Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Central American displacement and the politics of United States deterrence strategy Barrick, Leigh Christine 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


24-ubc_2018_november_barrick_leigh.pdf [ 1.54MB ]
JSON: 24-1.0372320.json
JSON-LD: 24-1.0372320-ld.json
RDF/XML (Pretty): 24-1.0372320-rdf.xml
RDF/JSON: 24-1.0372320-rdf.json
Turtle: 24-1.0372320-turtle.txt
N-Triples: 24-1.0372320-rdf-ntriples.txt
Original Record: 24-1.0372320-source.json
Full Text

Full Text

CENTRAL AMERICAN DISPLACEMENT AND THE POLITICS  OF UNITED STATES DETERRENCE STRATEGY by  Leigh Christine Barrick  B.A., The University of Puget Sound, 2009 M.A., The University of British Columbia, 2013  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Geography)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  September 2018  © Leigh Christine Barrick, 2018 ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled:  Central American Displacement and the Politics of United States Deterrence Strategy  submitted by Leigh Christine Barrick  in partial fulfillment of the requirements for the degree of Doctor of Philosophy  in The Faculty of Graduate and Postdoctoral Studies (Geography)  Examining Committee: Dr. Juanita Sundberg Supervisor  Dr. Jim Glassman Supervisory Committee Member  Dr. Geraldine Pratt Supervisory Committee Member Dr. Erin Baines University Examiner Dr. Rima Wilkes University Examiner  iii  Abstract  This dissertation examines the amplification of United States efforts to “deter” the arrival of asylum seekers from El Salvador, Guatemala, and Honduras in recent years. I focus on: (1) the forced separation of asylum-seeking families through detention; (2) the gendering and denial of common Central American asylum claims; and (3) knowledge production that depoliticizes conflict in El Salvador. I ask: how does forced separation impact families’ well-being and access to asylum? What makes this practice politically possible? What obstacles do young men face in making their asylum claims heard? What might more complex stories of displacement sound like, if permitted?  I analyze displacement through a coloniality/modernity lens, meaning that I foreground how power inequalities rooted in colonial conquests contribute to uneven mobility in the Americas today. I draw from qualitative research, including interviews with asylum seekers and advocates, textual analysis of court filings and policy documents, and observation of asylum processes. In Chapters 3 and 4, I suggest forced separation harms families by threatening their well-being and access to asylum. I conceptualize this practice as a form of racialized governance. In Chapters 5 and 6, I demonstrate that detention throws countless hurdles into the path of asylum seekers, while adjudication tends to feminize, depoliticize, and thereby reject common Central American claims. I conclude that the political nature of conflict in El Salvador defies such depoliticizing asylum narratives, demanding a more complex analysis. I argue that the amplification of deterrence strategy expands a racialized system of governance over mobility in the Americas. It limits public debate by depoliticizing the causes of displacement from Central America, while distancing United States actors from any culpability. iv  This dissertation contributes to a growing critique of deterrence strategy by elaborating a coloniality/modernity analytical approach to the study of displacement, which creates a fuller picture of the power imbalances that oblige people to leave their communities. The dissertation serves as a counterweight to deterrence strategy – challenging the current politics of mobility in the Americas and providing insights into strategies for change.    v  Lay Summary  This dissertation examines United States efforts to “deter” the arrival of asylum seekers from El Salvador, Guatemala, and Honduras in recent years through deportation practices and foreign aid. My goal is to shed light on how deterrence strategy can impact the well-being of asylum seekers, and shape popular knowledge about the causes of their displacement and potential solutions. Although deterrence strategy promises to protect asylum seekers by discouraging them from a dangerous, clandestine journey north, I find that it can have harmful effects, including: (1) the forced separation of family members in detention centers, (2) the denial of common asylum claims, and (3) the creation of depoliticizing narratives about the causes of displacement. This dissertation contributes to a growing critique of deterrence strategy by foregrounding how power inequalities rooted in colonial conquests contribute to uneven mobility in the Americas today – an insight that should inform strategies for change. vi  Preface  As the author of this dissertation, I identified and designed the research program, performed the research, and analyzed the research data. My research program was approved by the University of British Columbia (UBC) Behavioural Research Ethics Board under Certificate Number H15-03186. Chapter 3 is a revised and updated version of the following published report: Barrick, Leigh (2016, August 31) Divided by detention: Asylum-seeking families’ experiences  of separation. Special report. American Immigration Council.  I conducted the research for this report. Research participants included former clients of the CARA Family Detention Pro Bono Project, which generously made referrals for me. I wrote the report, which was then revised and edited by staff members with the American Immigration Council. I then further revised the report for its inclusion in this dissertation as Chapter 3.  vii  Table of Contents  Abstract ......................................................................................................................................... iii Lay Summary .................................................................................................................................v Preface ........................................................................................................................................... vi Table of Contents ........................................................................................................................ vii List of Figures ............................................................................................................................... xi List of Abbreviations .................................................................................................................. xii Acknowledgements ......................................................................................................................xv Chapter 1: Introduction ................................................................................................................1 1.1 Research design .............................................................................................................. 9 1.2 Research methods ......................................................................................................... 15 1.2.1 Family separation ...................................................................................................... 16 1.2.2 Asylum adjudication ................................................................................................. 18 1.2.3 Conflict in El Salvador.............................................................................................. 20 1.3 Analytical framework ................................................................................................... 21 Chapter 2: A history of North-South entanglement .................................................................29 2.1 Cold War militarism ..................................................................................................... 31 2.1.1 “Low-intensity” warfare ........................................................................................... 33 2.1.2 Deterrence strategy ................................................................................................... 36 2.2 Peacetime militarism ..................................................................................................... 40 2.2.1 Insecurity in El Salvador ........................................................................................... 42 2.2.2 Perpetual temporariness ............................................................................................ 45 viii  2.3 Conclusion .................................................................................................................... 47 Chapter 3: Divided by detention: Asylum-seeking families’ experiences of separation .......49 3.1 Overview ....................................................................................................................... 52 3.1.1 Research methods ..................................................................................................... 54 3.1.2 Analytical framework ............................................................................................... 55 3.2 DHS separates families ................................................................................................. 62 3.2.1 CBP separates families in its holding cells ............................................................... 62 3.2.2 ICE follows a narrow interpretation of family unity ................................................ 65 3.2.3 Bed quota encourages arbitrary detention practices ................................................. 66 3.3 Separation harms families’ well-being ......................................................................... 70 3.3.1 Participants report negative mental health impacts .................................................. 70 3.3.2 Inconsistent case timelines and decisions cause distress and extend separation ...... 73 3.3.3 Separation creates an undue caretaking burden ........................................................ 74 3.4 Separation harms families’ ability to access humanitarian protection .......................... 76 3.4.1 Separation interferes with families’ ability to present asylum claims ...................... 77 3.4.2 Separation divides asylum cases ............................................................................... 78 3.4.3 Separation divides evidence ...................................................................................... 79 3.4.4 Separated family members can face inconsistent timelines and case decisions ....... 80 3.5 Conclusion .................................................................................................................... 83 Chapter 4: A familiar state violence ...........................................................................................87 4.1 Family separation as racialized governance ................................................................. 90 4.1.1 The separation of enslaved families for profit .......................................................... 93 4.1.2 The separation of indigenous families as a technique of genocide ........................... 96 ix  4.1.3 Family unity and racial hierarchies within the US immigration system ................... 99 4.2 A threat narrative emerges .......................................................................................... 102 4.2.1 The rumors narrative informs policy debate ........................................................... 104 4.2.2 DHS uses rumors narrative to justify prolonged detention ..................................... 106 4.2.3 Litigation challenges prolonged detention as “deterrence" .................................... 108 4.3 Silence and doublespeak around family separation .................................................... 109 4.3.1 An eerie silence on adult family unity .................................................................... 110 4.3.2 Separation as threat and punishment....................................................................... 113 4.3.3 Evaluating effect and intention ............................................................................... 116 4.4 Conclusion .................................................................................................................. 118 Chapter 5: Bad hombres? Gendered scripts and exclusion in the US asylum system .........122 5.1 Analytical framework and methods ............................................................................ 124 5.2 Pursuing asylum behind bars ...................................................................................... 129 5.2.1 The status quo ......................................................................................................... 131 5.2.2 Not everyone gets their day in court ....................................................................... 132 5.2.3 Courtroom performance .......................................................................................... 136 5.3 The “bad scripts” of the asylum system ...................................................................... 139 5.4 Asylum adjudication depoliticizes conflict in Central America ................................. 142 5.4.1 The denial of gang recruitment claims.................................................................... 143 Adjudicators deny claims based on gender, youth, and resistance ................. 145 Neutrality in conflict is not recognized as a political opinion ........................ 147 Gang-related violence is not recognized as political ...................................... 149 5.4.2 The struggle for recognition of women’s gender-based violence claims ............... 150 x  5.4.3 The narrow recognition of family claims ................................................................ 154 5.5 Conclusion .................................................................................................................. 156 Chapter 6: Complicating black-and-white narratives of conflict in El Salvador ................159 6.1 The politics of knowledge production ........................................................................ 162 6.1.1 Revising my research plan ...................................................................................... 164 6.1.2 Reviewing my analytical framework ...................................................................... 166 6.2 Shades of gray ............................................................................................................. 170 6.2.1 Between comprehensive and repressive security strategies .................................... 171 6.2.2 Place-based stigmatization ...................................................................................... 174 6.2.3 Resurgent militarism ............................................................................................... 179 6.3 Conflict and displacement as political ........................................................................ 182 6.3.1 Reconciling contradictions...................................................................................... 184 6.3.2 Constraints to social reform .................................................................................... 187 6.4 Conclusion .................................................................................................................. 191 Chapter 7: Conclusion ...............................................................................................................193 7.1 Historical continuity and complexity .......................................................................... 195 7.2 Impunity and incoherence ........................................................................................... 199 7.3 Future research directions ........................................................................................... 202 7.3.1 Conceptual analysis: Displacement, complexity, and masculinity ......................... 203 7.3.2 Methodological approach: Deterrence as foreign and domestic policy .................. 204 7.3.3 Empirical focus: Family separation through detention ........................................... 206 7.4 Conclusion .................................................................................................................. 207 References ...................................................................................................................................209 xi  List of Figures  Figure 1: Poster in daycare near San Salvador, February 2015; photo taken by Leigh Barrick ..... 5 Figure 2: Two versions of the same page of an internal Border Patrol memo (US CBP 2008) ... 18 Figure 3: Eloy Detention Center, October 2016; photo by Leigh Barrick .................................. 130  xii  List of Abbreviations  ABC: American Baptist Churches v. Thornburgh (1991)  ADAA: Anti-Drug Abuse Act  AEDPA: Antiterrorism and Effective Death Penalty Act  AILA: American Immigration Lawyers Association   ARENA: Alianza Republicana Nacionalista [Nationalist Republican Alliance]   BIA: Board of Immigration Appeals  BREB: Behavioural Research Ethics Board   CAFTA-DR: Dominican Republic-Central American Free Trade Agreement   CARA (Family Detention Pro Bono Project): a collaboration between CLINIC, the American Immigration Council, RAICES, and AILA  CARSI: Central America Regional Security Initiative  CAT: Convention Against Torture   CBP: Customs and Border Protection  CCA: Corrections Corporation of America/CoreCivic  CDS: Consequence Delivery System  CIA: Central Intelligence Agency  CISPES: Committee in Solidarity with the People of El Salvador  CLINIC: Catholic Legal Immigration Network   DACA: Deferred Action for Childhood Arrivals  DAPA: Deferred Action for Parents of Americans and Lawful Permanent Residents  DHS: Department of Homeland Security   xiii  EOIR: Executive Office for Immigration Review  FBI: Federal Bureau of Investigation   FMLN: Frente Farabundo Martí para la Liberación Nacional [Farabundo Martí National Liberation Front]   FOIA: Freedom of Information Act  FSLN: Frente Sandinista de Liberación Nacional [Sandinista National Liberation Front]   FY: Fiscal Year  HHS: Department of Health and Human Services  ICCPR: International Covenant on Civil and Political Rights   ICE: Immigration and Customs Enforcement  IIRIRA: Illegal Immigration Reform and Immigrant Responsibility Act   ILEA: International Law Enforcement Academy  INA: Immigration and Nationality Act  INS: Immigration and Naturalization Service  LOP: Legal Orientation Program   NACARA: Nicaraguan Adjustment and Central American Relief Act   NTA: Notice to Appear  ORR: Office of Refugee Resettlement  PDDH: Procuraduría para la Defensa de los Derechos Humanos [Human Rights Ombudsman’s Office]  PNC: Policía Nacional Civil [National Civil Police]  RAICES: Refugee and Immigrant Center for Education and Legal Services   SSPAS: Servicio Social Pasionista  TPS: Temporary Protected Status xiv   UBC: University of British Columbia  UN: United Nations  UNHCR: United Nations High Commissioner for Refugees   USAID: United States Agency for International Development  USCIS: United States Citizenship and Immigration Services  US: United States  xv  Acknowledgements  This dissertation is inspired by the social movements working towards a fairer politics of mobility in the Americas, and I thank all the mentors, colleagues, and friends who have welcomed me into this work over the years. While acknowledging my gratitude, it is important to note that the views expressed throughout this dissertation, and any shortcomings, are my own.  I wish to thank research participants for generously sharing their insights into deterrence and displacement in the Americas. I also thank colleagues who helped make this project possible in various ways. I am especially grateful to team members with the CARA Family Detention Pro Bono Project and the Florence Immigrant & Refugee Rights Project. I thank friends with the Committee in Solidarity with the People of El Salvador for their invaluable insights and for meaningful opportunities to be part of the solidarity movement.   It is hard to articulate my gratitude to Juanita Sundberg, whose dedication to politically-engaged scholarship inspires my own work in countless ways, for years of generous mentorship. I thank Geraldine Pratt and Jim Glassman for providing crucial guidance at key moments. I also thank Alison Mountz, Erin Baines, and Rima Wilkes for their insights as examiners. I am grateful to the University of British Columbia for the Four Year Fellowship and Faculty of Arts Graduate Student Research Award. I thank mentors at the University of Puget Sound for an undergraduate education that challenged me to question social inequalities.  I am especially grateful to Alex Pysklywec and Paige Patchin for shaping this project for the better from beginning to end. I also thank Anne Lenzini, Sage Ponder, Sarah Brown, Rebekah Kartal, Julia Koch, and Ryan Heath. Finally, I thank my family – Miriam, Stephen, Ron, Cindy, Lindsey, Kayla, Kunal, and Mira – for their continual encouragement. 1  Chapter 1: Introduction  In June 2014, the growing number of Central American mothers and children crossing the United States-Mexico border, many in pursuit of humanitarian protection, made headlines. According to the US Department of Homeland Security (DHS), between Fiscal Years (FY) 2013 and 2014, the number of “unaccompanied children” apprehended near the border jumped from 38,759 to 68,541, while the number of “family units” apprehended grew from 14,855 to 68,445 (US DHS 2016). FY 2014 was also unusual because it marked the first year that DHS apprehended more people from Central America than from Mexico near the southwest border (ibid). Media coverage often compared the children to an invading army or a natural disaster, for example, proclaiming, “Border centers struggle to handle onslaught of young migrants” (Santos 2014, emphasis added), and “As child migrants flood to border, US presses Latin America to act” (Archibald 2014, emphasis added), giving a sense of a nation under siege.  The Obama administration responded to the changing demographics of migration by amplifying existing efforts to “deter” the arrival of asylum seekers and migrants from El Salvador, Guatemala, and Honduras.1 For example, DHS opened two large new detention centers in Texas dedicated entirely to locking-up mothers and children, and warned all would-be Central                                                  1 I use the term “asylum seekers” throughout this dissertation to refer to people who have fled their countries in pursuit of humanitarian protection. I use this term interchangeably with “displaced people.” I also use the term “migrant” here and elsewhere in the dissertation – both to acknowledge that not everyone who moves to the US from Central America without authorized immigration status pursues humanitarian protection, and to refer to people who move across borders more generally. My intention is not to reinforce a dichotomy between the terms, particularly given that asylum seekers are often rejected as “economic migrants.”  2  American migrants of its plans to detain and deport them as quickly as possible (Johnson 2014).2 DHS defended its most controversial practices as an exceptional, but necessary, response to an overwhelming “influx” of children (Flores v. Lynch September 30, 2016: 3-4). The Obama administration also amplified its deterrence strategies beyond United States borders. For instance, DHS launched “an aggressive Spanish language outreach effort” to warn would-be migrants of the dangers of the journey (US CBP 2014). The US State Department has provisioned $24 million in recent years towards Mexico’s Southern Border Plan, an archipelago of checkpoints and other enforcement tools to intercept people on their way north (Ribando Seelke 2016). These enforcement efforts dramatically increased Mexico’s apprehensions of noncitizens with unauthorized immigration status – by 85 percent after just two years (Isacson, Meyer, & Smith 2017: 3). Central Americans continue to travel north. In other words, they have not been deterred, despite an increasingly treacherous journey, excessive force from authorities, and lack of accessible means to seek refuge in Mexico (Isacson, Meyer, & Smith 2015). By 2016, the US Congress had also approved a new foreign aid package for Central America that promised to address the root causes of migration through economic development and security initiatives, but with a few major strings attached – including that recipient countries ramp up enforcement efforts along their regional borders (Biden 2015; Meyer 2017). This aid package thus outsourced United States deterrence strategy even further south. Deterrence strategy is a border enforcement paradigm that seeks to prevent the arrival of certain noncitizens in the United States, and to discourage them from leaving their countries in                                                  2 I use the term “Central America” throughout this dissertation as shorthand to refer to three specific countries: El Salvador, Guatemala, and Honduras, and the term “Central American” to refer to citizens of those countries.  3  the first place. Some of its methods are overtly-coercive, like the segments of fencing that mark the US-Mexico border, and the physical removal of deportees on planes and buses. Other tactics are less blunt or direct, or work at a slower pace, but are no less harmful. For example, many US detention centers are located in remote areas, which isolates detained people from legal counsel and loved ones (Martin 2012a). The multi-sited nature of United States deterrence strategy, which is deployed within and beyond the US-Mexico borderlands, demands a multi-sited analysis. This dissertation is a study of a border enforcement paradigm, but is not fixed at the border. Rather, I trace “circuits of knowledge production and racialized forms of governance” launched by deterrence strategy about people displaced from Central America (following Stoler 2006: 32; Coleman 2008a; Coleman & Stuesse 2016; Hyndman 2001; Loyd 2014; Mountz 2011; Mountz & Loyd 2014). The amplification of deterrence strategy, I argue, expands a racialized system of governance over mobility in the Americas. It limits public debate by depoliticizing the causes of displacement from Central America, while distancing United States actors from any culpability. I provide evidence for my argument by examining three specific expressions of deterrence tactics, which first came to my attention through my participation in solidarity work (as I will elaborate in Section 1.2): (1) the forced separation of detained asylum-seeking families through detention; (2) the gendering and denial of common Central American asylum claims; and (3) knowledge production that depoliticizes conflict in El Salvador. Several themes run throughout my analysis of the techniques of governance and knowledge produced by deterrence as a border enforcement paradigm. Through a focus on normative family life, masculinity, and femininity, this dissertation offers insights into the racialized and gendered dimensions of deterrence tactics. It also sheds light on the impunity held by the state actors and institutions that bring this 4  enforcement paradigm into being for resultant harms. At the same time, the analysis reveals an internal incoherence that troubles any notion of an always-unified ‘state’ agenda. Finally, this dissertation illustrates that knowledge production framing conflict in Central America as apolitical fuels rationales for exclusion. These findings point to an urgent need to complicate the popular conceptualizations of displacement that animate public policy. Despite its harmful tactics, the US executive branch marketed its 2014 amplification of deterrence strategy as a way to protect children, while absolving itself of any culpability for harms they face on their journey north. This messaging is hard to miss in the poster pictured in Figure 1, which I saw hanging in a children’s daycare in El Salvador in early 2015. It is marked with insignias of the US Embassy, Salvadoran government, International Organization for Migration, and United Nations Children's Fund. The text of the poster chastises parents: “Don’t put their lives at risk. The overland journey to the United States without a visa isn’t easy. Don’t expose you daughters and sons to a dangerous journey that could lead to their death.”  5   Figure 1: Poster in daycare near San Salvador, February 2015; photo taken by Leigh Barrick At the center of the illustrated image stand a boy and a girl with flowing black hair, facing away, looking off into the distance. They hold what looks like a balloon string tied to a stark white outline of Canada, the United States, and Mexico floating above. A howling, brown coyote fills some of the empty space between the children and their destination, symbolizing the threat of hiring a coyote [smuggler]. The empty space gives the impression that Central America is socially and spatially distinct from the rest of North America. The messaging boils migration down to the product of poor individual choices.  6  As this poster illustrates, deterrence strategy invokes a process of subject formation that assigns relational identities to people and places. A central premise of this dissertation, which I will elaborate later in this chapter, is that these are racialized identities that assign value to people and places in an asymmetrical way, while naturalizing any perceived differences. In doing so, deterrence strategy, as one paradigm of border enforcement, draws on an archive of representation and techniques of governance that over time have rendered such asymmetrical treatment commonsense (see Braun 2003; Goldberg 2002; Stoler 2013). The most literal function of deterrence strategy is to govern people’s movement – for example, through this poster’s effort to discourage parents in El Salvador from allowing their children to migrate north to the United States. At the same time, these acts of governance also produce knowledge about who belongs where, which places are safe and which are dangerous, and assigns culpability for the resultant risks (see Stoler 2006).  What is missing from the poster are historical context and the structural constraints in which people decide to stay or move. For instance, during the Cold War, the US military trained and armed brutal right-wing regimes across Latin America to eliminate “communists,” which in practice meant “almost any critic of the status quo” (Gill 2004: 10). People who fled then, and those displaced by the aggressive neoliberal reforms adopted soon after, have faced restrictive US immigration policies that, for decades, have sought to exclude them.3 The United States continues to play an outsized role in setting the regional economic and security agenda, while                                                  3 By neoliberal reforms, I mean policies that put the ‘free market’ at the center of social, economic, and political life – for example through deregulation, privatization, and the adoption of free trade agreements. Neoliberal policies were widely implemented across Latin America during the 1980s and 1990s at the behest of powerful US actors and international institutions like the World Bank and International Monetary Fund. In practice, neoliberal reforms have often been imposed through repressive state actions (Glassman 2009: 497).  7  restricting certain people’s mobility. Bringing this context into the story of displacement challenges the official narrative that deterrence is about protecting children, and that the US executive branch’s policies following June 2014 were benign or exceptional.  Deterrence strategy is not new. In fact, it has been central to US border enforcement for decades, and has only expanded geographically, while pulling more people into its dragnet (Andreas 2005; De Genova 2011; Dunn 1996, 2009; US GAO 1992; Mountz & Loyd 2014; Martin 2012a; Nevins 2010; US OIG 2000). This enforcement paradigm only seems to grow, despite its unproven efficacy and harmful outcomes – not least the deaths that have occurred regularly in the US borderlands with Mexico as enforcement has pushed border crossings into remote, harsh terrain (Andreas 2000; Cornelius 2001; Massey 2005; US GAO 2001). This dissertation contributes to a growing critique of deterrence strategy, with an empirical focus on the amplification of deterrence tactics in 2014. As mentioned, this year marked a demographic shift. Migration across the US-Mexico border was increasingly composed of citizens of El Salvador, Guatemala, and Honduras – with a growing number of children (traveling with or without parents) among them (US DHS 2016). 2014 also marked a political moment that warrants analysis. As migration from Central America was politicized in novel ways, the Obama administration targeted this demographic for exclusion (see Chapter 4 for analysis of the political moment). For these reasons, my empirical focus centers on 2014 and its aftermath, even though the deterrence enforcement paradigm itself is not new.   This dissertation also contributes to critiques of deterrence strategy by elaborating a coloniality/modernity analytical approach to foreground how power inequalities rooted in colonial conquests contribute to uneven mobility in the Americas today. The goal is to counter “coloniality,” meaning “the logic of domination in the modern/colonial world” imposed by 8  European colonialism, which endured even as the colonies in the Americas gained formal independence, and lives on today, including through contemporary expressions of imperialism such as United States interventionism across the Western Hemisphere (Mignolo 2005: 7).  As Walter D. Mignolo (2005: 11) argues, a logic of domination embedded within European colonial conquests in the Americas (and globally) associated “modernity” with Eurocentric ways of knowing and being, while non-hegemonic epistemological and ontological standpoints were rendered non-legible. European colonial powers developed a new way of thinking about social difference through hierarchical categories of ‘race,’ which dictated people’s place in a stratified labor system that was increasingly tied to a capitalist global market (Maldonado-Torres 2007: 243-244; Quijano 2000). This logic of domination – centered on race as well as other axes of social difference like gender, sexuality, and class – continues to be expressed through at least four overlapping realms:  (1) the economic: appropriation of land, exploitation of labor, and control of finance; (2) the political: control of authority; (3) the civic: control of gender and sexuality; (4) the epistemic and the subjective/personal: control of knowledge and subjectivity (Mignolo 2005: 11).  Following activist-scholar Harsha Walia (2013), a colonial logic of domination defines the contemporary politics of mobility between the Global North and South. This logic of domination shapes not only paradigms of border enforcement, but can also permeate migrant justice movements, generating advocacy arguments that “perpetuate divisions between the worthy, deserving, and desirable migrant and the disposable, undeserving, and undesirable migrant” (Walia 2013: 257-258). In this way, an argument for the inclusion of one group can reinforce the exclusion of another. Recent efforts to ‘decolonize’ migrant justice advocacy have led some social movements and policymakers to make ambitious demands that leave no one behind: to abolish mass detention and deportation altogether (ibid). Such demands can be seen as 9  ‘decolonizing’ advocacy in that they do not reinscribe a colonial logic of domination. Rather than arguing that a punitive border enforcement paradigm is appropriate for certain people (e.g. people with criminal records) and not appropriate for others (e.g. children), an abolitionist approach simply argues against the punitive enforcement paradigm itself. Inspired by this abolitionist approach, one central goal of this dissertation is to challenge the logic of domination that confines mobility in the Americas. To this end, I draw from anti-colonial and anti-racist analytical tools developed by the Latin American coloniality research program, Chicano/a studies, and critical race, feminist, and postcolonial theory. My analysis is also inspired by the efforts led by undocumented people, immigrants, and people of color within migrant justice movements to put anti-colonial and anti-racist theory to work – exemplified by the #Not1More movement’s fight to end all US deportations, whose approach is aligned in many ways with the prison abolition movement, in which women of color play a leading role (see Davis 2003; Gilmore 2007). As I explain next, the anti-colonial, anti-racist approach I pursue in this dissertation is not something a researcher can simply subscribe to when writing their final analysis. Rather, it must be woven throughout the research process, beginning with research design.    1.1 Research design My undergraduate education in political economy taught me to ask the following question when analyzing any social relationship: “who benefits?” (see Balaam & Veseth 2005: 480). This simple question unearths a labyrinth of interests. In 2008, I traveled to Chiapas and Oaxaca, Mexico to research whether and how small-scale farmers benefit from selling their coffee to specialty markets. This research project arose from my sense that fair trade was not fair enough. I 10  quickly realized that lurking just below the surface of this social landscape were complex power relations: enduring colonial histories, the rise of neoliberal policies and ideology, cultures of consumption of the Global North, my own access to a research grant and passport that allowed me to travel, and much more.  As I tromped alongside forest-shaded coffee plants with farmers who kindly shared their time and insights with me, I turned the microscope onto myself. As a white woman, and US citizen of a privileged economic background, what role did I play in this unequal status quo? Who was my research for, and who would benefit from it, besides me? These unsettling questions, which are revealing of the historical complicity of academic research in larger power disparities, have pushed me to align my work and studies with a politics of solidarity ever since. My goal as a researcher of inequalities between the Global North and South is thus, in whatever small way possible, to serve as an accomplice of movements for social justice.  Aspiring to a politics of solidarity does not resolve these difficult questions, but it does create a starting point to plan research that is more accountable, and less extractive (Sundberg 2015a). This demands that a researcher begin with careful self-analysis, asking how their social position and geographical origins shape their worldview, and how that worldview and positionality influence their project from beginning to end (ibid). As feminist scholars have long argued, no one possesses an objective “gaze from nowhere,” which means that no knowledge production sits outside of politics (Haraway 1991; also see Rose 1993, 1997, cf Sundberg 2005: 19). Thus, it is crucial for researchers to situate the political conditions of their research. I begin this reflexive analysis here, and revisit it throughout this dissertation.   Trying to do research in alliance with social justice movements is laden with challenges and contradictions. One challenge is that this type of research is often accused of lacking rigor, 11  and of pursuing “reductive, politically instrumental truths at the expense of social complexity” (Hale 2008: 2). For me, the tension between instrumental truths and social complexity was so palpable in my PhD research process that I chose to make it a subject of analysis. I encountered various, sometimes clashing narratives of displacement, and found that each one was shaped by an instrumental agenda. Chapter 4 looks at the instrumental narratives deployed by the Obama administration, policymakers, and the media about displaced people. Chapter 5 examines how the US asylum system’s narrow legal bounds box-in people’s narration of their displacement, and limit adjudicator decisions. Chapter 6 delves into what a more complex story of displacement from the vantage point of El Salvador might sound like, if permitted.     Ultimately, I found that my political orientation (and my consciousness of it) only sharpened my analysis. Chapter 3 is a revised version of a policy report I wrote, which examines forced family separation and advocates for DHS to keep loved ones together, and out of detention. My goal was to contribute, if in a small way, to advocacy for more dignified and caring treatment of asylum seekers. My analysis for the report opened a host of new questions for me about the conditions of possibility for forced family separation, which I address in Chapter 4. Further, my analysis also benefited from careful editorial review to make it suitable for a practitioner audience in its publication as a report. At the same time, I was highly motivated to do my best work precisely because I care about the issue. I see this as the production of knowledge that is instrumental, but not reductive or lacking rigor.  My political commitment to the topic of this dissertation prompted me to stay involved with solidarity work alongside my formal research. In early 2015, I traveled to El Salvador as an international elections observer, and to learn about social movements, on a delegation led by the 12  Committee in Solidarity with the People of El Salvador (CISPES).4 I have continued to volunteer with CISPES from Seattle, where I wrote this dissertation. In July 2016, I helped coordinate a CISPES delegation focused on migration and US-led border militarization. I spent much of the summer of 2015 in US detention centers. That July, I volunteered with the CARA Family Detention Pro Bono Project in Texas.5 And that August, I worked as a contractor for a United Nations High Commissioner for Refugees (UNHCR) research project, interviewing Central American and Mexican women about the barriers they face to accessing asylum (UNHCR 2015).  I am thankful for these opportunities to be part of migrant justice and Latin American solidarity movements. I should note that though informed by these experiences, the views expressed throughout this dissertation, and any shortcomings, are my own. This exposure allowed me to develop a research plan that is accountable to my politics. It also deepened my background knowledge of contemporary displacement from Central America and United States deterrence strategy, which I believe created a project that was more rigorous than it otherwise might have been. Finally, this exposure called my attention to the specific dimensions of deterrence strategy that are the focus of this dissertation, of which I otherwise may have been unaware. Beyond my involvement with United States-based solidarity movements, carrying out my social justice-oriented project with a Canadian institution deepened my consciousness of settler                                                  4 CISPES is a United States-based “grassroots solidarity organization that has been supporting the Salvadoran people’s struggle for social and economic justice since 1980” (CISPES 2017).  5 The CARA Family Detention Pro Bono Project (CARA), a collaboration between the Catholic Legal Immigration Network (CLINIC), the American Immigration Council (Council), Refugee and Immigrant Center for Education and Legal Services (RAICES), and the American Immigration Lawyers Association (AILA), is dedicated to promoting and strengthening the rights of immigrants, with a particular focus on ending family detention. 13  colonialism as an active, ongoing form of oppression across the Americas. The University of British Columbia is located on the traditional, ancestral, and unceded territory of the Musqueam people – a reality that is often verbally acknowledged at events on campus. This attention to place, and the colonial history and present it carries, was something I had not experienced at US institutions. Hearing the land acknowledgement countless times prompted me to reflect critically on my own position as a settler and how I contribute to colonization, whether living in Canada or the US.  In 2015, the Truth and Reconciliation Commission of Canada released its final report detailing the heritage of the residential school system that, for upwards of a century, removed children from their families with the intention of eradicating indigenous cultures (TRC of Canada 2015). The report called my attention to the ways that this heritage in both Canada and the US has been buried and dismissed. It also pushed me to see the forced separation of asylum-seeking families today in a different light – not as an isolated incident, or solely an immigrant rights issue, but as an expression of state violence that has been invoked repeatedly across time and space. Chapter 4 explores how family separation has been invoked as a technique of state violence across three historical moments in the United States. The vantage point I gained in Canada made my project more rigorous by rendering the role of settler colonialism impossible to ignore in my analysis of a US border enforcement paradigm.  Another challenge that social justice-oriented researchers must contend with is the question of who, exactly, they aspire to be in solidarity with (Nelson 1999: Chapter 2). Despite working towards a common goal, social movements are never homogenous. Just like society at large, they can be fractured internally along countless lines of difference (ibid; Mohanty 2003; Sangtin Writers & Nagar 2006; also see Crenshaw 1991). Further, no matter how well-14  intentioned, the desire of critical scholars to represent a marginalized population in an ‘authentic’ way can be silencing (Spivak 2010; Mohanty 2003; Krog, et al 2009). The sometimes-clashing narratives of displacement that I encountered between the US and El Salvador prompted me to discuss these dilemmas of solidarity and representation in Chapter 6. Dominant research traditions of the Global North have long played a role in helping justify and sustain harmful imperial ventures, from mapping projects that facilitate militaristic interventions, to the reproduction of Eurocentric knowledge (Tuhiwai-Smith 2012; Mendieta 2008; Sundberg 2015a; Wainwright 2013). Anti-racist, critical indigenous, and feminist scholars have explored a number of methods to make their research more reciprocal towards the communities they wish to align themselves with. This can range from direct collaboration, to putting powerful actors under the microscope rather than marginalized people, to working in political alliance (Sundberg 2015a: 119-120). But as Juanita Sundberg (2015a: 120) notes, these efforts are inevitably incomplete, given that “As scholars trained in the Global North, we are always already marinated in and complicit with geopolitical relations and institutional knowledge that bears traces of imperial histories.”  One way these enduring inequalities express themselves is in the void between the contribution a researcher is able to make, and the urgent needs of participants. For example, the policy report I wrote does not directly assist the families torn apart by detention who participated in the research. Further, the language of such a report can be alienating to the people for whom it seeks to advocate (Hale 2008: 23). As I discuss in Chapter 4, I was quite cognizant of this disconnect, and uncomfortable with it. I decided to pursue the research anyway, because although imperfect, doing something to contribute to the public debate seemed better than staying silent (see Nagar 2002: 181, cf Sundberg 2015a: 120). A policy report can ideally add 15  one more voice to the chorus of a much broader movement for social change, alongside the urgent legal services that I am unable to provide (see Pulido 2008).   In sum, my academic training, social position, and geographic origins in the Global North all shape how I navigate the social world. I have approached this dissertation from a position of solidarity, while doing my best to acknowledge the limitations of my approach. As my analysis in this section demonstrates, striving for an anti-racist, anti-colonial approach to knowledge production is not simply about how the final analysis is written. It is also a methodological process – requiring that a researcher be deliberate about their research design, including research methods.  1.2 Research methods Following Gloria Anzaldúa (1987: Preface), “the Borderlands are physically present wherever two or more cultures edge each other, where people of different races occupy the same territory, where upper, lower, middle and upper classes touch, where the space between two individuals shrinks with intimacy.” Efforts to enforce these divides are just as diffuse as the borderlands that Anzaldúa describes – a reality that guided me towards a multi-sited analysis of deterrence strategy. Over the course of 2016 and into 2017, I carried out qualitative research in the United States and El Salvador, guided by questions I generated through my solidarity work: how does forced family separation impact well-being and access to asylum? What makes this practice politically possible? What obstacles do young men face in making their asylum claims heard? What might more complex stories of displacement sound like, if permitted?  To answer these questions, my primary methods included interviews with twenty advocates and five asylum seekers, textual analysis of court filings and policy documents, and 16  observations of the asylum process in Arizona. The interviews with asylum seekers and advocates each lasted between 40 and 90 minutes, and were conducted in English or Spanish. About half were in person, while the rest were over the phone or Skype. Throughout this dissertation, I refer to all research participants anonymously, with pseudonyms or generic titles like “advocate” or “attorney,” as per my University of British Columbia (UBC) Behavioural Research Ethics Board (BREB) approval. I am deeply grateful to research participants, and to colleagues who otherwise helped me with the research process. I spent three months traveling for this project: one month in Arizona, and two months in El Salvador. I completed the rest of the research and writing from Seattle. Below, I explain the specific methods I used to investigate each thematic area of this dissertation, which I have summarized as: (1) family separation, (2) asylum adjudication, and (3) conflict in El Salvador.  1.2.1 Family separation Through my volunteer work with the CARA Family Detention Pro Bono Project in July 2015, I learned that DHS generally detains mothers and children together as “family units,” but splits off other relatives like fathers, aunts and uncles, grandparents, adult siblings, and adult children, who it detains separately. That fall, I consulted with CARA-affiliated staff members about whether research on forced family separation would be politically useful for advocacy purposes. As mentioned, I ended up writing a report, which draws in part from interviews with five former CARA clients: asylum-seeking women who had been released from the Texas family detention centers to continue their cases.  During March and April 2016, a CARA-affiliated staff member connected me with former clients. I did phone interviews by necessity, given that CARA’s former clients live all over the US. I asked participants when, where, and how their family had been split up, and about 17  what had happened to each relative since then. I inquired about the impacts on their well-being, and the status of each relative’s asylum case. I also asked participants about their policy recommendations for DHS to better attend to the needs of asylum-seeking families. To learn more about how separation impacts families’ access to asylum, I also interviewed (by phone, for the same reason) four attorneys who serve asylum-seeking families. I drew on my connections within the family detention context, and referrals, to request these interviews. I also set out to investigate DHS policy on family unity and separation in detention. Finding little information in publicly-available policy documents, I submitted several Freedom of Information Act (FOIA) requests to Immigration and Customs Enforcement (ICE, which oversees long-term noncitizen detention) and Customs and Border Protection (CBP, which arrests noncitizens and holds them in short-term detention). I also emailed the ICE Office of Public Affairs, but never heard back (Email 2016, April 7). My FOIA responses from ICE arrived in a matter of weeks, while my CBP response took more than a year. These responses answered some of my questions. Their silences and redactions also offered other indirect insights, as I will discuss in further detail in Chapters 3 and 4. FOIA redactions can be intriguingly inconsistent. Figure 2 illustrates two versions of the same page of the same 2008 internal Border Patrol memo, but the version on the left features large redactions marked with the FOIA code “Exemption 7(E).” This code indicates that revealing the hidden text “would disclose techniques and procedures for law enforcement investigations or prosecutions” (US FOIA 2017). My FOIA response contained the heavily-redacted version. Yet the same Border Patrol memo can also be accessed from the CBP FOIA website (meaning that it had been requested through FOIA requests prior to my own and thereby became public record), but with a different pattern of redactions, as pictured on the right. Such discrepancies raise questions about 18  why law enforcement techniques and procedures would at one point in time be approved for public consumption, and later be deemed secret.    Finally, to make sense of the political context in which DHS splits families up through its custody determinations, I analyzed relevant court filings, public statements made by policymakers, and media reports. I contextualize my analysis within secondary literature, including policy reports and historical analyses of forced family separations. 