Printer Friendly

Refugee credibility assessment and the "religious imposter" problem: a case study of Eritrean Pentecostal claims in Egypt.


Credibility assessment in refugee status determination (RSD) poses unique challenges when the outcome of asylum applications turns on the question of whether an asylum seeker is actually a member of a persecuted religious minority. These cases require secular adjudicators to delve into matters of religious identity and faith that are, by their nature, subjective and beyond the realm of objective analysis. This Article explores practical means of addressing this challenge through a case study of the RSD interviews of Eritrean asylum seekers in Egypt who based their refugee claims on Pentecostal religious associations. Analysis of the interview methods used in RSD interviews indicates that RSD decision makers operated from several implicit assumptions about how to conduct religious-credibility assessment. Attempts to test the sincerity of religious faith via knowledge quizzes and inquiries into subjective beliefs have questionable logical justifications and are fraught with significant risks. By contrast, the most logically defensible approaches are based on the "eye of the persecutor" test, which focuses on observable triggers of persecution that put individuals at risk.


 A. A Gap in Standards
 B. Defining the Question
 C. Analyzing Questions in Refugee Status
 Determination (RSD)
 D. Source of Interview Transcripts

 A. Pentecostalism in East Africa
 B. Religious Repression in Eritrea
 C. Eritrean Refugee Claims Under International
 Refugee Law

 A. The Sincerity Test
 B. Neutral Questions on Religious Beliefs and
 C. Religious Knowledge Quiz
 D. Theological Clarification or Dispute
 E. Weaknesses of the Sincerity Test
 F. The Eyes of the Persecutor Test
 G. Narrative Focus
 H. Open Versus Closed Questions



Refugee status determination (RSD) is a special type of adjudication where officials from secular governments and the United Nations conduct formal hearings into religious faith. The interviews are lengthy and often intense, sometimes forcing people to recite Bible citations, to tell how and why they started going to one church instead of another, or to explain for the record why they pray in a particular way. These awkward scenes are not anticipated by any doctrinal dispute in refugee law. While substantial controversy has arisen about the international definition of a refugee, (1) few governments would contest that a person whose life or freedom is in danger because of her religious beliefs is a refugee under international law. Given the 1951 Refugee Convention's roots in the aftermath of the Nazi Holocaust, one might consider religious persecution cases to be prototypical refugee claims, especially when the persecution emanates from a central government.

Yet, when asylum seekers from Iran, China, Eritrea, and elsewhere file refugee claims based on their adherence to a banned religious movement or fear of punishment for conversion, they encounter significant practical challenges in winning refugee protection. When asylum adjudicators set out to decide whether to accept such refugee claims, they can quickly find themselves administering a process akin to a religious trial. In one case, for example, an American appellate court admonished a government attorney for "administer[ing] a sort of mini-catechism." (2) The source of this difficulty is not the substantive legal criteria for refugee status, but rather evidentiary difficulties in convincing adjudicators that asylum seekers are indeed members of a targeted religious group. Other writers call this the problem of the "religious imposter." (3)

This Article examines how adjudicators can address these cases systematically and objectively, ensuring as well as possible that refugees who are genuinely at risk will find protection without endangering the integrity of the asylum system or infringing on the religious liberty of asylum seekers. Part II explains the gap in standards that this Article attempts to address, its analytical approach, and the methodology of the case study that will be presented. Part III provides background on the situation of Pentecostal Eritrean asylum seekers in Egypt and the reasons why they present an ideal case study for examining the religious imposter problem. Part IV examines various approaches to religious-credibility assessment that have been proposed in literature and jurisprudence and attempts to match these approaches to specific types of questions used in RSD interviews in Egypt. In the course of this analysis, this Article reaches conclusions about the best approaches to use in religious-based RSDs, and it concludes in Part V by suggesting how a more systematic framework for assessment of credibility in religious cases might be developed.


A. A Gap in Standards

The religious imposter problem develops when oppression of a particular religious group in a particular country rises to a sufficient level of severity that any person of that nationality who subscribes to the targeted faith is likely to have a well-founded fear of being persecuted, thereby meeting the substantive requirements for refugee status. (4) However, it is difficult, as an evidentiary matter, to determine whether a claimant is actually a member of the persecuted religious group, and states naturally do not want to open their asylum systems to any person who simply declares herself to be a member of an oppressed group. (5) As the New Zealand Refugee Status Appeals Authority wrote, "[i]n the absence of any truly independent evidence, it would be easy to manufacture a claim based on personal religious belief." (6)

Because evidence of persecution in foreign lands is often lacking, refugee cases are commonly determined by assessment of the credibility of an applicant's testimony. (7) If an applicant's testimony is credible, it will be entitled to the "benefit of the doubt," but if the testimony is not credible, the refugee claim will usually fail. (8) Credibility assessment in refugee cases is nearly always a challenge because it requires judging the veracity of a vulnerable person from another country (9) in circumstances in which fraudulent applicants have an incentive to lie (10) and genuine applicants may appear unconvincing. (11) In easier cases, an applicant's testimony can be judged against independent evidence, such as documents (and assessment of their validity), reports about the applicant's country of origin, or the testimony of other witnesses. (12) This process is often called external credibility assessment. (13) However, the most challenging cases are those that rely extensively on affording asylum seekers the benefit of the doubt because of the lack of corroborating or contradictory evidence. (14) In these cases, adjudicators assess credibility mainly by analyzing the applicant's testimony in reference only to itself, looking for consistency, detail, vagueness, and contradictions, among other factors. (15) This process is sometimes referred to as internal credibility assessment. (16)

The process of credibility assessment in RSD encompasses many of the fundamental tensions and conflicts at the core of the refugee system. In theory, the definition of "refugee" should be applied liberally in light of its purpose, but states are not obligated to pursue refugee protection at all costs. (17) Mariham Iskander Wahba, one of the author's students at the American University in Cairo, interviewed refugee service providers in Egypt. (18) Wahba observed that refugee fraud is a pervasive concern in official interactions with refugees in Cairo, and this fear extends beyond a concern that a refugee will lie during status determination. (19) Focusing especially on gender-related concerns, Wahba notes that obtaining monetary, medical, and other services requires refugees to have been victimized, which creates official suspicion that refugees will fake victimhood. (20) For instance, a refugee might lie about being raped in Egypt to obtain services, even after she has passed the test of RSD. (21)

A previous study of the UN High Commissioner for Refugees's (UNHCR) RSD decisions in Egypt found that negative credibility assessment accounted for 77 percent of rejections, and studies of other jurisdictions have found similar patterns. (22) Despite its importance, credibility assessment remains a relatively undeveloped aspect of international refugee law for several reasons. (23) First, in many legal systems, credibility assessment is treated as a question of fact rather than a question of law, and as a result, credibility determinations are not subject to rigorous appellate review. (24) Therefore, standards for credibility assessment do not evolve gradually through jurisprudence to the same extent as other central questions in refugee law. (25) Second, as James A. Sweeney observed recently, legal culture tends to encourage critical thought about evidence and fact-finding less than it encourages the refinement of abstract normative rules. (26) Lawyers are inclined to seek "truth," while contemporary social science has been built, to a great extent, on the understanding that empirical certainty is often unrealistic. (27) Third, for a considerable amount of time, it was normal for credibility assessment to be considered a subjective matter of impression for individual adjudicators, and determinations were made by intuition rather than explicit analysis. (28) Fourth, the development of comprehensive guidelines on credibility assessment is challenging because the specifics of each case are unique, and assessing the believability of refugee testimony is not a simple matter of analyzing answers given by a witness. In real life, credibility assessment involves many more factors, including not just the answers but also the questions, the way the questions are asked, and the environment in which they are asked. Even vague and incoherent testimony may not definitively indicate fraud, because cultural barriers, language and interpretation problems, mental health issues, and the general limitations of human memory and communication can produce honest testimony that nevertheless appears superficially incredible. (29)

The credibility assessment challenge in religious-belief cases is more specific and more perplexing than credibility assessment in other refugee cases. Religious affiliation often cannot be confirmed by independent evidence and must be proved only by an applicant's own testimony. (30) The U.S. Court of Appeals for the Seventh Circuit wrote that "[w]here asylum turns on apostasy, word of conversion is not enough. We need evidence that corroborates this turn from Islam." (31)

However, in the same case, the court said that "[d]etermination of a religious faith by a tribunal is fraught with complexity as true belief is not readily justiciable." (32) Because religion is, by nature, a subjective matter that is beyond the powers of state adjudication, there is considerable confusion and debate about what adjudicators should do when forced to assess the credibility of a religion-based refugee claim. Adjudicators disagree about how to meet the challenge posed in these cases, and some attempts made by officials at the frontlines have been subject to severe criticism later. For example, in 2006, the U.S. Court of Appeals for the Tenth Circuit criticized a lawyer for the Department of Homeland Security for forcing an asylum seeker to answer a series of quiz-like questions on religious doctrine and accused the immigration judge of essentially setting a rigged game:
 To the extent Mr. Yan was unable to answer the questions posed to
 him, some of which were phrased as "trick" questions, the IJ
 concluded that Mr. Yan was not really a Christian. To the extent he
 was able to answer the questions, the IJ concluded that Mr. Yan had
 been coached. (33)

