, by virtue of their precarious legal status and confinement in the quasi-criminal system, constitute one of the most stigmatized and excluded social groups in the United States. Their marginalized status and institutional confinement make it difficult, if not impossible, for the public to gain knowledge of their legal attitudes. Yet, this basic knowledge has the potential to critically inform immigration-related public discourse and policy debates, which are often based on widespread assumptions of immigrant criminality and disrespect for the rule of law and legal authorities (Reference Ewing, Martínez and RumbautEwing, Martínez, and Rumbaut 2015; Reference RyoRyo 2015).
In addition, understanding immigrant detainees’ legal attitudes may have long-term implications for domestic and international governance more generally. Past studies have shown that people's legal attitudes, once acquired and absent significant intervening conditions, are relatively stable (see, e.g., Reference BrandlBrandl et al. 1994; Reference GauGau 2010; Reference PiqueroPiquero 2005; Reference RosenbaumRosenbaum et al. 2005). Moreover, as (Reference Levitt and Nadya JaworskyLevitt and Jaworsky 2007: 130) have noted, “migrants, to varying degrees, are simultaneously embedded in the multiple sites and layers of the transnational social fields in which they live.” The transnational nature of immigrants’ social networks suggests that their legal attitudes may have broad diffusive effects. This is particularly true for immigrant detainees, all of whom must either be deported to their countries of origin or released back into their communities in the United States, following an intensive period of confinement that requires them to navigate the U.S. legal system and to interact with legal authorities on a sustained basis. Immigrant detainees thus have the potential to widely disseminate expressions of deference and trust, or cynicism and delegitimating beliefs about the U.S. legal system and authorities—not only within the United States, but also around the world.
This study contributes to two major bodies of research. First, this study advances the longstanding research on procedural justice and legal compliance. A key model that has been the focus of much empirical investigation in this research tradition is the process-based model of regulation (Reference Sunshine and TylerSunshine and Tyler 2003; Reference TylerTyler 2006a; Reference Tyler and HuoTyler and Huo 2002). This model posits that judgments about procedural justice, independent of outcome favorability, are a significant determinant of the perceived legitimacy of legal authority, which in turn promotes voluntary compliance with the law and/or cooperation with legal authority. Voluminous research—predominantly focused on citizen-police interactions and citizen-court interactions to a lesser degree—offers evidence in support of this model (for reviews, see Reference MazerolleMazerolle et al. 2013; Reference TylerTyler 2006b). One of the most well-established and commonly used measures of legitimacy is people's perceived obligation to obey the law and/or decisions of legal authority (Reference TylerTyler 2006a; see also Reference Johnson, Maguire and KuhnsJohnson, Maguire, and Kuhns 2014).Footnote 1 I follow (Reference BakerBaker et al. 2015) in construing perceived obligations to obey as a crucial concept on its own terms and assess the relationship between the detainees’ procedural justice judgments and their felt obligations to obey the law and legal authorities.
My analysis extends prior research on procedural justice and legal compliance in two key respects. Consistent with the work of (Reference Murphy, Tyler and CurtisMurphy, Tyler, and Curtis 2009), which distinguishes between “legitimacy of law” and “legitimacy of authority” (see also Reference Murphy and CherneyMurphy and Cherney 2011), I examine perceived obligation to obey the law separately from perceived obligation to obey a particular legal authority. My analysis indicates that these perceptions do not converge among immigrant detainees, suggesting that these perceptions have analytically distinct components. Moreover, this study extends an emerging body of research on procedural justice perceptions of prison inmates (see, e.g., Reference Reisig and MeškoReisig and Meško 2009; Reference Sparks and BottomsSparks and Bottoms 1995). Empirical studies on the process-based model of regulation in the incarceration context remain relatively scarce despite such studies’ critical importance given the rise of mass incarceration in the United States (Reference WesternWestern 2006). I contribute to this line of research by demonstrating that, in a closed system requiring “batch living” (Reference GoffmanGoffman 1961) such as immigration detention, vicarious experiences of unfair treatment may be as salient as personal experiences of unfair treatment. For the purposes of this study, vicarious experiences refer to learning about or witnessing other detainees’ interactions with authority.
This study also contributes to research on legal attitudes of immigrants. Prior research in this area has focused primarily on immigrants’ attitudes toward the police (see, e.g., Reference Chu, Huey-Long Song and DombrinkChu, Song, and Dombrink 2005; Reference CorreiaCorreia 2010). Although the focus on immigrants’ attitudes toward the police is critical, especially in light of the growing involvement of local police in immigration enforcement (Reference Gulasekaram and RamakrishnanGulasekaram and Ramakrishnan 2015), understanding immigrants’ attitudes toward the law and immigration authorities remains an equally pressing task. Building on recent studies that examine current and prospective immigrants’ cooperation with legal authority and compliance with immigration laws (Reference KirkKirk et al. 2012; Reference RyoRyo 2006, Reference Ryo2013), I show that there is a significant relationship between immigrant detainees’ fair treatment perceptions and their perceived obligation to obey U.S. immigration authorities. I integrate an important insight from research on immigrants’ attitudes toward the police, which suggests that immigrants’ experiences with crime and the criminal justice system in their origin countries play an important role in shaping their current attitudes toward the police (see, e.g., Reference Davis, Erez and AvitabileDavis, Erez, and Avitabile 1998; Reference Menjívar and BejaranoMenjívar and Bejarano 2004). My analysis includes an index measure of immigrant detainees’ prior experiences with the law and legal system in their countries of origin, and I find that the relationship between detainees’ evaluations of fair treatment and perceived obligations to obey remains robust.
Background on Immigration Detention
I begin by providing a brief overview of the relevant legal background for this study. A removal process begins when an immigration enforcement agent takes a noncitizen into custody.Footnote 2 The noncitizen may seek relief from removal, such as asylum or cancellation of removal (8 U.S.C. § 1229a (2012)). The immigration judge's decision on the noncitizen's application for relief may be appealed to the Board of Immigration Appeals (BIA); the BIA's decision in turn may be appealed to the federal court of appeals.