1.2.2 Asylum adjudication Most asylum claims from El Salvador, Guatemala, and Honduras get rejected in the US (TRAC 2016, December 13). To make sense of why, I analyzed about 40 asylum case decisions, most from within the past 15 years or so. The Executive Office for Immigration Review (EOIR, which Figure 2: Two versions of the same page of an internal Border Patrol memo (US CBP 2008) 19  oversees the US immigration court system) does not publish its decisions on asylum cases. The Board of Immigration Appeals (BIA), on the other hand, does publish some of its decisions, as do the federal Courts of Appeals and the Supreme Court. I thus analyzed the texts of publicly-available cases on appeal, reviewing several types of claims common to applicants from El Salvador, Guatemala, and Honduras: those based on gang-related violence, gender-based violence, and family membership. I chose to review these specific types of claims not only because they are common, but because they are gendered differently (either typically submitted by men or by women). I examined how dominant gender norms animate these rejections. To select relevant cases, I began by reviewing secondary legal scholarship and practitioner guides that advise attorneys on strategies to defend Central American claims. The citations within one case often led me to another, particularly back to precedential decisions. To triangulate my analysis of case texts, I also interviewed two more attorneys and one legal advocate, asking them about trends in detention and the adjudication of Central American asylum claims. I drew on my existing connections to request these interviews. I had envisioned pursuing more such interviews, but decided to stop there out of respect for advocates’ time, as these interviews seemed to serve a similar purpose to my review of legal guides. To learn more about the asylum process – including the specific obstacles that young men face to accessing asylum – I served as a as a Direct Services Volunteer with the Florence Immigrant & Refugee Rights Project’s (Florence Project) adult program.6 The Florence Project’s adult program serves people detained in three adult detention centers located in Florence and                                                  6 “The mission of the Florence Project is to provide free legal and social services to detained adults and unaccompanied children facing immigration removal proceedings in Arizona” (FIRRP 2017). 20  Eloy, Arizona. These detention centers are each run by the private prison company CCA (formerly known as the Corrections Corporation of America, recently rebranded as CoreCivic), and/or by ICE. The Florence Project hosted me as a volunteer for the month of May 2016. Under attorney supervision, I assisted with know your rights trainings, intakes, asylum applications, guided pro se assistance, Spanish-English language translation, and research regarding country conditions for asylum applications. I signed a confidentiality waiver to protect Florence Project client information. To ensure that I have adhered to the terms of the waiver, my writing in Chapter 5 that refers to the asylum processes that I observed was reviewed by a Florence Project staff member. 1.2.3 Conflict in El Salvador From Arizona, I took a flight directly to San Salvador, where I spent June and July 2016. My plan was to investigate the early outcomes of a new US foreign aid package for Central America promising to address the root causes of displacement through public security and economic development initiatives. I was planning to seek interviews with advocates who do violence prevention work, or who otherwise serve communities experiencing conflict and displacement. I wanted to inquire about whether the new US aid package was at odds with the Salvadoran government’s recently-launched comprehensive security plan, which sought to move away from the repressive, mano dura [iron fist] public security strategies of years past. Yet it turns out that the aid package had not yet been disbursed. Further, the comprehensive security plan had not been fully implemented, while repressive public security strategies persisted.  As these dynamics became clear to me, I revised my plan. I made my interview questions more complimentary to my research about the adjudication of gang-related asylum claims, asking advocates, for example, about the barriers that young people face to maintaining 21  neutrality in communities impacted by both gang presence and repressive security practices. Colleagues involved in the Latin American solidarity movement helped me make connections with advocates, while participants provided further referrals. I interviewed thirteen advocates, who are affiliated with organizations that receive funding from a variety of sources, including from the US Agency for International Development (USAID), but none affiliated with Salvadoran government programs. As mentioned, I also helped coordinate a solidarity delegation to El Salvador in July 2016. We had about 15 meetings with civil society leaders and government officials on the topics of displacement, migration, security and violence prevention, and border enforcement. At the US embassy, we met with top officials from the US Agency for International Development (USAID), the US International Law Enforcement Academy (ILEA), and the US Bureau of International Narcotics and Law Enforcement Affairs. Finally, we traveled to Guatemala City, where we met with advocates at a migrant shelter, and with military officials who patrol the Guatemalan borders, with US funds and training. Although I do not refer specifically to these meetings, they helped inform my base understanding for this dissertation.   1.3 Analytical framework A central premise of this dissertation is that the amplification of United States government efforts to deter Central American people since 2014 is neither exceptional nor benign. Rather, it is an expression of a decades-old security paradigm that exposes people to harm en route to the US, threatens them not to take a journey north because they might face harm, and banishes them through an ever-growing system of mass deportation. Further, this is a highly racialized system that echoes longstanding inequalities across the Americas.  22  As biological notions of race gained traction at the end of the eighteenth century, the logic of domination in the Americas established by European conquests hardened into a racial grammar centering whiteness (Bonnett 1997: 197; Goldberg 2002). The deepening of US imperial power during the nineteenth century helped entrench Anglo-Saxonist beliefs that North American protestant traditions were the pinnacle of racial and cultural superiority (Berger 1995: 30). Anglo-Saxonism developed, in part, through an imaginary of North America as an industrious, democratic space in contrast to Latin America as a space of brutality, corruption, and excess left by Iberian and Catholic colonialism (ibid: 16). The rise of eugenics in the early twentieth century lent a new air of scientific authority to Anglo-Saxonism. Looking south to Latin America, leading US eugenicists theorized miscegenation as reproductive of racial inferiority, and that spaces of racial hybridity were bound to experience conflict. These racist beliefs directly informed the adoption of racialized immigration restrictions (e.g. the 1924 US Immigration Restriction Act) to preserve the purity of the “supposedly superior American stock at home” (Stepan 1991: 173). The enduring idea of the “illegal alien” developed in this era and was mapped selectively onto Asian populations targeted for exclusion, and soon onto Latin American populations as well (Ngai 2004).  Since the 1970s, US media and policy analyses have frequently portrayed the undocumented immigration of Mexicans, and of Latinos more broadly, as a type of “invasion,” in which an allegedly unassimilable population threatens white, protestant, Anglo people’s ways of life. This threat narrative has stoked nativist anxieties about a diminishing white population in relation to people of color in the United States (Chavez 2013: Chapter 1; Santa Ana 2002). As Leo Chavez (2013: 45) explains, the “Latino Threat Narrative works so well and is so pervasive precisely because its basic premises are taken for granted as true.” This pervasive threat narrative 23  continues to frame Latinos as a homogenous population that is “out of place” in the United States and therefore as contaminating the purity of those who are “in place” (ibid: 46, drawing on Douglas 2003; also see Sundberg & Kaserman 2007; Sundberg 2008).         Anglo-Saxonist views have thus contributed to an enduring imaginary of the United States as a space of racial purity (and superiority) in relation to Latin America. The idea of race as a biological attribute has been discredited, but a central premise of this dissertation is that race remains a constitutive organizing feature of United States social, political, and economic life. Following David Theo Goldberg (2002: 130) race can be understood as an “the embodiment and institutionalization of [power] relations,” in which whiteness has come to be associated with power and privilege. Conversely, people racialized as nonwhite have been subject to myriad forms of state-sanctioned exclusion and violence over time – from acts of genocide towards indigenous peoples, to slavery and Jim Crow, to immigration restrictions and criminal justice policies that disproportionately target people of color (ibid: 5). State institutions mediate this asymmetrical treatment by imagining clear-cut racial boundaries and defining people into those groups (e.g. through the census), regulating how those groups relate to each other, overseeing the economy and thus access to livelihoods, and governing racialized groups differentially (e.g. historically defining colonized peoples and people of color as less than human, less than citizens, or unable to self-govern) (ibid: 110). Given that ‘the state’ is not one homogenous actor, and thus lacks internal coherence, this differential treatment is not always carried out in an entirely intentional way (see Mountz 2004). The complex role of the state is a theme that is woven throughout this dissertation, which I return to in my conclusion (Chapter 7).  Systems of racialization continue to dictate US modes of knowledge production and governance towards people and places both within and outside of its borders. Such systems are 24  reproduced over time through “citationality,” meaning “the reiteration of statements, images, and narratives that have achieved a sort of hegemonic or commonsense status through their continuous repetition and circulation” (Braun 2003: 183, drawing from Butler 1993; Derrida 1988; also see Sundberg & Kaserman 2007). If racialized knowledge production is citational, so too are the tools of governance forged by European colonialism that continue to be invoked in the present – what Ann Laura Stoler (2013) calls “imperial formations” (also see Sundberg 2015b). This is not to say that racial projects exist in a perpetual feedback loop. Because they are never fully achieved, they are constantly vulnerable to disruption (Braun 2003: 185).   Tracing imperial formations is an urgent task, but not because “the contemporary world can be accounted for by colonial histories alone” (Stoler 2013: 7). Instead the purpose is to illuminate how logics of domination rooted in colonialism attribute value and protect certain lives at the expense of others, and to theorize ways to interrupt such violent patterns. This analytical lens is crucial to make sense of why the vast majority of people who DHS apprehends today are nationals of Mexico, Guatemala, El Salvador, and Honduras (accounting for 93 percent of apprehensions in 2015) (Zong and Batalova 2017). As this dissertation demonstrates in the case of Central American asylum seekers, the results can be devastating – ranging from forced family separation, to the denial of common asylum claims. This contemporary reality demands attention to the racialized history of displacement in the Americas, and of forced removal and exclusion from the United States.  US legal scholar Daniel Kanstroom (2007) demonstrates that deportation does much more than simply enforce immigration law. Deportation also serves as “a powerful tool of discretionary social control, a key feature of the national security state, and a most tangible component of the recurrent episodes of xenophobia that have bedeviled our nation of 25  immigrants” (ibid: 5, emphasis added; also see Coleman 2008a). Forced removals have served as a tool of social control throughout United States history. Those most often targeted include people of color, indigenous peoples, poor people, criminalized people, and other marginalized populations.  The legal architecture behind today’s system of mass deportation has deep roots, linking back to “the legitimating theories of the brutal removal of the Cherokee and other American Indians from their lands and to the laws governing thousands of fugitive slaves, captured and sent back to their masters” of the eighteenth and nineteenth centuries (Kanstroom 2007: 7). Litigation that unsuccessfully sought to challenge the Chinese Exclusion Acts of 1882 and 1888 allowed the Supreme Court to grant the federal government “plenary power” over immigration. This withheld full constitutional protection from “aliens” in deportation proceedings, while limiting judicial overview (Varsanyi 2008). Key to the court’s rationale for granting such expansive authority was the logic that mass immigration is a form of foreign aggression, akin to an invasion (Chae Chan Ping v. United States 1889: 606, cf Varsanyi 2008: 884).  These are just a few impactful moments in a larger lineage of stolen land and labor, and xenophobic exclusion that created a roadmap for mass deportation today, while upholding systems of white supremacy. The concept of white supremacy is commonly used to refer to individual racist views – recently personified by the torch-carrying “alt-right” demonstrators in Charlottesville, Virginia. In contrast, anti-racist thinkers like Charles W. Mills (1997: 1) explain white supremacy as a global system of domination forged through European conquest, which has allowed white people to consolidate socio-economic privilege. Mills calls this an “unnamed political system” because it renders its harmful effects commonsense, as though they exist outside the realm of politics. One of its nefarious features is that it imparts an “epistemology of 26  ignorance” upon its beneficiaries, “producing the ironic outcome that whites will in general be unable to understand the world they themselves have made” (ibid: 18).  Despite their ubiquity, systems of white supremacy, and their ties to colonial conquest, often remain unnamed within policy and popular analyses of displacement, migration, and border enforcement. This was true of the Obama administration’s analysis of Central American displacement put forth in 2014, and of public debate more generally, as I find in Chapter 4.  The enduring role of colonial conquest is generally not foregrounded in theories of sovereignty and subjectivity either (Sundberg 2015b: 212). This is the case for some critical scholarship that theorizes border zones and detention centers as a “state of exception,” meaning a site where liberal democratic norms are suspended, and legal protections thereby abandoned for the targeted individuals (in this case, certain noncitizens) (e.g. Doty 2011; Jones 2009, Salter 2008, cf Sundberg 2015b: 211; also see Bigo 2007). This work expands on Giorgio Agamben’s (1998; 2005) theory of sovereign power. Following Sundberg (2015b), Razack (1998), and Walia (2013), I think it is crucial for critical border and migration studies to account for how colonial legacies naturalize seemingly exceptional treatment, and to advocate against these legacies (also see Stoler 2013; 2016). As Sundberg (2015b: 215) demonstrates, the United States borderland with Mexico is a site where “violent forms of control and legal exceptionalism have been the rule, showing the exception to be a constitutive modality of US imperial genres of rule.” As a result, it is not only noncitizens crossing the border who have long faced state violence, but also racialized citizen populations like Latinos and indigenous peoples who reside in the borderland (ibid). A coloniality/modernity analytical lens can help account for distinct, but related imperial formations that have targeted different marginalized populations over time, as in the US 27  borderland with Mexico. This lens also accounts for empire as “a way of life” throughout the United States, and in sites where the US acts as an imperial power, including in Latin America (following Williams 1980, cf Gill 2004: 3; Kaplan 1993; Sundberg 2015b; Stoler 2016). By United States empire, I refer to wide-ranging efforts to consolidate geopolitical power, control of resources, and economic and cultural influence in the world by powerful US actors, including but not limited to territorial conquest (Lutz 2006; Stoler 2006). Crucially, inequality across the Americas driven by US imperial formations inflects everyday life “not only for the ‘foreign’ subjects of US domination, but for the US citizens who benefit from it, who are subjugated to it, and who resist it” (Kaplan 1993: 14; drawing from Williams 1980). In this sense, United States efforts to restrict the mobility of Central American people is bound up with the privileges that US citizens like myself enjoy, including to travel freely across the very same borders. My analytical point of departure is that the culture of United States empire shapes the knowledge produced about Central American displacement since 2014, and the solutions proposed and enacted. In this dissertation I seek to tell a different story of displacement – scripting US actors and institutions into a position of entanglement, and thereby abandoning any myth of disconnection and benevolence (Razack 1998; Nevins 2016; Loyd, et al 2012; Walia 2013). An awareness of this history is crucial to formulate advocacy efforts that do not simply reinforce depoliticizing narratives or oppressive social norms (e.g. in crafting arguments in defense of Central Americans’ access to asylum, or in favor of “family unity”). For families that have been historically denied the right to be together, and for communities whose mobility has long been restricted, a carefully-crafted demand for “family unity” or for access to immigration status, for example, can be quite radical. As Lisa Marie Cacho (2012: 141) argues, reading “the contemporary immigrant rights movement as part of the still-ongoing international rebellion 28  against imperialism, rather than as an emergent movement solely against United States immigration and deportation law” links the struggles of distinct groups marginalized by related systems of oppression (e.g. white supremacy). Tracing specific expressions of white supremacy, as they relate to colonial conquest, invites an analysis that is empirically-grounded, but also attentive to trends across time and space (Bonds & Inwood 2016). Recognizing these links sets the groundwork for inclusive advocacy strategies that leave no one behind.   To give an overview of the structure of this dissertation, in Chapters 3 and 4, I find that forced separation harms families by threatening their well-being and access to asylum. I conceptualize this practice as a form of racialized governance. In Chapters 5 and 6, I demonstrate that detention throws countless hurdles into the path of asylum seekers, while adjudication tends to feminize, depoliticize, and thereby reject common Central American claims. I suggest the political nature of conflict in El Salvador defies such depoliticizing asylum narratives, demanding a more complex analysis. Although President Obama’s final term ended in January 2017, his amplified deterrence tactics largely endure, while displacement from Central America continues at a high pace (see US DHS 2016). Critical analyses of how this harmful response came into being can productively inform struggles for more dignified and caring treatment of displaced people moving forward. By foregrounding the legacies of colonial conquest and related systems of white supremacy in this dissertation, I seek to contribute to advocacy efforts to unsettle the historical continuity that has made the Trump administration’s dragnet, overtly-racist deportation agenda possible – a point I explore in Chapter 7. I now turn to Chapter 2, which delves further into the conditions of possibility for today’s deterrence paradigm, including its connection to a long history of United States interventionism in Central America.  29  Chapter 2: A history of North-South entanglement  One way that systems of white supremacy produce hierarchies between people is by mapping racialized identities onto bodies and places. As Charles W. Mills explains (1997: 48, emphasis original), “part of the purpose of the color bar/the color line/apartheid/jim crow is to maintain these spaces in their place, to have the checkerboard of vice and virtue, light and dark space, ours and theirs, clearly demarcated.” Such performances of spatial and social separation have historically allowed white people not only to lay claim to the lands and resources of their choosing through colonial conquest, but also to craft their identities. This is clear in the Anglo-Saxonist thinking popularized in the nineteenth century, which imagined North America as a space of cultural superiority and racial purity in relation to Latin America. Although invoked in less overtly-racist ways today, this imaginary has had an enduring effect on the US academic and foreign policy orientation towards its southern neighbors (Berger 1995; Stepan 1991). Yet always brimming beneath the surface of these careful delineations of social and spatial separation is the fragility of racial purity – threatened by figures at the margins, and by the intimate connections that inevitably entangle people across lines of difference (Stoler 1995; Anderson 2007; McClintock 1995; Heron 2007).  Migration and border scholarship that is attentive to colonial legacies has critiqued the performance of spatial and social separation that is common to migrant-receiving nations of the Global North. Harsha Walia (2013) sees this performance as an expression of “border imperialism,” while Joseph Nevins (2016) connects it to “global apartheid:” racialized regimes of mobility and labor hierarchies that lock-down borders (for certain people, though not for trade or investment) without creating conditions that give people meaningful choices about whether to 30  move or stay. The harms that unauthorized migrants face en route, and their criminalization and precarious labor status upon arrival, are treated as a product of poor personal choices. This framing obscures the role of Global North actors in the “destruction of means of subsistence” that tends to spark displacement from the Global South, and splinter kin and community across continents as a result (Tadiar 2015: 150).  The myth of North-South disconnection allows nations of the Global North like the United States to craft a benevolent identity for themselves – as uninvolved, not obligated to accept displaced people, and even as charitable when they do (Razack 1998; Walia 2013). It is important to note that a number of critical scholarly traditions focused on ongoing colonial and imperial histories do conceptualize the fates of the Global North and South as intertwined. Some notable examples include dependency theory, world systems theory, postcolonial theory, and coloniality/modernity theory. Yet the insights about North-South connections generated by this scholarship tend to be absent from US policy and media analyses of Central American displacement (a point I will elaborate in Chapter 4). Further, as I suggest in Chapters 5 and 6, the US asylum system limits how claimants can narrate their experience of displacement, with the unfortunate side effect of reinforcing an imaginary of the United States as superior and disconnected from the countries they fled. In this chapter, my goal is to provide a counternarrative to this imaginary. I argue that the US deterrence paradigm is one piece of a larger, disingenuous performance of social and spatial separation from the inequalities that displace people from Central America today. To make this argument, I trace a history of US militarism across the Americas, beginning in the nineteenth century, though mainly focusing on the transition from the Cold War into the present. I provide examples from across Central America, with closest attention to El Salvador to provide some 31  background for Chapter 6, which focuses on my research there. Following Jim Glassman (2005: 1541), a central point of this analysis is that “militarism is an unsurprising and routine handmaiden of capitalist development.”   2.1 Cold War militarism Before turning to the Cold War period, it is important to mention that United States companies and investors have long coveted the resources, labor, and markets to the south – an interest the US government declared explicitly in its 1823 Monroe Doctrine. During the nineteenth century, “capitalists poured billions into the region, first in mining, railroads, and sugar, then in electricity, oil, and agriculture” (Grandin 2006: 17). The return on these investments helped some of the largest US corporate players, like the Rockefellers, to build their fortunes. Like the Evangelical missionaries of the time, these companies also sought to spread United States values, such as individualism and consumerism, driven by a racist belief in the biological superiority of Anglo Saxons and a sense of hemispheric entitlement. In US company towns in Latin America, workers “were to produce not just sugar, bananas, or ore but, as in Ford’s Amazon endeavor, self-disciplined American-style workers” (ibid: 18-19). Despite the “civilizing” goals of some of its proponents, the entrenchment of United States economic and cultural influence across the Americas was not peaceful. Rather, it was facilitated by overt militarism, including upwards of 6,000 gunboat invasions of Latin American ports by 1930, not to mention the US interventions that established Panama as a nation and gave the US control over the Panama Canal in the early twentieth century, and the outright territorial conquests of northern Mexico in 1848 and Puerto Rico in 1898 (Grandin 2006: 3). Nonetheless, Latin American elites tended to welcome foreign investment and expanding trade relationships 32  that allowed them to build their wealth. These North-South connections also helped facilitate the growth of small middle classes across Latin America by generating new jobs, such as work in government posts and in the import/export business (Chasteen 2006: 195).  The benefits of North-South foreign investment and trade did not trickle down evenly. They were filtered through the social hierarchies that endured from Spanish colonialism into independence in Central America, while setting the stage for violently-enforced systems of capitalist development that continued to marginalize indigenous peoples, Afro-descendants, and peasants (see Paley 2018). In El Salvador, for example, a small number of families got rich in the late nineteenth century through coffee production and export. This oligarchy first built its wealth by displacing peasants from communal land, including from the fertile volcano slopes where coffee shrubs grow well, and converting it into privately-owned plantations. With the onset of the Great Depression, these wealthy growers slashed workers’ wages, spurring an uprising of indigenous peasants led by communist leader Agustín Farabundo Martí, which killed an estimated 20 to 30 elite growers. In 1932, the Salvadoran military responded by massacring as many as 30,000 civilians, who soldiers profiled as indigenous or as peasants (Binford 1996: 28-33). As this brutal attack unfolded in El Salvador, the terms of US interventionism towards Latin America were shifting. The United States faced the ire of powerful nationalist movements like the Mexican Revolution (1910-1920), and the anti-imperial armed struggle led by Augusto Sandino in Nicaragua against the US Marines and their local allies (1927-1933). In 1933, US President Franklin Roosevelt launched his Good Neighbor Policy, which pledged to end outright military interventionism in Latin America. Instead, he sought to maintain influence through “soft power” tactics like trade, regional agreements, and foreign aid (Grandin 2006: 27-39). These 33  tactics to maintain hemispheric dominance have lived on, although the Cold War also created the opportunity for the US to renew its military influence in Latin America, as I explain next. 2.1.1  “Low-intensity” warfare As World War II wound down, the United States government deepened its relationship with Latin American armed forces. In 1946 it established the School of the Americas in the Panama Canal Zone. By the early 2000s, this training facility, in its various locations and guises, had trained at least 60,000 Latin American soldiers. Cold War graduates, who were instructed in counterinsurgency doctrine, include the perpetrators of some of the most vicious political violence of the era (Gill 2004).  The US military developed its counterinsurgency doctrine with the goal of preventing Cold War-era revolutionary movements from arising or spreading. It was deployed in force in Central America in the 1980s. As Timothy Dunn (1996) explains, the US military labeled these tactics a type of “low-intensity conflict,” as they did not require wide deployment of US troops, and therefore posed a low risk to the lives and well-being of its soldiers. The intensity, of course, was not low by any means for the Central American populations targeted for terror, torture, disappearance, death, and displacement. The objective of this warfare strategy was to preemptively ‘defend’ the nation from revolution by asserting social control over the population. These efforts combined coercive force with measures that on the surface seem milder, like economic aid, but were equally an assertion of social control. In its often-indiscriminate pursuit of “internal enemies” to the nation, low-intensity warfare blurred the lines between the military, police, and paramilitary death squads (ibid: 19-25). El Salvador provided a key testing ground for US counterinsurgency doctrine. By the 1970s, growing inequality had converted the country into “a social pressure cooker” (Chasteen 34  2006: 302). Since the 1932 massacre, a series of military-led governments that were closely aligned with the US had worked to preserve the unequal status quo in the country. Yet people were increasingly organizing for change in El Salvador, including through an armed uprising. A coalition of guerilla forces united as the Farabundo Martí National Liberation Front (FMLN, after the communist leader of the 1932 uprising). They took up arms to demand social reform, and as defense against the brutal violence being perpetrated by the Salvadoran state. From 1980 to 1992, the FMLN fought to hold ground in the countryside, while the army sought to eliminate them, wielding United States funds, equipment, and training (ibid: 302-305). The victims of the wartime political violence in El Salvador ranged from Archbishop Oscar Romero, who was shot and killed by a death squad operative while giving a sermon denouncing state violence in 1980, to the more than 1,000 people massacred by the Salvadoran military in the department of Morazán in 1981. The massacre at el Mozote, as it is referred to after the rural town in which half of the assassinations took place, was perpetrated by a Salvadoran military battalion trained and armed by the United States. Despite the eyewitness accounts of survivors, and careful documentation and reporting by journalists, the US embassy denied that the massacre had even taken place until a United Nations (UN) Truth Commission confirmed it more than a decade later (Binford 1996). Among the thousands of cases of torture, disappearance, and murder that it documented in El Salvador over the course of the war, the UN Truth Commission reported that the armed forces and death squads were responsible for 85 percent of the violence, while 5 percent was attributed to the FMLN (Americas Watch 1993: 4). The report concluded that the bloodshed committed by Salvadoran state actors “originated in a political mindset that viewed political opponents as subversives and enemies,” while the indiscriminate targeting of people in rural 35  areas was a “deliberate strategy of eliminating or terrifying the peasant population in areas where the guerillas were active, the purpose being to deprive the guerrilla forces of this source of supplies and information” (cf ibid: 16). At least 70,000 Salvadoran people were killed or disappeared in the 12-year conflict (Meyer, et al 2016: 2-3). As I will elaborate later in this chapter, the armed conflict in El Salvador also led to massive displacement, with the US as a key destination, while fostering an increased reliance on remittances from Salvadorans in the diaspora.  The United States government also directly influenced the course of Cold War-era Guatemalan politics. In 1954 it sent a proxy force from Honduras to help Guatemala’s military overthrow its leftist president. President Jacobo Arbenz had threatened US interests by standing up to foreign companies and investors like the United Fruit Company when he seized and redistributed large landholdings to peasants. His overthrow ushered in an era of political violence under a series of post-coup regimes (Chasteen 2006: 260-261). The violence intensified in the 1970s and 1980s, as the Guatemalan military sought to eliminate a leftist guerilla uprising, while targeting civilians indiscriminately – particularly indigenous people in rural areas (ibid: 298). Following a 1994 peace agreement, a UN Truth Commission reported that at least 200,000 Guatemalans were murdered or disappeared in the conflict. As in El Salvador, government actors were responsible for most of the violence (Meyer, et al 2016: 2-3), which also resulted in a great deal of displacement and growing reliance on remittances from Guatemalans in the diaspora.   The US government similarly helped install a dictatorial regime in Nicaragua. Anastasio Somoza was at the helm of the National Guard when he orchestrated the 1934 killing of Augusto Sandino, who had led a guerilla war against the US Marines. Somoza took control of Nicaragua, and his family did not let go for four decades. During the 1960s and 1970s, the Sandinista 36  National Liberation Front (FSLN, named after Sandino) fought to overthrow the Somozas, and succeeded in 1979. In response, the Reagan administration sent the Central Intelligence Agency (CIA) to Honduras to support the Somoza-led counterrevolutionary force. After a decade of attacks and efforts to undermine the Nicaraguan economy, tens of thousands of deaths, and with a still-polarized population, the armed conflict ended with the Sandinistas’ loss of the 1990 Nicaraguan presidential election (Chasteen 2006: 299-301). As mentioned, the United States government made Honduras a staging ground for its proxy wars in Central America. This cozy relationship allowed the Honduran military to grow in size and strength, and to wage its own counterinsurgency operations at home. For instance, School of the Americas graduate Gustavo Álvarez Martínez founded the infamous Honduran military death squad Battalion 3-16 in the late 1970s, which pursued civilians critical of socio-economic inequality, with the support of the Federal Bureau of Investigation (FBI) and the CIA (Gill 2004: 86). The increasingly powerful Honduran military thus engaged in a violent strategy of social control over its own population with impunity (Pine 2008: 50-55).  In sum, during the Cold War the United States government trained Central American armed forces in its counterinsurgency doctrine – a massively-destructive form of so-called “low-intensity” warfare that used terror to consolidate social control over the population, whose brutal effects persist.  2.1.2 Deterrence strategy Just as the Reagan administration backed political violence across Central America, it also experimented with ways to “deter” people displaced by the very same conflicts from ever setting foot in the United States. Deterrence strategy is a border enforcement paradigm that the United States executive branch first developed in the Caribbean in response to the arrivals of people 37  displaced from Haiti and Cuba on the Florida coast during the late 1970s and early 1980s (Mountz and Loyd 2014; Loyd 2014). The Carter administration’s 1978 Immigration and Naturalization Service (INS, the predecessor of DHS) Haitian Program required that all Haitian asylum seekers who arrived on US shores be detained. The Reagan administration took this a step further by requiring that asylum seekers in general be detained. It also began carrying out credible fear interviews (meaning the initial review of asylum eligibility) on US Coast Guard boats, and summarily returning the people who received a negative outcome on their interview to their countries. The Bush and Clinton administrations continued this practice, and also established offshore detention centers in the Caribbean, including at Guantánamo. By the mid-1990s, deterrence in the Caribbean region drew from a mosaic of strategies designed to prevent people from arriving on United States shores (Mountz & Loyd 2014: 393-394). It is important to note that asylum seekers have never been treated uniformly in the US. Under the logic of the Cold War, “Communist states have been presumed guilty of human rights violations, such that anyone who flees such a country is a ‘defector’” (Coutin 2001: 81). Thus, the asylum system developed in a way that was more receptive to people fleeing left-wing regimes like Cuba or Nicaragua than to those fleeing right-wing regimes such as Haiti, El Salvador, or Guatemala (ibid). For instance, from 1966 to 2017, Cubans who stepped foot on US soil were automatically granted humanitarian protection under a policy known as “wet-foot/dry-foot” (USCIS 2017, January 12: 2). Yet the 1980 arrival of 125,000 Cubans by boat created tension around this Cold War policy. At issue was that President Fidel Castro was said to have encouraged the mass departure of marginalized groups like “people with criminal records, mental health issues, and gay people” (Mountz & Loyd 2014: 392). Jonathan Simon (1998: 582-583) explains the tension that grew towards these marginalized Cuban asylum seekers as a 38  product of the racist and classist analysis of poverty taking root in the US in the late twentieth century: “the new nexus was not the superpower rivalry, but the mass of poor (both domestic and foreign) perceived as a dangerous class whose unconstrained needs and desires threatened to overwhelm the nation.” The threat, according to policymakers at the time, was that allowing Caribbean asylum seekers to enter the country with ease and recognizing them as refugees would send the wrong message. It would roll out the welcome mat, encouraging more people to come (Kennedy 1981: 142).  The US executive branch instead sought to send a message of “deterrence.” It would accomplish this not only by physically blocking people from stepping foot on dry land, but also by locking-up those who did arrive (Simon 1998: 584). The logic behind deterrence is that “to penalize those who seek to bypass the legal routes of entry into the United States is to honor those who pursue them and to treat the broader community of immigrants as rational actors who respond to incentives” (ibid: 601). As Simon (1998) contends, among the issues with this logic is that it does not account for the extreme conditions under which people fled their countries, including from the US-funded “low-intensity” conflicts raging across Central America. Further, if the imprisonment of people who entered the US irregularly was crucial to national security, why did the INS, in practice, selectively target particular nationalities for detention?   During the 1980s and early 1990s, the principle targets for indefinite detention (lasting days, weeks, or months) were people from Central American countries arriving in South Texas, whom the then-INS Commissioner accused of presenting frivolous asylum claims; people from Haiti arriving in Florida; and people from China, who faced accusations of entering the United States with fraudulent documents (US GAO 1992: 35-37).  39  In 1985, a coalition of immigrant rights organizations and religious groups sought to end a clear pattern of discrimination, in which the vast majority of asylum claims from El Salvador and Guatemala were being denied, by filing a lawsuit against the US government (American Baptist Churches v. Thornburgh 1991, known as the ABC case, Coutin 2007: 53-55). The ABC lawsuit was finally settled in 1991. It placed some restrictions on INS detention practices towards Salvadoran and Guatemalan citizens and created a process of reevaluation for the asylum claims that had been unfairly denied (US GAO 1992: 32). Yet somehow, the ABC settlement seemed to give the INS little pause about the ethics or efficacy its detention practices. Rather, the INS deemed the discriminatory deterrence strategies it had been using towards Salvadorans and Guatemalans a success, and called for more detention bed space “to gain better control over the flood of aliens entering the country illegally,” as the US General Accounting Office put it in 1992 (US GAO 1992: 4). This has become a familiar refrain in the years since: a declaration of the need for more detention beds, to lock-up an ever-growing population of noncitizens targeted by the INS, and later by DHS, for deportation.  The US-Mexico border has been undergoing a process of militarization since the late 1970s, which drew inspiration, in part, from the low-intensity warfare techniques tested in Central America (Dunn 1996). In 1994, the Border Patrol launched a National Strategic Plan centered on “prevention through deterrence.” It aimed to better enforce the US-Mexico border by stationing personnel and infrastructure in the areas that are easiest to cross (Haddal 2010; US OIG 2000). This pushed people into “more remote and less accessible locations in mountains, deserts, and untamed sections of the Rio Grande” (Massey 2005: 6). Unauthorized migrants became more vulnerable to death en route, yet also faced lower odds of being caught (Cornelius 40  2001; Massey 2005; US GAO 2001). Despite its tragic outcomes and apparent ineffectiveness at achieving its stated mission, the scope of deterrence strategy has only expanded. The Department of Homeland Security was established in 2003, in the fallout of 9/11. DHS moved away from the longstanding INS norm of releasing (rather than detaining) most deportable noncitizens who it apprehended, to await their hearing before an immigration judge. DHS adopted a new norm of “catch and remove” (Martin 2012a). To this end, the agency expanded its use of “expedited removal:” a fast-track deportation process carried out by DHS with little oversight. DHS began applying expedited removal to people it apprehended within a wide, 100-mile zone along the US perimeter, and to those caught through interior enforcement less than fourteen days after they arrived in the country. The agency folded “Other Than Mexicans,” such as Central Americans, who logistically cannot be deported quickly, into this dragnet. This created a rationale to demand more detention bed space (ibid: 322).  To illustrate the growth in detention over recent decades, on a typical day in 1994, about 5,000 noncitizens were detained in the United States. By 2001, that daily average reached 19,000 people, and then surpassed 34,000 people by 2014. From 1990 to 2014, the number of people deported annually from the US also skyrocketed, from 30,039 to 407,075 people (Zong and Batalova 2017). Lauren Martin (2012a) theorizes that beyond a logistical effort to deport noncitizens, one goal propelling this massive growth in detention and expedited removal was to create a spectacle to deter people who might consider migrating in the future.   2.2 Peacetime militarism The Cold War-era armed conflicts and their afterlives have displaced a vast number of people from Central America. Many fled to the United States, due in part to existing family and 41  community connections there (Menjívar 2000). From 1980 to 1990, the Central American immigrant population in the US rose from 354,000 people to more than 1.1. million. By the year 2000, this number surpassed 2 million people, growing to more than 3 million by 2010. Among the 3.4 million Central American immigrants who lived in the United States by 2015, about 85 percent were estimated to originate from El Salvador, Guatemala, and Honduras (Lesser & Batalova 2017). It is important to consider – why has migration to the US escalated after the wars ended? For one, countless families were split up across borders by the conflicts, and sought to reunite in the years after. Natural disasters in the ensuing years also spurred further displacement. Finally, and crucially, economic inequality and violence have endured, and in some cases even worsened, following the formal declarations of peace (ibid). The armed conflicts across Central America ruptured the livelihoods and social support networks of countless communities. A deep sense of distrust, and dynamic of impunity endured in the aftermath. Political violence had systematically dismantled collective organizations like cooperatives and unions, through which people had previously organized themselves to make demands of powerful actors like employers, state entities, and international organizations (Feitlowitz 1998; Menchú 1984, cf Gill 2004: 14). In this climate of vulnerability, neoliberal economic policies were introduced by governments across the region that “mandated lower tariff barriers, cut social services, privatized public utilities, aggravated unemployment, and increased the gap between rich and poor” (Gill 2004: 14). The United States was keen to push this neoliberal agenda in the aftermath of the Cold War, for example, by negotiating free trade agreements, and by encouraging the adoption of “light industry” like garment production in Central America and the Caribbean. As Cynthia Enloe (1993: 103) notes, this political maneuver demanded a “feminization of cheap labor” that 42  reinforced labor hierarchies between the Global North and South, and also within the countries that adopted light industry. I next explain how neoliberal reforms played out in El Salvador, and their relationship to the maintenance of a peacetime militarism. I foreground the role of United States foreign policy and punitive immigration reform in this shift.  2.2.1 Insecurity in El Salvador A United Nations-organized peace process officially ended the armed conflict in El Salvador in 1992. The Peace Accords democratized key public institutions – by reducing the military’s scope, creating a new National Civilian Police (PNC) force, reforming electoral processes, establishing a Human Rights Ombudsman’s Office, and more. Through the peace process, the FMLN transitioned from an armed resistance group to a recognized political party. Yet what remained largely unresolved by the Peace Accords was the economic inequality that has long polarized Salvadoran society, and had helped fuel the armed conflict in the first place. The main issue was that “at the negotiating table, the post-conflict government would not concede to economic demands, namely the redistribution of land and wealth away from the hands of a few elite families” (Pineda & Stoumbelis 2017: np; also see Binford 2010).   At the helm of El Salvador’s post-conflict government was the far right-wing Republican Nationalist Alliance (ARENA) party. ARENA was co-founded in the early 1980s by infamous death squad leader Roberto D’Abubuisson, who orchestrated the assassination of Archbishop Romero. During the twenty consecutive years that ARENA held the presidency, from 1989 to 2009, its leadership privatized key public industries and slashed crucial forms of state support like agricultural subsidies, shifting the country from an exporter to importer of food staples like rice and beans. As job opportunities shrunk, people continued to emigrate from El Salvador (Zilberg 2011: 6; Binford 2010). Remittances from Salvadorans in the diaspora, as well as 43  income from informal and extralegal activities like extortion within El Salvador, became increasingly important sources of livelihood in the void of economic opportunities (Zilberg 2013: 228). ARENA found an ally in the United States, which also pushed for neoliberal reforms, for example through the Dominican Republic-Central American Free Trade Agreement (CAFTA-DR), implemented in 2006. The Congressional Research Service reports that today, the US government continues to push its “strategic interests” in Central America by “encourag[ing] the spread of neoliberal economic policies and the consolidation of democratic governance” (Meyer & Ribando Seelke 2015: 15-16).  After the Peace Accords, violent death rates rivaled those of the war in El Salvador (Cruz & González 1997, cf van der Borgh & Savenije 2015: 157). The violence and harm that followed were often attributed by the Salvadoran state and media to criminalized young people, or written-off as accidental (Bourgois 2001; Zilberg 2007). The state’s partial retreat through privatization under ARENA leadership meant that it could not as easily be implicated in the population’s suffering as during the war (Moodie 2006: 74). At the same time, punitive immigration and criminal justice reforms taking place in the US during the mid-1990s made Salvadorans increasingly vulnerable to deportation. The deportation of marginalized young people, combined with the climate of social inequality in El Salvador, allowed a structure of gang activity familiar to California to take root in low-income communities around El Salvador (UNHCR 2016: 4-5). The response was largely repressive – ranging from assassinations conducted by death squads, to mano dura public security strategies implemented by the Salvadoran state. As in other Central American countries, El Salvador’s war-time death squads did not entirely disappear after peace was declared. In the mid-1990s, for example, a death squad called La Sombra Negra [the Dark Shadow] executed alleged gang members as part of a “social 44  cleansing” of El Salvador’s department of San Miguel. Human rights advocates continued to document death squad activity into the 2000s (Gutiérrez 2007). As I will explain in Chapter 6, these types of extrajudicial executions persist today in El Salvador. The people targeted are not limited to alleged gang members. For instance, government officials and human rights advocates have been targeted as well (Bargent 2014; Gutiérrez 2007).  Meanwhile, the Salvadoran state first deployed mano dura public security strategies in the mid-1990s, and then aggressively expanded them in the early 2000s. This approach came about through an exchange of punitive security theories, policies, and practices across the Americas. For example, the zero tolerance policing tactics popularized by Rudolph Giuliani and Police Chief William Bratton in New York City during the 1990s have been hugely popular among Latin American leaders, including in El Salvador (Swanson 2013). The US government remains deeply involved in Salvadoran security policy in official capacities as well, for instance, establishing an International Law Enforcement Academy (ILEA) in San Salvador to train public security agents, which critics compare to the notorious School of the Americas (Enzinna 2007). In contrast to neoliberal philosophy’s promise to minimize the state, in practice neoliberal policies between the US and El Salvador have been defined by the mutual growth of a punitive policing apparatus (Zilberg 2013: 229).  In 2008, the US government launched the Central America Regional Security Initiative (CARSI), which promised to combat drug trafficking, gang activity, and other forms of organized crime throughout the region (Meyer & Ribando Seelke 2015: 1). Critics argue that CARSI has expanded a failed, US-funded strategy first tested in Colombia and then Mexico that militarizes public security in the name of combatting crime. In those two countries, this model 45  has yielded increased violence, and facilitated neoliberal reforms, while failing to achieve its stated goals (in the case of Colombia and Mexico – to decrease drug trafficking) (Paley 2014).  2.2.2 Perpetual temporariness A punitive interplay between immigration policy and criminal justice policy has been underway in the United States since the 1980s – driven by the domestic “drug war” and xenophobia (García Hernández 2014: 1414). A series of laws have broadened the types of criminal offenses that put noncitizens at risk for deportation, criminalized certain forms of unauthorized entry to the country, and enrolled state and local law enforcement agencies into deportation practices. Some of the most impactful laws include the 1986 Anti-Drug Abuse Act (ADAA) and its 1988 modification, the 1996 Antiterrorism and Effective Death Penalty Act (AEDPA), and the 1996 Illegal Immigration Reform and Immigrant Responsibility Act (IIRIRA) (ibid).  These federal laws, alongside the zero tolerance policing practices adopted in major cities around the same time, led to increased arrests and incarceration of noncitizens who had committed minor offenses or were involved with gangs – which is what made certain Salvadorans vulnerable to deportation in the 1990s (Zilberg 2007). Key to this vulnerability is that many Salvadorans living in the US have long been denied a pathway to citizenship, and being a noncitizen leaves one perpetually at risk of deportation. As mentioned, Salvadoran and Guatemalan asylum claims were systematically denied until the settlement of the ABC lawsuit. Yet the reevaluation of asylum claims demanded by the ABC settlement was slow-moving, and remained largely unsettled until the Nicaraguan Adjustment and Central American Relief Act (NACARA) was passed in 1997, finally providing a pathway to citizenship for certain asylum seekers. In the meantime, people with precarious status became vulnerable to deportation (Coutin 2011: 585).  