The 1995 Asylum Reforms

By David A. Martin on May 1, 2000

As regulations go, the 1995 asylum reforms are among the most successful the immigration system has known, concretely achieving diverse goals embraced by observers from all parts of the political spectrum. But it is useful to remember that this is only the latest of several rounds of important changes to the asylum system in the United States. My aim here is to provide historical and global perspective on the 1995 reforms.

The Policy Tension

A substantial policy tension dominates the refugee field, at least in those countries that take asylum claims seriously — a tension felt by government officials, non-governmental organizations, and advocates alike, as well as by the broad public in most democratic nations. We genuinely want to protect those who are at risk of persecution, torture, or other serious harm. But at the same time we want reassurance that immigration is manage- able, that it remains subject to reasonable control.

Balancing the objectives of protection and control obviously presents a significant challenge. Persecution and civil strife occur beyond our shores, in outbreaks we do not regulate and often cannot predict. The extent to which such episodes trigger significant migration to the United States is also highly variable. U.S. or U.N. human rights and peacekeeping initiatives promise only limited effectiveness in curbing persecution or ending civil wars, although we have become better and bolder at this business over the last decade.

The Legal Framework

It might have been logical to leave the balancing as a question of pure policy, to be set case by case. That is, nations could simply decide ad hoc how to respond to each new refugee-generating event — how much to offer protection, how much to emphasize control. They could consider the range of threats posed to the asylum seekers, their likely numbers, the responses other than relocation that seem to be available, and the current economic and political conditions in the source state and the potential receiving states, in order to craft a situation-specific response.

Instead, since at least the 1950s, the world community has ambitiously treated these issues — in significant part, but not exclusively — as questions of law. Both international treaty law and domestic statutory law in most countries now set forth a legally protected entitlement for those who meet specified criteria. They establish an entitlement to asylum or at least to assurance against return to the place where the threat exists — the so-called nonrefoulement guarantee based on Article 33 of the 1951 U.N. Convention relating to the status of refugees. This turn toward law does not mean that the balancing process is over, or that control no longer matters. Instead, it merely shapes the arena where the struggle between our conflicting impulses goes on.

The Convention provided the key legal standard for judging refugee claims: whether the person has a "well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion." This is a fertile legal concept, with a definite lean toward protection. But it does not cover all human rights abuses, and the Convention's imprecise terminology leaves much room for dispute over its exact application.

In the 1990s, we have seen the flowering of debate over interpretation of these crucial, delphic words — meaty legal disputes over precisely how the standard applies to novel circumstances. How does it fit, for example, in a situation where central government has broken down, where the state simply fails? Does it cover objectors to compulsory population control measures? What about objectors to military service under corrupt or abusive regimes? How should it apply to restrictive forms of gender discrimination? Does it make a difference if the objectionable practice is deeply rooted in the culture or religion of the home society? Does the treaty's protection cover victims of spousal abuse when the home government provides no recourse?

Prominent cases in the courts and before the Board of Immigration Appeals are now wrestling with these issues, as are the courts and administrators (and indeed legislators) of most other western countries. But before we could even reach this stage of asylum debates, the countries involved had to solve other, more basic riddles. I will review those stages, because they help to put the achievements of the 1995 reforms in context.

At the risk of some oversimplification, I would identify three basic stages that countries go through in wrestling with the legal institution of political asylum. They are: 1) mastering the facts; 2) coping with numbers, which means deterring abuse while still giving security to bona fide refugees; and 3) refining interpretation — the stage reflected in the questions above. Nearly all countries go through them. Exactly when depends on the timing of the country's decision to treat these issues in a legal framework, rather than as ad hoc policy, and especially on when that legal structure finds itself challenged by mushrooming caseloads. Ireland, for example, though a party to the treaty since 1957, is still in the early stages. As recently as 1992 it received only 39 applications, but last year it met with over 7,500.


