¿HOY MARCHAMOS, MAÑANA VOTAMOS? EFFECTS OF INCREASING DEPORTATIONS ON LATINA/OS’ POLITICAL ENGAGEMENT By Kelly Marie Birch A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Sociology—Master of Arts 2015 ABSTRACT ¿HOY MARCHAMOS, MAÑANA VOTAMOS? EFFECTS OF INCREASING DEPORTATIONS ON LATINA/OS’ POLITICAL ENGAGEMENT By Kelly Marie Birch Over the past twenty years, the detainment and deportation of Central American and Mexican migrants have risen exponentially in the United States, and migrants’ fear of deportation has increased accordingly. Fear of deportation (FOD) has repercussions for migrants and their families, friends, and communities, psychologically and culturally as well as economically and politically. However, few scholars have addressed the effects of deportation threats and migrant “deportability” (De Genova 2002) on Latinos’ political practices. This project unites “gendered geographies of power” (Mahler and Pessar 2001) and “deportability” (De Genova 2002) with scholarship on political engagement and social citizenship to explore the relationship between detention, deportation, and political participation using a nationally representative sample from the Pew Hispanic Center’s 2010 National Survey of Latinos (n=1,104). Latinos’ acquaintance with deportees or detainees and their experiences of discrimination increase their FOD. Consistent with previous literature, these relationships are mediated by citizenship and immigration status. Furthermore, encounters with discrimination or immigration control agents increase Latinos’ predicted odds of attending protests or demonstrations in support of immigration reform and their odds of discussing immigration policy, while FOD is predictive of attendance at immigration reform marches and protests but not talking about reform or plans to vote in the 2010 election. ii   For Peter, always. iii   ACKNOWLEDGMENTS This project would not have been possible without the help of my committee and mentors. Thank you especially to my chair, Dr. Stephanie Nawyn, committee members Drs. Soma Chaudhuri and Brendan Mullan, and mentors Dr. David Wiley and Professor Sophia Koufopoulou. Your feedback and encouragement are unmatched and invaluable. Thank you also to Drs. Tom Dietz, Cathy Liu, and Leslie Moch for your statistical and theoretical expertise. I am also blessed with brilliant and supportive colleagues. Shannon Shen and Cam Whitley, I am forever grateful for your help with Stata. To my cohort—Jess, Mellissa, Shannon, and Riva—I could not have done it with you. Thank you also to the Pew Hispanic Center, which rigorously collects and generously shares its data at no cost. Finally, to my family, you are my hope and joy. Thank you for inspiring and believing in me always. iv   TABLE OF CONTENTS LIST OF TABLES ........................................................................................................................ vii LIST OF FIGURES ..................................................................................................................... viii INTRODUCTION ...........................................................................................................................1 HISTORICAL BACKGROUND ....................................................................................................4 Early Deportations to Mexico and Central America............................................................4 Increased Post-entry Social Control, 1990 to Present ..........................................................5 Immigration Act of 1990, AEDPA, and IIRIRA of 1996 ........................................5 The Homeland Security State, 2001-Present ...........................................................7 HR 4437 ...................................................................................................................8 SB 1070....................................................................................................................8 LITERATURE REVIEW ................................................................................................................9 Citizenship ...........................................................................................................................9 Social Citizenship ..................................................................................................10 Cultural Citizenship ...............................................................................................11 Political Engagement and Participatory Citizenship..............................................12 Migrant Counter-conducts, Counterpublics, and Gendered Geographies of Power .........................................................................................................13 Deportation Fears and Effects ............................................................................................15 Deportability in the Homeland Security State .......................................................15 Effects of FOD on Families and Communities ......................................................15 Effects on Deportees and their Families ................................................................17 RESEARCH QUESTIONS AND HYPOTHESES .......................................................................18 METHODOLOGY ........................................................................................................................19 Data and Sample ................................................................................................................19 Measures ............................................................................................................................20 Dependent Variables ..............................................................................................20 Political Engagement .................................................................................20 Independent Variables ...........................................................................................21 Encounters with the Homeland Security State...........................................21 Experience with Racial or Ethnic Discrimination .....................................21 Controls ..................................................................................................................22 Immigration and Citizenship Status ...........................................................22 Demographic Variables .............................................................................22 Analyses .............................................................................................................................24 RESULTS AND DISCUSSION ....................................................................................................25 Descriptive Statistics..........................................................................................................25 Discussion ..........................................................................................................................32 LIMITATIONS ..............................................................................................................................33 CONCLUSION ..............................................................................................................................34 v   REFERENCES ..............................................................................................................................36 vi   LIST OF TABLES Table 1: Summary of Descriptive Statistics, Ordinal and Continuous ..........................................22 Table 2: Summary of Descriptive Statistics, Nominal...................................................................23 Table 3: Odds Ratios for Reporting Higher v. Lower Fear of Deportation Using Ordinal Logistic Regression, 2010 (N=1,104) ............................................................................................28 Table 4: Odds Ratios for Attending an Immigration Reform Protest or March using Binary Logistic Regression (N=1,104) ........................................................................................30 Table 5: Odds Ratios for Talking about Immigration Reform using Binary Logistic Regression (N=1,104) .........................................................................................................................31 Table 6: Odds Ratios for US Citizens’ Plans to Vote in the 2010 Presidential Election using Binary Logistic Regression (N=652) ...............................................................................32 vii   LIST OF FIGURES Figure 1: Apprehended and Returned Migrants by Region, 2009-2012 ..........................................2 Figure 2: Total Number of Removed and Returned Migrants, 1927-2012......................................6 Figure 3: Immigrants Apprehended for Criminal v. Non-criminal Reasons, 2003-2012 ................7 viii   INTRODUCTION Deportations among undocumented immigrants and legal permanent residents from the United States have risen dramatically in the past twenty years (US Department of Homeland Security [DHS] 2013). Between 1900 and 1990, approximately 20,000 migrants were removed annually from the US to their home countries. By 2005, this number had climbed to 208,000, and by 2012, removals had almost doubled to 409,849 (DHS 2013). Additionally, 2011 was the first year since 1941 during which more deportees were removed from within the interior of the US than were returned after being detained along national borders1, a trend which became still more pronounced in 20122. Furthermore, deportation patterns have been distinctly racialized, with Latinos much more likely to be deported than their African, Asian, and European counterparts (DHS 2013; Golash-Boza and Hondagneu-Sotelo 2013). Figure 1 illustrates this trend; compared with all other regions combined3 the rates of apprehension and return for Latino and Caribbean migrants are substantially higher. Additionally, while the DHS does not record the gender of apprehended, removed, or returned persons, studies by Hagan, Eschbach and Rodriguez (2008) and Golash-Boza and Hondagneu-Sotelo (2013) suggest that deportation from the U.S. is a gendered process. Men are deported at much higher rates than women. Fear of deportation has increased among immigrant communities and families with the rise in detainments and deportations for immigration-related offenses (Rodríguez and Hagan 2004; NLS 2010). The Pew Research Center’s 2013 surveys of Hispanics and Asian Americans                                                                                                                         1 “Returned” migrants are denied entry at US borders by immigration agents, while “removed” immigrants are sent to their home country on an order of removal either after an immigration hearing or after waiving their right to a hearing (DHS 2013). Because returned migrants are often apprehended near U.S. borders while removed migrants are frequently arrested within the U.S. interior, the changing ratio of removals to returns signifies an increase in interior enforcement and extended border control (Kanstroom 2007; Golash-Boza and Hondagneu-Sotelo 2013). 