46  Political pressure from the migrant justice movement pushed Congress to extend Temporary Protected Status (TPS) to citizens of El Salvador in 1991. By 2014, approximately 212,000 Salvadorans held TPS (Messick & Bergeron 2014). This status provides protection from deportation for people displaced by conflict or natural disaster. TPS recipients have access to a work permit, but not a path to permanent residency or citizenship, and their status must be renewed every 18 months. Their temporary status means that recipients cannot bring relatives to join them. Recipients must maintain continuous residence in the United States, and if they travel back to their country may be denied reentry. The result is that many families have faced extended separation, and remain vulnerable to deportation, given that the executive branch holds the discretion to revoke their temporary status at any time (Menjívar 2006: 1012-3).  Cecilia Menjívar (2006: 1018) uses the term “liminal legality” to describe the perpetual temporariness and its toxic effects inflicted upon the Salvadoran diaspora in US, while Bailey and colleagues (2002: 127) argue that this uncertainty creates a disciplining power that extends across borders. This argument still rings true. In 2017 the Trump administration revoked TPS status for citizens of Haiti and several other countries (US DHS 2017). In the first few days of 2018, the administration revoked TPS status for El Salvador as well, giving Salvadoran TPS holders a September 2019 deadline to leave the country (Miroff & Nakamura 2018). Migrant justice advocates argue that this move will be profoundly destabilizing not only to TPS holders, but also to family members in the US and in El Salvador. It will likely be destabilizing to El Salvador on the national scale as well, given that as of 2016, remittances sent to El Salvador amounted to $4.58 billion, or 17.1 percent of El Salvador’s gross domestic product (Welsh 2017). 47  2.3 Conclusion As the historical context presented in this chapter demonstrates, the United States is deeply entangled in Central American affairs. During the Cold War, the US armed and trained militaries across the region in “low-intensity” warfare techniques in the name of anticommunism. This agenda propelled brutal political violence and stymied efforts to affect social change. William Robinson (2003, cf Moodie 2006: 66) argues that the geopolitical imperative of US Cold War-era intervention in Central America was to keep the region open for capitalism. This goal was certainly achieved in postwar El Salvador, where far right-wing leadership enacted aggressive neoliberal reforms, with the support of US allies. People displaced by economic disenfranchisement, and by the renewed conflict and militarization of public security in El Salvador that accompanied it, continue to move to the United States, and continue to support their loved ones by participating in the $4.58 billion/year remittance economy (see Welsh 2017). The violence of “low-intensity” warfare has continued to haunt people displaced from Central America during the wars and in the years since. Its tactics of social control have animated the militarization taking place at the US-Mexico border in recent decades (Dunn 1996). Deterrence strategy, the border enforcement paradigm built on this foundation, promises to prevent its targets from arriving in the US, or from leaving their communities in the first place.  At first glance, deterrence strategy’s efforts to discourage unauthorized migration sit in tension with the longstanding desire among some US employers for a disenfranchised noncitizen workforce. From its inception in the Chinese Exclusion era, to the mid-twentieth century Bracero Program that intermittently invited Mexican workers then subjected them to mass deportation, the idea of the “illegal alien” has always been tangled up with efforts to control labor (see Ngai 2004). As Nicholas De Genova (2002: 438; also see Kanstroom 2007) explains, categories of 48  “illegality” do much more than facilitate deportation: “It is deportability, and not deportation per se, that has historically rendered undocumented migrant labor a distinctly disposable commodity.” Deterrence strategy often fails to discourage migration as promised, suggesting that this enforcement paradigm has perhaps continued to expand, in part, because it serves other purposes. Apprehending and deporting every person who violates US immigration law would be logistically unfeasible and undesirable to the employers and consumers who benefit from a workforce with limited rights. Josiah Heyman’s (1995: 261) ethnography of INS officials demonstrates that no one may be more aware of this contradiction, which “balances publicly visible arrests and invisible but effective perpetuation of undocumented labor migration,” than the officials themselves.  Another apparent, and related, contradiction worth noting is the entrenchment of militarized policing and immigration enforcement tactics within the US and El Salvador during a neoliberal era. This seems like a contradiction because neoliberal philosophy demands a shrinking role for state institutions and expenditures, but in practice, a punitive transnational security apparatus has only grown. As Elana Zilberg (2013: 230) finds, ‘neoliberal security’ regimes can serve capitalist interests by maintaining a population of disenfranchised workers.  Ultimately, the frequent exchange of militarized tactics of social control between powerful actors in the US and in Central America, and their often-aligned efforts to keep the region open for capitalism, defies any myth of North-South disconnection. This dynamic exemplifies the concepts of border imperialism and global apartheid mentioned at the outset of this chapter – meaning the maintenance of an unequal, racialized regime of mobility that spans the Americas (Nevins 2016; Walia 2013). The forced separation of Central American families arriving in the US today that I describe next in Chapter 3 stems directly from this history.  49  Chapter 3: Divided by detention: Asylum-seeking families’ experiences of separation  In 2014, the number of asylum-seeking families from Central America (and from Mexico) arriving in the United States soared. Many people coming then, and those who continue to arrive, express a fear of returning to their countries, and a desire to pursue humanitarian protection in the form of asylum, withholding of removal, or relief under the Convention Against Torture (UNHCR 2014; 2015; US EOIR 2016). Between 2008 and 2015, the United Nations High Commissioner for Refugees (UNHCR) documented a dramatic increase in asylum applicants from El Salvador, Guatemala, and Honduras—by roughly fivefold in the US, and by thirteenfold in Mexico and other Central American countries (UNHCR 2015). In the United States, the Obama administration did not roll out a welcome mat. It did the opposite – throwing asylum seekers, including entire families traveling together, into fast-track deportation processes, and aggressively expanding family detention in an attempt to “deter” more people from coming in the future (ACLU 2015, February 20; Rosenblum & Ball 2016: 7, note 1; RILR v. Johnson 2015; US DHS 2014, November 20).  Families and advocates have exposed the numerous ways that detention and fast-track deportation jeopardize the well-being and due process of asylum-seekers, including families traveling together (e.g. Berks Mothers 2016; LIRS & WRC 2014; Sheppard & Murray 2017; USCIRF 2005; 2016; WCRWC & LIRS 2007). The issues abound – from the failure of Department of Homeland Security (DHS) officials to respond appropriately to families’ expressions of fear over returning to their countries and requests for humanitarian protection, to 50  the obstacles that detention creates to accessing legal counsel and due process, to the re-traumatization of families fleeing persecution who are held in detention centers that look and feel like jails (LIRS & WRC 2014; USCIRF 2005; 2016).   In this chapter, I examine what happens when “family detention” does not actually keep loved ones together. Through its custody determinations, DHS splits up family members – sending them to different facilities around the country, while failing to track and reunite those who arrive separately (LIRS 2016: 2). While DHS claims that family detention keeps families together, in practice, a mother and child who are sent to family detention will often have been separated by DHS from other loved ones with whom they fled, including husbands, fathers, grandparents, older children, and siblings. Minors who arrive with non-parent caretakers are often removed from their custody. I analyze how this separation occurs, as well as its impact on families’ well-being and ability to access humanitarian protection. To this end, I profile the experiences of five asylum-seeking families separated by detention. Tracing these separations reveals that the waiting and uncertainty inherent to the asylum process are only compounded for those who are released to continue their cases, but must worry about their detained and deported loved ones. This leads me to argue that forced separation extends the temporal and spatial bounds of confinement for all family members involved.  The detention of families arriving at the US-Mexico border is not mandatory. DHS has the discretion to place asylum-seekers directly into standard immigration court proceedings, instead of subjecting them to fast-track deportation and detaining them (AILA 2016: 12). Using this discretion would keep families together, free from detention. Despite this simple solution to the issue of forced family separation that I will describe in this chapter, DHS continues to detain and split up family members (LIRS 2017; Al Otro Lado, et al 2017; WRC, LIRS, & KIND 51  2017). I conclude this chapter with a policy recommendation. Separating families has countless negative impacts, while allowing them to stay together has numerous benefits. Doing the latter would allow the US government to better uphold its various commitments to family unity and parental rights in immigration-enforcement activities,7 support the well-being of families, give them more effective access to humanitarian protection, and prevent the unnecessary waste of government resources.  At the end of each section of this chapter, I explain in detail the experience of each research participant and her family. I begin below with Vanessa’s family. Vanessa’s family In the afternoons, when Vanessa’s 10-year-old daughter would come home from school, she always asked “¿Y papi? [And daddy?]” The two of them were waiting, staying with Vanessa’s mother-in-law in California, for her husband to be released from a Texas detention center. They were also waiting just for a phone call, as their attempts to purchase phone credit for him had been unsuccessful. Since Vanessa and her daughter were first separated from him in a Customs and Border Protection (CBP) holding cell, she did her best to distract her daughter from the uncertainty and sadness of waiting by giving her books to read. The family fled extortion and death threats from a gang in Soyapango, El Salvador. Although the gang specifically targeted her husband, Vanessa said the risk extended to the whole family:                                                  7 Among these commitments, this chapter focuses on the 1997 Flores settlement agreement, which prioritizes family reunification and the expedient release of minors from detention (Flores v. Reno 1997). Also see the 2004 memorandum of understanding that establishes that the unity of Mexican families should be maintained during repatriation from the US (US DHS & the Secretariat of Governance and Secretariat of Foreign Affairs of the United Mexican States 2004). The 2013 Immigration and Customs Enforcement (ICE) Parental Interests Directive instructs ICE personnel to avoid any unnecessary disruptions of parental rights through immigration enforcement actions (US ICE 2013). A 2014 DHS memorandum calls for discretion regarding the detention of “primary caretakers of children or an infirm person” (US DHS 2014, November 20). On US commitments to family unity under international human rights law see Ginatta (2014).  52  They were extorting my husband, and they would have killed him if he stayed. And then, if they didn’t do so immediately, they would kill everyone—they kill whole families. They rape the girls. I would not allow this for my daughter. I only have one daughter, and I have to seek the best for her.   In March 2016, Vanessa and her daughter had been released after five days in CBP holding cells, plus two weeks in the South Texas Family Residential Center (a long-term detention center for noncitizen women and their minor children). They continued to pursue their asylum case. Vanessa only learned where her husband was detained because a volunteer attorney, who assisted her while she was detained, located him through an online search. When we spoke, Vanessa believed that her legal case was still separated from that of her husband, even though she had specifically asked a government official to unite their cases. The family fled El Salvador together for the same reason, yet the initial outcome of their cases was uneven. Although Vanessa had no certainty that her asylum case would ultimately be granted, in detention her husband remained imminently vulnerable to deportation.   From Vanessa’s point of view, the separation from her husband was hard on the whole family, but her daughter suffered the most from being split up. “At least adults can bear this, or try to understand it,” Vanessa reasoned, but “in my case, my daughter is used to having both of us. She is very close with [her father]. At home, it has always been the three of us. It has been very difficult.”  3.1 Overview This chapter investigates one pernicious side of the Obama administration’s targeting of asylum-seeking families: family separation. There are two principal ways that arriving asylum-seeking relatives are split apart en route from El Salvador, Guatemala, and Honduras to the US: 53  1. Some families are split up during their journey to the US. The family members may arrive in the country separately, be apprehended separately by CBP, and be subject to different Immigration and Customs Enforcement (ICE) custody determinations.   2. Other families arrive in the US together. They are apprehended together by CBP, and then are split through the process of ICE custody determinations. In this chapter, I focus on the second way that families are divided: in the custody of CBP and ICE, which are two component agencies of DHS that carry out immigration enforcement. Yet, it should be noted that United States immigration policy and enforcement fractures families in a number of other ways. The fear of separation hangs over an estimated 9 million children in mixed-status families living in the US. “Mixed-status” refers to families made up of a combination of citizens and noncitizens who are vulnerable to interior immigration enforcement, detention, and deportation (Taylor et al 2011; also see HRW 2007).  Further, many of the Central American families arriving in the United States today are already fragmented across borders, particularly since the US-funded armed conflicts of the Cold War. This means that some family members have been living in the US, often for many years, while others have remained in Central America. Those with precarious legal status in the US (resulting from a long history of exclusion and liminal status at the hands of US immigration law) may be ineligible to petition for their family members to join them, or may face long waits—leading to prolonged separation or precluding reunification altogether (Menjívar 2006). This is to say that a heritage of United States interventionist foreign policy and punitive immigration policy, as outlined in Chapter 2, divides Central American families in a number of ways. Here, I focus on one very specific mode of separation that happens near the US-Mexico border. 54  3.1.1 Research methods During March and April 2016, I interviewed five women who had been detained with their children in the South Texas Family Residential Center, and released to continue their asylum cases. Four families originate from El Salvador and one from Honduras. I refer to research participants with pseudonyms. I interviewed participants over the phone (rather than in-person), given that they live all over the US – from Virginia, to Texas, to California. The participants are former clients of the CARA Family Detention Pro Bono Project, which generously made referrals for me, as explained in Chapter 1.8  My goal in the interviews was to learn about how and why families are split up in detention, and about the impacts on their well-being and legal cases. The interviews did not focus on the reasons why participants and their families fled their countries. This only came up in the interviews when I asked participants if their entire family had fled together, and for the same reason. Most participants then elaborated briefly on that reason, though not every participant did, and I deliberately chose not to linger on the topic. Thus, I do consistently provide this information about each of the five families.  My analysis is also aided by expert interviews with four attorneys highly experienced in representing detained asylum-seeking families, who I refer to anonymously. These were phone interviews, for the same reason mentioned above. I also draw empirical data for this chapter from federal court filings, Freedom of Information Act (FOIA) responses, other government documents, and advocacy reports. The American Immigration Council, one of the CARA                                                  8 As noted in Chapter 1, the views expressed throughout this dissertation, and any shortcomings, are my own. 55  partners, published my analysis as a report in August 2016 (Barrick 2016). This chapter is a revised and updated version of that report.  My goal in this chapter is not to generalize as to how family separation always works, but to provide insights into how it can happen and what the experience is like for some families. It is not easy to generalize about this practice, or even about its prevalence. Service providers and legal advocates have observed a growing number of separations near the US-Mexico border in recent years as families have been increasingly targeted by DHS for fast-track deportation. Advocates report that many families have filed reports about their separation with the DHS Office of Civil Rights and Civil Liberties, yet because DHS and other federal agencies involved in deportation do not have a system to track family separations, “quantitative data is not available and the full scope of the problem is unknown” (WRC, LIRS, & KIND 2017: 18, endnote 5). The analysis I present here is thus qualitative by necessity, and carried out with the referrals that CARA had the capacity to provide.  3.1.2 Analytical framework A contentious debate has simmered for decades in the United States about whether noncitizen children and families should be detained, and, if so, under what circumstances (Heidbrink 2014). The 1997 Flores settlement agreement mandated the expedient release of juveniles from noncitizen detention and “underscor[ed] the principle of family unity” (Jones & Obser 2015). This binding agreement, reached after more than ten years of litigation, sets out the conditions under which the then-Immigration and Naturalization Service (INS), now DHS, can detain noncitizen children. It establishes a “general policy favoring release” of minors (Flores v. Reno 1997: 9), and requires that the government “make and record the prompt and continuous efforts on its part toward family reunification and the release of the minor” (ibid: 12). It dictates “the 56  best interest of the child must be considered when contemplating family separation” (US ICE 2016, September 20: 6; also see US CBP 2008: 1). Flores litigation is ongoing, as Flores class members and their advocates push DHS to live up to the promises of the 1997 agreement. The Flores agreement came on the tail of the 1996 Illegal Immigration Reform and Immigrant Responsibility Act (IIRIRA), which lumped asylum-seekers in with a vast population subject to “expedited removal”—a form of fast-track deportation without an opportunity to present one’s case before the immigration court. When asylum seekers express fear of returning to their countries, however, the government has an obligation to refer them for a credible or reasonable fear interview with an asylum officer. These interviews provide an initial review of eligibility for asylum and other forms of humanitarian relief from removal, and thus the potential for subsequent review by an immigration judge in a full merits hearing (Campos & Friedland 2014).  The Flores agreement and DHS fast-track deportation practices represent conflicting agendas. In the name of the best interest of the child, the Flores agreement seeks to keep noncitizen children with their families and out of detention. In contrast, the goal of expedited removal is to detain and deport certain noncitizens, including asylum seekers, as quickly as possible. This tension underscores Alison Mountz’s (2004: 325) point that “‘the state’ does not contain or enact a unified series of agendas, objectives, or actors. State practices encompass, rather, a series of diverse interests and bodies that are often themselves in conflict.” Asylum-seeking families find themselves at the crossroads of this conflict. As described in Chapter 2, the turn of the twenty-first century saw a rapid expansion of fast-track deportation and detention in the United States, as part of a broader strategy to “deter” noncitizens, including families with minor children. To this end, DHS repurposed Hutto, a Texas 57  medium security prison, into a family detention center in 2006. Yet by 2009, litigation that exposed its inadequacy as a place to house children pushed DHS to convert Hutto into an adult facility, reducing the total number of family detention beds nationwide to under 100 (Libal, Martin, & Porter 2013: 253, 264). In 2014, the Obama Administration made “recent arrivals” a top enforcement priority, generally detaining families with the goal of “deterring” the arrival of others (Rosenblum & Ball 2016: p. 7, note 1; US DHS 2014, November 20). DHS opened new family detention centers in New Mexico, and then Texas, bringing the family detention capacity nationwide swiftly up from 96 to 3,326 beds (ACFRC 2016: 150-151). The five women who participated in this research and their children were all detained in the South Texas Family Residential Center: a 2,400-bed facility opened in December 2014, and operated by ICE and the private prison company CCA (CCA 2017C). It is located in Dilley, Texas, which is a rural area more than an hour drive southwest of San Antonio. Dotting the flat, dusty red landscape are sprawling clusters of trailers that make up the living quarters of the women and children locked inside the complex. This family detention center is not far from Crystal City, Texas – one of the sites where the United States government incarcerated families of Japanese descent during World War II. Internment camp survivors were quick to draw parallels between the everyday indignities and trauma they suffered, and that occurring within today’s family detention centers like the one in Dilley (Ina 2015; Takei 2015). After visiting noncitizen women and children detained in Dilley, Satsuki Ina, a mental health practitioner, professor emeritus, and internment camp survivor compared their trauma with that of her own family, noting (Ina 2015): When I discovered my mother’s prison diary 10 years ago, I learned that she lived with anxiety daily, not knowing how long we would be imprisoned. Would it be a few months? A few years? She never knew where we would be sent next. And she lived in fear of my father’s well-being 58  after he was sent to a separate prison in North Dakota. Finally one day she wrote, “I wonder if today is the day they’re going to line us up and shoot us.”  Without conflating two distinct modes of United States family incarceration that are not precisely the same, I find some striking overlaps. Ina’s family experienced anxiety and uncertainty from being imprisoned indefinitely for obscure reasons of “national security,” and from having their family forcibly, and carelessly, split apart. This resonates through the families’ experiences with contemporary detention that I profile in this chapter, who were atomized from intact families into individual detainees by DHS. For me, this atomization of families into individual subjects with differential access to rights is an expression of the individualism characteristic of liberal democracies like the US. Judith Butler (2004: 24-25) notes that liberal democratic legal systems accord rights to the individual, and at times to a whole class of people who unite to seek protection against discrimination (e.g. the children protected by the Flores agreement). Yet this framework “does not do justice to passion and grief and rage, all of which tear us from ourselves, bind us to others, transport us, undo us, implicate us in lives that are not our own, irreversibly, if not fatally” (Butler 2004: 25). Indeed, the isolation of rights to the individual, alongside the removal of individuals from family and community through of imprisonment, denies the realities of “human relationality” (McKittrick 2011; Gilmore 2007).  This is not to suggest that liberal democratic systems are incapable of recognizing and protecting social ties. On the contrary, as Elizabeth A. Povinelli (2006: 188-189, drawing on Habermas) explains, “the humanist subject was forged out of the intimate recognition that passed between two people in the conjugal household.” Although liberal democratic systems were built around the individual, the individual was made legible, in part, through the proper performance 59  of family life. The European Enlightenment thinking that gave rise to the US political system valorized a certain expression of love – namely, that between a European heterosexual husband and wife (and their children). Enlightenment thinking equated this expression of love with the highest form of civilization, while denigrating its alleged opposite: “a particular kind of illiberal, tribal, customary, and ancestral love” (ibid: 225-226). This exceptionality was built, in part, through European imaginaries of difference between themselves and colonized peoples (ibid: 215). In Chapter 4, I will elaborate this point further, and suggest that this racialized exceptionality – which only recognizes the social bonds of certain families – gets to the heart of what makes the forced separation of asylum-seeking families today politically possible. For now, I turn back to the Flores agreement, which was born out of a class action lawsuit, and thus granted certain rights to all noncitizen children in DHS custody – including the right to family unity and to expedient release from detention. The agreement grants no rights to adults (whether rights to be with their children, or to other loved ones like partners, siblings, etc.) (Flores v. Lynch 2016, June 7). As a result, public debate about family detention has long centered on concerns about the well-being of children, rather than adults (see Flores v. Sessions 2017; Martin 2011). In this chapter, my goal is to bring the fate of adults into the frame as well, by recognizing that the well-being of loved ones is intertwined. This interconnected nature of well-being comes through clearly in the experiences of Wendy’s family, described below. Wendy’s family In Sonsonate, El Salvador, Wendy says, she and her husband had a stable life. In addition to caring for her 10 and 5-year-old sons, Wendy worked and studied to be a teacher. This solid ground was thoroughly shaken when a gang began extorting the family. The final straw was the murder of her brother-in-law. Around October 2015, the family began a month-long journey to 60  the US, guided by coyotes [smugglers]. They traveled in a group, some days by bus, other days by car. It was a harsh trip, plagued by hunger and fatigue, with the ever-present threat of being apprehended by the Mexican authorities. At one precarious juncture, the coyotes split the group apart, placing Wendy’s 10-year-old son in a separate walking group. Wendy walked on with her husband and 5-year-old for a day and night. “Don’t worry, your son is in another group. You’ll be reunited when they pick you up,” the coyotes said. But by dawn, the story had changed. “Your son was caught by Mexican immigration,” they told her. Wendy says her first impulse was to turn herself in to the Mexican authorities so her son would not be alone. However, the coyotes would not let her. So the remaining family members carried on and were eventually apprehended together by CBP after they arrived in the United States. Wendy only later learned that her older son had been deported from Mexico back to El Salvador, where he now stays in hiding with Wendy’s parents. When he is not in school, he keeps himself locked in the house, too afraid to go out.  CBP further separated the family in its holding cells, sending Wendy and her 5-year-old son to one cell, and her husband to another. Wendy did not see her husband again after that. Wendy reports that her son was sick and suffered in the icy air-conditioning of the facility. She also reports abusive behavior from the officers in the holding cell— they laughed at detainees and increased the air-conditioning to punish them. In her initial interview notes officers recorded that she was not afraid to return to her country, although she had expressed fear. Wendy and her younger son were transferred to the South Texas Family Residential Center. The conditions there were far from desirable for Wendy, but less grim than those in the CBP temporary holding cells (a distinction I will explain further below, in Section 3). The US Citizenship and Immigration 61  Services (USCIS) Asylum Office granted Wendy a positive credible fear determination and she was told that she and her husband’s cases would be linked.  It was only with the help of a CARA pro bono attorney that Wendy traced the whereabouts of her husband, although at the time we spoke she was still unsure of the name of the detention center where he had been detained. While Wendy was released with an electronic monitor on her ankle, her husband spent three months detained, she believes in Washington. There, he was ordered “removed” (which is the sterile language the US government uses to describe deportation).  In our interview, Wendy expressed confusion about whether their cases ever were linked, what this meant for her husband, and why they received inconsistent outcomes. She and her husband had different immigration histories, in which he had a previous deportation and she did not – differences that likely affected their inconsistent case decisions. What is notable, though, is how being detained and separated limited Wendy’s knowledge of her husband’s whereabouts and of their case processes. The inconsistent timeline of their cases also extends the family’s separation. On his trip back to El Salvador, Wendy’s husband found himself on the same plane with gang members who recognized him. Fearing for his life, he again fled El Salvador as soon as he arrived, to a neighboring country, where he remained in hiding when Wendy and I spoke. The stability that Wendy’s family once had remained out of reach. Wendy felt anxious and desperate, while her 5-year-old son was sad and missed his father. Without work authorization, she could not support her older son in El Salvador.   62  3.2 DHS separates families This section begins by examining how CBP and ICE, two component agencies of DHS charged with enforcing US immigration law, separate intact families. It then delves into the policies, practices, and logistical constraints of detention facilities that propel this division of loved ones.  3.2.1 CBP separates families in its holding cells For asylum-seeking families who arrive in the US together and who are apprehended together, the first point of separation is likely to be CBP’s temporary detention facilities near the US-Mexico border. These holding cells, which are designed for short-term custody of 12 hours or less, yet regularly detain people for days at a time, are commonly referred to by guards and detainees alike as hieleras (iceboxes) and perreras (dog kennels) because of their frigid temperatures and harsh conditions (Cantor 2015).  CBP first takes detainees to one of its hieleras, which are extremely cold, overcrowded, and unsanitary. Detainees are denied showers and supplies like soap, diapers, sanitary napkins, and sufficient toilet paper (AIC 2016, January 13). At night, the lights stay on while people sleep on the floor or benches without bedding. They are denied medical care and given inadequate meals and drinking water. Detainees are isolated from their loved ones, their consulate, and legal counsel. They report abusive and coercive behavior from CBP officers, such as pressure to accept their deportation (AIC 2016, January 13; Cantor 2015). Some detained families are subsequently taken to a perrera (meaning that they are transferred to a different CBP short-term holding cell) for an additional day, or even for several more days. It is in this frigid and coercive climate that asylum-seeking families report being segregated by gender and age. Research participants described the painful experience in the hielera as the beginning of a prolonged and indefinite separation from their husbands or partners, 63  from other adult relatives, and from minor relatives who are not their biological children. Below, Wendy recounts her experience, in which CBP held her partner apart, did not attend to her son’s medical needs, and punished detainees by increasing the air conditioning: We were walking when the Border Patrol caught us. From there, they separated my son and I and put us in one place, and my partner in another, and from there I did not see [my partner] again…My son was sick in the hielera, and I asked for help from an official who was there. I said, “Listen, my son has a fever.” “Wait, the doctor will come,” [he said], but the doctor never came. The other people even complained that it was too cold, and what they did was put the [air-conditioning] on stronger.   In the CBP holding cells, research participants also reported being separated temporarily from their children around the age of 11 and older, whom CBP placed in a separate area with other children roughly their age. The women separated from their minor children report being eventually reunited and transferred together to a long-term family detention center. Their adult relatives, on the other hand, were eventually transferred to separate adult facilities. This is what Mariana reports happened to her family after a disorienting series of transfers and gender segregation within the holding cells: When we arrived [in the United States], I am not sure where because I am not familiar, but they took us to the hielera. There, they separated us—men on one side, and then my mom, my sister, my daughter, and I. Later, they took only my baby and me to another hielera. I am not sure where. They moved us there with my husband. Later, but that same day, my mom and sister arrived. We spent a whole day there, and a night. Then, that day if I am not mistaken, they sent me to a perrera.…From that day on, I did not see my mom, or my husband, or anyone in my family—it was just my daughter and me.   The CBP holding cells also can mark the beginning of a lengthy separation of children from their caretakers who are not their biological parents. Daniela traveled to the United States with her two daughters and her eight-year-old nephew. In the hielera, officers removed her nephew from her care, and eventually sent him into the custody of the Office of Refugee Resettlement (ORR): When we got to the hielera, they took everything that we had brought with us, and recorded our information again. Then, they told me, “Your nephew cannot stay with you, because he is not your son.” So they took him off to one side, and my daughters and I stayed in the other. From there, I did not see him again.  64   This CBP practice of separating family members in its hieleras and perreras conflicts with the agency’s national detention policy, as expressed in a 2015 policy document: “CBP will maintain family unity to the greatest extent operationally feasible, absent a legal requirement or an articulable safety or security concern that requires separation” (US CBP 2015: 4). CBP further states, “Generally, family units with juveniles should not be separated” (ibid: 22).  I obtained this document through a FOIA request that I submitted to CBP in March 2016, inquiring about the agency’s policy on family separation. More than a year passed before I received a response, which contained several documents – some created for public consumption, like the 2015 national policy document mentioned above, and others that were not – including a 2008 memo about “Hold Rooms and Short Term Custody.” In this memo, the US Border Patrol chief defined a “family group” as: “A group of closely related adults (parent, legal guardian, grandparent, brother, sister, aunt, uncle) and juveniles (son, daughter, grandchild, sibling, niece or nephew) in custody at the same time and place” (US CBP 2008: 2, emphasis added). The memo explains which family groups must be detained together in CBP custody: “a. Grandmother and/or grandfather with juvenile(s) – grandchildren; b. Mother and/or father of juvenile(s) – children; c. Adult sibling with juvenile(s) – siblings; d Legal guardian with juvenile(s); e. Aunt and/or uncle with juvenile(s) – nephew/niece” (ibid: 12-13). The presence of a juvenile (meaning a child under age 18) seems key to CBP’s definition of family, as articulated in the 2008 Border Patrol memo. The policy seems to prioritize the unity of family groups composed of children and adults, but does not make any specific promises about the unity of adults taken into custody together like Mariana, her adult sister, and their mother. Regardless, in practice, CBP does in fact separate adults from children – whether 65  temporarily, as in the case of Rosa from her 11-year-old son, or for the long-term, as happened to Daniela and her 8-year-old nephew.  It is clear that CBP is not fulfilling its promise to keep families together, nor to maintain records of any separations in its holding cells (see LIRS 2016). I next examine how broader DHS interpretations of family unity and ICE definitions of the “family unit” shape custody decisions regarding where to send detainees after the CBP hielera and perrera.  3.2.2 ICE follows a narrow interpretation of family unity A family’s right to unity is a widely-recognized principle under international humanitarian law and international human rights law (Jastram & Newland 2003: 566). Being party to the International Covenant on Civil and Political Rights (ICCPR), the United States has committed to avoiding arbitrary detention and respecting family unity (Ginatta 2014). As mentioned, the 1997 Flores settlement agreement also prioritized family unity—a standard that should guide current DHS custody decisions for arriving asylum-seeking families (Jones & Obser 2015). Ultimately, though, DHS defines what counts as a family.  In addition to my CBP FOIA request, mentioned above, I also sent two FOIA requests to ICE about the agency’s policies on family separation and family unity. I heard back in a matter of weeks. An April 2016 FOIA response I received states that ICE policy within its family detention centers is “to maintain family unity wherever possible” (US ICE 2016, April 13: 1). However, ICE defines the “family unit” narrowly as “a group of detainees that includes one or more non-United States citizen juvenile(s) accompanied by his/her/their parent(s) or legal guardian(s)” (US DHS 2014, August 12: 2). As a result, ICE’s detention practices actually work to separate families in many instances, particularly with respect to spouses, older children, siblings, and grandparents who cannot be accommodated in family detention centers.  66  ICE policy also dictates that a parent or guardian and minor child can be separated after their arrival to family detention. In my FOIA response, ICE lists some examples – such as medical emergencies, allegations of physical abuse or violence, if parents are not able to care for their children, or if ICE uncovers a history of violence, criminal activity or gang affiliation (US ICE 2016, April 13: 1). This type of separation in the name of child welfare does happen. Advocates report that such decisions to split a parent from their child are not properly documented, are made arbitrarily, and are carried out by officials who may not have expertise in child welfare (WRC, LIRS & KIND 2017: 7; Al Otro Lado, et al 2017). Despite my multiple inquiries, I was not able to obtain information about any ICE policies on the unity or separation of adult family members (US ICE 2016, April 13; US ICE 2016, September 20; Email April 7, 2016, on file). This silence left me with the impression that ICE is indifferent to family relationships unless a child is involved. The only apparent configuration of family members that ICE seeks to keep together is the parent/child “family unit” (see WRC, LIRS & KIND 2017: 5). As I explain next, mothers become the de facto parent in ICE custody, given that its “family” detention beds are reserved almost exclusively for mothers and their children under 18, whose ages range from infancy to adolescence.  3.2.3 Bed quota encourages arbitrary detention practices In practice, the majority of “family units” that DHS detains together are mothers and their minor children. DHS currently houses family units in three detention centers: the South Texas Family Residential Center in Dilley, Texas (with a 2,400 bed capacity); the Karnes County Residential Center in Karnes City, Texas (with a 830 bed capacity); and the Berks County Residential Center in Leesport, Pennsylvania (with 96 operational beds). Only Berks, the smallest facility, can accommodate adult men. The remaining 3,230 family-bed capacity of the two Texas facilities 67  combined is explicitly reserved for mothers and their children (ACFRC 2016: 150-151). Yet as of the summer of 2016, ICE was generally not detaining fathers even in Berks (ibid: 12). This practice demonstrates that “family” detention has become a euphemism for locking-up noncitizen mothers and their minor children. The attorneys who I interviewed identified this reservation of the majority of family bed space for mothers and children as one reason why DHS makes differential custody decisions across asylum-seeking family members (also see ACFRC 2016: 4). One attorney interviewee told me about her struggle to explain to her clients why the mother and baby in the family were subjected to prolonged detention, while the father and two oldest children were released: Well, even if immigration had some sort of rationale, but they don't, so you can't explain it because there is no answer, the only way to explain it is to say, “You're right, it makes no sense, and there is no reason why you're here and they're not here.” Or “They had a bed open for you, and they didn't for him, and they needed to fill that bed space.” That's really what it comes down to.   Differential custody decisions across the same family brings the arbitrariness of ICE custody decisions into sharp relief, as well as its interest in keeping detention beds full. The incentive to keep beds full can be traced back to a 2010 DHS Appropriations Act in which Congress introduced a controversial directive to maintain a minimum of 33,400 long-term detention beds at all times. Since then, two large private prison companies have reaped the profits. ICE contracts a major share of its detention beds out to CCA and the GEO Group (Carson & Diaz 2015: 3) These private prison companies are notorious for the miserable conditions caused by their imperative to make profits by detaining people. They scrimp on food, medical services, training and wages for employees, and on their facilities (ibid: 7).  According to advocates, “The role the quota has played in artificially stabilizing these corporations’ revenue from federal immigration enforcement has helped them to double in value 68  since 2010, at the expense of taxpayers, detained immigrants, their families, and communities” (Carson & Diaz 2015: 14). GEO and CCA continue to lobby aggressively to keep the bed quota in place, against vociferous immigrant rights advocacy to end it. In 2013, Congress raised the detention bed minimum to its current capacity of 34,000 beds. To be clear, a bed quota is not common practice for law enforcement agencies. No other US agency is incentivized to keep its beds full at the whim of a Congressional directive (ibid: 3).  As Antonio Ginatta (2014: 4) wrote for Human Rights Watch in a letter to former DHS Secretary Jeh Johnson, “Arbitrariness pervades US immigration detention policy,” contrary to its commitments as party to the International Covenant on Civil and Political Rights. Ginatta pinpoints two main sources of this arbitrariness: mandatory detention laws and the detention bed mandate. The former INS, and DHS as its successor agency, have maintained broad authority since the enactment of the 1996 immigration reform to detain noncitizens for prolonged periods, with little oversight. Ginatta notes that the decision to detain is often made without consideration of an individual’s flight risk or any risk they may pose to public safety. The bed quota, Ginatta argues, raises further questions about whether custody decisions are actually tied to an individual’s specific circumstance (ibid). This arbitrariness is hard to miss when trying to make sense of differential custody decisions across the same family, such as that of Rosa. Rosa’s family In June 2015, Rosa and her two youngest sons—ages 7 and 3—spent several days together in a packed hielera. At night, they found a patch of bare floor to sleep on. Like the cell where they were held, the sandwiches they were given to eat were frozen, according to Rosa. CBP apprehended Rosa’s family together, yet in the hielera they were kept apart. Rosa’s oldest son, who is 11, was locked in a cell dedicated to boys around his age and older. Rosa’s husband was 69  also split off from the family, locked in a separate cell. This was the first time the family had been apart, and it was a reception to the United States that Rosa did not anticipate when the family fled gang-related threats in San Salvador: I never imagined that all this would happen to us. It never crossed my mind. When you are in a difficult situation, all you want to do is get out, running for your life. But we never imagined that we would find ourselves with so many problems. It never crossed my mind that they would separate us.   In the hielera, what worried Rosa the most was the well-being of her three sons. “As an adult,” she explains, “you can tolerate anything, but when your children suffer, this hurts you a lot.” On top of the cold, hard floor that served as their bed, and the painful separation, Rosa reports abusive comments from the CBP officers. “You’ll probably be deported,” they said, and “You shouldn’t be here.” Finally, Rosa was reunited with her 11-year-old son, and without a word as to where they were headed, she was transferred with all three children to the South Texas Family Residential Center. There, Rosa reports slightly improved conditions, “But we were always prisoners. It was like we were in a golden cage. We were always detained, with the uncertainty of what would happen to us.” By the end of July 2015, Rosa felt a fleeting sense of relief when the Asylum Office gave her a positive credible fear determination, and she was released with her three children and an electronic ankle monitor to pursue her case before the immigration court. They went to stay with Rosa’s aunt in California. Her husband, however, still faced an uphill battle from an adult detention center in Georgia. Thankfully, Rosa explained, she was at least able to speak to him every day on the phone. Based on their conversations, she told me her husband felt a sense of desperation, bolstered by abusive comments from the officers and his perception that many of his fellow detainees were being deported. Nine months dragged by. Rosa was perplexed. Her entire family had fled El Salvador for the same reason. Ultimately, an immigration judge ordered her 70  husband removed, and in March 2016, the Board of Immigration Appeals (BIA) denied his appeal.  When we spoke in March 2016, Rosa’s husband’s safety upon his imminent return to El Salvador weighed heavily on her mind. She also worried about how the separation was impacting her sons psychologically. She reported that the younger two cry frequently, and that the oldest had become rebellious. While struggling with her own anxiety and difficulty sleeping, Rosa lamented that she did not yet have work authorization. With this constraint, along with the absence of her husband, she worried about how she will provide for her family.  3.3 Separation harms families’ well-being Being split up can profoundly affect a family’s well-being. As I outline below, research participants spoke to this topic in terms of the mental and material well-being of all family members involved. They described negative mental health impacts in their own analyses of themselves and of loved ones. Inconsistent case timelines also caused distress and extended separation. Finally, some interviewees found themselves with the emotional and financial burden of being the sole caretaker when their partner was detained or deported. 3.3.1 Participants report negative mental health impacts The women who I interviewed described their families’ separation in the hielera as just the beginning of an indefinite time apart from loved ones. It was often a moment of disorientation, given that families may not be told where they are going, nor informed of their loved ones’ destinations (see O’Connor, Thomas-Duckwitz, & Nuñez-Mchiri 2015: 8-9). Interviewees and their loved ones experienced a persistent emotional fallout of separation—especially feelings of sadness, uncertainty, and anxiety, as well as difficulty sleeping. Vanessa described feeling 71  shocked when she and her daughter were split apart from her husband, and argued that forced separation causes all family members to suffer: I had heard that sometimes [the US government] help[s] whole families. When you are separated, you suffer, including the children, especially when they are little. I did not know [that they could separate us], because if I had, I would have been somewhat prepared, although no one is prepared for this.   Mariana also highlighted the abruptness of her family’s separation, and the persistent sadness it caused for all her relatives involved: At [that] moment, they called my mom in for an interview, my sister stayed in the hielera, and they said to me, “Grab your things, you’re leaving.” I could not even say goodbye to my mom. My sister was crying. It was one of the saddest moments that I have had. My daughter and I have been separated for about five months from my husband and from my sister. This is something that had never happened before.   When we spoke in the spring of 2016, Rosa was distressed and losing sleep over the uncertainty of her husband’s fate while he was detained for nine months. This compounded the uncertainty she felt about her own asylum case and that of her sons:   This affects you, psychologically. In my case, I get little sleep, am distressed, worried, and anxious, because every day my husband calls, and I hope that good news will come—that we can buy him a plane ticket, that they have given him an opportunity. But nine months, without anything. Nothing has happened. It is hard to be living with the uncertainty of what is going to happen. Even with me, what might happen to my sons and I, because we are also in an immigration process. We are not one hundred percent sure that we can stay in the United States. Only a judge can decide. So we remain with this uncertainty.   In June 2015, the American Academy of Pediatrics wrote to DHS to express its concern that detention exposes asylum-seeking families to unnecessary mental and physical health risks, while exacerbating the trauma they fled in their countries (Hassink 2015). The numerous risks that detention poses to the well-being of both adults and children are well-documented (e.g. AILA, WRC, & AIC 2015; CARA 2016; O’Connor, Thomas-Duckwitz, & Nuñez-Mchiri 2015; Physicians for Human Rights 2011; Zayas 2014).  72    As the medical literature illustrates, the indefinite nature of detention in and of itself can negatively impact the mental and physical health of asylum seekers (see Physicians for Human Rights 2011). The uncertainty and uncontrollability of being detained can provoke chronic anxiety and dread, pathological stress levels with long-term ramifications for immune and cardiovascular health, depression and suicide, post-traumatic stress disorder, persistent changes in personality, disconnection from family and community, and a worsening of previous trauma (ibid). The stress, uncertainty, and despair that detention causes in children invite negative long-term outcomes for their cognitive and intellectual development, while putting them at greater risk of developing chronic illnesses (Zayas 2014). Similar negative health outcomes have been found for children who are not detained, but have a parent at risk of detention or deportation (Zayas, et al 2015).    