International Law and Administrative Insulation: A Comparison of Refugee Status Determination Regimes in the United States, Canada, and Australia


Rebecca Hamlin ( is an Assistant Professor of Political Science at Grinnell College. She has a PhD from the University of California, Berkeley. I thank the Canadian Embassy in Washington, DC, as well as the Pacific Rim Research Program and Institute for International Studies at Berkeley, for generously funding portions of my research. I am grateful to Irene Bloemraad, Ming Hsu Chen, Ken Haig, Robert Kagan, Leila Kawar, Gordon Silverstein, and Philip Wolgin, as well as three excellent and insightful anonymous reviewers, for their feedback at various stages. I also thank Elisabeth Rennick for her valuable research assistance, and my interviewees for their time and candor.


International law provides nations with a common definition of a refugee, yet the processes by which countries determine who should be granted refugee status look strikingly different, even across nations with many institutional, cultural, geographical, and political similarities. This article compares the refugee status determination regimes of three popular asylum seeker destinations—the United States, Canada, and Australia. Despite these nations' similar border control policies, asylum seekers crossing their borders access three very different systems. These differences have less to do with political debates over admission and border control policy than with the level of insulation the administrative decision-making agency enjoys from political interference and judicial review. Bureaucratic justice is conceptualized and organized differently in different states, and so states vary in how they draw the line between refugee and nonrefugee.


International law provides nations with a common definition of a refugee, and with guidance on how asylum seekers asking for refugee protection should be treated. In this way, the process of identifying people who are in need of protection from persecution—a process known as refugee status determination, or RSD—is a universal concept linked to an international legal category. However, RSD can differ greatly depending on where it takes place, even among states with many institutional, cultural, geographical, and political similarities. This article compares the refugee status determination regimes of three popular asylum seeker destinations—the United States, Canada, and Australia—in order to isolate which factors explain the striking cross-national variation in RSD outcomes.

These cross-national differences are perhaps best illustrated by a direct comparison of the acceptance rates of asylum claims made by people originating from the same place. The People's Republic of China has long been a dominant source country of asylum seekers in the United States, Canada, and Australia,1 but the acceptance rates in the three countries run on three remarkably distinct tracks (see Figure 1). Canada has always been more generous to Chinese claims, and has also become more generous over time. In the United States, the variation over time is striking, and stands in contrast to the other two countries, whose rates have been more stable. In Australia, the acceptance rate has remained steadily low until very recently, when it began to climb slowly (Amnesty International 2000; Jacobs 2009).2 In 2009, Australia accepted 14 percent of Chinese applicants, whereas the United States accepted Chinese claims more than twice as frequently, with an acceptance rate of 35 percent. Even more puzzling is the fact that Canada's acceptance rate for Chinese applicants was 58 percent, a proportion greater than the other two countries added together.3

Figure 1.

Asylum Grant Rates for Chinese Claims in Three Countries, 1997–2009

All three states have adopted the UN treaties outlining a refugee definition, and thus RSD decision makers in all three countries are interpreting an identical text in determining the validity of asylum claims. So, why do they frequently come to different conclusions, even in similar cases? Why are so many more Chinese asylum seekers categorized as refugees in Canada? Or, to put it another way, why are so few given refugee status in Australia?4 Why are rates in Canada and Australia more stable, while they fluctuate sharply from year to year in the United States?

In this article, I argue that the relative generosity and stability of refugee status determination rates can be explained by the different RSD regimes that have developed in each country. I view an RSD regime as both the set of institutions that are responsible for conducting RSD and the relationships and power dynamics between those institutions. The concept of a regime highlights the fact that RSD outcomes are part of a larger system that must be studied holistically, rather than by comparing isolated elements, such as parallel exclusionary policies or high court decisions on similar questions. The regime approach helps us understand why some issues are noncontroversial in one country, yet become contentious jurisprudential conundrums in another.

My comparison of RSD regimes in the United States, Canada, and Australia suggests that the most important variable in explaining varied RSD outcomes is the degree of insulation the administrative agencies conducting front-line RSD decision making enjoy. Structural differences across RSD regimes allow for differing levels of insulation from the exclusionary politics of deterrence, which in turn affects the role of courts in the regime. When administrative insulation is low, courts engage in turf wars as they try to limit legislative and executive interference. Thus, RSD outcomes depend on the institutional players that end up dominating the process, the level of contention between actors, and the degree of centralization within the decision-making processes.

A systemic institutional approach identifies additional influences on decision makers beyond international norms or migration politics. Each state determines its own RSD process not according to international norms, but by domestic patterns of administrative and bureaucratic justice. Conceptions of due process, administrative fairness, and judicial review are debated and resolved differently in each country, and so key players often act according to the institutional identity created by the constitutional and administrative framework of each country, or according to their place in a larger interbranch conflict. Ultimately, the process of determining how RSD should work has an impact on the separation of powers and the shape of procedural justice more generally.

After outlining a comparison of the three RSD regimes, I use the empirical case of Chinese asylum seekers (and those who claim to be victims of China's coercive population control policies in particular) as a lens for examining how differences in the way that RSD regimes handle difficult legal and procedural questions matter for refugee status determination outcomes. The case study of Chinese coercive population control demonstrates that RSD regimes do not just matter at the margins; they can vary in ways that affect vast numbers of asylum seekers.


This article aims to link the conversations of comparative constitutional law, comparative administrative law, and migration policy studies. While political science has long embraced comparison of every other aspect of political life, the field of comparative constitutional law is only now beginning to flourish. Traditionally, public law scholars interested in constitutions have assumed that legal institutions reflect the distinct culture and history of each nation (Alford 1986; Schauer 1993). This particularist approach downplays the existence of cultural commonalities across jurisdictions, and ignores the fact that many legal institutions are transplanted into new countries through colonialism and globalization. By suggesting that there is nothing universal about the way law or courts work, this approach has had the unfortunate effect of discouraging scholars from engaging in comparison.

The more internationally oriented strain of public law scholarship concerned with comparative administrative law draws a distinction between adversarial and inquisitorial styles of decision-making procedures, a conceptual dichotomy dating back to medieval times (Damaska 1986, 3). The adversarial style takes the shape of a triad: two disputants arguing their respective cases before a passive judge, who must resolve the dispute by deciding which case is more persuasive (Shapiro 1981; Stone Sweet 1999). RSD procedures are adversarial if a government prosecutor is used to challenge asylum claims, and a judge decides between the competing positions. In contrast, the inquisitorial form requires much more active decision makers, combining the role of investigator and judge into one. The actors form a dyad between the person whose fate is to be decided and the person deciding it. RSD is inquisitorial if the asylum seeker goes before a decision maker who both researches and decides the claim.

The literature on comparative administrative decision making reminds us that “different conceptions of procedural fairness are associated with different types of decision making” (Adler 2003, 335). However, this field has not used asylum seeking as a case study, despite the fact that RSD is one of the most frequent forms of administrative decision making that an advanced democracy must conduct. Administrative agencies are the main point of contact between asylum seekers and the state, the central component of the domestic implementation of international refugee law. They are also precedent setters, laying the groundwork for the interpretation of asylum law before the courts see it in particular appeals. Yet there has been limited academic study of how administrative RSD processes work, how they differ from one another, and how they interact with other players in the policy arena.

Migration scholarship tends to focus on admissions policies or major developments in judicial interpretation of the law. Many comparative migration scholars point to the existence of a common international legal category, and the widespread practice of RSD as evidence of a globalization of law. These scholars argue that the proliferation of international human rights has shifted the locus of power away from the state in defining the boundaries of membership, giving otherwise vulnerable people a stake for their rights claims (Soysal 1994; Jacobson 1996; Sassen 1996; Guiraudon 2000; Spiro 2007). Jacobson argues that immigrants are increasingly “becoming the object of international law and institutions” (1996, 10), suggesting that international law enables noncitizens to assert rights that they would not otherwise have.

This line of scholarship implies that international human rights norms can trump the exclusionary tendencies of states and make it difficult for courts to distinguish between citizens and noncitizens when protecting rights. This proposition of international convergence helps explain why there are 147 countries that have adopted international refugee protection treaties and have pledged to consider the claims of asylum seekers who arrive uninvited at their borders. But a focus on international convergence ignores many state practices that run counter to this trend, and paints an overly rosy picture of the globalization of law.

In contrast to the international convergence school, a line of more pessimistic scholarship has recently emerged, pointing to a different kind of cross-national merger: exclusionary convergence. These scholars observe “broadly similar patterns in the response of these receiving countries to asylum seekers and refugees over the last decade” and the development of “a similar legal doctrine, namely the sovereign right to exclude” (Kneebone 2009, 281, 292; see also Gibney 2006). Pellerin suggests that this “global agenda on migration” is the result of a “new consensus among industrialized states” reacting against internationalism (2008, 190).

Exclusionary convergence theorists are correct in noting that since the end of the Cold War, the asylum policies of the United States, Canada, and Australia (among others) have converged on a politics of deterrence. States have become more focused on keeping asylum seekers out with more rigorous border control measures, interdiction at sea, visa requirements, multinational collaboration, and widespread use of detention. In all three countries, asylum policy making is dominated by an awareness of the cost of conducting RSD and concerns about border security. This shift explains why it is increasingly difficult for asylum seekers to access the RSD programs of the United States, Canada, and Australia. However, it does not explain the divergence of outcomes when they do.

Both types of convergence theory are concerned with the degree to which states are embracing the tenets of international law, but they come to opposite (contradictory) conclusions about how much international protection norms are influencing domestic politics. Convergence theories also locate immigrant rights outside the state by assuming that when noncitizen rights are protected, it is despite opposition from the exclusionary state and that these rights have no domestic source.

Other migration scholars make the argument that both exclusionary and inclusive forces are at play within states. Joppke suggests that “conflicts over asylum policy are . . . domestic conflicts over the dual mandate of liberal nation-states” (1998, 110; see also Schuck 1998). States balance majoritarian, exclusionary forces against internal pressures to protect the rights of vulnerable minorities in what Hollifield (2004) calls “the liberal paradox.” Such scholarship gives us a sense that nations might strike this balance differently from one another, positing a theory of domestic divergence that stands in contrast to both varieties of convergence proponents.

When domestic divergence scholars account for cross-national differences, their explanations tend to be judicially focused, based on the notion that if states have become reluctant asylum seeker hosts, enacting policies that reflect this reluctance, cross-national differences must be explained by looking beyond legislative politics. Joppke, for example, suggests that the US Constitution has made it more difficult for courts in the United States to ignore violations of noncitizen rights than their German and British counterparts (1998, 140–41). Soennecken (2008) suggests that NGOs in Germany and Canada have had varying levels of success in assisting refugees in court because the two countries have very different traditional levels of access to the judiciary. It makes sense that domestic divergence scholars would choose to examine nonlegislative institutional players. However, by focusing on one institutional player, comparative studies run the risk of assuming that courts play an equivalent role in the overall RSD regime of each state.