2 In 2012, the ratio of removals to returns was 1.824, compared with a ratio of 1.206 in 2011 and 0.808 in 2010 (DHS 2013). 3  Here “not Latin American/Caribbean” includes immigrants from Europe, Asia, Africa, Australia, Oceana, Canada and “other,” based on DHS classifications and data.   1   found that 46% of Latino respondents worried that they, a friend, or a family member would be deported compared with 16% of Asian Americans (Lopez et al. 2013). This number includes native- and foreign-born Latinos, as well as legal citizens, residents, and undocumented immigrants, although it increases to 59% among foreign-born respondents. Latino citizens’ fear of deportation may be caused by the racialization of illegal immigration as “Mexican” within the US or it may reflect their social networks. First, the racialization of illegal immigration as “Mexican” translates into risks for both native-born and naturalized Latinos, who may be stopped and asked about their immigration status because they are suspected of being undocumented due to their skin color and phenotype (Golash-Boza 2012; Gonzales 2014). As Golash-Boza notes, border patrol agents often rely on “Mexican appearance” when deciding whom to interrogate (2012:88). Additionally, Latino citizens frequently belong to networks or mixed-status families that include migrants (Jiménez 2009; Yoshikawa 2011), so their fear may stem from the relationships they form in these communities. For example, Pew’s 2012 National Survey of Latinos (NSL) found that 23% of native-born respondents personally knew someone who had been detained or deported by the federal government for immigration reasons in the past year, compared with 28% of their foreign-born counterparts (NSL 2012). Persons Figure 1: Apprehended and Returned Migrants by Region, 2009-2012 900000 800000 700000 600000 500000 400000 300000 200000 100000 0 2009 2010 2011 Aliens Apprehended - Latin American/Caribbean Aliens Apprehended - not Latin American/Caribbean Aliens Returned - Latin American/Caribbean 2012 Figure 1 was created using data from the 2012 Yearbook of Immigration Statistics (DHS 2013). 2   Despite mounting agitation surrounding detainment and deportation, few scholars have explored the effect of FOD on Latinos’ political engagement in the U.S. (Beltrán 2009 is an exception). Conversely, scholars have found that the threat of deportation has a decidedly negative impact on migrant physical and mental health (Arbona et al. 2010; Hacker et al. 2011; Inda and Dowling 2013), family relationships (Golash-Boza and Hondagneu-Sotelo 2013), access to social and educational services (Yoshikawa 2011), economic achievement (Rodriguez and Hagan 2004; Brabeck, Lykes and Hershberg 2011), and sense of belonging in the US (Beltrán 2009). Moreover, extensive deportations have intensified the risks attached to migrants’ political and civic participation. Inda and Dowling explain, “[T]he creation of insecurity among immigrants—by depriving them of the ability to participate meaningfully in quotidian life— appears to be a willful production designed to isolate this population from society and render them utterly powerless” (2013:23). This study contributes to scholarly understandings of political engagement by addressing whether or not deportation threats and encounters with the homeland security state affect Latinos’ political participation. Specifically, I use the Pew Hispanic Center’s 2010 National Survey of Latinos to investigate how FOD, acquaintance with a detainee or deportee, and personal experience with representatives of the homeland security state4 (see De Genova 2007) influence Latinos’ political engagement in the US. After a brief historical sketch of significant US migration policies and a review of relevant research on citizenship, political engagement, and deportation, and following previous scholarship (for example Flores 2003; Oboler 2006; Bosniak 2006; Beltrán 2009), I argue that citizenship is more than the provisioning of suffrage or other legal rights. Rather, it is a set of                                                                                                                         4 These variables—FOD, knowing someone who has been deported or detained for immigration reasons, and being stopped by police and asked about one’s immigration status—are indicators of being aware of or personally affected by the homeland security state in the US. The variables could not be combined due to a low alpha (0.341), but I cluster and label them as “encounters with the homeland security state” throughout the paper. 3   practices and processes that are enacted by persons who are and are not legal citizens. I ground my analysis in definitions and indicators of social, civic, and political citizenship (Marshall 1964) as well as cultural citizenship (Flores and Benmayor 1997; Flores 2003). Next, I turn specifically to political and civic engagement, also known as participatory citizenship5 (Bloemraad 2006; Brettell and Reed-Danahay 2012), to demonstrate how Latinos with and without legal status engage in formal and informal citizenship practices. I address two key themes within deportation scholarship that may influence participatory citizenship: (1) the effects of FOD on family and communities’ relationships and cohesion and (2) the effects of actual deportation and detainment experiences on deported persons and their families. Finally, I analyze the effects of FOD on Latinos’ political engagement using binary and ordinal logistic regression on the Pew Hispanic Center’s 2010 NSL. HISTORICAL BACKGROUND While the number of Latinos deported from the US has increased in recent years, a review of US immigration policy reveals a historical reliance on exclusionary border control policies to define national identity, belonging, and citizenship. In the 1880s, US federal and local governments passed laws that specifically stopped Chinese migrants from entering the US, most notably the Chinese Exclusion Act of 1882 (McKeown 2008). McKeown (2008) asserts that the act was unique because it was the first time that the US exercised its power to deny entry based on racial criteria (also see Golash-Boza 2012). Early Deportations to Mexico and Central America Throughout the Great Depression, as many as 600,000 Mexicans were returned from the US; this number included many legal US residents and American citizens (Gutiérrez 1995;                                                                                                                         5 Following Bloemraad (2006) and Brettell and Reed-Danahay (2012), I use civic engagement and participatory citizenship interchangeably. 4   Balderrama 2005). Balderrama notes that what has historically been labeled the repatriation of these citizens was actually “[u]nconstitutional deportation and coerced immigration” (2005:108). This was one of the first instances of racialized border control within the US (Kanstroom 2007). Not only were the expulsions unlawful, but they also demonstrate stereotypes about Mexican “illegality;” as the economic situation in the US worsened, Mexicans were increasingly depicted as scapegoats who carried diseases, committed crimes, displaced American workers, and were “un-American” (Gutiérrez 1995:72). Similarly, from 1954 to 1955 “Operation Wetback” aimed to remove undocumented Mexican immigrants from US soil through the widespread round up and deportation of Mexican men (Kanstroom 2007). Increased Post-entry Social Control, 1990 to Present Immigration Act of 1990, AEDPA, and IIRIRA of 1996. Deportations continued to increase gradually from 1955 into the late 1980s (see Figure 2). The Immigration Reform and Control Act, passed in 1986, offered a legalization option to migrants who had lived in the US for five years or more while it also imposed sanctions on employers who hired undocumented migrants. Four years later, the Immigration Act of 1990 expanded the definition of an aggravated felony to include previously minor offenses, increasing the pool of deportable migrants by criminalizing what had previously been considered misdemeanor offenses (Kanstroom 2007; Golash-Boza 2012). The definitional reach expanded still further in 1996 with the passage of IIRIRA and the Anti-Terrorism and Effective Death Penalty Act (AEDPA). Together, these laws increased extended border control and post-entry social control, which focus attention and resources on removing migrants from the interior of the US rather than from along its borders. IIRIRA and AEDPA expanded the crimes that rendered one deportable to include several nonviolent crimes (Golash-Boza 2012), increased the grounds for deporting legal 5   permanent residents, and instated mandatory detention for asylees while they waited for their case to be decided (Kanstroom 2007). Moreover, they were retroactive, so migrants who had previously committed an aggravated felony and completed their sentence were subject to deportation. As Figure 2 illustrates, the total number of deportees reached its peak in 2000— only four years after IIRIRA and AEPDA were passed. The laws also legalized a “fast-track deportation system” known as expedited administrative removal (Kanstroom 2007:11). Under this process, migrants have no right to an in-person hearing; are not provided with the evidence against them; are given only 10 days to respond to charges; must respond in English; and are permanently banished from the US if they are found guilty (Kanstroom 2007). These rules send migrants a clear message: they have no legal rights in the US and, consequently, do not belong there (Inda and Dowling 2013). Figure 2: Total Number of Removed and Returned Migrants, 1927-2012 2,000,000 1,800,000 1,600,000 Persons 1,400,000 1,200,000 1,000,000 800,000 600,000 400,000 0 1927 1932 1937 1942 1947 1952 1957 1962 1967 1972 1977 1982 1987 1992 1997 2002 2007 2012 200,000 Removals Returns Total Figure 2 was created using data from the 2012 Yearbook of Immigration Statistics (DHS 2013). 6   The Homeland Security State, 2001-Present. Immigration policy became more draconian after September 11, 2001. The Homeland Security Act of 2002 restructured the INS, and the Department of Homeland Security (DHS) became responsible for immigration control and enforcement. This effectively housed immigration control with anti-terrorism efforts, with the goal of removing potentially dangerous criminals through deportation (Kanstroom 2007). However, from 2009 to 2012 more migrants were apprehended for noncriminal immigrationrelated offenses than for criminal ones (see Table 3). Furthermore, while the DHS has a branch dedicated to helping migrants attain citizenship, Bloemraad argues, The question … is whether a more proactive and service-friendly atmosphere can be maintained or expanded when naturalization and immigration services are housed within a department dedicated to ‘homeland security,’ especially when security is defined largely as against foreign threats. (2006:250) Again, migrants may interpret the DHS’s title and mission as reminders that their presence is suspect and foreign, rather than feeling encouraged to participate and naturalize (Inda and Dowling 2013). Figure 3: Immigrants Apprehended for Criminal v. Non-criminal Reasons, 2003-2012 250000 200000 150000 Persons 100000 50000 0 -50000 -100000 -150000 -200000 -250000 Non-criminal Criminal Figure 3 created using data from the 2012 Yearbook Immigration Statistics (DHS 2013). 