Falling Outside: Excavating the History of Central American Asylum Seekers

This article takes a retrospective look at legal advocacy on behalf of Central American asylum seekers, which has been influential in the development of US asylum law and in the creation of an infrastructure to address immigrants' needs. The article considers three time periods when Central Americans have been deemed to fall outside of the category of refugee: (1) the 1980s, when US administrations argued that Central Americans were economic immigrants; (2) the 1990s, when civil wars in El Salvador and Guatemala came to an end; and (3) the 2000s, when some Salvadoran youths in removal proceedings have argued that they faced persecution as perceived or actual gang members. This retrospective analysis highlights the ways in which law can be creatively reinterpreted by legal actors, as well as how legal innovations carry forward traces of prior historical moments.

believing that by doing so they would earn the right to return to the United States m quickly. In 2008, when I interviewed them in El Salvador, the three brothers were sti hoping that eventually, through US citizen relatives in the United States, they w be able to return to the country legally.
The Ramirez brothers' experiences, which are not unusual, raise a number perplexing questions. Why wasn't a family who fled a civil war in fear that their childr would be forcibly recruited able to obtain political asylum in the United States? H despite not being granted asylum, was this family eventually able to qualify for permanent residency? And why were three of the children ultimately deported, though they feared being subjected to gang-related violence? Addressing these q tions requires excavating the legal history of Central American asylum seekers in United States, a history that has been influential in the development of US asylum law and practice (Churgin 1996).
My excavation of this history focuses particularly on the political and legal impli cations of falling outside the category of political asylum. On one hand, the U government's treatment of Central Americans as generally undeserving of polit asylum gave rise to considerable political and legal advocacy, which in turn impr protections for asylum seekers and contributed to the formation of an infrastructur immigrant-rights organizations in the United States. On the other hand, the failure t award asylum in a timely fashion left Central Americans, such as Pablo, Francisco, an Jorge Ramirez, without US citizenship and therefore still vulnerable to deportat many years (in this particular case, sixteen years) after they had entered the Un States (Kanstroom 2000(Kanstroom , 2007. Furthermore, in countries such as El Salvador, t character of violence has changed from overtly political to seemingly crimina nature. This change has led to continued renegotiation of the category of polit asylum, as youths who fear gang-related persecution seek to avoid deportation (Zilber 2004, forthcoming;Godoy 2005;Moodie 2006). A retrospective analysis of Cent Americans' efforts to obtain safe haven in the United States sheds light on bo the power of asylum's promise and the histories through which people are rend deportable.
In excavating the history of Central American asylum seekers, I attend to law's archeology, that is, to the layering of documents, statutes, court cases, notices, and records that take form at particular historical moments (see also Merry 2004).2 Atten tion to such layering is critical, I argue, because the documents and texts of which law is composed do not remain confined to a single historical context (or stratum). Rather, these items have ongoing impact in several ways. They can influence subsequent legal innovations; some of them may be retroactively reinterpreted in light of new develop ments; and, in the case of court decisions and statutes, they are projected into the future in that they are expected to shape or govern events that have not yet occurred.
Law combines the qualities of being produced over time (and thus being a product of history) and moving through time (being a means through which history is pro duced). Law's complex temporal character is further complicated by the fact that legal 2. Merry defines the archaeology of law as "an historical analysis of layers of legality and the historical contexts of their deposition." She further notes, "the archaeology metaphor suggests simple continguity in chronological order, but in practice each system affects the operation of the others" (2004,570).
Falling Outside 571 claims are staked in a range of forums. With regard to the status of Central American asylum seekers, coalitions of attorneys filed class-action suits on their behalf, individual attorneys represented clients in deportation proceedings, advocates sought legislative and administrative remedies, and individual Central Americans applied for asylum, seeking, among other things, to obtain work permits while their applications were pending.3 Legal artifacts produced in one forum can also reshape the claims being staked in another (see also Mather and Yngvesson 1981), thus enabling legal actors to put forward creative reinterpretations but also potentially foreclosing possibilities or limit ing reinterpretation. In the present case, devices that gave unauthorized Central Ameri cans temporary status eventually were reinterpreted in ways that permitted the majority of such persons to become legal residents but that also bore the trace of earlier denials.
The carrying forward of this trace reproduced disparities between Central Americans from different nations and made numerous Central American youths available for deportation.
My analysis in this article draws on more than two decades of fieldwork among Central American-and particularly Salvadoran-asylum seekers and immigrants in the United States. From 1986 to 1988, I did fieldwork within Arizona and California segments of the US sanctuary movement, a network of congregations that declared themselves sanctuaries for Salvadoran and Guatemalan refugees. This fieldwork included observing and participating in sanctuary activities, helping to document Central Ameri cans' asylum claims, and interviewing more than one hundred movement participants. In the mid-1990s, I examined the evolving legal strategies of Central American advocates as civil wars in El Salvador and Guatemala had come or were coming to a close. As part of that research, I observed the legal services programs of three Central American community organizations; attended immigration hearings; and interviewed advocates, migrants, and legal services providers. In the early 2000s, I analyzed the shifting significance of the Salvadoran population in the United States at a time when migrant remittances had become critical to the Salvadoran economy. For that project, I inter viewed policy makers, advocates, and migrants in El Salvador, Washington, DC, and Southern California. Lastly, I am now completing a study of the power and limitations of nation-based membership categories in the lives of people who were born in El Salvador but who grew up in the United States. This project has entailed interviewing approxi mately 130 Salvadoran youths and individuals who work with these youths, in both El Salvador and Southern California, including 41 individuals who were deported to El Salvador after having lived the bulk of their lives in the United States.
This article contains four sections. The first section further develops my rationale for examining law's archeology. The second section discusses how legal advocacy on behalf of Central American asylum seekers during the 1980s gave rise to temporary remedies. The third section examines how these temporary remedies fared in the 1990s, when civil wars in Central America concluded. The fourth section considers the long-term effects of temporary remedies on Salvadorans who grew up in the United States but who later faced deportation. Throughout the article, I attend to the ways in 3. Frequently, such applications were prepared by a notary who, unbeknownst to applicants, was not actually authorized to practice immigration law.

LAW & SOCIAL INQUIRY
which being deemed outside the category of political asylum impacts individuals wh in many cases, have experienced extreme forms of violence, including the violence associated with migration itself.

LAW'S ARCHEOLOGY
In attending to law's archeology, I treat law as an artifact that is const large degree out of preexisting material. Crafting a statute, writing an op a file, or issuing a document entails entextvudization, that is, excerpting ele texts, documents, or records to be redeployed in a new case or contex Briggs 1990; Richland 2008). Textual redeployments invoke texts tha been deemed authoritative, make use of agreed-upon language, ensure that applies to a previously delineated population, and occur as part of co making cycles (Riles 1998;Halliday and Carruthers 2007). Each instan builds on prior instantiations, even as redeployment in a new context law, making it part of a new legal conversation and a new administrative m also Urciuoli 2008). Law can therefore be seen as the residue of prior n residue that leads forward as well as into the past.