The debates between international/exclusionary convergence and domestic divergence have dominated scholarly inquiry about migration in a globalizing world. They set up an image of two competing forces—the centrifugal force of exclusion and the centripetal force of inclusion—as a political battle. However, if we look closely at the dynamics of RSD regimes, we see that there are many other forces in play besides the politics of migration. Instructive comparisons can be drawn only by comparing the overall political and institutional context of the three countries: constitutional, administrative, and legislative.


The United States, Canada, and Australia are fascinating cases for a comparative study of RSD regimes because they have so many important commonalities. These three countries are naturally and frequently grouped together by immigration scholars as “classic countries of immigration” because of their history of population growth via large-scale immigration and their national identities as migrant destinations (Freeman 1995; Joppke 1998; Cornelius et al. 2004). All three countries have extended protection to displaced people via refugee resettlement, particularly in the post World War II and post Vietnam War eras. All three countries have federal systems in which immigration policy is dictated at the level of the national government. Though Australia and Canada are parliamentary systems and the United States has a president, all three countries have large immigration bureaucracies, and all three allow courts to review decisions made by those agencies.

The United States, Canada, and Australia are also extremely popular asylum seeker destinations that conduct RSD on a large scale. The United States consistently receives more asylum applications than any other country; in 2009, it received 13 percent of the total number of asylum applications lodged worldwide. Canada was the third most common receiving country, with 9 percent of overall applications. Australia was sixteenth with 2 percent, despite its inaccessible location (UNHCR 2010). Canada receives by far the most asylum seekers per capita; between 2003 and 2009 it had one asylum application per every 168 current residents. The ratios for the same period were one asylum application per every 813 Americans and every 730 Australians respectively.5

Another important similarity between the three states lies in their common legal traditions and strong courts. All three states drew heavily on British common law when designing their legal systems. Australia (in 1900) and, to a lesser extent, Canada (in 1867) drew on the US Constitution of 1787 as a guide when drawing up their own, and courts in all three jurisdictions use similar common law logics today. Until after World War II, the United States, Canada, and Australia were the only three states in the world to have well-established judicial review, and so their courts share a uniquely high level of independent power (Shapiro and Stone Sweet 2002). More recently, all three court systems have achieved “increasing constitutional independence,” with more attention to the “progressive elaboration of rights” for citizens than in the past (Scheppele 2003, 14).

Because the United States, Canada, and Australia have so much in common, I consider these states to be “most similar systems” for the purposes of comparison. By holding many common domestic factors reasonably constant, I can isolate the factors that lead to divergent outcomes (Mill [1843] 2002; Przeworski and Teune 1970). Despite all the similarities mentioned above, and the fact that RSD is conducted in each country based on the same international standard, the RSD regimes in each state look remarkably different.

To build a systemic comparison of RSD regimes, I used a combination of qualitative research methods. Drawing on a year of fieldwork, I used in-depth elite interviews, courtroom observation, and case analysis to demonstrate why RSD regimes differ, and how those cross-national differences matter. To build a historical narrative of the development of each regime, I also created a media database of asylum policy coverage by selecting two major broadsheets from each country and searching their archives for the years 1980–2010. Any article that was primarily devoted to a discussion of asylum seeker policy was included in the database and analyzed further for a total of 2,260 articles.6 Not only was this collection of news accounts extremely helpful in piecing together the complex timelines of asylum policy making in each country, it was essential for identifying interview subjects and significant court cases.

I conducted 103 in-person interviews (thirty in the United States, thirty-five in Canada, and thirty-eight in Australia).7 These interviews fell into three categories. First, in each capital city, I met with policy elites, including staff at the relevant agencies: Department of Homeland Security, Department of Justice (both the Executive Office of Immigration Review and the Office of Immigration Litigation), and Department of State in the United States; Citizenship and Immigration Canada, the Immigration and Refugee Board, and the Department of Justice in Canada; and the Department of Migration and Citizenship, the Refugee Review Tribunal, and the Federal Court of Australia. The conversations focused on perceived challenges and successes in implementation and relations with the advocacy community and with other actors responsible for asylum policy implementation. I also interviewed representatives of the UN High Commissioner for Refugees in each capital to discuss coordination with government, and how their offices handle instances of noncompliance with international law.

Second, I interviewed key refugee advocates from organizations that the media had mentioned as having participated in lobbying efforts on asylum policy during the past ten years and that list asylum as an organizational focus. I interviewed a senior staff person at almost all these organizations, and used these contacts to expand my list via snowball sampling—asking each contact if there were any significant organizations I had missed. My conversations with policy elites focused on their biggest priorities and challenges, relations with government, and strategies for achieving reform. Many policy elites had been working on this issue for decades, and provided me with useful historical perspective and contemporary analysis of their policy environments. I collected literature from each advocate I interviewed and each government agency, many of which conduct their own research on asylum seeker policy. While these documents have a distinct political slant, they are useful guides to the most pressing debates within the policy area.

Third, I interviewed a sample of people who represent asylum seekers in RSD hearings and were usually, but not exclusively, lawyers. Since there are hundreds of people who represent refugees in each nation, I focused on one major refugee destination city per country: Toronto, San Francisco, and Sydney. I chose these cities because the majority of asylum hearings in Canada are conducted in Toronto and the vast majority of Australian hearings are conducted in Sydney. Hearings in the United States are distributed more evenly over a broader range of cities than in the other two countries, but San Francisco is one of the top four cities where asylum claims are heard in the United States. In each city, I took a random sample of representatives from a list I compiled using telephone books and Internet searches. I asked them about their experiences with the RSD regime, the advice they give their clients, the tactics they use to gain favorable decisions, and the concerns they have about the system as a whole.8 I conducted fifty interviews in this category (fifteen to twenty in each country).

Another key element of my fieldwork was observation of RSD hearings. In each country, I visited the administrative agencies responsible for RSD, and I collected statistics, training materials, and operational regulations from the agencies. I observed multiple hearings at the San Francisco immigration court, and several appeals at the Ninth Circuit Federal Court of Appeal. In Toronto, I observed hearings at the Immigration and Refugee Board, and appeals at the Federal Court of Canada. In Australia, all hearings at the Refugee Review Tribunal are closed, but I observed multiple first-level appeals of these hearings at the Federal Magistrates Court and further appeals at the Federal Court of Australia in Sydney.

To supplement my ethnographic data, I collected descriptive statistics from multiple sources to compare basic numbers across countries and over time. First, I used the UN High Commission on Refugees (UNHCR) statistical yearbooks from 1980–2009 for information on the numbers of applications for asylum and acceptance rates reported by each nation. Second, I collected data from the federal courts on applications for judicial review and success rates. Third, I collected data on acceptance rates, reversal rates, and rate of legal representation from the administrative agencies responsible for RSD in each country. Many of these data are not publicly available, but were provided to me after visiting the agencies personally and making a special request.

The final element of my research consisted of a detailed content analysis of 137 significant administrative, federal court, and high court decisions regarding asylum law.9 I selected these cases based on mentions in the media database, interviewee responses about which cases they viewed as important, and a survey of law review articles using the terms “refugee/asylum,”“administrative,” and “procedure.” This three-pronged approach was extremely helpful in identifying the most significant cases for defining the roles of each institutional player and setting the parameters of procedural justice for asylum seekers. In analyzing the case, I looked for patterns in case law development across the three states, and tracked the frequency and duration of unsettled legal questions in each place. I took particular note of instances of cross-fertilization (reference to an equivalent decision in another country), and the level of deference courts in each state afforded to the administrative agencies. The case analysis revealed a common theme across the three countries. Courts in all three jurisdictions have dealt to some degree with cases involving forced sterilization claims from the People's Republic of China, albeit in very different ways. Thus, the case analysis was particularly informative to my discussion about these types of claims at the end of this article.


In the United States, controversies about the poor quality of administrative decision making have led to serious tensions between the executive and judicial branches. For example, on December 26, 2005, The New York Times ran a story about the RSD regime that included some notably fervent public bickering. The article was about a Chinese asylum seeker named Qun Wang, but the larger dispute was over the relative ability of different agencies to make a high-quality refugee status determination (Liptak 2005). Mr. Wang had applied for asylum for himself and his wife based on his claim that she had been sterilized against her will. They both feared retribution from the Chinese government for fleeing if they were forced to return to their home country. The controversy arose over the handling of Mr. Wang's claim at the administrative level. His application had been denied by an immigration judge (IJ), who stated in her decision that his sexist comments during the RSD hearing had led her to conclude (in something of a non sequitur) that “he's a horrible father as far as the court's concerned.” The federal judge who reviewed the case was outraged by the personal bias displayed in the administrative decision, and sent the case back for a rehearing, complaining in the opinion that “the tone, the tenor, the disparagement, and the sarcasm of the IJ seem more appropriate to a court television show.”

The president of the National Association of Immigration Judges defended her colleague to The New York Times by responding that “the vast majority of IJ's do an excellent job given such a large caseload. . . . To go name-calling and having an open season on [administrative] judges, it's crossing the line of civility.” However, the article went on to quote the Chief Judge from the US Second Circuit Court of Appeals, asserting that “we're the first meaningful review that the petitioner has.” In other words, according to some powerful members of the US judiciary, administrative decision making is meaningless in its lack of quality.

As this incident suggests, a person who applies for asylum in the United States enters a terrain that is fraught with turf wars and interbranch conflict. The contemporary US RSD regime is permeated by a domestic institutional politics of wrangling for control of the process that brings the courts into the center of the issue and is reliant on lawyers to make a good case. This structure makes the US RSD regime the paradigm of an adversarial, legalistic decision-making process in four key ways emphasized by Kagan (2001). First, decision making is fragmented across multiple government agencies. Second, asylum policy in the United States is highly participatory and allows for multiple perspectives to be presented and weighed, which can serve as a powerful check against bias or error. Third, decisions at all but one of the key stages in the process are made in a typical adversarial style involving a lot of rules and procedures that constrain decision makers. Fourth, advocates often criticize decisions for being unpredictable, inconsistent, and costly, three pejorative hallmarks of adversarial legalism.

There are two distinct points of entry to the US RSD regime (see Figure 2). The first is known as an affirmative asylum application, which is begun when a person proactively files a claim within one year of entering the country, regardless of whether that person initially entered illegally or legally. Once the form is filed, the applicant will be asked to appear for an RSD hearing at the Asylum Office, which now falls within the Department of Homeland Security. The hearing is inquisitorial and takes place in a small hearing room with a bureaucrat who is trained in international and domestic asylum law.