While the removal proceedings are pending, ICE may detain the noncitizen on either a discretionary or a mandatory basis. For noncitizens held under the discretionary detention provisions of the Immigration and Nationality Act (INA), ICE may release the noncitizen on conditional parole or on a bond if they do not pose a danger to society nor present a flight risk (8 U.S.C. § 1226(a)(2) (2012)). In contrast, noncitizens subject to mandatory detention are typically ineligible for release or parole pending their removal hearings.Footnote 3 Mandatory detainees include, for example, (1) certain classes of “arriving aliens,” including those seeking asylum who have not yet passed their credible fear determination, and (2) noncitizens, including lawful permanent residents (LPRs), convicted of certain crimes enumerated in the INA (8 U.S.C. § 1226(c) (2012)).
Beginning in the late 1980s, Congress enacted a series of laws, closely tied to the war on drugs, mandating the detention of a certain class of noncitizens convicted of crimes, and depriving federal immigration officials of the authority to release such detainees on bond pending their removal proceedings (Reference SayedSayed 2011). In 1996, the Antiterrorism and Effective Death Penalty Act and the Illegal Immigration Reform and Immigrant Responsibility Act significantly broadened the use of mandatory detention by casting its net over a larger class of noncitizens. Since 2007, Congress has required ICE to “maintain a level of not less than 34,000 detention beds” at any given time (Reference MorgenthauMorgenthau 2014: 5). Together, these laws and policies have produced a significant increase in the number of noncitizens detained. For example, the total number of noncitizens who entered ICE detention facilities more than doubled from a little over 200,000 in 2001 to more than 440,500 in 2013 (Reference SimanskiSimanski 2014: 5; U.S. Department of Homeland Security 2011: 3).
The length of detention varies from case to case depending on a variety of factors, including, for example, the type of detention, whether the noncitizen has sought legal relief, and the type of legal relief sought. The Transactional Records Access Clearinghouse (TRAC)'s Immigration Project estimates that in 2013, ICE detained over 30,000 individuals for three months or longer, and over 10,000 individuals for six months or longer (TRAC Immigration 2013).Footnote 4 In 2013, ICE contracted with over 244 state and county jails to house about 70 percent of the country's immigrant detainees (National Immigration Forum 2013: 4). In addition, ICE now increasingly contracts with private correctional corporations to house detainees in their private facilities (Reference TorreyTorrey 2015).
Immigration detention facilities manage the detained population in ways that are generally indistinguishable from the treatment of criminal inmates. According to a report written by the former Director of the Office of Detention Policy and Planning (Reference SchriroSchriro 2009: 4), immigrant detainees are “ordinarily detained in secure facilities with hardened perimeters in remote locations at considerable distances from counsel and/or their communities.” The report further noted: “[The facilities’] design, construction, staffing plans, and population management strategies are based largely upon the principles of command and control. Likewise, ICE adopted standards that are based upon corrections law and promulgated by correctional organizations to guide the operation of jails and prisons.” Taking this background knowledge about immigration detention as a starting point, I now develop a theoretical framework for analyzing the legal attitudes of immigrant detainees.
Theoretical Framework
I integrate insights from two major bodies of research: (1) research on procedural justice and legal compliance,Footnote 5 and (2) research on legal attitudes of noncitizens in the United States.
Research on Procedural Justice and Legal Compliance
A substantial body of research shows that people's legal attitudes can have wide-ranging behavioral consequences (Reference CohnCohn et al. 2010, Reference Cohn2012; Reference Eisner, Nivette, Tankebe and LieblingEisner and Nivette 2013). For example, studies have found that individuals who view authorities as legitimate are more likely to voluntarily defer to the law and to cooperate with legal authorities (for reviews, see Reference MazerolleMazerolle et al. 2013; Reference TylerTyler 2006b). In these studies, legitimacy has been most commonly conceptualized as “the perceived obligation to comply with the directives of an authority, irrespective of the personal gains or losses associated with doing so” (Reference TylerTyler 2006a: 27; see also Reference Johnson, Maguire and KuhnsJohnson, Maguire, and Kuhns 2014: 950).
What factors shape perceived legitimacy? The instrumental model posits that people will perceive the authority to be legitimate to the extent the authorities’ decisions favor them. In contrast, the process-based model posits that the key determinant of legitimacy perceptions is people's judgments about procedural justice (Reference TylerTyler 2006b). According to the relational theory of procedural justice, people care about procedural justice because fair treatment signals to them they are full and valued members of the group (Reference PaternosterPaternoster et al. 1997; Reference Tyler, Smith, Tyler, Kramer and JohnTyler and Smith 1999). Wide-ranging studies involving such diverse groups as the general U.S. adult population (Reference Sunshine and TylerSunshine and Tyler 2003; Reference Tyler and JacksonTyler and Jackson 2014), criminal offenders (Reference BakerBaker et al. 2014, Reference Baker2015; Reference PaternosterPaternoster et al. 1997), adolescents and juvenile offenders (Reference Carr, Napolitano and KeatingCarr, Napolitano, and Keating 2007; Reference PiqueroPiquero et al. 2005; Reference Trinkner and CohnTrinkner and Cohn 2014), immigrants (Reference KirkKirk et al. 2012; Reference RyoRyo 2013), and minorities (Reference Tyler and HuoTyler and Huo 2002; Reference Tyler, Schulhofer and HuqTyler, Schulhofer, and Huq 2010), offer empirical evidence that is consistent with the process-based model.
Are these findings applicable to incarceration contexts in which the imposition of government authority is at its zenith and the stakes for the individuals are higher given the deprivation of their personal liberty? Scholars have theorized about the importance of legitimacy and procedural justice in maintaining order in prisons (Reference JacksonJackson et al. 2010; Reference SparksSparks 1994; Reference Sparks and BottomsSparks and Bottoms 1995; Reference TylerTyler 2010). A small but growing number of empirical studies on prison inmates’ procedural justice perceptions of correctional institutionsFootnote 6 offer some evidence—albeit sometimes indirect—consistent with the process-based model. (Reference BierieBierie 2013) examines grievance processing systems in federal prisons and finds a positive and significant relationship between inmate violence and the study's two procedural justice measures: (1) the volume of late replies to prisoners’ complaints, and (2) the number of cases in which a prisoner's complaint was ignored because the complaint was deemed “irrelevant, moot, or a statement rather than a specific complaint requiring action” (Reference BierieBierie 2013: 20).