Archeologically, legal artifacts are defined both synchronically, by the relation to other artifacts in a given stratum, and also diachronically, by t they occupy in a series. Importantly, though, the elements that make up l put, but rather move to new strata as they are invoked or cited. Entextual enables law to move forward and backward in time through its redep exhibits what Judith Butler describes as the "double-movement.. . wh stituted' means 'to be compelled to cite or repeat or mime'" (1993, 220), as Bruno Latour calls "backward causation," the retroactive creation of someth always already there (1999).
Attending to law's archeology highlights the role of legal form in both and precluding legal indeterminacy. Law's indeterminacy, which is often em critical legal scholars in particular, has been attributed to the contradic law (i.e., statutes conflict with each other) and to the fact that laws "on th be applied in situations that lawmakers could not have fully anticipate see also Hagan, Ferrales, and Jasso 2008). This indeterminacy is crucial within political struggles, as meanings are contested by actors and a according to legal realists, can usually find a precedent to justify a de (Llewellyn 1962;Collier 1973;Comaroff and Roberts 1981;Greenhous and Engel 1994;Matoesian 1997). At the same time, scholars have not rights are used as a vehicle for social change, associated notions of propert and agency are simultaneously reproduced. Such reproductions limit l gemonic potential (Hunt 1990;Hirsch and Lazarus-Black 1994) and als those who seek to redeploy law are constrained by earlier meanings th forward. The persistence of such earlier meanings limits law's indetermina My own focus on law's archeology builds on this work by examinin tualization enables a kind of temporal movement. The reproduction of pree elements allows law to return to and reconstitute a prior moment but also Falling Outside 573 legal moments forward through time. Thus, the future may anticipate the present, or the present may retroactively constitute the past (Yngvesson andCoutin 2006, 2008). The redeployment of legal artifacts creates potential reinterpretations, but it also reproduces a historical trace or shadow (see also Corsrn Jimenez and Willerslev 2007). In essence, law's form facilitates innovation while also making it difficult to leave prior moments entirely behind.
In excavating the history of such temporal movements, I also draw on the notion that law is located within a variety of formal and informal legal transactions (see, e.g., Merry 1990;Yngvesson 1993;Munger 1996, 2003). Filing a class-action suit and applying for asylum through a notary both have legal effects, and the cumulative impact of these effects may reshape law, regardless of actors' legal knowledge and intentions. Focusing on effects makes it possible to examine the senses in which documents themselves become historical actors, not only in that they are capable of temporal movement, but also in that they can be imbued with a kind of authoritative force. For example, the possession of a work permit enables a person to be hired legally, not only because the permit confers legal authority, but also because the document itself is the authority. There are, of course, limits to documentary authority, such as when green card holders discover that they can still be deported, or when a work permit fails to authorize travel across national borders. Nonetheless, there are instances in which the mere existence of a document becomes more important than the claim that the document presents or the argument that it is designed to substantiate (see also Valverde 2005).
Excavating law's archeology is particularly instructive in the case of asylum, which was designed to address the exceptional instance of a person who has left his or her country of citizenship due to that country's failure to enforce the minimal human-rights protections that citizenship is supposed to secure (Arendt 1966). With the number of refugees and displaced persons rising worldwide (Zolberg 1990;Churgin 1996), this "exceptional" circumstance has, unfortunately, become all too common, making asylum an area of intense advocacy on the part of human-rights activists as well as a target for restrictionists who contend that it is subject to abuse (Hathaway and Neve 1997;Walters 2002;Fassin 2005;Junker 2006;Bohmer and Shuman 2008). Despite the pervasiveness of political violence, asylum law's grounding in exceptionality requires applicants to make the case that they have been singled out, subjected to a risk of violence higher than that experienced by the population at large, and forced to move (Harris 1999;Coutin 2001). Efforts to make this case also contend with the heightened securitization of immigration laws in a post-9/11 context (Swanwick 2006;Schoenholtz 2007;Settlage 2009;Macias 2010). Understanding how advocates and asylum seekers negotiate the meaning of political asylum in such adverse circumstances requires attending to the possibilities created by law's history, the potential to double-back and to carry forward, and the ways in which the identification of particular individuals or groups as "exceptional" creates bases for future actions.
These temporal movements both make asylum law powerful and leave particular groups outside the category of asylum. Drawing attention to how asylum law has evolved thus also helps to explain the inconsistencies that have plagued this area of law.
Analysts have attributed disparate outcomes of asylum claims to various factors, includ ing the indeterminacy of asylum law, political and ideological biases, restrictions on judicial review, the lack of legal representation, and asylum seekers' more limited process rights (Legomsky 2007;Ramji-Nogales, Schoenholtz, and Schrag 2007). A archeological approach suggests as well that doubling back to earlier legal moments not only creates the opportunity for reinterpretation, but also enables that which co before to haunt or shape that which follows, even if earlier iterations are submerged not visible. These traces of what went before both enable and limit reinterpretat and thus are not unlike what Scheffer, Hannken-Illjes, and Kozin (2007, 8) refer t the binding nature of earlier statements: "Binding does not resemble determination, p obedience, or even physical force. It emerges as a growing limitation by involvement." As we shall see, tensions between the supposedly "exceptional" nature of violen and its widely pervasive presence, between protection and control of migrants, a between expansion and restriction of rights have pervaded Central American asyl seekers' efforts to secure permanent legal status within the United States. These effor have resulted in a number of important legal developments that, in many instan prevented deportation but awarded only liminal legality, such as Temporary Prote Status, or the right to remain in the United States while an application was pend (Mountz et al. 2002;Menjfvar 2006). At the same time, given both sheer numbers the fact that increased Central American immigration coincided with the passage of th 1980 Refugee Act (Kennedy 1981;Churgin 1996), Central Americans' experiences h been central to the development of US asylum law more generally.4 Their story is one success against great odds but also of success tempered by the fact that many were gran remedies other than asylum and by the redefinition of immigration policy as a matte security (rather than labor or justice-see Calavita 1992). Central American you vulnerability to removal is a product of this history, a history that I will now excavat THE 1980S: ASYLUM DENIED During the 1980s, the primary challenge faced by Central Amer seekers was the need to establish that the violence they had experi tional in the ways anticipated by US asylum law. This challenge was political considerations, the prior history of US refugee law, and t inclusive and restrictive asylum policies. Preexisting definitions of "refu officials, when confronted with Central American asylum seekers, to int law in ways that precluded granting them status. Thus, during the government generally denied political asylum to Central American a basis that these migrants were fleeing economic conditions and gener of violence rather than targeted political persecution, and that they safely remain within their countries of origin or the Central Amer 2000). This government stance was shaped by Cold War politics and b the pre-1980 definition of "refugee," which, in the United States, individuals from the Middle East or those fleeing communist countrie Falling Outside 575 Kennedy 1981;Zolberg 1990;Swanwick 2006).5 Although the 1980 Refugee Act had made asylum available to individuals who were fleeing non-Communist regimes as well, Salvadorans and Guatemalans, who sought asylum from governments that the United States supported, were particularly disadvantaged. In contrast, Nicaraguans fleeing a country ruled by the Sandinistas, a leftist force that had ousted right-wing dictator Anastasio Somoza through the 1979 Nicaraguan revolution, received more favorable attention.6 The US government's contention that Salvadorans and Guatemalans did not deserve asylum was contested through individual asylum claims, multiple class action lawsuits, lobbying for legislative relief in the form of Extended Voluntary Depar ture Status, or EVD (the precursor to Temporary Protected Status, or TPS), and a broad-based solidarity movement. These actions sought not only to expand the category of asylee to include Salvadorans and Guatemalans, but also to compel asylum proce dures to adhere to principles embedded in notions of due process. Creative lawyering, the cumulative effects of the onslaught of asylum claims filed by Central Americans, and changed political circumstances in Central America eventually led the US government to take account of these asylum seekers-though not in the way that advocates had originally anticipated.