Figure 2.

The RSD Process in the United States

Only about half of asylum applicants have access to the Asylum Corps decision makers for their initial decision. The other half of applicants is funneled directly into the highly adversarial and legalistic administrative immigration court system called the Executive Office of Immigration Review (EOIR), which is housed within the Department of Justice. These hearings are known as defensive asylum hearings because they are defending against deportation, and asylum seekers are tracked into them in three ways. First, they may be found by Immigration and Customs Enforcement at a port of entry to be without valid travel documents. Second, they may be found by immigration officials to be illegally in the United States, and placed in a removal proceeding. If anyone in these two groups of people expresses a desire to seek asylum, he or she is sent to a defensive hearing. The third path to the EOIR is via the Asylum Office. All applications that are not accepted at that stage are automatically referred to a defensive hearing.

Defensive hearings are held in a courtroom setting before an administrative IJ. A lawyer represents the Department of Homeland Security by arguing that the applicant does not qualify for asylum and should be deported. The applicant is encouraged to represent himself or herself with counsel, though the government does not provide one because deportation hearings are civil proceedings, not criminal trials. The burden of proof in a defensive hearing is on the asylum seeker and he or she may be cross-examined by the government attorney about the details of the claim.

In interviews, advocates complained that IJs, who are political appointees, differ widely in their approach. For example, a lawyer complained: “It's a lottery draw of which judge you get at EOIR.”10 Another lawyer suggested that there are some IJs he does not even bother trying with, explaining: “You can find out who the judge is ahead of time, and this absolutely influences our decision of whether or not to even take the case, and how to prepare for it.”11 These charges found support in a 2008 study of grant rates across IJs, which found several instances of different judges in the same court with grant rates ranging from under 5 percent to over 80 percent, even when examining cases from the same country (Ramji-Nogales, Schrag, and Schoenholtz 2007).12 Presidents can appoint IJs with backgrounds that make them more likely to be sympathetic or tough on RSD applicants, but IJs from previous administrations remain.

If an applicant is unsuccessful at EOIR, he or she will be deported unless the applicant files an appeal with the central headquarters of the office, called the Board of Immigration Appeals (BIA). The board, located in Falls Church, VA (a suburb of Washington, DC), does not hold in-person hearings, but conducts on-paper reviews of the decisions of IJs. In 2002, under pressure from the Bush administration, Attorney General John Ashcroft initiated procedures that were designed to reduce the immense backlog of cases at the BIA. These “streamlining” reforms included a policy of “deference” to the conclusions drawn by IJs, an end to independent BIA fact finding, and the use of “Affirmances without Opinion” (AWO), which uphold the decision of the IJ with no elaboration of the legal issues involved. The reforms also included a reduction of the board from fifteen to eleven members, a change designed to increase internal consistency and “collegiality” by removing Clinton-appointed members.13 These reforms quickly resulted in a dramatic rise in the number of cases that the Board of Immigration Appeals was able to process in a year.14 Since then, the BIA has continued to move through its caseload very quickly, even as the number of cases processed at the EOIR has remained steady.

The BIA has an even worse reputation than the immigration courts. Judge Posner has called the BIA's work “dismal” and “perfunctory” (Posner 1996, 265). Lawyers said that they expect the decision to be upheld at that review stage, and view the BIA as “a pit stop on the way to” federal court for judicial review.15 One called the BIA “a rubber stamp” and said that the AWOs are “a copout; we used to get real judicial review there, but now the appeals we send are just sent back within a matter of months.”16 Another said that “every case gets appealed [to federal court], even if it's the weakest, because we know the BIA doesn't grant anything.”17 According to the American Bar Association (2003), these accusations have some truth to them: the ABA found that before the 2002 reforms, the BIA decided in favor of the asylum applicant 25 percent of the time; afterward, the rate was 10 percent. Beyond that stage, the last remaining recourse is to apply for judicial review in whichever federal circuit the initial asylum claim was filed.

A doctrine of administrative deference and the history of congressional plenary power over migration matters have historically limited the role of courts in US immigration law. However, courts today are reviewing administrative immigration decisions with increasing frequency. This shift is partly due to a change in US administrative law more generally. As the administrative state expanded, the number of people affected by administrative action grew, as did the range of people eligible for judicial review (Stewart 2003). This expansion of due process rights to include many new groups has been called a due process revolution because it swept through the courts and forced judges to decide how best to assess the actions of administrative agencies (Mashaw 1985). The legal rule governing this issue is that when a statute is ambiguous, the agency interpretation should be presumed to be correct because administrative decision makers are specialists. However, courts are allowed to strike down agency interpretations that they deem unreasonable.18

The judicialization of US immigration policy extends beyond these shifts in administrative law and has had an undeniable impact on the US judiciary, as cases dealing with the rights of noncitizens have become a huge proportion of the work of the federal courts. Appeals from administrative agencies are almost 20 percent of the caseload of the federal circuit courts of appeals, and immigration cases have risen to a startling 90 percent of all administrative appeals.19 This phenomenon began in the 1980s, but has continued since the 1990s on a different scale (Posner 1996; Law 2010). In 2009, the federal courts of appeals received only 5,749 more total cases than in 1996. Over the same period, cases originating from the Board of Immigration Appeals went from 1,063 in 1996 (2 percent of the total federal court of appeals caseload) to 7,518 in 2009, or 13 percent of the total. Thus, while other areas of litigation declined, including criminal cases and bankruptcies, immigration appeals accounted for most (if not all) of the rise in caseload over this fourteen-year period.20

The immigration litigation explosion has not been evenly distributed across the federal circuit courts of appeal. In 2007, the Ninth Circuit (which includes California and Arizona) received 47 percent of the total immigration caseload, and the Second Circuit (which includes New York) received 24 percent.21 Federal circuit courts do not keep statistics on the specific issues raised in cases before them, and could not provide hard numbers on the proportion of the appeals from the BIA each year that are RSD matters.22 However, since applications for asylum are an exception to the jurisdiction-stripping legislation passed by Congress in 1996, asylum cases likely make up a disproportionate set of the immigration cases under federal court review.23 In fact, in his 2006 testimony to Congress about the rise in immigration litigation, Chief Judge Walker of the Second Circuit claimed that over 90 percent of their immigration cases “raise asylum issues.”24

Judicialization has led to some dramatic turf wars between the administrative agencies and the federal courts in the circuits where the bulk of appeals are brought. This tension has led the Supreme Court to police the dispute by disciplining the lower courts on several occasions. In 2002, the Supreme Court issued a brief per curium (summary) ruling in INS v. Ventura, in which it reprimanded the Ninth Circuit for granting asylum to an applicant for a reason that the BIA had not considered, instead of returning the case to the BIA for rehearing. In his petition to the Supreme Court, the Solicitor General cited at least eight other similar cases, calling this kind of power grab by circuit courts “a recurring error.” When the same issue reached the Supreme Court again in Gonzales v. Thomas (2006), it stated sternly that “the Ninth Circuit's failure to remand is legally erroneous, and that error is obvious in light of Ventura, itself a summary reversal.” This deliberate contravening of Supreme Court decisions is evidence of the level of disdain with which some in the federal courts view the BIA. As a senior BIA staff member told me in an interview: “Some circuits, like the Ninth and Second, don't actually see the BIA as experts.”25

Ironically, the congressional response to judicial criticism of the administrative RSD process was to pass the REAL ID Act of 2005, making the process even more reliant on lawyers. The REAL ID Act raised the burden of proof for asylum and increased the requirements for verifying identity and providing documentary evidence of persecution. In interviews, lawyers reported on the challenges that this new standard has created in proving certain claims of persecution. One told me: “[T]here has been increased rigidification. In the past, we could win on testimony. Now we have to get documents that may not exist.”26 Another said that REAL ID had noticeably “tightened the screws” on the asylum process.27 A third said that “the expectation for supporting documentation has gone up, and there is less reliance on testimony.”28 These changes do not seem to have reduced the number of cases that are appealed to the federal court system, but they make it much more difficult for asylum seekers to win their cases.

Despite the fact that Asylum Office hearings are nonadversarial, because they are only one small part of a much larger system, this stage ultimately becomes subsumed into the overall picture of a regime that is highly adversarial, legalistic, and contentious. While some asylum seekers only experience the initial stage, the tenor of the regime as a whole is of adversarial legalism. There are different points of access to the process, different sites and forms of decision making, and low levels of insulation from political tinkering, as evidenced by the 2002 streamlining of the BIA and the REAL ID Act of 2005. The US RSD regime also has high levels of contentious judicial oversight, which has led to turf battles, inconsistency, and unpredictability. Taken together, these features raise important questions about the direction of US administrative law and the newly powerful role of the judicial branch in migration policy making.


In 2010, after many years of debate, the Canadian Parliament passed the Balanced Refugee Reform Act, introducing some changes to the Canadian RSD regime. After the Act was passed, the Canadian Bar Association, which had opposed previous versions of the bill, offered cautious praise for the final version. “Some of the concerns about the bill have been adequately addressed,” it said in a statement. “This speaks well of the willingness of the minister and MPs from all parties to respond to legitimate criticism” (Toronto Star 2010b). In a similarly supportive vein, well-known refugee advocate and former Chairman of the Immigration and Refugee Board Peter Showler wrote an op-ed about the Act. While he had some concerns and cautioned that “the devil is in the details,” he stated that “the government has made a practical and legitimate attempt to balance fairness with prompt refugee claim processing” (2010).

Compared to the fragmentation and institutional conflict of the US RSD regime, the Canadian regime is very centralized and the interactions between the courts and the administrative tribunal are less frequent and less fraught. As a senior official at the Immigration and Refugee Board put it: “[W]e are Canadian, so the relationship between the two is cordial. We don't have the same tensions and pressures that exist in the United States.”29 The Canadian RSD regime is characterized by the vertical accountability and legal informality of the professional judgment model of decision making (Kagan 2001, 10). The bureaucrats who conduct RSD in Canada have high levels of discretion to make decisions without legislative tinkering or judicial oversight and the low level of court involvement is uncontroversial. In sharp contrast to reform proposals in the United States and Australia, in its most recent legislation, the Canadian Parliament consolidated even more authority and invested even more resources in the administrative tribunal.