Some decisions focus extensively on an applicant's lack of knowledge about religion and demand that the applicant be able to provide convincing explanations for their religious faith. (34) For instance, a Canadian adjudicator refused to believe that an Iranian woman was Baha'i because:
 She only knew a couple of the important dates on the Baha'i
 religious calendar--no more than could be expected, as the Refugee
 Claim Officer (RCO) observed, for someone who had Baha'i
 relatives[--and s]he was unable to explain the basis of her
 assessment that the Baha'i religion was the best. Her comparisons
 with Islam were vague even though she declared there were many
 differences. (35)

Other courts severely criticize the assessment of an applicant's level of religious knowledge (36) and instead urge attention to an applicant's testimony about his personal religious experiences. (37) Although some adjudicators seem inclined to disbelieve claims of religiosity that would benefit a refugee claim, (38) others are willing to believe these claims based on positive impressions of the applicant's demeanor. (39) Some decisions give asylum seekers significant credit for joining religious organizations in exile, (40) while others regard such activities as self-serving. (41) Some tribunals argue that the central issue is whether a professed faith is sincere or genuine, (42) while other courts hold that sincerity is not the relevant question. (43)

There are no extensive guidelines from either UNHCR or governments that provide adjudicators with guidance about how to conduct such credibility assessment in religious persecution cases. There are three published articles on this subject in American law reviews, but all of the articles focus on particular domestic and political questions that are peculiar to the American legal system. (44) UNHCR's 2004 guidelines on religion-based refugee claims contain six paragraphs of advice on this subject and appear to constitute the most extensive guidance available. (45) One of the challenges facing adjudicators in this field is that, aside from the 2004 UNHCR guidelines, there are few clearly established standards of practice.

B. Defining the Question

The central difficulty of credibility assessment in religious cases is that the underlying question is different than in standard credibility cases. In what could be called the classic form of refugee-credibility assessment, an applicant makes an assertion, and the adjudicator must decide whether to accept it as true for the purposes of RSD. (46) If an applicant describes a past event of persecution, the adjudicator must determine whether to believe that the event actually occurred.
Classic Credibility Assessment

Applicant (A) assertion: A says that "I was in prison."
an event (P) of persecution
happened in the past.

Credibility: We accept that P We accept that she was in prison.

In religious cases, the challenge is that the core assertions by applicants are matters of conscience and are not subject to objective or neutral adjudication regarding their truth or believability. An adjudicator would have to adjudicate the believability of a religion to rule on whether it is true or believable that Mohamed is the last prophet or the Book of Mormon is a sacred text. There is no objective way to decide these questions, and to attempt to decide these questions would interfere with religious liberty by bringing secular adjudication into the world of faith. As a result, it is not immediately obvious what question an RSD adjudicator should ask when a refugee claim turns on religious credibility.
Religious-Credibility Assessment

Applicant (A) assertion: A makes a "God sacrificed his only son."
statement of religious faith (F).

Credibility [not equal to] F is true. We have no means by which to
 judge the truth or believability
 of this statement.

Although this dilemma is especially clear when an applicant makes a statement of faith, it similarly occurs when an applicant makes an assertion of religious identity. The risk lies in the adjudicator attempting to determine whether person X is indeed a member of religion A and relying on an explicit or implicit judgment about what a genuine member of religion A would be like. If an applicant says, "I am a Jew," an adjudicator can judge the credibility of this statement only by making an initial determination about what it means to be a Jew. This is a controversial question, and one that depends on subjective belief nearly to the same extent as a statement of religious faith. Must a Jew have a Jewish mother? Do real Jews always keep kosher and pray on the Sabbath? Who can convert a non-Jew into the religion, and how? The only objective conclusion is that Jews disagree about the answers to these questions, (47) and this disagreement leaves an RSD adjudicator in a difficult position. The adjudicator could call religious scholars to provide expertise, but any such attempt will ultimately position an official of the state between two quarrelling rabbis.

Credibility assessment in these cases requires a government or UN adjudicator--in a situation where the adjudicator's judgment is effectively an assertion of power over an individual--to wrestle directly with the ambiguities of religious identity and faith. This poses a danger that a process intended to protect victims of religious repression may actually impair religious freedom and compromise the state's neutrality. As Tuan N. Samahon warns, "[t]he act of definition entails exclusion or inclusion." (48) If the adjudicator adopts a narrow conception of the alleged religious identity, refugees with a genuine fear of persecution will feel pressure to conform to the adjudicator's orthodox view to avoid deportation. (49) This pressure runs afoul of the rule of international human rights law that religious freedom is not limited to traditional or institutionalized religions, (50) and it arguably violates the prohibition against compelling a person to adhere to a certain form of religious belief. (51)

C. Analyzing Questions in Refugee Status Determination (RSD)

Most training guides and articles on refugee-credibility assessment focus on the way that adjudicators analyze the information provided by asylum seekers. (52) Others focus on surveying the reflections of decision makers. (53) Psychologists have produced clinical studies documenting the gap between evidentiary expectations in RSD and the frailties of human memory. (54)

This case study, however, focuses primarily on the questions asked by RSD adjudicators. In an RSD proceeding, the testimony comes to light during the course of an interview--a two-way process in which the obtained information typically consists of answers to questions asked by adjudicators. (55) The premise of this case study is that it is artificial to examine an adjudicator's analysis of an asylum seeker's answers without also paying attention to the questions asked. It is essential to focus on the questioning strategies that adjudicators use, because the questioning is the aspect of the interview that is most within the adjudicator's control and, therefore, the area where guidance will be most effective.

Dr. Jane Herlihy, a psychologist with the Centre for the Study of Emotion and the Law in London, concludes that adjudicators make assumptions, explicitly and implicitly, when they decide asylum cases, (56) and this Article adopts her conclusion. Herlihy and her colleagues analyze British asylum decisions and use a coding framework to identify and categorize the assumptions made by the decision makers. (57) In an initial survey of thirty RSD decisions, Herlihy found that adjudicators made more than 400 separate identifiable assumptions. (58) In conducting credibility assessments, adjudicators made explicit and implicit assumptions about how they believed foreign governments would behave, how they believed that people and families who experience danger and trauma would act, how genuine refugees would behave in the U.K. asylum system, and what a truthful refugee account should look like. (59) Several of these assumptions contradicted each other, and the assumptions draw on subjective understandings of human behavior. (60) In a similar vein, Hilary Evans Cameron compared Canadian adjudicators' assumptions about how a genuine refugee would behave in the face of a danger of persecution. (61) She wrote that "[t]hese assumptions are not based on any evidence, and yet evidence is close at hand," because social science has developed considerable knowledge about how human beings respond to danger. (62) To combat these problems, Herlihy calls for the development of an analytical methodology to identify those assumptions that should be considered valid based on actual empirical knowledge. (63)

Building on Herlihy's work, this study argues that adjudicators' assumptions will surface and impact refugee adjudications long before the final judgment on the application and before the applicant even answers. The questions asked of asylum seekers should have an underlying implicit logic, and, thus, by asking certain questions and not asking others, a decision maker reveals underlying assumptions about the kind of information that is relevant to the credibility assessment. For instance, if interviewers ask applicants about their knowledge of the Bible, they are acting on the implied assumption that knowledge of the Bible, or lack thereof, proves something relevant to credibility assessment. Analysis of the interview questions reveals what the adjudicator assumes to be true before encountering the applicant's actual testimony.

The ideal study following this approach would be to analyze the questions asked in RSD interviews, the decisions about those applications, and quantitative correlations between types of questions and assumptions and the positive or negative result of refugee applications. However, sufficient data for this more ambitious study is not accessible. As a result, this Article's approach is more modest.

D. Source of Interview Transcripts

The research in this Article was conducted in cooperation with Africa Middle East Refugee Assistance (AMERA), a nongovernmental organization that provides free legal aid to asylum seekers applying for refugee recognition in Egypt. (64) AMERA allowed the author to analyze the transcripts of its clients' RSD interviews at UNHCR on the condition of compliance with the Nairobi Code: Model Rules of Ethics for Legal Advisors in Refugee Cases, which prohibits publication of any unique identifying details about the individuals involved. (65)

In order to understand the nature of the transcripts, it is essential to put them in the context of Egyptian RSD. Egypt is a party to the 1951 Refugee Convention, but it has never established its own asylum system and has no refugee legislation. Instead, as in many countries in Africa, the Middle East, and Asia, the UNHCR conducts RSD. (66) In principle, the Egyptian government grants residence permits to refugees recognized by the UNHCR and also to asylum seekers with applications pending at the UNHCR. (67) Once in the country, an Eritrean (68) can register as an asylum seeker at UNHCR and receive an interim UNHCR identity document known locally as a "yellow card," as well as a renewable six-month residence permit from the Ministry of Interior for as long as the RSD application remained pending for adjudication or appeal. A UNHCR RSD eligibility officer normally interviews asylum seekers and then issues a first instance RSD decision, which recognizes or denies the asylum seeker's status as a refugee. Applicants can appeal rejections within thirty days, but appeal applications typically remain pending for months or years. Recognized refugees receive UNHCR papers known as "blue cards," which entitle them to renewable residence permits. Blue card holders are eligible for UNHCR-provided medical care and other social services funded by the UNHCR. (69)

AMERA's transcripts provide the only record of some RSD interviews because of UNHCR policies. In 2003, for the first time, the UNHCR issued a set of procedural standards governing its own RSD operations. (70) These standards recommended, but did not require, that UNHCR field offices give rejected applicants individualized reasons for rejection in writing. (71) During the studied period, UNHCR's office in Cairo declined to provide written explanations. Instead, the Cairo office gave rejected applicants oral explanations during meetings known as appeals counseling sessions. AMERA legal advisors attended some appeals counseling sessions and made their own record of the reasons given. The absence of written explanations for the decisions made it impossible to replicate Herlihy's approach to British refugee adjudications.