7   The Homeland Security Act, the USA PATRIOT Act of 2001, and the REAL ID Act of 2005 also greatly augmented the DHS budget, increased detentions and deportations, and enacted a record number of raids in community spaces, homes, and workplaces (Capps et al. 2007; Golash-Boza 2012). The REAL ID Act requires proof of legal residence to obtain a driver license, which greatly limits migrants’ physical mobility and autonomy. Additionally, immigration raids legitimized the growing threat of deportation for Latino undocumented immigrants throughout the US. During workplace raids, hundreds of workers are arrested simultaneously. These raids have emotional, economic, and physical consequences for arrested migrants, their families, and their communities6. HR 4437. A still more stringent bill, the Border Protection, Antiterrorism, and Illegal Immigration Control Act of 2005, sought to make undocumented immigration a felony, in addition to expanding border security measures, redefining human trafficking to include all transportation of undocumented persons in an automobile, and increasing the list of deportable crimes (Gonzales 2014). Known as the Sensenbrenner bill, HR 4437 spurred a national boycott, marches, and walk-outs during the spring of 2006. The 2006 protests successfully prevented HR 4437 from passing, though similar policies continue to be introduced. Nonetheless, the protests were a momentous display of participatory citizenship, as hundreds of thousands of Latinos and their allies mobilized to claim rights, many for the first time (Beltrán 2006; Gonzales 2014). SB 1070. In April 2010, Arizona passed SB 1070, which again sparked national protests, boycotts, and debates about immigration policy, enforcement, and racial profiling (Golash-Boza 2012; Sáenz, Menjívar and García 2013). The bill required that law enforcement officials check the immigration status of all individuals present during a lawful stop, detention, or arrest, as well as when the officers suspect that a person is breaking a law that renders them deportable. Sáenz,                                                                                                                         6 Notably, large-scale workplace raids have declined since 2008 (Golash-Boza 2012). 8   Menjívar and García (2013) argue that SB 1070 “bars undocumented immigrants (and their families) from a wide range of public spaces where their presence is visible and people are reminded that the demographic landscape of the country is changing” (2013:167). In other words, the law upholds the rights and social citizenship of white, native-born Americans while simultaneously warning Latino migrants and their communities that they do not have the right to be seen or heard publicly. The authors further assert that SB 1070 directly violates individuals’ human rights and increases Latinos’ FOD within and outside of Arizona. Like the Sensenbrenner bill and 2006 protests, SB 1070 arguably influenced the context and political incorporation of Latinos before and during the 2010 NSL. LITERATURE REVIEW Citizenship Traditionally, western thought treated citizenship as a legal status bestowed by the nation-state on its members. McKeown (2008) has demonstrated that, as white settler states struggled for the power to define national membership and control their borders, they began to categorize individuals as either aliens or citizens, a distinction that determined whether or not they were eligible for state support and public resources. Similarly, Kanstroom asserts that the contemporary deportation regime relies on “‘citizenship-as-membership’ … [in which] citizens are full members of the United States’ constitutional community, whereas noncitizens are something less” (2007:17, emphasis added). The Western nation-state assumes the right to define citizenship, who belongs, and who does not (Kanstroom 2007; McKeown 2008). This legalistic, state-based understanding treats citizenship as a dichotomous, unchanging status that renders people worthy or undeserving of certain rights and privileges. Bosniak argues that this is the basic dilemma of citizenship: 9   The idea of citizenship is commonly invoked to convey a state of democratic belonging or inclusion, yet this inclusion is usually premised on a conception of a community that is bounded or exclusive. Citizenship as an ideal is understood to embody a commitment against subordination, but citizenship can also represent an axis of subordination itself. (2006:1) To counteract static, state-centered conceptions, scholars have disaggregated citizenship into several branches, including social (Marshall 1964), political and civic (Bloemraad 2006; Brettell and Reed-Danahay 2012), and cultural (Flores 2003; Beltrán 2009). Bosniak (2000) defines citizenship as all of the above: a legal status, a set of rights, a political project, and a collective identity. These categories are not mutually exclusive and often coincide. I focus on the civic and political aspects of citizenship and ways in which participatory citizenship is affected by deportation threats. Social Citizenship. Marshall (1964) proposed that citizenship be divided into three elements: civic, political, and social. Social citizenship refers to individuals’ sense of belonging in society and their right to participate fully and equally (Marshall 1964; Grace 2013). Included in social citizenship is the right to access social welfare programs, such as the public education system in the US. This branch of citizenship is linked to political and cultural citizenship, because a sense of belonging increases one’s likelihood of civic participation (Flores 2003). Significantly, social citizenship does not always align with legal citizenship (Flores 2003; Grace 2013). Flores (2003) and others (Oboler 2006; Brettell and Reed-Danahay 2012) have argued that racial, ethnic, and sexual minorities in the US have had to struggle for social citizenship rights in ways that heterosexual white males have not. Thus, the provision of legal citizenship does not automatically guarantee social citizenship. Conversely, Bosniak (2006), Lundy (2011), and Coutin (2013) emphasize that social citizenship can exist without its legal counterpart. Bosniak (2006) notes that some social welfare 10   programs in the US are offered to persons regardless of their citizenship status, though she emphasizes that undocumented migrants’ access to these programs is constrained. Lundy (2011) found that deported 1.5 generation immigrants who migrated as minors developed a strong sense of social citizenship and belonging in the US before they were deported to countries with which were much less familiar7. Coutin’s (2013) interviews with deported 1.5 generation Salvadoran deportees offer similar findings; these forced migrants recalled “normal” childhoods in the US and experiences of voting, working, paying taxes, and speaking English fluently. Thus, social citizenship—the feeling that one belongs to a nation and deserves to participate in it fully—can exist with or without a legal status. Cultural Citizenship. Rosaldo and Flores (1997) and others (Flores and Benmayor 1997; Beltrán 2009; Cisneros 2013) use cultural citizenship to describe a combination of social and political citizenship that can be enacted by citizens, undocumented migrants, and legal permanent residents. According to Flores (2003), Latino cultural citizenship is the creation of and claim to a distinctly Latino space and rights. Rosaldo and Flores assert that Latino cultural citizenship refers to the “right to be different … with respect to the norms of the dominant national community, without compromising one’s right to belong, in the sense of participating in the nation-state’s democratic processes” (1997:57). This process is enabled by a collective memory, common stories, and a shared vision for the future, as well as solidarity amongst Chicanos and Mexican-Americans (Flores 2003). Cisneros (2013) further argues that cultural citizenship is performative, enacted through symbols and actions such as displaying a flag, participating in debates, attending public ceremonies and protests, and voting.                                                                                                                         7  For example, two of Lundy’s (2011) interviewees, Chico and Claudio, insisted that they were Americans, dressed in American fashions, and were nostalgic for American holidays like Thanksgiving.   11   Political Engagement and Participatory Citizenship. The above activities listed by Cisneros (2013) are indicators of immigrants’ political engagement in their host nation. Ramakrishnan and Bloemraad (2008) note that immigrants’ political incorporation remains understudied and that it has received scant attention in literature on immigrant adaptation as well as studies of political and civic engagement. Reed-Danahay and Brettell (2008) add that, when political incorporation is addressed, voting behavior is often the sole measure included (see also Ong and Meyer 2008). Indeed, much of the rich research on Latinos’ political engagement addresses their potential as a collective voting bloc (Pachon and DeSipio 1994), political views (Leal 2007; Fraga et al. 2012), and party affiliation (Barreto 2007; Fraga et al. 2012). However, immigration scholars have recently expanded the ways that they address migrants’ political engagement, particularly within the arenas of community organizations and ethnic associations (Vanderooy and Nawyn 2011; Brettell and Reed-Danahay 2012), comparisons between migrant groups (Mahler and Siemiatycki 2011), transnational political participation (Viramontes 2008; DeSipio 2011), and migrant claims-making (Flores and Benmayor 1997; Abrego 2011). According to Ong and Meyer, political incorporation is “the capacity to mobilize effective political action in response to perceived political opportunities in a host country” (2004:4, emphasis in the original). Bloemraad (2006) and Brettell and Reed-Danahay (2012) distinguish between legal citizenship and participatory, substantive citizenship; for Bloemraad, legal and substantive citizenship together constitute political incorporation. Much of her work concentrates on three indicators of political incorporation: citizenship acquisition, community advocacy, and immigrant groups’ success in electing political representatives from their racial or ethnic group. The range of these factors allows her to explore