All Canadian RSD takes place within one agency: the Refugee Protection Division of the Immigration and Refugee Board (IRB), which Kernerman calls “one of the most sophisticated refugee determination systems in the world” (2008, 232) (see Figure 3). The IRB was established in 1989 after the Supreme Court of Canada read the 1982 Charter of Rights and Freedoms to require an in-person hearing for asylum seekers.30 Since that time, Canada has invested resources in creating a tribunal that would meet Mashaw's four standards for bureaucratic justice: accuracy, active investigation, effective management, and a systemic perspective (1985, 171). The IRB conducts inquisitorial hearings and does not use lawyers to represent the government's position. Rather, the tribunal engages in extensive research of its own prior to the hearing. The staff of the IRB is divided into regional teams that specialize in cases from one area of the world, including board members and tribunal officers who are assigned to them and assist them with research and preparation. The IRB houses a large documentation center with reports on country conditions, and the board shares the information it collects on the case with the claimant ahead of time. As a senior IRB official explained it, under an inquisitorial process, “we are not forcing the claimant to hire a lawyer because we are bearing the cost of the research.”31 This commitment to an active investigatory process stands in contrast with adversarial legalism, which is based on the premise that an impartial judge will decide whether the claimant is a refugee after hearing both sides argued forcefully.

Figure 3.

The RSD Process in Canada

The IRB has several locations, but the majority of claims are heard in the Toronto location.32 This centralization makes a system perspective and active management more feasible because decision makers interact regularly and receive policy directives via a hierarchical leadership structure. Once an asylum claim is filed, it is referred to the appropriate regional team within the RPD, where a staff member conducts a basic triage on the case. Using their specialized knowledge about the refugee claimant's country of origin and the facts of the case that are laid out in the detailed Personal Information Form, the tribunal officer assigns each case to one of three paths: expedited, regular, or extended hearing. Expedited hearings are used for seemingly clear-cut cases in which the applicant comes from a country with a very high acceptance rate and/or fits a basic profile that can be easily corroborated. These hearings do not involve board members directly; instead, the officer meets with the claimant in order to verify his or her story and, as a result of the conversation, will recommend that the person either be granted refugee status or continue on to a regular hearing. Most claimants go directly to a regular, or in complex cases, extended, hearing presided over by a board member.

The IRB enjoys a lot of flexibility and freedom to develop its own procedures and create its own jurisprudence. According to a senior official at the IRB: “[A] case cannot be overturned in court because it contradicts a similar case . . . so we must find ways to build mechanisms for consistency for ourselves internally. That's why we make the guidelines, to give a sense of where we think we should be going with this.”33 For example, Canada was the first country in the world to create a specialized policy on handling gender-based RSD claims.34 The IRB developed the guidelines autonomously and the courts granted the tribunal great deference and leeway to innovate proactively. The guidelines are not binding on courts, but courts have upheld them repeatedly, making it the controlling policy on the issue at all levels of the RSD regime. Today, gender-based asylum claims are a fairly settled area of Canadian refugee law. As one lawyer who regularly takes cases to the IRB put it: “The Board has been pretty good about that because the gender issue has been absorbed over there.”35 The IRB was not always as enlightened in the way it handled gender-based asylum claims, but it adapted to change quickly and definitively and has maintained a stable position on the issue ever since.

Another strategy the IRB uses is to identify “lead” or precedential cases to guide the decision makers without resorting to court intervention. The IRB first used this tactic in the late 1990s when the number of Czech Roma refugee claimants began to climb rapidly. IRB management designated a January 1999 case as “lead,” the member rejected the claim, and the decision was distributed to all the members. The rate of acceptance for Roma claimants dropped from 71 percent in December 1998 to 27 percent by March 1999 and to 9 percent by June 1999.36 The IRB has recently used a similar approach for managing the rapidly rising caseload from Mexico. In December 2006, IRB management labeled three decisions “Persuasive Decisions,” all of which were rejections. “Persuasive Decisions” differ from “lead” cases because they are selected after the fact and it is clear that the board members did not intend to set broad precedent when they wrote their reasons because there is very little language in the decisions that is generally applicable. These cases do promote the idea that state protection is available in Mexico, and suggest that Mexican claimants are not true refugees. In one case, the member explicitly states that the claimant “seems to be attempting to use the refugee system in order to gain entry into Canada.”37 In another case involving a homosexual claimant who had escaped violent persecution in rural Mexico, the board member concluded that the applicant could probably “relocate to Mexico City and access state protection if needed.”38 The acceptance rate for Mexican claims at the IRB went from 35 percent in 2006 to 15 percent in 2007, and by 2009 it was under 9 percent.39 The post hoc designation of very particularized cases as Persuasive Decisions may be a questionable management strategy, but it illustrates clearly the willingness of the IRB to step in and fill the legal void when there is no guiding federal court jurisprudence.

Some advocates saw a benefit to the informality of the IRB because it often led members to accept applications that were not solid fits with the legal definition of a refugee. As one lawyer put it: “If the member believes the person, they will find a way. You may get a ‘sorry, you are just not a refugee’ from a Federal Court judge, but not from the board members who are lay people. They don't get training in being bound by law.”40 This type of flexibility is less possible in an adversarial setting where the government is represented by legal counsel who will likely intervene if an asylum seeker is granted refugee status without a clear fit with the official definition.

In 2002, Parliament revamped the IRB with the aim of reducing costs by shifting from two-member to single-member hearings. In exchange, the Act created the Refugee Appeals Division (RAD), which would conduct an on-paper review of negative decisions on their merits. However, after the Act was passed, the Department of Citizenship and Immigration announced that due to “pressures on the system” implementation of the RAD would be delayed.41 Despite several attempts by some members of Parliament to pass another act forcing the implementation of the RAD, including a very near success in the summer of 2008, Parliament did not force the IRB to add the RAD for almost a decade (Taylor 2008). The Department of Citizenship and Immigration did not push for its implementation; the head of asylum policy told me: “What we have is a very good front-end system that is very strong and thorough, which makes the appeal a bit redundant.”42

When Parliament passed the Balanced Refugee Reform Act in 2010, it finally added the Appeals Division as a second layer of RSD within the administrative tribunal. In fact, the whole Act was focused on keeping claims within the agency; the reforms are designed to make the process more expedient and reduce the number of applications for judicial review. Under the new Act, the IRB will invest resources into processing claims from countries with low RSD acceptance rates more quickly than before (Toronto Star 2010a). The Act is not a huge overhaul and preserves all the key components of the previous system. The most controversial of the proposed changes, the inability of asylum seekers from designated “safe countries” to appeal their initial decision if it is negative, was not included in the final bill because too many members of Parliament were concerned about the fairness and the constitutional ramifications of such a move (Clark 2010).

A rough comparison of the caseload trends in the United States and Canada suggests an ostensible parallel between the two states. Between 1995 and 2000, the Federal Court's immigration workload approximately doubled, and after the passage of the Immigration and Refugee Protection Act of 2002, there was a further temporary surge in immigration cases as courts clarified new provisions in the law.43 Each year since 2000, immigration cases have made up at least 45 percent of the total caseload of the Federal Court Trial Division, and in five of those years that percentage rose above 60 percent.44 By the end of 2006, it had become clear that “the workload of the Federal Court is intrinsically related to the volume of refugee claimants that the IRB is handling,” and as a result, the Department of Citizenship and Immigration and the IRB pledged a “permanent transfer” of funds from their budgets to help offset the cost of processing the additional caseload.45

Despite this heavy caseload, a closer look makes clear that courts play a minor role in the Canadian RSD regime. Unlike in the heavily judicialized US RSD regime, once Canadian courts receive an application for judicial review in an immigration case, they use a paper screening process that weeds out about 85 percent of cases and lets the IRB decision stand without a court hearing.46 This requirement that claimants must apply for “special permission (or leave)” to have their case heard does not apply to any other area of law, and the Federal Court is not required to give reasons for why a case is granted judicial review or not. This lack of transparency makes it impossible to assess whether applications are rejected because they are not seen as having merit, or whether, as the statistics would imply, the Federal Court has adopted an unofficial policy of only hearing a specific ratio of the cases available, which hovers steadily around 15 percent.47 A recent study found that applicants who are represented by big law firms are more likely to be granted leave (Gould, Sheppard, and Wheeldon 2010), though these firms are also less likely to take on weak claims. The claimants who are granted leave go on to a full court hearing, but the prospect of reversal is extremely slim. The standard of review of an administrative decision is no more deferential than in the United States, and as in the United States, between 10 and 15 percent of cases that are granted a full hearing are successful in Federal Court. However, because so many cases are screened out, only about 1 percent of claims that are rejected by the IRB end up being overturned by a court. Essentially, administrative decision makers operate without fear of reversal in Canada because courts weigh in on the actions of the IRB very rarely.

If the Federal Court decides against a refugee claimant, the recourse for continued legal action is, again, extremely limited. The Federal Court of Appeal can hear a case only if the Federal Court judge decides that “a serious question of general importance is involved” and agrees to certify that question.48 The question must be proposed by either the applicant or the respondent prior to knowledge of the judgment in the case. As this process requires a judge to choose to open himself or herself up to reversal, it is not surprising that only about two-hundred questions have been certified on any immigration matters since the Immigration and Refugee Protection Act went into force in 2002, even though the Federal Court has heard thousands of immigration cases during that period.49 Given the cost of taking a case to the Federal Court of Appeal, it is also not surprising that out of these two hundred questions, the Federal Court of Appeal has answered only about 35 percent. In the rest of the cases, the appeal was never filed or the case was dismissed. Thus, out of the tens of thousands of RSD cases that the IRB has decided since 2002, only a handful have become a precedent-setting case at the Federal Court of Appeal, leading to a jurisprudential durability that is not seen in the United States.

One of the reasons that courts play a relatively small role in the Canadian RSD regime is that Canada's era of judicial power in regard to individual rights is quite recent. Prior to the adoption of the Canadian Charter of Rights and Freedoms in 1982, courts were much more focused on policing the separation of powers between the federal and provincial governments and on adjudicating economic matters than on protecting individual rights (Epp 1998).50 Scholarship on the effect of the charter has not examined how Canadian courts have reconciled the charter's new procedural fairness requirements with a long tradition of administrative deference (Sossin 2004). RSD matters make up the bulk of cases on the front line of this question, and this study suggests that courts are working hard to maintain a hands-off approach.

The IRB is not totally immune to political tinkering, as the delay of the RAD and controversies about appointment of board members illustrate, but relative to other regimes, the administrative insulation of the Canadian RSD regime is very high. A recent analysis of international legal standards concluded that the IRB meets international standards for independence in RSD (Heckman 2008). Advocates seem to value this insulation even while leveling harsh criticisms at other aspects of the regime. The head of the biggest and most powerful refugee advocacy group in Canada told me: “The basic core of the system is good though. If I were to change it, it would be to increase the independence of the IRB. I have low faith in the federal courts. They don't have refugee experience.”51


Australia is a relative newcomer to RSD, but it has generated more than enough legislation, jurisprudence, and controversy over the past few decades to make up for lost time. Since 2000, RSD has been brought to the center of a tense battle between Parliament and the courts, and a heated debate about the nature and source of rights more generally. The issue of power over RSD in Australia has traveled back and forth between Parliament and the courts much like a “bouncing ball” crossing over the net in the Australian Open (Kneebone 2003, 2).