When asylum seekers attend RSD interviews with a UNHCR eligibility officer, a legal advisor may accompany them. (72) There is no independent stenographer or recorder; other than the UNHCR eligibility officer, the applicant, and the legal advisor, the only other person present is normally an interpreter (if one is needed). (73) The eligibility officer types a transcript of the interview, including both the questions asked and the answers given, but UNHCR policy explicitly prohibits providing asylum seekers with copies of their own interview transcripts. (74) This policy has been criticized for violating basic principles of due process and for contradicting UNHCR's own advice to governments. (75) The UNHCR denied access to its transcripts and case files for this research.

In an effort to develop a detailed record of the proceedings, AMERA directed its legal advisors to record their own detailed transcripts of the RSD interviews that they attended with clients. (76) These AMERA transcripts form the raw data for this study. Because most of an RSD interview consists of questions and answers involving the eligibility officer and applicant, the legal advisor can concentrate on making a high quality transcript. The AMERA transcripts are generally highly detailed, including, in many cases, a record not only of the substance of the questions and answers but also the filler words ("uh," "hmm"), half sentences, and grammatical errors that are normal parts of oral communication. They also occasionally reference interruptions, external noises, and other ambient factors in the interviews.

In total, the author reviewed thirty AMERA interview transcripts (the Cairo transcripts) of Eritrean asylum seekers who claimed to be in danger of persecution because of associations with Pentecostalism. Sixteen of the interviews took place in 2007, thirteen in 2008, and one in 2009. All but three (twenty-seven of thirty) were at the first instance stage, while the others were appeals or, in one case, a re-opening of a previously closed file. For confidentiality, the UNHCR cases are renumbered (one through thirty), and, where examples are given, they are referenced by these arbitrary applicant numbers.

The interview transcripts included 1,144 separate questions about religion or religious persecution, which were classified by topic.
General topics Specific categories

Narrative questions Past religious activities
 Religious persecution

Faith and knowledge Theological knowledge
 Theological clarification or dispute
 Religious belief or practices
 Religious motivation

Other topics Religious practice in exile

Each of the narrative and faith and knowledge questions was further categorized based on whether the questioner framed them in an open or closed manner. The precise definition of these categories will be described in more detail later in this Article.

Because it was possible to analyze a large number of questions but only a small number of cases, this Article makes no effort to draw correlations between questions and results. It reports proportions, indicating that one type of question was asked more or less than others. However, this study depends less on statistics than on finding real world examples by which to analyze different approaches to credibility assessment. Therefore, this Article primarily attempts to categorize different types of questions and evaluate their relative utility in order to isolate the interview strategies most likely to produce a reliable and fair determination of credibility.

UNHCR's RSD procedures have been subject to criticism for failing key benchmarks of procedural fairness (77) but have also recently been reformed. (78) In some ways, this analysis holds up aspects of UNHCR-Cairo interview techniques as a positive model, especially the technique of relying on narrative questions. However, this study also raises concerns about some interview techniques used by UNHCR in Cairo, and these concerns may add to the criticism of the UNHCR's RSD operations. During 2007 and 2008, UNHCR-Cairo posted a relatively low recognition rate of Eritrean refugee claims compared to other RSD systems.

RSD for Eritrean Applicants in 2007-2008: (79) UNHCR-Cairo Compared to Selected (80) Asylum States
 Number of
 Convention Grants of
 New RSD Refugees Complimentary
 Applications Decisions Recognized Protection

NHCR-Cairo 1,461 901 497 -
Canada 376 186 165 -
German 710 697 307 109
Italy 5,194 4,127 395 3,482
Malta 515 525 10 461
Netherlands 389 407 31 259
Sweden 1,735 1,792 255 846
Uganda 3,080 2,251 1,150 492
United Kingdom 4,250 4,150 2,355 145
United States 588 575 352


NHCR-Cairo 55%
Canada 89%
German 60%
Italy 94%
Malta 90%
Netherlands 71%
Sweden 61%
Uganda 73%
United Kingdom 60%
United States 61%

However, this data reflects only a rough measure (81) of RSD tendencies at a particular point in time. UNHCR-Cairo claims that it made efforts in late 2008 and early 2009 to improve credibility assessment in religious cases. (82) In the first half of 2009, UNHCR-Cairo reports that approximately 85 percent of Eritrean asylum seekers were recognized, a marked increase from 2007-2008. (83) However, it is difficult to pinpoint a single cause for this change. Moreover, this study was not designed as an assessment of the quality or fairness of UNHCR's RSD procedures per se, but rather uses the case study of a particular group of asylum seekers at a particular UNHCR office as an example of the wider challenge that confronts any RSD system.


A. Pentecostalism in East Africa

Pentecostal Christianity is notably difficult to define, which makes the task of determining the credibility of someone's claim to be Pentecostal more daunting. The movement's origins are traced to early twentieth-century America, beginning in Topeka, Kansas in 1901s4 and moving in 1905 to Los Angeles, where Pentecostalism spread from a revival meeting on Azusa Street (known as the Azusa Street Revival). (85) A century later, Pentecostals may constitute as much as one-quarter of all Christians worldwide, second only to Catholicism. (86) The fastest areas of growth of Pentecostalism are in Africa and East Asia. (87) Theologically, Pentecostalism appears most closely related to evangelical Protestantism, (88) but it includes members of Catholic and Orthodox denominations as well. (89) This Article uses the term Pentecostal somewhat loosely to include a range of Christians who are variously called "renewalists," "charismatics," and "spirit-filled movements." (90)

The origin of the name "Pentecostal" is in the book of Acts, chapter 2, which tells how, on an autumn harvest festival fifty days after Passover, believers were filled by the Holy Spirit, began to speak in tongues, and received a message that in the end times the Spirit would lead men and women to prophesy. (91) If there is a single unifying belief among the diverse strands of the Pentecostal movement, it is faith "in the spiritually renewing gifts of the Holy Spirit"; for this reason, the general movement is sometimes called "renewalist." (92) Part of the challenge of defining the Pentecostal movement is the fact that these core beliefs are evident in other streams of Christianity as well. (93) Although one can identify themes that form the theological foundations of the movement, (94) it is difficult to isolate a definitive specific belief or practice that distinguishes Pentecostalism, aside from a focus on the presence of the Holy Spirit in modern-day believers. (95) Many beliefs are very common among Pentecostals, but few beliefs are universal. For example, in a ten-country survey, more than 80 percent of Pentecostals and renewalists said they believe in the Rapture, when the Christian faithful will be rescued to heaven before the end of the world. (96) Large majorities believe in modern-day miracles (97) and in the active intervention of angels and demons in the world. (98) However, not all Pentecostals hold these beliefs; many reported that they never speak or pray in tongues, even though this is often perceived as a definitive feature of Pentecostalism. (99)

The Azusa Street Revival led to two new denominations, the Assemblies of God and the Church of God in Christ, but tens of thousands of other Pentecostal groups (many of which are independent local churches) have emerged in other countries, and these churches lack formal affiliation to a larger organization or denomination. (100) Pentecostalism arrived in Eritrea in the late 1970s, during the rule of the Ethiopian Dergue regime. (101) Ethiopian Pentecostal missionaries came to Asmara to preach, (102) but the resulting movement includes a number of churches and groups that vary in approach. Some renewalists and charismatics do not form new religious organizations at all. The Pew Center explains:
 Charismatics share many of the experiences that are distinctive
 to Pentecostalism but remain members of mainstream Protestant,
 Catholic and Orthodox denominations. This movement, sometimes
 referred to as "second wave" pentecostalism, emerged in significant
 numbers in the 1960s as part of what its members considered to be a
 much-needed spiritual renewal with these older churches. (103)

Scholars doubt the utility of applying specific labels to a movement that evolved so rapidly. (104) Around the world, charismatics tend to affiliate with the Christian denominations most prevalent in their countries. (105) As a result, the movement eschews the classic Catholic-Orthodox and Catholic-Protestant distinctions that typically categorize Christianity. The result is a movement that "is difficult for even religious scholars to describe." (106)

B. Religious Repression in Eritrea

In 2007, Eritreans submitted more individual asylum applications around the world than any other nationality except Iraqis and Somalis. (107) In 2008, Eritreans filed more asylum claims than Iraqis; only Zimbabweans filed more applications. (108) The scale of the Eritrean search for asylum is notable not just for its gross size but also for its proportionality to the small total population of the country.
New Asylum
Applications TOTAL Country
Global 2007 2008 2007-2008 Population

Zimbabweans 20,800 118,500 139,300 11,392,629
Iraqis 52,000 43,900 95,900 28,947,657
Somalis 46,200 51,900 98,100 9,832,017
Eritreans 36,000 62,700 98,700 5,647,168

Amnesty International summarizes some of the reasons for this mass exodus:
 Two-thirds of the [Eritrean] population remained dependent on
 international emergency food aid. The government did not allow
 opposition parties, independent civil society organizations or
 unregistered faith groups and tolerated no dissent. Thousands of
 prisoners of conscience were held. There was no recognizable rule
 of law or justice system, civilian or military. Detainees had no
 means of legal redress and judges were unable to challenge or
 question arbitrary detentions or government or military actions
 violating human rights. Constitutional and legal protections of
 human rights were not respected or enforced. (109)