(Reference Franke, Bierie and MacKenzieFranke, Bierie, and MacKenzie 2010) examine adult inmates randomly assigned to serve their sentence at either a traditional prison or a military-style correctional boot camp in Maryland. They find that perceptions of legitimacy of the criminal justice system became more positive among those assigned to the boot camp, likely due to the camp's “procedural safeguards,” such as consistent, impartial and ethical treatment of inmates. Likewise, studies of prison inmates in other countries find a significant negative relationship between procedural justice perceptions and prisoner misconduct (Reference BeijersbergenBeijersbergen et al. 2015; Reference Beijersbergen, Dirkzwager and NieuwbeertaBeijersbergen, Dirkzwager, and Nieuwbeerta 2016; Reference LieblingLiebling 2004; Reference Reisig and MeškoReisig and Meško 2009; Reference Sparks, Bottoms and HaySparks, Bottoms, and Hay 1996; Reference van der Laan and Eichelsheimvan der Laan and Eichelsheim 2013). Taken together, these studies suggest that procedural justice evaluations might be significantly related to immigrant detainees’ perceived obligation to obey the law and immigration authorities.
Two additional issues inform my analysis. First, many studies that examine legitimacy either combine measures of perceived obligation to obey the law with perceived obligation to obey a particular legal authority (see, e.g., Reference Johnson, Maguire and KuhnsJohnson, Maguire, and Kuhns 2014: 959; Reference Sunshine and TylerSunshine and Tyler 2003: 539; Reference Tyler, Schulhofer and HuqTyler, Schulhofer, and Huq 2010: 390), or focus solely on perceived obligation to obey a particular legal authority (see, e.g., Reference JacksonJackson et al. 2012: 1066; Reference Reisig, Tankebe and MeškoReisig, Tankebe, and Meško 2014: 272; Reference TankebeTankebe 2013:116). However, (Reference Murphy, Tyler and CurtisMurphy, Tyler, and Curtis 2009) emphasize the conceptual distinction between “legitimacy of an authority” and “legitimacy of the laws” (see also Reference Murphy and CherneyMurphy and Cherney 2011). Consistent with this distinction, (Reference TylerTyler 2007: 661) has observed: “[T]he public continues to support the ideas underlying the rule of law. There are no signs that trust and confidence in these principles are declining. However, the results of public opinion polls suggest that many Americans believe that legal authorities do not actually act in accord with these ideas.” Likewise, the obligation to obey the law may reflect “diffuse support” or a “reservoir of favorable attitudes or good will” (Reference Gibson, Caldeira and SpenceGibson, Caldeira, and Spence 2003: 356) toward the law or the legal system—an overall orientation that may not be directly contingent on immediate interactions with specific legal authority. Thus, I examine perceived obligation to obey the law separately from perceived obligation to obey U.S. immigration authorities.
Second, although research on procedural justice has tended to focus predominantly on people's personal experiences with legal authority, awareness of others’ experiences with legal authority may be an important component of people's procedural justice judgment (see Tyler Reference Tyler2006a: 153). For example, a number of studies on police/citizen interactions have documented the importance of not only direct but indirect or vicarious encounters in shaping citizens’ perceptions of the police (Reference BrunsonBrunson 2007; Reference RosenbaumRosenbaum et al. 2005; Reference WarrenWarren 2011). In this literature, vicarious encounters involve “learning about other group member's police contacts” (Reference BrunsonBrunson 2007: 73). As discussed earlier, immigration detention is functionally equivalent to criminal incarceration—a closed, constant-surveillance environment whereby “a large number of like-situated individuals, cut off from the wider society for an appreciable period of time, together lead an enclosed, formally administered round of life” (Reference GoffmanGoffman 1961: xiii). In such an environment, “news travels fast and people know each others’ business” (Reference BottomsBottoms 1999: 256). Moreover, it may be commonplace for detainees to learn about their fellow detainees’ interactions with authority by personally witnessing those interactions. Thus we might expect vicarious experiences with legal authority to be as salient and impactful for immigrant detainees as their direct or personal experiences with legal authority.
Research on Legal Attitudes of Immigrants
Research on legal attitudes of immigrants in the United States (broadly construed) has followed two major lines of inquiry. The first body of research consists of qualitative studies that implicitly or explicitly examine the “legal consciousness” of unauthorized immigrants, focusing on how these individuals understand their illegality and the impact of those understandings on their identity formation, legal mobilization, and integration (Reference AbregoAbrego 2011; Reference GleesonGleeson 2010). In contrast, the second body of research focuses more directly on evaluative judgments that immigrants make about the law and legal authority, with the goal of understanding the nature and determinants of those attitudes, and their effects on cooperation or compliance with legal authority (Reference KirkKirk et al. 2012; Reference Menjívar and BejaranoMenjívar and Bejarano 2004; Reference RyoRyo 2006, Reference Ryo2013). Although these two bodies of research overlap (see Reference RyoRyo 2015), for purposes of analytical clarity, my review will focus on the second body of research.
With notable exceptions I discuss below, research on immigrants’ legal attitudes has tended to focus on the police (Reference Chu and HungChu and Hung 2010; Reference Chu, Huey-Long Song and DombrinkChu, Song, and Dombrink 2005; Reference CorreiaCorreia 2010; Reference KirkKirk et al. 2012; Reference Menjívar and BejaranoMenjívar and Bejarano 2004; Reference Wu, Triplett and SunWu, Triplett, and Sun 2012). By and large, this body of research has been concerned with legal attitudes of Asian and Latino immigrants in the United States. I highlight two key aspects of this research that inform this study. First, studies on immigrants’ attitudes toward the police generally find that these attitudes may be shaped in large measure by the immigrants’ prior experiences with crime and interactions with the police not only in the United States, but also in their countries of origin (see, e.g., Reference Davis, Erez and AvitabileDavis, Erez, and Avitabile 1998; Reference Menjívar and BejaranoMenjívar and Bejarano 2004; Reference Pogrebin and PoolePogrebin and Poole 1990). For example, based on their analysis of in-depth interviews with immigrants from (Reference Menjívar and BejaranoCuba, El Salvador, Guatemala, and Mexico, Menjívar and Bejarano 2004: 129) conclude that immigrants use a “bi-focal lens” in evaluating U.S. police authorities. The bi-focal lens refers to the comparative framework in which immigrants use their home-country experience as a point of reference in interpreting their current experiences in the United States.