The 1980s debates over the immigration status of Salvadorans and Guatemalans were occasioned by the rapid deterioration of conditions in El Salvador and Guatemala, which resulted in a dramatic rise in the number of Salvadorans and Guatemalans immigrating to the United States. Although emigration from Central America to the United States had existed previously, the Salvadoran civil war (which raged from 1980 to 1992) and the intensification of civil conflict in Guatemala displaced huge segments of the population of each country. In El Salvador, the government launched bombing campaigns to drive civilians out of areas under guerrilla control, while in Guatemala indigenous groups, who were particularly suspected of supporting the opposition, were relocated or massacred (Montgomery 1995;Binford 1996;Schirmer 1998;Green 1999;Nelson 1999). In each country, the targets of repression were widespread, as the armed forces "equate[d] the government's critics with the enemy, repressing trade unionists, campesino leaders, opposition politicians, and student protesters with the same or more force than they use[d] on the real insurgents" (Schwarz 1991, 25).7 By 1984 there were 5. Through the 1980 Refugee Act, the US definition of "refugee" was brought into conformity with international law, which defined a refugee as a person who "owing to well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable, or owing to such fear, is unwilling to avail himself of the protection of that country; or who, not having a nationality and being outside the country of his former habitual residence as a result of such events, is unable or, owing to such fear, is unwilling to return to it" (United Nations 1951 Convention Relating to the Status of Refugees, Article 1). 6. Through the Nicaraguan Review Program, created in 1987 by the administration of Ronald Reagan, Nicaraguans who were denied asylum were allowed to remain in the United States, whereas Salvadorans and Guatemalans who did not receive asylum were ordered deported or granted voluntary departure. See Wasem (1997) for further details.
7. FMLN forces also committed human-rights violations, though these were fewer in number. The Commission on the Truth for El Salvador (1993) documented cases of summary executions of mayors, extrajudicial executions, and abductions by FMLN members. The FMLN also reportedly engaged in forced recruitment and forced requisition of food and other material goods. The Commission on the Truth for El Salvador nonetheless attributed 85 percent of human-rights abuses committed during the civil war to the Salvadoran armed forces or to paramilitary death squads (Kaye 1997).

LAW & SOCIAL INQUIRY
more than 1.2 million displaced Salvadorans-468,000 within the country, 244,000 Mexico or elsewhere in Central America, and 500,000 more in the United States representing 25 percent of the total population of El Salvador (Byrne 1996, 115). 1992, when peace accords were finally signed, Salvadoran community groups in L Angeles estimated that there were 1 million Salvadorans in the United States alon The majority of these migrants had entered without authorization.
At the time, advocates concerned about widespread human-rights violations i Central America sought asylum for the victims, but the Reagan administration h that Salvadorans and Guatemalans were economic immigrants rather than refuge The issues at stake in this debate included whether asylum was an appropriate rem for victims of generalized violence, whether these migrants were really victims human-rights abuses, and whether the earlier definition of refugee as a victim communism still dominated US refugee policy. Consistent with these attitudes, the US State Department, which was required weigh in on asylum applications, routinely advised INS district directors to den Salvadoran and Guatemalan asylum cases (see also Swanwick 2006). These recomm dations were generally followed. During the early 1980s, asylum applications filed Salvadorans and Guatemalans were denied at rates of 97 and 99 percent, respectiv (US Committee for Refugees 1986).
This broader political context played out at the level of individual asylum hearings where government representatives depicted Central Americans' accounts of perse tion as something other than the threat of violence, or as a type of violence that could not be linked to race, religion, nationality, or social group membership or to polit opinion. The influence of Cold War definitions of "refugee" is visible within th depiction, but the politically neutral language of the 1980 Refugee Act demand politically neutral rationales for denials. Transformation and reinterpretation of asylu candidates' claims were therefore accomplished through a number of devices, including challenging witnesses' credibility, requiring nonexistent or dangerous documentat (such as copies of death threats), delinking the decision to emigrate from the experien of violence, treating individual experiences as instances of generalized sufferin 8. For additional population estimates, see Aguayo and Fagen (1988), Montes Mozo and Gar Vasquez (1988), and Ruggles et al. (1985).
Falling Outside 577 defining violence as criminal rather than political in nature, and defining "indirect" threats, such as the assassination of neighbors or family members, as not rising to the level of persecution (Anker 1992;Durst 2000; see also Smith 1989).9 For instance, at an asylum hearing that I attended in Tucson, Arizona, in the mid'1980s, a Salvadoran applicant testified that, when he was a soldier in El Salvador, his life had been threatened by a sergeant who had previously been involved in a relative's assassination and who accused him of being a guerrilla sympathizer. The applicant had deserted and attempted to hide, but, upon being recognized, had fled to the United States. In this case, the INS attorney countered that Salvadoran authorities had a legitimate interest in apprehending deserters; that the applicant was at no greater risk than the general population; and that the government, had it wanted to persecute the applicant, would have had the opportunity to do so while he was in the armed forces. This last argument presented a frightening type of catch-22 for asylum seekers: if they escape, then their survival suggests that their lives are not really in danger; if they are killed, the danger is proven-but too late for them to gain asylum. At another hearing that I attended, an INS attorney asked an applicant who had been threatened by the guerrilla forces, "But they in fact didn't kill you?" as though merely being alive undermined her claim (see also Coutin 1993, 99-102;2001).
Such attempts to undermine accounts of alleged persecution were challenged during the 1980s by US religious activists who declared their congregations "sanctuar ies" for Salvadoran and Guatemalan refugees. Sanctuary practices invoked and reinter preted law in significant ways. As US law has increasingly held individuals legally accountable for the immigration status of those whom they transport or house, sanctu ary workers acted in ways that defined Central Americans as refugees rather than as illegal aliens. Specifically, movement participants brought Central Americans across the US-Mexico border without going through immigration inspection points, housed them in religious institutions or congregants' homes, transported them to places of safety around the United States, and publicized refugee "testimonies," that is, public accounts of the circumstances that had led Central Americans to flee their countries of origin. Sanctuary workers in Tucson, Arizona, one of the sites where I did fieldwork, argued that, through their actions, they fulfilled US and international legal require ments that the US government itself was violating.
To remedy what they saw as violations, sanctuary activists asked the US govern ment to grant temporary status to Central Americans either administratively or through legislation. Advocates also helped to arrange pro bono legal representation for Central Americans who were in deportation proceedings. For many of them, such advocacy was part of an effort to prevent further US military involvement in Central America, a goal that, they reasoned, would be furthered by the recognition that Salvadoran and Guatemalan regimes were violating their citizens' rights. Conflicting visions as to whether the US should offer military assistance to Central American governments were at stake in these legal strategies.