Eritrea is a religiously diverse country, described by Amnesty International as "a highly religious population, with some 98% of its 3.7 million people belonging to a long-established branch of a major world religion." (110) Estimates differ as to Eritrea's precise demography, but the vast majority of the population is affiliated either with Sunni Islam or the Eritrean Orthodox Church. (111)

In 1995, Eritrea's Proclamation on Religious Organizations required all religious organizations to register with the government. (112) The government initially allowed registrations for only Sunni Islam, the Eritrean Orthodox Church, the Roman Catholic Church, and the Lutheran Church, which is the only Protestant denomination to be officially recognized in Eritrea. (113) The government postponed action on all other registration requests. (114) Not all religious groups submitted registration applications. (115) According to Eritrean law, religious groups seeking registration must prove the "uniqueness" or benefit that their religion offers relative to other religions and disclose their sources of funding from abroad, as well as the names of the religious groups' leaders. (116) The government judges the applications on the basis of whether they conform to local culture. (117)

In 1997, the National Assembly approved a new constitution that provides for freedom of religion, (118) but the new Constitution has had little apparent impact on the policies of the Eritrean state. Until 2002, the government tolerated unregistered religious groups to some extent, often allowing them to hold prayer meetings in private homes. (119) However, in May 2002, the government decreed that all unregistered religious groups must cease operations until they are permitted to register. (120) From that day forward, only the four registered religions have been able to operate openly. (121) Since 2002, the government has closely monitored the activities of unregistered religious groups, the members of which are often arrested and held without charge. (122) There were estimates that by the end of 2008, the government had detained more than 3,225 Christians from unregistered groups. (123) Therefore, close to one in ten members of dissident Christian churches in Eritrea were in prison at the end of the year, given that these churches are estimated to account for only 2 percent of the population. (124) In addition to the Pentecostal and unregistered evangelical groups, Amnesty International reports that authorities have targeted a Bible-study group within the Eritrean Orthodox Church called Medhane Alem (Savior of the World). (125) Some Muslim dissidents (126) and the head of the Eritrean Orthodox Church have been detained despite membership in recognized religious groups. (127)

Detainees reportedly endured brutal treatment, including being held for long periods in crowded metal shipping containers (128) or underground cells. (129) Some were reportedly subjected to a torture technique known as "the helicopter," (130) during which a person is forced to lie on his stomach with his hands clutching his feet for hours or days until he or she agreed to not participate in the banned religion again. (131) The shipping containers were "often swelteringly hot in the day and very cold at night, with no sanitary facilities and very poor food provided." (132) Amnesty International reports that detainees were often forced to sign documents agreeing to stop worshipping. (133) "Some were reportedly forced to recant their faith and agree to rejoin the Orthodox Church." (134)

The Government of Eritrea is not theocratic in the manner of Iran, Saudi Arabia, or Spain at the time of the Inquisition, and its repressive policy cannot be explained by any particular religious orientation of the state. The government has both authorized and repressed a wide variety of religious groups. (135) Officially recognized national holidays include several Christian festivals (Christmas, Epiphany, Good Friday); Meskel, which is specific to the Ethiopian and Eritrean Orthodox Churches; and several Muslim holidays (Eid al-Adha, the Prophet's Birthday, and Eid al-Fitr). (136) Most sources report that Jehovah's Witnesses have been treated most harshly since 2002, mainly because of their objection to military service. (187) However, this explanation cannot explain the other targeted religions, which do not object to military service. (138) Amnesty International reports that evangelical movements have grown rapidly in Eritrea since the 1990s and have received converts from the registered religions, and these conversions have led to conflict with the main religious groups. (139) Amnesty International asserts that the targeting of religious minorities "reflected the government's general repression and intolerance of freedom of opinion and association. The government appeared to be punishing any kind of expression of dissent, religious or political." (140)

C. Eritrean Refugee Claims Under International Refugee Law

This study assumes that if a decision maker accepts as fact that an Eritrean claimant is in fact a Pentecostal Christian or is otherwise a dissenter from the four registered religions, the decision maker would likely conclude that the claimant has a well-founded fear of being persecuted for his religion and is thus a Convention refugee. (141) The Eritrean policy that limits religious practice to only four religions is a fairly straightforward violation of international human rights law. (142)

In Li v. Gonzales, an outlying decision by the U.S. Court of Appeals for the Fifth Circuit, the court considered a man who had been tortured for belonging to an unregistered church in China. (143) Li, the asylum seeker, did not want to register his church because doing so would give the Chinese government control and force endorsement of socialism. (144) Reasoning that Li was punished for violating the religious registration law rather than for religious reasons, the Board of Immigration Appeals (BIA) overruled an immigration judge's grant of asylum. (145) The U.S. government insisted that China's aim was to maintain social order, not to harm religion per se. (146) The court of appeals initially upheld the BIA's decision: "While we may abhor China's practice of restricting its citizens from gathering in a private home to read the gospel and sing hymns, and abusing offenders, like Li, who commit such acts, that is a moral judgment not a legal one." (147) Widespread outcry and the intervention of the U.S. Commission on International Religious Freedom persuaded the government to reverse its opposition to granting Li asylum, and the court of appeals vacated its original judgment. (148) The Li judgment appears to be an aberrational error that was corrected, and, in any case, it is out of line with most international jurisprudence on the issue. (149)

International human rights law does not prohibit requiring religious organizations to register per se, provided that the registration system is subject to law, does not have the aim or effect of creating discrimination, is subject to judicial review, and maintains the state's neutrality in matters of belief and practice. (150) However, the UN Special Rapporteur on Freedom of Religion or Belief observed an increasing tendency of states using administrative regulations to restrict the exercise of freedom of religion. (151)

Distinguishing religious persecution from persecution for lack of religious registration appears to be a distinction without a difference. In Hasan v. Bulgaria, the European Court of Human Rights held that the organization of religious communities is often just as much an expression of religious faith as more abstract matters of spirituality. (152) Therefore, restrictions by the state on religious organizations may be infringements on the right to freely manifest one's religion. (153) The Court held that a state must remain neutral on matters of religion, "meaning that the right to freedom of religion as guaranteed under the Convention excludes any discretion on the part of the State to determine whether religious beliefs or the means used to express such beliefs are legitimate." (154) The duty of neutrality forbids states from favoring one religious leadership over another and prohibits using state authority to prevent religious groups from dividing or splitting from each other. (155) Elsewhere, the European Court has warned that "the State's duty of neutrality and impartiality is incompatible with any power on the State's part to assess the legitimacy of religious beliefs or the ways in which those beliefs are expressed." (156) The UN Human Rights Committee has expressed concern about discrimination against religions or beliefs based on "the fact that they are newly established, or represent religious minorities that may be the subject of hostility on the part of a predominant religious community." (157) Official recognition of a religion or religions should not lead to impairment of rights to minority religions or nonbelievers, (158) and there can be no restriction on the "right to replace one's current religion or belief with another." (159)


In the Cairo RSD interviews, the questions typically fell into two broad categories. The first type of questions were narrative, asking applicants to describe their experiences. These questions addressed religious experiences, but instead of focusing on abstract belief, many of the questions focused on concrete actions, activities, and tangible interactions with other people. UNHCR's guidelines favor narrative questions, (160) and they accounted for 66 percent of the questions in the sample. (161) By contrast, 34 percent of the questions focused on abstract theology or belief. (162) Interviewers asked applicants to describe their religious beliefs and practices, to articulate their religious motivations, or to display knowledge of their professed religion. These questions reflect an approach to religious credibility known as the sincerity test. I will focus on this second set of questions first because it has a long history and remains widely accepted, though I will argue that it is of limited utility.

A. The Sincerity Test

The sincerity test has roots in the U.S. Supreme Court's 1944 decision in United States v. Ballard and is a common means of avoiding the problem of defining religion or ruling directly on the truth of religious beliefs. (163) Ballard did not deal with refugee protection, but it dealt with an analogous attempt by the government to adjudicate religion-based behavior without judging religious faith. The Ballard family was prosecuted for mail fraud for recruiting people to contribute money to a purportedly religious group called the "I Am" movement. (164) The prosecution believed that the Ballards were con artists who peddled quasi-religious faith as a scam. (165) However, by accusing the Ballards of fraud, the trial could have been construed as an adjudication of whether the "I Am" movement was a false religion. To avoid this problem, the jury was instead asked to decide whether "these defendants honestly and in good faith believe[d] those things." (166) The Court upheld this approach as a legitimate means of avoiding judicial inquiries into questions of theology. (167) Under Ballard, therefore, if the government can prove that a preacher does not sincerely believe what he is preaching, it may punish the preacher for conning people into donating money.

Some scholars suggest the sincerity test as a useful way of framing the credibility issue in religious-based refugee cases. (168) In theory, if an applicant states that "God sacrificed his only son," one need not investigate the truth or believability of this statement to determine whether the applicant genuinely believes it to be true. One can instead assess whether the applicant believes the statement to be true, which is potentially a more objective and neutral inquiry.

Sincerity Test
Applicant (A) assertion: A makes a "God sacrificed his only son."
statement of religious faith (F).

Credibility = A genuinely believes We accept that she believes that
F. God sacrificed his only son.

This approach is based on an underlying understanding of refugee status that boils down to the following logic:

* All members of X religion have a well-founded fear of being persecuted.


** If A is sincere in her belief in X, she has a well-founded fear of being persecuted.

** If A is not a sincere believer in X, she is not a refugee.

Following this logic, an adjudicator would need to develop an interview and assessment technique that provides a basis for deciding whether A is sincere in her claim to belong to X.