Research in the criminal justice context has similarly demonstrated that male incarceration can negatively impact the mental health of non-incarcerated family members like mothers and partners (Green, et al 2006; Wildeman, Schnittker, & Turney 2012, cf Lee, et al 2014: 421). Public health researchers also report that the incarceration of a relative can increase a woman’s cardiovascular health risks (Lee, et al 2014). The non-incarcerated partner may be exposed to chronic stress and social isolation, alongside reduced social support and family income. She may be left to provide economically for the household, while caring for children and supporting her incarcerated partner (ibid: 421).    Mentioning this literature on family separation caused by the United States criminal justice system is not to conflate the experiences of penal incarceration and noncitizen detention, which are not precisely the same. My purpose is to point out that having a loved removed from family life and imprisoned can have wide-ranging impacts a family’s well-being, both in the 73  short-term and long-term. To me, this literature makes the point that an imprisoned person’s well-being cannot be understood in isolation, given that it is bound up with that of their loved ones (see Martin 2012a: 328; Gilmore 2007).   In their 2015 study of trauma among detained asylum-seeking families, mental health scholars O’Connor and colleagues (2015: 9) find that forced family separation only exacerbates the trauma of being detained, while increasing the risk of depression, anxiety, and post-traumatic stress. While it is outside the scope of my research to make clinical assessments, in their own self-analyses research participants expressed that separation negatively impacted their families’ mental health. Further research into precisely how forced separation contributes to trauma for asylum-seeking families in the detention context would be valuable. I next consider how the inconsistencies in case timeline and decisions after families are split up can impact their well-being.  3.3.2 Inconsistent case timelines and decisions cause distress and extend separation The deportation process has particularly high stakes for asylum-seekers, for whom returning to their country can be a matter of life or death (see Brodzinsky & Pilkington 2015). The women who I interviewed were deeply worried about the return of their family members who were separated from them and deported from the US while their own cases remained pending. They described the precarious strategies their deported loved ones have adopted to protect themselves, such as making themselves prisoners in their own homes, hiding out in a different part of the country, or fleeing the country again. The interviewees also expressed confusion about why their family received inconsistent legal decisions on their cases.  It is not possible to know the precise role that separation played in the inconsistent decisions received by the families of research participants — in other words, whether all family 74  members would have received consistent outcomes had they never been separated.9 However, it is clear that separated families can receive uneven decisions on their cases, which are adjudicated on distinct timeframes. It is also clear that this inconsistency can be a source of distress, and that it can extend a family’s separation. Some family members pass their credible or reasonable fear interview, while others do not, even when both face the exact same danger at home. Some are released, while others face lengthy detention that makes it much harder to prepare their cases. Some are deported, while others continue their cases. Ultimately, the family members who remain in the United States with pending cases are left with the uncertainty of what will happen in their own cases, while also forced to worry about their deported loved ones, who in turn must worry about their own safety. In the meantime, the family is subject to lengthy separation. This separation is especially difficult when the economic and emotional labor of caring for children is involved.  3.3.3 Separation creates an undue caretaking burden Being split up can compromise families’ economic well-being. The women interviewed whose partners were subjected to prolonged detention or deportation by default became the custodial parent. They described the financial and emotional burden of becoming sole caretakers, particularly after being released from detention, but before receiving their work authorization. Rosa struggled to pay rent and keep food on the table for herself and her three sons in California: Yes, it has been very hard for us. When we left El Salvador, it was because of danger—threats from the gangs. And thank God I have a lot of family here, but it is not the same. My sons need their father. I need my husband. Besides God, he is my pillar. They are three boys, and these nine                                                  9 It is important to note that separation is one of a variety of factors that may influence inconsistent decisions across a family. For example, certain family members may be subject to reinstatement of removal, while others are not—affecting what type of relief DHS considers them to be eligible for. For an explanation of reinstatement of removal, see Realmuto (2013). In other circumstances, family members may face different risks upon return to their country, resulting in different decisions on their asylum cases.  75  months have been very difficult.…I have to be here at home, because I do not have a work permit and I cannot take risks. So here I am—I need to pay rent, pay for food—and I am alone with my three sons. So it has been very difficult.   Like Rosa, Wendy was still waiting for her work permit to be issued when we spoke. She was unsettled by not being able to work, and therefore support the family, including her son who had been deported from Mexico back to El Salvador: “[I am] anxious and desperate, because I cannot work with this ankle bracelet. It is not feasible. I have been spending time doing nothing, and I am used to working. I also have my son there [10-year-old son deported from Mexico to El Salvador], and imagine—since I am not working, I cannot help him.…[My 5-year-old son] is sad because he is very close with his father, and he misses him a lot. He wants to see his daddy.   As Wendy and Rosa’s experiences indicate, forcibly removing an adult relative, like a father, from asylum-seeking families can place an undue burden on the parent who becomes the sole caretaker by default. This emotional and economic burden only further compromises the family’s wellbeing.   Daniela’s family In early 2016, Daniela entered the United States with her 11-year old daughter, 2-year-old daughter, and 8-year-old nephew. A business administrator from Comayagua, Honduras, Daniela had been caring for her nephew in addition to her own children for the past year and a half. The family fled Honduras because a gang was forcibly recruiting and threatening the older children. The afternoon they arrived in Texas, CBP officers apprehended the family and took them to the hielera. There, the officers took Daniela’s 8-year-old nephew off to one side. The hours ticked by, and Daniela did not see him again until the next morning, but only from afar. During the night, Daniela and her daughters struggled to sleep in the cold and discomfort of the hielera, her younger daughter crying. The next day, Daniela once again saw her nephew briefly in passing, as the family was transferred to a different holding cell. There, the family was fractured further, as 76  Daniela was told that her 11-year-old daughter could not stay with her. The officers led her daughter to a separate area to sleep. Daniela and her two daughters were next taken to the South Texas Family Residential Center, where they were detained for 18 days. Daniela says: “It was very, very horrible. They do attend to you there, yet you still feel like a prisoner, because you do not feel free; you always feel sad. They have activities for the children, but it is not the same as being free. It was very hard.” Daniela did not see her nephew again after the CBP holding cell, nor did she have any say in where he was taken. DHS officials did not even keep her informed of his location. Even though Daniela was his caretaker, he was treated differently than her daughters – like an unaccompanied minor, as though he had traveled alone.  Only later did Daniela learn that her nephew stayed with a foster family in San Antonio for a week before he was released into the custody of his mother, who lives in Texas. Daniela was relieved to know that her nephew was ultimately released to his mother’s care. She only figured this out once she was able to place a call to her sister from detention. Her nephew reported that the foster family treated him well, yet he lost weight during his stay with them. Daniela had been her nephew’s primary caretaker in Honduras, and she wanted to continue caring for him, or at the very least know where he was and that he was safe, but instead was left to worry about his well-being.    3.4 Separation harms families’ ability to access humanitarian protection Splitting asylum-seeking families up clearly has negative effects on their well-being. It can also throw new hurdles into their ability to present their asylum claim, and thereby access humanitarian protection, as I detail below. Some of these hurdles stem directly from the 77  emotional fallout of being separated, when families are distracted from their case by the stress of not knowing where their loved ones are. Dividing up family members also splits up their case. The initial results of their case (and potentially the final results as well) can thus turn out differently. Finally, dividing people also divides evidence, which can make harder for families to present their case and for adjudicators to rule fairly on it. 3.4.1 Separation interferes with families’ ability to present asylum claims For asylum-seeking families, the high stakes of being granted humanitarian protection in the US, and thereby avoiding a dangerous return to their countries, makes their capacity to advocate for themselves of the utmost importance. Most families have no other advocate besides themselves, given that—in contrast to the criminal justice system—the US government does not guarantee legal counsel in removal proceedings, thereby leaving approximately 70 percent of families and 50 percent of children without legal representation (AILA 2016: 4; TRAC 2016, June). The impact cannot be overstated. A national study of 1.2 million deportation cases found that “similarly situated” respondents with attorneys are fifteen times more likely to seek relief from removal (meaning to defend themselves against deportation), and five and a half times more likely to be granted relief, than those without attorneys (Eagly & Shafer 2015: 2). The vast majority of families without legal representation are at a major disadvantage.  As one attorney interviewee explained to me, splitting asylum-seeking families only deepens this disadvantage, given that the urgent task of locating relatives can take energy away from their legal case: So if somebody is locked up, they do not know where the other kid, or spouse, or anybody is. They cannot hop online and search on the online detainee locator, and…there is no way to put money in their account so they can call, even if they knew where the person is. So this effective separation creates…in addition to the trauma and the stress of having that uncertainty of “Where is my loved one? What is happening to them?” that alone may derail your case, the [in]ability to 78  be able to focus on representing yourself—you are so concerned about what's going on with the other person you love.  For the minority of family members who do have access to legal counsel, being split up still takes their energy away from their own asylum cases. Attorneys working in the family detention context report the difficulty their clients have in focusing on their own cases when they are preoccupied with investigating the whereabouts and well-being of their loved ones. In the words of one attorney: We are trying to give them all this really nuanced legal advice, and they [say] “No, I just want to find my husband.” And you [say] “Well, but I am your lawyer, I am not your husband's lawyer, so I am trying to do my best to help you. I understand how you feel, but you have to focus on this right now.” And they [say], “No, where is my other kid?”   Being split up in a sudden, and disorienting way through DHS custody decisions can leave relatives scrambling to find each other. This is a daunting task, especially while detained. The burden of finding each other leaves family members distracted from the urgent legal matters requiring their attention. 3.4.2 Separation divides asylum cases Advocates report that “when families are detained in different federal facilities, there is no way to regularly monitor this or inform the detainee where another family member is located, making it nearly impossible to reunite or pursue a joint asylum claim without counsel” (LIRS 2016: 2; WRC, LIRS & KIND 2017). A 2015 study of family separation experienced by Mexican citizens in the expedited removal process likewise finds that DHS does not have a systematic way of tracking familial relationships (Danielson 2015: 1). Divided families must navigate the dense web of government agencies that may be involved in their removal cases—including ICE, CBP, ORR, USCIS, and the Executive Office for Immigration Review (EOIR), which houses the immigration courts.  79  Rosa describes her unsuccessful attempt to link her case with that of her husband, even though they fled El Salvador together and for the same reason: We tried to have them link the case, but they were never able to. They said that [my husband] would have to fight his case there. I would have to keep fighting my case here. We did not come for different reasons. We came for the same reason—due to threats from the gangs. I really do not know what happened, but this is our situation—nine months detained. The same situation, the same problem.…We came together as a family. But it seems that it doesn’t not matter much that [my husband] came with me. They have taken it as a separate case.   In the void of administrative support, even families who are able to get released from detention face an uphill bureaucratic battle in trying to unite their cases for humanitarian protection, as outlined by one attorney: “Assuming that you all even get out of jail, you [have] to file all these motions to change venue and consolidate your cases—you are creating tons more work for these families that I do not think anyone is able to do, unfortunately.” Faced with this difficult task, families who fled their country together and for the same reason may continue to have their cases adjudicated by different judges, on distinct timeframes, and even in separate federal circuits governed by disparate case law. This also creates administrative inefficiencies for the immigration courts and worsens the courts’ tremendous backlog (EOIR is currently scheduling final merits hearings for non-detained asylum cases for years away) (Campos & Friedland 2014: 12; TRAC 2016, July 19). 3.4.3 Separation divides evidence For people who have fled persecution in their countries and find themselves in removal proceedings as they pursue asylum in the United States, being detained in and of itself makes it hard to obtain records and testimony to corroborate their claims. When asylum-seeking families are divided between different detention centers, certain family members may have difficulty obtaining evidence that may be crucial to their cases. As a result, adjudicators may only hear fragments of their stories and rule on incomplete facts. An attorney interviewee gave me an 80  example of such a situation in which relatives’ isolation from each other makes it difficult for them to tell their story: If you are the mother and child, who left with your husband because he was having problems—he was facing death threats—all you might know is…“He came home and he was pale and said we have to leave right now, they are going to kill us.” And you may not have the information that is necessary to put in your claim. You might not be able to testify about the kinds of threats that were going around.   Many Central American asylum-seekers have family-based claims, which may be more difficult to present when families are separated.10 One attorney interviewee explained to me that breaking up families weakens their ability to support their family-based claims, which may require them to demonstrate that an entire family is at risk because one family member was threatened: So if the mara [gang], the MS-13, is persecuting the Alonso family because the father refused to pay extortion, and then the Alonso family comes to the border, and then you split up the Alonso family, and you divide them between detention centers, so their ability to corroborate their claim, and say— “No, my dad, he [has] this testimony about how he stood up to them,” but you have put dad in, like the middle of Florida without any legal representation—it just becomes really hard for them to corroborate their claims, because you have broken everyone up.   Separating relatives can fragment crucial evidence, like testimonies or documents, between them. When a person has been persecuted because of their membership in their family, taking the family apart makes it hard to present key pieces of their story.  3.4.4 Separated family members can face inconsistent timelines and case decisions Trends in the adjudication of asylum claims vary wildly between immigration judges and across geographic regions of the United States. As of Fiscal Year (FY) 2015, immigration courts across the country granted asylum claims, on average, 45% of the time. Yet in certain regions and                                                  10 To prove eligibility for asylum, an individual must demonstrate that, if forced to return to her home country, she has a well-founded fear of persecution on account of race, religion, nationality, membership in a particular social group, or political opinion; her government is unwilling or unable to protect her; and she cannot safely relocate within her country. A family can constitute a particular social group, which is thus a protected ground for asylum. See section 101(a)(42) of the US Immigration and Nationality Act for the legal definition of refugee that is used to adjudicate asylum claims (US INA). 81  detention facilities, adjudicators grant claims at dismally low rates. For example, in Atlanta, only 2% of asylum claims were granted in FY 2015, and in Stewart Detention Center in Georgia, only 5% (US EOIR 2016: K2). This takes place in a climate in which adjudicators have an extremely low and inconsistent record of recognizing the most common Central American claims—such as those based on domestic violence and gang-related persecution (Campos & Friedland 2014: 12-13). Given these inconsistencies, when different judges adjudicate the same case in distinct locales, it is possible that the final outcomes will be different. Some of the families of the women who I interviewed faced inconsistent timelines and decisions on their cases, as described in the previous section about well-being. For example, Mariana, her children, and her mother were released from detention to continue their cases, while her sister and husband were deported directly from detention (their story is described below). Given that the women interviewed for this report had not yet received final decisions on their cases, it is not possible to know precisely how being separated might influence final outcomes, or cause distinct final outcomes, for different members of their families. The question of how separation influences different final outcomes would be a valuable area for further research.  Mariana’s family Mariana described a trip to the airport to pick her mother up as one of the most difficult moments she has experienced. She was living in Virginia with two of her siblings and her 2-year-old daughter. Mariana’s mother had finally been released from a Hidalgo, Texas detention center to join the family in Virginia and pursue her asylum case. Mariana recalls the stillness that pervaded their meeting in the airport when her mother learned of the deportation of Mariana’s sister, Noemi: 82  When [my mom] arrived, she says “Now, we’re only missing Noemi to be here with us.” We all fell silent. My brother says, “She needs to know what happened. Mom, they deported Noemi.” And my mom became so upset, because she had no idea. My father passed away two years ago, and although my sister is an adult, she is the youngest and we always think of her as the baby. Believe me, this was very intense because we had never been separated.   This was just the latest of a series of painful separations that Mariana’s family had suffered since fleeing El Salvador in September 2015. A nursing student in San Salvador, Mariana was about to begin her fifth and final year, motivated to provide for her family and embark on a career path of helping others. She left these dreams behind as she traveled on buses with her husband, daughter, mother, and sister (Noemi) for 20 days. Upon arriving in the United States, they were apprehended together by CBP, but subsequently separated. Mariana and her daughter were transferred to the South Texas Family Residential Center, while her mother and sister went to a detention center in Hidalgo and her husband to a facility in Atlanta. Twelve days later, Mariana was released with a positive credible fear determination, and an electronic monitor latched to her ankle.  Mariana reports that neither her husband nor Noemi felt they were given an adequate opportunity to express themselves in their own credible fear interviews. Her husband wanted to discuss Mariana and their daughter, but was told to focus on himself. To Mariana, this narrow focus makes little sense: “I believe that since we are a family, what happens to me, happens to him.” Both her husband and Noemi were each ordered removed. When Mariana and I spoke, they were back in El Salvador, but remained internally displaced, as it was not safe for them to return to their community. With their safety a continual concern, Mariana struggles to build a new life in Virginia while she awaits a final decision in her own case. Given her daughter’s young age, she laments that she likely will no longer recognize her father whenever she next sees him. 83   3.5 Conclusion The lives of asylum seekers who are locked up, or deported back to the dangers they fled, are confined in every sense of the word. The evidence presented in this chapter also suggests that the lives of non-detained relatives, such as mothers and children, can continue to be confined by the detention and deportation of their relatives, such as partners, fathers, and adult siblings.  The experience of imprisonment can exceed the spatial and temporal bounds of the prison walls, demanding a conceptual framework that accounts for the spatially- and temporally-diffuse impacts, including on kinship and community networks (Gill et al. 2016: 5; citing Comfort 2002; da Cunha 2008; Harman, Vernon, & Egan 2007). For instance, Comfort (2002: 467) finds that US women’s efforts to bring family activities into the visitation room of their partners’ prison can lead to “the paradoxical ‘institutionalization’ of their own family life and thus extend the reach and intensity of the transformative effects of the carceral apparatus.” Family and imprisonment collide in complex ways.  Drawing on Ruth Wilson Gilmore’s (2007) work on prison expansion in California, Kathryne McKittrick (2011: 959) suggests, “being locked in and being locked out are two sides of the same coin.” In other words, imprisonment is much more than an individual experience, even if only one person is physically behind bars. As McKittrick (2011) details, Gilmore’s work centers “human relationality” – foregrounding not only that the harmful outcomes of imprisonment are relational across family and community, but also that contesting imprisonment can be a relational process of community members observing the similarity of their harmful experiences, and coming together to demand more dignified treatment.  84  Research participants expressed that for their families, well-being and protection are best interpreted as relational, rather than individual experiences. UNHCR research supports this point, illustrating that family unity and reunification are of critical importance to asylum seekers. “Protection at its most basic level derives from and builds on the material and psychological support that family members can give to one another” (Jastram & Newland 2003: 557). This insight – that a person’s well-being does not exist in isolation from their community and kin – seems obvious in many ways. Yet it runs counter to the current way that the US detention system atomizes certain families into individuals, and endows those individuals with differential access to rights. This atomization can be bizarre in practice, facilitating uneven treatment across individuals within the same family (e.g. when detained children have the right to be with their parents, but parents do not have the same right to be with their children). The promise to support the well-being of a child, while disregarding that of their parent (or of any other close loved one, for that matter), is untenable in practice. It simply implies that the well-being of no one in the family is prioritized, because children and adults alike will undoubtedly suffer from the prolonged detention or deportation of their loved ones. This is not some temporary inconvenience – it a harm that may persist indefinitely into the future. Judith Butler (2004: 23) speaks to this indefinite temporality, and to the way that our interdependence becomes impossible to ignore in the face of loss:  When we lose certain people, or when we are dispossessed from a place, or a community, we may simply feel that we are undergoing something temporary, that mourning will be over and some restoration of prior order will be achieved. But maybe when we undergo what we do, something about who we are is revealed, something that delineates the ties we have to others, that shows us that these ties constitute what we are, ties or bonds that compose us.   Although the women interviewed for this research are no longer behind bars, when we spoke they were still constrained in significant ways. Some were still shackled to ankle monitors, while 85  others were not yet permitted to work. All were confined by their sadness, worries, and the interruption to their social support network caused by the ongoing detention or deportation of their relatives. The physical confinement of one person continues to confine others through the emotional and economic toll it brings upon them. I thus argue that the spatial bounds of confinement extend far beyond the heavy cement walls, bars, and barbed wire of the detention center. The temporal bounds of confinement can also extend indefinitely – days, weeks, months, or years into the future that loved ones may wait to be together again.  The problem described in this chapter is upheld by the United States executive branch’s targeting of Central American asylum seekers, a congressional mandate to fill detention beds, and DHS custody determination practices that systematically separate family members. Yet as mentioned, the detention of families arriving at the US-Mexico border is not mandatory. When CBP apprehends people who express fear of returning to their country, it has the discretion to issue them a Notice to Appear (NTA) before the immigration court, rather than placing them in fast-track removal and detaining them (AILA 2016: 12). Prior to the summer of 2014, DHS often used this discretion, opting to release rather than detain many intact families (Flores v. Lynch 2016, September 19: 5). Utilizing this discretion, combined with judicious use of proven community-based alternatives to detention when necessary in individual cases, would keep families to together, free from detention. Allowing asylum-seeking families to stay together would yield multiple benefits. It would help the United States government better uphold its various commitments to family unity. It would also support families’ mental and material well-being by freeing them to support each other. Further, maintaining family unity would give families fairer access to the system of humanitarian protection created precisely for people in their circumstances. Presenting the facts 86  and evidence of their case together, before the same judge, and in the same location would create the best conditions for adjudicators to understand the family’s claim and thus rule fairly. Prioritizing their unity would support the well-being of asylum-seeking families, facilitate efficient adjudication of their cases, and allow them fairer access to protection.  Despite concluding with this policy recommendation, I acknowledge that the United States government has often failed to uphold its stated commitments to family unity – and in the following chapter I explore why.   87  Chapter 4: A familiar state violence  “Personally, I don’t like interviews,” Vanessa revealed as we finished our interview in March 2016. She chose to speak to me, she said, because a legal services coalition that she greatly appreciates had connected us. Vanessa, her husband and their daughter had fled extortion and death threats in El Salvador, in pursuit of asylum in the United States. Despite the gravity of what had happened to Vanessa’s family, it was not the topic of our interview. When she and her family crossed the US-Mexico border together the Department of Homeland Security (DHS) split them apart and sent them to different detention centers, marking the beginning of their painful and indefinite separation. As described in Chapter 3, I interviewed Vanessa and several other women to write a report advocating for family unity and freedom from detention. Despite the generosity of participants, and my optimism about contributing to advocacy, I was worried about doing the interviews. The topic felt intrusive. I needed to ask how it feels to be forcibly separated from loved ones. The answer seemed painfully obvious – sad, shocked, distraught, uncertain, worried, anxious. Mariana, another research participant, turned the question back to me and to the potential readership of the advocacy report, asking us to imagine ourselves in her shoes: “If you have a mother, or if you have children, and they were separated from you, how would you feel? This is my point. If, for a moment, you said, wow, how would I feel without my daughter? I would feel bad. Or how would I feel without my husband? I would feel very bad.” As I wrote the report, I felt deeply troubled (to say the least) that this government practice continued unabated. DHS could easily avoid separating asylum seekers by releasing them, rather than following a general policy of detention. This practice is likely to persist into the foreseeable 88  future given the Trump administration’s harsh stance on immigration (US ICE 2017; Trump 2017), while remaining in the shadows for those not directly affected or otherwise attuned to the issue. This chapter asks: what makes the forced separation of asylum-seeking families, which is clearly harmful and could be easily avoided, politically possible? In other words, why does this practice generate little public dissent, and why does DHS keep separating families? I offer three partial answers. One answer is that this practice is normalized by a vast historical precedent of forced family separation in the US that tends to target people of color and indigenous peoples. A second answer is that in 2014, a threat narrative began circulating about Central American asylum-seekers. It asserted that releasing entire asylum-seeking families together from detention was causing mass migration. DHS used this narrative, which framed family unity and freedom as a threat, to rationalize its expansion of family detention. A third answer is that, as I began to outline in Chapter 3, DHS follows a narrow definition of the “family unit” that obscures the separation of any combination of family members beyond the mother/child unit. DHS is not fixing this problem, and in some cases actively uses family separation against asylum seekers.  This analysis leads me to argue that forced family separation, and certain arguments for “family unity,” operate as a tool of racialized governance. As outlined in Chapter 1, this system of racialized governance has long framed Latinos as “illegal aliens,” and therefore as criminally-suspect and perpetual foreigners (Ngai 2004). Latinos continue to be framed in public discourse as a threat to the racial and cultural purity of the US, thus naturalizing punitive immigration enforcement (Chavez 2013; Santa Ana 2002; Sundberg & Kaserman 2007; Sundberg 2008). This racist logic has been rendered commonsense through a legacy of Anglo-Saxonist thinking about Latin American people and places as immutably inferior (see Berger 1995; Stepan 1991). The 89  amplification of deterrence strategies towards Central American asylum seekers since 2014, and the family separation it causes, offers a clear example of how a system of racialization lives on.  I acknowledge that my argument here – that forced family separation, and certain arguments for “family unity,” operate as a tool of racialized governance – sits in tension with my own advocacy for family unity in Chapter 3. I analyze this tension, and the politically-fraught nature of formulating an argument for “family unity” in Section 4.4 of this chapter. To make my argument, I draw on some of the same empirical sources cited in Chapter 3, including Freedom of Information Act (FOIA) responses and interviews (see Chapters 1 and 3 for a full discussion of my methods). I also analyze government documents, public addresses of elected officials, federal court filings, policy reports, news articles, and secondary historical scholarship. I begin this chapter by outlining three expressions of family separation that have historically taken place in the United States, either carried out directly by state actors, or sanctioned by law: (1) the separation of enslaved families from the early days of settler-colonial rule through abolition; (2) the separation of indigenous families through the boarding school system launched in the late-nineteenth century; and (3) discretionary administration of immigration law in the early-twentieth century that separated Mexican and Asian families, while prioritizing the “family unity” of European and Canadian families. Accounting for histories like these and their enduring effects demonstrates that the right to family unity – today, a well-established human rights norm among liberal democracies – has never been open to all. This revelation would likely be unsurprising to communities targeted for this type of state violence, and to those otherwise attuned to the paradoxes embedded within liberal democratic nation-states, which have always hidden violence behind the promise of 90  universal rights (Mills 1997; Goldberg 1993). I call attention to this history because it gets to the heart of what makes the forced separation of asylum-seeking families today politically possible.  I also call attention to this history because it reveals the limitations of a liberal, human rights approach to advocating for family unity. This point struck me while reading a United Nations report about the forced separation of asylum seekers in noncitizen detention centers around the world, which suggests, “Detention practices are one of the rare areas in which States commonly take direct actions that divide intact families” (Jastram & Newton 2003: 602, emphasis added). The idea that forced separation is rare stood out to me because my experience in the migrant justice movement told me that this is a devastatingly-routine fate for mixed-status families in the US. Further, the historical overview that I provide next illustrates that the state-sanctioned division of intact families has been quite common over the course of United States history. A liberal, human rights approach can render this reality invisible because the racialized families often subject to separation have historically been treated as less than human. I conclude that an awareness of this history is crucial to advocate for “family unity” in an inclusive way.   4.1 Family separation as racialized governance Scholarship on historical and contemporary forms of family separation demonstrates that family unity has never been a right open to all. Concepts like “social death” (Patterson 1982; also see Cacho 2012; Kim 2009), “legal violence” (Menjívar & Abrego 2012; also see Abrego, et al 2017), and “slow death” (Lee & Pratt 2012) explain the enduring pain and cross-generational marginalization that can result when families and communities are systematically dismantled (Pratt 2012). This work is careful to critique state practices, rather than the constrained choices that families make, to avoid blaming families, pathologizing non-nuclear kinship arrangements, 91  and ignoring families’ resilience in continuing to care for each other (Pratt 2016: 11, citing Aguilar 2013; McKay 2007; Olwig 2007; also see Zentgraf & Chinchilla 2012). As Geraldine Pratt (2012: xxvii), referencing Walter Benjamin, writes: When a temporary migration program leads to the systematic marginalization of (future) citizens and their children…this reveals “something rotten” within the legal migration system and the exercise of a kind of state violence, less spectacular and decisive than extrajudicial killing to be sure, but one that destabilizes assumptions about geographies of order and goodness, and opens a space to question the (Canadian) state’s monopoly on assessments of and claims of justice. It might jolt (some of) us out of our passive consent.  Pratt writes about Canada, but her point that immigration policy that systematically marginalizes particular families by keeping them apart is a type of state violence applies equally well to the United States context. I find concepts like legal violence, social death, and slow death helpful to name state-sanctioned family separation for what it is: violent. These concepts foreground the temporal complexity of forced separation, tracing its “afterlives” (Hartman 2007; Ybarra & Peña 2016). For instance, historian Saidiya Hartman (2007) describes the alienation from their ancestry that some African American people may continue to feel today as part of the afterlife of the institution of slavery that systematically broke families apart. To conceptualize forced separation as a form of racialized governance it is important to first explain that giving meaning to and regulating what counts as a family is a mode through which nation-states govern their populations (Lee & Pratt 2012: 892, citing Povinelli 2006; Puar 2007; Stoler 1995; also see Martin 2012b). This is a central conceptual insight of Michel Foucault’s (1978/1990) work, which recognizes that power can be wielded not only in an overtly-repressive manner, but also in an “affirmative, knowledge-producing form” (Stoler 1995: 22). Stoler (1995) has extended Foucault’s analysis to consider how European imperial conquests, and the racial orderings of everyday life they generated, informed the production of modern norms around sexuality, including family life. 92  By the mid-nineteenth century, European imperialism helped cement a new middle-class (or ‘bourgeois’) identity that allowed people of European descent, whether in Europe or the colonies, to differentiate themselves from the upper and lower classes, as well as from colonized peoples classified as nonwhite. For women, part of occupying this identity meant staying home to care for their children (Stoler 1995; Heron 2007; McClintock 1995). Of course, this domestic arrangement was not possible for enslaved, colonized, and working-class women who were obliged to work outside the home, often in the homes of bourgeois women (Roberts 1997: 10). Aníbal Quijano (2000) suggests that black and indigenous populations in the Americas, as well as women sex workers in Europe, served as key foils for the development of the bourgeois family ideal. Family unity developed as a standard the bourgeois reserved for itself, but could deny to others: Familial unity and integration, imposed as the axes of the model of the bourgeois family in the Eurocentered world, were the counterpart of the continued disintegration of the parent-children units in the “non-white” “races,” which could be held and distributed as property not just as merchandise but as “animals” (Quijano 2000: 378, cf Lugones 2016: 18-19).  In North America, the rise of the bourgeois family ideal was also facilitated by the displacement of the variety of family, kinship, and governance arrangements that preceded European colonialism, including the women-led and clan-oriented structures of some indigenous peoples (Gunn Allen 1986/1992, cf Lugones 2016: 24-27). This historical diversity in family structure has largely been erased in settler-colonial legal structures. The United States legal system and dominant social norms continue to center a white heteropatriarchal nuclear family ideal in countless ways, and thereby regulate families, erase other kinship structures, and marginalize those who cannot, or refuse to fit the mold (Arvin, Tuck & Morrill 2013: 14-15; also see Butler 2002; Byrd 2011; Collins 2000). Efforts to restrict the reproductive autonomy of black, indigenous, and Latina women, such as the widespread use of coerced sterilization during the 93  1960s and 1970s, provide a clear example (Lawrence 2000; No Más Bebés 2016; Roberts 1997). Another example is the ongoing struggle for marriage equality. The governance of family life in the United States has created conditions that foster the well-being and unity of particular families, while allowing for or even encouraging the fragmentation of others. Below, I provide three historical examples. My intention is not to imply that these are the only such examples, nor that they are all the same – whether in how they took place, the specific justifications used at the time that allowed them to happen, or in their precise effects. They each came about at specific moments through a unique set of events. What connects them are their enduring negative effects, state involvement or complicity, and their function in upholding a white supremacist social order. I conceptualize these historical junctures as ‘citational’ in that they have created an archive of racial imaginaries and techniques of governance that can be revived in the present, precisely because their repetition over time has rendered them commonsense (see Braun 2003; Goldberg 2002; Stoler 2013; Sundberg 2015b).  4.1.1 The separation of enslaved families for profit After enslaved people were forcibly displaced from the African continent to the settler-colonial society that became the United States, they found themselves isolated not only from the social history of their ancestors, but in many cases from their closest living family members. Plantation society did not recognize the kinship and community ties of enslaved people – giving slave owners the authority to separate couples, parents from children, extended family members, and people from their community of origin (Patterson 1982: 5-6). In his classic study of the power relations that have defined systems of slavery in the US and globally, Orlando Patterson (1982: 7) terms this practice “social death” – the alienation of enslaved people from “both ascending and descending generations.”  94  As historian and attorney Heather Andrea Williams (2012) notes, very few legal restrictions were established in slaveholding states on an owner’s ability to sell individuals for profit, and therefore ability to separate relatives. The result was that forced migration and therefore family separation were enduring threats for enslaved people from the 1600s through the abolition of slavery in 1865. The end of the Atlantic slave trade in 1808, alongside the US conquest and settlement of lands further south and west, led to a boom in the domestic slave trade, and the forced separations it brought (Williams 2012: 39). Not all enslaved families were divided, yet the constant possibility served as a technique of domination – a way to terrorize and thereby discipline enslaved people. Patterson (1982: 6) notes:  Even if such forcible separations occurred only infrequently, the fact that they were possible and that from time to time they did take place was enough to strike terror in the hearts of all slaves and to transform significantly the way they behaved and conceived of themselves.   This institutionalized assault on family life was a defining feature of United States slavery, often taking place when a debt or death in a slave-owning family prompted them to sell or divide their “property” (Williams 2012: 90). Preventing couples from legally marrying or being together was also a way that slave owners asserted their power. The legal justification was that enslaved people were property, and thus could not to enter a legal contract like marriage. Perversely, some slave owners argued against allowing the marriage of enslaved people in order to protect them from the risk of eventual separation (ibid: 49-63). The reality that their family could be suddenly forced apart provided a deeply traumatizing lesson to enslaved children about the value assigned to their lives and well-being. In analyzing the historical narratives of people enslaved as children, Williams (2012: 24) summarizes what losing a parent was like – an utterly unforgettable experience in a child’s life, a loss akin to death but with the glimmer of hope to one day find each other and be together again: 95  “the jolt of sudden loss, holding on to a faint hope of reunification, and the searing, lasting memory of confusion and pain.” This comparison to death illustrates the enduring emotional fallout that separation brought upon enslaved children.  White settlers reacted in a number of ways to the horrors of slavery. Abolitionist Harriet Beecher Stowe saw family separation as one of the most deplorable aspects of slavery and thus made it a central feature of her writings (Williams 2012: 102). Although some slave owners had ambivalent or remorseful feelings about their part in dividing enslaved families, many had absorbed the dominant racist thinking of the era (held even by some abolitionists) that black people did not feel pain as deeply as white people. Many white people remained silent, whether out of an inability to empathize, or reluctance to risk their own privilege or well-being by questioning the status quo (ibid: 98). Despite institutionalized separation, enslaved people of course still forged loving relationships. They went to great lengths to maintain family connections, prevent separation, and reunify. They also improvised family structure, with extended family networks and communities taking up caring responsibilities as needed. These efforts to maintain, reunify, and establish kinship flourished after abolition as well, as emancipated people worked to locate and reunite with loved ones, create families, and legalize marriages (Hine & Thompson 1998: 20, 151). An utter disregard for the family life of enslaved people was sanctioned by settler-colonial leadership and later by the United States government. This disregard was a central feature of a deepening white supremacist social order, in which black families’ pain in being apart could be brushed aside in favor of profits for white slave-owners. There are many ways that the legacy of forced separation, as just one of the many cruelties of slavery, may reverberate through the present. Alongside the sense of alienation from their ancestry that some African 96  Americans continue to experience, Hartman (2007: 6) summarizes the legacy of slavery as: “skewed life chances, limited access to health and education, premature death, incarceration, and impoverishment.” As a direct result of this legacy many black families continue to be subjected to new forms of forced separation today. The criminal justice system effectively ‘disappears’ black men, one million of whom are locked-up, apart from their loved ones (Muwakkil 2005, cf Alexander 2012: 179). Further, US Child Services has long criminalized the parenting practices of low-income black women and targeted their children for removal – earning the nickname of “Jane Crow” (Clifford & Silver-Greenburg 2017). 4.1.2 The separation of indigenous families as a technique of genocide Indigenous communities have been targeted in a number of ways for forced separation over the years, often in the name of assimilation into dominant white culture or in the name of the best interest of the child. Families have been taken apart through the channels of boarding schools, foster care, and adoption (Strong 1999; 2001). The generations of indigenous children removed from their families and communities are sometimes called Lost Birds, named after a Lakota baby who was orphaned by her mother’s execution in the Wounded Knee Massacre of 1890, and then adopted by a white general (Powell, Friedman, & Herrman 2014). Roxanne Dunbar-Ortiz (2014: 9) classifies this heritage of children being taken from family and community as central to an ongoing genocide against indigenous peoples. The boarding school system was devised at the end of the nineteenth century as a method of forced assimilation. Captain Richard Henry Pratt experimented with imprisoned indigenous people from the Plains region, who were removed to the Fort Marion Prison in 1875, to develop what would become the boarding school model. Pratt infamously followed a philosophy of “kill the Indian and save the man” (Dunbar-Ortiz 2014: 151). He treated the captive people like 97  soldiers – assigning them drills to complete and making them cut their hair short and dress in military garb. The US Office of Indian Affairs (the predecessor of today’s Bureau of Indian Affairs) followed Pratt’s model in establishing numerous federal boarding schools around the country. A variety of Christian missionary schools soon followed (ibid). The schools, often located far from Native American reservations, instructed children in vocational training and assigned them manual labor. Generations of families from tribes across the United States attended the schools (Child 1998: 173, 205). By isolating children from their families, communities, traditions, languages, and even their own names and identities, the schools sought to “civilize” and thereby assimilate them into dominant white culture by indoctrinating them with a worldview defined by Christianity and wage labor (ibid: 271).  As historian Brenda J. Child (1998) explains, the ‘need’ for vocational training at the close of the nineteenth century flowed directly from other genocidal policies enacted by the US government towards indigenous peoples. Legislation like the General Allotment Act of 1887 had abolished communal landholding structures on reservations, thereby shattering self-sufficient economies and impoverishing affected communities (ibid: 215). In 1891, Congress passed a law requiring children to attend the boarding schools. Families that did not comply were met with coercive measures – such as the denial of crucial public benefits (ibid: 273). Speaking to the experiences of Ojibwe families in the Midwest, Child (1998) describes how government policy in the early years of the schools prevented visits home. Boarding school administrations worked to limit contact between parents and children, often keeping them apart for four years or more at a time. Homesickness was endemic, but administrators did not accept this as a reason to release children for a visit (ibid: 752-753). Loneliness was one of many hardships and risks weathered by boarding school students, in addition to a strict military-like 98  environment, vocational training that funneled them into low-wage work, surveillance, conditions like overcrowded housing that let diseases like tuberculosis flourish, insufficient healthcare, inadequate meals, and even death (ibid). At boarding schools across the US, students were exposed to “physical, sexual, cultural, and spiritual abuse and neglect, and experienced treatment that in many cases constituted torture for speaking their Native languages” (National Native American Boarding School Healing Coalition 2017).  Parents and other loved ones tirelessly petitioned school administrators for the opportunity to see their children, while students frequently ran away to visit home. Running away was so common that Child (1998: 182) calls it “a universal thread that united boarding school students through the decades.” As Child (1998: 473, 1517) illustrates, the boarding school system was immensely harmful, but never fully succeeded in its mission to tear kin and community apart. Likewise, in her analysis of an Oklahoma boarding school, K. Tsianina Lomawaima (1994) demonstrates that the boarding school system at times even had the unintended effect of uniting indigenous students across diverse backgrounds in resistance.  Nonetheless, it is important to recognize the schools’ enduring negative impacts on the estimated 100,000 children who attended the schools between 1879 and the 1970s, and on their families and communities. The National Native American Boarding School Healing Coalition (2017) reports that even after the schools were shuttered, survivors continued to struggle with their sense of identity, safety, and self-esteem, as well as their ability to establish relationships. Extended family and community structures were eroded, while some tribal traditions and language abilities were lost. This legacy undoubtedly contributes to the health and education disparities, as well as high suicide rates, that persist for Native American communities today (ibid). This legacy also lives on in the alarming rates at which indigenous children are removed 99  from their communities. This reality necessitated the Indian Child Welfare Act of 1978, which created legal protections to better prioritize keeping indigenous children with their relatives (Strong 1999; 2001). Despite this crucial protection, Native American children are still taken from their family and put into foster care four times more often than white children (NICWA 2017). To this day, the United States government has not apologized in a meaningful way, nor attempted to reconcile the harms wrought by the boarding school system (Pember 2015).  