Second, research on immigrants’ legal attitudes has been relatively slow to integrate research on legal compliance. To my knowledge, (Reference KirkKirk et al. 2012) is one of the first empirical studies to apply the process-based model to analyze the impact of immigrants’ procedural justice perceptions on their legal cynicism and willingness to cooperate with the police.Footnote 7 That study analyzes data from the 2002 survey of New York City residents to explore whether and to what extent the residents’ perceived injustices perpetrated by the criminal justice system shape their level of legal cynicism and willingness to cooperate with the police in immigrant communities. (Reference KirkKirk et al. 2012) find that in general, legal cynicism is less prevalent in immigrant communities than communities populated with native groups. They also find that legal cynicism varies inversely with not only perceptions of procedural justice of the police in the United States, but also with fairness of and confidence in the law in origin countries. (Reference KirkKirk et al. 2012) also conclude that legal cynicism is a powerful predictor of willingness to cooperate with the police.
My previous work (Reference RyoRyo 2006, Reference Ryo2013) also challenges the sufficiency of the instrumental model of legal compliance; but these studies, unlike (Reference KirkKirk et al. 2012), examine immigrants’ attitudes toward immigration law and immigration authorities. In addition, my previous studies focus on current and prospective unauthorized immigrants’ decisions/intentions to engage in unauthorized migration rather than cooperation/compliance with the police. The first of these two studies (Reference RyoRyo 2006) uses archival data to examine illegal border crossings of Chinese laborers during the Chinese exclusion era (1882–1943). This analysis suggests that noncompliance with U.S. immigration laws during this period was not only a product of instrumental factors, but normative factors—the widespread perceptions among the Chinese that the exclusion laws lacked social and moral legitimacy, and thus not worthy of obedience. In a subsequent study (Reference RyoRyo 2013), I apply an extension of this theoretical framework in a quantitative analysis of contemporary unauthorized migration from Mexico to the United States. My analysis of survey data from prospective unauthorized migrants from Mexico shows that these individuals’ procedural justice perceptions are significantly related to their legitimacy perceptions, which in turn is inversely related to intentions to cross the border illegally from Mexico to the United States.
I build on these studies by applying the process-based model to analyze the legal attitudes of immigrant detainees. I also extend the research on immigrants’ attitudes toward the police by shifting the focus of inquiry from attitudes toward the police to attitudes toward immigration authorities. Consistent with one of the key findings from research on legal attitudes of immigrants toward the police, which emphasizes the importance of immigrants’ legal experiences in their origin countries, my analysis of the relationship between immigrant detainees’ procedural justice perceptions and their perceived obligations to obey takes into account the detainees’ experiences with the legal system in their origin countries.
Data and Method
Data
The data for this study comes from the Rodriguez Survey, an in-person survey of long-term immigrant detainees in Southern California.Footnote 8 All respondents had received a bond hearing notice pursuant to Rodriguez v. Robbins (2013), a class action litigation in which the U.S. District Court ordered the government to provide bond hearings to noncitizens who have been held in detention continuously for longer than six months.Footnote 9 The Rodriguez class formally consists of all noncitizens within the Central District of California who: (1) are or were detained for longer than six months pursuant to one of the general immigration detention statutes pending completion of removal proceedings, including judicial review, (2) are not and have not been detained pursuant to a national security detention statute, and (3) have not been afforded a hearing to determine whether their detention is justified (Rodriguez v. Robbins 2013:Note 1).
Between May 2013 and March 2014, 565 detainees who were 18 years of age or older participated in the in-person survey. The survey was conducted as soon as practicable after the detainees’ scheduled bond hearings; as a result, all but 36 detainees (6 percent) had a substantive bond hearing at the time of the survey. The interviewers provided each eligible detainee a detailed set of information about the survey, and only those detainees who voluntarily consented to participate were surveyed. More than 92 percent of the detainees who were provided information about the survey by the interviewers completed the survey; refusal rates did not vary significantly by gender nor by country of origin. The top three countries of origin represented in the sample are Mexico (50 percent), El Salvador (21 percent), and Guatemala (12 percent). All of the interviews were conducted in English or Spanish and lasted on average about 60 minutes. The survey captures diverse information, including the detainees’ demographic background, case background, pre-detention criminal and employment history, household status and family relationships, detention experiences, health conditions, views about the law and legal authorities, and bond hearings.
At the time of the survey, the detainees were held in four facilities across the Central District of California pending their removal proceedings. These facilities are James A. Musick Facility (Musick), Theo Lacy Facility (Theo Lacy), Santa Ana City Jail (Santa Ana), and Adelanto Detention Facility (Adelanto). Approximately 23 percent of the respondents were held at Musick; 21 percent at Theo Lacy; 13 percent at Santa Ana; and 43 percent at Adelanto. Musick and Theo Lacy are county jails operated by the Orange County Sheriff's Department. Santa Ana is a city jail operated by the Santa Ana Police Department. Adelanto is operated by a private prison company called the GEO Group, and houses only immigrant detainees. ICE contracts with each of these facilities to confine immigrant detainees pending their removal proceedings. The detainees at each of these facilities must wear government-issued uniforms and wristbands with identifying information at all times. In addition, the detainees are subject to daily regimens, surveillance, and living conditions that are generally indistinguishable from those imposed on criminal inmates.
Rodriguez class members likely differ from other immigrant detainees in a number of respects. In practice, all Rodriguez class members are contesting their removability and/or seeking legal relief from removal,Footnote 10 whereas this is not the case for all short-term detainees (defined in this study as noncitizens detained by ICE for less than six months). On average, Rodriguez class members are more likely to have a criminal record than short-term detainees, as many of the Rodriguez class members may have been mandatorily detained due to their statutorily-enumerated criminal offenses. For example, based on self reports, approximately 96 percent of 565 respondents in the Rodriguez Survey had a criminal conviction; the two most common convictions were drug related and traffic related (46 percent and 43 percent, respectively).Footnote 11 Supporting Information Table A1 provides detailed information on the respondents’ criminal history. Finally, Rodriguez class members may have a higher rate of legal representation than other immigrant detainees. For example, according to a national study of removal cases decided between 2007 and 2012, only 14 percent of those detained throughout the pendency of their removal proceedings were legally represented (Reference Eagly and ShaferEagly and Shafer 2015: 32). In contrast, 38 percent of respondents in the Rodriguez Survey who had been denied bond at the time of the survey (and thus likely to remain detained throughout the pendency of their removal proceedings) reported having legal representation.