9. In contrast, Janis Jenkins notes that "the systematic deployment of terror as a means of coercion defies distinction between actual violence and the threat of immanent violence. Is not the display of mutilated bodies more than the result of violence or the threat of violence, but a form of violence itself?" (1998,124).

LAW & SOCIAL INQUIRY
Sanctuary activities did not go unchallenged by US authorities. There were several prosecutions of sanctuary workers in New Mexico, Texas, and Arizona; in 19 following a lengthy undercover investigation and criminal trial, eight moveme participants were convicted of conspiracy and alien smuggling (Coutin 1993). Th convictions were a means of delegitimizing participants' claims to be carrying US law.
The legal initiatives spawned by sanctuary workers and other Central Ameri advocates included not only assistance to individual asylum applicants, but also le actions designed to influence how the US government implemented asylum. On important case, known as the "young male case," sought to establish that all young me in El Salvador were at risk of forcible recruitment and were therefore should be eligib for asylum (Sanchez-Trujillo v. INS 1986). An attorney involved in this case descri the legal strategy in an interview with this author: In 1980 when the Refugee Act was passed, they added a category to the act . . . membership in a particular social group. And there had never been any definition of what that was and we decided that basically, this was what it w meant for, was people who . . . didn't necessarily have their own political opinions but the government suspected them.. . . And so we developed what was really imputed political opinion theory but couched it in terms of young men of militar age from El Salvador as a social group and who the government suspected of being guerrillas or guerrilla supporters.
Although ultimately unsuccessful, this case laid the groundwork for a subseque class-action suit, American Baptist Churches v. Thornburgh, filed in 1985 in dire response to the prosecution of sanctuary workers. In the "ABC case," as it came to known, plaintiffs argued that Central American asylum applicants had been deni equal treatment under the law. As a remedy, plaintiffs sought an award of temporary status to class members, as well as a prohibition on future sanctuary prosecutions. Thi case drew on the embedded notion of equal treatment while seeking an expansiv remedy, namely, temporary asylum status for a large number of persons.
Another case, Orantes-Hernandez v. Meese (1988), challenged such coerci practices as forcing detainees to stand in the Arizona sun all day in order to persu them to sign deportation papers. Yet another, Mendez v. Reno (1993), challenged perfunctory nature of asylum interviews and "alleged that interviewers were n trained and were ignorant of applicable asylum law, interpreters were not provid and sessions were rushed with little privacy" (Churgin 1996, 321 Now tell me the grounds on which someone's eligible for and entitled to get political asylum." And they would say, "What do you mean?" And I'd say, "Well, you know, there's five grounds in the statute on which someone's eligible or entitled to get asylum. Can you name them?" "Uh, no I can't do that right now." "Well, take your time. Think about it." They got through the entire deposition, they couldn't say, they didn't know a single thing. The success of attorneys' legal advocacy depended as well on the hundreds of thousands of claims filed by individual asylum seekers. Many of these claims were likely filed without a full awareness of the US government's stance toward Central American migrants, or even an understanding of political asylum. Rather, the passage of the 1986 Immigration Reform and Control Act (IRCA), which for the first time imposed sanc tions on employers who hired immigrants without work authorization, created a new demand for employment authorization documents (EADs) and therefore fueled a dra matic rise in asylum applications (Hagan 1994 notaries for assistance in obtaining a work permit. These notaries responded by f asylum applications, which, though often poorly prepared, entitled the migrants to w authorization while their applications were pending. Cumulatively, these applicat created a massive backlog, producing lengthy delays in adjudication. Thus, these cations were successful in one sense, in that they secured work authorization temporary permission to remain in the United States for the applicant, even though poor quality of the work may have harmed individuals with strong asylum claims wh could have benefited from properly prepared applications. By the decade's end, circumstances had changed in ways that made resolution Central Americans' asylum claims feasible. The INS was under pressure to reform asylum procedures (Chen 2006(Chen /2007, and the infamous assassination of six Jes priests in El Salvador had drawn international attention to human-rights abus making it more difficult for the United States to continue to provide military aid to Salvadoran government. Moreover, the war in El Salvador was at a stalemate, mig remittances had become a crucial boon to the Salvadoran economy, the Salvado government had begun to advocate allowing Salvadorans to remain in the Unit States, and the ABC case was entering the discovery phase, which was likely to p embarrassing to the US government (Blum 1991). In this context, the US Cong passed the 1990 Immigration Act, which created TPS and awarded Salvadorans teen months of this new status. Also, in 1991, the ABC case was settled out of co thus enabling Salvadoran and Guatemalan asylum applicants to reopen their ca have de novo hearings on the merits, submit new materials, or apply for the first tim ABC asylum hearings were to be governed by special rules designed to ensure consideration of applicants' claims. Some 300,000 Salvadorans and Guatemalans istered for the benefits of the settlement agreement.
Eligibility for TPS was bounded by a date-applicants had to have entered United States prior to September 19, 1990-and TPS was incorporated into settlement agreement, in that applying for TPS defined an individual as an ABC member (Guatemalans, who were not eligible for TPS, had to register as c 10. The number of asylum applications then dropped back to 1981 levels in 1991, after the adopted a "last-in, first-out" asylum adjudication policy and instituted a six-month delay in the issuan a work permit to new applicants. According to the "last-in, first-out" policy, the most recently subm asylum applications would be processed before asylum applications that were part of the backlog. Fu more, asylum applicants had to wait for six months before being eligible for work authorization. The g the "last-in, first-out" policy was to adjudicate new asylum applications before the six-month waiting p expired, thus preventing individuals who were denied asylum from ever becoming eligible for work au rization. Prior to the last-in, first-out policy, lengthy delays in adjudication had often allowed as applicants to obtain work authorization for years, regardless of the merits of their asylum claims. With new policy, in contrast, it became impossible to apply for asylum simply as a means of obtaining a permit. (Chen 2006(Chen /2007. After 1991, the number of asylum applications again skyrocketed, to more t 150,000 in 1995, as ABC class members rushed to submit their applications before a January 1996 dead Falling Outside 581 members). The award of TPS recognized that victims of generalized violence also had a need for safe haven but did not grant such victims refugee status, while the ABC settlement agreement set the stage for a fair adjudication of ABC class members' asylum petitions. Significantly, TPS built on but renamed EVD status, which had been awarded administratively rather than legislatively (Harris 1999). Although the ABC case was rooted in the 1985-86 prosecution of sanctuary activists, the settlement agreement addressed discrimination against asylum seekers, not future prosecutions of sanctuary workers.
Together, ABC and TPS created a potential remedy and applied this remedy on a mass scale, making all Salvadorans and Guatemalans already present in the United States eligible to apply for benefits. Advocates' goal of obtaining a remedy for victims of large-scale violence had been fulfilled. But this remedy was limited by the temporary nature of TPS and by the need for individual asylum adjudications. Furthermore, the population eligible for benefits was limited by eligibility dates, thus excluding indi viduals who entered the United States later and who also may have been subjected to persecution (Aita 2000).