political citizenship without relying solely on suffrage and voting habits. Bloemraad’s (2006) project focused on naturalized 12   migrants, so her list of indicators is not entirely appropriate here. However, undocumented citizens and legal residents can participate in citizenship acquisition and community advocacy, as well as other practices such as attending public protests (see for example Ong and Meyer 2008; 2004b) and leading community organizations and campaigns (Mahler and Siemiatycki 2011). DeSipio (2011) adopts this latter approach, adding civic and community engagement and transnational political participation to Bloemraad’s list. While DeSipio notes that immigrants’ civic and political participation is generally lower than that of their native-born counterparts, he emphasizes that migrants will often become engaged “without even realizing that they are being ‘political’” by getting involved in local organizations such as parent-teacher associations and trade unions (2011:1191). Migrant Counter-conducts, Counterpublics, and Gendered Geographies of Power. Flores (2003) and Inda and Dowling (2013) label migrants’ assertions of social and cultural citizenship “counter-conducts,” activities which can range from safe, mundane activities—alerting one’s friends about migration roadblocks or carpooling—to dangerous actions like coming out as undocumented or participating in a public protest. Beltrán (2009) asserts that noncitizen migrants form “counterpublics” to offset their vulnerable status as noncitizens; they are unable to participate in electoral politics, but they enact citizenship through “active participation rather than ascriptive belonging” (2009:598). Here Beltrán (2009) refers to the 2006 immigration marches, during which migrants and their allies entered the public sphere—many of them for the first time—and became visible despite the risk of deportation. Flores (2003) and Inda and Dowling (2013) add that Latinos have formed communities that counteract discrimination and legal vulnerability in a variety of ways, practicing migrant “counter-conducts” that offset punitive migration policy and enforcement. Migrant counter-conducts include participation in 13   marches, “Know Your Rights” campaigns, and labor organizing, among others. Recently, migrant activists also have begun “coming out” as undocumented, enacting a unique and especially risky counter-conduct (Nicholls 2013). These activists assert that they are “undocumented and unafraid” in order to demonstrate that they are human beings with dignity (Nicholls 2013). However, migrant counter-conducts also can be basic strategies for survival, such as text-messaging undocumented peers with alerts about police roadblocks or being driven by citizen children to avoid being detained during a traffic stop (Inda and Dowling 2013). Beltrán (2009), Flores (2003), and Inda and Dowling (2013) emphasize that noncitizens are politically engaged agents in spite of—or perhaps due to—their vulnerable status. The authors’ works resonate with Mahler and Pessar’s (2001) “gendered geographies of power,” a conceptual model that centralizes migrants’ often conflicting social locations, such as gender, class, sexuality, ethnicity, and citizenship status. Mahler and Pessar’s (2001) framework fosters analysis of people’s social agency by accounting for multiple structural factors—such as racism, sexism, and nativism—as well as individual initiative. For undocumented migrants and their allies, counter-conducts offer a chance to assert one’s agency while maintaining awareness of and often protecting against the threats posed by deportation, nativism, and prejudice against Latinos. For example, Inda and Dowling (2013) note that protesters wore all white during some 2006 demonstrations so that documented and undocumented marchers were indistinguishable; this placed a buffer between undocumented participants and immigration control agents. Throughout this project, political incorporation will be characterized by participatory and cultural citizenship; participation in electoral politics, membership in Latino and immigrant counterpublics, and counter-conducts all will be treated as indicators of political engagement. 14   Deportation Fears and Effects Deportability in the Homeland Security State. SB 1070 and other punitive policies suggest that the US is a “homeland security state” that characterizes undocumented migrants as “illegal” and “deportable” (De Genova 2002; 2007). The fear of migration and terrorism have shaped the US into such a state, in which the polity, economy, and civic sector converge to advocate stringent migration policies that theoretically protect the nation from destructive outside forces (De Genova 2007; Gonzales 2014). Within a homeland security state, the militarization of national borders, detainment, and deportation of migrants are accepted as necessary components of a safe society. De Genova (2002) and Golash-Boza (2012) add that the DHS does not expect or even hope to deport all illegal immigrants from the US. Rather, the goal of deportation is to remind migrants that their presence in the US is temporary, deportable—in effect, disposable. Deportability causes migrants to remain a vulnerable group and a malleable, obedient workforce (De Genova 2002). This adds another important dimension to migrants’ geographies of power and subsequent political participation. As Mahler and Pessar (2001) assert, migrants’ social agency is constrained not only by substantive factors but also by cognitive processes. In this case, migrants’ imaginary, and hence their ability to make political claims and participate meaningfully in the political system, is greatly influenced by their conception of themselves as deportable. Effects of FOD on Families and Communities. Fear of deportation has tangible negative effects on Latino individuals, families, and communities. First, FOD heightens Latino migrants’ stress level (Arbona et al. 2010) and decreases their mental and physical health (Hacker et al. 2011; Yoshikawa 2011; Saénz, Menjívar and García 2013). Significant for this study, Hacker and colleagues (2011) have revealed that migrants’ fear sometimes prevents them from seeking 15   necessary health services; in other words, FOD reduces their ability to enact social citizenship through their decreased access to available resources. Latinos’ limited access to social services is especially troubling among mixed-status families8, who face unique challenges and increased fear due to the threat of deportation (Hagan, Rodríguez and Castro 2011). Nearly one in every three children of immigrants has at least one undocumented parent (Yoshikawa 2011), which may explain why many Latinos express FOD for themselves, a family member, or a friend, even as legal US citizens. Yoshikawa (2011) emphasizes the paradox of mixed-status social citizenship: citizen children of undocumented immigrants are eligible for social programs and resources offered by the same government that may deport their parents. When parents are too afraid of deportation to claim these resources, their children’s cognitive skills lag behind those of their peers. Yoshikawa’s study of young children with noncitizen parents reveals that children’s development is negatively affected by their parents’ status as early as two years old. Undocumented parents’ psychological distress, coupled with their limited resources and economic instability, translate into cognitive—and therefore participatory—losses for their children. Similarly, Abrego and Menjívar (2011) have demonstrated that immigrant mothers’ relationships with their children are affected by deportation-related stress, and Brabeck, Lykes and Hershberg (2011) have found that 44% of parents reported that FOD caused their children to experience academic problems, depressive symptoms, anxiety, and poor behavior. Additionally, the workplace raids that occurred until 2008 affected immigrants’ children. After raids, many families hid in their homes for days or weeks, which increased children’s stress levels and disrupted their learning and school schedule (Capps et al. 2007). These findings together suggest that deportation threats limit Latino                                                                                                                         8  Mixed-status families are those in which family members hold different citizenship statuses, such as legal citizenship, permanent or temporary residency, and/or undocumented status.   16   immigrants’ ability to enact social and cultural citizenship: not only are they denied social citizenship through decreased access to social welfare programs, but they also are less able to claim space and rights, key components of cultural citizenship (Flores and Benmayor 1997). Immigrant communities also have been affected by migrants’ FOD (Rodríguez and Hagan 2004; Hagan, Rodríguez and Castro 2011). Rodríguez and Hagan’s (2004) interviews with community leaders and social service providers revealed that these individuals experienced increased stress as deportations rose, despite their own secure citizenship status. The respondents worried that their clients and community members would withdraw from schools and healthcare programs to avoid deportation. Hagan, Rodríguez, and Castro (2011) additionally note that some migrants’ fear caused them to socially withdraw from their community, so local businesses have lost Latino customers and report decreasing sales. Clearly, FOD affects not only noncitizens but also their communities and loved ones. Effects on Deportees and their Families. As stated above, deportees and detainees serve the function of upholding non-deported migrants’ awareness of their deportability (De Genova 2002). Deportees and their families also are the group most directly affected by US deportation policy, even when families include US citizen children or spouses. Within the current genderedracial removal regime, deportations cause family separations in which women and children are left in the US and their husbands and fathers are returned to Central America or Mexico (GolashBoza and Hondagneu-Sotelo 2013). When men are deported, families often lose economic and emotional support in the forms of a breadwinner and father figure9. Some families even begin receiving financial support from relatives in their country of origin (Golash-Boza and                                                                                                                         9 There are exceptions to this rule, of course; some women and children are deported, and others may be single mothers and/or breadwinners before their partner is removed from the US. I follow Golash-Boza and HondagneuSotelo’s (2013) analysis, which reflects overall trends. Further research is needed to show the effects of FOD on gay, lesbian, bisexual, and transgender migrants, as well as deported women. 