Graph: Asylum Applications and Grants, FY 1984-1997


Mastering the Facts

The drafters of the Convention did not think hard about the administrative difficulties the treaty standard portended, nor did the U.S. Senate in 1968 when it approved U.S. ratification of the treaty. The same is true of western European states and of most other countries as they became parties. Not surprisingly, then, the early implementing machinery was not well adapted to the potentially enormous challenges of the fact-finding task.

Asylum decisions may require the most difficult adjudication known to administrative law, owing both to the high stakes involved and the unique elusiveness of the facts. The risks for the individual are potentially higher than anything known in other U.S. adjudications, including most criminal trials. An erroneous denial could send the applicant back to torture, death, or slow starvation in a filthy cell. Viewed from the opposite perspective, the stakes are high in another sense. The outcome of a successful asylum claim — normally, permanent status in a country far more wealthy and stable than the country of origin — is so attractive that the system draws fraudulent claims by individuals. Worse, it gives rise to schemes by entrepreneurs who tempt migrants with cynical offers to deploy the magic power of asylum to gain permanent status, for a hefty fee.

Despite the weighty stakes involved, the basic facts in any particular case are highly elusive. The adjudicator has to decide what happened in a distant country, but usually has only two imperfect sources for judging what really has happened to the applicant or others similarly situated. First, general human rights reports on conditions in the country of origin can provide genuine assistance, but they rarely mention the applicant or the precise events at issue in the individual claim. The second source is live testimony, but quite often the only witness to the central personal events now available in the country of haven is the applicant himself or herself. That witness has incentives to exaggerate. But at the same time an effective asylum judge must always remember that true victims of past persecution can have real trouble providing a convincing account of what occurred. Numerous studies document the post-traumatic effects suffered by victims of torture or other severe abuse. Victims may find it difficult to speak at all about the critical historical events. Or they may render the account in a flat and matter-of-fact tone often mistaken for falsehood — because the listener can hardly believe that such grisly details could be recounted without more visible emotion. Finally, even when the adjudicator has puzzled out the historical facts, the job is not done. The ultimate decision requires a further prediction about future threats of persecution.

These problems were masked in most of the West in the first decades after the 1951 Convention was adopted. In the early years, asylum adjudication was essentially folded into the responsibilities of existing immigration functionaries or administrative courts. The caseload was small, most of it aligned with Cold War divisions, few were ultimately turned away, and the system attracted little attention. By the 1970s and especially the 1980s, when the number of applicants leaped and the origin points diversified, demands for better sorting of worthy from unworthy claims came to the fore, and fundamental problems could no longer be ignored.

The first challenge, then, was to develop a system that could effectively marshal the information needed for quality fact-finding. Europe generally moved well ahead of the United States in this process. In the 1970s and 1980s, while the INS still employed ordinary immigration examiners on rotation to decide the claims, many countries in Europe began assigning the task to specialist adjudicators with no other responsibilities. Specialization enabled them to acquire better interviewing skills for this especially sensitive caseload, and to develop certain revealing country-specific or even city-specific lines of factual inquiry. Moreover, it enabled the creation of information resource centers that pulled together detailed human rights information on countries of origin. (It had become clear that reliance on the applicant for that kind of broad country-condition information led to spotty results.) And one should also remember that human rights reporting itself was also in its infancy in those days. The State Department's first country reports on human rights appeared only in 1977 — reluctantly and under congressional orders — and for the first couple of years State produced a slender volume. NGOs were also somewhat late into the business of systematic reporting. Amnesty International, founded in 1961, established its comprehensive central documentation center in London only in 1975.

The United States finally joined Europe's structural trend in 1990, when the Department of Justice adopted hard-fought and long-delayed regulations that created a separate specialist asylum corps within the INS. While it was pursuing broad-based recruitment and training of new officers, and opening seven — later eight — new asylum offices around the country, INS also established a central Resource Information Center in Washington upon which those officers could draw. The result is a system now far better equipped to perform the crucial task of assessing the facts.