4.1.3 Family unity and racial hierarchies within the US immigration system The turn of the twentieth century was a time when exclusions were being cemented into United States immigration law and enforcement practices – of Chinese citizens and other Asian populations, of poor and criminalized people, of women projected to become “public charges,” and of people with radical politics (e.g. “communists”). These exclusions reflected the biological determinist thinking of the era suggesting “social undesirability derived from innate character deficiencies, which were perceived to be rooted biologically in race, gender, or ‘bad blood’” (Ngai 2004: 78). This was also a time of large-scale immigration of working-class people from southern and eastern Europe to the United States, who were popularly seen as racially inferior and unable to assimilate into white Anglo-Saxon Protestant society (ibid: 17-20). Following World War I, the fear of a mass arrival of southern and eastern Europeans to Ellis Island led Congress to create quotas to restrict their immigration. The Johnson-Reed Immigration Act of 1924 cloaked the long-standing racist logic of US immigration law in less overt terminology, maintaining hierarchies of immigrant admission based on “national origins,” while continuing to exclude people already rendered ineligible for citizenship (namely, Asian people). Nations of the Western Hemisphere were not subject to the quotas, reflecting, in part, an economic demand for Mexican workers (Ngai 2004: 21-25). Nonetheless, a powerful nativist 100  movement sparked a massive repatriation and deportation of people to Mexico (including many US citizens) in the 1920s and 1930s (ibid: 71-75).  As historian Mae Ngai (2004) has documented, European and Canadian immigrants also became vulnerable to deportation at this time, yet an influential social welfare and legal advocacy movement coalesced in their defense, insisting that deporting a noncitizen with deep family and community roots in the United States (e.g. for committing a minor crime) would create undue “hardship.” Advocates argued for greater discretion in the administration of immigration law in the name of the unity of mixed-status families, yet largely did not challenge the racial discrimination at the law’s core. The reforms that resulted from this advocacy were uneven: “while European immigrants with criminal records could be constructed as ‘deserving,’ Mexicans who were apprehended without proper documents had little chance of escaping either the stigma of criminalization or the fate of deportation” (Ngai 2004: 75-82).  The Immigration and Naturalization Service (INS) also created a special procedure that, by the mid-1930s, allowed Canadians living in the United States with unauthorized immigration status to easily regularize their status. This practice was generally applied to Canadians with US citizen family on the grounds their separation would be a hardship. The INS soon extended this leniency to Europeans as well, while denying it to Asians and Mexicans (Ngai 2004: 84-86). Ngai (2004: 87) writes: The racism of the policy was profound, for it denied, a priori, that deportation could cause hardship for the families of non-Europeans. In stressing family values, moreover, the policy recognized only one kind of family, the intact nuclear family residing in the United States, and ignored transnational families. It failed to recognize that many undocumented male migrants who came to the United States alone in fact maintained family households in their home country and that migration-remittance was another kind of strategy for family subsistence.   As Ngai’s analysis illustrates, this early twentieth century use of administrative discretion reinforced racial hierarchies. European and Canadian mixed-status families were constructed as 101  deserving of unity, while Mexican and Asian families were not. Although the national quota system was ultimately abolished in 1965, it undoubtedly contributed to a racialized imaginary of the “illegal alien” – an imaginary that continues to center Latinos today. Further, structural shifts in United States immigration law have created a massive backlog in the issuance of family-based visas to citizens of Mexico and other Latin American countries. This can translate into many years of separation for transnational families, while relatives are obliged to wait in the country of origin – leading to the difficult choice to keep waiting, or to attempt an unauthorized entry into the United States. Many choose the latter, which has only expanded the number of people living in the US with undocumented status (Kanstroom 2007: 225).  The historical overview presented throughout this section illustrates that forced family separation has long served as a tool of racialized governance in the United States. Before the abolition of slavery, slave-owners routinely divided enslaved family members and sold them for profit. The removal of Native American children from their families and communities and placement in far-flung boarding schools was used as a technique of genocide for roughly a century. Protections for family unity within the US immigration system were devised around the needs of white mixed-status families, while denied to people of color. I suggest this history of state-sanctioned violence helps normalize the separation of Central American asylum-seeking families today. Building on this historical trajectory, I now examine the emergence of a threat narrative about asylum seekers in 2014. I find that this threat narrative, combined with DHS’s nearly-unchecked authority over noncitizen detention practices, further normalizes the forced separation of asylum-seeking families today.  102  4.2 A threat narrative emerges In June 2014, news headlines broke that large numbers of Central American children and families were arriving in South Texas. This population did not simply slip into the national spotlight by chance. Breitbart News was integral to breaking the news and shaping how the story would be told. With Stephen Bannon, who would become a political advisor for the Trump administration, in charge this media outlet has made a name for itself by publishing the openly racist, misogynist, Islamophobic, homophobic, and anti-Semitic thought that has been termed the “alt-right” (Latino USA 2016; also see Victor & Stack 2016).  On June 5, 2014, Breitbart News published a series of “internal federal government photos” depicting the conditions in which the US government was detaining recently-arrived noncitizens in Texas (Darby 2014). In the photos, stark white and gray cinder-block cells are crowded full of people dressed in jeans and t-shirts. In some cells, women and young children lay across the floor. Those trying to sleep have crinkled silver Mylar blankets draped across them, while most lay uncovered. People are so crowded in some cells that their bodies bend around each other. Other cells are full of men – sitting cross-legged, squatting, extended on the floor, or perched against the wall on hard benches. In pictures that are taken through foggy windows into the cells, the detained people’s faces are obscured, leaving an undifferentiated crowd appearing to fold back endlessly into the depths of the cell.  Rather than raising questions about the well-being of the people in these pictures, the purpose of their detention, or the conditions in which they were detained, the Breitbart caption accompanying the photos focuses on the “overwhelming task the Border Patrol is facing” to protect the “porous border” from so many “illegal immigrants” (Darby 2014). Another Breitbart article published on June 5 cites a Breitbart editor and “border security expert” named Sylvia 103  Longmire, who asserts that people are coming to the United States from Central America because of “gossip” that the border is open (Tate 2014, June 5): Many Americans don’t understand the power of word-of-mouth in Latin America; it’s like gossip in a small American town times ten. Word about anything, especially friends or family members going [through] the northbound migration or southbound deportation process, spreads very quickly.  Longmire goes on to elaborate the alleged power of gossip, and its relationship to family unity and release from detention: “Now that DHS doesn’t want to split up families and is releasing so many non-criminal illegal immigrants with only orders to return in 15 days for a court hearing, those [being] released are calling home” (cf Tate 2014, June 5, emphasis added). DHS had only one family detention center, Berks, with capacity to detain less than 100 people at the time. Thus, it could not detain many of the “family units” who it was apprehending because it did not have the space. According to Longmire, word of this “perceived amnesty” was spreading like wildfire back to affected Latin American communities (ibid).  Breitbart continued to push a narrative that asylum seekers and migrants were motivated by rumors of an open border and abundant public benefits for the taking. In this logic, President Obama was to blame for encouraging people to come, while the overwhelmed Border Patrol and a nation under siege were the victims (e.g. see Tate 2014, June 1). Some tenets of this narrative would also circulate through mainstream media and public policy debates – allowing lawmakers like Jan Brewer (then governor of Arizona) and Rick Perry (then governor of Texas) to demand that DHS stop releasing children and families into the United States and that they be deported as quickly as possible (see Carlsen 2014), as I describe next.  104  4.2.1 The rumors narrative informs policy debate In July 2014, Governor Perry testified at a House Committee on Homeland Security hearing about unaccompanied minors, suggesting the “short-sighted and tragic decision to essentially turn them loose into the United States” was not a humane option (US Congress 2014): Allowing them to remain here will only encourage the next group of individuals to undertake this very, very dangerous and life-threatening journey. Those who come must be sent back to demonstrate, in no uncertain terms, that risking your lives on the top of those trains and the ways that they are coming here, it is not worth that. Even those who have survived this very treacherous journey are still at risk. We already had one confirmed case of H1N1.  After suggesting that quick deportation was the most humane response, and implying that the children were bringing disease into the country (H1N1 refers to “swine flu”), Perry went on to argue that this was also a security crisis, because “Officials who should be guarding the border are dealing with the overflow instead of fulfilling their primary task” (US House of Congress 2014: np). This, he suggested, was making the border less secure, as drug cartels and transnational gangs were taking advantage of the opening. After visiting a McAllen, Texas Border Patrol detention facility, Perry wrote in a June 25, 2014 op-ed (cf ibid, emphasis added): the Federal officials who operate these facilities daily are doing the best they can with what they have, trying desperately to keep up with a seemingly unending tide of immigrants coming to our border because they’ve heard current US policy will quickly reunite them with loved ones in our country.  On June 20, 2014, then-House of Representatives Speaker John Boehner implored President Obama to deploy the National Guard to the US-Mexico border, as well as personnel to process deportations more quickly. Boehner asserted, “The policies of your administration have directly resulted in the belief by these immigrants that once they reach US soil, they will be able to stay here indefinitely” (Boehner 2014; also see Gogolak 2016).  It is easy to guess the policies to which Boehner was referring to: Deferred Action for Childhood Arrivals (DACA). President Obama established DACA through an executive order in 105  2012, making certain young people who were brought to the United States as children with undocumented status eligible for a work permit and temporary relief from deportation. DACA is not a path to citizenship, nor is it open to recently-arrived young people (USCIS 2017). By the summer of 2014, President Obama was on the cusp of signing additional executive orders: Deferred Action for Parents of Americans and Lawful Permanent Residents (DAPA), as well as an extension of DACA, which were ultimately blocked by a federal court after a coalition of twenty-six states challenged the orders (Texas v. United States 2015; USCIS 2015).  In November 2014, President Obama gave a national address about immigration. He rebuked Republicans for obstructing the immigration reform agenda he had worked with Congress to create, which passed as a Senate bill, but never made it through the House. President Obama rationalized his executive orders as necessary in the void of a legislative solution. In making the case for DAPA, he put the principle of family unity squarely at the center of his speech, asking, “Are we a nation that accepts the cruelty of ripping children from their parents’ arms? Or are we a nation that values families, and works together to keep them together?” (Obama 2014). President Obama went on to assert that the rightful targets of immigration enforcement are “Felons, not families. Criminals, not children. Gang members, not a mom who’s working hard to provide for her kids” (Obama 2014). These dichotomies give the impression of two categories that are mutually exclusive, with one category deserving of family unity and legal immigration status (family, children, working mom), and the other undeserving (felon, criminals, gang members). Although the arriving Central America families would easily seem to fit into the “deserving” categories, people who had entered the country recently did not qualify for DACA 106  or DAPA (ibid). The president stated this plainly in his speech, as though to clear up any rumors of an open border.  4.2.2 DHS uses rumors narrative to justify prolonged detention Months earlier, Department of Homeland Security leadership had already integrated the rumors narrative into its policy and practice. On June 10, 2014, DHS Secretary Johnson characterized the recently-arrived noncitizen children, families, and adults, as “illegal migration,” rather than as asylum seekers. He vowed to construct new detention centers to imprison them until their deportations were finalized. Directing his message to any would-be migrants, Johnson (2014) threatened, “you will be sent back home.” DHS has worked tirelessly to fulfill Secretary Johnson’s threats. One of the agency’s first moves was to hastily open new family detention centers, in which it detained women and children for prolonged periods. In January 2015, a group of formerly-detained women filed a class-action lawsuit, alleging that ICE had launched a “no-release” policy towards noncitizen mothers and children by denying the release of each “family unit” in order to deter other families from coming to the US in the future (RILR v. Johnson 2015: 5). They asserted that making individual custody determinations based on a philosophy of general deterrence was unlawful on multiple fronts – violating federal immigration law, undermining due process, and imposing an “arbitrary and capricious” change from DHS regulations (ibid: 7). DHS officials justified the agency’s policy of detaining mothers and children for prolonged periods as a necessary deterrence strategy by charting dubious connections between indefinite detention, national security, and the prevention of mass migration. The agency did so by invoking the rumors narrative, and also by drawing inspiration from previous denials of release to asylum seekers who had entered the United States in large groups, which it cited as 107  legal precedent (Manning 2015). One such denial is from 2007, when the Board of Immigration Appeals (BIA) ordered the respondent, a Haitian asylum seeker who had entered the country by sea with a group of other people, detained without bond. The BIA cited “national security interests” to deny this person’s release – the purpose being to deter “future surges in illegal migration.” (US DHS 2014, August 7: 50). In its arguments against the release of individual “family units” under the banner of general deterrence, DHS cited not only the cases of Haitian asylum seekers, but it also repeatedly filed the same affidavits from two DHS officials, Phillip T. Miller and Traci A. Lembke (Manning 2015). Miller’s affidavit states, “According to briefings of Guatemalan, Honduran, and Salvadoran detainees, the high probability of a prompt release, coupled with the likelihood of low or no bond, is among the reasons they are coming to the United States.” Thus, Miller argues, posting high bonds or simply denying bond altogether “would significantly reduce the unlawful mass migration” (US DHS 2014, August 7: 55). One risk of an “unlawful mass migration,” DHS official Lembke elaborates, is that “violent extremists and criminals can hide within this larger flow of migrants who intend no harm” (ibid: 60).  To drive home their view that releasing detained families would serve as an invitation to other families to come from El Salvador, Guatemala, and Honduras, both DHS officials cite a 2014 academic study (Hiskey, Malone, & Orcés 2014). The officials argue that releasing the families would be a magnet for more migration due to the “friends and family effect” (US DHS 2014, August 7). This social science theory suggests that migration begets more migration when migrants who have arrived and established themselves can support the arrival of others from their community by sending remittances and sharing information (Hiskey, Malone, & Orcés 2014: 5; citing Massey, et al 2005; also see Hiskey, Montalvo, & Orcés 2014). 108  This reading of the “friends and family effect” frames asylum seekers’ kinship and community connections as a source of threat. This DHS interpretation, combined with prolonged detentions of Haitian asylum seekers in the recent past, provided the basis for DHS to lock Central American mothers and children up for weeks, months, a year, or more in the name of national security. 4.2.3 Litigation challenges prolonged detention as “deterrence" In February 2015, a federal court determined that DHS was using the principle of general deterrence to guide its individual custody determinations, and granted a preliminary injunction to halt the practice (RILR v. Johnson 2015). The court asserted that general deterrence is not a permissible rationale to deny an individual’s release. The court further mused that even if this justification were permissible, DHS had failed to provide evidence of a national security need, or even of the efficacy of this practice in discouraging migration (ibid: 35-36): The court questioned the DHS interpretation of the “friends and family effect,” citing Jonathan Hiskey, one of the co-authors of the academic report utilized by DHS to justify its practices. In an affidavit submitted to the court, Hiskey (2014: 4) argued that: The idea that the women and children currently being detained at the border will, if released, begin sending remittances back to family members back home is highly unlikely given that the children will not be working and the women will be focused on simply feeding their own children and will be dependent on the income of sponsors or family members to which they are released.   The “friends and family” effect, Hiskey clarified, was not likely to apply to the asylum-seeking women and children coming from Central America. Hiskey and other scholars who provided declarations for the litigation further argued that rumors of an open border were unlikely to be the principle driving cause of Central American migration (RILR v. Johnson 2015: 36-37). This lawsuit shed light on some of the most egregious and unlawful elements of DHS practices towards “family units” imposed in the aftermath of June 2014. DHS rationalized its 109  nearly categorical denial of release to detained asylum-seeking mothers and children as a national security measure – a policy to deter other families from coming during a time of mass migration. In doing so, DHS officials invoked the same narrative peddled by the alt-right – that families were coming to the US because they heard rumors of an open border. The shaky use of social science theory at the heart of this justification ignores the well-documented fears of persecution expressed by the arriving population (see UNHCR 2014; 2015). As a result of this lawsuit, in May 2015 ICE established a new policy that it would not use general deterrence as a factor in its individual custody determinations for “family units” (ACLU 2015), while DHS began releasing mothers and children from family detention centers, on average, more quickly than before (Flores v. Lynch 2016, September 30: 4). Ultimately, though, DHS continues to detain mothers and children en masse in its three family detention centers, and other relatives in separate adult facilities. Some mothers and children still do spend months, a year, or longer in family detention (Berks Mothers 2016).  In 2014, a threat narrative emerged about the Central American asylum seekers arriving at the US-Mexico border – asserting that rumors of entire asylum-seeking families being quickly released from detention, together, was causing mass migration. DHS used this logic to justify its expansion of family detention as a deterrence tactic. This tactic causes family separations that are obscured by the narrow DHS definition of the “family unit,” as I explain next.   4.3 Silence and doublespeak around family separation The Department of Homeland Security, and its component agencies CBP and ICE, are notoriously opaque with the public about their practices. Given that DHS is characteristically silent and enigmatic about its practice of splitting families up in custody near the US-Mexico 110  border, I submitted several Freedom of Information Act (FOIA) requests to learn about its policies on family unity and separation. Although the results did not entirely answer my question, and were sometimes riddled with redactions, the silences were revealing of who counts as a family, and whose unity matters to DHS. 4.3.1 An eerie silence on adult family unity I sent a carefully-worded, paragraph-long FOIA request to ICE inquiring about its policy on separating “family units" in its custody determinations for long-term detention. I clarified that by “family unit,” I meant any given combination of adult and/or minor relatives apprehended together (US ICE 2016, April 13). When I sent my request, I assumed that the term “family unit,” which I had seen used by DHS, was bureaucratic jargon to refer to any configuration of family members who were apprehended together. When the ICE FOIA office wrote back, I quickly gleaned was that for ICE, a “family unit” solely refers to a parent accompanied by a minor child. My FOIA response did not explain how or why the agency separates adult family members, nor did it speak to any policy about their unity. Instead, it described the conditions under which ICE separates children from parents. It included some chilling examples presented without context, such as when “An adult female resident ingested shampoo and was sent to the hospital for evaluation and stabilization” (US ICE 2016. April 13: 2). Redacted pieces of an email thread entitled “Mother Cut Veins in Karnes” describes a mother detained in Texas for nine months who attempted suicide after the appeal of her deportation order was denied (ibid: 3-4).  Disturbed by ICE practices towards mothers and children, yet still without an answer as to any ICE policy on adult family separations, I submitted another FOIA request. This time I inquired very specifically into ICE policy on the unity of adult family members like spouses or 111  siblings taken into custody together (US ICE 2016, September 20). The weeks went by, only for ICE to once again send results that focus on the parent/minor child relationship, but this time detailing what happens when a minor child reaches age eighteen while in custody alongside their parent (ibid). Hoping to speak to an ICE official, I also sent my questions to the ICE Office of Public Affairs, which did not respond (Email April 7, 2016). This correspondence left me with the sense that for ICE, and DHS more generally, the familial bonds of loved ones arrested together do not carry much importance, or perhaps are not even recognizable as such unless there is an accompanying child. My correspondence with ICE also raises questions about intentionality in the practice of family separation, which I examine in Sections 4.3.2 and 4.3.3. As outlined in Chapter 3, if DHS lacks an incentive or means to prioritize and track adult family relationships, there is a good reason that it concerns itself with child welfare (at least in written policy, if not in practice). The protections for the family established by the 1997 Flores agreement, as Lauren Martin (2011) has demonstrated, center on a child’s right to be with their parents, without challenging the broader criminalization of noncitizen adults.  Flores litigation has compelled DHS to generally keep children with a parent, for the sake of the child. I have not found evidence that the agency is similarly required to keep any other configuration of family together, or even to track separations that occur in its custody. DHS is particularly silent about whether adults have any right to family unity. As I suggested in Chapter 3, the provision of the right to family unity to children, and denial of that same right to their parents, embodied within the Flores agreement demonstrates the limits and contradictions of a liberal system of individual rights. In liberal democratic systems such as that of the United States, the individual subject becomes legible, in part, by properly 112  performing their role in the nuclear family. Over time, this idealized family structure has been implicitly framed around the white, heterosexual, patriarchal, bourgeois family. Thus, a liberal democratic system can function in a paradoxical way – at once promising to prioritize the social bonds of loved ones, while regularly dismantling families that fall outside this ideal. The result is a disregard for the “human relationality” (McKittrick 2011; Gilmore 2007; also see Butler 2004) of families targeted for separation. The well-being of such families is treated as an individual matter, thereby obscuring the role that social support networks play in keeping the individual happy, healthy, and safe. Further, centering the child/parent nuclear family as the family unit whose unity matters also reinforces the heteronormative structure integral to the bourgeois family imaginary (Martin 2011). This configuration disregards the separation of any other grouping of loved ones apprehended together, and serves as a reminder that “family unity” has never been a right open to all. This is hard to miss in my correspondence with ICE, in which the agency could only make sense of my questions about “family” as questions about parents and minor children.  This is also evident in recent Flores litigation. In February 2015, Flores class members filed a motion in federal court to enforce the Flores agreement. The litigation involved a conflict over two promises within the agreement: to release children quickly from detention, and to prioritize releasing them to a parent. Flores class members argued that this provision should favor the release of a parent who is detained with their child. The judge who heard the case agreed (Flores v. Johnson 2015: 24-25). When DHS appealed, the Ninth Circuit reversed those findings in part, arguing, “The fact that the Settlement grants class members a right to preferential release to a parent over others does not mean that the government must also make a 113  parent available; it simply means that, if available, a parent is the first choice” (Flores v. Lynch 2016, June 7: 19, emphasis added).  The Ninth Circuit’s interpretation obscures DHS responsibility in making the parent unavailable in the first place by opting to detain them. Further, I believe the DHS silence on whether adults have any right to be with their family, and the prevailing narrow readings of the Flores agreement, could give the impression that the division of adult family members is a bureaucratic accident resulting from the uptick in detention since 2014. In other words, one could look at it as collateral damage caused by a network of agencies that lack the means and incentives to track and maintain family unity (see LIRS 2016).  However, as I outline next, there are times when the state actors involved in deportation actively use separation against families. I further suggest that whether most separations are accidental or not, even when separations are unintentional the result is effectively still a punishment. Ultimately, the effect is to cause harm to the targeted family and reinforce the separation of Central American asylum-seeking families as commonsense, rendering them illegible as a family unit whose unity should matter. 4.3.2 Separation as threat and punishment The Flores agreement requires that children be released from detention within 5 days, which the US government has generally failed to achieve since 2014 (Flores v. Lynch 2016, September 19; also see RILR v. Johnson 2015). DHS has justified this failure by categorizing the current arrival of noncitizen children as exceptional – an “influx of minors,” which allows for an exception to the 5-day limit (Flores v. Lynch 2016, September 30: 3-4). The California federal court hearing the recent Flores litigation granted DHS an exception to the 5-day limit, extending the maximum processing period to 20 days (Flores v. Lynch 2016, September 19: 9). 114  In September 2016, a coalition of immigrant rights organizations filed an amicus brief with the court in support of Flores class members (Flores v. Lynch 2016, September 19). The advocates asserted that DHS is placing most family units into expedited removal, rather than standard deportation proceedings, as part of its broader deterrence strategy. When DHS places children and their parents into expedited removal, it then tends to keep them detained for weeks or months – a period that exceeds that allowed by the Flores agreement (ibid: 2-3).  DHS responded to this brief with a sinister threat, asking whether advocates wanted ICE to release minors from family detention into the custody of the Department of Health and Human Services (HHS, which holds custody of unaccompanied children through its component agency, the Office of Refugee Resettlement [ORR]), thereby splitting them apart from their parents. Or, DHS asks, should it maintain family unity by detaining them with their accompanying parent (Flores v. Lynch 2016, September 30: 4-5, emphasis added)? DHS stated that minors spend an average of 34 days in ORR custody, so splitting families up would prolong the process, with a “negative impact on family unity” (ibid: 7), given that ICE was releasing family units from family detention more quickly – on average within 11.8 days (ibid: 4).  In this way, DHS implies that family detention is a desirable way to keep loved ones together, while overtly threatening to institutionalize the separation of parents from children. These are not empty threats, nor is it the first time that DHS officials have made this argument.  During a previous experiment with mass family detention a decade ago, DHS regularly divided the relatives who it arrested together when it chose to detain the adults, while placing the children into ORR custody (WCRWC & LIRS 2007: 1). DHS did not have enough space to accommodate all the families it sought to detain (having only one family facility at the time), and separating adults from children was the initial solution. Advocates and some members of 115  Congress vocally protested (Libal, Martin, & Porter 2013). As a result, DHS established the Hutto Family Residential Center in 2006 to detain mothers and children together. DHS thus expanded family detention under the banner of family unity (ibid: 262; ACFRC 2016: 150).  Advocates report that one way staff disciplined family-detainees in Hutto was by threatening to separate them – for example, if a detained child fought with a staff member or with another child. In Hutto, “The prevailing belief among families [was] that they [would] be separated if anyone misbehaves, which can create an environment of extreme psychological distress” (WCRWC & LIRS 2007: 30). The Obama administration shuttered Hutto as a family detention center in 2009. Yet since 2014, DHS has recycled some of the same logic to justify its revival of mass family detention. Advocates also report that DHS has used separation as a method of punishment and deterrence for a wide variety of families apprehended near the US-Mexico border since at least 2005 through its Consequence Delivery System (CDS) enforcement strategy. CDS subjects people with previous deportations to criminal prosecution, which can initiate the separation of family members who have different immigration histories (WRC, LIRS & KIND 2017: 10). Further, DHS has a practice of intentionally deporting Mexican family members to different locations, which doubles the odds that deportees will encounter abuse shortly after their return to Mexico (ibid: 11; Danielson 2015). The most recent, high profile threat to further entrench family separation within deportation policy came directly from Trump advisor John Kelly (then-DHS Secretary) in March 2017. In a televised interview with CNN, Kelly confirmed that DHS was considering a policy of systematically splitting children apart from parents in order to deter them from taking a dangerous journey north through Mexico (CNN 2017). Although Kelly denied making such a 116  proposal just a few weeks later (Kopan 2017), the threat speaks for itself. Splitting families up or simply threatening to do so can be a way of exerting control over those families.  4.3.3 Evaluating effect and intention As outlined above, there is clear evidence that certain cases of family separation are carried out intentionally as an effort to punish or deter families. The existence or degree of intentionality perhaps bears on what strategies are most effective to challenge each specific practice. However, these intentional acts coexist in a complex way with separations that are likely more ad hoc, or in which intentionality would be difficult to prove. In Chapter 3, I presented evidence that splitting relatives up can weaken their well-being and access to humanitarian protection. For Angela, an attorney who I interviewed, these negative effects matter more than trying to parse out whether each separation is accidental or intentional:  I don’t think you even have to go as far as whether it’s intentional.…You shouldn’t have frameworks that have perverse consequences with [them], where they weaken people’s claims so that they do get returned to the country where they face persecution. So as soon as you become aware that that is happening, it should be fixed!   There is no evidence of a concerted Department of Homeland Security effort to fix this problem. Further, the attorneys whom I interviewed had themselves witnessed DHS officials and adjudicators actively using family separation against their clients, even if the family’s initial division was perhaps unintentional, or occurred before the family arrived in the United States.  Angela described her experience representing an indigenous, Quiche-speaking family from Guatemala whose coyote split them up before they crossed the US-Mexico border. When Angela met the mother in a Texas family detention center, she was desperate to find her 19-year-old daughter, who had crossed the border separately. The day that Angela located the daughter, who was imprisoned in a different detention center, Angela began calling and faxing that facility, requesting that they provide the daughter with a credible fear interview. 117  At 6:00 AM the next morning, Angela learned that despite her repeated efforts to contact the facility over the course of fourteen hours, the daughter had been deported in the early hours of the morning. Angela’s next move was to petition CBP to rescind its deportation order against the daughter, given the circumstances. CBP refused to do so, evading responsibility for DHS ignoring Angela’s efforts to inform the facility of the Quiche-speaking daughter’s need for a credible fear interview, and for the arbitrarily-different legal outcomes that DHS had imposed upon the family (in which the daughter was summarily deported without an opportunity to explain her fear to return). “It’s just completely illogical, makes you want to pound her heard against a concrete wall,” Angela said about the experience. Another attorney, Matthew, had a similar experience. A mother and her two youngest children were detained in one of the Texas family detention centers. The mother received a negative finding on her credible fear interview. She appealed, and an immigration judge reversed it, granting her a positive finding. As a result, the mother and the two youngest children were able to pay bond and be released to continue their case before the immigration court. However, her 19-year-old son was detained separately at an adult facility in Stewart, Georgia. After months in detention, he withdrew his asylum application because he could not bear to be detained any longer. Shortly thereafter, Matthew and allied attorneys in Georgia tried to reopen the son’s case because there were new circumstances (his mother’s positive credible fear finding). An immigration judge denied Matthew’s motion to reopen the 19-year-old son’s case under the rationale that he was in separate proceedings from his mother. Matthew clarified to me that the immigration judge erred in this decision, because the son was entitled to be a derivative on his mother’s asylum application until the age of 21. And 118  crucially, “the only reason he’s in separate proceedings is because they sent him to a separate detention center. They’re creating these circumstances and then using them against people.”  Even when families are unintentionally split by state actors, the result is still effectively a punishment. This evidence runs counter to DHS arguments that mass family detention is in the best interest of the child. In this exceptional moment of mass migration, family detention keeps families together and allows for an efficient processing of their case, the logic goes. DHS uses this logic to justify a form of “family unity” that often splits relatives apart when the agency arrests a family composed of anyone beyond the mother/child “family unit.”  The argument that a form of family detention, which, in practice, routinely splits families up is in the best interest of the child (because it keeps families together) is deeply contradictory. It has the feel of Orwellian doublespeak. For me, this doublespeak suggests that family separation in detention is something more than a bureaucratic accident. Ultimately, DHS wields the power to define the kinship relations that count as family. The executive branch also makes crucial choices about which noncitizen families will be prioritized for immigration enforcement, and thus which families will be forced apart. Social isolation through detention, whether intentional or not, is not simply the outcome of a large, uncaring bureaucracy at work. It is an assertion of power and control, that contributes to a broader deterrence strategy (see Martin 2012b). This reality leads me to argue that the forced separation of Central American asylum-seeking families is a form of racialized governance.  4.4 Conclusion State-sanctioned family separations rooted in the past, from the division of enslaved families for profit, to the boarding school system that systematically removed indigenous children from their 119  homes as a technique of genocide, live on in countless ways. They set an enduring precedent by reserving family unity as a right only available to certain families, particularly to the normative white, heterosexual, bourgeois, nuclear family. This point holds true in the history of immigration policy as well, in which discretionary relief from deportation for the sake of “family unity” was built around the needs of European and Canadian families. Today’s forced separation of asylum-seeking families near the US-Mexico border is a practice that, at first glance, could look like an exception or aberration (see Menjívar & Abrego 2012: 1388). This, combined with the commonsense embedded within human rights norms that state separation of intact families is rare, could make the division of asylum-seeking families seem unusual. Further, this painful separation happens behind bars, not visible to the public. Finally, because harmful long-term impacts unfold across time and space, they are difficult to track for those not directly impacted, and state responsibility is often occluded. Formulating an argument for “family unity” is politically fraught. DHS rationalizes its mass detention of noncitizen mothers and children in the name of family unity – as a desirable way to keep families together and children safe. The only viable alternative, DHS has threatened, would be to systematically split parents and children apart. These are the only options that DHS concedes, despite the fact that it could simply release asylum seekers instead of detaining them. Thus, an argument for family unity is not necessarily liberatory. It can also be used as a tool of racialized governance, much like the act of separating family members in DHS custody. My realization that arguments in favor of the “the family” can be twisted in many directions (and thereby risk reinforcing an ideal that is oppressive towards women and marginalizes those who fall outside), and of the historical banality of forced family separation in the United States, left me struggling with how to advocate for family unity. As Wendy Brown 120  (2000) argues, the pursuit of rights in a liberal regime is rife with paradox. The extension of rights may establish crucial material improvements for the rights-bearing subject, but often does not address systemic root causes, and may allow the state to continue defining and regulating that subject (in this case, family members in detention). Further, as Geraldine Pratt (2012: 40) notes, liberalism “has been a mechanism to legitimate and mystify social stratification and racial, gender, and other subordinations.” But following Pratt and Brown, some political potential can exist within liberalism’s paradoxes. Brown (2000: 240) poses the question, “How might paradox gain political richness when it is understood as affirming the impossibility of justice in the present, and as articulating the conditions and contours of justice in the future?”  Ultimately, I have concluded that an awareness of the historical banality of forced family separation in the US can encourage carefully-crafted arguments for family unity, which avoid delimiting what counts as a family, or how families must live. I believe that more expansive models of family life could serve as inspiration for the “contours of justice in the future” (Brown 2000: 240). For instance, inspired by an artistic exploration of “queer family making,” Maggie Nelson (2015: 72-73) writes that “nothing we do in this life need have a lid crammed on it…no one set of practices or relations has the monopoly on the so-called radical, or the so-called normative.” In other words, the family and the home can be what you make them. This point has also long been recognized by black feminist scholars and activists who argue that the home is not simply a site of gender oppression. It can also be a source of refuge and a wellspring of transgressive politics (hooks 1990, cf Pratt 2002: 197; also see Collins 2000; Lorde 1984). Loved ones’ care for one another in the face of formidable barriers to being together, like imprisonment, can help generate life – enacting a counterweight against social death (Gilmore 2007; McKittrick 2011; also see Tadiar 2015). 121   An awareness of the racialized history of family separation is crucial to crafting arguments for family unity that contribute to, rather than detract from, struggles against the multiple, overlapping forms of oppression linked to the governance of family life. This awareness could productively inform a range of advocacy efforts – from the writing of a report meant to inform public policy, to the framing of scholarly research. In practice, I believe this awareness could prompt advocacy that portrays “the family” in as broad a manner as possible. In the case of Central American asylum seekers, this means advocating not only for the unity of mothers and children, but also any other combination of loved ones who desire to be together. Ultimately, advocacy for family unity can be quite radical if also situated within broader decolonizing struggles, such as the #Not1More movement to end all deportations in the United States. I elaborate this point further in Chapter 5. 122  Chapter 5: Bad hombres? Gendered scripts and exclusion in the US asylum system  Most of the Guatemalan, Honduran, and Salvadoran citizens deported from the United States and Mexico tend to be young men. Between 2010 and 2014, upwards of 80 percent of deportees over age 15 from these three countries were male; more than 60 percent of Guatemalan, Honduran, and Salvadoran deportees were below age 29. Deportees also tended to have limited education, and work in low-skilled jobs or be unemployed. Most had no criminal convictions, which runs “contrary to the stereotype of the young Central American gang member” (Dominguez-Villegas & Rietig 2015: 2). This demographic is deported at high rates, and can face severe harm or even death upon their return (see Brodzinsky & Pilkington 2015), yet the barriers that young men face to accessing asylum are generally not at the forefront of advocacy efforts.  When the spotlight does fall on this demographic, it often invokes stereotypes about criminality. In countless examples, President Donald Trump imagines a vast noncitizen criminal population gendered as male and racialized as Latino – perhaps most notoriously through his depiction of Mexican people as “rapists” during his June 2015 candidacy announcement (Ross 2016). In an October 2016 presidential debate, he promised to “get all of the drug lords, all of the bad ones,” the “bad hombres” out of the country (Levine 2016). In July 2017, Trump dubbed people involved with the Mara Salvatrucha, or MS-13, “animals,” and launched immigration raids against alleged gang members – targeting teens based on their clothing preferences, tattoos, and gang presence where they live (Democracy Now 2017). 123  Although the loudspeaker that Trump has given to this racist rhetoric is somewhat novel for a president, the punitive policing of young Latino men, and broader criminalization of Latinos in the United States are nothing new (Rios 2011; Chavez 2013; Ngai 2004). Over the last two decades, predominant threat narratives about noncitizens have tended to focus on men – conjuring a specter of alleged “terrorists” and “criminal aliens” trying to force their way into the country (Golash-Boza & Hondagneu-Sotelo 2013: 273). This chapter is guided by the following question: what barriers do Central American men negotiate to make their asylum claims heard? In analyzing this question, I draw from studies of gender, political violence, and displacement, which have illuminated the “bad scripts” that women must negotiate to make their experiences heard – scripts circumscribed by normative femininity (Taylor 1997: 184). In the United States asylum system, harms commonly directed at women, such as domestic violence, were long rejected by adjudicators as “personal” or “cultural,” outside the realm of politics (Calavita 2006: 111). Further, the most successful women’s gender-related claims tend to place culpability on foreign, cultural “backwardness,” blaming foreign men for “subjugating helpless women” (Sinha 2001: 1581). This creates a dynamic in which, following the insights of postcolonial feminist scholar Gayatri Spivak (cf Oxford 2005: 22), “white men are saving brown women from brown men.” In this chapter, I examine the role of normative masculinity in a broad practice of exclusion towards Central American asylum seekers. After explaining my research methods and analytical approach, including what I mean by normative masculinity and femininity, I consider the difficult environment where many recently-arrived asylum seekers compile and present their cases: within the confines of fast-track deportation and detention centers, which expose detainees to myriad barriers before they even present their claim. I then juxtapose two types of Central 124  American asylum claims that have had some limited success (family-based claims and women’s gender-based violence claims), with one that is strikingly unsuccessful (opposition to gang claims).  I find that in their denial of gang opposition claims, which are often, although not exclusively presented by young men,11 adjudicators echo some of the same gendered scripts long used to deny women’s gender-based violence cases – by framing this violence as outside the realm of politics. I argue that this slippage of gendered exclusion reflects a broad climate of hostility towards Central American asylum claims. It also underscores the point that systems of male supremacy tend to benefit men who embody hegemonic ideals of masculinity, yet are also oppressive towards men who are unable to, or refuse to fit into the role prescribed to them (Harris 2000; Jackson 1991). I elaborate this conceptual point next.  5.1 Analytical framework and methods Masculinity and femininity refer to the binary, normative expressions of gender identity that have come to be associated with men and women within Western thought. From Hegel, to Rousseau, Plato, Descartes, and Marx, some of the most influential Western thinkers have theorized social difference around a dualistic notion of reason and nature, in which reason is seen as superior to nature (Plumwood 1993: Chapter 2). The capacity for reason is gendered as a masculine trait – attributed to elite men, who are framed as fully human. In contrast, the nature                                                  11 Contemporary gang recruitment in Central America has gendered dimensions. As UNHCR (2016: 36) reports in the case of El Salvador, boys and young men are generally targeted for gang membership, while girls and young women are often, although not exclusively, targeted to become the partners of gang members. 125  side of the binary is gendered as a feminine trait, and has also been mapped onto other marginalized groups like enslaved and colonized peoples (ibid: 44).  Intersectional feminist theory, such as Kimberlé Crenshaw’s (1991) classic analysis, has demonstrated that gender cannot be understood in isolation from other axes of social difference, like race and class. This is a central insight of black feminist thought, which has shown that predominant ideals of femininity in the United States developed around the white, middle-class woman, who was valued for her passivity, fragility, and her role in the home serving as a wife and mother. This ideal has historically excluded colonized women, women of color, and working-class women obliged to work outside the home (Roberts 1997: 10; Collins 2000; Lugones 2016; McClintock 1995; Stoler 1995).  Critical race scholarship on masculinity has also demonstrated the need for an intersectional understanding of masculine ideals and the differential ways that men embody gender roles. Speaking to the United States context, sociologist Karen Pyke defines “hegemonic masculinity” around a white, heterosexual, upper-class, male identity (cf Harris 2000: 783). Men of color have historically been excluded from this norm, and therefore excluded from the upper echelons of power in the US. For instance, since the abolition of slavery, white society has perpetuated stereotypes around black men as “violent, unable to control their physical and sexual urges, and unintelligent” (Harris 2000: 783; also see Alexander 2012: Chapter 1).  As legal scholar Angela P. Harris (2000: 781) notes, “Manliness is one of those ideas that is often made real through violence.” People who commit crimes overwhelmingly tend to be men, while law enforcement agencies have historically been male-dominant in the US. This is not to suggest that men are naturally predisposed to conflict. Rather, violence can offer a viable way for men to perform their masculinity – whether in a form that is criminalized or applauded 126  by society (ibid; Ríos 2011). Harris (2000) makes a key point: hierarchies of difference between men render certain men vulnerable to the violence of other men. Cold War constructions of masculinity in Central America speak to the vulnerability of both men and women to violence committed by men. In Guatemala, for instance, soldiers used rape, as well as the assassination of indigenous women and children as weapons of war. These techniques were likely intended, in part, as an assault on men’s self-perception as protectors (Enloe 1993: 121). Given that soldiers were often recruited unwillingly, Cynthia Enloe (1993: 121) raises the question, “Were new recruits pressed to participate in rape during military maneuvers as a technique for socializing them into a kind of brutality, thereby severing their ties from their civilian compatriots?” As the Cold War wound down, Enloe (1993) argued that such militarized constructions of masculinity meant to provoke conflict would not just vanish with formal declarations of peace. This prediction rings true across contemporary El Salvador and the US (see Hume 2008) – suggesting that Cold War masculinities endure in certain ways.      