B. Neutral Questions on Religious Beliefs and Practices

One type of question that clearly focuses on testing religious sincerity is a query about what a person believes or how she practices her religion. An applicant's ability to provide such explanations is sometimes cited as bolstering her credibility, and in some cases it makes a profound impact on decision makers. For example, in a New Zealand decision, the court found a family credible even though the family's previous application was fraudulent:
 All of the adult appellants gave consistent evidence. They did
 not dissemble. They each spoke at length about their own attraction
 to Christianity and what it meant for them. They acknowledged their
 earlier deceit with contrition, but inviting consideration that
 their earlier, 'pre-Christian' selves were responsible.

 The involvement of the appellants in the church has been intense.
 It dominates their lives on a daily basis. Such wholesale,
 sustained commitment to the church is probably genuine. (169)

By the same logic, an inability to explain one's religious beliefs and practices might make an applicant appear less sincere.

This category of questions is distinguished from two others that also address religious beliefs and practices. Unlike the category of questions that this Article calls "theological clarifications or disputes," this form of question maintains neutrality toward the answer, implying no sense of confusion, disapproval, or disagreement. This type of question asks a person to describe their beliefs and practices in the abstract. Moreover, these abstract questions are different from those that seek out a narrative of religious practice. A narrative question could focus on the same subject but would ask specifically about concrete actions rather than abstract beliefs, focusing specifically on what a person has done, how they have done it, etc.

Of the questions asked in the Cairo transcripts, 12 percent qualify as neutral questions about religious beliefs and practices, although in general these questions were used only a few times per interview, and the exchanges in this category were typically brief. (170) For example:

Q: How can a person join the Pentecostal church?

A: First, one has to believe in Jesus.

Q: And?

A: When he believes, he has to be baptized. This means you die as a sinner and are reborn as a holy person. (171)

In order to use the substance of these questions and answers to assess credibility, an adjudicator would need a preconception about the beliefs held by sincere followers of the religion, and it is difficult to pin down the distinguishing beliefs of Pentecostals relative to other Christians. (172)

To the degree that genuine beliefs can be defined, these questions also contain a weakness similarly inherent in religious knowledge quizzes: they press only for abstract information, and an insincere person can learn the right answers prior to an interview. (173) For instance, an interviewer could ask a Muslim a neutral but abstract question about religious practice, such as "how many times a day do you pray?," and the person could answer, "I should pray five times." However, many non-Muslims know that Muslims pray five times per day. If the interviewer follows up by asking "how many times do you actually pray?," the answer would reveal relatively little because many genuine Muslims do not pray five times a day every day even if they believe that they should.

C. Religious Knowledge Quiz

Among the most common and controversial means of discerning religious sincerity is to ask a question or line of questions that tests whether a witness knows certain information about a religion. Well-Founded Fear, a 2000 documentary about the U.S. asylum system, (174) contains a vivid example of this type of question. An American official asks a self-proclaimed Anglican if she knows the name of the head of the Church of England. (175) The Anglican is first rejected and later granted asylum because of confusion about the accuracy of her answer. (176) Similar questions were commonly asked of purported Pentecostal applicants at UNHCR's office in Cairo. The following series of questions from the Cairo transcripts illustrates the form:

Q: Do you know what the name Pentecostal means?

Q: Was Pentecostal mentioned in the Bible?

Q: Can you tell me the verses?

Q: Do you remember what was mentioned in these verses?

Q: How many Apostles gathered?

Q: So what happened when the Apostles were gathered?

Q: Why did they gather?

Q: Do you know who Michael and Gabriel are?

Q: Do you know who is Saint Paul?

Q: Can you tell me how many letters Saint Paul wrote?

Q: Do you know who wrote the Book of Acts? (177)

Q: What is baptism?

Q: What is the relevance of baptism?

A: The Bible says you have to believe and then get baptized.

Q: Where does it say that?

Q: Who wrote the Corinthians?

A: Saint Paul.

Q: How many others did he write?

Q: Who else wrote in the Bible, other than John?

[After the applicant volunteered three more names, the interviewer asked four more follow up questions seeking additional names of authors of the Bible.] (178)

Questions that tested knowledge of specific religious information constituted 11 percent of the total questions analyzed for this study, (179) but the use of these questions was notably inconsistent. In a few cases, knowledge quizzes appeared to be a major focus of the interviews; five or more knowledge questions were used in seven different transcripts. In one case, the interviewer asked twenty-one separate knowledge questions, (180) and in another case, a different interviewer asked seventeen knowledge questions. (181) However, no knowledge questions appeared in six of the transcripts, and in five cases only one or two knowledge questions appeared. Statistically, it appears that UNHCR-Cairo did not rely extensively on knowledge-based questions in most interviews, but interviewers relied heavily on knowledge quizzes in a significant minority of cases.

Knowledge questions resemble a quiz or test because of the way in which they were used in the interviews. When used heavily, knowledge questions frequently appeared consecutively, giving the appearance of an oral exam. In six cases, interviewers asked at least five religious knowledge questions in a row. In the most extreme case, an interviewer asked separate consecutive series of ten questions, five questions, and four questions in a row at different points in the course of the interview. (182) In total, the interview contained twenty-one knowledge questions. (183)

One question--What is the origin or meaning of the word "Pentecostal" or the Pentecostal faith?--was asked in some form in fourteen of the transcripts, (184) and no other specific question was repeated nearly as often. Yet the themes and forms of the questions were fairly consistent, especially in their focus on specific Biblical sources for religious beliefs or practices. Applicants were asked about the rules governing baptisms; about the names, number, and works of the Apostles; and about particular angels. (185) The form of the questions was consistent. The interviewer asked the applicant for a specific substantive piece of knowledge and then asked for the specific scriptural source of the knowledge. Applicants were often pressed to relate the chapter, verse, and content of the reference:

Q: Do you have any idea about who established the Pentecostal faith in the world?

A: I don't know who established it. The only thing I know is that it's from the Bible, from the Apostles.

Q: Where was the word Pentecostal in the Bible?

A: Book of Acts.

Q: Where in the book of Acts exactly?

A: Chapter 2, from verses i to 3.

Q: What are these verses talking about?

A: It was a feast of fifty days that the Holy Spirit falls or comes upon the disciples.

Q: What else is mentioned in this chapter? (186)

The transcripts suggest that this frequent demand for specific chapter and verse citations intimidated some applicants. For example, in one interview, the applicant stated that she converted after a friend preached the Bible to her. The interviewer followed up by asking for a specific Bible reference, leading to this exchange:

Q: Which part of the Bible did she, I mean your friend, preach to you?

A: The Book of John.

Q: Do you remember which part?

A: I couldn't remember which part, but it was around Chapter 3.

Q: What was the chapter about?

A: You want me to tell you one verse or the whole chapter?

Q: What you know. I don't want you to worry. It's okay if you don't remember.

A: I am very stressed.

Q: Okay, are you willing to continue the interview today or do you want to reschedule?

A: I am not in a good mood now.

Q: I am here for you. You have the choice to reschedule or take another break.

A: I know I make you tired, but maybe we can reschedule. (187)

Knowledge quizzes ask for objective information about externally verifiable facts, and the answers, therefore, do not require actual religious belief. The advantage for adjudicators is that knowledge quizzes are objective and do not delve into subjective matters of faith. Moreover, the same set of knowledge questions can be asked of all asylum seekers because the questions are easily standardized. This objectivity exists only on the surface, however. For knowledge quizzes to have relevance in RSD, the adjudicator must make an assumption that a certain type of religious person would necessarily know a certain piece of information and an insincere applicant would be unlikely to possess the knowledge. If this assumption is true, a decision maker could conclude that lack of knowledge is a factor indicating lack of credibility. A decision maker need not believe that this assumption is always true; knowledge quizzes could be used as one factor among many rather than as the decisive factor. Nevertheless, heavy reliance on an assumption about what religious people usually or always know is inescapable. Without such an assumption, knowledge quizzes would be irrelevant.

Unfortunately, the underlying assumption is itself difficult to prove. An adjudicator would have to possess an empirical study showing that the relevant group of people usually or always knows a certain thing. Moreover, not every religion prioritizes abstract knowledge in the same way, and not every religious person is drawn to a particular religion for the same reason. Thus, knowledge quizzes involve a significant risk of a false negative in credibility assessment: the risk of errantly judging a genuine applicant to be insincere because she does not know a particular piece of information.

Much of the UNHCR's guidance about credibility assessment in religious persecution cases warns against the over-use of knowledge quizzes:
 While decision-makers will often find it helpful during research
 and preparation to list certain issues to cover during an
 interview, extensive examination or testing of the tenets or
 knowledge of the claimant's religion may not always be necessary or
 useful. In any case, knowledge tests need to take account of
 individual circumstances, particularly since knowledge of a
 religion may vary considerably depending on the individual's
 social, economic or educational background and]or his or her age or
 sex .... Women, in particular, are often denied access to religious
 education. 188

The UNHCR also warns that the repression of religious practice may impede adherents' ability to learn about their faith; as a result, people targeted for persecution may actually have less religious knowledge than other members of their religion. (189) This warning is especially pertinent to Eritrea, where churches are closed and Pentecostals are often arrested while attempting to hold Bible-study sessions in secret. (190)

Knowledge quizzes also contain a danger of false positives by aiding fraud. As UNHCR's guidelines warn, "a claimant's detailed knowledge of his or her religion does not necessarily correlate with sincerity of belief." (191) Knowledge quizzes ask about abstract information, such as the structure of a church or the content of a section of scripture, rather than personal experience. This knowledge can be learned and does not require a fraudulent applicant to invent a coherent and detailed story from scratch. Once it becomes known in a migrant community that adjudicators are asking questions of this nature, fraudulent asylum seekers can prepare for credibility assessment by studying their assumed religion. One need not have genuine faith to answer questions about another person's religion; an atheist can learn about Christianity, a Christian can learn about Hindu gods, and a Buddhist can learn to answer knowledge questions about Islam. In general, knowledge quizzes may simply favor educated people who know how to find and learn abstract information and who thus perform well on a test.