Deterring Abuse While Securing the Status of Genuine Refugees

The 1990 changes brought significant improvements in the quality of U.S. asylum decisions. But soon thereafter a second challenge became apparent. Despite the reforms, the overall decision process was still slow, cumbersome, and inadequately staffed. As had also happened in many European countries after comparable changes, the new office could never seem to get out ahead of the rising curve of new applications. Moreover, nearly all national systems proved to be ill-equipped to enforce negative decisions — that is, to deport those people whose asylum claims failed and who had no other defense to removal.

In a cycle that mirrored the experience of several other western countries, the U.S. asylum backlog mounted. Those in the backlog received work authorization and faced little risk of removal even if the claim were in fact reached and denied, itself a remote prospect. Delay thus fed on itself, as the process attracted more and more weak claims from people who sought work authorization. At the same time, the system was deeply dysfunctional for its intended beneficiaries, those who truly did flee persecution. For them an unending wait in a backlog, with no way to assure an early decision and the granting of a secure status, only prolonged their uncertainty and hindered family reunification.

One should not be surprised, I suppose, that the mounting numbers and the worries about fraud, more than concern for genuine claimants, provided the political impetus for the round of changes — both here and abroad. We in the United States sometimes think we had difficult years in 1994 and 1995 when the INS received some 140,000 to 150,000 new asylum claims. (This number does not include those that were heard only in immigration courts.) But Germany, with one-third the population of the United States, received over 430,000 applications in 1992. Faced with a political backlash, the main left and right parties resolved old differences and acted together to amend Germany's constitutional guarantee of asylum, thereby enabling several new measures thought likely to deter ill-founded claims. And nearby Switzerland received over 40,000 applications in 1991 and again in 1998, both times triggering legal reforms. That number may sound modest, but for Switzerland it is the equivalent of 1.5 million applications in a single year in the United States. Imagine how such numbers would affect our political climate.

In response to these second-stage challenges, many European countries adopted parallel new measures. For example, they disqualified persons who had come through "safe third countries," and they worked out readmission agreements with the chief transit countries so that the person could be returned quickly in order to pursue the claim there. Many central and eastern European countries emerging from Communism proved eager, in order to build links to European Union members, to adopt the 1951 Convention and its 1967 Protocol. They were also willing (though less eagerly) to accept readmission from west European target states. Because the former countries were considered less attractive politically and economically, forcing applicants to seek asylum there was deemed a deterrent to weak claims.

Several European states also adopted formal lists of "safe countries of origin." Applicants from those countries were not per se disqualified, but faced a strong presumption that the claim should be denied. Many west European democracies also adopted special fast-track procedures at airports (and sometimes internally) to weed out "manifestly unfounded" applications. And finally, when all these measures proved insufficient in the face of the crisis in ex-Yugoslavia, Europe departed from its tradition of offering durable status to those who proved their refugee claims, instead tendering only temporary protection until the wars subsided.

The U.S. Response to the Second Stage

The United States managed to avoid some of the more problematic measures used in Europe when events combined in 1993-94 to give us the political will to confront these second-stage challenges. We have no "safe country of origin" lists, instead considering each case on its individual facts, without the use of broad presumptions. The 1995 reforms also placed little reliance on "safe third country" provisions, although they did make room for discretionary denial of asylum under more circumscribed conditions than is the European norm. And we still provide full permanent resident status routinely to persons who prove their refugee claims under the standards of the 1951 Convention, rather than pulling back to offer them only temporary shelter.

Instead, the U.S. reform effort focused on other steps to deter abuses and ultimately bring down the backlog. Most important of all was the provision of adequate resources, through an eventual doubling of the asylum officer corps and the ranks of the immigration judges. This augmentation assures that the full initial procedure, often including two rounds of consideration of the asylum claim, is nearly always completed within 180 days from filing.