The most recent volley occurred on November 11, 2010, when the High Court handed down the twin decisions of Plaintiff M61 and Plaintiff M69 v. Commonwealth of Australia. The two asylum seekers in the cases were Sri Lankans who attempted to arrive by boat in October 2009 and were interdicted at sea. In keeping with Australia's policy toward so-called offshore asylum seekers, the two men were sent to a detention and processing center on Christmas Island, an Australian territory south of Indonesia. While the asylum seekers were in detention, the Department of Immigration and Citizenship (DIAC), in conjunction with a private contractor, determined them to be ineligible for refugee protection. However, the process they used was far less transparent than the RSD process for typical “onshore” asylum seekers, who file for refugee protection once they arrive in Australia. In the unanimous M61 and M69 opinion, the High Court ruled that the RSD process DIAC had offered these men was inadequate, “denied [them] procedural fairness,” and “had the consequence of depriving them of their liberty.” The Court also found that the guiding legislation, the Migration Act, requires future RSD processes to comply with principles of procedural fairness, regardless of whether the asylum seekers make their claims on- or offshore.

This decision called the longstanding Australian two-tired RSD system into question, but because it was based on a statutory interpretation and not the Australian Constitution, the decision pushed the issue back into Parliament's hands for clarification. Writing in the Sydney Morning Herald, law Professor George Williams assessed the government's dilemma following the decision.

It is now open for Parliament to, in turn, amend the [Migration] Act to prevent this. The problem with this path is that it would not preclude future High Court interpretation of the act. History suggests that courts may find a way around even express words . . . When Parliament closes off one avenue of appeal to the courts, judges may readily find another. (2010)

The Australian Parliament has frequently amended the Migration Act in order to prevent asylum seekers from accessing thorough RSD hearings, to limit the power of courts to review RSD decisions, and to narrow the ways the refugee definition is interpreted by decision makers. The High Court has walked a fine line in response; it has deferred to Parliament by upholding many of its border control policies and limitations on asylum seeker rights within the Australian territory. However, the Court has been aggressive in policing the separation of powers and asserting its right to review administrative decisions. Parliament has only redoubled its attempts to keep the courts out of this policy area altogether.52 However, struggles with the High Court continue because the Court holds a trump card—a constitutionally entrenched right to review all decisions made by “Officers of the Commonwealth.”53 Because this power is vested in the High Court alone, Australia's RSD regime is characterized by a politically charged, high-profile interbranch dispute.

The Australian RSD regime is a hybrid of the US and Canadian systems. As in the United States, the administrative phase is fragmented, involving multiple distinct agencies (see Figure 4). However, as with Canada, every stage of Australian RSD prior to federal court review is nonadversarial and nonlegalistic, without the active participation of lawyers or reliance on formal legal rules. While the Australian regime has high levels of bureaucratic discretion, it differs from the Canadian system because it lacks the centralization designed to ensure consistency, and resource investment aimed at ensuring accuracy that characterizes a true model of bureaucratic justice (Mashaw 1985, 172). It also lacks the teeth of the US system to vet cases thoroughly and set legal precedents at the administrative level.

Figure 4.

The RSD Process in Australia

The first stage of Australian RSD is conducted by bureaucrats at DIAC, which is based in the capital city, Canberra. DIAC decision makers have access to a large clearinghouse of country of origin information, but unlike in the United States and Canada, the vast majority of these cases are processed without a hearing. Department decision makers may interview applicants on a discretionary basis, but do so less than 20 percent of the time.54 DIAC decisions are shrouded in secrecy and are the subject of ire among refugee advocates because decision makers are designated representatives of the Minister of Immigration and are not completely independent of the other more political functions of the department. As one advocate put it: “[T]he distinction between policy and law is a fundamental principle of public administration, but today I feel like public servants are serving the Minister, not the public. And that's dangerous.”55 A high-level DIAC official insisted that the minister does not have any direct influence over individual decisions and is kept at “arms length” from the RSD that is conducted within the department.56 However, the same official acknowledged that DIAC is in a delicate political situation because it is responsible for conducting both RSD and enforcement. The M61 and M69 case suggests that the dual role of enforcement and adjudication has been difficult for DIAC to navigate when dealing with offshore applicants.

In 1992, Parliament amended the Migration Act and added the Refugee Review Tribunal (RRT) in an attempt to keep RSD cases out of the federal courts. The RRT is an administrative body that reviews DIAC's decisions and is similar in many ways to the Canadian IRB. Hearings are conducted as inquisitorial conversations, and applicants may bring a friend, relative, or counsel, but the process is designed not to require legal representation. A recent comparison of migration policy in Canada and Australia stated that the two administrative tribunals are “highly comparable and similarly situated” without exploring the fact that in Australia the tribunal is part of the executive branch, which “removes the ideology of judicial independence” (Dauvergne 2005, 96–97). In addition, the scope of the RRT operations is much smaller than the Canadian IRB and because the RRT is only one part of a multifaceted RSD regime, it is far less involved than its Canadian counterpart in designing proactive RSD policy. These differences may stem from the fact that tribunals have an ambiguous place in Australian constitutional law (Creyke 2007). Unlike in Canada, where they are distinct and powerful bodies whose independence is fiercely protected by law, there is much debate in Australian law over where tribunals fall between the executive and the judicial branches.

The RRT's independence from DIAC is not well established; in 2009–2010, it upheld the negative DIAC decision 71 percent of the time.57 In 1998, the Immigration Minister publicly warned members of the RRT not to interpret the refugee definition in a way that was counter to “the sovereignty of parliament and the will of the Australian people.”58 That year, the tribunal's reversal rate of DIAC decisions was found to drop precipitously in the months leading up to the reappointment process (Legomsky 1998, 250). More recently, advocates have acted to reduce the discretion of RRT decision makers to reject asylum claims because they lack credibility. In 2005, asylum lawyers and advocates lobbied the RRT at a stakeholder meeting, leading it to release guidelines for decision makers on how to assess credibility properly. The guidelines state that members “should focus on what is objectively or reasonably believable in the circumstances.”59 The creation of guidelines is a move typical of an organization designed in the model of bureaucratic legalism because it is meant to create a more uniform standard for decision making (Kagan 2001). However, there is not much evidence that the tribunal's sudden attempt to provide guidance to its members and achieve consistency was successful. Over a year after the guidelines took effect, a lawyer who attends several hearings per week at the Sydney office said, “I never feel totally confident, even if it is a really strong case, because you never know who you will get.”60 Another lawyer commented: “Often, the RRT tries to make judge-proof decisions, by basing them on credibility.”61

Despite these efforts, since 2005, 40 percent of RRT decisions have been appealed to the federal courts.62 As a result, in a striking parallel to the US RSD regime, refugee matters have become a dominant part of the business of the Australian federal courts. In 2007, migration matters were 35 percent of the federal court workload, and 90 percent of those cases were refugee matters.63 As the head of the RRT told me: “[W]e have a jurisprudence in Australia that is second to none. There is a lot of judicial scrutiny of the process . . . and courts will find it if there is any failure to deliver procedural fairness.”64

Parliament has, on several occasions, tried to strip courts of their jurisdiction to hear appeals of RRT decisions and the courts have responded by routinely reasserting their constitutionally entrenched right to review, even on the most narrow grounds. For example, in 2001, the Australian Parliament passed the Migration Legislation Amendment (Judicial Review) Act, which added a privative clause to the Migration Act: a statutory removal of a court's ability to review decisions made under that piece of legislation. Specifically, Parliament amended the Migration Act to say that decisions under that Act are “final and conclusive” and “must not be challenged, appealed against, reviewed, quashed or called in question in any court.”65 This was a clear signal to the courts that their involvement in migration decision making was against the will of Parliament.

The High Court responded with a nuanced and calculated decision in Plaintiff S157/2002 v. Commonwealth of Australia (2003). In a nod to Parliament, the Court found that the addition of a privative clause to the Migration Act was constitutional. However, it went on to explain that any decision that was made in “jurisdictional error” would be “regarded, in law, as no decision at all.” In this way, the Court did not directly overturn an act of Parliament, but it eviscerated the intent of the Act by reasserting the right of courts to review all decisions that have been made in error. Since the S157 decision, Australian courts have defined jurisdictional error very broadly. In the case of SAAP & Anor v. Minister for Immigration and Multicultural and Indigenous Affairs (2005), the High Court found that a common RRT practice was a jurisdictional error, and remanded five-hundred similar cases to the RRT for reconsideration.66 A lawyer who has represented asylum seekers in many of the major High Court cases told me: “[T]he High Court is fairly involved, and overturns decisions reasonably often. Which means that tribunals are shown to be wrong, and this creates an in for litigation.”67

Other examples of the back and forth between Parliament and the courts have included, first, the courts' use of class actions to reverse large numbers of cases at once and the subsequent banning of class actions via the Migration Legislation Amendment Act (No. 2) 2001; second, the creation of time limits for bringing review requests in the Migration Litigation Reform Act 2005 and the subsequent invalidation of those time limits in the case of Bodruddaza v. MIMA (2007);68 and third (in many ways the most substantively significant skirmish), the amendment by Parliament of the Migration Act to narrow and specify the meaning of persecution to cases where there is a risk of “serious harm” and the persecution is “systematic.”69 This move was a direct response to High Court decisions interpreting the refugee definition more broadly. As the registrar of the Federal Court put it: “The judiciary is not accepting that parliament can limit their power . . . . The Court says that parliament is limited by the Constitution.”70

The most interesting aspect of the High Court's insistence on its right to review RSD is the fact that, according to Australian lawyers, the vast majority of the remands are based on very narrow procedural grounds, so that when the case is reheard by the tribunal, the outcome is the same. One lawyer told me: “[W]e often end up seeking review based on minute technical grounds that don't have much to do with the claim.”71 The Registrar of the Federal Court echoed this statement, saying: “Usually the jurisdictional error has nothing to do with the substance of the case, so a lot of them, even if they win in Federal Court, do not get a visa at the end of the day. Their long term prospects do not change here.”72 Statistics bear these comments out; in 2009–2010, the federal courts upheld RRT decisions 84 percent of the time.73 In the remaining 16 percent of cases, where the court remanded the matter back to the RRT for a rehearing, recent statistics reveal that the RRT changed its decision and granted the refugee visa 32 percent of the time under reconsideration after judicial review.74 Together, these numbers suggest that less than 3 percent of rejected applicants who seek judicial review end up being granted a refugee visa. So the overall impact for individual asylum seekers of going to federal court is limited.