Measures
Supporting Information Table A2 provides details on the variables used in the multivariate analyses, including the survey items that measure each of the relevant concepts. Each of the items has been pretested with a subset of the study sample.
Dependent Variable
I analyze two main dependent variables: (1) perceived obligation to obey the law (generally, rather than any specific law) and (2) perceived obligation to obey U.S. immigration authorities. Following a long line of studies that have examined perceived obligations to obey (see, e.g., Reference BakerBaker et al. 2015; Reference Papachristos, Meares and FaganPapachristos, Meares, and Fagan 2012; Reference TylerTyler 2006a), I analyze the detainees’ perceived obligation to obey the law using an item that measures their level of agreement or disagreement with the following statement: “In general, people should obey the law even if it goes against what they think is right.”Footnote 12
I analyze the detainees’ perceived obligation to obey U.S. immigration authorities using an item that measures their level of agreement or disagreement with the following statement: “In general, people should accept the decisions made by U.S. immigration authorities.”Footnote 13 This item wording was adapted from one of the most common ways of measuring perceived obligations to obey legal authority in prior studies (see, e.g., Reference Papachristos, Meares and FaganPapachristos, Meares, and Fagan 2012; Reference Šifrer, Meško, Bren, Meško and TankebeŠifrer, Meško, and Bren 2015; Reference Sunshine and TylerSunshine and Tyler 2003). This item seeks to tap people's attitudes toward authorities as institutional actors generally, rather than particularized reactions to specific actions taken by individual authority figures.
Although both obligation-to-obey items have been used successfully across many different legal contexts and study populations, including criminal offenders (Reference BakerBaker et al. 2015; Reference Papachristos, Meares and FaganPapachristos, Meares, and Fagan 2012), we might be concerned that the detainees’ responses to these items may have been influenced by fears of retribution or concerns about social desirability. I am not aware of any existing studies in this area of research that have directly assessed the nature and extent of these types of issues. Nonetheless, countervailing such concerns in this study are a number of factors: (1) the interviews were conducted privately or out of the earshot of others; (2) the respondents were assured of strict confidentiality before the interviews began; (3) the instructions accompanying the obligation-to-obey items assured the respondents that there were no right or wrong answers, and that the interviewers were only interested in their opinions; and (4) it is difficult to explain why only 39 percent of the effective sample agreed with the statement that people should accept the decisions made by U.S. immigration authorities if fear of retribution or social desirability bias had been a significant issue.
Explanatory Variables
There are two main categories of explanatory variables. The first set of explanatory variables captures procedural justice evaluations. Studies show that there are two major components to people's procedural justice evaluations (Reference Sunshine and TylerSunshine and Tyler 2003; Reference TylerTyler 2009): (1) perceived fairness in decision making (whether the process is participatory, neutral, and transparent), and (2) perceived fairness in interpersonal treatment (whether the individual was treated with dignity, care, and concern). This study focuses on the latter component, drawing on four survey items.Footnote 14 These items asked the detainees whether (1) the guards and staff address and talk to them in a respectful manner, (2) they are treated as human beings, (3) they have been verbally insulted, humiliated, or threatened by a guard or staff, and (4) they have seen another detainee verbally insulted, humiliated, or threatened by a guard or staff.Footnote 15 These fair treatment items were adapted from studies of procedural justice perceptions among prison inmates (Reference BeijersbergenBeijersbergen et al. 2015; Reference HendersonHenderson et al. 2010; Reference LieblingLiebling 2004). The last of the four items captures the detainees’ vicarious experience of unfair treatment. The process-based model suggests that each of these treatment evaluations will be significantly related to the detainees’ perceived obligations to obey, controlling for other relevant factors.
The second set of explanatory variables consists of outcome-related factors that might affect the detainees’ perceived obligation to obey under the instrumental model. These outcome-related factors are measured by survey items that asked the detainees (1) whether they have been granted bond by an immigration judge, (2) whether their immigration case is before an immigration judge, and (3) length of time in detention. That a detainee's immigration case is before an immigration judge (1 = yes; 0 = otherwise) means that his or her case has not yet been denied by the judge; if the case is not before an immigration judge, the case has been denied by an immigration judge and is in an appeal process at the time of the survey. The instrumental model suggests that detainees who have been denied bond and those whose immigration cases have been denied by the immigation judge are less likely to perceive an obligation to obey. The instrumental model also suggests that the longer the detention length, the less likely the detainee is to perceive an obligation to obey.
Control Variables
The multivariate analyses include the following control variables representing the detainees’ background characteristics that may be related to their perceived obligations to obey: gender, age, race, education, English fluency, current legal status, length of stay in the United States, whether the detainee has a U.S. citizen or LPR child or spouse, pre-detention employment status, number of prior felony convictions, number of prior misdemeanor convictions, and whether or not the detainee has legal representation in his or her pending removal proceeding.
I also include a measure of detainees’ experiences with the legal system in their origin countries. Following (Reference KirkKirk et al. 2012), I draw on the World Bank's World Governance Indicators (WGI) (www.govindicators.org). There are six composite country-level indicators based on over 30 underlying databases that capture several dimensions of governance. The standard normal units for the composite indicators range from approximately −2.5 to 2.5, with higher values corresponding to better outcomes (World Bank 2015). One of these indicators, the rule of law indicator, captures “perceptions of the extent to which agents [individuals and organizations] have confidence in and abide by the rules of society, and in particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence” (Reference Kaufmann, Kraay and MastruzziKaufmann, Kraay, and Mastruzzi 2010: 4). (Reference KirkKirk et al. 2012: 88) use this rule of law indicator to measure immigrants’ perceptions about the “fairness of the legal system” in their origin countries. The WGI data is available for 1996 through 2014 (with the exception of 1997, 1999, and 2001); I calculated the average across these years for my analyses.Footnote 16
Analytical Strategy
I collapsed the response categories on some of the survey items based on theoretical considerations and a series of Wald tests to avoid problems arising from categories with too few cases and to maximize the power and parsimony of the multivariate models. My overall analytical strategy was threefold. First, I examined the univariate results on the dependent variables. Given that other studies have used the same survey item as this study in measuring the perceived obligation to obey the law, I comparatively analyzed this study's univariate result on that item and the results from some of those other studies. Second, I examined the bivariate test results to determine whether there were any significant differences between detainees who expressed an obligation to obey the law and those who did not. Likewise, I examined whether there were significant differences between detainees who expressed an obligation to obey U.S. immigration authorities and those who did not.