Beyond calling for more staff, the team that worked on what became the 1995 reforms went carefully over all steps in the existing asylum process, to identify and eliminate inefficiency and duplication. Written denial letters from asylum officers were replaced with a short checklist and a quick referral to the immigration court, if the officer was not persuaded to grant asylum. The referral makes use of the initial application form already filed, instead of waiting while a new application is completed for the court proceedings — as had been the former practice. The new process also requires applicants to come back approximately two weeks after the interview, in order to pick up the decision in person. They have a strong incentive to appear, because the outcome might indeed be favorable. But if the decision is not a grant, this procedure assures effective service of a charging document that initiates deportation proceedings, in a form that would support a removal order, even if the applicant fails to appear in immigration court. (No-shows had been a problem under the old system.) The basic message is this: we are serious about protection if you meet the refugee standard. But we are equally serious about deportation if you do not.

Finally, in order to break the psychological connection, exploited by unscrupulous operators, between work authorization and mere applications for asylum, the 1995 reforms delay work authorization until asylum is granted, or until 180 days pass without a final decision in immigration court, whichever comes first. (The new resources have meant that few cases stretch beyond that timetable.) If the immigration court has rejected the claim in a timely manner, an appeal to the BIA is available, but without work authorization, no matter how long the appeal takes.

The other side of the coin, the protection side, was equally important. The 1995 reforms carefully preserve a fully expert engagement with the merits of each asylum claim; that is, they do not cut back on the gains of the first stage of reform. Weak claims are deterred, but strong claims still lead to the security of asylum status and eventually to permanent residence. Importantly, genuine claimants now can be assured of an interview within six weeks of filing, in a nonadversarial office setting. And unlike the old system where they might stay in the limbo of the applicant stage for years, strong claims should generally result in recommended grants within 60 days of filing, with work authorization to follow close behind.

These regulatory changes had compiled just enough of a successful track record by 1996 to head off most of the damaging restrictions on political asylum initially proposed for the Illegal Immigration Reform and Immigrant Responsibility Act, enacted that September. When the legislative dust settled, Congress wound up essentially incorporating the 1995 regulations into statutory law. But not all worrisome restrictions were avoided. I regret the addition of the one-year deadline on filing for asylum. Delayed application may properly be taken into account in assessing the credibility of a claim, but should not wholly bar it. The assumption that true refugees invariably claim protection just after crossing the border — though widely voiced in 1996 — is ill-informed. Many studies document the complexities of personal choice when people respond to a threat of persecution, and they recount many victims' natural reluctance to cut ties decisively with the home country, even one where they have suffered. The assumption also overlooks the difficulties that some traumatized victims have in placing future trust in any government officials, even after crossing a border.

The 1995 reforms achieved measurable success in deterring abuse and reassuring control. New asylum applications received by the INS declined from 150,000 in 1994 to 35,000 in 1999. And while the number of applications is decidedly down, the percentage of claims granted has risen — a further sign of system health. The decline in applications was doubtless aided by the settlement of many of the violent conflicts in Central America, and there is no guarantee against new conflicts or new dictatorships in our hemisphere, of course. But even if events bring us a higher intake or harder cases in the future, we have a good structure in place to meet any new challenges. Such an increase, if it comes, should require mainly resource adjustments, not system redesign.

The Future

The United States has come far in mastering the challenges of the first stage, fact marshalling and quality fact-finding, and I do not foresee any major new initiatives in that realm. With regard to the second stage, the 1995 reforms have gone a long way toward deterring abuse — a tribute to the hard work of the asylum officers and immigration judges who staff the front lines of decisionmaking and assure that decisions remain both timely and well-informed. But it is also a tribute to the budget officers, regulation writers, appropriators, lawyers on both the government and applicant sides, and countless others whose work we justly celebrate here today, because all those skills, and all their on-going vigilance, have been necessary to the reform process and its continued vitality.