D. Theological Clarification or Dispute

A more worrisome variation on the religious knowledge question occurred where, in the context of asking about religious beliefs or knowledge, the interviewer sought clarification about a religious issue. These exchanges usually began innocently and were usually brief. Only in seven cases out of thirty did the interviewer ask more than two theological clarification questions, and in twelve out of thirty cases they asked none at all.

In one fairly innocuous example, (192) an applicant described how she converted from the Orthodox Church to Pentecostalism. When asked her religion, she said that she was Christian, and when asked how long she had been a Christian, she answered since the time of joining the Pentecostal movement. (193) The applicant thus equated Christian identity exclusively with Pentecostalism. When asked about her religion before conversion, she answered "Orthodox" but avoided the Christian label. (194) This answer appeared to confuse the interviewer, who asked her to confirm whether both she and her husband "were Christian before [conversion]," to which the applicant avoided saying yes and instead repeated her explanation of conversion. (195) The interviewer was attempting to clarify and confirm a potentially confusing chronology of events. The interviewer also asked for the clarification only one time, and once the interviewer understood the applicant's use of words she accepted the applicant's phrasing without further inquiry. (196) Similar inquiries for clarification occurred in other exchanges. (197) Seeking clarification is often essential to conducting a coherent interview, and it generally benefits applicants by lessening the risks of misunderstandings. Ideally, interviewers seeking clarification will ask applicants to explain their own beliefs in their own words and will not appear outwardly to dispute the applicant's beliefs.

However, in the minority of cases where several questions were asked that sought clarification on matters of faith, the tone sometimes shifted away from neutrality and the interviewer appeared to doubt or debate the asserted beliefs of the claimant. (198) The difference between seeking clarification and religious disputation is a matter of degree and interpretation, and the interviewer's intentions were not necessarily hostile. However, religious disputation could cause interviewers to seem provocative or hostile, which could lead applicants to feel hesitant about speaking. In several cases, interviewers asked applicants to explain why they left their previous churches for Pentecostalism. When the applicants gave general statements about a desire to be saved or to connect with God, the interviewers asked why they could not achieve these objectives in other churches. (199) Presumably, the intent of such questions is to probe for a compelling statement of genuine belief, but the questioning can appear to communicate skepticism or disapproval on the part of the interviewer about the conversion.

In one example, (200) the interviewer repeatedly invoked her own religious beliefs ("I am Christian and I don't want to intimidate you") and then engaged in a highly personalized argument with the applicant about whether speaking in tongues should be relevant for Christians today. (201) The interviewer pressed the applicant to explain why she had to turn to Pentecostalism in order to accept Jesus as her savior since "he is the savior of other beliefs, too." (202) When the applicant spoke of believing that there should be no mediator between herself and God, the interviewer objected that there are Biblical examples of mediators, and opened a Bible as she listed several. (203) In the same transcript, the interviewer asked the applicant why Pentecostals rejoice that Jesus died. (204) Evidently exasperated, the asylum seeker eventually said, "That's what I believe and no one can change it ... I'm telling you the verses, but you still don't believe me." (205)

This style of questioning appeared at this extreme level in three cases out of thirty, all with the same interviewer, and was clearly not the normal approach at UNHCR-Cairo. UNHCR-Cairo also stated that their approach more recently has been to avoid asking theological questions and to allow applicants to describe their beliefs in their own words if they desire to do so. (206) In the three cases at issue, the interviewer clearly lost any pretense of neutrality toward religion, repeatedly invoking her own faith (and that of the lawyer and interpreter). (207) Although these interviews were abusive in effect and offensive in tone, the interviewer may have been attempting to probe religious sincerity and may not have had any actual hostility toward the applicants' faith. In one case, the interviewer apologized when the applicant seemed offended by a question and said, "I must provoke you. That is my job. I have to provoke you to find out if you are Pentecostal." (208) If a genuinely religious Pentecostal would react in a certain way to certain types of provocative religious debate, staging provocations would be a logical application of the sincerity test.

E. Weaknesses of the Sincerity Test

Any application of the sincerity test must be built on an assumption about how a genuine believer would answer a particular question or act in a particular circumstance. Knowledge quizzes, for instance, are based on assumptions about what a sincere believer would know. However, because it is difficult to prove how a genuine religious person would behave or speak, the sincerity test may not offer quite as many advantages as promised. (209)

In Ballard, Justice Jackson dissented, arguing that focusing on good faith belief did not really free a court from investigating the truth of the underlying belief:
 I do not see how we can separate an issue as to what is believed
 from considerations as to what is believable. The most convincing
 proof that one believes his statements is to show that they have
 been true in his experience. Likewise, that one knowingly falsified
 is best proved by showing that what he said happened never did
 happen. How can the Government prove these persons knew something
 to be false which it cannot prove to be false? If we try religious
 sincerity severed from religious verity, we isolate the dispute
 from the very consideration which in common experience provides its
 most reliable answer. (210)

The facts of Ballard demonstrate a root weakness in the sincerity test. Even if one assumes that the Ballards acted in bad faith, religious fraud is different than other frauds. In a classic fraud case, a person makes an assertion similar to an asylum seeker's claim of a past event of persecution. (211) To say "I was in prison" is similar to claiming "the money is in my account" or "I own the title to this land." With statements such as these, the asserted fact must either be objectively true or objectively false. Good faith mistakes are possible, but ultimately, the ownership of title in a piece of property is not a matter of personal belief. Similarly, a person was either in prison, or she was not; the money is in the bank account, or it is not. Therefore, a judge or jury can assess good faith belief about these facts by asking standard questions, such as:

Q: What information did the applicant have access to before making the assertion?

Q: What should the applicant have known?

Q: Did the applicant's actions correspond rationally to the assertion that she made?

These questions yield fruit because the asserted fact is knowable, and it is therefore possible to investigate whether a person knew it. However, these inquiries yield little benefit in cases of religious fraud because (to paraphrase Justice Jackson) the assertion is not something that can be proven true or false based on available information. (212) It was always a matter of faith, and the government is actually prohibited from investigating the reasonableness of faith. (213)

The most tempting strategy to overcome this objection is to judge assertions against conduct, rather than against the way the applicant answers questions. In rare cases, this strategy may be effective. For instance, if a person says that she is a religious Christian, but she never goes to church, never prays, never reads the Bible, and never engages in any conduct remotely evocative of Christianity, it may be possible to conclude that her claim of religious belief is insincere. The opposite is also true; if she engages in Christian religious practice, it should be more believable that she is actually Christian. For instance, in a New Zealand case, the Refugee Status Appeals Authority concluded that an asylum seeker was a genuine convert in part because he was "a serious and dedicated member of an Evangelical Christian church in New Zealand." (214) In this case, and in several of the examples described in the Introduction, an applicant's participation in religious activities (or lack thereof) in the country of asylum carried significant positive or negative weight with judges. (215) However, other tribunals are reluctant to place much reliance on religious practice in exile for fear that applicants will pretend to practice in order to strengthen asylum claims. (216)

Using behavior to assess the sincerity of belief can also lead to problematic results. For example, the Australian Refugee Review Tribunal rejected a Ukrainian man's claim of homosexuality in part because he also claimed that he was a Roman Catholic, concluding that a genuine homosexual would reconsider his Catholic faith. (217) The Catholic Church might not even accept this logic, (218) but the nuances of Church teachings are not the central issues. The basic problem is objective and empirical. Unless the adjudicator can prove empirically that there are no Catholic homosexuals, he cannot say that professed Catholicism discredits a claim of homosexuality, or vice versa. Sincerity of religious membership "is not easily judged by outward criteria" (219) because religions often strive for self-improvement and tend to focus on the conflicting impulses of the human condition. (220) Insisting that asylum seekers conform perfectly to religious doctrine assumes that a sincerely religious person will not assimilate to society or engage in any kind of heterodoxy, and this assumption is ultimately no more empirically accurate than an assumption that all genuine Catholics go to confession and oppose birth control.