Today’s US criminal justice system provides another example of how hierarchies of difference between men render certain men vulnerable to violence committed by men. Legal scholar Michelle Alexander (2012) links the rise of mass incarceration in the United States back to the Jim Crow era following the Civil War, in which white people and the state institutions at their disposal continued to assert violent social control, including by enacting laws that strategically criminalized black people. When the Civil Rights Act ended legal segregation in 1964, these efforts to maintain white supremacy gave rise to punitive criminal justice policies promising to enforce “law and order.” The “drug war,” from the Nixon administration onward, set the groundwork for a system of mass incarceration that continues to disproportionately target 127  black men for minor offenses, as compared to white men (Alexander 2012: Chapter 1; US Sentencing Commission 2017).  The rise of punitive US immigration policy in recent decades, which has made an ever-expanding population of noncitizens vulnerable to detention and deportation, was directly enabled by the rise of “drug war” criminal justice policy (García Hernández 2014). The demographic of noncitizens most criminalized, and most directly targeted by this system tend to be working-class men from Latin America and the Caribbean, creating “a rupture with previous restrictionist immigration regimes that had focused on excluding women (from reproduction) and including men for labor (production)” (Golash-Boza & Hondagneu-Sotelo 2013: 273). Since the turn of the twenty-first century, with the confluence of the criminalization of young men of color, climate of Islamophobia after 9/11, and male unemployment due to economic restructuring and the 2008 financial crisis, “the new danger is masculine” (ibid: 273).   In making this point, my intention is not to ignore cisgender and transgender women’s uniquely-difficult experiences with detention and deportation (e.g. see HRW 2016; Rabin 2009; UNHCR 2015; WRC 2017). This is no small matter. Between 2009 and 2016, the percentage of women among noncitizen detainees nationwide grew from 9 percent to 14.5 percent (WRC 2017: 2). I see a focus on normative masculinity and criminalization as complementary to the important research being done about women. Taken together, analyses of differently-gendered experiences shed light on broader United States practices of exclusion towards Central Americans at large.  To this end, I draw from two key research methods in this chapter: (1) an analysis of asylum adjudication trends, and (2) observations of the asylum process. To deepen my understanding of trends in the adjudication of Central American asylum claims, I analyzed roughly forty asylum case decisions on appeal issued by the US Board of Immigration Appeals 128  (BIA), the Federal Circuit Courts of Appeal, and the Supreme Court – mostly issued within roughly the last 15 years. To begin my search for relevant cases, I first reviewed practitioner guides and legal scholarship. The citations within one case decision often led me to other similar cases, and particularly back to precedential decisions.  I follow Susan Bibler Coutin’s (2011: 570, citing Merry 2004) approach to the legal ‘archaeology’ of Central American asylum claims. Coutin looks at the “layering of documents, statutes, court cases, notices, and records that take form at particular historical moments.” As Coutin notes, these developments are not static. They carry forward, creating openings and barriers for future asylum seekers in the legal precedent they set (ibid). I follow a similar approach, although with a central focus on the role of normative masculinity (following Mascini & van Bochove 2009; Spijkerboer 1999), looking at both material practices of governance and knowledge production. I contextualize my analysis of the case decisions that I reviewed and my observation of asylum processes by drawing from secondary legal and advocacy scholarship and from a small number of interviews with attorneys and legal advocates.     Focusing solely on the adjudication of asylum claims would leave out a crucial part of the picture for Central American asylum seekers today: many people do not even reach this final stage in the legal process. Thus, I think my analysis would be incomplete without illustrating what goes on before an asylum seeker is even allowed to present the merits of their claim before an immigration judge. My understanding of these barriers is informed by a month that I spent as a volunteer with the Florence Immigrant & Refugee Rights Project’s (Florence Project) adult 129  program in May 2016.12 The Florence Project is a non-profit legal services organization that provides crucial services to adults detained in three Arizona detention centers: two in Florence, Arizona, and one in Eloy, Arizona. As a volunteer, I sought to learn about the conditions in which many adult Central American asylum seekers seek relief from deportation, and the process they must follow to do so. My time in Arizona informs my analysis in the following section, which illustrates some of the initial barriers to accessing asylum.  5.2 Pursuing asylum behind bars The Eloy Detention Center is a sprawling, pallid-gray complex encircled by a fence brimming with loops of barbed wire (see Figure 3). The muted buildings almost blend into their dusty, flat surroundings along the highway halfway between Tucson and Phoenix, Arizona. A lonely cactus stands between signs naming the two proprietors: Corrections Corporation of America (CCA, the private prison contractor that runs the facility), and the Department of Homeland Security (DHS). Inside, as many as 1,500 noncitizen men and women are detained on any given day (CCA 2017b).                                                  12 As noted in Chapter 1, I signed a confidentiality waiver to protect Florence Project client information. To ensure that I have adhered to the terms of the waiver, my writing in this chapter that refers to the asylum processes that I observed was reviewed by a Florence Project staff member. Further, as mentioned in Chapter 1, the views expressed throughout this dissertation, and any shortcomings, are my own. 130   Figure 3: Eloy Detention Center, October 2016; photo by Leigh Barrick I first entered this complex in May 2016. Guards buzzed me inside through several rounds of locked doors, and, after clearing a metal detector, I entered the visitation room. Waiting expectantly for a legal orientation with the Florence Project, sitting on rows of seafoam green and lavender benches, were about thirty men. They were mostly dressed in forest green prison scrubs, indicating their ‘low risk’ security classification, while several men wore tan scrubs to signify ‘moderate risk.’ That day, almost all the participants were Latin American, save several men originating from South Asia and Africa. Some men had recently arrived in the US, while others had lived in the country for many years. Although they were there to learn about their rights, including to seek relief from deportation, the odds that the detainees seeking asylum would avoid deportation were slim. Throughout this section, I explain why. 131  5.2.1 The status quo Location matters a great deal for the outcome of an asylum case. There are few worse places in the United States to be an asylum seeker than Eloy. In Fiscal Year (FY) 2015, the four immigration judges in Eloy granted only 11 asylum claims, and denied 115 claims (a grant rate of 9%). In contrast, the nationwide average grant rate that year was much higher, at 48% (US EOIR 2016: K1-K2). It is hard to win an asylum case while locked up in a place like Eloy, not only due to the Eloy court’s strikingly low grant rates, but because simply being detained anywhere makes it difficult to succeed. A detained person can struggle to gather evidence for their case. They also may be split apart from loved ones, and face health problems, inadequate medical care, trauma, language barriers, and other factors that sap their attention and energy away from their legal case (Sheppard & Murray 2017).  In recent years, ICE has become increasingly resistant to giving detained people the option to pay bond for their release, while raising its median bond rate (in the cases when it does offer bond) (TRAC 2016, September 14). Thus, the trend is towards prolonged detention – keeping people locked-up for weeks, months, a year, or longer, until their case is completed (AIC 2016, August: 2-4; TRAC 2016; July 19). Detained people and their allies have fought back against ICE’s prolonged detention, gaining some crucial and hard-won, but limited protections (TRAC 2016, September 14; e.g. Demore v. Kim 2003; Robbins v. Rodriguez 2015). For example, in June 2015, then-DHS Secretary Jeh Johnson announced reforms to the agency’s family detention policy, including that ICE had created “criteria for establishing a family's bond amount at a level that is reasonable and realistic, taking into account ability to pay” (US DHS 2015, June 24). It is unclear whether ICE has actually implemented this policy for families (HRF 2016: 25), while no such policy exists for most detainees in adult facilities like Eloy. 132  One attorney who I interviewed about family separation pointed out that, while family detention has received a high level of public scrutiny since 2014, DHS practices towards adults have not. Thus, mothers and children have gained some crucial protections against prolonged detention, yet “single” adults (including those separated from family), are stuck in the same system. Speaking about families that are split up, the attorney speculated as to why detained adults do not provoke the same level of attention: Mothers and children in a private jail is more appealing from a mass advocacy standpoint than, you know, a mother and child who are out but the father who came with them is separated and detained. I don't think that it's like a gender thing, but it's just, you just don't hear about it. Either people don't know, or it's not as offensive to their…sense of what's right and wrong, I guess, for one of the parents to be detained. Wherever the child's at, whoever has the kid is really going to benefit from the public outcry. And the other person just gets forgotten.   The person who gets forgotten – whether a father separated from his family, or any person categorized as a single adult – steps into the status quo: the largest noncitizen detention system in the world, which over the course of a year detains roughly 400,000 people (DWN 2017), and deports 407,075 (as of 2014, Zong and Batalova 2017). Being detained makes people unlikely to succeed in their case. But as I explain next, even getting one’s foot in the door to secure an immigration hearing in the first place is no easy feat. 5.2.2 Not everyone gets their day in court Asylum seekers who are either arrested by the Border Patrol after clandestinely crossing the US-Mexico border, or those who present themselves at a port of entry and declare their intention to seek asylum to a Customs and Border Protection (CBP) official, immediately face hurdles to making their fear known and taken seriously.  Border Patrol officers have the authority to place people apprehended near the US borders into a fast-track deportation process like expedited removal. However, if the detainee says they are afraid to return to their country, they should be given the opportunity for a credible 133  or reasonable fear interview with a US Citizenship and Immigration Services (USCIS) Asylum Officer (AIC 2016, August: 2).13 I say should because legal advocates have documented that some CBP officials actively discourage people from seeking asylum, and turn people away who declare their fear. They do so by providing misinformation, abusing asylum seekers verbally and physically, and trying to intimidate them, including by threatening to divide family members (Campos & Friedland 2014: 10; AIC 2016, January 13; Al Otro Lado v. Kelly 2017).   When CBP officials do attempt to perform their screening duty, they may fail to ask the required questions about fear, or may record a fake answer. One telling example is that of “Y-F,” who was interrogated under oath by Border Patrol agents in 2014. The agency’s written record ends with the following exchange: “Why did you leave your home country or country of last residence?” Y-F answered, “To look for work” (Matter of M-M-R 2015: 3-4). However, Y-F is an unlikely ‘economic migrant,’ given that he was only three-years-old. Adjudicators and DHS attorneys regularly use unreliable written records like this one against people (ibid: 5; USCIRF 2005; 2016).  Asylum seekers must dodge hurdles like these to secure a credible fear interview. The official purpose of this non-adversarial interview is “to quickly identify potentially meritorious claims to protection and to resolve frivolous ones with dispatch” (USCIS 2014: 11). Yet in 2014, USCIS suddenly raised its standard for “what should be a screening interview closer to full asylum adjudication” (USCIRF 2016: 2; Campos & Friedland 2014). It is important to note that most asylum seekers are detained when they have their credible fear interview, making it hard or                                                  13 People with a prior deportation order, or specific types of criminal convictions, are subject to a different process called reinstatement of removal, and have a “reasonable,” rather than “credible” fear interview as screening for their eligibility for humanitarian protections from deportation (AIC 2016, August: 3). 134  even impossible for them to secure an attorney first, or to assemble evidence to prove the merits of their case (AILA 2017: 1). If the asylum officer grants a positive decision on the fear interview, the applicant’s case is finally referred to the immigration court for a series of hearings before an immigration judge, culminating in a final merits hearing. Unlike in the criminal justice system, the US government does not provide a public defender for people in deportation proceedings. Those who cannot afford an attorney find themselves at a major disadvantage.14 Having an attorney makes a respondent five and a half times more likely to win their case (Eagly & Shafer 2015: 2).  As punitive immigration and criminal justice policies have expanded in lockstep in recent decades, certain violations of immigration law make noncitizens vulnerable to criminal prosecution, while certain criminal offenses trigger deportation. In light of this punitive drift, demands are growing for the US government to establish a public defender system for deportation proceedings (Eagly 2013; AILA 2016). One legal advocate who I interviewed, who is experienced in the family detention context, made this argument. The advocate also highlighted both the urgency and challenge of trying to achieve this advocacy goal in a piecemeal fashion (i.e. first establishing a public defender system for the most vulnerable populations), given that the benefit may be slow to trickle down to the rest: So the argument is that we start with the children, and I think it’s good, right, because they’re vulnerable in extra ways. But at the same time, one of the few circumstances where you’re really going to find me pulling for adult men, but they’re straight-up treated like an already problematic criminal characterization, which I don’t agree with. Adult men come here seeking asylum and they’re treated as if they killed somebody. And the government is really doing everything essentially in their power to make sure they get deported.                                                   14 It was only in the past few years that litigation finally obliged EOIR to appoint legal counsel free of cost to certain people with mental disabilities, and to certain children, but only within particular regions of the country (ACLU 2013; Franco-Gonzalez v. Holder 2013; NWIRP 2016; F.L.B. v. Lynch 2016). 135  This advocate’s point was that a public defender system is needed for all low-income people, including the most easily-criminalized like the adult male demographic, to ensure due process. The Florence Project also shares this view that a public defender system is needed for all low-income people, and advocates for the creation of such a system.   In the meantime, in the void of government services, any low-cost or free legal services tend to come from one of three domains: nonprofits, pro bono private attorneys, and law school clinics (Eagly 2013). One way that nonprofits try to reach as many people as possible is through the Legal Orientation Program (LOP) (Vera Institute of Justice 2017). Through its work in the 1990s and early 2000s to establish a “rights presentation” that would enable detained people to more meaningfully access their rights and participate in the legal process, the Florence Project created the blueprint for the LOP model used in detention centers across the country today (Siulc, et al 2008: 7-8). The LOP is available to any detained person who wants to participate, making this a daily ritual for nonprofits like Florence Project. The program is composed of pro se workshops, meaning orientations that are designed to help detained people represent themselves. Nonprofits also refer cases that appear to have merit (meaning they are likely to succeed) to private attorneys who are willing to take a case pro bono (Eagly 2013: 2291). This is true of the Florence Project, which provides in-depth mentorship to pro bono attorneys, as well as in-house representation whenever possible.  Back in May 2016 on my first day in the Eloy Detention Center, I sat on the pastel-colored bench of the visitation room, watching Florence Project staff members present animated, easy-to-follow LOP group orientations on complex topics. They held up big poster-boards with graphics and flow charts as visual aids, while they gave an overview of the common defenses against deportation, and explained how to seek release (e.g. through bond or parole). The 136  detainee-participants asked astute questions, participated, and joked around at times. The mood was light, despite the claustrophobic surroundings bearing down on us. The participants who requested an individual pro se orientation to learn more about their own potential defenses against deportation stuck around the visitation room. After several hours of these individual orientations, I stumbled out into the piercing mid-day Arizona sun, overwhelmed by the outsized need of detainees who cannot afford a private attorney.  5.2.3 Courtroom performance Asylum seekers who persevere through all the hurdles mentioned above will eventually get their final merits hearing before an immigration judge – an adversarial proceeding in which they are pitted against a DHS trial attorney, and finally give their testimony (USCIRF 2016: 54). What do asylum seekers need to prove when they finally get their day in court?15 They must explain how their experience fits the legal definition of “refugee” ([US INA] Act 101[a]15P):  Any person who is outside any country of such person's nationality or, in the case of a person having no nationality, is outside any country in which such person last habitually resided, and who is unable or unwilling to return to, and is unable or unwilling to avail himself or herself of the protection of, that country because of persecution or a well-founded fear of persecution on account of race, religion, nationality, membership in a particular social group, or political opinion.  The asylum seeker must demonstrate that the severity of the past or future harm meets the benchmark of “persecution” – for instance, threats, beatings, kidnappings, or murder. They also must show that the direct source of the threat is either a government actor, or that that the government will not protect them from the harm they face. Further, they must demonstrate that                                                  15 Asylum is one of three types of humanitarian relief from deportation available to certain noncitizens in the US who are afraid to return to their country (the other two being Withholding of Removal, and Withholding/Deferral of Removal under the Convention Against Torture [CAT]) (FIRRP 2013). In this dissertation, I focus on asylum for the sake of simplicity, although there is much to be said about who is excluded from applying for asylum. 137  they cannot simply move and be safe in a different part of their country. Crucially, the asylum seeker must also show that their persecutor has targeted them on account of at least one of the five protected grounds: their race, religion, nationality, membership in a particular social group, or political opinion (FIIRP 2013: 3-5).  This “on account of” requirement, known as the nexus, demands a theory of the persecutor’s motive. Why did this person single you out for harm, and not someone else? The asylum seeker must connect the harm they fear to at least one of the five protected grounds – a complicated legal maneuver that is difficult even for experienced attorneys and adjudicators, given that “the statute remains open to interpretation and the case law is constantly developing” (Sheppard & Murray 2017: 17). One of the most ambiguous, and embattled aspects of the nexus is the protected ground “membership in a particular social group.” What counts as a social group for protection, and who is a member? In 1985, the BIA defined the “particular social group” as a group of individuals with “a common, immutable characteristic, i.e., a characteristic that either is beyond the power of the individual members of the group to change or is so fundamental to their identities or consciences that it ought not be required to be changed” (Matter of Acosta 1985: np). In 2008, the BIA tacked on two new requirements: that the social group also be (1) “socially visible;” and (2) defined with “particularity” (NIJC 2016). These complex and ambiguous requirements were precisely what I needed to help explain when I accompanied a Florence Project staff attorney to deliver a Spanish-language pro se orientation about asylum and related forms of humanitarian protection. That day, we headed to the Florence Correctional Center. A group of about 35 men, dressed in green and tan prison scrubs, sat in rows of plastic chairs in a room that could almost be mistaken for a public-school 138  classroom – with its florescent lights, white board, and shiny tile flooring – save for the fact that we were locked inside a secure compound.  Most orientation participants were from Mexico, El Salvador, Guatemala, and Honduras. The attorney and I presented on a range of topics – from basic legal requirements, to the court and appeals processes, to strategies for articulating the nexus. We explained that testimony is key, as it is often the primary source of evidence in an asylum case. An applicant can build credibility by presenting their story in a chronological order, with relevant details (e.g. who harmed you? When, and how many times? Where did this happen? How did they harm you? Why did they target you?). We explained that a testimony should be as focused and vivid as possible, alive with details.  A concern arose from one participant about the difficulty of telling his story because it makes him extremely sad. This launched us into a brief discussion about emotion, in which we noted that it is okay to openly express how you feel, including to cry. A few uncomfortable laughs erupted across the room at the idea of crying in the courtroom. But some participants spoke up in defense of tears, suggesting that expressing your feelings openly might help the judge understand the gravity of your fear.  The legal, emotional, and communication nuances about presenting one’s story in court that we had discussed came to life when we ended the orientation with a scripted role-play of the final merits hearing. We placed three plastic chairs in the front of the room for the key courtroom characters. I threw on a black robe to play the judge, while one participant played the respondent, and another played the DHS trial attorney. Following our script, the respondent fumbled some of his testimony, for example, explaining his story in an unchronological order, being imprecise 139  about how many times he was harmed in his country, and straying from the question that I, as the judge, asked him.  In our debrief after the role-play, the participants caught these issues. They suggested ways for the respondent to be more precise, and to listen carefully to the judge’s questions. To me, the role-play felt like a hopeful moment, brimming with the possibility that a careful and truthful performance of one’s story could resonate with the judge. This glimmer of hope stands in contrast with the bleak numbers, which project that many of the orientation participants were unlikely to prevail in their cases. Between FY 2011 and 2016, US adjudicators denied 49 percent of all asylum claims. The denial rate for certain nationalities were much higher than the overall average: 89.6% of claims from Mexico were denied, as well as 82.9% from El Salvador, 77.2% from Guatemala, and 80.3% from Honduras (TRAC 2016, December 13). For Central American asylum seekers who do dodge all the hurdles posed by detention and fast-track deportation to get their day in court, why are so many cases denied?   5.3 The “bad scripts” of the asylum system Feminist and postcolonial scholars have observed that there is a performative dimension to winning recognition as a legitimate refugee (e.g. Bhabha 1996; Malkki 1996; Oxford 2005; Sinha 2001). Asylum seekers must learn an opaque legal lexicon, defined by Western-centric notions of persecution and fear, that demands a concise, chronological explanation of a disruptive life experience (Ordóñez 2008; Schuman & Bohmer 2004). There are limited scripts that yield success: “Asylum seekers need to present themselves as truthful and quintessential victims, either with little agency in the activities that led to their displacement, or hyperagency (in the case of political and other activists)” (Ordóñez 2008: 43).  140  By drawing on this work to conceptualize the asylum process as performative, I am not implying that people present their stories in an untruthful way. As Diana Taylor (1997: 184) suggests, performance is not the opposite of reality. Taylor writes about the Mothers of the Plaza de Mayo, an activist group that developed to protest the disappearance of their loved ones by the Argentinian military dictatorship. Taylor sees the mothers’ expression of grief in public space as performative, in that they intentionally sought to invoke public scrutiny of the dictatorship’s political violence. The mothers were deliberate in how they presented themselves, but this does not make their pain any less real, or their actions untruthful. If the mothers were performers, so too was the dictatorship, which crafted its own script to justify its political violence. As Taylor writes, the mothers were “trapped in bad scripts.”  Following Taylor (1997), I see Central American asylum seekers as being “trapped in bad scripts.” I call these scripts bad, not only because United States immigration law leaves limited avenues for a person to prove their claim, but also because a number of other factors conspire to exclude Central American asylum seekers, as I have explained thus far. Adjudicator decisions are limited by existing legal precedent (which is unfavorable to many Central American applicants, as I will explain in Section 4), while reflecting broader cultural sensibilities, gendered and racialized exclusions, and geopolitical imperatives. Cold War geopolitics ensured that the most successful asylum claimant of that era was an elite man fleeing a communist regime that he actively denounced (Sinha 2001: 1576). In this sense, asylum adjudication developed along “bad scripts” by centering a universal male norm, and by serving as a theater of geopolitics. Despite the 1980 US Refugee Act’s attempt eliminate gender and regional biases by adopting the UNHCR definition of “refugee” (Kennedy 1981), 141  decisions to grant asylum are inevitably shaped by the granting state’s understandings of fear, persecution, and failure of the home country to protect its citizens (Coutin 2007: 157).  The gendering of asylum scripts is likely also rooted, in part, in scholarly migration ‘expertise’ of the Global North, which until the 1980s, largely conceptualized women’s decisions to migrate as ‘personal’ (not ‘political’ or ‘economic’) – focusing on emotional and familial factors, such as women’s efforts to escape gender discrimination and domestic violence, to flee marriages, and to seek greater property rights (Morokvasic 1984: 898). Feminist studies have since demonstrated that women’s ‘personal’ issues are at once structural, economic, and global, yielding “narratives of constrained agency, complexity, and contradiction” (Pratt & Rosner 2012: 16; also see Mohanty 2003; Spivak 2010; Faier 2009). Humanitarian narratives of displacement remain comparatively blunt, continuing to frame women as passive victims. For example, women who hire guides to cross militarized borders are interpreted as trafficking victims, while men who do the same are seen to be soliciting the services of smugglers (Van Liempt 2011). Unaccompanied child migrants are likewise figured as passive in their movement and contrasted with the active decisions of adults to move (Martin 2011). There is a pragmatic reason to be blunt. As mentioned, harms commonly directed at women, such as domestic violence, were long rejected by US adjudicators as “personal” or “cultural,” and therefore not fitting the standard of “political” persecution (Calavita 2006: 111). Three decades of legal advocacy and precedent-setting cases have paved a path, although never a certain one, towards adjudicators better recognizing women’s gender-based asylum claims (Musalo 2010; 2014/15).  Drawing from postcolonial feminist theory, socio-legal scholars have argued that women’s gender-based asylum grants provide crucial protections for the individual, but may 142  reinforce social and spatial differences – between women and men, deserving and undeserving victims, and Global North and South (see Bhabha 1996; Luibhéid 2002; Malkki 1996; Oxford 2005; Razack 1998; Sinha 2001). For instance, Oxford (2005) finds that an “exotic” harm like female genital cutting makes for a compelling claim, even it is not the central reason a woman fled her country. This can “create a new gendered victim based on a cultural act the asylum seeker may not consider to be persecution” (ibid: 29).  I see this as a “bad script” not only because it reinforces oppressive gender norms, but because it reproduces negative, racialized stereotypes about the places that people left behind, while framing the United States as somehow above the fray. I believe there is also much to be learned about how exclusions and criminalization function by looking specifically at the “bad scripts” that Central American men navigate, in relation to those faced by women. To do so, I analyze three common types of Central American asylum claims – those based on gang opposition, domestic violence, and family membership.  5.4 Asylum adjudication depoliticizes conflict in Central America In this section, I examine how adjudicators either recognize, or deny certain types of common Central American asylum claims. This denial is not necessarily straightforward. For example, an immigration judge may grant asylum, but then a DHS attorney appeals that decision to the Board of Immigration Appeals (BIA), which may or may not agree with the judge. Likewise, if an immigration judge denies a claim and orders deportation, the respondent can appeal. The appeals process can reach the federal circuit courts, or even the Supreme Court. Thus, the decisions analyzed here do not speak for a single voice of adjudicators. Rather, they arise through a messy, lengthy, and contested legal process. 143  5.4.1 The denial of gang recruitment claims Adjudicators tend to deny Central American asylum claims based on opposition to gang activities, such as resistance to recruitment and refusal to pay extortion (Zedginidze 2016: 236; Harris & Weibel 2010; Schulman 2016). Below, I focus on resistance to gang recruitment cases, highlighting three narratives that run through these denials, that adjudicators tend to: (1) not recognize young men or women who resist recruitment as a particular social group; (2) not recognize neutrality in conflict as an active expression of political opinion; and (3) construe gang-related violence as apolitical. These narratives did not appear out of thin air. They have deep roots in Cold War-era asylum adjudication (see Coutin 2011), as I briefly outline here. In 1986, the Ninth Circuit Court of Appeals decided that “young, urban working class males of military age who have never served in the military or otherwise expressed support for the government” was too broad to be a cohesive social group for asylum (Sanchez-Trujillo v. INS 1986). The two Salvadoran respondents in the case testified to surviving personal attacks, accusations of rebel group membership, as well as public events in which Salvadoran state security forces violently attacked all participants (ibid: 11-12), yet the court declared the following about the war in El Salvador (ibid: 7):   Although a substantial number of the victims have been young males – which would hardly be surprising in any violent conflict – the [immigration judge] and the BIA reasonably found that the evidence was inconclusive to establish that mere age and gender, even when combined with labor class background, urban residence, or political neutrality, had any bearing on the likelihood of persecution.  In one breath, the Ninth Circuit acknowledged that young men were a key target of Cold War-era state violence in El Salvador, but denied that their identity mattered. Rather, the court asserted that the people at risk were those who most actively and openly opposed the government, finding that the two respondents in this case did not meet that standard (Sanchez-Trujillo v. INS 1986: 7). 144  By denying the respondents’ claim that they were persecuted because of their identity as young, working class men trying to stay out of the conflict in El Salvador, the Ninth Circuit universalized this demographic – implying that their suffering was common and unfortunate, but not legible for asylum. It also made a judgement about what it means to be politically-active.  In its 1992 decision on INS v. Elias-Zacarias, the Supreme Court also weighed-in on the meaning of political activity in Cold War Central America. The respondent, Elias-Zacarias, was a Guatemalan man who refused to join the guerillas. The Supreme Court decided that: Even a person who supports a guerilla movement might resist recruitment for a variety of reasons – fear of combat, a desire to remain with one’s family and friends, a desire to earn a better living in civilian life, to mention only a few. The record in the present case not only failed to show a political motive on Elias-Zacarias’ part; it showed the opposite. He testified that he refused to join the guerillas because he was afraid that the government would retaliate against him and his family if he did so (INS v. Elias-Zacarias 1992: 482).  Although the Supreme Court saw the guerillas’ motives as political (in that they sought recruits to fight against the government of Guatemala), it construed Elias-Zacarias’ refusal to join as apolitical. The court’s logic was that he acted out of fear, rather than his political convictions. Yet the court was not united in this opinion. Three Supreme Court justices dissented (INS v. Elias-Zacarias 1992: 486), arguing: A political opinion can be expressed negatively as well as affirmatively.…Even if the refusal is motivated by nothing more than a simple desire to continue living an ordinary life with one’s family, it is the kind of political expression that the asylum provisions of the statute were intended to protect.  Despite this dissenting opinion, the Supreme Court’s denial ultimately set forth a limited view of political action. Cold War-era adjudication like INS v. Elias-Zacarias (1992) and Sanchez-Trujillo v. INS (1986) helped set into motion the bad asylum scripts that persist for Central American asylum seekers today.  145 Adjudicators deny claims based on gender, youth, and resistance Central American asylum seekers face an uphill battle to convince adjudicators that young men and women who oppose gang activity like recruitment could form part of a particular social group legible for protection – particularly since 2008, when the BIA demanded that the social group be “socially visible” and defined with “particularity.” The BIA introduced these requirements through its precedential decisions on two Central American gang-opposition claims: Matter of S-E-G- and Matter of E-A-G-.  Matter of S-E-G- is the case of three siblings from El Salvador, who argued their claim to asylum around their anti-gang political opinion, and their membership in a social group of “Salvadoran youth who have been subjected to recruitment efforts by the MS-13 gang and who have rejected or resisted membership in the gang based on their personal moral, and religious opposition to the gang’s values and activities, and their family members” (Matter of S-E-G- 2008: 579). While the BIA acknowledged that MS-13 does retaliate against people who impede its activities, like the two brothers in this case, it holds that such risks can befall anyone. Further, the BIA finds “no evidence in the record that the respondents were politically active or made any anti-gang political statements” (ibid: 589). In this precedent-setting decision, the BIA universalized the experience of young people who resist recruitment. It found their proposed social group to be too broad and amorphous, not recognized by Salvadoran society. Nor was their resistance active enough to constitute a political opinion.  The case that the BIA released as a companion in 2008 fared no better. The respondent, “E-A-G-,” was a young Honduran man who feared two rival gangs that had murdered his brothers. The BIA rejected E-A-G-’s social group of “persons resistant to gang membership” for lacking social visibility. It noted that a gang may target someone due to their resistance to 146  recruitment, “But such a risk would arise from the individualized reaction of the gang to the specific behavior of the prospective recruit,” rather than in a systematic way (Matter of E-A-G- 2008: 594). The BIA also rejected E-A-G-’s proposed social group of “young persons who are perceived to be affiliated with gangs,” finding that asylum cannot be granted due to criminal affiliation, even if it is simply a perceived and mistaken affiliation (ibid: 596). Finally, the BIA argued that gang members would harm him due to their rivalry, and to expand their “power and influence,” not in response to his anti-gang political opinion (ibid: 597). The BIA’s decision thus pathologized and individualized E-A-G-’s experience, painting him at once as politically-inactive and criminally-suspect, and the gang members as greedy and irrational. Since 2008, adjudicators often use the “social visibility” and “particularity” requirements to deny gang opposition claims, at times simply by citing Matter of S-E-G-, without fully reviewing the facts of the given case. This has left legal advocates puzzling over what, if any, formulation of a social group would work for these claims (Frydman & Desai 2012: 7).   This negative fallout is evident in the Ninth Circuit’s 2009 decision on Ramos-Lopez v. Holder, in which the court argued, “The threats and harassment Ramos and other young men experience after being recruited is a sad part of the ‘general criminality and civil unrest’ perpetrated and perpetuated by the MS-13 in Central America” (ibid: np). This view posits young men as a universal subject, foreclosing the possibility that they are targeted by gangs in a strategic way due to their age and gender. It also gives the impression that each act of violence is inconsequential, simply because such acts are common.  In their universalization of young men, these types of denials obscure the many ways this demographic is criminalized. For instance, the Ninth Circuit argued that, “Ramos-Lopez attests to indiscriminate action taken by the police against children who are not in gangs in their efforts 147  to curb gang activity. This fact evinces a lack of social visibility” (Ramos-Lopez v. Holder 2009: np). Yet one could easily make the opposite argument – if this demographic is targeted by both gangs and the police, perhaps society does recognize it as a cohesive, visible social group. The Guatemalan, Honduran, and Salvadoran governments all adopted anti-gang laws and mano dura [iron fist] policing strategies in the early 2000s that targeted young, working-class people quite indiscriminately. Further, the targeted killing of suspected gang members by paramilitary groups across the region is well-documented (Paz 2014: 1106). Nonetheless, adjudicators continue to deny gang-opposition social groups for lacking “particularity” and “social visibility.” Neutrality in conflict is not recognized as a political opinion The Supreme Court’s 1992 finding in INS v. Elias-Zacarias, that neutrality does not count as an expression of political opinion, continues to echo through the adjudication of Central American gang-related cases, such as the Ninth Circuit’s decision in Santos Lemus v. Mukasey (2008). The respondent, Jose Nelson Santos-Lemus, fled El Salvador in 2004 after members of MS-13 threatened him and his family. When one of Santos-Lemus’ brothers confronted the gang members, they murdered him. The police did not investigate the crime, while the family did not report it because they feared retaliation, and believed the police and gang to be working together (ibid: np).  Santos-Lemus argued that he was persecuted on account of membership in his family, and in “the class of young men in El Salvador who resist the violence and intimidation of gang rule,” as well as his political opinion against gang activity (Santos Lemus v. Mukasey 2008: np). The Ninth Circuit asserted, “Nothing in the record establishes that he was a well-known anti-gang activist or even outspoken about gangs. On the contrary, he testified that he never spoke out against or insulted gang members” (ibid: np). The court found that the gang members targeted 148  Santos-Lemus for “economic and personal reasons” rather than due to his political opinion (ibid: np). Finally, the court brushed off his testimony about police-gang links as speculative, suggesting “it is undisputed that any torture Santos-Lemus fears would be committed by private individuals, not the government” (ibid: np, my emphasis).  In denying Santos-Lemus’ claim, the Ninth Circuit construed his neutrality as apolitical and inactive, the gang members involved as private individuals, and their motives as purely personal or economic. This logic is strikingly similar to arguments long used to dismiss women’s gender-based violence claims, for example, that domestic violence is a private harm committed by a private actor, and therefore does not count as political persecution (I will elaborate and provide examples in Section 5.4.2). My point here is that such depoliticizing narratives are also common in the adjudication of gang-opposition cases, even in those where the applicant has made their anti-gang stance crystal clear, such as in Rivera-Barrientos (Tenth Circuit, 2012). Minta del Carmen Rivera-Barrientos, a young woman from El Salvador, gave the following testimony: “In August 2005, members of MS-13 approached [her] and asked her to join the gang. She refused, stating ‘No, I don’t want to have anything to do with gangs. I do not believe in what you do’” (Rivera Barrientos 2012: 3). They responded “[i]f you don’t want to join with us, if you don’t participate with us, if you are against us, your family will pay” (ibid: 3). After months of harassment, five gang members abducted, violently attacked, and sexually-assaulted Rivera-Barrientos (ibid: 3). The BIA argued that the gang members reacted violently to her refusal to join the gang, rather than her anti-gang political opinion (ibid: 8). The Tenth Circuit agreed, and drew a parallel to the Supreme Court’s 1992 findings in INS v. Elias-Zacarias by asserting, “So, too, must Rivera Barrientos show that her attack was motivated by more than anger at her unwillingness to join MS-13 and a desire to coerce her into joining” (ibid: 149  9). Despite Rivera-Barrientos’ clear assertions of her anti-gang views, adjudicators construed her refusal to join the gang as something other than a political opinion, and provided a one-dimensional portrayal of the gang members’ motives. Gang-related violence is not recognized as political The adjudication I have mentioned thus far gives the sense that gangs do not target people because of their traits (like youth and gender), nor because of their resistance to a gang’s authority. So what does motivate this abuse? Adjudicators tend to argue that gangs target people indiscriminately, simply to bolster their “ranks, wealth or power” (Frydman & Desai 2012: 15; e.g. see Larios v. Holder 2010). They paint a picture of Central America as a place of “civil strife” where people may be subject to “indiscriminate violence” (Bonilla-Morales 2010; Herrera Flores v. Mukasey 2008). In this way, gang members are cast as one-dimensional characters in asylum proceedings – as private actors and pure villains, motivated solely by greed and anger. This depoliticizes violence in Central America. The outcome on Benjamin Rodas-Orellana’s (Tenth Circuit 2015) case provides an example.  Rodas-Orellana argued that he was persecuted as part of the social group of “El Salvadoran males threatened and actively recruited by gangs, who resist joining because they oppose the gangs” (Rodas-Orellana v. Holder 2015). He testified that at age 17, he fled “extreme poverty and gang violence” in El Salvador (ibid: 5). During an encounter with MS-13 members, they “bloodied his face” upon his refusal to join. Rodas-Orellana explained that they assaulted him “because they asked me for money and I didn’t have any” (ibid: 6). The Tenth Circuit argued, “His testimony suggests only that the gang wanted to take his money or have him join the gang. When he refused, the gang reacted not to his membership in a particular group but to his refusal to pay or join” (ibid: 23).  150  This analysis somehow disconnects Rodas-Orellana’s refusal to join or pay the gang from his claim that he was targeted for opposing them, running counter to the evidence that people who subvert the authority of the gangs in El Salvador can face severe reprisals, including death (UNHCR 2016: 28-30). The idea that gang-related conflict is fueled solely by greed, or that it is indiscriminate, is depoliticizing – a dynamic I will explore further in Chapter 6 of this dissertation. I next turn to the hard-won, but limited recognition of women’s domestic violence claims, which have long struggled against the same depoliticizing narratives that continue to afflict gang-opposition claims. 5.4.2 The struggle for recognition of women’s gender-based violence claims A legal battle has brewed for decades over the viability of women’s domestic violence asylum cases – with Central American and Mexican women’s claims at the eye of the storm. A major victory came in 2014, when the BIA issued a landmark decision on Matter of A-R-C-G, finding that “Depending on the facts and evidence in an individual case, ‘married women in Guatemala who are unable to leave their relationship’ can constitute a cognizable particular social group” (Matter of A-R-C-G- 2014: 388).  The lead respondent in that case, Aminta Cifuentes, survived more than ten years of severe abuse from her husband in Guatemala (Bookey 2016). When Cifuentes sought help from the police, they declined to get involved with the ‘marital relationship.’ Her husband threatened to kill her should she contact the police again. She fled multiple times, going to Guatemala City, and to her father’s home. Yet her husband always tracked her down (Matter of A-R-C-G 2014: 389). An immigration judge initially “found that the respondent’s abuse was the result of ‘criminal acts, not persecution,’ which were perpetrated ‘arbitrarily’ and ‘without reason’” (ibid: 389-390), but Cifuentes eventually won her case on appeal. 151  As legal scholar Karen Musalo (2010) outlines, women’s gender-based claims are one of the most embattled areas of US asylum adjudication – long rejected as personal or cultural matters, particularly given that the persecutor is often a non-state actor (e.g. a male partner committing domestic violence). The earliest women’s gender-based violence cases to prevail, in the mid-1990s, were those that centered non-Western cultural practices, such as female genital cutting (Sinha 2001). It would take longer for US adjudicators to begin to recognize more culturally-familiar forms of abuse against women, like domestic violence, as a legible basis for asylum (Oxford 2005).  In 1999, the BIA weighed-in on the viability of domestic violence cases when it rejected the claim of Rody Alvarado-Peña (Matter of R-A- 1999; Musalo 2014/15: 46-47). Alvarado-Peña sought asylum as a member of the particular social group of “Guatemalan women who have been involved intimately with Guatemalan male companions, who believe that women are to live under male domination” (cf Matter of A-R-C-G- 2014: 391). The BIA argued that Alvarado-Peña crafted this social group solely to win her claim, and “that it was unclear that ‘anyone in Guatemala perceives this group to exist in any form whatsoever,’ whether spousal abuse victims themselves or their male oppressors” (cf ibid: 391). Alvarado-Peña’s case remained unresolved for years – bouncing back and forth between a series of attorney generals, the BIA, and finally an immigration judge who granted her asylum in 2009. In the decade in between, the BIA’s 1999 rejection had a negative fallout among women’s gender-based claims (Musalo 2010).  In 2004, DHS took the unusual step of defending Alvarado-Peña’s case (US DHS 2004). This is unusual because the agency’s role in deportation proceedings is to argue in favor of deportation. The agency wrote a brief clarifying how women like Alvarado-Peña could better articulate their domestic violence claims: “Certainly Alvarado’s status as a wife is intricately 152  linked with her gender, but to define the social group broadly as ‘women’ ignores the evidence that a primary animus for violence arises from the abuser’s perception of the subordinate status his wife occupies in the domestic relationship” (ibid: 27). In other words, DHS asserted that this subordinate domestic position should be the crux of the claim, rather than gender identity. In its 2004 brief, DHS also argued that Rody Alvarado-Peña’s political opinion was not relevant to her asylum case, “because there was no evidence that the husband was aware of the applicant’s opposition to male dominance, or even that he cared what her opinions were. Rather, he continued to abuse her regardless of what she said or did” (ibid: 13; also see US DHS 2009: 22). Writing in 2001, legal scholar Anita Sinha observed that the US government was willing to accept women’s domestic violence claims through the social group nexus, but tended to dismiss women’s articulation of a political opinion. Sinha (2001: 1593) suggested, “Omitting analysis of ‘political opinion’ claims reflects a judgement on the part of the INS that domestic violence is not a political matter.” This obscures the possibility that surviving and escaping abuse, and seeking asylum, is “a form of political resistance” (ibid: 1594). This limited reading of political action remains in place, despite the hard-won progress that has been made for domestic violence claims under the protected ground of particular social group. The BIA’s 2008 introduction of the “social visibility” requirement for particular social group claims has clearly had a negative fallout for gang-related cases, as described in the previous section. It also had the potential to threaten the fragile viability of women’s gender-based claims. At stake, as legal scholars Lindsay M. Harris and Morgan M. Weibel (2010: 24) explain, is that “women who oppose and suffer domestic violence are certainly no more ‘visible’ than youth opposing gang persecution.” Further, “Social visibility ‘raises the specter of the private/public distinction by requiring members of a particular social group to have a public 153  face’” (ibid: 24, citing Marouf 2008: 94-99). Despite this threat, DHS has proven willing to accept domestic violence claims as meeting the “social visibility” requirement, even if such abuse may not be literally visible to society (Musalo 2010: 61-62; US DHS 2009: 14). Why hasn’t the “social visibility” requirement impeded women’s gender-based violence claims in the same way it has blocked gang-opposition claims? For Harris and Weibel (2010: 25), this discrepancy indicates that “social visibility is a tool adjudicators employ to avoid granting asylum in politically unpopular cases.” Adjudicators and the INS/now-DHS have faced political pressure since at least the mid-1990s to recognize women’s gender-based violence claims as valid, in order to remedy a historical gender bias in the asylum system. The same political urgency does not exist for gang-opposition cases.   