TPS and ABC incorporated and redefined preexisting law and legal events, and they also resulted in the production of new documentation, namely, the hundreds of thousands of claims filed by applicants. These claims initiated administrative processes: papers were filed, work permits were issued, records were kept, and databases were constructed. The establishment of TPS and the ABC settlement reflected advocates' hope that peace could be achieved, making TPS unnecessary, or that asylum claims would actually be adjudicated, thus resolving applicants' status. Before the adjudications could occur, however, events intervened, once again making asylum an unlikely immi gration remedy for these migrants.

THE 1990S: ASYLUM AVOIDED
During the 1990s, the Cold War meanings of "refugee," the sanctuary prose the ABC settlement, and TPS were incorporated into new legal developmen that brought forward the traces of prior legal decisions while also permitting legal moments to be redefined retroactively. Central American migrants began with renewed recognition of their need for safe haven and with new legal mec TPS and the ABC settlement-designed to meet this need. However, the adminis burden created by thousands of ABC asylum applications, all of them gov complex settlement agreement, led to delays in adjudications. In addition, the s peace accords in both El Salvador (in 1992) and Guatemala (in 1996) made asylum less likely, and changes in US immigration law in 1996 eliminated other of legalization. The temporary status that had been issued to TPS recipient class members marked these migrants as different from the broader immigran tion, but the meaning of this difference was ambiguous. Were they una migrants to be expelled once their status expired, or were they in the country knowledge and authorization of US authorities, and therefore deserving?
This ambiguity complicated the legal mechanisms that had been create 1980s. On one hand, the fact that these migrants had not been granted asy them in a position of continued vulnerability to deportation. On the other hand, persistence and legal ingenuity that had resulted in TPS and the ABC agreemen allowed them to depict themselves as a bounded, documented, and "known" pop tion, rather than as mere unauthorized migrants. A complex set of circumstan ultimately allowed the latter depiction to prevail and thus retroactively reinterpre the meaning of Central Americans' presence-though not without the continue influence of the Cold War politics that had contributed to asylum denials in the f place and that were embedded in the asylum laws that US officials would invoke.
Although the 1991 ABC settlement was written in anticipation that adjudicatio would begin shortly after the agreement was reached, actual hearings on Centr Americans' asylum claims did not begin until 1997. The complexity of the settlem agreement, which required sending particular notices to applicants, made ABC asy petitions a lower priority for the asylum unit than non-ABC cases. As one asyl official told me during an interview, "The very fact that you've got a potential 250,00 [ABC asylum] cases that are on the backlog of the backlog ... I think, in essence, t doesn't help but facilitate keeping current [with new asylum receipts] and even be able to begin to dig into your backlog of non-ABC cases." The population carved through TPS and ABC required different administrative treatment. Additionally, both the ABC class counsel and at least some immigration officials anticipated that a alternative remedy would be created for ABC class members, making adjudication their claims unnecessary. One official commented that "around here [a word] was with jocularity, but with a sense of 'boy, wouldn't it be nice,' and that was 'amnes Amnesty." In the meantime, when the eighteen months of TPS awarded to Salvadorans i 1990 expired, a new status, Deferred Enforced Departure (DED), was invented for these migrants. It would later be extended until 1996, the deadline ultimately set for applyin for asylum under the terms of the ABC settlement. DED, like TPS, was a le innovation, but one that incorporated key features of its predecessors, particular eligibility dates. DED allowed former TPS recipients to remain in the United St with work authorization but introduced greater ambiguity regarding their future offered only the deferral of what would otherwise become a forced departure.
By 1996, the filing deadline for ABC class members' asylum applications, th political and legal context in which claims would be adjudicated had changed cons erably. DED and pending applications under the ABC settlement agreement were t redefined by other temporally contiguous legal artifacts. Although the peace acco signed in Guatemala and El Salvador did not necessarily put an end to all violence, they did make it more difficult for migrants to argue that they would face persecution if th returned. Furthermore, as Salvadorans and Guatemalans had to have entered th United States prior to 1990 to qualify for the benefits of the ABC settlement, by 1996 a significant amount of time had passed since the events that they described in th asylum applications.
The 1996 Illegal Immigration Reform and Immigrant Responsibility Act (IIRIR also added to applicants' difficulties by making alternative forms of legalization m difficult to obtain. Advocates had hoped that ABC class members who were den asylum would be eligible to apply for suspension of deportation, which was available to immigrants in deportation proceedings who could prove good moral character, who had Falling Outside 583 lived continuously for at least seven years in the United States, and who could show that deportation would pose an extreme hardship on the applicant or on the applicant's US citizen or lawful permanent resident relative. Delays in adjudicating ABC asylum claims therefore enabled ABC class members to assemble their documentary files (including work permits, registrations for TPS and DED) as evidence of the requisite years of presence to qualify for suspension. But then the 1996 IIRIRA eliminated suspension and replaced it with "cancellation of removal"; qualification for this remedy required proving at least ten years of continuous presence in the United States and that removal would create an extreme and exceptional hardship on a US citizen or legal permanent resident spouse, parent, or child of the applicant. Hardship to the applicant was no longer relevant; aliens who lacked the requisite ten years of residency or a qualifying relative were ineligible to apply. Additionally, IIRIRA imposed an annual cap of 4,000 on the number of suspension or cancellation cases that could be granted in a single year, making these an unlikely remedy for the approximately 300,000 ABC class members with pending asylum cases. Finally, IIRIRA created a "stop-time" rule, accord ing to which individuals who were issued a Notice to Appear in court could no longer accrue additional time in the country toward the requisite ten years of residence. It was unclear whether Orders to Show Cause (an earlier version of Notices to Appear) that had been issued to Salvadoran TPS recipients in 1992 but never mailed would be considered to have stopped their immigration "clocks"-a retroactive redefinition that could not have been anticipated when the orders were first issued.
In this bleak legal context and against the advice of colleagues who thought that the political climate was not receptive to their concerns, Central American advocates launched a campaign to grant residency to ABC class members. This effort was assisted by the fact that IIRIRA's provisions adversely affected not only Salvadoran and Gua temalan ABC class members, but also Nicaraguans who had been permitted to remain in the United States without a permanent status even if they were denied asylum. Salvadorans, Guatemalans, and Nicaraguans joined forces to request relief, and the Central American governments, concerned about reduced remittances and the impact of deportations, pressured the US government to assist their nationals. The Clinton administration responded favorably to these governments' concerns. At a May 1997 summit meeting, US president Bill Clinton told Central American presidents that it would be problematic for the United States to deport Central Americans who had lengthy ties to the United States and who supported their native countries' economies through family remittances. In addition, at least some US immigration officials sup ported restoring suspension eligibility to ABC class members and Nicaraguan Review Program participants, who had benefited from a 1987 policy of providing an extra level of INS review for any Nicaraguan who had been denied asylum.