Besides judicial review, there is also another appeal that applicants may use: ministerial discretion. Under the Migration Act, the Immigration Minister has the purely discretionary power to grant protected status. While there is no official procedure for how ministerial discretion works, and no way of predicting who will be successful, almost every rejected asylum seeker makes an appeal to the minister, even if he or she is also appealing the case to the federal court. This extra pathway allows DIAC to be involved in RSD all the way through the process, something that does not happen in either the United States or Canada.75 According to the UNHCR in Canberra, asylum seekers who ask for ministerial discretion and have UNHCR support letters have been quite successful (around two-thirds have been approved), despite their rejections at earlier stages of the process. But because this process is so nontransparent, it has come under criticism from advocates. In early 2008, Immigration Minister Evans from the newly elected Rudd government announced a desire to reform this system, stating, “I have formed the view that I have too much power” (Metherell 2008). To date, however, appeals to the minister remain the last resort for failed asylum seekers in Australia.

The Australian RSD has high levels of judicial review that at first glance are very similar to those in the United States. However, unlike the high-profile US battle between the administrative agencies and the courts, where federal judges use their power of review to send a message about the low quality of administrative decision making, the Australian federal courts seem much more intent on sending a message to Parliament. In addition, unlike in the United States, the Australian High Court is a key player in asserting judicial power. Thus, Australia's RSD regime is a fascinating paradox: a full privative clause declaring tribunal decisions to be “final and conclusive” has resulted in a system in which the courts are deeply involved in setting the parameters of procedural justice for asylum seekers. In contrast, the Canadian system has restricted judicial review without banning it.


Some Chinese asylum seekers are high-profile journalists, activists, dissidents, and religious leader with relatively clear-cut cases.76 However, a large proportion are ordinary people fleeing generalized state repression such as the “one child” policy instituted by the Chinese government in 1979 (Zhang 2007). This policy is enforced primarily through economic incentives (families with one child receive free education and health care) and propaganda campaigns (Hsu 1996). However, governments in some regions use more aggressive enforcement strategies, including compulsory birth control through IUD insertion, forced sterilization and abortion, heavy fines, and intimidation by local officials. Experts estimate that since the 1980s, the Chinese government has sterilized millions of people per year, both male and female, as a punishment for having two children (Hsu 1996).

Chinese asylum claims based on coercive population control are a challenging type of RSD. As a matter of legal interpretation, it is not clear whether those affected by a general policy are UN Convention refugees, that is, targets of individual persecution, even if the policy violates human rights norms.77 Practically speaking, because there is much local variation in how the one child policy is implemented, and because many claims are based on fear of future action, the claims are almost impossible to corroborate. RSD adjudicators must decide difficult legal questions about the difference between oppression and persecution, the standard for establishing that fear is well-founded, and whether millions of people with little in common other than the desire to have more than one child are a “particular social group.” They must also determine credibility case by case with little evidence.

The United States

The position of the US government on whether coercive population control is grounds for granting refugee status has varied according to the political winds and is illustrative of the institutional tug of war, adversarial legalism, and political wrangling that characterizes the RSD regime. In 1988, driven by ideological concerns about forced family planning, the Reagan administration issued a memorandum to the Immigration and Naturalization Service instructing it to give “careful consideration” to forced sterilization or abortion claims.78 However, in 1989, when the Board of Immigration Appeals first considered the issue, it decided that such advisory memos were not binding and that coerced family planning did not amount to “persecution” under the refugee definition.79 Thus, from the beginning, the US RSD regime issued contradictory decisions: the BIA and the EOIR denied coercive population control claims while the INS (a precursor to the Asylum Office) followed the guidance to accept them.

In 1990, President Bush issued an executive order requiring adjudicators to give “enhanced consideration” to asylum seekers who expressed fear of persecution through forced sterilization or abortion (Rosenthal 1990).80 However, the Clinton administration was far less supportive of such claims (Arnold 1993). When the freighter ship Golden Venture ran aground in New York harbor with three-hundred Chinese stowaways in 1993, INS officials were quoted as saying: “The Bush Administration encouraged this. They sent a signal: If you get to the U.S., no matter in what manner, we'll accept your presence here” (Weiner 1993). In an attempt to discourage future smuggling, the INS detained the Golden Venture asylum seekers while their claims were processed, the vast majority of which were based on fear of forced sterilization. In sharp contrast to the high acceptance rates in previous years, the EOIR denied almost all these applications, and the courts upheld these decisions, finding that Bush's executive order had not overruled the BIA's original ruling on the question (Faison 1993).

To circumvent what they saw as Clinton administration stonewalling, the Republican-controlled Congress included a provision in the Illegal Immigration Reform and Immigrant Responsibility Act of 1996 overruling the BIA's decision and the opinions of the federal courts. It amended the Immigration and Nationality Act to include explicit protection for victims of coercive population control policies, and those who feared such policies.81 A few months later, in the Matter of C-Y-Z (1997), the BIA interpreted IIRAIRA's language to include spouses of people affected by coercive population control measures, granting asylum to a male applicant whose wife had been sterilized.82 That decision had a huge impact on the number of people eligible for asylum because many Chinese asylum seekers were men who had fled alone and hoped to sponsor their wives later.

The issue was not settled, however, because by 2006, the spousal question had reached federal court. In Shi Liang Lin v. Gonzales (2007), the Second Circuit concluded that the wording of IIRAIRA was ambiguous on whether it applied to partners. Because the statutory instructions to the administrative agency were not clear, the court concluded that it had the power to interpret the statute rather than grant the agency deference. It concluded that IIRAIRA protection did not extend to partners. Aside from contradicting ten-year-old BIA precedent, this decision also created splits in the circuits (i.e., subnational variation in court interpretation). In contrast to the Second Circuit, the Fifth, Sixth, Seventh, and Ninth Circuits have all adopted the C-Y-Z standard of including spouses.83 These circuits are also split on whether nonmarried partners can be included in the C-Y-Z reasoning, something the BIA does not favor.84

The Chinese case illustrates how policy decisions and legal outcomes are closely intertwined, and highly variable in the US RSD regime. An unmarried partner of a forced sterilization or abortion victim may be eligible for asylum in some parts of the country and not others. In the Second Circuit, where the majority of Chinese asylum claims are filed, only the victim herself is eligible. However, because immigration courts still follow the more generous BIA logic, men can be granted asylum by an IJ in the region covered by the Second Circuit and the decision will stand as long as the case is not appealed to federal court. This situation makes outcomes of asylum claims based on coercive population control extremely unpredictable and inconsistent, even among applications that were initially filed in the same office. The Supreme Court has not weighed in to resolve splits in the circuits, leaving RSD outcomes as unpredictable as ever.


Because Canada's RSD regime is so centralized and the Immigration and Refugee Board is so insulated, there is far less jurisprudence about Chinese claims than in the United States. Despite the fact that many Chinese applicants are male, from the earliest days, the IRB subsumed the coercive population control issue under its pioneering stance on gender-based refugee claims. Canada's gender guidelines, in listing the forms of persecution women face, specifically mention forced abortion, or compulsory sterilization.85 The IRB used this guideline to extend protection to men, while coercive population control claims in the United States and Australia were never connected to the concurrent evolution of gender-based claims in those countries.

Only a few weeks after the IRB released Canada's gender guidelines in 1993, a case dealing with the issue of forced sterilization reached the Canadian courts for the first time. The case concerned a Chinese woman named Ting Ting Cheung who fled China to avoid forced sterilization as punishment for having a second child. When her case was first heard, the decision maker at the IRB did not find her to be a UN Convention refugee, reasoning that, as a general law, the one child policy did not constitute persecution. By the time Cheung's case reached the Federal Court of Appeals, the gender guidelines had been issued. While guidelines for administrative decision makers are not legally binding on courts, the Federal Court of Appeals showed great deference to the tribunal. It found both Cheung and her daughter to be UN Convention refugees, stating that “sometimes the operation of a law of general application can constitute persecution.”86

In the aftermath of the Cheung decision, the question of whether males facing forced sterilization were UN Convention refugees was left wide open. But instead of the continued litigation of the US story, the IRB acted to fill the legal void. In 1997, it published a precedent-setting decision in which it considered the application of a Chinese man who had violated the one child policy and concluded that “to limit Cheung to women would have been to suggest that the reproductive rights of men were not on a par with those of women.”87 Thus, they found the claimant to be a member of a particular social group: men in China who face forced sterilization.

Since that time, the issue has, amazingly, not returned to the federal courts for clarification and the IRB has continued to consider male victims of forced sterilization to be members of a particular social group.88 The deference shown by the Canadian federal courts to tribunal-level innovation means that each new issue is not litigated endlessly, as it is in the United States. Also notable is the deference shown to the tribunal by Parliament and the Department of Citizenship. Both these political entities sat out and allowed the issue to be handled as an administrative matter. Thus, the position of the Canadian RSD regime on coercive population control claims has remained both stable over time and remarkably generous compared to other countries.


Chinese asylum claims have been central to the development of the Australian RSD regime. Government concern about China's proximity and large population, combined with the fact that the High Court has taken a narrow textual reading of the refugee definition, has resulted in very few successful claims based on coercive population control. The first major High Court case interpreting the refugee definition was a Chinese sterilization case, Chan Yee Kin v. Minister for Immigration and Ethnic Affairs (1989). In that decision, the Court stuck closely to the text of the refugee definition and made clear that in order for fear to be considered well founded there must be a “real chance” of persecution. This “real chance” test remains a foundational part of Australian refugee jurisprudence today.

The Chinese case Applicant A v. Minister of Immigration and Ethnic Affairs (1997) is also central to Australian refugee law and its trajectory is illustrative of the RSD regime. Applicant A's asylum claim based on fear of forced sterilization was rejected by DIAC, but on appeal the RRT overturned the department's decision and granted refugee status. Then, in a rare move, the Minister of Immigration appealed this positive decision. The Federal Court upheld the RRT decision, creating such a public controversy about opening the floodgates from China that Parliament drafted a bill intended to reverse the decision (Crock 2004). Senator McKiernan, in urging passage of the bill, stated, “I would anticipate that hundreds of thousands of people from China and some other Asian countries will shortly be making plans to get to Australia.”89 The bill was never put to a vote because after the Full Federal Court reversed the Federal Court decision, Applicant A appealed to the High Court.90 A majority of the High Court concluded that since there were millions of Chinese people who would fall into the category of potential victims of forced sterilization, “to construe the term ‘particular social group’ in that way would make it an almost all-encompassing safety net.” In a strict textual reading of the refugee definition, the majority of the Court argued that it would be circular to call people who feared forced sterilization members of a particular social group. China's one child policy is a general law and thus potential victims of forced sterilization in China lack the critical nexus between fear and the UN Convention grounds that would make them eligible for refugee protection.