Next, based on my review of the bivariate test results, I performed a series of binary logistic regressions on the perceived obligation to obey U.S. immigration authorities; each model contains one of the four fair treatment perception variables. Formally, each of these models takes the form:
where log represents the natural logarithm; 
is the probability that the dichotomous outcome variable Y = 1 (i.e., respondent agrees with the statements pertaining to the perceived obligation to obey); 
is the Y intercept; βs are regression coefficients; X 1 is a vector of fair treatment perceptions; X 2 is a vector of instrumental factors; and X 3 is a vector of detainee background characteristics. Each model is a detention-facility fixed effects model that includes dummy variables for the facilities in which the detainees were held at the time of the survey.Footnote 17 As a robustness check, I re-estimated the multivariate regression models on “preprocessed” samples using coarsened exact matching (CEM), which ensures greater balance in the distribution of characteristics between key comparison groups (Reference BlackwellBlackwell et al. 2009; Reference Iacus, King and PorroIacus, King, and Porro 2012). The basic goal is to compare cases that are as identical as possible but for their responses on the main explanatory variables. Appendix A discusses the CEM procedure and the supplemental analysis results.
Results
Descriptive and Bivariate Patterns
Table 1 presents means and standard deviations of all variables used in the analyses of perceived obligation to obey the law, and perceived obligation to obey U.S. immigration authorities, respectively. The first column labeled “Total Sample” provides univariate statistics on the pooled sample. While only 39 percent of the detainees believe that people should accept the decisions of U.S. immigration authorities, 82 percent of the detainees believe that people should obey the law even if they disagreed with the law.
Table 1. Descriptive and Bivariate Statistics for Variables Used in the Regression Models
Notes: Detention facility dummy variables are not shown.
a Reference category is “Speaks English Just a Little/Not at All.”
* p < 0.05; **p < 0.01; ***p < 0.001 (two-tailed tests); n.s. = not significant at p < 0.05.
It is instructive to consider how other sample populations in the United States have answered the same survey item measuring people's perceived obligation to obey the law. I thus compare this study's univariate result on this item to results from two other studies focusing on different sample populations. In a study analyzing a cross-sectional survey of 141 known gun offenders in Chicago, (Reference Papachristos, Meares and FaganPapachristos, Meares, and Fagan 2012: 427) find that about 64 percent of the respondents agreed with the statement, “People should obey the law even if it goes against what they think is right.” In contrast, (Reference Tyler and HuoTyler and Huo 2002: 109), in their study of 1,656 non-criminal residents in Oakland and Los Angeles, California, find that 77 percent agreed with the same statement. Two sample tests of proportions indicate that the detainees in the current study were significantly more likely to agree with the relevant statement than the active offenders in Papachristos et al.'s study and the non-offenders in Tyler and Huo's study (p < 0.001 and p < 0.05, respectively). These differences across study populations are shown in Figure 1.
Figure 1. Percentage of Respondents Who Agree/Disagree with the Statement, “People Should Obey the Law Even If It Goes Against What They Think Is Right.”
Note: Figure adapted from (Reference Papachristos, Meares and FaganPapachristos, Meares, and Fagan 2012: 427). [Color figure can be viewed at wileyonlinelibrary.com]
Table 1 shows that neither the fair treatment perceptions nor the instrumental factors are significantly related to detainees’ perceived obligation to obey the law. The only demographic characteristic that is significantly related to this item is English fluency; the more fluent the detainee is in English, the more likely he or she is to agree with the statement that people should obey the law even if it went against what they thought was right. This general pattern of non-significant results is unsurprising given the relative lack of variation on the perceived obligation to obey the law. For these reasons, I do not conduct multivariate analyses on obligations to obey the law.
Table 1 shows a very different pattern of bivariate results for the perceived obligation to obey U.S. immigration authorities. Unlike instrumental factors, all of the procedural justice perceptions are significantly related to the perceived obligation to obey U.S. immigration authorities. Specifically, the detainees who agree that people should accept the decisions of U.S. immigration authorities are significantly more likely to report being treated with respect, and being treated as a human being by the guards and staff. By contrast, the detainees who disagree with the statement that people should accept the decisions of U.S. immigration authorities are significantly more likely to report having been verbally insulted, humiliated or threatened by the guards or staff, and having witnessed other detainees receiving such treatment. The two groups also differ with respect to a number of demographic characteristics. The detainees who believe that people should accept the decisions of U.S. immigration authorities are significantly less likely to have a high school education and less likely to be fluent in English; they are also significantly less likely to be LPRs and less likely to have a U.S. citizen/LPR child or spouse; finally, their length of stay in the United States, on average, is relatively shorter. To further explore these results, I turn to multivariate analyses.
Multivariate Analyses
Table 2 shows the results of four binomial logistic regressions. For ease of interpretation, all results presented are in the form of odds ratios rather than coefficient estimates. The odds ratio represents the odds of agreeing with the statement relating to the obligation to obey U.S. immigration authorities, as compared to the reference group (those who disagree with the statement). An odds ratio higher than 1 indicates an increase in the odds associated with a one-unit increase in a given independent variable. An odds ratio between 0 and 1 indicates a decrease in the odds associated with a one-unit increase in a given independent variable.
Table 2. Odds Ratios from Logistic Regression Analysis of Obligation to Obey U.S. Immigration Authorities
Notes: Robust standard errors in parentheses.
a Reference category is “Speaks English Just a Little/Not at All.”
* p < 0.05; **p < 0.01; ***p < 0.001 (two-tailed tests).