During an interview, an Arizona attorney experienced in women’s gender-based violence claims explained to me that the growing recognition of domestic violence claims has created an important opening for Central American asylum seekers. This is especially true for cases that closely mirror the facts of the BIA’s landmark 2014 decision on Matter of A-R-C-G-. Yet speaking to the grim odds for asylum seekers in the Arizona detention centers like Eloy, this attorney cautioned that success on domestic violence claims is still far from guaranteed: I don’t think any claims are strong in Eloy, and I think that any claim where you don’t have directly controlling BIA precedent, it’s just going to be harder for the Eloy judges. I mean they really, really don’t – they basically come into the case not wanting to grant asylum – so you have to make it almost, I think some of it is a calculus about appeal. And so if you can show that you’re developing a really strong record, and I mean honestly I think it helps when there’s pro bono counsel that they actually really follow through on an appeal. Then I think there’s a little bit more hope for a grant.  The fact that adjudicators are now more inclined than in the past to approve women’s gender-based violence asylum claims is a huge change. It comes out of a long struggle for recognition of the common harms against women as political, yet significant barriers remain for domestic 154  violence claims (Bookey 2016). Women’s resistance to domestic violence is still not recognized as political, and success is limited to claims that follow the narrow path of acceptable social groups set forth by legal precedent. Finally, this progress comes within a broader climate of exclusion towards Central American claims.  5.4.3 The narrow recognition of family claims Adjudicators have long treated “kinship ties” as a “prototypical example” of an immutable characteristic that defines a particular social group for asylum (Matter of Acosta 1985; Sanchez-Trujillo v. INS 1986). As the INS explained in 1995, “claims based on family membership are frequently asserted by female applicants, particularly in countries where men tend to be more active politically than women” (US INS 1995). Family-based claims have thus historically offered a pathway to asylum for women, although they are not exclusively submitted by women. Legal precedent is relatively favorable for family-based claims (Quinn, Lee, & Boerman 2015), yet just like for other common types of Central American claims, success is never guaranteed.  As in gang opposition and domestic violence cases, the BIA’s 2008 “social visibility” and “particularity” requirements pose a threat to family-based claims, given that they are crafted around the particular social group nexus (in which the family is the social group). For instance, in the immediate aftermath of 2008, DHS and the BIA used the “social visibility” requirement to demand that a family be well-known within the society of origin in order to qualify as a “socially-visible” social group (Marouf 2008: 93-94).     Some other hurdles that family claims can face arose in the recent case of Luz Marina Cantillano Cruz, a Honduran woman whose claim hinged on her membership in the “nuclear family of Johnny Martinez,” her husband. Martinez was employed as a private bodyguard for a man named Avila, who claimed to be a “fisherman,” but Martinez learned was actually part of an 155  organized criminal group that trafficked drugs and arms. Just when he was on the verge of quitting, Martinez disappeared on a “fishing trip.” Cantillano-Cruz and Martinez’s uncle asked Avila where he was, and then began receiving threats. Avila and his associates loitered outside the family home, fired weapons, killed Cantillano-Cruz’s dogs, and threatened to kill her and her children (Cantillano Cruz v. Sessions 2017: 3-4).  After Cantillano-Cruz fled to the US, an immigration judge found that the “main reason” the employer threatened her was to prevent her from filing a police report. The judge reasoned that “‘although family ties likely motivated’ her search for Martinez and her decision to confront Avila, ‘that concern for [Martinez’s] well-being could exist outside their familial relationship’” (Cantillano Cruz v. Sessions 2017: 5). The BIA agreed. Further, “According to the BIA, Avila’s threats constituted ‘[h]arm meted out by a private actor for personal reasons or solely on general levels of crime and violence in Honduras’” (ibid: 7)  Although the Fourth Circuit ultimately granted Cantillano-Cruz’s petition to review her case, the initial hurdles she faced are illuminating. The immigration judge and BIA used gendered language, calling the persecutor a “private actor” motivated by “personal reasons,” and subsuming the harm experienced by the family within some vague general criminality in Honduras (also see Gomez-Romero v. Holder 2012). In this way, adjudicators can deny family-based claims in some of the same ways they frequently deny other varieties of Central American claims.  Even when adjudicators do grant family-based claims, the parameters can be quite narrow. The person most directly in the line of fire may not meet the requirements for asylum (e.g. see Crespin-Valladares v. Holder 2011; Zelaya v. Holder 2012). One attorney interviewee 156  mentioned this dynamic. Referring to asylum cases that center on gang extortion of bus drivers in El Salvador, the attorney explained that: [In] these bus cases, the primary person we’ll say is the dad, like in my case. They kill him and then they also go after the whole family. But had they not killed him, had the whole family run away, you have a weird situation which would be that the dad has not a very good asylum case because they're trying to target his money, or they want his business; they want him to pay the rent.   But the family members of the dad, who are not the main target, have a much better case, because they're being targeted because they're a particular social group called the “immediate family of the bus driver.” So you have a weird situation where they primary target can't get protection, but the secondary targets have a stronger case because they fit into a particular social group.  For me, this strange dynamic sums-up the overall state of United States asylum adjudication towards Central American asylum seekers: a climate of hostility, in which small glimmers of possibility exist for very narrow categories of people.   5.5 Conclusion The United States asylum system has historically favored a masculine subject – the elite political dissident who boldly and actively denounced a communist regime. I find that a masculinized standard continues to create barriers to the recognition of Central American asylum seekers at large, for women and men alike. Adjudicators tend to depoliticize, feminize, and thereby dismiss the common asylum claims from this region of the world – from resistance to gang recruitment claims, to domestic violence and family-based claims. I see this denial as a slippage of gendered exclusion, because it places all claimants on the passive and irrational (rather than active and rational) side of the gender binary entrenched within Western thought (see Plumwood 1993). Central American men’s claims can be denied, for example, because adjudicators commonly perceive neither gang activity nor resistance to recruitment to be political. Similarly, adjudicators long denied women’s domestic violence claims on the basis that such harm, and actions to resist 157  it, exist outside the realm of politics. This common depoliticization is why I suggest the denial of common men’s claims represents a feminization. I believe that this slippage reflects a longstanding political imperative to prevent Central American migration at large. This political imperative is hard to miss in a bill recently introduced to the US House of Representatives: the “Asylum Reform and Border Protection Act of 2017” (US Congress 2017, January 10; 2017, July 24). This bill would further entrench a number of structural barriers to accessing the asylum system that I have discussed in this chapter. It would write into law an increased standard for the credible fear interview, while mandating the prolonged detention of asylum seekers after they pass their interview. The bill would also prevent the government from paying for an attorney for anyone in deportation proceedings, under any circumstances, including for children or people with limited mental capacity (AILA 2017: 1-2). Beyond these structural hurdles, which this bill proposes to deepen, it would cement into law the precise barriers to the recognition of Central American asylum claims that I have outlined in this chapter, by creating a “limitation on eligibility for asylum based on generalized violence” (US Congress 2017, July 24: 15). This provision lumps together, and would categorically exclude three seemingly distinct populations: people with gang affiliations, those who resist recruitment, and crime victims (ibid).  This bill very well may never be enacted into law, yet I mention it because it exemplifies the contours of the current political strategy to exclude asylum seekers, particularly those arriving from Central America. This effort is waged not only through calls to expand the US-Mexico border wall, but also by throwing down countless logistical hurdles for people who do arrive. It is further waged by upholding the depoliticizing narratives that prevent adjudicators from recognizing the most common asylum claims from El Salvador, Guatemala, and Honduras 158  – framing these claims as “generalized” or “criminal” violence, in other words, as apolitical acts of brutality that fall outside the asylum framework.  This bill also illustrates that common Central American claims, regardless of how they are gendered, are connected, and are vulnerable to these types of political attacks. This connection suggests that the most effective arguments against Central American exclusion at large would not leave anyone behind – even the people who are most easily criminalized, like those who President Trump writes-off as “bad hombres.” With this in mind, moving beyond the “bad scripts” of the asylum system means arguing for dignified treatment for all people facing displacement from Central America. As I explore next in Chapter 6, moving beyond “bad scripts” would also mean allowing for more complex stories of displacement.       159  Chapter 6: Complicating black-and-white narratives of conflict in El Salvador  In June 2016, I flew straight from Arizona to El Salvador to continue my research. I arrived steeped in a certain understanding of displacement, shaped most recently by the asylum claims of the people I met who were locked-up in detention centers. My head was still fully in those cases – in trying to unravel complex and painful stories, and to understand how these experiences might articulate within the legal requirements of asylum. As discussed in Chapter 5, being granted asylum allows the recipient to obtain the legal protection they need, but with an unfortunate side effect. Asylum narratives tend to reinforce an assumed superiority of the United States, and draw a simplistic picture of people and places of the Global South. In a recent essay, immigration attorney Jawziya Zaman (2017: np) observes that asylum seekers may have complex relationships with the places where they are from,   But black-and-white keeps immigrant families together, not shades of gray. Over time, the names of our clients’ countries become sounds that call forth a series of images unanchored from political context and history—images of gang violence, hungry children, and oppressed women. We think we know the most important thing there is to know about these places: people leave.  When I arrived in El Salvador, I brought some of this black-and-white conceptual baggage with me, in my desire to write a dissertation that advocates for Central Americans’ access to asylum. At the same time, I also wanted to counter the asylum narrative of Central America as simply a place that “people leave.” San Salvador made international news in 2015 as the “murder capital of the world” (Muggah 2016), yet there is much more to be said about the constrained choices that people make to leave El Salvador or to stay. Simply framing the current government as unable to protect its citizens also risks fueling a long-brewing, polarized political conflict between El Salvador’s two main political parties.  160  For twenty years (1989 to 2009), the far-right wing Nationalist Republican Alliance (ARENA) political party held the presidency in El Salvador. ARENA was co-founded by Roberto D’Aubuisson, a graduate of the US military training School of the Americas, who directed death squads to torture, abduct, and execute thousands of people during the armed conflict (Gutiérrez 2007). Following the party’s presidential victory in 1989, ARENA pursued aggressive neoliberal reforms. Party leadership quickly moved to privatize or partially-privatize key public services like the national bank, telecommunications, the pension system, and the ports and airport. Next, the party sought to privatize components of the healthcare and education systems, as well as water and electricity (Schuld 2003; Avalos 2016).   The Farabundo Martí National Liberation Front (FMLN) transitioned from leftist guerilla movement to political party after the war. The FMLN first won the presidency in 2009 under President Mauricio Funes, and then again in 2014 under now-President Salvador Sánchez-Cerén. Since 2009, the FMLN administrations have sought to reduce inequality in a country that has long been vastly unequal, for example, by broadening access to public services like education and healthcare. The government now funds the basic items children need to attend school and sends healthcare teams to underserved rural areas (CISPES 2013). To revitalize the country’s food sovereignty, the Salvadoran government runs a national seed bank program that buys and sells local seeds, thus supporting small-scale farmers and bucking the previous stranglehold on seeds in the country held by Monsanto (Ritterman 2014; Young 2015).  The FMLN social agenda has faced some powerful obstacles, including actions by the United States. For example, recent US aid packages have obliged the Salvadoran government to open certain public services to private investors. As a source of crucial aid and remittances, as well as a key trading partner with El Salvador, the US constrains the FMLN’s ability to run too 161  far afield of its interests (Young 2015). As the Congressional Research Service reports, US “strategic interests” in the region include promoting the “spread of free-market economic policies” (Meyer & Ribando Seelke 2015: 15-16). Elite private and government actors in El Salvador have also obstructed key pieces of the FMLN’s social agenda in recent years. FMLN policymakers and civil society leaders have condemned this maneuver as a “soft coup,” meaning a deliberate effort to destabilize the FMLN’s ability to govern as the country approaches its legislative elections in 2018, and presidential elections in 2019 (Pineda & Stoumbelis 2017).  My research goals in El Salvador were hard to reconcile. I wanted to make an argument in defense of people’s access to asylum, but without reproducing a simplistic narrative about El Salvador as an unsafe place that people leave behind, and the United States as a beacon of safety and benevolent source of help. Ultimately, the brief two months I spent in El Salvador gave me a new vantage point – one that began to flesh out the stark black-and-white of asylum narratives into shades of gray, and pressed me to imagine what more complex stories of displacement might sound like, if permitted. As Sherene Razack (1998) writes, an asylum system based on a sense of responsibility from the Global North, rather than pity would not require the applicant to vilify their country, nor present a prescribed story of victimhood. It would allow for nuanced conceptualizations of violence and well-being, beyond basic survival. In this chapter, I advocate for complex stories of conflict and displacement, populated by multi-dimensional characters and animated by enduring imperial histories. I use this analytical approach to argue against the depoliticizing narratives of today’s conflict in El Salvador that circulate from the vantage point of the United States. To make my argument, I begin by situating my analysis within coloniality/modernity theories of knowledge and power in the Latin American studies tradition. Following this tradition, I focus on the epistemological dimensions of 162  displacement in the Americas, including by making my own research design a subject of analysis in this chapter. I elaborate this point below in Section 6.1.   6.1 The politics of knowledge production In planning my PhD research, I sought to analyze both the United States domestic and foreign policy responses to Central American displacement launched in 2014. These two domains were, and remain, connected in a concrete way by the Obama administration’s efforts to amplify deterrence strategy – most notably, by detaining and deporting people who arrived in the United States, and by discouraging would-be migrants from leaving their countries in the first place (including through foreign aid initiatives). These two domains are discursively connected as well, as they both rely on a logic that depoliticizes violence and poverty and frames these issues as “endemic” to Central America (see Biden 2015).  Domestically, I planned to focus my PhD research on issues arising from detention and asylum adjudication. As for foreign policy, my plan was to travel to El Salvador to analyze the nascent outcomes of the US Strategy for Engagement in Central America. President Obama proposed this aid package in late 2014 to help fund the Alliance for Prosperity – a plan crafted by top government officials from Central America, with close involvement of the Inter-American Development Bank (Palencia 2014). Groups representing elite business interests, and US leaders like Vice President Joe Biden and General John Kelly (then-head of the US military’s Southern Command), also gave input. In contrast, civil society groups were not consulted. The Alliance for Prosperity promised to address the root causes of migration with initiatives to bolster economic development, security, and institutional capacity (Paley 2016; White House 2015).  163  Vice President Biden promoted the US Strategy for Engagement in Central America in a January 2015 op-ed in the New York Times – explaining that the Obama administration sought to nearly triple United States foreign aid to El Salvador, Guatemala, and Honduras to address the “climate of violence and poverty that has held them back” (Biden 2015: np). Biden took a stern but benevolent tone, having “made it clear to those leaders [of El Salvador, Guatemala, and Honduras] that the United States was ready to support them – provided they took ownership of the problem” (Biden 2015: np). Intriguingly, Biden promoted this aid package as a ‘new Plan Colombia’ for Central America. This comparison was striking given that Plan Colombia, a US foreign aid package to fund the war against drugs launched in 2000, has been widely denounced for spurring human rights atrocities and displacement, while failing to achieve its stated mission to interrupt drug trafficking across the Americas. Dawn Paley (2015) finds that when Vice President Biden and others remember Plan Colombia so fondly, their standard of success “has little to nothing to do with drugs, but could be measured by examining growing levels of foreign direct investment and investor security.” The Alliance for Prosperity centers migration, not drugs, but follows the model of Plan Colombia and its offshoots, the Mérida Initiative in Mexico and the Central American Regional Security Initiative (CARSI), which since 2008 have militarized public security in the name of the “drug war,” with outcomes similar to Plan Colombia (Paley 2014).   In calling for “special economic zones” and “improved labor market conditions,” the Alliance for Prosperity deployed “euphemistic language that generally refers to the deregulation of labor markets and the dismantling of workers’ rights protections” (Main 2015). Its plans for economic development centered on extractive and infrastructure projects that looked more likely to benefit large corporations than working people. As Paley (2016: np) explains, this involves 164  “tax breaks for corporate investors and new pipelines, highways, and power lines to speed resource extraction and streamline the process of import, assembly, and export at low-wage maquilas.” The name, ‘Alliance for Prosperity’, itself bears an uncanny similarity to the ‘Alliance for Progress’ launched by the John F. Kennedy administration in the 1960s – a US aid policy that promised to prevent the spread of communism by spurring economic opportunity and encouraging institutional capacity-building (see Chasteen 2006: 281).  Civil society leaders across the region raised concerns that the Alliance for Prosperity would actually undermine existing efforts to address the root causes of displacement. For example, the historical record shows that big infrastructure and extractive projects tend to displace marginalized populations, such as indigenous peoples and Afro-descendants. Advocates also feared that the plan’s public security initiatives would make people less safe (as evident in the bloodshed following the militarization of Mexico’s public security), and could undercut existing community-based violence prevention initiatives (Mesoamerica Working Group 2015). Despite such compelling objections – that as crafted, the plan would do the opposite of what it promised – the US Congress ultimately agreed to help fund the Alliance for Prosperity (Meyer 2017). 6.1.1 Revising my research plan In early 2015, El Salvador’s president, Salvador Sánchez Cerén, announced a national security plan with a comprehensive focus on the root causes of insecurity, informed principally by non-repressive tactics like violence prevention and reinsertion (Consejo Nacional de Seguridad Ciudadana y Convivencia 2015). The plan seemed to mark a major break from El Salvador’s repressive, mano dura [iron fist] public security paradigm of recent decades. For my research in San Salvador, I intended to seek interviews with violence prevention program organizers. I 165  wanted to know how the new US aid package was changing things. Was it impeding the Sánchez Cerén administration’s efforts to move away from a repressive security paradigm?  It turns out that, although the US Congress had already appropriated funds for the US Strategy for Engagement in Central America ($750 million for Fiscal Year [FY] 2016, and $655 million for FY 2017), it would not be until the beginning of 2017 that funds were disbursed (Meyer 2017: 13). Thus, when I arrived in San Salvador in June 2016, it was too soon to trace the impacts of the new aid package. Further, I quickly realized that my driving question for this part of my research did not fit the political landscape. I learned that the Sánchez Cerén administration had been paradoxically scaling up a “war” on the gangs (El Norte 2016), just as it launched its comprehensive security plan.  As these dynamics became clear to me, I revised my plan. I still requested interviews with violence prevention program organizers and other advocates who work with communities facing violence and displacement. However, I decided to make my interview questions more complimentary to the research I had done for Chapter 5 about gang-related asylum claims – asking advocates, for example, about the barriers that young people may face to maintaining neutrality in communities impacted by both gang presence and repressive police activities.  I spent June and July 2016 in El Salvador and ended up interviewing thirteen advocates whose work is either directly focused on violence prevention and displacement, or deeply affected by these issues. This included several advocates whose work focuses on economic development or environmental protection, but who had found issues of security and out-migration becoming central to the communities they serve. Interviewees included advocates affiliated with organizations that receive funding from a variety of sources, including from the 166  US Agency for International Development (USAID). I was not able to interview anyone with an agency run directly by the Salvadoran government. I refer to interviewees with pseudonyms.  This chapter is also informed by insights from a solidarity delegation from the US that I participated in that July. We had about 15 meetings with civil society leaders and government officials about the topics of displacement, migration, public security, regional border enforcement, and the Alliance for Prosperity. Though I do not refer specifically to these meetings, they inform my base understanding.16 Throughout this chapter, I also contextualize my analysis with secondary scholarship, advocacy reports, and investigative journalism. 6.1.2 Reviewing my analytical framework As described in Chapter 1, my political orientation with the Latin American solidarity movement informed the way I formulated my research design for this dissertation. This orientation left me struggling with how to portray the revival of repressive security tactics in El Salvador. Being a United States citizen and based on my academic background and sense of research ethics, I felt best-situated to analyze the US role in El Salvador, and as mentioned, was wary of reinforcing narratives of El Salvador as simply an unsafe place that people leave.  I was wary, in part, due to the historical complicity of certain approaches to Latin American studies with US imperialism (Sundberg 2005; 2015a; Wainwright 2013). Research in this vein has long treated people and places of the Global North, or more specifically the US, as the implicit, unmarked category from which to judge differences observed in Latin America. This is evident from nineteenth and early-twentieth century studies that centered race as a purported cause of social difference, to the Cold War-era rise of modernization theory within                                                  16 As noted in Chapter 1, the views expressed throughout this dissertation, and any shortcomings, are my own. 167  area studies, which proposed that people in the Global South were developmentally behind the North and needed to catch-up (Sundberg 2005: 23, citing Berger 1995; Schoultz 1998).  As Sundberg (2005: 23) points out, such United States “representations of Latin America may say more about US interests and identities than about Latin American people and society.” This insight is central to the coloniality/modernity analytical approach developed by critical scholars of Latin America, which demonstrates that the underdevelopment framing is an assertion of power. It erases the role of enduring colonial histories in contemporary poverty and conflict, while centering “expertise” from the Global North as the locus of “development” solutions (see Stetson 2012). This is a form of cultural imperialism that has been used repeatedly over time to justify military and economic interventionism (e.g. to ‘help’ Latin America by preventing the spread of communism during the Cold War) (Mignolo 2005: 97-98; also see Moraña, Dussel, & Jáuregui 2008; Sheppard, et al 2009: Chapter 4).  As Eduardo Mendieta (2008: 289) summarizes, there is also a second, inverse side to the underdevelopment framing: “a Latinamericanism of Third Worldism, or a form of First World romanticization and exoticization of the Latin American.” In planning my PhD project, I had put a lot of thought into how the underdevelopment approach could influence my research trajectory. Yet when I was confronted with a heterogeneous social movement in El Salvador, I began to ask myself: how might this inverse tendency to romanticize and exoticize social movements in Latin America impact my thinking?  A thorny issue of solidarity politics is that they tend to “lean on ‘solid’ identities and clear-cut divisions between victim and victimizer” (Nelson 1999: 41). This binary imagines a clear subaltern community with whom North American researchers can align their solidarity. It also imagines a clear source of oppression that a researcher can critique, such as harmful 168  interventions from the Global North (or ‘the West’). But, “to only condemn Western society as repressive runs the risk of muting heterogeneities within both sides and bolsters the Western versus third-world contrast that underpins the very power relations that anticolonial studies seek to destabilize” (Nelson 1999: 70, citing Gyan Prakash).  In approaching this part of my research, my goal had been to turn the US asylum narrative on its head. As illustrated in Chapter 5, the asylum narrative frames El Salvador as a “failed state,” depoliticizes the causes of displacement, and paints the US as benevolent and uninvolved. However, my initial inclination – to simply label right-wing political actors in El Salvador and their US allies as the principal source of repressive public security strategies, and El Salvador’s recent social reforms as simply a source of liberatory politics – may reinforce an unhelpful binary. This is a decidedly right/left binary rooted in the Cold War, in which right-wing regimes across Latin America, aligned with elite classes and supported by the US, engaged in extreme political violence against people advocating for greater social equality. Of course, the Cold War was also defined by militant uprisings associated with the political left, such as the FMLN in El Salvador’s armed conflict. Yet a UN Truth Commission found that the (right-wing) Salvadoran armed forces and death squads were responsible for 85 percent of the wartime violence, while only 5 percent was attributed to the FMLN (Americas Watch 1993: 4). This history, alongside my orientation of solidarity towards social movements advocating for greater social equality, led me to associate repressive security tactics with the political right.  As I was reminded through my research, the politics of militarized public security, whether in El Salvador, the US, or elsewhere, do not necessarily follow a straightforward right/left binary. Further, there is not any one pure, unchanging expression of ‘right’ or ‘left’ politics. As Goodale and Postero (2013: 3) note, the sharp analytical binaries typical of Latin 169  Americanist scholarship like “neoliberalism v. socialism; the Right vs. the Left; indigenous vs. mestizo; national vs. transnational” have been complicated by the tensions and contradictions that define efforts at social change across the Americas in recent decades. For instance, both President Obama and President Funes came to power in 2009. They were both leaders oriented further to the left than their predecessors, who both offered the promise of hope and change. This sense of optimism came after years of US President George W. Bush’s administration’s ‘War on Terror,’ and of neoliberal reforms and mano dura security tactics under successive ARENA administrations in El Salvador – raising the question: would these two administrations seek to “reconfigure the hegemony of neoliberal politics on the one hand and hardline regional and global security agendas on the other?” (Zilberg 2013: 226). In the realm of security, in many ways the answer would be no. Although both administrations would utilize a discourse of human rights, neither would mark a radical shift from their predecessors. For instance, President Obama would continue to oversee an expansion of the ‘Homeland Security State,’ while President Funes would further expand mano dura policing tactics (ibid: 240-242, citing De Genova 2010a).   With this in mind, it is still my goal in this chapter to provide a counterweight to US asylum narratives, but in a more complex manner than I had initially conceived, as a way to advocate against depoliticizing narratives of today’s conflict in El Salvador. To this end, in Section 6.2, I examine some gray areas around what it means to be part of a gang in El Salvador today; whether to join, or to be seen by others as affiliated. Repressive practices on the part of public security agents under the current Salvadoran government are part of this. But as I assert in Section 6.3, making sense of this repressive turn requires deeper contextualization of the national politics in El Salvador and of United States geopolitical interests and involvement.   170  6.2 Shades of gray Being granted asylum in the United States tends to demand a blunt, black-and-white interpretation of political persecution, defiance, survival, and agency in conflict. In practice, however, these experiences can be much more complex. In theorizing everyday acts of defiance and survival, critical scholars have recognized this complexity, including subtle forms of resistance, through concepts like “weapons of the weak” and “resilience,” (Scott 1985; Scheper-Hughes 2008). Transitional justice scholarship has shed light on the “gray zones” of long-term conflict to theorize complicity and the constrained choices that people who have been victimized may face to participate in violence (e.g. child soldiers) (Levi 1959; also see Baines 2011; Card 2000; Coulter 2008; Leebaw 2011). Theories of “complex perpetrators” and “complex political victims” also seek to account for this complexity (Baines 2009; 2016; Bouris 2007). In contrast, the “bad scripts” of the United States asylum system leave no room for gray areas or complexity. Perhaps for this reason, when I arrived in San Salvador in the summer of 2016, I had come to think of Central American gang involvement in very stark terms (e.g. either someone is a gang member or they are not). I also came to think of recruitment as generally ‘forced,’ or taking place in a way that would appear overtly-coercive to an observer. Yet I learned that there are many nuances to the meaning of gang affiliation today in El Salvador. This is not to diminish the very constrained, undeniably coercive environment in which young people face choices about getting involved with gangs (or come to be perceived by others as involved) (UNHCR 2016: 36). Rather, my intention is to suggest that the US asylum system should better recognize these complexities and the constrained environment that many low-income young people navigate in El Salvador. 171  6.2.1 Between comprehensive and repressive security strategies Today’s insecurity in El Salvador has roots in the disappearances, death, and displacement of the United States-funded armed conflict. Not long after the 1992 Peace Accords, mass deportations of marginalized young Salvadorans living in the US helped reproduce a model of rival gang activity from California to El Salvador – centered around the Mara Salvatrucha (also known as MS-13) and Barrio-18 (or B-18; the 18th Street Gang). In the post-war climate of ongoing social inequality, the gangs took root in low-income communities around the country. The majority of local gangs remain integrated within MS-13 or Barrio-18 (UNHCR 2016: 4-5), although Barrio-18 split into two factions in 2005: Pandilla 18 Revolucionaria, and Pandilla 18 Sureña. Thus, there are three main rival gang structures intact today in El Salvador (Martínez, C. 2016). The Salvadoran legislature first introduced mano dura public security strategies in 1996, and then successive ARENA administrations expanded them in 2003 and 2004. These policies invited the military to participate in domestic policing, suspended key procedural rights, and criminalized gang membership, as well as minor offenses like loitering and vagrancy. Holland (2013: 45) argues that because “[crime] is a rare issue of national importance that cuts across class and ideological lines” in El Salvador, mano dura helped ARENA appeal to a wide variety of voters and maintain power. El Salvador’s mano dura strategies have been critiqued on many fronts – not least for their failure to reduce homicide rates as promised. Critics argue that repressive policing has actually pushed gangs towards more lucrative and sophisticated activities, such as extortion, drug sales, and murder for hire (Wolf 2012b: 191-192; Olate et al 2014). They also argue that mano dura violates the human rights of young people, especially boys and young men, while ignoring socio-economic root causes (see Holland 2013).  172  Under twenty years of far-right ARENA leadership, the Salvadoran government’s response to gang activity was almost entirely repressive. Grassroots violence prevention and rehabilitation initiatives, such as the nonprofit organization Homies Unidos, developed in the void (Wolf 2012b: 198). When Mauricio Funes was elected in 2009 as the first-ever FMLN president of El Salvador, finally “[there] was an openness in the government to the idea of seeing gang members not only as perpetrators of crime but also as victims of broader structural causes, most notably marginalization” (van der Borgh & Savenije 2015: 161).  President Funes appointed David Munguía Payés to the position of Minister of Justice and Security, and he reportedly helped broker a short-lived truce among the gangs by providing material benefits to imprisoned gang leaders. In May 2013 the Constitutional Chamber of the Supreme Court deemed Munguía Payés’ appointment to be unconstitutional – leading to his removal from the position, and the appointment of a new minister, who did not maintain the truce (SSPAS 2017: 22). The Funes administration went on to use a number of repressive security strategies – re-integrating the military into policing and prisons, and passing an anti-gang law (van der Borgh & Savenije 2015). Under the FMLN, El Salvador’s movement away from mano dura has been slow and contradictory, yet by 2014 the approach to public security appeared to be on the precipice of change as now-President Salvador Sánchez Cerén took office.  As mentioned, President Sánchez Cerén has moved to reorient El Salvador’s national security agenda around a comprehensive, prevention-oriented approach. Yet just as it introduced Plan El Salvador Seguro [Plan for a Secure El Salvador] in early 2015, the administration also ramped up a repressive response to the violence that had been escalating in the country since 2014 (Lohmuller 2014). In January 2015, National Civil Police (PNC) Director held a press conference in which he essentially gave his officers permission to engage in brutality, stating, 173  “All members of the PNC that have to use weapons against criminals due to their work as officers, should do so with complete confidence. …There is an institution that backs us. There is a government that supports us” (cf Gagne 2015).  In May 2015, the Sánchez Cerén administration deployed three new military brigades into communities with a gang presence, on top of the 7,000 soldiers already deployed to contain the gangs (La Prensa 2015). That August, El Salvador’s Supreme Court officially categorized Barrio-18 and MS-13 as terrorist groups, while also implicating any “collaborators, apologists, and financiers” as terrorists (cf Daugherty 2015). The court justified its decision on the escalating violence in the country and the gangs’ growing level of organization. For instance, gang members held a public transportation strike in July 2015 in San Salvador, which culminated in the murder of multiple bus drivers who did not comply. Their demand was that the Sánchez Cerén administration reopen truce negotiations (ibid; Reuters Staff 2015). After a March 2016 massacre of eleven people in San Juan Opico (Martínez 2016, July 11), the president declared a “war” with the gangs that would end all dialogue (cf El Norte 2016, parentheses original): All the (security) measures that we have taken are in the spirit of combatting (the gang members), although some might say we are in a war, but no other path remains, there is no space for dialogue, there is no space for truces, there is no space to for understanding with them.  The administration swiftly introduced a number of “Extraordinary Security Measures.” These measures sought to gain greater control of the prisons in order to prevent communication among imprisoned gang leaders, increase punishments, further criminalize gang activity, and develop joint police-military task forces meant to take control of areas with a gang presence and to arrest leaders. In taking these measures, the Sánchez Cerén administration framed the gangs as the key source of the escalating levels of violence in the country (SSPAS 2017: 18-19). Members of El 174  Salvador’s Legislative Assembly responded swiftly in April 2016, unanimously approving fourteen Extraordinary Measures for the coming year (Lohmuller 2016; Martínez, C. 2016).  6.2.2 Place-based stigmatization When I met with advocates in San Salvador, I asked them what options exist for young people living in territories with a gang presence to maintain neutrality, when they find themselves stuck between different expressions of authority (e.g. rival gangs, as well as public security agents deployed to their neighborhoods). I inquired into recruitment pressures, the resurgence of repressive security policies, the United States’ role in Salvadoran security policy, and the possibilities of alternative, non-repressive approaches to public security.  When I asked advocates specifically about “forced recruitment” in the communities where they work, I got some puzzled responses. The notion of overtly-coercive recruitment is pervasive in the types of gang opposition asylum claims I had been exposed to, yet several advocates who I interviewed in El Salvador explained to me that in their experiences working in territories with a gang presence, while extremely-coercive recruitment does happen, the pressure to join a gang often tends to be more nuanced – reflecting the constrained environment in which many low-income young people live. One advocate who I interviewed, Karen, who has long worked with civil society organizations on issues of security and displacement in El Salvador, and who has served as an expert witness in US asylum cases, noted a difference in discourse between the United States and El Salvador in how the causes of displacement are typically explained. Karen critiqued the US asylum system’s tendency to restrict the narrative, noting:  That idea of recruitment in the sort of standard sense – that there's any army marching through and they're forcing people to join the army and take up guns in a battle that's going to take place – that's not happening. So, what happens in terms of recruitment is very difficult here, because for 175  the most part, most people in communities that are dominated by one gang, and in contact with or in danger of being attacked or repressed by another gang, is that kids identify with the group that's in their communities.   There's no, they may not want to be gang members, or they don't see it as something appealing to them. And for the most part, people who participate in church or go to school, don't actively try to report to the police criminal activity on the part of the gang – they just [coexist] – don't have any reason for the gang to be recruiting them. The gang doesn't need them, and they can't really sustain them.  In this way, young people may come to identify in some ways with the gang in their community, or simply coexist, without a desire to be an active member. Further, Karen went on to clarify, there is more than one way to be involved with a gang (e.g. hanging out with gang members, being a gang member’s girlfriend, etc.), and distinct gangs and their local affiliates, known as cliques, operate in different ways as well. Crucially, she explained, living in a community that is territorially controlled by a gang marks the residents as affiliated in the eyes of others – whether to a police officer, or to an opposed gang.  Karen made the point that, while extremely coercive forms of recruitment definitely do happen, she worries that the concept of “forced recruitment” feeds into an “us against them” narrative that has propelled repressive public security tactics in El Salvador, framing the gangs as “an alien troop:” That the gangs are an alien troop or something that we can identify, contain, and eliminate, and then the rest of us will all be better – that story gets repeated over and over again. And that makes it very hard for people to be truth-telling about their own experience, both here and when they leave the country and seek asylum elsewhere. It's very difficult for people to say they think it's more complicated than that.  Indeed, the idea of the gangs as an alien troop becomes untenable when considering the economics of extortion, which has become a prevalent enterprise across El Salvador, Guatemala, and Honduras since the end of the Cold War (Fontes 2016: 595; Cruz 2010). 176  Extortion was not my research focus, however Anthony Fontes’ (2016) ethnography of gang activity in Guatemala City is instructive. There, as in El Salvador, culpability for the most common expressions of post-Cold War violence like murder and extortion tends to be attributed in a very black-and-white way to MS-13 and Barrio-18 members. In the mid-1990s, gangs began collecting regular “rent” or “taxes” from local businesses, institutions, and residents of their neighborhoods, in exchange for providing protection – akin to the role of security guard. By the early 2000s, the Guatemalan, Honduran, and Salvadoran governments introduced mano dura policies. The resultant mass imprisonment of gang members allowed extortion to evolve into a well-organized strategy of livelihood and territorial dominance directed from behind bars. Gang leaders increasingly sought to expand their terrain by moving their extortion operations outside the neighborhoods they already controlled. This expansion has been achieved through extreme violence against those who refuse to pay the rent, and through terrifying threats of violence (ibid; Cruz 2010). As Fontes (2016: 601) notes in the case of Guatemala City, gang members employ many people not formally affiliated with the gang to carry out this labor: “it is often their neighbors, relatives, girlfriends, and wives who deliver the written demands or hand over the cell phone with an incarcerated marero [gang member] waiting on the line.” The material benefits of the extortion economy can also extend far beyond the gangs. In Guatemala, state employees like prison guards and police officers collect bribes in exchange for allowing the gangs to run their businesses, banks earn a surcharge off the money they allow people to deposit anonymously, and unaffiliated individuals even emulate the gangs to carry out extortions of their own (ibid).  Beyond extortion economies, the idea of the gangs as an alien troop becomes untenable when simply considering the number of active gang members across El Salvador, with estimates ranging between 30,000 and 60,000 members. The gangs are present across most regions of the 177  country, including urban and rural areas (UNHCR 2016: 10). Approximately 470,000 people were estimated to have an affiliation with a gang (e.g. as family, friend, or other connection) as of 2013. Estimates of the number of people with an affiliation grew to 600,000 to 700,000 as of 2015 – equating to roughly 10 percent of El Salvador’s population (ibid: 12). Describing an incident that took place in January 2015, journalist Óscar Martínez (2016) captures the deeply-embedded nature of the gangs in low-income communities across El Salvador. Members of Barrio-18 threatened all the inhabitants of the San Valentín condominium building in Mejicanos, which is just outside of San Salvador, that they must flee the residence or be massacred. The residents faced a dilemma about how seriously to take the threats, which would require them to abandon their homes. One resident quoted by Martínez (2016: 3610) elaborates: “This is what it comes down to,” one of the young, dark-skinned windshield-washers tells me. “Where do we go if there are gangs every place we can afford to live? If there are gangs every place with rent below $300, where do we go? What do we say if the new gangsters ask us where we’re coming from? If they’re from the 18s they’re not going to like what went down here. If they’re from the other gang, they’re not going to want us either. Right now we’re just leaving, later on we’ll see how to live.”   This dilemma is fueled, in part, by the way that gang affiliation gets mapped onto a person due to where they are from. Not all the residents ultimately decided to flee their home in Mejicanos, but as Martínez explains, the decision was not an easy one, not only because of the abovementioned dilemma, but because the threat of a massacre was terrifyingly plausible. Barrio-18 members had committed a number of recent homicides in the area – from burning a bus full of people, to murdering the son of a building resident (ibid: 3610). The issue, Martínez (2016: 3554) sums up, is that the “Gangs don’t leave. They are part of the social fabric.”  This rootedness, and the way that gang affiliation gets mapped onto places and onto people (regardless of whether they are actively involved with the gang, and regardless of whether 178  they want to be associated with the gang), is crucial to making sense of the impacts of the Salvadoran government’s most recent repressive security measures. Some human rights advocates have denounced this repressive turn as part of a “social cleansing” going on in the country, which ranges from the everyday police harassment of young people profiled as gang members, to extrajudicial killings committed by uniformed public security agents, to assassinations carried out by death squads (see WOLA & DPLF 2016).  I asked advocates for their perspective on this notion of social cleansing, including what it looks like in practice, and who its targets are. When I asked one advocate, Andrea, who does violence prevention work near San Salvador, whether the police tend to know who is involved in a gang and target them (as opposed to engaging in more indiscriminate profiling), she responded by first complicating my framing of what it means to be involved with a gang. Andrea critiqued black-and-white discourse around who is ‘in’ and who is ‘out’ of the gangs. She then explained that in the communities where she works, public security agents target a broad range of people, contributing to a wide place-based stigmatization:  The problem that I’m talking about is the black-and-white [discourse]. It forgets about the grays that exist in the middle. Most of all, because these two extremes do not exist. Only the grays exist. There are young people who do not have a very strong link, but do have a link with the gangs. This could be my cousin, my neighbor, my brother, or my friend. And all of these could generate some type of link. There are young people who of course talk to them, sit with them, but don’t participate.…The police check over these young people in the community with a constant logic of stigmatization.  In sum, gang affiliation in El Salvador is much more nuanced than any notion of formal membership. These gray areas are not unique to the gangs in Central America. Shaylih Muehlmann (2013) makes a similar argument about drug trafficking structures in the context of Mexico’s rural northern borderland, where drug economies are deeply embedded in the everyday lives of working-class people. As with gang extortion profits in Central America, drug 179  trafficking profits radiate out to powerful actors far beyond the lowest-level people and businesses who facilitate the movement of drugs (ibid; Fontes 2016). Likewise, there is a whole spectrum of gang affiliation in El Salvador – including when a young person is simply affiliated in the eyes of others because they are from a stigmatized community, as I explore further below.  6.2.3 Resurgent militarism Over the past few years, public security agents in El Salvador, meaning the PNC and the armed forces, have increasingly targeted young people in low-income communities with a gang presence for harassment and abuse. The violence prevention organization Servicio Social Pasionista (SSPAS) serves people in Mejicanos, which has been treated as a “laboratory” for shifting security strategies – prioritized for Plan El Salvador Seguro, but also falling into the eye of the storm for the recent Extraordinary Security Measures (SSPAS 2017). In January 2017, SSPAS published a report on the impacts of these security strategies, based on focus groups with adolescents and youth in Mejicanos. The focus group participants reported they believe the authorities know who is actively involved with the gangs, and who is not, and thus concluded that police harassment of young people is a way to harass the community at large (SSPAS 2017: 64). Yet they also noted common characteristics among who is targeted and how, based on whether a young person has tattoos, their socio-economic status, and their gender (ibid: 46, 65). While young women and adolescent girls commonly face sexualized verbal abuse, young men and adolescent boys in Mejicanos report that public security agents make false accusations about “evidence” in their backpacks or bags, and threaten to transport them to a territory controlled by an opposing gang (SSPAS 2017: 46-48). Men and boys also report being beaten in a variety of ways – such as being kicked in the ribs, hit on the head, hit with a gun – and 180  observing public security agents firing a gun near people to intimidate them. This abuse can escalate to the level of torture (ibid: 49-50). Finally, men and boys report being targeted for arbitrary and illegal seizures, for example of personal items like national identification cards and cell phones (ibid: 53). The result of these arbitrary and excessive shows of force, the SSPAS report (2017: 62-63) summarizes, is that “some focus group participants felt less insecure when they were near the gangs, than when they were near the security forces.” On top of these everyday forms of violence that young people profiled as gang members may face, extrajudicial killings are another risk. One infamous incident took place in the middle of the night on March 25-26, 2015, when a group of fifteen officers and two investigators with the PNC shot and killed eight people on a coffee farm (PDDH 2016: 11). The PNC Control Unit would later explain that the police had responded to a report saying “between ten and fifteen subjects pertaining to the ‘Mara Salvatrucha’ who had lodged themselves at the San Blas Farm, in the municipality of San José Villanueva, were carrying high caliber arms and were planning their criminal activities” (ibid: 11). The PNC reported that when the police arrived, they identified themselves as officers, but then the subjects began firing shots from different directions. According to the PNC account of the incident, this spurred a forty-five-minute “armed confrontation” that ended in the death of “eight members of the Mara Salvatrucha,” some of whom had tattoos alluding to gang membership. In contrast, only one police officer was wounded (PDDH 2016: 12).  