The sincerity test will be, in most cases, a trap of circular logic that ultimately returns to the original conundrum of religious credibility. It purports to measure a person's statements of religious belief against her actual beliefs. The sincerity test can work only by making assumptions about how a religious person would talk or act, but ambiguity, ambivalence, conflicted deeds and words, and apparent incoherency (especially in the perception of outsiders) are all regular parts of religious experience. Moreover, by imposing a preconception about religious experience, adjudicators deprive asylum seekers of the freedom to experience religion through individual choice. (221) Inquiring about pious conduct and knowledge opens the door to disciplined frauds while casting suspicion on genuine but imperfect believers: as Samahon observes, "[a] well-practiced imposter will impersonate the perfectly orthodox, straight-arrow religionist whereas actual converts may still be learning the formalities of their newly chosen faith." (222)

Judges continue to disagree about the appropriate approach. For instance, in a split decision concerning an Iranian who outwardly practiced Christianity in the United States, the Ninth Circuit Court of Appeals rejected the asylum claim for insincerity. (223) Although the applicant, Toufighi, regularly attended church services in the United States, the immigration judge concluded that he was not genuine in his conversion to Christianity because he did not know the names of the twelve apostles. (224) The immigration judge wrote:
 With the Court's understanding that Christianity begins with the
 life and teaching of Jesus Christ in the New Testament, the 12
 apostles have some of the most important, if not the most
 important, writings of Christianity.... The respondent's knowledge
 about Christianity [was presented] to the Court in such general
 terms that any person of any religion can come up with that
 description of their religion, namely peace, tranquility, and
 love. (225)

Judge Singleton and Judge Ikuta of the Ninth Circuit upheld the denial of asylum, but, because of the standard of review in administrative law, they did not engage directly with the analysis that led to the negative credibility assessment. (226) Judge Berzon issued a dissent, arguing that the immigration judge asked the wrong question: "In short, the question is not what Toufighi believes but what Iran understands him to believe--or, more accurately, not to believe. It is thoroughly plausible that because he attends Christian services and belongs to a Christian church, Toufighi will be taken to have renounced Islam." (227)

On the one hand, one might question, as an empirical matter, how many of the world's Christians can actually name all of the apostles, and whether this list is necessarily the most important aspect of the life and teachings of Jesus. On the other hand, if Toufighi had recited the list correctly, one might object that, just as a person of any religion can describe devotion to general principles, any educated person can memorize twelve names and thus fake genuine Christianity. The hesitancy of some adjudicators to abandon the religious sincerity test is understandable because it seems that disregarding sincerity implicitly ignores the core state concerns about religious imposters by allowing asylum seekers to fraudulently win recognition as refugees. However, there is little reason to assume that the sincerity test really keeps out imposters because it actually opens the door to clever fraudsters who can effectively learn to look and talk like a genuine believer.

F. The Eye of the Persecutor Test

The alternative approach, which Judge Berzon suggested in Toufighi, proposes that refugee status does not depend on the genuine religiosity of refugee claimants. (228) Rather, refugee status is defined by the motives of persecutors. (229) As the Supreme Court of Canada held, "[t]he examination of the circumstances should be approached from the perspective of the persecutor, since that is the perspective that is determinative in inciting the persecution." (230) If a persecutor believes a person to be a member of a targeted faith, it matters little if the targeted person is a true believer. (231) This basic principle reframes the religious-credibility assessment and avoids forming decisions based on the credibility of the asylum seeker's subjective beliefs. The sincerity approach may still affect a judgment: if an asylum adjudicator believes the applicant to genuinely be a member of a prohibited religion, it is reasonable to assume that the persecutor would reach a similar conclusion. Moreover, an applicant's religious commitments would aid in assessing her likelihood of engaging in prohibited activities upon her return to her home country. Under this logic, the ultimate goal is to anticipate the perceptions of the persecutor, rather than to directly judge religious sincerity. (232)

In United States v. Bastanipour, an asylum seeker claimed that he had converted from Islam to Christianity in prison in the United States, and he feared execution as an apostate if returned to his native Iran. (233) The Board of Immigration Appeals concluded that he was not a genuine convert and rejected his asylum request. (234) The BIA based its reasoning on the fact that Bastanipour had never been baptized, had not formally joined a church, and continued to eat pork-free meals. (235) The court of appeals vacated the decision because the BIA had focused on its own perceptions of genuine Christian conversion rather than "what would count as conversion in the eyes of an Iranian religious judge, which is the only thing that would count so far as the danger to Bastanipour is concerned." (236) Samahon calls this approach "apostasy analysis" (237) because it directs the adjudicator to decide whether an Iranian religious judge would be likely to convict the asylum seeker of apostasy. This Article re-labels the Bastanipour approach as "the eyes of the persecutor test."

The eyes of the persecutor test guides credibility assessment in two critical ways. First, it keeps RSD focused on the actual criteria for refugee status (238) and on the critical question of whether an asylum seeker is at genuine risk of persecution. Second, it redirects the essential credibility assessment toward observable and objective factors, which are more comfortable grounds for adjudication. (239) Agents of persecution are no more capable than adjudicators of knowing what a person genuinely believes or does not believe. (240) Instead, persecution is triggered by the associations and observable behaviors that lead the persecutor to perceive that the victim is a member of the targeted group.

The starting point for this analysis is to develop an understanding of the triggers of persecution that, when perceived by the persecutor, lead to human rights violations. In Bastanipour, the court of appeals noted that apostasy in Islamic law was defined by abandoning Islam, (241) not by sincere adherence to Christianity, so that the trigger of persecution was the act of publicly renouncing Islam. (242) This approach made credibility assessment more straightforward because it focused on Bastinipour's public assertions of Christian faith while in the United States, even if it remained possible that he was less than sincere in his Christian faith. (243)

In Eritrea, the critical triggers might be attendance at certain churches or Bible-study groups. These actions are objectively observable, and interviewers can avoid interrogating subjective belief systems by focusing on objectively observable actions. In twenty-nine of the thirty Cairo transcripts, the applicant openly claimed to be Pentecostal, and the triggers of persecution might be difficult to distinguish from the basic declaration of belief. In one case, however, the applicant attended a Pentecostal Bible-study group with friends and decided not to convert, but she was still arrested, tortured, and accused of being a Pentecostal. (244) The interviewer asked fourteen questions seeking a narrative description of the arrest, interrogation, and detention. (245) However, the interviewer asked no questions seeking a description of the Bible-study group, which was the original trigger of persecution. (246) The interviewer, once satisfied with the applicant's account of actual past persecution, did not need to inquire into the applicant's religious status. (247) However, as an example, the interviewer could have further tested the credibility of the actions that provoked religious persecution, even though the applicant did not claim to be a genuine Pentecostal.

The eyes of the persecutor test should not allow an applicant to win asylum through a simple declaration. Credibility assessment is still essential, but it should be redirected toward testing whether the applicant can provide a detailed and coherent account of concrete actions and events that trigger a risk of persecution. Just as in refugee cases that do not involve religion, credibility assessment should focus primarily on the applicant's ability to put forward facts about her life. Questioning can and should be extensive, and fraud should require an unusually gifted storyteller.

G. Narrative Focus

Since the persecutor, like the adjudicator, cannot know an asylum seeker's genuine beliefs, inquiring about the applicant's genuine beliefs accomplishes nothing. If an applicant says, "God sacrificed his only son," the RSD adjudicator should note it but should draw no conclusions. Both a believer and a clever nonbeliever could utter a subjective statement of faith. Instead, the adjudicator should focus on observable factors that trigger persecution, such as actual attendance at church. If an applicant says, "I went to church on Sunday," then she has made an observable statement of fact analogous to "I was in prison" or "I own title to that land." It matters little why an Eritrean person went to church, whether for belief, for curiosity, for business connections, or to find a spouse. What matters is the impact of Pentecostal church attendance on the actions of the Eritrean government. The interviewer and the adjudicator should seek to determine if the applicant can describe her church attendance in detail and with consistency.

Narrative questions offer two advantages over theological questions. First, narrative questions accommodate different individual experiences and do not prejudge what a genuine refugee would know or believe. Unlike religious knowledge questions, narrative questions do not contain a bias in favor of educated people because narrative questions focus on personal experience. (248)

Second, personal narrative can be more easily assessed based on standard credibility criteria. If an applicant says, "God commands us to pray on the Sabbath," all that an adjudicator can attempt to discern is whether the applicant sincerely believes the statement to be true. Narrative questions avoid this trap, because they focus on concrete and objective assertions. If a person says, "I went to church on Sunday," the adjudicator must decide only whether to believe that the person was in fact present in church. Judging the credibility of these answers is not necessarily easy, especially if there is no independent information about the church. However, because these determinations closely resemble ordinary legal fact-finding, this approach allows adjudicators to rely on standard guidance about how to assess credibility and avoids the pitfalls of attempting to judge subjective faith.

Of the questions asked about religious persecution in the Cairo transcripts, 66 percent were narrative. (249) The majority of the narrative questions (250) were about the applicant's religious experience and activities, but these questions did not probe internal beliefs. Instead, they focused on chronology and concrete activities that formed the applicant's religious history; for instance, some questions asked about the applicant's activities in particular churches or Bible-study groups. Most of the remaining questions (251) focused on the applicant's direct experience of persecution, including threats, harassment, arrests, detention, interrogation, mistreatment, and similar abuses resulting from religious association.

The tendency at UNHCR-Cairo to focus on narrative questions, and specifically on questions about religious experience, corresponds to the UNHCR guidelines:
 Experience has shown that it is useful to resort to a narrative
 form of questioning, including through open-ended questions
 allowing the claimant to explain the personal significance of the
 religion to him or her, the practices he or she has engaged in (or
 has avoided engaging in out of a fear of persecution), or any other
 factors relevant to the reasons for his or her fear of being
 persecuted. Information may be elicited about the individual's
 religious experiences, such as asking him or her to describe in
 detail how he or she adopted the religion, the place and manner of
 worship, or the rituals engaged in, the significance of the
 religion to the person, or the values he or she believes the
 religion espouses. (252)

This approach is sensible if an applicant has not been persecuted before but claims to participate in religious activities that result in persecution. In these cases, the success of the refugee claim should depend on the applicant's ability to describe in compelling detail the nature of those activities.