17   Hondagneu-Sotelo 2013). Additionally, when migrant parents are suddenly detained or deported, their children may be left alone or with relatives who cannot afford to care for them (Capps et al. 2007). Capps and colleagues (2007) found that parental detainment or deportation after workplace raids caused children and other family members psychological, emotional, and economic hardship. Children lacked supervision, became increasingly socially isolated, and felt abandoned by their parents (Capps et al. 2007). When deportees have the option to bring their children with them, they often must choose between family unification in a foreign country or upholding their children’s citizenship rights in the US (Kanstroom 2007). In either case, children and adults’ lives and feelings of belonging to the US change dramatically. These studies demonstrate definitively that deportation fears and experiences frequently disrupt Latinos’ lives, whether they are undocumented migrants, permanent residents, or citizen members of mixed status families or communities. Arguably, political participation is affected as well. This paper adds to scholarship on migration and deportation by demonstrating the effects of FOD on political engagement and substantive citizenship. I hypothesize that Latinos’ political engagement and practices are affected by FOD and encounters with the homeland security state, as represented by police officers and border control agents. RESEARCH QUESTIONS AND HYPOTHESES I seek (1) to explore key predictors of Latinos’ FOD and (2) to investigate whether political participation is affected by FOD, acquaintance with detainees or deportees, experiences of discrimination, and encounters with representatives of the homeland security state. Specifically, I test the following hypotheses: H1: Controlling for migration and citizenship status, respondents’ odds of fearing deportation will increase if they know a detainee or deportee; if they have been stopped 18   by police and asked about their immigration status; or if they report experiences of racial or ethnic discrimination. H2: Encounters with the homeland security state and experiences of racial discrimination will increase Latinos’ odds of attending a protest or demonstration in support of immigrant rights. H3: Similarly, encounters with the homeland security state and experiences of racial discrimination will increase Latinos’ estimated odds of talking about the immigration policy debate. H4: Conversely, increased encounters with the homeland security state and experiences of discrimination will decrease Latinos’ odds of planning to vote. METHODOLOGY Data and Sample This study uses the Pew Hispanic Center’s publicly accessible 2010 NSL, which was conducted from August 17 to September 19, 2010. The year 2010 was a significant year for U.S. immigration policy, specifically due to SB 1070, which was introduced to the Arizona State Senate in January 2010 and signed into law by April 23 of the same year. The policy change received national press coverage and increased FOD among Latinos throughout the US (Sáenz, Menjívar and García 2013). Over 90% of the 2010 NSL respondents were aware of SB 1070, 79% of whom reported that they disapproved of it. All interviewers for the NSL were bilingual, and respondents had the option of completing the interview in Spanish or English. The annual survey employs random digit dialing via cell phone and landline and includes respondents who self-identify as Latino or Hispanic and report an age of 18 or older. When weighted, the data provided are nationally 19   representative of the US Latino population. The initial 2010 sample contained 1,375 participants, but 174 cases were dropped due to missing data. This was unavoidable, due to the sensitive nature of the survey questions, which addressed immigration and citizenship status, as well as political attitudes. Missing cases for income and education were imputed with mean scores to avoid losing data points. Puerto Rican respondents (n=97) were removed from the sample, because their citizenship status offers atypical access to and denial of political and social rights, such as suffrage and freedom from deportation (Fraga et al. 2012). The resulting sample included 1,104 respondents, approximately 80% of the original. However, the sample’s demographics were not substantively different before and after cases were dropped. Measures Dependent Variables Political Engagement. Political engagement was measured using (1) attendance at a protest or demonstration in support of immigration reform; (2) participation in discussions about immigration policy with family, friends, or coworkers; and (3) plans to vote in the 2010 presidential election. These three items were included to signify a range of political engagement, from safer, undemanding activities (i.e. talking about immigration policy), to time-consuming and potentially risky actions (i.e. attending a public demonstration). The three variables allow for activities accessible to undocumented persons and legal residents, as well as those in which citizenship is required for participation; only citizens can vote, but talking about the election can happen inside anyone’s home. Attendance at a protest or demonstration, discussions of immigration policy, and plans to vote in the 2010 election were coded as dummy variables, where 1 indicated affirmative answers. The majority of respondents (64.1%) had discussed immigration reform, while fewer planned to vote (46.4%) or had attended a protest (16.6%). 20   Independent Variables Encounters with the Homeland Security State. Encounters with the homeland security state were measured with three indicators: (1) fear of deportation (FOD), (2) acquaintance with someone who has been deported or detained for immigration reasons in the past year, and (3) personal experience being stopped by police or other authorities and asked about one’s immigration status in the past year. These indicators had low internal consistency (alpha=0.341), so they are treated separately. Fear of deportation includes fear for oneself, a family member, or a close friend. It is an ordinal scale ranging from 1 (“not at all”) to 4 (“a lot”), with a mean of 2. 258 for citizens and 3.174 for non-citizens. Acquaintance with a deportee or detainee and having been stopped and asked about immigration status are dummy variables, in which 1=yes. Approximately 33% of all respondents knew someone who had been detained or deported, while only 5.7% reported being stopped and asked about their immigration status. When divided by citizenship status, non-citizens were more likely to know someone who had been detained or deported (roughly 43% to 27%, respectively) but not substantially more likely to have been stopped and asked about their immigration status (6.1% to 5.5%, respectively). For all variables, higher scores represent greater awareness of or experience with the homeland security regime. Experience with Racial or Ethnic Discrimination. Respondents also were asked whether they themselves, a close friend, or a family member had experienced racial or ethnic discrimination in the past five years. This was treated as a dummy variable; respondents who reported direct or indirect experience of discrimination were coded as 1 and those who did not report these experiences were coded as 0. 34.1% reported experience with discrimination. 21   Table 1: Summary of Descriptive Statistics, Ordinal and Continuous Variable U.S. Citizens Non-citizens Weighted Standard Weighted Standard Mean Dev. Mean Dev. Fear of deportation (1-4, least to most fear) 2.258 1.261 3.174 1.110 Political Ideology (1-5, conservative to 2.957 1.001 3.045 0.946 liberal) N=324,866 N=179,927 Controls Immigration and Citizenship Status. Respondents’ immigration status was coded with three dummy variables: first-generation, second-generation, and third-plus generation. Firstgeneration migrants include all foreign-born respondents; 61.2% of the sample reported being born in a country other than the US or Puerto Rico. To further disaggregate immigration status, I defined second-generation migrants as all those native-born US Americans who reported that their mother, father, or both of their parents were born in a country other than the US or Puerto Rico. Approximately 21% of native-born respondents had at least one parent born abroad. Native-born participants whose parents were both born in the US were coded as third-plus generation migrants. Throughout this study, third-plus generation is treated as the reference category. Additionally, citizenship status was coded as a binary variable, where 1=citizen and 0=noncitizen. 64.4% of participants reported that they have US citizenship; this includes both native-born and naturalized US Americans. Demographic Variables. Several confounding variables were controlled for, based on migration theory and previous research. Initially, confounding variables included sex, family heritage, language preference, educational attainment, income, age, marital status, parent status, and homeownership. Based on DHS data (2012), I grouped respondents’ family heritage into two 22   categories: Mexican/Central American and Other10. Similarly, marital status was treated as series of dummy variables, including married, partnered, previously married, and never married. Married was the reference category. Language preference was included because limited English proficiency has been a central barrier to immigrants’ civic and political engagement (Ramakrishnan and Viramontes 2006). Language was classified based on respondents’ ability to read, write, and converse in Spanish and English. Respondents who could read, write, and speak Spanish and English equally were coded as bilingual, while those who reported more ease in English were considered English-dominant and those who reported more ease in Spanish were considered Spanish-dominant. Spanish dominance was used as the reference group. Sex, parenting status, and homeownership were all treated as dummy variables (1=female; 1=parent or guardian of a child under the age of 18; and 1=homeowner). Education, income, and age were treated as ordinal variables, coded in ascending order. Education was divided into five categories (1=middle school or less, 2=less than high school, 3=high school, 4=some college, and 5=college degree or beyond). Similarly, total family income was broken into nine groups, ranging from “less than $10,000 annually” to “$150,000 or more.” Age was separated into four categories (18-29, 30-49, 50-64, and 65 and older). Political ideology was measured with a 5-point Likert Scale from very conservative to very liberal. Variable Table 2: Summary of Descriptive Statistics, Nominal Weighted %: Citizens (64.36%) Weighted %: Noncitizens (35.64%) Sex Female Male Immigration Status First generation migrant 51.76% 48.24% 47.74% 52.26% 39.76% 100%                                                                                                                         10 The DHS Immigration Yearbook (2012) illustrates that Mexicans and Central Americans have the highest rates of deportation from the US. For this reason, I consider them at greater risk for deportation than their South American, Cuban, and Caribbean counterparts. 