Models 1 through 4 in Table 2 contain each of the four procedural justice items, respectively. In addition, each model also contains instrumental factors, basic detainee background characteristics, and dummy variables representing detention facilities (results not shown). The effects of fair treatment perceptions, including vicarious fair treatment perceptions, are large and significant in each of the four models. Model 1 shows that detainees who report being treated with respect have 2.5 times higher odds of agreeing with the statement that people should accept the decisions made by U.S. immigration authorities (odds ratio = 2.520), which is statistically significant at p < 0.01. Model 2 shows that those who report being treated as human beings have even higher odds (3.7 times) of agreeing with the obligation-to-obey statement. Conversely, Model 3 shows that those who have personally experienced verbal insult, humiliation, or threat from a guard or staff are 51 percent less likely to agree with the statement ([1–0.490] × 100). Likewise, Model 4 shows that those who have witnessed other detainees receiving such treatment are 44 percent less likely to agree with the statement ([1–0.556] × 100). My supplemental analysis using matched samples provides similar results (see Appendix A).
In contrast to the large and significant effects of fair treatment perceptions on perceived obligation to obey U.S. immigration authorities, none of the instrumental factors are statistically significant in the four models presented in Table 2. Among the detainee background characteristics, the same two variables are statistically significant across the four models. Specifically, the odds of agreeing with the obligation-to-obey statement decrease by more than 50 percent for the detainees who are English-fluent compared to those who are not English-fluent (Models 1 through 4). The odds of agreeing with the obligation-to-obey statement decrease even more—about 60 percent—for the detainees who have a U.S. citizen/LPR child or spouse, compared to those who do not have a U.S. citizen/LPR child or spouse (Models 1 through 4). Notably, the origin rule of law variable is not statistically significant in any of the models in Table 2.
Discussion
Three notable findings emerged from my analysis of original survey data on long-term detainees. I find that the majority of detainees express a felt obligation to obey the law (over 82 percent), and do so at a level that is higher than those of other U.S. sample populations (64 percent and 77 percent, respectively). I also find that the detainees’ perceived obligation to obey U.S. immigration authorities is positively related to their evaluations of procedural justice, as measured by their assessments of fair treatment while in detention. This finding remains robust controlling for a variety of instrumental and detainee background factors, including the detainees’ experiences with the legal system and legal authorities in their origin countries. Finally, I find that vicarious procedural justice evaluations based on detainees’ assessments of how other detainees are treated are as important to their perceived obligation to obey U.S. immigration authorities as their personal experiences of fair or unfair treatment. I consider the broader implications of each of these findings in turn below.
As (Reference TylerTyler 1998: 865) has explained, “people can feel dissatisfaction about the operation of legal institutions and the actions of legal authorities without losing their feelings of obligation to obey the law.” However, as I have noted earlier, much of the research on procedural justice and legal compliance typically does not make a clear analytical distinction between people's perceived obligation to obey the law generally, and their perceived obligation to obey specific legal authorities. This study demonstrates that analyzing these two concepts separately can deepen our understanding of the complex and nuanced nature of people's legal attitudes. My analytical approach does not deny the related and convergent properties that these two types of obligations to obey might share under certain conditions. For example, it is possible that over time, perceived obligations to obey certain legal authorities may promote perceived obligations to obey the law more generally; the converse may also be true.
Contrary to popular assumptions, the majority of immigrant detainees expressed obligations to obey the law—in even greater proportions than what other studies have found among non-incarcerated populations in the United States. How should we understand this finding? The existing data does not allow me to test alternative explanations. However, it might be instructive to consider two possibilities. First, as I have highlighted earlier, the detainees in the current study are likely different from short-term detainees in a number of ways, including that the former are more likely to be seeking legal relief from removal. That the detainees are seeking legal relief and are willing to pursue such relief over the period of their long-term detention might suggest that either they are more hopeful about the outcome of their immigration cases, or are more predisposed toward accepting legal processes. Both optimism about one's legal outcome and trust in the legal system are likely to be positively related to a felt sense of obligation to obey the law.
Second, to understand why the majority of detainees expressed an obligation to obey the law in the current study, it might be helpful to test the applicability of the system justification theory in future research. According to the system justification theory, people are motivated to view their existing social system as fair and just because doing so satisfies basic epistemic, existential, and relational needs (Reference Jost and HunyadyJost and Hunyady 2005; van der Toorn and Jost 2014). Further, people who are highly dependent on the system experience an even greater need to system justify to cope with the uncertainty and threat inherent in their subordinate positions (Reference KayKay et al. 2009). Consistent with this hypothesis, (Reference van der Toorn, Tyler and Jostvan der Toorn, Tyler, and Jost 2011) find that dependence on an authority figure is positively associated with appraisals of legitimacy, measured in terms of trust and confidence in, empowerment of, and deference to, authority.
Immigrant detainees, given their complete deprivation of liberty and their precarious legal status, likely experience a high level of dependence on the existing legal system. Such dependence may motivate detainees as a group to more readily view and embrace the rule of law as the desired or aspirational governing principle, resulting in heighted perceptions of obligations to obey the law. The same dynamic may also explain my other finding that detainees with English fluency and ties to family with U.S. citizen/LPR status are less likely to express an obligation to obey. To the extent this subgroup of detainees feel less dependent and vulnerable in the legal system, their need to system justify may be proportionately less than that of detainees who lack English skills and U.S. family ties. Investigating these possible dynamics and processes will advance our knowledge of the context-dependent nature of legal attitudes among subordinated groups, particularly in incarceration contexts.
(Reference Sparks and BottomsSparks and Bottoms 1995: 60) have argued: “[E]very instance of brutality in prisons, every casual racist joke and demeaning remark, every ignored petition, every unwarranted bureaucratic delay, every inedible meal, every arbitrary decision to segregate or transfer without giving clear and well founded reasons, every petty miscarriage of justice, every futile and inactive period of time—is delegitimating.” Consistent with these observations, my analysis shows that immigrant detainees’ evaluations of fair treatment are significantly related to their perceived obligation to obey U.S. immigration authorities. This finding aligns with much of the existing research on procedural justice and legal compliance, illustrating the broad applicability of the process-based model beyond the traditional focus on citizens in policing and court contexts.
I hasten to add, however, these findings do not negate the importance of further investigating outcome-related factors. As critics of the procedural justice model have noted, under certain circumstances and for certain groups, outcome-related judgments may be more central to people's legitimacy perceptions than procedural justice evaluations (for a review, see Reference MacCounMacCoun 2005). For example, (Reference Epp, Haider-Markel and Maynard-MoodyEpp, Haider-Markel, and Maynard-Moody 2014: 5) argue in their study of investigatory police stops, “official politeness could not convert an otherwise offensive police stop into a fair and legitimate one.” My analysis includes a number of key intermediate outcome-favorability measures, but not the final outcomes of the detainees’ removal proceedings given the lack of available data.Footnote 18 Nor do I examine the detainees’ evaluations of outcome fairness, which is theoretically distinguishable from outcome favorability.Footnote 19 Systematic examinations of these kinds of outcome-related factors promise a fuller understanding of how detainees might perceive, judge, and respond to law and legal authorities.