As a result of these factors, an unlikely alliance of immigrant-rights advocates, US and Central American officials, and Nicaraguan, Guatemalan, and Salvadoran immi grants and their supporters was able to secure passage of NACARA.11 This legislation incorporated earlier eligibility dates and thus allowed the population affected by ABC and TPS to remain in the country. Passing this legislation was an extraordinary 11. Tellingly, this act was originally called the "Victims of Communism Relief Act." See Coutin (2007) for further details. accomplishment, coming only one year after the adoption of highly restrictive im gration measures. Yet, for advocates, this accomplishment was marred by the fact tha the Cold War ideology that seemingly had influenced asylum approval rates during th 1980s produced a disparity within the legislation. ABC class members, TPS recipien and certain other Salvadorans and Guatemalans with pending asylum applications w to apply for "special rule cancellation of removal"-basically the equivalent of t former suspension of deportation. In contrast, NACARA allowed Nicaraguan Revi Program participants to adjust their status to that of legal permanent residents.
Nicaraguans, who had fled the leftist Sandinista government, fared much better, as th remedy created for Salvadorans and Guatemalans, who had fled right-wing gover ments, was lengthier, more expensive, more cumbersome, and less certain.
This disparity of treatment was galling to advocates and to Salvadoran and G temalan leaders, resulting in tremendous pressure to resolve the perceived injust administratively through the regulations that would implement NACARA.12 T pressure focused on three questions: First, could the application process be streamlined by permitting asylum officials to evaluate ABC class members' claims, even thou these claims would be for both suspension eligibility and asylum? Second, could im gration officials specify the ways in which ABC class members might meet hards criteria? And, third, could immigration officials grant ABC class members a blan finding of hardship, thus virtually guaranteeing a grant in almost every case? Each of these issues was resolved largely in ABC class members' favor. Over the opposition immigration judges, who reportedly argued that only they could adjudicate suspension claims, asylum officers were granted authority to assess NACARA applications. On the regulations' authors explained the rationale for this decision in an interview w me: "[The asylum unit] had the files, and asylum had to do the interviews anyway. Mo would lose their asylum cases but be granted NACARA. It was a time-saver to do t together." The asylum unit's possession of the files that had been opened during 1980s and 1990s, when ABC class members had filed for asylum, TPS, and DED, t influenced this outcome. Likewise, although immigration officials concluded that hardship criteria were defined by relevant case law rather than the particular situatio of ABC class members, they nonetheless took the unprecedented step of specifying th criteria; the NACARA regulations thus codified case law. Finally, although immig tion officials determined that they could not grant ABC class members a blanket find of hardship (as, under the statute, suspension had to be evaluated on a case-by-ca basis), they concluded that ABC class members were sufficiently similarly situate justify granting them a rebuttable presumption of hardship. As an official explain "[ABC class members] were people we knew, people who had strong equities. They lived here a long time, they had kids, they were working, they had work autho tion. ... So we looked at it, and we decided we could do it. The presumption hardship shifts the burden from the applicant to the INS." Under the NACARA regulations, formerly prohibited actions-living and working in the United St illegally-retroactively became grounds for awarding status, and the liminality 12. There were also efforts to pass legislation that would resolve the disparity. See Coutin (200 chap. 7) for a discussion of that process. Falling Outside 585 temporary status was resolved in favor of documentation. Law thus retraced and retro actively redefined Central Americans' histories in the United States.
In an extraordinary turnaround, the passage of NACARA and the drafting of its implementing regulations caused an immigrant group previously defined as economic migrants to be considered deserving of protection, because of the lives that they had created in the United States and the circumstances in which they left their country of origin. NACARA cases were overwhelmingly approved, at a rate of 95 percent (Chris tian 2004). I had the opportunity to witness this process when I accompanied one Guatemalan applicant to her NACARA interview as an interpreter. When the inter view began, it was discovered that the asylum unit had no record of her original registration as an ABC class member. Suddenly, the applicant's status as an ABC class member, and thus her eligibility for NACARA, was unclear. Yet, instead of immediately denying her petition, the asylum official informed the applicant that, because immigra tion officials had lost numerous records, they were giving applicants the benefit of the doubt. The official advised her to draft an affidavit then and there, describing how she had registered for benefits. When she responded that she did not know how to draft an affidavit, the official helped her compose it-actually helping her create the record that would allow him to approve her request.
The extraordinary benefits that Central American asylum seekers obtained through NACARA came at some cost, as the NACARA regulations did not entirely resolve the disparate treatment of different groups of Central Americans. As a result, most Salvadoran and Guatemalan NACARA beneficiaries did not become legal per manent residents of the United States until the early 2000s, meaning that most were not eligible to naturalize until the late 2000s. And, as some of these migrants learned, until one naturalizes, one can be subjected to deportation. The trace of the 1980s denial of asylum therefore remained, within liminal statuses and bureaucratic delays, to be reactivated by stripping away the legal permanent residency of noncitizens convicted of crimes.

THE 2000S: ASYLUM BEYOND REACH
Salvadorans and Guatemalans who obtained legal permanent residency NACARA found that their new status--along with their green cards-wa redefined by other contiguous legal artifacts, specifically those focused aliens. In 1996, IIRIRA and the Anti-Terrorism and Effective Death Pen (AEDPA) expanded the range of crimes that could result in immigration c These laws also eliminated the 212(c) waivers previously issued to petition equities (such as evidence of rehabilitation or ties in the United States) outwei severity of their crimes (Kanstroom 2000(Kanstroom , 2007Morawetz 2000; Nevins population carved out by ABC and TPS and carried forward through NA 1997 was impacted by this 1996 legislation, as were other immigrant popu In the case of individuals with certain criminal convictions, IIRIRA an preexisting focus on criminal histories trumped the factors, such as circu departure, work histories, family ties, and period of residency, that al NACARA beneficiaries to be considered deserving. Although noncitizens c crimes had faced possible deportation prior to the 2000s, IIRIRA and AEDPA crea a new class of deportable migrants. This class was new in two senses: (1) the expan definition of crimes that bore immigration consequences led immigrants who would n previously have faced deportation to be placed in proceedings, and (2) the eliminat of waivers meant that many noncitizens who faced deportation were ineligible to apply for relief. Had Central American migrants and their families been granted asylu during the 1980s, when they first fled their countries of origin, or even in the 1990s when the ABC settlement was reached, then they might no longer have been non zens who could become vulnerable to deportation under the 1996 laws. One of the remedies available to individuals who are facing deportation is asylum, yet Cen American noncitizens who had been convicted of crimes and who feared being pe cuted in their countries of origin faced obstacles not unlike those that had faced Cent American asylum seekers in the 1980s. These would-be asylees feared gang violence and police repression, actions that were pervasive in El Salvador but that differed from th forms of political violence that asylum has traditionally addressed. Once again, it woul be difficult for these individuals to define themselves as exceptional or to constit themselves as a distinct class.
My analysis of the legal circumstances of this new class of potential deportees dra heavily on interviews that I conducted during summer 2008 with forty-one Salvadoran who had immigrated to the United States as children, grown up there, and then b deported. These deportees' experiences of the Salvadoran civil war were much li those of ABC class members and NACARA beneficiaries, suggesting that many co have been eligible for asylum had it been available. For instance, Enrique Lemus, whom I interviewed in El Salvador after he was deported, recounted that, as a child, he and h friends used to climb trees to pick mangos in an area near a guerrilla hideout. recalled, We used to see helicopters from the army actually go down. One time we saw an execution when we were on top of the tree.... They put four guys out onto their knees. They had a bag [over their heads]. And they just executed them there. And afterwards, they left on the helicopter, and we got off the tree. We actually went and played with the bodies. ... I would look at the blood spilled, and sometimes we would see guts spilled out. Something that a normal seven-year-old kid shouldn't be watching. But the environment that I was in, it was kind of becoming normal for me to see that.