23   Table 2 (cont’d) Second generation migrant Third+ generation migrant 32.55% 27.69% 0% 0% 27.36% 5.51% 34.94% 68.36% 72.03% 13.87% 64.31% 43.10% 43.28% 6.10% 32.68% 0% 0% 21.39% 63.70% 74.78% 21.81% 30.38% 47.82% 73.98% 0.38% 25.64% 11.66% 12.42% 28.83% 22.09% 25.00% 29.20% 14.82% 34.51% 11.50% 9.96% 8.28% 13.50% 13.50% 28.68% 9.51% 9.82% 7.36% 5.83% 3.53% 13.27% 24.56% 19.03% 30.31% 5.53% 3.54% 1.99% 1.33% 0.44% Immigration Attitudes and Political Engagement (Binary measures) Knows someone who has been deported Has been stopped by police & asked about immigration status Has experienced racial or ethnic discrimination Registered to vote Plans to vote in the 2010 election Has attended a march or protest for immigration reform Has talked about immigration policies Fears Deportation (Some or a lot) Language Preference Spanish English Bilingual Education Middle school or less Less than high school High school graduate Technical, trade, vocational school, or some college College graduate or beyond Annual income Less than $10,000 $10,000-19,999 $20,000-29,999 $30,000-39,999 $40,000-49,999 $50,000-74,999 $75,000-99,999 $100,000-149,999 $150,000 and above Analyses Ordinal and binary logistic regression models tested the effects of Latinos’ experience with and awareness of the homeland security state on political engagement. The effects of knowing a deportee or detainee, being stopped and asked about one’s immigration status, and experiencing racial discrimination on respondents’ FOD were examined using ordinal logistic regression. Binary logistic regression models tested the effects of encounters with the homeland security state and experiences of discrimination on (1) previous participation in a protest or 24   demonstration for immigration reform, (2) discussions of immigration policy with friends, family members, or coworkers, and (3) respondents’ plans to vote in the 2010 election. Progressive adjustment was used throughout to explore the relationship between Latinos’ encounters with the homeland security state and their odds of participating in a protest or demonstration, discussing immigration policy, and planning to vote. Using progressive adjustment permits the removal of several confounding variables that did not affect these main effects, including age, family heritage, marital status, parent status, and homeownership11. Furthermore, for each hypothesis, likelihood-ratio (LR) and AIC/BIC tests were used to ensure that the final models improved the fit of the earlier models. All final models were improvements over earlier counterparts. AIC and BIC scores are reported in Appendix II. RESULTS AND DISCUSSION Descriptive Statistics Tables 1 and 2 disaggregate the sample into citizens and noncitizens. The noncitizen population includes all foreign-born who have not yet naturalized, and the citizen population includes all native-born US Americans and those immigrants who have become naturalized citizens. Table 1 demonstrates that noncitizens have substantially more FOD (3.174 on a scale from 1 to 5) than their US citizen counterparts (2.258). This is even more evident when FOD is combined to include “Some” or “A lot” of fear; 43.10% of citizens report some or a lot of fear, compared with 74.78% of their noncitizen counterparts. This is unsurprising because noncitizens face a direct risk of deportation, rather than a more generalized fear for one’s friends or family members. Many more noncitizens also knew someone who had been deported or detained compared with citizens (43.28% of noncitizens compared with 27.36% or citizens).                                                                                                                         11 These variables were found to have a negligible effect within these models. However, previous research suggests that they may have a qualitative impact. Parenthood status especially affects FOD. Further research should explore these relationships to test their substantive as well as statistical significance. 25   Significantly, however, citizens and noncitizens’ rates of being stopped and asked about their immigration status (5.51% and 6.10% respectively) and experiencing racial or ethnic discrimination (34.94% and 32.68%) are much more comparable. This supports De Genova (2002; 2006) and Golash-Boza’s (2012) argument that migration is racialized as “Mexicanness” in the US. Approximately 21% of noncitizens reported attending protests or marches for immigration reform, compared with 13.87% of citizens, which is indicative of noncitizens’ participation in counter-conducts and cultural citizenship performances (Beltrán 2009). Citizens also had relatively high rates of voter registration (68.36%), and over 70% of citizens reported that they planned to vote in the 2010 election. Self-reported rates of voter registration and turnout are usually inflated (see Fraga et al. 2012), but these findings corroborate Pachon and DeSipio’s (1994) assertion that Latinos are engaged in politics and register to vote at high rates. Table 2 also corroborates established trends among Latino migrants to the US. The weighted sample confirms that noncitizens are slightly more likely to be male, much more likely to speak Spanish and have lower incomes and educational attainment than their citizen counterparts. Roughly 75% of noncitizens report being Spanish-dominant, compared with only 22% of their citizen peers. Less than 1% of noncitizens report English-dominance. 21.46% of noncitizens report some college or a college degree, compared with 47.09% of citizens. Furthermore, 24.08% of citizens and 44.02% of noncitizens report less than a high school education. These numbers are likely mediated by age (were migrants too old to attend school when they arrived in the US?) and income (were participants required to begin working immediately to provide for themselves and/or their family members?). However, the numbers also suggest that a substantial minority of Latinos—citizens and noncitizens alike—face barriers that limit their access to public education, a key indicator of social citizenship. Noncitizens also 26   report lower income rates than their citizen peers. For example, 57.13% on noncitizens report an income of less than $30,000 annually, compared with 35.28% of US citizens. H1: Controlling for migration and citizenship status, respondents’ odds of fearing deportation will increase if they know a detainee or deportee; if they have been stopped by police and asked about their immigration status; or if they report experiences of racial or ethnic discrimination. Hypothesis 1 is supported (see Table 3). Knowing someone who has been detained or deported is a statistically and substantively significant predictor of FOD; in Model 1, the predicted odds of reporting higher fear of deportation were 2.934 times higher for respondents who reported knowing a deportee or detainee, holding consistent all other covariates. The effect of experiencing racial discrimination also increases respondents’ predicted odds of reporting higher levels of deportation fear. The odds of having a higher score on the FOD scale were 39.5% higher for those who reported discrimination compared with those who did not, holding consistent all others. Conversely, being stopped and asked about one’s immigration status is not a significant predictor of FOD (p=0.644), holding all others constant. Additionally, immigration and citizenship status, income level, and language preference all significantly predict FOD. US citizens are less likely to report FOD than their non-citizen counterparts, while immigrants are much more likely to report high levels of deportation fear. Model 2 reveals that being a first-generation migrant increases a respondent’s predicted odds of reporting higher levels of fear by 5.981 times, holding constant all other covariates. This was expected, especially given the high number of noncitizens in the sample (35.64%). Secondgeneration migrants are also more likely to express FOD than third-plus generation participants, 27   even with the addition of confounders. Again, this may be a function of mixed-status families or social networks, in which the FOD is for family members or friends, rather than one’s self. Table 3: Odds Ratios for Reporting Higher v. Lower Fear of Deportation using Ordinal Logistic Regression, 2010 (N=1,104) Model 1 Model 2 Model 3 Knows someone who has been deported 2.934*** 2.573*** 2.676*** Has been stopped and asked about status 1.142 1.251 1.146 Has experienced racial discrimination 1.395** 1.617*** 1.735*** U.S. Citizen --0.529*** 0.645** First generation immigrant --5.981*** 3.873*** Second generation immigrant --2.670*** 2.261*** Sex-female ----1.116 Income ----0.851*** Education ----0.914 Language - English ----0.530** Language – Bilingual ----0.888 Log Likelihood -625,801.38 -577,142.35 -565,299.33 Pseudo R-square 0.036 0.111 0.129 AIC 1,251,615.0 1,154,303.0 1,130,627.0 BIC 1,251,645.0 1,154,348.0 1,130,697.0 Two-tailed tests: ***p<0.01; **p<0.05; *p<0.1 The relationships between FOD and acquaintance with a deportee or detainee, having been stopped and asked about one’s immigration status, and experience with discrimination do not change drastically with the addition of confounders (sex, citizenship, immigration status, marital status, parent status, homeownership, income, education, and years spent in the US12). Holding all other covariates constant, the strength of the effect of knowing a deportee or detainee changes slightly, but it retains its statistical significance. Moreover, the effect of experiencing discrimination becomes stronger with each model: in Model 2, discrimination experience increases one’s estimated odds by 61.7% and in Model 3, it increases one’s estimated odds by 73.5%, holding all others constant. Being stopped and asked about one’s immigration status                                                                                                                         12 For the sake of parsimony, marital status, parent status, homeownership, and years spent in the US do not appear in the final models. They were tested based on their substantive significance in previous research; however, they do not change the main effects. 