The logic of Applicant A is still dominant, but it has been tempered by a recent High Court decision on the eligibility for refugee status of so-called black children—children born in contravention of the one child policy. In April 2000, the High Court affirmed the RRT decision that “black children” are a particular social group in China because they are refused basic education and health care and face social discrimination. While Chen Shi Hai v. Minister for Immigration and Multicultural Affairs (2000) is technically more generous than the Applicant A decision, both opinions rely heavily on a technical, textual reading of the refugee definition and are logically consistent with one another.

The High Court's textualist approach is generally compatible with the concerns about floodgates frequently expressed by members of Parliament. The Chen Shi Hai case illustrates, however, that the Court is not simply bowing to the will of the legislative branch. The role of the RRT stands in contrast to the proactive policy making of the Canadian IRB, and the bold precedent setting of the US BIA. When the RRT made decisions that were potentially precedent setting, the minister appealed, and ultimately the High Court decided. In the story of Chinese coercive population control claims, the administrative tribunal is an un-insulated fulcrum in the battle between the courts and parliament. Political fears about unbridled Chinese migration have dominated, but the High Court has shown that its commitment to legally consistent decision making is ultimately more powerful than its willingness to go along with government priorities.


Every time RSD is conducted, a shift occurs from the global to the particular, as the lofty ideals of international law take the more mundane form of bureaucratic decision making. Domestic divergence theory is right to suggest that we cannot assume this shift to be uniform or international human rights law to work the same way in every place it settles. Divergence theory recognizes that each state will react differently to the rise of global governance institutions and the human rights norms they perpetuate. However, this line of thinking can fall into a view of domestic politics as a binary struggle between exclusionary legislators and inclusive courts. This three-country comparison suggests that state responses are much more internally complex. The concept of an RSD regime provides a mechanism for explaining striking cross-national differences between states with much in common. Below, I describe three important additions to domestic divergence theory that the comparison illuminates.

First, domestic struggles over asylum policy are only partially about how generous to be. The dynamics of RSD regimes are also about protecting institutional turf, such as asserting the right to review administrative decisions or reemphasizing the position of administrative agencies as expert specialists. In many interbranch conflicts, international human rights norms are not part of the conversation. In others, they are used strategically to shore up power. When the Australian High Court refused to acknowledge the privative clause that Parliament had added in order to strip jurisdiction over RSD, it was not because the Court had a preference for a more expansionist migration policy. The Court was protecting its power. In the United States, the RSD regime has been subsumed by political battles over abortion, struggles over administrative agency authority, and conflicting court interpretations. Political pressure to limit large-scale migration exists in all three countries, but power struggles are often unrelated to the desired migration policy outcome held by a particular actor. Nevertheless, the outcomes of power struggles have unintended consequences that affect the rights of migrants.

Second, courts are not always a strong counterbalance to the exclusion and restrictionism of the political branches. Courts must be strategic about when they push against strong political will, and they need ammunition to support bold stances. A determining factor in predicting what role courts will play in migration policy is the traditional level of power that courts have in politics more generally. The United States, Canada, and Australia all have strong courts relative to other countries, but even among these three powerful judiciaries, there is variation in the level and style of court involvement in policy making. US courts have a much longer tradition of activism and are more involved in reviewing the substance of RSD, even granting refugee status to individual asylum seekers. Canadian courts were empowered by the passage of the Charter of Rights and Freedoms in 1982, but beyond requiring a hearing, are still reluctant to become involved in the substance of RSD. Australian courts pick their battles: they are willing to protect their turf, and step in when political interference in the administrative process has gone too far, but they have no eagerness to push for an expansionist reading of the refugee definition.

Third, this study illustrates that international norms can bubble up into an RSD regime via administrative agencies; they need not always trickle down from courts. Thus, courts are not the only forces of expansion and rights protection in an RSD regime, just as the political branches are not the only force of restriction. Administrative agencies are the front-line interpreters of international law as well as the front-line enforcers of migration policy. In this sense, individual, low-level bureaucrats are the agents of legal globalization, not judges or politicians. These decision makers must constantly respond to new types of claims and the innovative legal arguments supporting them. An RSD regime with high administrative insulation puts a great deal of pressure on the one hearing, requiring the tribunal to invest resources into fact finding and training, to ensure decision makers have as much information at their disposal as possible. It also results in questions about the refugee definition being more settled, either in a generous or restrictive direction, than they are in other, less centralized regimes. The centralized Canadian system has evolved much more quickly and unidirectionally than that of the United States or Australia. Because Canada has high levels of administrative insulation, the IRB develops detailed guidance for its decision makers that is not contradicted later. By contrast, both the Australian Parliament and the US Congress have acted to narrow or specify the refugee definition; this type of legislative intervention has not happened in Canada.

The claims of many modern asylum seekers fall into categories that were unanticipated by the framers of the 1951 Refugee Convention (Martin 1988). As the exhilic model, in which refugees are typically political dissidents, becomes less dominant in the post–Cold War period, many asylum claims fall into the gray middle zone between obvious UN Convention refugees and economic migrants. As the examples of Chinese victims of coercive population control illustrates, RSD regimes can have a particularly stark effect on asylum seekers with more ambiguous claims because regimes vary widely in their ability to adapt to change in a clear and consistent manner. Bureaucratic justice is conceptualized and organized differently in different states, and so states vary in how they draw the line between refugee and nonrefugee.


  • 1

    China has been the top source country for asylum applications in the United States every year since 1999 except 2001, when it was second. In Canada, China was in the top three source countries from 1998–2007 and has since been number five. China has been the top source country in Australia since 2002 (“Asylum Levels and Trends in Industrialized Countries,”UNHCR Reports: 1980–2009).

  • 2

    The upward trend in acceptance rates, despite other differences, is most likely due to China's outlawing of the popular spiritual movement Falun Gong in 1999.

  • 3

    UNHCR Statistical Online Population Database:

  • 4

    Six other countries sent at least two-hundred asylum seekers to the United States, Canada, and Australia in 2009. The acceptance rates were as follows: India: 27 percent (US), 33 percent (CAN), 2 percent (AUS); Iraq: 77 percent (US), 75 percent (CAN), 67 percent (AUS); Iran: 53 percent (US), 76 percent (CAN), 89 percent (AUS); Pakistan: 29 percent (US), 59 percent (CAN), 39 percent (AUS); Sri Lanka: 35 percent (US), 91 percent (CAN), 65 percent (AUS); Zimbabwe: 40 (US), 85 percent (CAN), 59 percent (AUS).

  • 5

    To calculate these numbers, I compared the UN asylum application statistics for 2003–2009 with the most recent census estimates for the population of each country.

  • 6

    In the United States, I searched the The New York Times and Washington Post for the terms “asylum” and “congress” and kept 445 articles. In Canada, I searched The Globe and Mail and The Toronto Star. Because the term asylum is never used in Canada, I searched for “refugee” and “parliament” and kept 453 articles. In Australia, I searched The Age and The Sydney Morning Herald for the terms “asylum” and “parliament” and kept 1,362 articles. I included all news items, features, editorials, and op-eds.

  • 7

    The interviews averaged seventy-four minutes long. For a full list of interviews, see Appendix 1.

  • 8

    I did not include decision makers as an interviewee group because attaining official permission would have been next to impossible and would have inhibited candor. In Canada, I was given permission to interview one sitting member of the Immigration and Refugee Board, but I do not pretend that he/she was representative. Five interviewees (three in Canada and two in Australia) were former decision makers, and I discussed this experience with them at length. In the United States, I had several informal, confidential discussions with current immigration judges (two in particular) as a result of my visits to the EOIR to observe hearings. I do not list those as official interviews, but they were extremely informative conversations.

  • 9

    I read seventy-eight cases from the United States, twenty-five from Canada, and twenty-five from Australia.

  • 10

    Author interview, 8/24/2007.

  • 11

    Author interview, 9/14/2007.

  • 12

    This study also found variation across individual Asylum Officers, but to a far lesser degree.

  • 13

    Board of Immigration Appeals: Final Rule: Providing Quality Service Through More Expeditious Review, August 23, 2002, US Department of Justice Fact Sheet; BIA Restructuring and Streamlining Procedures, March 9, 2006 (revised), US Department of Justice Fact Sheet.

  • 14

    Executive Office of Immigration Review, Statistical Year Books, 2000–2007.

  • 15

    Author interview, 8/24/2007.

  • 16

    Author interview, 8/23/2007.

  • 17

    Author interview, 8/27/2007.

  • 18

    See Chevron v. Natural Res. Def. Council (1984) and United States v. Mead Corp. (2001).

  • 19

    Administrative Office of US Courts, Judicial Business, 1996–2009.

  • 20

    These numbers do not reflect the dramatic increase in the number of people charged with immigration-related crimes. See the Transaction Records Access Clearinghouse:

  • 21

    Administrative Office of US Courts, Judicial Business, 2007.

  • 22

    I have repeatedly requested this information from the US Administrative Office of the Federal Courts, Statistics Division, and from individual circuit courts, but have been told that the data do not exist.

  • 23

    Illegal Immigration Reform and Immigration Responsibility Act, 1996, amending US Code 1252a(2)(B)(ii).

  • 24

    Senate Hearing 109–537, “Immigration Litigation Reduction.”

  • 25

    Author interview, 2/14/2007. See also Law's (2010) interviews with Ninth Circuit judges, who repeatedly expressed disdain for the IJs and the BIA.

  • 26

    Author interview, 8/23/2007.

  • 27

    Author interview, 8/24/2007.

  • 28

    Author interview, 8/23/2007.

  • 29

    Author interview, 3/29/2007.

  • 30

    Singh v. Minister of Employment & Immigration (1985).

  • 31

    Author interview, 3/29/2007.

  • 32

    The Toronto office of the IRB processes 55 percent of the total national caseload.

  • 33

    Author interview, 3/29/2007.

  • 34

    Guideline Four: Woman Refugee Claimants Fearing Gender Based Persecution, Immigration and Refugee Board of Canada, February 1993. Updated in 1996 and 2003.

  • 35

    Author interview, 5/29/2007.

  • 36

    IRB statistics presented in the case at the Federal Court. Corroborated by the UNHCR Statistical Database. Kozak v. Canada (Minister of Citizenship and Immigration) (2006). For a detailed discussion of this controversy, see Kernerman (2008).

  • 37

    IRB Decision TA4-18833, February 27, 2006.

  • 38

    IRB Decisions TA4-10802 and TA4-10803, February 24, 2005.

  • 39

    UNHCR Statistical Online Population Database:

  • 40

    Author interview, 5/7/2007.

  • 41

    CIC Press Release, “Refugee Appeal Division Implementation Delayed,” April 29, 2002.

  • 42

    Author interview, 3/30/2007.