Likewise, Edgar Ramirez, a deportee who later became a gang member in Los Angeles, recounted that as a child, he had seen "buses on fire. Shots everywhere. Headless bodies.. . . And on the way to school, I saw two or three dead bodies thrown there... . Psychologically, I was traumatized." After fleeing to the relative safety of the United States, interviewees such as Enrique and Edgar became caught up in the securitization of US deportation and border enforcement policies that occurred during the late 1990s, and that also rede fined political asylum as a weak point within border control practices. Several factors are responsible for this intensification. During the 1990s, US public concern over unauthorized migration grew, even as long-term undocumented residents formed Falling Outside 587 integral parts of US communities. This concern over immigration intersected with the war on drugs and other criminal-justice policies that subjected youths of color and low-income neighborhoods to intensified policing (Simon 2007). Targeting criminal aliens was an easy way to simultaneously increase removal statistics and "fight crime," and, as a government report noted, removing aliens immediately after they had com pleted their prison sentences was more efficient than releasing them into the interior, where they would then have to be apprehended (US General Accounting Office 1999). Improved apprehension and record-keeping techniques also saddled larger numbers of illicit border crossers with criminal and immigration records (Heyman 1995). The attacks of September 11, 2001, increased security concerns and derailed advocates' efforts to create a new guest worker or legalization program for unauthorized migrants. Detention center populations grew, the numbers of individuals removed from the United States increased, and migrants were increasingly prosecuted and The deportees whom I interviewed in El Salvador in 2008 had lived through this transformation. During the 1980s and early 1990s, when the securitization of US immigration law was just beginning, a simple entry without inspection or reentry was unlikely to lead to prosecution, detention was not mandatory for people apprehended on immigration violations, waivers could be granted to convicts facing deportation, border enforcement was not as stringent, and the fees charged by alien-smugglers were lower. Noncitizens accustomed to these earlier practices found it difficult to believe that, after 1996, criminal convictions would result in irreversible exile from the United States.
For example, deportee Roberto Orellana had immigrated to the United States in 1989 legally, at the age of seven, when his father obtained a family-based visa through Roberto's grandfather. Roberto saw himself as like everyone else: "I was feeling free, I was confident. I feel like an American because I had the same rights. I had no issues going places, like to TJ (Tijuana, Mexico) or to a bank." Roberto planned to become a US citizen. He joined a gang, however, and, after getting into a fight, was charged with assault and battery. He pled guilty to the charges in exchange for a reduced prison sentence of one month. After being released, he was caught riding in a stolen car, and, under pressure from his public defender, again pled guilty.
Believing that, because his father had naturalized when Roberto was underage, he had automatically become a citizen himself, Roberto did not anticipate that his con viction would affect his immigration status. Conversations with his fellow inmates convinced him otherwise. Roberto explained, "They told me that if I did more than a year, I qualified for deportation. I told them, 'No, man, I've got papers.' They said that it doesn't matter, that if you've got a criminal record, you've got felonies... 'Okay, that's it.. . You're going back!' I couldn't believe I was going back."

LAW & SOCIAL INQUIRY
The 1996 changes to US immigration law left noncitizens who had been convi of crimes with few means of challenging their deportation. Detention was mandatory most cases. As there is no right to a state-appointed attorney during immigra proceedings, detainees' families had to shoulder the costs of legal counsel. Detain who sought to appeal a judge's ruling had to remain in detention while the appeal considered, and detention center practices seemed designed to convince them that the had no hope of prevailing. One deportee, Mario Lopez, described his experience detention as follows: They wouldn't even let me see the judge. I requested it so many times. ... W the detective, officer, from INS took me to the headquarters of INS in Baltimore I told him, "I'm married to a US-born citizen." He said, "We don't care. That's the way we work." And we got there, [he] fingerprinted me. He said, "Would you like to see a judge?" I say, "Yes." He said, "If my supervisor approves it, you're ab to see [the judge]." He did not. They denied it. I would send letters from t detention center requesting a judge or a trial or something to fight the case. The would never respond. They would just be a pain to us. They even made you sign t papers without you [being] willing to sign the paperwork. I remember that, when I got in the detention center, they said, "We're gonna transfer you into Phoenix. I said, "I want to fight the case." They said, "No, you can't fight the case." And t just literally woke me up one day and said, "You're getting transferred." You didn get notice or anything. They would make new fingerprints, they would make co of it, if you didn't want to sign the papers. They would force you to sign your o deportation, saying that you are agreeing to get deported.... So that's what they'd say. "Okay, if you don't want to sign, just stay here. You're going to be here 12 ye if you want to." Sometimes they would just make copies of fingerprints. You wo just ask them a question, say, "Can I file this? Can I file that?" They would just s "I don't know." They would never give you an answer. And the treatment w you get deported is like you're a dog. To them, it is like we are clowns. Almost li we are from another planet. That's how they treat you.
As this passage suggests, the combination of mandatory detention and officials' tinual predictions of failure undercut migrants' abilities to fight or appeal their cases In this limited legal context, asylum once again emerged as a possible aven through which to challenge removal. One interviewee who tried to utilize this op was Cesar Rojas, who faced deportation after being convicted of riding in a stolen car credit card theft, drug possession, and violating his parole conditions. Cesar had ente the United States at age three when his parents, like many ABC class members, had f the onset of the Salvadoran civil war. At age eight he had obtained legal perman residency through his father. Cesar recounted, "I knew I was Salvadoran but I tho I was from the United States. Because I got there so little." Terrified of returning to Salvador, Cesar spent a year in detention fighting deportation. He explained, "Se prison, you get a lot of tattoos. They're not gang-related, just art. I have tattoos on m arm, my leg.... But over here [in El Salvador], they don't look at it as art. They at it as like you're a gang member." Cesar's applications for political asylum an withholding of deportation were denied, and he was ordered deported by a judge to his horror, reportedly told him, "I don't care if they are going to kill you when you Falling Outside 589 off the plane, you are still going back." Cesar appealed this decision but was again denied. Deeming further appeals pointless, he agreed to the deportation.
In El Salvador, where "zero tolerance" gang policies have been imported from the United States, Cesar's fears of persecution proved well founded. There he was arrested for attempting to kidnap a taxi driver, a charge that he says was fabricated. He experi enced continual harassment from police and private security guards who demanded that he lift his shirt so that they could inspect his body for tattoos. Cesar lived in continual fear of being mistaken for a gang member and killed, either by security forces or by gangs themselves. As further evidence that such fears were justified, Manuel Urquilla, another deportee, told me of two friends who had been murdered following their deportation.
One was killed at a party and the other while he sold CDs at a street stand. Presumably both were killed by gang members.