28   remains insignificant regardless of controls, however. Neither is sex a significant predictor of FOD; in other words, women and men report similar levels of FOD. H2: Encounters with the homeland security state and experiences of racial discrimination will increase Latinos’ odds of attending a protest or demonstration in support of immigrant rights. Hypothesis 2 is supported (see Table 4). Greater levels of deportation fear and other encounters with the homeland security state increase one’s predicted odds of attending an immigration reform protest or demonstration. In Table 4, Model 1 indicates that knowing a deportee or detainee, being stopped and asked about one’s status, and experiencing discrimination all significantly increase a respondent’s odds of attending an immigration reform protest or march. Similarly, Model 2 reveals that FOD increases one’s odds of attending a march by 55.7%, again a statistically significant effect. Consistent with previous studies, sex and language preference also significantly predict odds of attending a protest or march for immigration reform. The predicted odds of attending a march are 36.2% percent lower for women compared with their male counterparts, accounting for all other covariates. English-dominant respondents also are much less likely to attend a march than their bilingual and Spanish-dominant counterparts. Conversely, immigration status, citizenship, income, and education do not affect respondents’ odds of participating in a march for immigration reform. These findings indicate that encounters with the homeland security regime may supersede other factors. On the other hand, these numbers also may reveal some level of autocorrelation; noncitizens are more likely to report FOD and know someone who has been detained or deported than their citizen peers. However, the comparable rates of being stopped 29   and asked about one’s immigration status and experiencing racial or ethnic discrimination (see Table 2) suggest otherwise. Qualitative, in-depth research is needed to fully unpack these results. Table 4: Odds Ratios for Attending an Immigration Reform Protest or March using Binary Logistic Regression (N=1,104) Model 1 Model 2 Model 3 Knows someone who has been deported 2.610*** 2.011*** 1.857*** Has been stopped and asked about status 2.782*** 2.794*** 2.427** Has experienced racial discrimination 2.049*** 1.884*** 2.014*** Fear of deportation --1.557*** 1.452*** U.S. Citizen ----1.013 First generation immigrant ----0.983 Second generation immigrant ----1.805 Sex – female ----0.638** Income ----0.948 Education ----0.994 Language – English ----0.296** Language – Bilingual ----0.688 Constant 0.090*** 0.029*** 0.066*** Log Likelihood -206,745.14 -198,656.03 -192,649.64 Pseudo R-square 0.087 0.123 0.149 AIC 413,498.30 397,322.10 385,325.30 BIC 413,518.30 397,347.10 385,390.40 Two-tailed tests: ***p<0.01; **p<0.05; *p<0.1 H3: Encounters with the homeland security state and experiences of racial discrimination will increase Latinos’ estimated odds of talking about the immigration policy debate. Knowing a detainee or deportee and reporting experiences of discrimination significantly increase one’s odds of discussing immigration reform. Table 5 shows that knowing detainees or deportees increases one’s predicted odds of talking about reform by at least 65% in each model, while reporting discrimination consistently doubles one’s odds. However, FOD becomes statistically significant only in Model 3 with the addition of immigration status, citizenship, sex, income, education, and language preference. In Model 3, FOD increases one’s likelihood of talking about immigration reform by 28.7% percent. Again, being stopped and asked about 30   one’s status remains insignificant, perhaps due to the small cell size. Sex is also insignificant, though higher levels of education and income increase one’s likelihood of talking about immigration reform. Therefore, Hypothesis 6 is partially supported. Table 5: Odds Ratios for Talking about Immigration Reform using Binary Logistic Regression (N=1,104) Model 1 Model 2 Model 3 Knows someone who has been deported 1.780*** 1.667*** 1.653*** Has been stopped and asked about status 1.285 1.269 1.383 Has experienced racial discrimination 2.343*** 2.295*** 1.983*** Fear of deportation --1.100 1.287*** U.S. Citizen ----0.845 First generation immigrant ----0.648 Second generation immigrant ----0.731 Sex – female ----1.092 Income ----1.196*** Education ----1.391*** Language – English ----0.716 Language – Bilingual ----0.884 Constant 1.141 0.919 0.217*** Log Likelihood -314,326.99 -313,556.73 -296,587.07 Pseudo R-square 0.046 0.049 0.100 AIC 628,662.0 627,123.50 593,200.10 BIC 628,682.0 627,148.50 593,265.20 Two-tailed tests: ***p<0.01; **p<0.05; *p<0.1 H4: Increased encounters with the homeland security state and experiences of discrimination will decrease Latinos’ odds of planning to vote. Only citizens are able to participate in the electoral process, so Hypothesis 4 was tested using a subsample of participants who are US citizens (N=652). Among this group, the only significant predictor of planning to vote is experience with discrimination (see Table 6). Respondents who report experiencing racial discrimination are twice as likely to plan to vote than their counterparts who have not encountered discrimination, holding constant all other covariates. This effect retains its statistical significance even with the addition of confounders (these additional independent variables include immigration status, sex, income, education, and 31   language preference). Increased income and education also cause higher odds of planning to vote, though immigration status, sex, and language preference are not significant indicators. Again, more research is needed to disentangle these relationships; confounders such as political views or religious affiliation may affect citizens’ likelihood of voting. Table 6: Odds Ratios for US Citizens’ Plans to Vote in the 2010 Presidential Election using Binary Logistic Regression (N=652) Model 1 Model 2 Model 3 Knows someone who has been deported 0.750 0.810 0.809 Has been stopped and asked about status 1.046 1.073 1.192 Has experienced racial discrimination 2.183*** 2.251*** 1.996*** Fear of deportation --0.894 0.953 First generation immigrant ----1.392 Second generation immigrant ----0.679 Sex – female ----1.011 Income ----1.135** Education ----1.276** Language – English ----1.265 Language – Bilingual ----1.371 Constant 2.181*** 2.725*** 0.551 Log Likelihood -188,521.46 -187,943.89 -179,284.55 Pseudo R-square 0.021 0.024 0.069 AIC 377,050.9 375,897.8 358,593.1 BIC 377,068.8 375,920.2 358,646.9 Two-tailed tests: ***p<0.01; **p<0.05; *p<0.1 Discussion The statistical models explaining Fear of Deportation provide additional evidence supporting much of the existing literature and, in addition, suggests that FOD and encounters with the homeland security state complicate Latinos’ relationship with political incorporation. As Beltrán (2009), Cisneros (2013), and Gonzales (2014) have asserted, Latino migrants resist discrimination and alienation by participating in counter-publics and highlighting their pseudocitizenship status. Perhaps for this reason, participants who report higher levels of FOD are more likely to participate in immigration reform marches and protests. This also may be an 32   effect of mixed status families; citizens or residents may attend rallies to show support for their undocumented relatives and friends or their loved ones who have been detained or deported. Similarly, experiences of discrimination significantly increased Latinos’ plans to vote in the 2010 presidential election, while FOD and encounters with detainees, deportees, and immigration control agents did not have a significant effect in either direction. This relationship was upheld when citizenship and immigration status were introduced. As De Genova (2002; 2006) and Golash-Boza (2012) have argued, the racialization and criminalization of young Latino males has contributed to the persistence of racial discrimination regardless of immigration status; “Mexican-looking” citizens, residents, and undocumented immigrants all may be treated poorly due to their skin color and phenotype. LIMITATIONS This study is limited by its indicators of experience with the homeland security state as well as political incorporation. The use of secondary data analysis necessarily limits scholars’ choice of variables. For example, within this study FOD includes fear for oneself or fear for a friend or family member; ideally these would be disaggregated to better address the deportation threat and explore the potential difference between those who fear for themselves compared with those who are afraid for family members or friends. Political incorporation here is divided into (1) participation in protests for immigration reform, (2) discussions about immigration reform, and (3) plans to vote. While immigration is a particularly salient issue for many Latinos (Ontiveros 2014), it is not the only issue. Better indicators of political integration and participation would surpass specific issues, as well as address a wider variety of actions, such as letters written to congressional leaders and funds raised for political parties. Following DeSipio (2011), future studies should also incorporate seemingly apolitical acts such as attending parent- 33   teacher association meetings or belonging to a trade union. Furthermore, qualitative research is needed to clarify participants’ responses. For example, respondents may define “talking about immigration reform” in a variety of ways. Inconsistency among definitions may invalidate certain responses. However, the sample is nationally representative and all major assumptions were tested by the author to ensure reliability. CONCLUSION This project explored the effects of Latinos’ fear of deportation, encounters with the homeland security state, experiences with racial discrimination on their self-reported political engagement. I anticipated that Latinos’ FOD would be influenced by their acquaintance with deportees or detainees, their having been stopped by immigration agents, and their experiences of discrimination. Of these, acquaintance with deportees or detainees and experiences of discrimination were significant. Consistent with previous literature (Lopez et al. 2013), these relationships were mediated by citizenship and immigration status. Hypotheses 2 and 3 asserted that FOD and encounters with immigration control agents would increase Latinos’ odds of attending immigration reform protests and engaging in conversations about immigration, respectively. My findings provided some support for these hypotheses. Hypothesis 4, that FOD and other encounters with the immigration control regime would decrease Latinos’ estimated odds of planning to vote, was not supported after accounting for citizenship. However, experiences of discrimination were found to positively affect respondents’ plans to vote. These findings corroborate Inda and Dowling’s (2013) assertion that Latino migrants are engaging in migrant counter-conducts in spite of—or perhaps due to—their alienation from the American civic and political realm. Furthermore, this study supports DeSipio’s (2011) argument that Latinos are politically engaging in the US and adds that their participation is uniquely 34   shaped by FOD and other encounters with the immigration control regime, as well as experiences of discrimination. Together, these findings suggest that providing undocumented immigrants with a path to citizenship would enhance and enliven the US political realm as well as lessen the stress and anxiety that deportation creates within Latino communities and families. This study has begun to address Latino migrant communities’ political incorporation and citizenship in the US. However, future research should interrogate native- and foreign-born Latinos’ attitudes toward detainment, deportation, and migration reform, as well as their political practices and civic participation. Qualitative research will need to comprehensively examine the deportation threat from the perspectives of those who live with it every day, paying special attention to gender differences and variation among ethnic groups and LGBTQ respondents. Finally, longitudinal analysis also should be explored to determine whether Latinos’ political participation and fear of deportation have changed over time. 35   REFERENCES 36   REFERENCES Abrego, Leisy J. 2011. “Legal Consciousness of Undocumented Latinos: Fear and Stigma as Barriers to Claims-Making for First- and 1.5-Generation Immigrants.” Law and Society Review 45(2): 337-369. Abrego, Leisy J. and Cecilia Menjívar. 