  • 43

    Canadian Courts Administration Service Report, 2003–04, Federal Court of Canada Statistical Reports, 2000–2010.

  • 44

    In 2009, 79 percent of proceedings commenced in Federal Court were immigration matters. Two-thirds of those cases, or 52 percent of the total caseload, were RSD appeals.

  • 45

    Canadian Courts Administration Service Report, 2005–06.

  • 46

    Federal Court of Canada Statistical Reports, 2005–2010.

  • 47

    The leave rate for refugee cases was 15 percent in 2005, 14 percent in 2006, 18 percent in 2007, 15 percent in 2008, and 13 percent in 2009 (Federal Court of Canada Statistical Reports).

  • 48

    Immigration and Refugee Protection Act, 2002, Section 74.

  • 49

    Federal Court of Canada Web site, “Immigration—Certified Questions”: (accessed August 5, 2010).

  • 50

    While Epp discusses the shift in the Supreme Court of Canada's docket away from economic matters and toward individual rights between 1960 and 1990, he makes no mention of administrative law or the relative proportion of the Court's docket it represented during that period.

  • 51

    Author interview, 4/7/2007.

  • 52

    When the interbranch turf wars between Parliament and the High Court first emerged, they seemed like a reaction to the aggressive politics of deterrence that the Liberal Party pursued under Prime Minister Howard from 1996–2007. However, since the Labor Party took control in 2007, both parties seem to agree on maintaining the basic framework of that approach. See Narushima (2010).

  • 53

    Commonwealth of Australia Constitution Act of 1901, Section 75(v): Original Jurisdiction of High Court. The text gives original jurisdiction over writs or injunctions, but has been interpreted to mean all orders made by an officer of the Commonwealth.

  • 54

    According to the head of the RSD program, DIAC conducts interviews “if there is a credibility issue, or for client facilitation reasons, like if the person has major obstacles to expressing themselves on paper.” Author interview, 10/5/2007.

  • 55

    Author interview, 10/16/2007.

  • 56

    Author interview, 10/5/2007.

  • 57

    Migration Review Tribunal and Refugee Review Tribunal Annual Report 2009–2010, p. 36.

  • 58

    Philip Ruddock, “Immigration Reform: Unfinished Agenda” (speech delivered at the National Press Club, Canberra, March 18, 1998).

  • 59

    Guidance on the Assessment of Credibility, Refugee Review Tribunal, October 2006.

  • 60

    Author interview, 11/9/2007.

  • 61

    Author interview, 10/16/2007.

  • 62

    Refugee Review Tribunal Annual Reports, 2005–2010.

  • 63

    Author interview, 10/29/2007.

  • 64

    Author interview, 11/2/2007.

  • 65

    Migration Act 1958, Section 474, Part 1, (a) and (b).

  • 66

    In response to this decision, Parliament revised the Migration Act via the Migration Amendment (Review Provisions) Act of 2007.

  • 67

    Author interview, 10/26/2007.

  • 68

    The Migration Litigation Reform Act also gave the Federal Magistrates Court jurisdiction to hear claims that the RRT had made a jurisdictional error. Parliament intended this new layer of review to be a buffer; however, it seems to have done nothing to slow the rate of appeal into the Federal Court.

  • 69

    Section 91R of the Migration Act, amended by the Migration Legislation Amendment Bill (No. 6) 2001.

  • 70

    Author interview, 11/5/2007.

  • 71

    Author interview, 10/16/2007.

  • 72

    Author interview, 11/5/2007.

  • 73

    “Controlling Migration Litigation.” Denis O'Brien, Principal Member, Migration Review Tribunal and Refugee Review Tribunal. August 7, 2009. National Administrative Law Forum, Canberra.

  • 74

    Migration Review Tribunal and Refugee Review Tribunal Annual Report 2009–10, p. 38.

  • 75

    Unsuccessful asylum seekers can seek protection on “humanitarian and compassionate” grounds in Canada, and a form of relief called “withholding of removal” in the United States, but both processes are more formalized and transparent than ministerial discretion.

  • 76

    US Department of State, Bureau of Democracy, Human Rights, and Labor. Country Report on Human Rights Practices: China. 1999–2008. Available at

  • 77

    The five grounds are “race, religion, nationality, membership of a particular social group, or political opinion.” Article 1(a). Convention Relating to the Status of Refugees. July 28, 1951.

  • 78

    Memorandum. Office of Attorney General Meese to Alan Nelson (August 5, 1988).

  • 79

    Matter of Chang, Interim Decision #3107, Board of Immigration Appeals, May 11, 1989.

  • 80

    Executive Order No. 12,711, 55 Fed. Reg. 13,897 (April 11, 1990).

  • 81

    The text added by IIRIRA states: “A person who has been forced to abort a pregnancy or to undergo involuntary sterilization, or has been persecuted for failure or refusal to undergo such a procedure, or for other resistance to a coercive population control program, shall be deemed to have been persecuted on account of political opinion, and a person who has a well-founded fear that he or she will be forced to undergo such a procedure or subject to persecution for such failure, refusal, or resistance shall be deemed to have a well-founded fear of persecution on account of political opinion” (8 U.S.C. § 1101(a)(42)).

  • 82

    In re C-Y-Z, Applicant. Interim Decision # 3319, Board of Immigration Appeals, June 4, 1997.

  • 83

    See, for example, Zhang v. Gonzales (7th Cir. 2006); Huang v. Ashcroft (6th Cir. 2004); He v. Ashcroft (9th Cir. 2003); Li v. Ashcroft (5th Cir. 2003).

  • 84

    Zhu v. Gonzales (7th Cir. 2006) and Ma v. Ashcroft (9th Cir. 2004) support claims of partners who have been unofficially married in traditional ceremonies. The Third Circuit recognizes only official marriages (Chen v. Gonzales, 3d Cir. 2005). The remaining circuits have not ruled on these questions.

  • 85

    Women Refugee Claimants Fearing Gender Related Persecution, Immigration and Refugee Board, Chairperson's Guidelines, March 9, 1993.

  • 86

    Cheung v. Canada (Minister of Employment and Immigration) (1993).

  • 87

    Immigration and Refugee Board, Decision # U93-08191, Mora, April 21, 1997.

  • 88

    Immigration and Refugee Board, Decision # V99-03480, April 11, 2000.

  • 89

    Senator Jim McKiernan, “Interdiction May Have to Be Considered,” Media Release, March 2, 1995 (quoted in Crock [2004]).

  • 90

    Applicant A v. Minister for Immigration & Ethnic Affairs (Federal Court of Australia 1994).



Organization/Role Date
US Citizen and Immigration Services, Asylum Office2/05/2007
UNHCR DC2/06/2007
Human Rights First2/07/2007
Refugee Council USA2/07/2007
US Conference of Catholic Bishops2/07/2007
Human Rights Watch2/08/2007
International Rescue Committee2/08/2007
Women's Commission for Refugee Women and Children2/08/2007
American Immigration Lawyers Association2/09/2007
Amnesty International—DC2/09/2007
Migration Policy Institute2/09/2007
Refugees International2/12/2007
US Advocate2/12/2007
Office of Immigration Litigation, Dep't of Justice2/12/2007
Office of Immigration Litigation, Dep't of Justice2/12/2007
Board of Immigration Appeals2/13/2007
State Department—Bureau of Population, Refugees and Migration2/13/2007
American Bar Association Committee on Immigration2/14/2007
US Advocate2/15/2007
Lutheran Immigration and Refugee Services2/16/2007
Immigration and Refugee Board—Operations3/29/2007
Citizenship and Immigration Canada—Asylum Policy3/30/2007
Citizenship and Immigration Canada—Refugee Branch3/30/2007
CIC—International Coordination3/30/2007
Former Immigration Refugee Board Member4/02/2007
Amnesty International—Montreal4/03/2007
CIC—Senior Counsel4/03/2007
Canadian Council for Refugees4/07/2007
Can Advocate5/09/2007
Can Advocate5/09/2007
Immigration and Refugee Board—Regional Director5/10/2007
Immigration and Refugee Board—Legal Services5/10/2007
Can Advocate5/11/2007
Can Advocate5/11/2007
Immigration and Refugee Board—Training5/14/2007
Can Advocate5/14/2007
Immigration and Refugee Board—Coordinating Member5/15/2007
Can Advocate5/15/2007
Can Advocate5/15/2007
Can Advocate5/18/2007
Can Advocate5/22/2007
Can Advocate5/22/2007
Can Advocate5/23/2007
Can Advocate5/24/2007
Department of Justice—Canada5/25/2007
Can Advocate5/25/2007
Can Advocate5/28/2007
Can Advocate5/28/2007
Can Advocate5/28/2007
Can Advocate5/29/2007
Can Advocate5/29/2007
Refugee Law Office—Toronto5/30/2007
Can Advocate5/30/2007
Can Advocate5/31/2007
US Advocate8/23/2007
US Advocate8/23/2007
US Advocate8/24/2007
US Advocate8/27/2007
US Advocate8/28/2007
US Advocate8/29/2007
US Advocate9/12/2007
US Advocate9/13/2007
US Advocate9/14/2007
US Advocate9/15/2007
Federation of Ethnic Communities Australia10/03/2007
Dep't of Immigration and Citizenship—Legal Policy10/05/2007
Dep't of Immigration and Citizenship—Legal Officer10/05/2007
Dep't of Immigration and Citizenship—Protection Policy10/05/2007
Dep't of Immigration and Citizenship—Litigation10/05/2007
Dep't of Immigration and Citizenship—Onshore Protection Branch10/05/2007
Refugee Action Committee10/08/2007
AUS Advocate10/15/2007
A Just Australia10/16/2007
AUS Advocate10/16/2007
AUS Advocate10/16/2007
Refugee Council of Australia10/18/2007
Edmund Rice Center10/18/2007
Jesuit Refugee Services10/19/2007
Refugee Advice and Casework Service10/22/2007
Asylum Seeker Center10/22/2007
Bar Association of New South Wales10/23/2007
AUS Advocate10/24/2007
AUS Advocate10/25/2007
AUS Advocate10/25/2007
AUS Advocate10/26/2007
Legal Aid Office—Sydney10/29/2007
AUS Advocate10/31/2007
AUS Advocate10/31/2007
AUS Advocate11/01/2007
Refugee Review Tribunal—Registrar11/02/2007
Refugee Review Tribunal—Legal Services11/02/2007
AUS Advocate11/05/2007
Federal Court of Australia—Registrar's Office11/05/2007
Federal Court of Australia—Registrar's Office11/05/2007
AUS Advocate11/06/2007
AUS Advocate11/06/2007
AUS Advocate11/07/2007
AUS Advocate11/08/2007
AUS Advocate11/09/2007
AUS Advocate11/13/2007