But we have not quite overcome all the challenges posed by the second stage. I see three major problem areas, which may initially seem separate but whose solutions are, I believe, closely related. The first is the issue of work authorization. Delaying work authorization by as much as 180 days was thoroughly necessary in 1995, to break the link that then attracted and to defeat the operators who encouraged boilerplate filings or other kinds of manipulation. But that psychological decoupling has now occurred, and we should take stock. Since our system makes no other pro-vision for aid to applicants awaiting a decision, the conscientious asylum seeker is reduced to living off the charity of friends, family, or assistance organizations, for as much as six months. The time may have come when the United States can provide time-limited work authorization early in the asylum application process and rely on other features of the reformed system to deter weak claims. The most important such alternative feature is the close tie we have now created and publicized between affirmative filings and the start of deportation proceedings if the claim fails. The system's better fact-finding capacity is less likely to let a bad claim through, and the system is now geared toward prompt issuance of a deportation order if asylum and all other defenses fail.

But that insight brings up the second problem area. An early deportation order is not necessarily much of a deterrent if it rarely results in actual deportation. Through much of the 1990s, actual removal of nondetained aliens ordered deported has been quite low, as low as 10 percent. For this reason, the working papers leading up to asylum reform emphasized the need to increase removals of those whose claims are not successful (and who have no other basis for a stay in the United States). Some genuine progress has been made in increasing removals in recent years, but the percentages are still low. In fact, frustration with these statistics goes a long way toward explaining why Congress embraced crude detention mandates so strongly in the 1996 legislation — on the belief that only start-to-finish detention results in actual removal at the end of the process. I think that belief is mistaken. There are other ways to improve the incidence of actual removals, but they require great tenacity to implement and maintain. The Vera Institute of Justice has been running a pilot project for the INS to test some of these methods, through supervised release of nondetained respondents in immigration proceedings. The project has shown promising results that should both save detention money and result in a better removal record, while also minimizing hardship on those who prove to have valid claims. But the endeavor has been hampered by resistance in some INS quarters, and wide application of its lessons appears a distant prospect.

In any event, the removal problem does not belong solely to asylum. It is an issue that touches on nearly all aspects of INS enforcement. But I suspect that we will not be able to ease up on work authorization restrictions for asylum applicants or on the wide use of early detention (the third problem area) until we show better results in enforcing deportation orders. I would hope to see more immigration advocacy groups engage on this issue. Counterintuitive though it may seem, they too have a considerable stake in improving removal statistics and thereby providing a key reassurance to those in the electorate and in Congress who are in the grips of the control impulse, in order to improve the climate for easing of other restrictions. And there should be no hesitation in placing an emphasis on this subset of our illegal migrant population. These are people who have been through the full procedures and have had opportunities at several levels to present their claims and defenses. In short, refinements regarding work authorization, detention, and removal are still needed to complete the work of the second stage.

And then the decks will be cleared for a focus on the third stage, the interpretive stage applying the "well- founded fear" standard to the challenging new legal questions now finding their way into the courts and the BIA. This task is equally demanding, for it essentially asks us to decide what abuses can successfully be addressed through asylum and what must be left to human rights and other global initiatives.

David A. Martin is Doherty Professor Law, University of Virginia. This essay is an edited version of remarks prepared for the Department of Justice conference in observance of the fifth anniversary of asylum reform, held in Washington, D.C., on February 1, 2000. The reforms were adopted in regulations promulgated at 59 Fed. Reg. 62,284 (1994), taking effect January 5, 1995. For a more detailed treatment of the history and the regulations, see David A. Martin, "Making Asylum Policy: The 1994 Reforms," 70 Wash. L. Rev. 725 (1995).

Martin served as a consultant to the Immigration and Naturalization Service (INS) in1993 during a key stage of the reform's development, and served from 1995 through 1997 as INS General Counsel. The remarks here convey his personal views and do not necessarily reflect the position of the INS or the Justice Department.