Although this study draws on the best and the only available source of data on immigrant detainees’ legal attitudes, this study has a number of limitations. As with many if not most studies in this area of research, the data I analyze is cross-sectional and do not allow direct causal inferences about the relationship between procedural justice evaluations and perceived obligations to obey.Footnote 20 In addition, as discussed earlier, perceptions of fair treatment constitute only one component of procedural justice evaluations; this study does not examine the other major component of procedural justice evaluations—evaluations of fair decision making. Finally, this study's survey sample warrants caution in generalizing the current findings to the broader immigrant detainee population. As I noted earlier, long-term detainees may differ in important respects from short-term detainees. Recognizing these differences is especially important in this line of research because as (Reference Berrey, Hoffman and Beth NielsenBerrey, Hoffman, and Nielsen 2012: 6) have argued, “people's sense of fairness is formed through their particular experiences within the legal system and in relation to the litigants’ embeddedness in institutional contexts.”
Before concluding, I highlight three especially promising lines of inquiry for future research. First, this study demonstrates the importance of examining the role of vicarious procedural justice evaluations in shaping perceived obligations to obey. Whether, to what extent, and under what conditions such vicarious evaluations may matter for other incarcerated and non-incarcerated populations remain important questions for research on procedural justice. Second, social network theories suggest that individuals’ attitudes and beliefs can influence the normative values and behaviors of those in their immediate social networks (Reference Papachristos, Meares and FaganPapachristos, Meares, and Fagan 2012). Moreover, legal attitudes may be transmitted culturally and intergenerationally (Reference Kirk and PapachristosKirk and Papachristos 2011). Exploring the possible diffusion or contagion effects of legal attitudes will provide new insights and opportunities for future research on legal attitudes of noncitizens. Finally, given the growing convergence of the criminal justice system with the immigration enforcement system, an important task for future research is to better understand how these two systems might be working in tandem to shape the procedural justice perceptions and the legal attitudes of noncitizens.
Conclusions
Studies show that people's perceived obligations to obey the law and legal authorities have wide ranging behavioral consequences—from willingness to report crimes to law-abiding behavior (Reference BakerBaker et al. 2015; Reference Johnson, Maguire and KuhnsJohnson, Maguire, and Kuhns 2014). These studies span many different disciplines and are of interest to wide-ranging groups of policymakers and law enforcement, because these studies relate to complex problems of legitimacy, order, and power, which are central issues of governance. I advance this important area of research by exploring procedural justice judgments and perceived obligations to obey among immigrant detainees—a rapidly growing population that faces multiple forms of social, economic, political, and legal exclusion. This study's findings demonstrate the importance of nuanced analytical approaches to understanding the legal attitudes of individuals that recognize the varying social and institutional contexts in which they are embedded.
To conclude, I return to my earlier discussion on the broader implications of this study. Scholars have begun to explore how international migrants channel and facilitate cross-border diffusion of ideas, beliefs, knowledge, and practices. (Reference Pérez-Armendáriz and CrowPérez-Armendáriz and Crow 2010: 120), for example, have argued that migrants are “agents of democratic diffusion who help strengthen democracy in their countries of origin.” Immigration detention, however, might be functioning as a major institutional apparatus for the national dissemination and global exportation of delegitimating beliefs about our legal systems and legal authorities (Ryo forthcoming). This broader, transnational lens through which we may understand the significance and implications of noncitizens’ legal attitudes may be useful for evaluating long-term impacts of not only immigration enforcement policies, but also enforcement policies in many other areas of law in which noncitizens may have substantial contacts with U.S. legal systems and legal authorities.
Appendix A
Supplemental Analysis Using Matched Samples
I performed a supplemental analysis using CEM, which allows the independent and exact matching of comparison groups across multiple characteristics of interest. This “preprocessing” of the data produces a smaller sample, as unmatched observations are discarded. Regression adjustment is then used to “clean up” any residual covariate imbalance between the groups (Reference StuartStuart 2010: 13). The basic idea with matching is to approximate randomized experiments as much as possible by pairing observations that are similar or identical (on relevant “pretreatment” covariates) but for the “treatment” condition (Reference Ho and RubinHo and Rubin 2011: 20). The technical details of CEM and its advantages over other matching techniques are well documented elsewhere (see, e.g., Reference BlackwellBlackwell et al. 2009; Reference Iacus, King and PorroIacus, King, and Porro 2012).
Table A Odds Ratios from Logistic Regression Analysis of Obligation to Obey U.S. Immigration Authorities Using Matched Samples
Notes: aPre-match N = 434.
* p < 0.05; **p < 0.01; ***p < 0.001 (two-tailed tests).
Here, I used the CEM routine in Stata to produce matched samples that differ on the fair treatment perceptions but are balanced on key covariates that are related to the dependent variable (see Reference StuartStuart 2010: 6). Thus, I matched the respondents on High School Degree or More (2 groups), Speaks English Very Well/Pretty Well (2 groups), Lawful Permanent Resident (2 groups), Length of Stay in U.S. (7 groups), and Has a U.S. Citizen/LPR Child or Spouse (2 groups). L1 is an index of the degree of global imbalance across the covariates. A value of 0 on L1 indicates perfect balance between comparison groups of interest; a value of 1 on L1 indicates that no overlap exists between the two groups. As shown in Table 5 above, each multivariate L1 is reduced post-matching (reductions range from 28 to 50 percent), indicating a substantial improvement in the overall balance of the sample.
Using these matched datasets, I re-estimated the original multivariate regression models (with the same set of covariates) as presented in Table 2. Table A shows similar results as those shown in Table 2, indicating that my main findings are robust to matching.
Supporting Information
Additional Supporting Information may be found in the online version of this article at the publisher's web site:
Table A1. Type of Criminal Convictions.
Table A2. Description of Measures Used in the Bivariate and Multivariate Analyses.