This official version of the story would be challenged by a report in El Faro, an investigative news outlet based in El Salvador. El Faro presented evidence that this was an arbitrary execution, rather than an “armed confrontation” (PDDH 2016: 1). This evidence included an interview with Consuelo Hernández, the mother of one of the people killed at the 181  San Blas farm, twenty-year-old Dennis Alexander Hernández Martínez, who worked at the farm as a clerk. Consuelo Hernández held that her son Dennis was not armed, did not know how to fire a gun, and was not a gang member. She asserted that he followed the police officers’ commands and tried to explain his role at the farm. Further, Hernández clarified that her son died from a gunshot to his head, fired from above, in the style of an execution (ibid: 4). In response to the investigative reporting on the deaths at the San Blas farm, as well as a similar “armed confrontation” committed in August 2015 by a joint police/armed forces team, the Human Rights Ombudsman’s Office of El Salvador (PDDH, Procuraduría para la Defensa de los Derechos Humanos) launched an investigation. In the case of San Blas farm, the PDDH determined that the police had indeed massacred unarmed people and attempted to cover it up (Martínez & Valencia 2016). The eight deaths were “extrajudicial killings” (ibid; PDDH 2016).  What happened at the San Blas farm was not an isolated incident. As of April 2016, the PNC reported two to three “confrontations” each day between public security agents and alleged gang members, yet most deaths were occurring on the side of the alleged gang members (Martínez & Valencia 2016). These “confrontations” persist, with most deaths taking place on the side of civilians. After a November 2017 visit to El Salvador, the UN High Commissioner for Human Rights, Zeid Ra’ad Al Hussein (2017), reported that, “According to civil society groups, from January 2015 to February 2017, more than a thousand civilians and 45 police officers were killed in armed confrontations between the police and alleged gang members.” This trend towards extrajudicial killings led El Salvador’s Human Rights Ombudsman, David Morales, to caution that “If the State begins again, as it did in the decade of the 80s, to exercise illegal violence, and this is tolerated, all we will have is a scaling-up of violence, a response more atrocious than the gangs” (cf Martínez & Valencia 2016).  182  The heritage of the armed conflict also echoes across reports of increased death squad activity in recent years (Bargent 2014). War-time death squads, which included military officers and businessmen in their ranks, tortured, abducted, and executed thousands of people in El Salvador (Gutiérrez 2007). Post-war death squad activity has targeted alleged gang members in the name of “social cleansing,” as well as politicians, government officials, and human rights advocates. In 2014, the Human Rights Ombudsman’s office began investigating possible police involvement in the recent resurgence of death squad activity (Bargent 2014). By August 2017, the investigative news outlet Revista Factum presented evidence that death squads are indeed embedded among uniformed police officers (Avelar & Martínez d’Aubuisson 2017). In sum, by the summer of 2016 when I was in El Salvador, state actors were playing a complex and contradictory role in the public security landscape. President Sánchez Cerén had first introduced a comprehensive security plan for the country, but then declared a “war” on the gangs at a time of escalating violence. Some uniformed public security agents have been implicated in abuse, harassment, and extrajudicial killings of civilians, while others have potentially played a covert role in death squads. At the same time, other sectors of the state, like the Human Rights Ombudsman’s Office, were denouncing and investigating these activities, alongside civil society groups and investigative journalists.   6.3 Conflict and displacement as political In discussing some gray areas around what it means to be affiliated with a gang, and in attending to the resurgence of repressive public security practices in El Salvador, my intention is not to lump all experiences of insecurity together as an indistinguishable “generalized violence,” as proponents of restrictive US immigration policy tend to do (see US Congress 2017, July 24: 15). 183  My goal is the opposite: to illustrate that today’s conflict in El Salvador, and related displacement from the country, cannot be dismissed as apolitical, “generalized violence.” I say this not only because the Sánchez Cerén administration has called a “war” on the gangs, but because today’s conflict carries traces of Cold War politics, US deportation policies, and twenty years of dual efforts on the part of ARENA and US leadership to keep El Salvador open for business, to the detriment of working-class people (see Goodfriend 2017; Zilberg 2011).  Under the neoliberal reforms that followed the Peace Accords, El Salvador proved to have rates of violent deaths that matched or exceeded those from the war (Cruz & González 1997, cf van der Borgh & Savenije 2015: 157). The harms that ensued – shootings and knifings in the cities, road accidents, and massacres in the countryside – were framed by dominant state and media narratives as random, accidental, or as the outcome of youth delinquency (Bourgois 2001; Moodie 2006; Zilberg 2007). The state’s retreat through privatization under two decades of ARENA leadership meant that it could not as easily be implicated in the population’s suffering as during the war (Moodie 2006: 74).  Following critical scholarship on security in Central America and Mexico (regarding gang activity, the “drug war,” and other areas of criminalization), contemporary regional violence cannot be divorced from politics (Coutin 2007: 164-165; Bourgois 2001; Corva 2008; Godoy 2004; Holland 2013; Moodie 2006; Paley 2014; Pine 2008; Wolf 2012a; Zilberg 2007). This is clear in El Salvador not only because militarized public security practices continue to play a part in today’s conflict, but also because the post-war withdrawal of crucial public services and failure to provide access to dignified living conditions for the majority was a political maneuver, not the absence of politics (following Moodie 2006; Bourgois 2001).  184  This political maneuver is inherently an economic maneuver as well. Most recently, conservative political actors have thrown a wrench in the enactment of the FMLN’s social programs, for example by blocking the funds needed to run them. At the same time, elite business interests have endeavored to keep El Salvador’s minimum wage at one of the lowest rates in Central America (Goodfriend 2017; Pineda & Stoumbelis 2017). In this section, I explore these themes, which came up in my interviews with advocates in 2016. 6.3.1 Reconciling contradictions When an FMLN candidate was first elected to the presidency in 2009, the party did not have a clean slate upon which to enact its policy agenda. In the realm of public security, this meant the heritage of mano dura – years of repression, which led to a growing complexity of the gang structures, and an often-contentious relationship between public security agents and stigmatized communities. One thing I asked advocate research participants about was how today’s repressive security policies and practices in El Salvador under a left-leaning government compare to the mano dura strategies of the early 2000s under ARENA leadership. One advocate with extensive experience in violence prevention, Javier, explained that public security agents confront a much more complex landscape of gang activity today than they did in the early 2000s. Since that time, the gangs have expanded their territorial reach and increased their use of extortion – activities that have proven difficult for the government to contain (also see Cruz 2010). Javier also observed that the FMLN has taken a complex position on repressive public security: [in the early 2000s] mano dura policies were adopted, precisely, publicly. Or you could say that mano dura did not involve a discourse of the government violating human rights.…On the other hand, now, the government’s discourse is that they have prevention policies that are more comprehensive, which they do – it is not that they reject the theme of prevention – but in practice, they deny that there is a practice of territorial cleansing.  185  Elana Zilberg (2011: 169) details that although the political left in El Salvador generally denounced the repressive security regime of the early 2000s, after the Peace Accords the FMLN did not adopt a dramatically different posture on public security than ARENA. The FMLN was not unique in this way. As Zilberg (2011: 169) explains, despite the apparent contradiction between repressive public security policies and a social equality agenda, “with the exception of Brazil, in the aftermath of their military dictatorships and dirty wars, the Latin American Left in general did not extend its human rights agenda for political prisoners to the domain of common criminals or to prisoners’ rights in general.” An advocate who I interviewed, Gabriela, who has long worked in violence prevention, discussed the Sánchez Cerén administration’s complex security practices, arguing that the recent repressive turn can generate distrust in stigmatized communities, making it difficult for the government to carry out its prevention work: They have created a special battalion to go and kill gang members. They have a different focus, even though they do at least maintain a discourse that there are preventative programs. But yes, frankly it’s complicated, because if there is aggression, the people know [about it]. The community is witness to what happens; that sometimes [public security agents] come and remove people from the house, whether they are gang members or not. In either case, this generates more violence. So, in contrast, then comes the good face, but there is a lot of distrust among the people.   Although repressive security strategies can generate distrust in stigmatized communities, other communities and individuals welcome these strategies. Carlos, an advocate who does economic development work in rural areas that have experienced a growing gang presence in recent years, explained to me that both gang extortions and death squad killings increased in the communities where he works. He reports that some people fled, while others shut themselves in their homes out of fear. By 2016, the state response was an increased public security presence. Carlos observed general support for repressive policing in the communities where he works. He 186  speculated that this support comes, in part, in response to media narratives that associate public security agents with safety, and popular assumptions of the guilt of alleged gang members: There was an ideological construction in the media identifying the security corps as protection. So the population [of these communities], generally, supports these types of methods. This is not something that people reject. On the contrary, people want the presence of these security corps in their community…because the common thought is that the solution would be to exterminate all the gang members. So, where there is a killing of a gang member, people celebrate. No one is worried if they investigate to have justice; because they accuse the guilty because they know they’re guilty. People support and easily justify this type of behavior from the police and the armed forces.   The media narratives that Carlos refers to are firmly-entrenched within the mainstream media outlets in El Salvador, which are owned by a small number of elite families. In an analysis of media representations of gang activity and mano dura practices in the early 2000s, Wolf (2012a: 44) finds that “gang members were dehumanized and their threat distorted and overstated such that suppression appeared inevitable.” This narrative, Wolf theorizes, allowed ARENA to garner widespread public support for mano dura, despite its failure to stop the growth of the gangs.  Like Carlos, Javier also spoke to the public pressure on elected officials (and on those running for office) to generate immediate results in the realm of public security, given the fear and palpable risks that people face. He also argued that a broad social agreement would be necessary in order to forge a different, less repressive path forward: In the majority of electoral debates in recent years, it has to do with this, to demonstrate that they are going to resolve the issue of homicides – to lower them, they say, with mano dura, and later with the [gang] truce. Unfortunately, the methods of prevention and reentry are not short-term methods. They don’t give next-day results. So the people demand results. But I think there has to be a more general agreement in the country, because all sectors are involved: the media, in the way they present the news; the education system; private business. I think we could move forward if there was a national agreement to get out of this scheme, because we already have more than 20 years of this repressive scheme.   Beyond a public demand for results, Gabriela critiqued the US government’s tendency to finance repressive security programs, rather than comprehensive social initiatives, and to push regional 187  economic policies that reinforce inequality, and thus contribute to conflict and displacement. She asserted that economic inequality is a crucial factor in El Salvador’s security situation:  Killing all the gang members is impossible. It’s like sweeping garbage under the chair, and later it’s still there.…The problem is not just the gangs; the problem is the economic structure that we have. The problem is the inhumane structure that does not permit a possibility of living with dignity to the majority of people. So instead of killing gang members, what needs to be done is find a solution to the causes for people joining the gangs. This is why we have the problem that everyone goes to the United States.  Gabriela held the Sánchez Cerén administration accountable for its repressive tactics (as did the other advocates I spoke to), while pointing to the constraints the party faces in enacting its agenda for social reform. I next take a closer look at some of those constraints. 6.3.2 Constraints to social reform As mentioned in Chapter 2, the 1992 Peace Accords ended El Salvador’s armed conflict, and established key democratic institutions. However, the ARENA government in power at the time did not allow for substantial economic reforms (Pineda & Stoumbelis 2017). Since the FMLN first won the presidency in 2009 and sought to counter El Salvador’s long heritage of socio-economic inequality through social reform, private actors and policymakers aligned with elite interests in El Salvador have sought to obstruct this agenda (ibid).  One example of how this has played out, which I heard about while I was in El Salvador, is through the obstruction of funds that have been appropriated for social and security programs – a maneuver that has often been led by four magistrates within the five-person Constitutional Chamber of the Supreme Court (Pineda & Stoumbelis 2017). For instance, in July 2016, this branch of the Supreme Court blocked the distribution of $900 million in government bonds that had been approved by the legislature. It also interfered in a planned increase in taxes destined to fund renewable energy initiatives. FMLN policymakers argued that this was an overstep by the 188  Supreme Court, and a political move from within the judiciary to undercut the other branches of government’s ability to govern (ibid; CISPES 2016, July 28).  The seeds of today’s obstructionist politics were planted even before the FMLN first won the presidency in 2009, with keen support from US state actors. This support is hard to miss in a cable (later released by Wikileaks) that then-US Ambassador to El Salvador, Charles Glazer, sent to the US State Department in 2008 (Goodfriend 2017). In the cable, Ambassador Glazer describes feeling reassured after meeting with a conservative analyst who “told us of a ‘Plan B’ in the works to insulate El Salvador from (leftist) FMLN mischief should Mauricio Funes win the March 2009 election” (cf ibid). One component of “Plan B” was to install sympathetic magistrates in the Supreme Court’s Constitutional Chamber (ibid).  This exchange sheds light on an obstructionist agenda led by conservative actors in El Salvador, and applauded by top US diplomatic leadership. It also hints at a broader trend taking shape in Latin America in recent years that has been labeled a type of “parliamentary coup” aimed at destabilizing left-leaning presidents. The most extreme examples have ended in the removal of the presidents of Honduras in 2009, Paraguay in 2012, and Brazil in 2012. These removals took place under a thin “veneer of legality,” and allowed conservative political actors to re-consolidate control in their wake (Pitts, et al 2016: 335). For instance, when President Manuel Zelaya was forcibly removed from Honduras by the military, the Honduran legislature quickly cobbled together questionable legal justifications to replace him. The US State Department swiftly recognized the post-coup regime (ibid: 337; cf Pineda & Stoumbelis 2017), echoing the support expressed by US policymakers and media coverage for the 2002 attempted coup against President Hugo Chávez in Venezuela (see Clement 2005; Young 2013).    189  In El Salvador, efforts to obstruct the FMLN’s social reform agenda have taken many forms, including in recent struggles over the country’s minimum wage. While I was there in 2016, the National Minimum Wage Council approved a plan for a cost of living increase to El Salvador’s minimum wage. The council, which is composed of representatives of the labor sector, private business sector, and government, approved an amount that remained below the cost of living. The two labor representatives on the council voted in favor of the lowest possible increase, alongside the business representatives – giving credence to longstanding concerns that the so-called labor representatives were colluding with the business sector (CISPES 2016, June 16).  The Labor Minister chose not to ratify this proposal – instead creating mechanisms to make the council more transparent, and calling an election for the National Minimum Wage Council (given that the prior representatives’ terms on the council had technically ended). The business sector responded by trying to undermine the Labor Minister’s actions, including by pressing the Constitutional Chamber of the Supreme Court to classify the council election as unconstitutional. Ultimately, in December 2016, the council finally approved an increase to the country’s minimum wage that was more in step with the country’s cost of living. This newly-increased minimum wage ranges from $200 to $300 monthly (CISPES 2016, December 16). The minimum wage conflict was on my mind while I was in El Salvador, in part because I heard from advocates that low wages can limit the possibilities of violence prevention work. During my interview with Lily, an advocate with a USAID-funded violence prevention program, she explained to me that one component of her organization’s work is to provide job training for low-income youth. The organization works to place participants with large companies like Walmart and with transnational corporate call centers. Lily explained that gaining English skills 190  was an especially attractive option for participants because call centers tend to offer a monthly wage of about $500 – far above the minimum wage. Lily argued that raising the minimum wage, and therefore expanding the availability of dignified work opportunities that are in-step with the cost of living was crucial to make job training effective as a violence prevention tactic.   USAID plays a complex role in the violence prevention landscape in El Salvador. The agency does not fund Salvadoran government violence prevention initiatives. Rather, it funds a number of initiatives run by major nonprofits, businesses, and other non-governmental groups. This has opened USAID to critiques for creating a competing violence prevention model to that of the Salvadoran government, but with more funds. Further, the USAID-funded violence prevention model has been critiqued for guiding participants into employment with large businesses run by United States corporations and the Salvadoran elite (Goodfriend 2017).  I heard about this tension between differing models of violence prevention while I was in El Salvador. For instance, I learned that USAID was on the cusp of launching a project called Puentes para el Empleo [Bridges to Employment], a $42.2 million project for 2015 – 2020 meant to bolster the employability of young people in municipalities prioritized by Plan El Salvador Seguro, in support of the Alliance for Prosperity. In doing so, the program promised to prevent violence and reduce migration (Secretaría Técnica de Planificación 2016). At issue was that the USAID program bore a striking similarity to an already-existing government violence prevention program, JóvenES con Todo, but with an ample budget and without an explicit link to small and local businesses (see JóvenES con Todo 2016).  It is puzzling why the US government would fund a competing program that is similar, though not entirely the same, to the Salvadoran government’s initiatives that are struggling for the funds to operate. As I have outlined in this section, the FMLN’s ability to enact its social 191  reform agenda – from obtaining funds for public programs, to establishing a minimum wage that meets the basic costs of living – has faced formidable barriers from elite actors in El Salvador. These obstructionist efforts have, at times, found support from powerful United States actors.  6.4 Conclusion From the United States vantage point, explanations of the causes of displacement, as well as justifications for foreign aid like the US Strategy for Engagement in Central American, tend to portray places like El Salvador as a chaotic space of “generalized violence” (US Congress 2017, July 24: 15) – a place that “people leave” (Zaman 2017). These one-dimensional framings depoliticize and simplify the causes of displacement, while erasing any messy role that the US might play. This is evident in Vice President Biden’s promise that the new foreign aid package for El Salvador, Guatemala, and Honduras would help “change the climate of endemic poverty and violence that has held them back” (Biden 2015, emphasis added). Labeling these issues as “endemic” to Central America strategically forgets a long history of US interventionism.  In this chapter, I have sought to complicate this framing by illustrating the political nature of today’s conflict that helps fuel displacement from El Salvador – politics that cross borders. The complex public security landscape in El Salvador prompted me to rethink my initial analytical lens as a researcher, which was stuck in a right/left political binary. I was reminded that in practice, the politics of militarized public security cross the political spectrum, and do not necessarily adhere to the left or right in a simple way.  I believe that it should be possible to make an argument in defense of Central American people’s access to asylum without reproducing a depoliticizing narrative of El Salvador as a “failed state” and the United States as a safe haven. Simplistic stories, for example that boil the 192  causes of displacement down to “the gangs,” lend themselves to simplistic solutions. These solutions could range from calls to eliminate gang activity solely through repressive tactics, or to address the causes of migration with foreign aid that promotes major infrastructure and extractive projects, which do not tend to benefit the most marginalized populations. Such solutions risk having the opposite effect – of causing further marginalization and displacement, and do not do justice to the complexity of what it means to live in a stigmatized community in El Salvador today, nor the nuanced role that elite actors in El Salvador and the US continue to play in the politics of public security, even with a left-leaning government in power.  Chapter 5 demonstrated that even the asylum seekers whose stories most readily fit into the bad scripts of the US asylum system are often rejected. This raises the following question: what is the point of telling complex stories if adjudicators deny even the strongest, most straightforward claims? Here it is important to point out that legal advocates and DHS officials alike regularly draw on academic publications and policy reports as evidence in asylum cases – whether arguing for or against deportation. Scholarly research thus informs the asylum system, whether researchers intend it to or not. This means that researchers have a clear inroad to build a conceptual language that better reflects the shades of gray that this chapter has only begun to describe. Ultimately, I argue for nuanced portrayals of displacement as a way to move away from the myth of benevolence so prevalent in the United States – and hopefully away from pity and closer to a sense of connectedness and political responsibility (following Nevins 2016; Razack 1998; Walia 2013). A deeper sense of responsibility, not to mention humility, from the United States would productively inform US public policy approaches – whether in the realm of immigration policy or foreign policy – that more genuinely seek to allow Central American people meaningful choices about whether to move or to stay. 193  Chapter 7: Conclusion  From the vantage point of United States public discourse, the countries that Central American asylum seekers leave behind are often portrayed as chaotic spaces defined by brutal violence. Divorced from politics and history, this violence appears inexplicable or even inevitable, and entirely separate from the US. This worldview was on full display in January 2018 when, during negotiations with members of Congress over immigration reform, President Trump railed against humanitarian protections for people from “shithole countries,” referring to Haiti, El Salvador, and the entire African continent (Dawsey 2018).  Asylum narratives contribute to this public discourse. They invoke an ‘archive of representation,’ embedding themselves as commonsense precisely because the imaginary that such places are inferior to the US, and to the Global North at large, already exists (see Braun 2003). This is a racialized imaginary with origins in the logic of domination set into motion by European conquests (see Maldonado-Torres 2007; Mignolo 2005; Quijano 2000). In the US context, it is an imaginary that draws strength from a heritage of racial thinking about Latin America as a space of brutality and excess, in contrast to ideas of the US as an industrious, democratic, and racially-pure space (see Berger 1995; Stepan 1991). This thinking underpins contemporary ‘imperial formations’ between the US and Central America, including deterrence strategy – a system of racialized knowledge production and governance over the mobility, well-being, and access to livelihoods of people in the Americas (see Stoler 2006; 2013).   Although the Obama administration’s amplification of deterrence tactics in 2014 promised to protect children from a dangerous journey north, this strategy has proven to expose Central American people to harm. This dissertation has analyzed three interconnected harms: (1) 194  the forced separation of families through detention, which harms their well-being and access to protection; (2) the funneling of asylum seekers into mass detention and denial of common asylum claims, which leads to deportation; and (3) the circulation of a depoliticizing narrative about the causes of displacement, which fuels a rationale for exclusion. These harms, which are often inflicted with impunity and in an arbitrary manner by US state actors and institutions, challenge the imaginary of the US as a beacon of safety in the Americas. In this way the United States can embody the violence that President Trump decries elsewhere in “shithole countries.”   My central argument in this dissertation is that the amplification of deterrence strategy expands a racialized system of governance over mobility in the Americas, while limiting public debate and distancing United States actors from any culpability for the harms that ensue. This dissertation contributes to a growing critique of US deterrence strategy by elaborating a coloniality/modernity analytical approach to the study of displacement. This approach creates a fuller picture of the longstanding power asymmetries that limit mobility in the Americas – namely, systems of white supremacy and their ties to colonial conquest that often go unnamed in popular and policy analyses of displacement, migration, and immigration enforcement. Such ongoing legacies can also go unnamed in critical migration and border studies. The analysis that runs throughout this dissertation centers the racialized and gendered dimensions of deterrence strategy through a focus on normative family life, masculinity, and femininity. The dissertation advocates for a coloniality/modernity lens to better account for the distinct, but related imperial formations that have targeted different marginalized groups over time (following Stoler 2013; Sundberg 2015b; also see Razack 1998; Walia 2013). This lens sheds light on how US imperial formations inflect the everyday lives of people targeted for domination, as well as those who benefit from imperialism (Kaplan 1993: 14, citing Williams 195  1980; also see Sundberg 2015b; Stoler 2016). This implicates US citizens like myself, who have the privilege to breeze across borders in the Americas, in the harmful efforts to restrict the mobility of Central American people. In tracing imperial formations, my goal is not to suggest that the repetition of violent historical patterns is inevitable. Rather, my goal is to theorize ways to interrupt these patterns (see Braun 2003; Stoler 2013). To this end, I next draw some conclusions about the role of state power in the harms caused by deterrence strategy and advocacy strategies to disrupt this status quo. The chapter concludes with a discussion of the dissertation’s strengths, limitations, and the ideas it generates for future research.    7.1 Historical continuity and complexity One theme that runs throughout this dissertation is the stubborn growth of militarized approaches to keeping people “in their place” in the Americas (see Mills 1997: 48). I mean this in two senses. First, deterrence strategy seeks to prevent Central American asylum seekers and migrants from arriving in the United States, or from leaving their communities in the first place. Broadly speaking, it does so by locking people up, banishing them from the country, and threatening them not to come. In the second sense, a North-South exchange of violent techniques of social control has long served to impede demands for social equality in the Americas. This ranges from the US-instigated “low-intensity” warfare tactics that fueled political violence and displacement throughout Central America during the Cold War, to today’s mano dura [iron fist] public security tactics that respond with repression to the insecurity that has taken root alongside post-war neoliberal reforms. These militarized tactics castigate marginalized members of society, whether the targets are “guerillas” or “gang members” (Zilberg 2007). In failing to address basic socio-economic needs like access to dignified livelihood opportunities, these tactics serve to 196  “keep people in their place” and then punish them for the outcomes (e.g. displacement). State violence across the Americas can thus impede both the physical and social mobility of targeted communities – embodying a racialized system of governance that has been conceptualized as “global apartheid” and “border imperialism” (Nevins 2016; Walia 2013).     Chapter 2 of this dissertation demonstrates that United States-led militarism plays a long-standing, ongoing role in the perpetuation of state violence and impunity in Central America. Chapter 6 explores some of the complexities and contradictions of efforts to demilitarize public security in El Salvador under a left-leaning government. This serves as a reminder not only of the polarized political landscape in El Salvador, in which powerful right-wing and US actors continue to obstruct efforts at social reform, but also that the use of militarized approaches to public security can span the political spectrum. This point holds true in the US as well (De Genova 2010a; Zilberg 2013). One telling example is deterrence strategy itself – a militarized approach to border security inspired, in part, by Cold War “low-intensity” warfare, which has steadily grown since the late 1970s (Dunn 1996; Mountz & Loyd 2014). This growth has been overseen by both Democratic and Republican US leadership, spanning the administrations of Presidents Jimmy Carter, Ronald Reagan, George H.W. Bush, Bill Clinton, George W. Bush, Barack Obama, and now Donald Trump. Immigration reforms that have expanded pathways to legal immigration status have often been accompanied by punitive enforcement measures.  Most recently, President Obama played the dual role of “deporter in chief,” as some migrant justice advocates called him, and “champion in chief of comprehensive immigration reform,” as he dubbed himself (Krogstad 2014). Although President Obama championed the cause of young undocumented people, he also intensified the policing and detention of noncitizens more broadly – particularly those marked as criminals and security threats – while 197  deporting more people than any prior president (NDLON 2014; Herz 2016; US ICE 2011; also see Chacón 2012; 2014). This pattern continues today. In early 2018, for example, the fate of Dreamers (DACA recipients and other young people brought to the US as children without legal status) and Trump’s demand for an expanded border wall and extreme new limits on family-based immigration became bargaining chips in congressional debates. The bartering of protection for a small population of “deserving” immigrants in exchange for further criminalization of a much larger “undeserving” population invokes a type of respectability politics familiar to the US immigration system (see Cacho 2012; Ngai 2004; Kanstroom 2007). Achieving the status of respectability is not possible for the majority of people with undocumented status, given that “For the last half century, economic restructuring has exacerbated poverty for the poor of color in the United States and abroad,” while public policies have criminalized the strategies that marginalized communities are able to access for survival (e.g. participation in informal and illicit economies; undocumented migration) (Cacho 2012: 119).  It is quite possible that President Obama’s deportation record will pale in comparison to his successor. Yet President Trump has the reigns to a well-oiled ‘deportation regime’ (De Genova 2010b) because his predecessors across the political spectrum built it. This historical continuity offers crucial insights for migrant justice advocacy – of the complicity of ostensibly progressive leadership and the pitfalls of respectability politics. The dilemma posed by the tradition of protecting “deserving” immigrants and criminalizing the rest is exemplified by the comments of a Democratic representative from New York following President Trump’s January 2018 State of the Union address. The representative asserted, “you would think coming out of this that every undocumented alien is actually a member of MS-13…That is not reflective of the overall immigrant community and I think that was disgusting that [Trump] continued to make 198  reference to them as if every immigrant was a member of the gang” (cf Everett, Kim, & Schor 2018: np). This critique of Trump’s speech reproduces the noncitizen “gang member” population as the foil to truly “deserving” immigrants. Such an argument can fuel punitive treatment of the most criminalized noncitizens, as demonstrated throughout this dissertation. These risks are well-recognized by abolitionist approaches to advocacy like the #Not1More movement, which demand an end to all deportations and refuse trade-offs like protection for Dreamers in exchange for an expanded US-Mexico border wall – insights that should inform scholarly, legal, and activist advocacy efforts at large.  Recognizing the historical continuity in the growth of a mass deportation regime and the need for solutions beyond protection for the most “deserving” paints a complex picture. This picture illustrates that the ‘solution’ to displacement for the past four decades has generally been further militarization and therefore criminalization of those not readily seen as “deserving,” with harmful results, including across members of the same family. Chapter 3 illustrates that protections for noncitizen children in custody are insufficient when they grant no rights to adults, such as the parents or caretakers of the very same children. Even when select asylum-seeking family members, such as children and mothers, are released to continue their cases, the ongoing detention or deportation of their loved ones, like partners and fathers, is harmful to all. This suggests that humanitarian protections are most meaningfully applied when loved ones can support each other (Jastram & Newland 2003: 557). In other words, “protection” loses much of its meaning when granted only selectively to certain individuals and withheld from their family.  Chapter 5 concludes that common Central American asylum claims – whether they hinge on harms that may be seen as “deserving” (e.g. gender-based violence) or harms generally seen as “undeserving” (e.g. resistance to gang recruitment) – are vulnerable to the same political 199  attacks against Central Americans at large. These political attacks frame conflict in the region as apolitical, “generalized violence,” and therefore as a non-legible basis for asylum. Depoliticization helps facilitate the high rejection rate of asylum claims from El Salvador, Guatemala, and Honduras. Chapter 6 illustrates some of the gray areas around what it means to be affiliated with a gang in El Salvador – complicating any simplistic notion of gang members as a discrete, “alien troop” – and questions tendencies to center the gang member as the sole source of violence. In contexts of long-running conflict, victimhood has complex meanings (see Baines 2009; 2016; Bouris 2007). Conflict and displacement from El Salvador, I argue, cannot be written off as apolitical, “generalized violence.”  The themes of complexity and historical continuity threading through these chapters lead me to suggest the most effective arguments in defense of Central Americans’ access to asylum cannot simply advocate for politically-palatable populations – meaning people who fit into the most legible categories of victimhood, such as children and women. Advocacy efforts should leave no one behind, whether adult asylum seekers, or the most criminalized populations easily written off as “bad hombres.” Highlighting these complexities and continuities is one contribution offered by this dissertation. This complex picture sets the groundwork for more durable, less divisive solutions (see Loyd, et al 2012). As I suggest next, this complexity also demonstrates that the US mass deportation regime is perhaps not as well-oiled as it appears at first glance, revealing opportunities for interruption.   7.2 Impunity and incoherence Alison Mountz (2004) suggests that, from the outside, immigration bureaucracies seem like disembodied institutions. They tend to be opaque with the public and offer few opportunities for 200  meaningful interaction between state personnel and the people targeted for enforcement and their allies. Some written policies are accessible to the public, while others can only be accessed in bits and pieces through mechanisms like Freedom of Information Act (FOIA) requests. Either way, making sense of everyday practices is no easy feat, given that “Policy on paper…narrates only a partial story, the idealized ways in which events should take place, rather than the ways that things actually happen on the ground” (ibid: 329). This inaccessible chasm between policy and practice can give ‘the state’ the veneer of a powerful, unified actor with an always-coherent agenda. As this dissertation demonstrates, US state actors and institutions do hold a great deal of power over the fate of Central American asylum seekers. For instance, Congress maintains an arbitrary quota of 34,000 detention beds to be filled (Carson & Diaz 2015). Within detention facilities, officials routinely carry out harmful and arbitrary actions, such as blasting the air conditioning at excessively high levels in the temporary Border Patrol holding cells – a practice that yields the nickname hielera [icebox]. Another key example is when officials choose to detain and separate family members, despite having the discretion to release them together. Given the inaccessible nature of detention centers, limited avenues for recourse, and the vulnerable position of detainees, harmful actions like these are often carried out with impunity.  In some ways, the evidence presented in Chapter 4 illustrates a seemingly clear state agenda, given that state actors do, at times, intentionally deploy family separation as a deterrence strategy. Former Department of Homeland Security (DHS) Secretary John Kelly even said in a March 2017 interview that DHS conceives of family separation as a form of deterrence (CNN 2017). On the other hand, my experience interpreting policies on adult family unity by reading between the lines of FOIA responses suggests a certain incoherence of policy and practice. With the increased arrival of asylum-seeking families in the summer of 2014, which was framed as a 201  border security crisis by right-wing and mainstream media accounts, the Obama administration rapidly expanded family detention. This sudden change makes it likely that officials have made ad hoc decisions on the ground along the way, as occurred a decade ago during the first DHS experiment with mass family detention (see Libal, Martin, & Porter 2013). It is plausible that the routine separation of adult family members outside of the mother-child “family unit,” such as fathers and male partners, stems partially from a lack of clear policies or protocols to process families, rather than an always clearly defined (or intentionally harmful) state agenda. As Mountz (2004: 339) finds in her analysis of the Canadian state response to human smuggling in 1999, “Everything changes in times of crisis, including bureaucratic operations. Policy that appears neatly on paper is more convoluted when implemented on the ground, when decisions are made in haste, without much time for discussion.”  Mountz’s insight that the state is always embodied, and therefore may be less coherent and powerful than it appears, is highly instructive. These insights also resonate with institutional ethnographies that have been conducted of US immigration bureaucracies and enforcement paradigms (e.g. Dunn 2009; Heyman 1995; Maril 2004). While not an institutional ethnography, my research (and my volunteer and solidarity work) afforded access to state officials involved in detention and deportation – from detention guards, to immigration judges, and DHS officials – which constantly reminded me of the complex, contradictory, and always embodied nature of the state.    Recognizing that some degree of discretion on the part of individual officials, along with a spectrum of intentionality, propels the harmful practices of state institutions and actors is not an invitation to absolve them of responsibility. Rather, such an analysis is useful because it illustrates a certain incoherence in state actions, and therefore possibilities to interrupt the 202  historical continuity of the US deportation regime. One advocacy strategy is to illuminate the gaps between the promises made by state institutions and their everyday practices. Chapter 3 makes this contribution, concluding with the point that DHS officials have the discretion to release entire families together, rather than detain and separate them. Doing so would allow the US government to better uphold its commitments to family unity, while supporting the well-being of families, giving them fairer access to protection, and allowing their cases to be processes more efficiently. To the extent that family separation does occur in an ad hoc way, I believe that calling for state accountability to the promise of family unity is a useful advocacy strategy. Of course, this advocacy strategy stands in tension with the evidence presented in Chapter 4 – that forced family separation has represented a stunningly banal form of racialized state violence throughout US history. Arguments in favor of “the family” have often been quite exclusionary – centering the white, heteropatriarchal family. This point illustrates that advocacy for “family unity” must be done in the most expansive way possible, to avoid reinscribing exclusionary norms around what counts as a family, or how families must live.   7.3 Future research directions  Before offering some final concluding thoughts, I outline three areas highlighted by this dissertation that demand further research: (1) a conceptual focus on displacement, complexity, and masculinity; (2) a methodological focus on deterrence as a matter of both foreign and domestic policy; and (3) an empirical focus on forced family separation through detention. Along the way, I reflect on this dissertation’s strengths and limitations in researching these three areas. 203  7.3.1 Conceptual analysis: Displacement, complexity, and masculinity Young men are a key demographic facing violence and displacement from El Salvador, Guatemala, and Honduras, yet little conceptual language exists to interpret the harms they experience and the barriers they face to accessing safety. This is one area highlighted by this dissertation that warrants further research. Most critical social science literature on gender, displacement, and humanitarian protection focuses on women. In the realm of asylum law and adjudication, the term ‘gender’ is often used as shorthand for ‘women,’ even though men do face gender-specific harms like forced recruitment, sex-selective massacre, rape as political prisoners, and forced sterilization (Oxford 2005). One strength of this dissertation is that it begins to extend this conceptual language by examining the dual roles of normative masculinity and femininity in a broad practice of exclusion towards Central American asylum seekers. Chapter 5 illustrates that asylum claims commonly submitted by men (e.g. forced recruitment), are denied in much the same way that women’s domestic violence claims have long been denied. Winning an asylum claim requires that applicants fit their experiences of harm into rigid black-and-white categories. These categories tend to dismiss as apolitical the most common harms faced by Central American men and women alike. In contrast, Chapter 6 shows that these harms, as expressed in El Salvador, are unquestionably political – shaped by the heritage of Cold War violence and the neoliberal reforms and punitive public security measures that followed.  Although this dissertation demonstrates the insufficiency of black-and-white asylum narratives and begins to conceptualize some shades of gray, my research in El Salvador was limited in its time and scope. It is clear that the US asylum system is sorely in need of conceptual language that more accurately reflects today’s conflict in El Salvador, Guatemala, and Honduras – to shake off the bad asylum scripts of the Cold War, which never sufficiently recognized 204  common wartime experiences of persecution and have only become less adequate in the aftermath of the Cold War. Social scientists are in a position to help build conceptual language that better accounts for the “gray zones” of long-term conflict (see Levi 1959), the barriers that criminalized young men face to accessing safety, and the ongoing United States role in conflicts in Central America. Legal advocates and DHS officials alike use scholarly research as evidence in asylum cases, whether arguing for or against an asylum seeker’s deportation. This creates an opening for scholars to help reshape the conceptual language that defines asylum adjudication. It also implies that social scientists are actively shaping asylum narratives, whether they wish to or not. The need for more nuanced narratives of displacement only grows more urgent as the Trump administration continually poses new threats to Central American asylum seekers, particularly those targeted as “bad hombres” (e.g. see Dreier 2018).   7.3.2 Methodological approach: Deterrence as foreign and domestic policy A second area demanding further scholarly attention that this dissertation points to is methodological. Since the beginnings of the Chinese Exclusion era in the late nineteenth century, US immigration enforcement has stood at the threshold of foreign and domestic policy. Adjudicators of that era conceptualized mass immigration as a form of foreign aggression or invasion (Chae Chan Ping v. United States 1889: 606, cf Varsanyi 2008: 884). This rhetoric of invasion persists, although now applied to Latinos (Chavez 2013; Ngai 2004). Following Mathew Coleman (2008a), the figure of the unauthorized migrant continues to link US foreign and domestic policy – a connection that studies of displacement and immigration enforcement regimes should account for methodologically, as I have endeavored to do in this dissertation.  As noted in Chapter 1, the Obama administration expanded deterrence tactics both within and beyond US borders in 2014. Mexico’s Southern Border Plan, which draws on US funds and 205  training to interdict migrants passing through Mexico, is one prime example (Ribando Seelke 2016). Another example is the US Strategy for Engagement in Central America, which tasks aid recipient countries with greater policing of their regional borders (Meyer 2017). This outsourcing of deterrence persists under Trump. For instance, by April 2018, DHS had expanded its collaboration with Mexican authorities to collect the biometric data of Central Americans and other noncitizens apprehended in Mexico. The goal, according to a Chicago Tribune report, is “to identify criminals, gang members and potential terrorists long before they reach the US border” (Partlow & Miroff 2018: np).  The continued growth of deterrence strategy within and beyond the United States, and the novel ways that people are criminalized across borders, illustrates the ongoing need for research methodologies to address displacement that are not confined by borders. I found such an approach to strengthen my research in many ways, while limiting it in others. By conceiving of my project as a matter of both foreign and domestic policy, I was constantly reminded that the US asylum system is connected to geopolitics abroad, while the US asylum system informs how displacement is conceptualized and policed in Central America.  Attention to these links between United States foreign and domestic policy creates an expansive picture of how mobility is governed across the Americas. At the same time, I realized the limits of my own capacity as a researcher when I arrived in El Salvador and was not able to trace deterrence strategy in the way I had originally planned. As outlined in Chapter 6, this was partly because the new US aid package had not yet been disbursed, and because public security politics in El Salvador were different than I had imagined. Conducting a multi-sited project can limit the depth of research in each site. I dedicated the majority of my time and energy in this project to researching US domestic detention and asylum processes (see Chapters 3, 4, and 5). I 206  found it challenging to prepare for my research in El Salvador while still immersed in the United States side of my research. Thus, my desire to trace US deterrence strategy as both foreign and domestic policy faced the limits of my time and energy. 7.3.3 Empirical focus: Family separation through detention A third area for further scholarly attention that this dissertation points to is empirical: the issue of family separation through detention. As I complete this conclusion in the spring of 2018, forced separations in DHS custody persist near the US-Mexico border, while families and advocates continue to demand an end to the practice, or even clarity on how the practice works. In February 2018, a group of 75 members of Congress sent a letter to DHS urging the agency to reconsider its use of forced separation as a technique of deterrence, and to clarify its policies on family separation (Roybal-Allard, et al 2018). In April 2018, a coalition of several major immigrant rights advocacy organizations submitted FOIA requests to DHS and other relevant federal agencies seeking to clarify any “policies, guidelines, or procedures” that govern the separation of adults from minor children near the US-Mexico border (AIC 2018).  These demands for accountability and even the most basic information show the ongoing need for research about this issue, as well as the brick walls that stand in the way – impeding not only academic researchers like myself, but even elected officials an