2011. “Immigrant Latina Mothers as Targets of Legal Violence.” International Journal of Sociology of the Family 37(1): 9-26. Arbona, Consuelo et al. 2010. “Acculturative Stress Among Documented and Undocumented Latino Immigrants in the United States.” Hispanic Journal of Behavioral Sciences 32(3): 362–84. Balderrama, Francisco E. 2005. “The Emergence of Unconstitutional Deportation and Repatriation of Mexicans and Mexican Americans as a Public Issue.” Radical History Review 93:107-110. Barreto, Matt A. 2007. “The Role of Latino Candidates in Mobilizing Latino Voters: Revisiting Latino Vote Choice.” Pp. 63-89 in Latino Politics: Identity, Mobilization, and Representation, edited by Rodolfo Espino, David L. Leal, and Kenneth J. Meier. Charlottesville, NC, and London, UK: University of Virginia Press. Beltrán, Cristina. 2009. “Going Public: Hannah Arendt, Immigrant Action, and the Space of Appearance.” Political Theory 37(5): 595-622. Bloemraad, Irene. 2006. Becoming a Citizen: Incorporating Immigrants and Refugees in the United States and Canada. Berkeley and Los Angeles, CA: University of California Press. Bosniak, Linda. 2000. “Citizenship Denationalized.” Indiana Journal of Global Legal Studies 7:447-508. Bosniak, Linda. 2006. The Citizen and the Alien: Dilemmas of Contemporary Membership. Princeton, NJ: Princeton University Press. Brabeck, Kalina M., M. Brinton Lykes, and Rachel Hershberg. 2011. “Framing Immigration to and Deportation from the United States: Guatemalan and Salvadoran Families Make Meaning of Their Experiences.” Community, Work & Family 14(3): 275–96. Brettell, Caroline B. and Deborah Reed-Danahay. 2012. Civic Engagements: The Citizenship Practices of Indian and Vietnamese Immigrants. Stanford, CA: Stanford University Press. 37   Capps, Randolph, Rosa Maria Castañeda, Ajay Chaudry, and Robert Santos. 2007. “Paying the Price: The Impact of Immigration Raids on America’s Children.” Report for the Urban Institute for the National Council of La Raza. Retrieved April 29, 2014 (http://www.urban.org/publications/411566.html). Cisneros, Josue David. 2013. “(Re)bordering the Civic Imaginary: Rhetoric, Hybridity, and Citizenship in La Gran Marcha.” Pp. 253-268 in Governing Immigration Through Crime: A Reader, edited by Julie A. Dowling and Jonathan Xavier Inda. Stanford, CA: Stanford University Press. Coutin, Susan Bibler. 2013. “Place and Presence within Salvadoran Deportees’ Narratives of Removal.” Childhood 20(3): 323–36. De Genova, Nicholas. 2002. “Migrant ‘Illegality’ and Deportability in Everyday Life.” Annual Review of Anthropology 31:419-447. De Genova, Nicholas. 2006. “The Legal Production of Mexican/Migrant ‘Illegality.’” Pp. 61-90 in Latinos and Citizenship: The Dilemma of Belonging, edited by Suzanne Oboler. Hampshire, UK, and New York, NY: Palgrave MacMillan. De Genova, Nicholas. 2007. “The Production of Culprits: From Deportability to Detainability in the Aftermath of ‘Homeland Security.’” Citizenship Studies 11(5): 421-448. DeSipio, Louis. 2011. “Immigrant Incorporation in an Era of Weak Civic Institutions: Immigrant Civic and Political Participation in the United States.” American Behavioral Scientist 55(9): 1189-1213. Flores, William V. and Rina Benmayor, eds. 1997. Latino Cultural Citizenship: Claiming Identity, Space, and Rights. Boston, MA: Beacon Press. Flores, William. 2003. “New Citizens, New Rights: Undocumented Immigrants and Latino Cultural Citizenship.” Latin American Perspectives 30(2): 295-308. Fraga, Luis R., John A. Garcia, Rodney E. Hero, Michael Jones-Correa, Valerie Martinez-Ebers, and Gary M. Segura. 2012. Latinos in the New Millennium: An Almanac of Opinion, Behavior, and Policy Preferences. Cambridge, UK, and New York, NY: Cambridge University Press. Golash-Boza, Tanya. 2012. Immigrant Nation: Raids, Detentions, and Deportations in Post-9/11 America. Boulder, CO: Paradigm Publishers. Golash-Boza, Tanya and Pierrette Hondagneu-Sotelo. 2013. “Latino Immigrant Men and the Deportation Crisis: A Gendered Racial Removal Program.” Latino Studies 11(3): 271292. 38   Gonzales, Alfonso. 2014. Reform without Justice: Latino Migrant Politics and the Homeland Security State. New York, NY: Oxford University Press. Grace, Breanne L. 2013. “Rights of Refuge, Rights of War: Social Citizenship After Refugee Resettlement.” Ph.D. Dissertation, Department of Sociology, Michigan State University. Gutiérrez, David. 1995. Walls and Mirrors: Mexican Americans, Mexican Immigrants, and the Politics of Ethnicity. Berkeley, CA: University of California Press. Hacker, Karen et al. 2011. “The Impact of Immigration and Customs Enforcement on Immigrant Health: Perceptions of Immigrants in Everett, Massachusetts, USA.” Social Science & Medicine 73(4): 586–94. Hagan, Jacqueline Maria, Karl Eschbach, and Nestor Rodriguez. 2008. “U.S. Deportation Policy, Family Separation, and Circular Migration.” The International Migration Review 42(1):64-88. Hagan, Jacqueline Maria, Nestor Rodriguez, and Brianna Castro. 2011. “Social Effects of Mass Deportations by the United States Government, 2000-10.” Ethnic and Racial Studies 34(8): 1374-1391. Inda, Jonathan Xavier and Julie A. Dowling. 2013. “Introduction: Governing Migrant Illegality.” Pp. 1-36 in Governing Immigration Through Crime: A Reader, edited by Julie A. Dowling and Jonathan Xavier Inda. Stanford, CA: Stanford University Press. Jiménez, Tomas. 2009. Replenished Ethnicity: Mexican Americans, Immigration, and Identity. Berkeley and Los Angeles, CA: University of California Press. Kanstroom, Dan. 2007. Deportation Nation: Outsiders in American History. Cambridge, MA, and London, UK: Harvard University Press. Leal, David L. 2007. “Latino Public Opinion: Does it Exist?” Pp. 27-43 in Latino Politics: Identity, Mobilization, and Representation, edited by Rodolfo Espino, David L. Leal, and Kenneth J. Meier. Charlottesville, NC, and London, UK: University of Virginia Press. Lopez, Mark Hugo, Paul Taylor, Cary Funk, and Ana Gonzalez-Barrera. 2013. “On Immigration Policy, Deportation Relief Seen as More Important than Citizenship: A Survey of Hispanics and Asian Americans.” Pew Hispanic Trends Project, December 19. Retrieved February 6, 2014 (http://www.pewhispanic.org/2013/12/19/on-immigration-policydeportation-relief-seen-as-more-important-than-citizenship/). Lundy, Brandon D. 2011. “Dearly Deported: Social Citizenship of Undocumented Minors to the US.” Migration Letters 8(1):55-66. 39   Mahler, Sarah J. and Myer Siemiatycki. 2011. “Diverse Pathways to Immigrant Political Incorporation: Comparative Canadian and U.S. Perspectives.” American Behavioral Scientist 55(9): 1123-1130. Mahler, Sarah J. and Patricia R. Pessar. 2001. “Gendered Geographies of Power: Analyzing Gender Across Transnational Spaces.” Identities 7(4):441-459. Marshall, Thomas H. 1964. Class, Citizenship, and Social Development. New York City, NY: Doubleday and Company. McKeown, Adam M. 2008. Melancholy Order: Asian Migration and the Globalization of Borders. New York, NY: Columbia University Press. Nicholls, Walter J. 2013. The DREAMers: How the Undocumented Youth Movement Transformed the Immigrant Rights Debate. Stanford, CA: Stanford University Press. Oboler, Suzanne. 2006. “Redefining Citizenship as a Lived Experience.” Pp. 3-30 in Latinos and Citizenship: The Dilemma of Belonging. Hampshire, UK, and New York, NY: Palgrave MacMillan. Ong, Nhu-Ngoc T. and David S. Meyer. 2004. “Protest and Political Incorporation: Vietnamese American Protests, 1975-2001.” UC Irvine: Center for the Study of Democracy. Retrieved October 14, 2014 (http://escholarship.org/uc/item/6q8631rv#). Ong, Nhu-Ngoc T. and David S. Meyer. 2008. “Protest and Political Incorporation: Vietnamese American Protests in Orange County, California, 1975-2001.” Journal of Vietnamese Studies 3(1): 78-107. Pachon, Harry and Louis DeSipio. 1994. New Americans by Choice: Political Perspectives of Latino Immigrants. Boulder, CO: Westview Press, Inc. Pew Hispanic Center. 2010. National Survey of Latinos [data file and codebook]. Retrieved March 3, 2014 (http://www.pewhispanic.org/category/datasets/). Ramakrishnan, S. Karthick and Celia Vilamontes. 2006. Civic Inequalities: Immigrant Volunteerism and Community Organizations in California. San Francisco, CA: Public Policy Institute of California. Ramakrishnan, S. Karthick and Irene Bloemraad, eds. 2008. Civic Hopes and Political Realities: Immigrants, Community Organizations, and Political Engagement. New York, NY: Russell Sage Foundation. Reed-Danahay, Deborah and Caroline B. Brettell, editors. 2008. Citizenship, Political Engagement, and Belonging: Immigrants in Europe and the United States. Piscataway, NJ: Rutgers University Press. 40   Rodríguez, Nestor and Jacqueline Maria Hagan. 2004. “Fractured Families and Communities: Effects of Immigration Reform in Texas, Mexico, and El Salvador.” Latino Studies 2: 328-351. Rosaldo, Renato and William V. Flores. 1997. “Identity, Conflict, and Evolving Latino Communities: Cultural Citizenship in San Jose, California.” Pp. 57-96 in Latino Cultural Citizenship: Claiming Identity, Space, and Rights, edited by William V. Flores and Rina Benmayor. Boston, MA: Beacon Press. Sáenz, Rogelio, Cecilia Menjívar, and San Juanita Edilia García. 2013. “Arizona’s SB 1070: Setting Conditions for Violations of Human Rights Here and Beyond.” Pp. 165-179 in Governing Immigration Through Crime: A Reader, edited by Julie A. Dowling and Jonathan Xavier Inda. Stanford, CA: Stanford University Press. U.S. Department of Homeland Security. 2013. Yearbook of Immigration Statistics: 2012. Washington, D.C.: U.S. Department of Homeland Security, Office of Immigration Statistics. Retrieved February 7, 2014 (https://www.dhs.gov/yearbook-immigrationstatistics). Vanderkooy, Patricia and Stephanie J. Nawyn. 2011. “Identifying the Battle Lines: LocalNational Tensions in Organizing for Comprehensive Immigration Reform.” American Behavioral Scientist 55(9): 1267-1286. Yoshikawa, Hirokazu. 2011. Immigrants Raising Citizens: Undocumented Parents and Their Young Children. New York, NY: The Russell Sage Foundation. 41