Although the risks of gang-related persecution appear frightfully real, few asylum applications relying on this argument have been successful.13 Beyond the difficulty of securing competent legal representation, these applicants face two more challenges: showing that a gang should qualify as a "social group" and defining gang-related vio lence as persecution. In Matter of Acosta, a 1985 decision involving a Salvadoran asylum seeker, the term "social group" was defined broadly as "a group of persons all of whom

LAW & SOCIAL INQUIRY
"persons resistant to gang membership" were persecuted in Honduras was not dealin with "a socially visible group within Honduran society" (2007,594). Additionally, social groups must now demonstrate "particular and well-defined boundaries" and must be "discrete" (Matter ofS-E-G-et al. 2008, 582, 584). The BIA therefore determined that, in El Salvador, a group made up of "male children who lack stable families and meaningful adult protection, who are from middle and low incom classes, who live in the territories controlled by the MS-13 gang, and who refuse recruitment" is too "amorphous" to demonstrate the requisite particularity (585). Only very narrowly defined groups, such as a family, can meet the definition of a social grou for asylum purposes (see LopeZ'Sosto v. Ashcroft 1985). Once again, a risk of becoming a victim of violence is deemed outside the protections promised by political asylum Furthermore, distinctions between political and criminal violence are used to make case law regarding political opinion irrelevant to these cases.
Clearly, the challenges faced by noncitizens convicted of crimes are shared by a broad swath of migrants, not only the Salvadorans and Guatemalans who were denie asylum due to Cold War politics in the 1980s. Nonetheless, it is precisely this seemin commonality that makes an excavation of law's archeology necessary. Individuals ar not intrinsically deportable or removable; rather, they are made removable by particular histories, policies, moments, and movements. In the case at hand, the illegality tha tainted the status of Pablo, Francisco, and Jorge Ramirez and their families at earlie moments in their lives was reactivated at the moment of their deportation. Cold Wa definitions of "refugee" that rendered Salvadorans and Guatemalans "merely" economic immigrants resulted in liminal legal statuses such as TPS and DED, and also influence the disparity within NACARA, which in turn delayed adjudication of NACAR applications. When legal permanent residency is stripped away, individuals' earlie liminality and alien status reemerges. Individuals who fear being persecuted as suspected or actual gang members find that, once again, the violence they seek to escape is defined as something other than political, classes of people turn out to be something other than social groups, and legal statuses other than citizenship become markers of undeserving ness rather than steps on the way to inclusion (Motomura 2006). Current deportatio policies retrace prior denials.

This excavation of the histories of Central American asylum seekers
States demonstrates both the promise and peril of asylum in immigratio are increasingly defined by security concerns. On one hand, Central able to carve out extraordinary exceptions to the more restrictive immi adopted in the United States in 1996, exceptions that resulted in a permanent residency for Salvadoran, Guatemalan, and Nicaraguan had entered the United States during civil conflicts in Central Ame tions were linked to the reform of US asylum procedures and to a innovations, including TPS, the codification of hardship criteria, an presumption of hardship. On the other hand, the Cold War concerns tha led Salvadoran and Guatemalan asylum claims to be denied also detr Falling Outside 591 remedies. Numerous Central Americans with long histories as US residents became vulnerable to insecurity, prolonged family separations, and eventual deportation.
Asylum holds out the promise of safe haven, human-rights protections, and the rule of law, but, when security concerns raise suspicion regarding the validity of asylum claims, then the desire to close borders, limit legal access, and keep out undesirables trumps the promise of protection.
As we have observed, legal artifacts both bring forward prior legal meanings and return to and reconfigure the past. In the situation at hand, Central Americans were issued work authorization in the 1990s, through TPS and pending asylum applications.
Documents that originally promised merely a temporary authorization to work were retroactively reconfigured as a tacit authorization of presence. At the same time, the details of the ABC settlement were designed in anticipation of asylum adjudications that, for the most part, did not occur. Nonetheless, the complexity of these procedures (which included requirements to mail particular notifications to applicants) con tributed to delays that lengthened the period of time in which applicants lived in the country with legal authorization.
Temporal movements (in which the future anticipates the present or in which the past is redefined retroactively) can result from both formal and informal legal actions.
Many ABC class members submitted applications through notaries who may have had a limited understanding of asylum law or through organizations that, to facilitate mass submissions, prepared only skeletal applications. In the end, the act of filing often proved more important than the content of applications, as filing defined one as a class member, and as most cases were decided according to suspension rather than asylum criteria. Each instantiation or iteration of law-the filing of an application, the firing of an individual who lacks a work permit, the deportation of a noncitizen following a criminal conviction-gives rise to new potentialities and foreclosures.
Attention to how law's form facilitates and precludes temporal movement dem onstrates that exceptionality eventually conferred US residency on many asylum seekers but also kept a significant portion outside of the protections promised by asylum.
Central Americans' efforts to obtain asylum were continually hampered by the legacy of the Cold War, which had defined these applicants as economic immigrants and unde serving. In the 1990s, when legal remedies were created, the delays occasioned by earlier denials had weakened Central Americans' asylum claims, while Cold War politics were reproduced through the disparate treatment of Nicaraguan, Salvadoran, and Guate malan asylum seekers. In the 2000s, even though most NACARA petitions were ultimately approved-with the important exception of individuals with criminal convictions-processing delays and the lack of availability of asylum during the 1980s and 1990s left a number of Central American youths without US citizenship and therefore, as noncitizens, vulnerable to potential deportation. Like their parents, who struggled to define wartime violence as grounds for political asylum, some of these youths have struggled to define police repression and gang violence as reasons to approve asylum petitions.
As the meaning of violence in formerly war-torn countries shifts from explicitly political to criminal in nature, asylum law itself may need to shift in recognition of the risks that security policies themselves create. In Central America, for example, gang activity has increased at least in part due to the deportation of gang members from the United States (Zilberg, forthcoming). It remains to be seen whether asylum can retooled for new situations of widespread "criminal" violence whose political imp tions are as yet unclear.
The archeological approach developed in this article to identify connections between asylum policies during the 1980s and deportations in the 2000s can be fruitfu employed to understand other legal phenomena as the residue of prior contests, d sions, and administrative actions. Examining how law is constructed over time rev the multiple ways in which law not only seeks to address the present but also re figures the past and haunts the future. Attention to such temporal movements sh light on the mechanisms that produce and preclude innovation, making it possible see the histories that are present on law's surface. Treating law as an artifact refocuse law's meaning within its material and textual form, within documents, actions, o ions, statutes, rulings, and regulations. Such a focus also overcomes distinction between law and implementation, between inert books and active judicial or enfo ment practice.
Indeed, this case study suggests that the potential for movement is embedded law's form, in the deployment of excerpted elements that come from and lead so where else. Law itself acts, through the accumulation of documents, files, and reco that not only are interpreted by others, but also make assertions through their m presence. Thus, the database of ABC class members, the records created by asyl applicants, and the work permits issued by immigration authorities helped to prod NACARA. Excavating histories denaturalizes the present-people are not intrinsica deportable; rather they have to be made deportable. Such excavations not only rev hauntings (such as the continued influence of Cold War politics, long after this war w over), but also hold out hope, namely, the possibility that pasts will be reformulated i ways that promote just futures.