Surprisingly, however, RSD interviewers asked few questions about applicants' religious practices in Egypt. Only 27 out of 759 narrative questions asked for information about religious practice in Egypt, and in twenty of thirty transcripts, there were no questions about religious practice in Egypt. In other jurisdictions, as previously noted, cases dealing with religious credibility frequently focus on religious practice in exile. (253) Moreover, a focus on religious practice in exile offers the potential advantage of local witnesses who can corroborate the testimony. Religious practice in exile may not directly trigger persecution, but it provides some information as to the likelihood that the applicant, if returned to Eritrea, would engage in similar persecution-triggering behavior.

Inquiring about religious practice abroad offers some advantages. First, unlike most central issues in RSD, much of what occurs in the country of asylum should be subject to additional evidence, including locally available witnesses. Second, an applicant raises legitimate doubts about her credibility if she claims to be devout but makes no effort to practice her religion in the asylum state. However, the danger remains that a smart liar could attend church just to win asylum.

H. Open Versus Closed Questions

One of the key distinctions in refugee interviewing techniques is the difference between open questions and closed questions. A UNHCR training manual explains that "[a]n open question is one that asks for general information and cannot be answered by 'yes' or 'no.' ... By using open questions, the interviewer provides the applicant the opportunity to relate events in his or her own way." (254) In contrast, closed questions can and usually must be answered with either a "yes," a "no," or a very specific piece of data. (255) The following chart illustrates how open and closed questions can be framed regarding the same subject matter:
Open Questions Closed Questions

What did you do after high school? Did you go to college after
 high school?

What do you remember most about What did you study in college?
your college education?

 For what years were you

The UNHCR published its training manual on RSD interviewing techniques in 1995, and a new version is due. (256) The 1995 version disfavors open questions:
 Open questions can be time-consuming. If the applicant is nervous,
 emotionally upset or has misunderstood the type of information you
 are seeking he or she may become confused and talk at length about
 irrelevant details. If this occurs the interviewer must gain
 control of the interview by politely intervening and changing the
 line of questioning. (257)

More recent UNHCR guidance, which specifically applies to refugee claims based on religious persecution, substantially relies on open questions. (258) Nonetheless, the Cairo interviewers preferred closed questions: 39 percent (259) of the Cairo questions were open, while 61 percent (260) were closed. (261)

Although open questions may increase the time required to complete RSD, RSD interviews, when done properly, are always time consuming, and fair RSD requires considerable time. (262) Using only general open-ended questions is potentially problematic because it may not produce adequate detail. For instance, in one Cairo interview, the officer asked almost exclusively general open questions and did not give the applicant significant guidance about other information that might be required:

Q: Can you tell me more about your religious problems?

A: They closed our church ... They also stopped our programs. Besides I had problems with the military and at college.

Q: What problems did you face at college?

A: That college is inside the camp called Sawa. But it was not only at college but also at the military. Because of my religion, they did not give me my rights and they did not treat me equally. For example, if we were running and I stopped because I was tired, he used to hit me. (263)

In this exchange, the applicant gave little concrete detail that would assist credibility assessment, and the topic jumped from church closures to discrimination in the Eritrean military without adequate exploration of either. Open questions are essential to elicit an applicant's own account of her experiences, but if only general open-ended questions are used, the interview may still fail to produce the necessary detail and coherent narrative. The 1995 UNHCR training manual warns that "[w]hen asking open questions you must also take into account the education level and cultural background of the applicant in order to assess whether he or she is capable of providing clear relevant information in response to your questions." (264) Lack of detail may be a sign of a lack of honesty, but it also may simply mean that the applicant failed to understand the question or did not understand the type of information required. (265)

At the same time, overreliance on closed questions requires interviewers to make assumptions in advance about what is important in an applicant's story, and, because the questions are closed, these questions prevent applicants from giving information unless the interviewer asks for it specifically. Consider this series of questions in one of the Cairo transcripts, where the interviewer attempts to solicit information about the church that the applicant attended in Eritrea:

Q: In which church did you used to go to while you were in Asmara?

A: Full Gospel.

Q: Do you know approximately how many believers used to attend the prayers in this church?

A: Yes, I know that there were many.

Q: Around?

A: I can't say in figures. I only know that there were many.

Q: Around 100, 200, 1,000?

A: I'd say more than 1,000, because for worship many people visit the church for worshipping.

Q: Where is the Full Gospel Church located in Asmara?

A: In Geza Manda.

Q: Could you tell me please the names of the pastors of this church?

A: Ok.

Q: Ok, tell me.

A: [Gives the names of five pastors]. (266)

The interviewer did not ask any other questions about the church. By pressing repeatedly for the number of parishioners, the interviewer erred and increased the risk that the applicant would eventually decide to simply say what the interviewer wanted to hear. Moreover, the questions all focused on trivial details that could easily be memorized. Because the interviewer asked only for very specific pieces of information, the applicant was deprived of the chance to present her own description of the church, and the interviewer did not learn whether the applicant was capable of providing a detailed account of the prayer services that she claimed to have attended. A better line of questioning might have asked the applicant to describe what the church looked like, its weekly or daily schedules, and the nature of its worship programs.

One of the reasons that interviewers may prefer closed questions is that closed questions probe for very specific information, thus filling in gaps and eliciting detail (or lack thereof) that provides a basis for assessing credibility. (267) However, it is possible to ask open questions that are very specific.
 Open Closed

General Can you describe the Who were the pastors at
 prayer services at your the church?

Specific What was the first thing What was the name of the
 that would normally first prayer recited on
 happen at the beginning Sunday?
 of prayer services on

Specific open questions focus on details but give applicants the ability to answer with a range of different types of information. Although closed questions can elicit detail, they test a person's ability to provide only one type of data, and they may be founded on false assumptions about what information would be significant enough that a genuine applicant would be able to recall it accurately. As Juliet Cohen explains in a study of the cognitive aspects of asylum seeker testimony, human memory does not function in such a standardized manner; people tend to remember the details that have some subjective importance to them. (268) A better alternative is to ask specific but open questions with possible follow-ups to prompt a person to provide whatever type of sensory detail they can actually recall. For example:

Q: Can you describe the room?

Q: What did it look like?

Q: Do you remember if it had a particular smell?

Q: Do you remember any sounds that you heard in that room?

Q: Can you draw a diagram or picture of the room?

This type of questioning gives genuine applicants more opportunity to provide compelling detail. Further, if applicants fail to produce any specificity, the fact that they have been given more of an opportunity to do so gives adjudicators a firmer basis to treat vagueness as a negative credibility factor.


In Egypt, the UNHCR generally focused on narratives more than on abstract matters of faith and asked more questions about concrete actions or events than about personal religious knowledge, practice, or belief. Yet, while there was a general tendency to focus on narrative, it is difficult to find any uniform approach to credibility because interview tactics varied substantially. Many interviews did not focus on religious knowledge and beliefs, but a significant minority of interviews included knowledge quizzes that seemed to carry substantial weight. (269) This variability in interview tactics is likely to produce an inconsistency in results.

The specific form of the questions poses a different problem. Because the UNHCR eligibility officers in Cairo generally asked few open questions, the value of focusing on narrative was undermined. (270) Without open questions, an applicant going into such an interview would be unlikely to take the initiative in describing the events that she believes are most important. The applicant would instead need to be able to provide specifically requested information, and the efficacy of the interview would become dependent on the accuracy of the interviewer's assumptions about what information is important.

This Article suggests that to have a high chance of being found credible under the approach used by UNHCR-Cairo in 2007-2008, an Eritrean Pentecostal asylum seeker needs to be skilled at answering very specific closed questions about her experiences in Eritrea. If called upon, she should also be able to provide a recitation of her religious beliefs with chapter and verse Bible citations to support them. However, it is possible that she will not be asked about the Bible at all, or might only be asked a few questions about her experiences in Eritrea. If she performs better on one type of question than another, it is unclear how the relative strengths and weaknesses of her answers will be judged, just as it is difficult to discern why the approaches used in different interviews vary so much.

The different types of questions that interviewers used in the Cairo interviews are all potentially useful in certain circumstances, with the exception of questions that appear to contest or debate applicants' asserted religious faith. In cases concerning religion, it is essential that RSD steer clear of invasive questions that violate religious freedom. Although each question might be useful in certain circumstances, the RSD process needs a more systematic approach to determining which questions are relevant in which cases. As Herlihy argues cogently, (271) RSD interviewers need yardsticks by which to assess different approaches to credibility assessment, as well as a greater awareness of the logical steps that adjudicators take on their way to reaching credibility decisions. Although there is general agreement that RSD cannot adjudicate matters of theology, (272) there is much less agreement about how to meet the challenges posed by the religious imposter problem.

The best place to start appears to be the refugee definition itself, which is the foundation of the eye of the persecutor approach. The core question in RSD must ultimately be about what triggers persecution. Credibility assessment should start, therefore, with providing applicants an opportunity to explain the experiences that make them feel at risk, typically through use of open questions combined with country of origin information that explains the behavior of the persecutor. This approach will normally begin, and sometimes end, with narrative questions only. It should assume that persecutors are no more capable than asylum adjudicators of looking into men's hearts. Instead, the focus must be on the observable risk factors that raise the chances of persecution.

Questions about religious knowledge and beliefs are useful sometimes, but they should be used as supplements only when factors in the narrative make them relevant. If an applicant says that she engaged in years of Bible study, it is reasonable to follow up by asking her about what she studied and to replicate some of what she learned. Here, too, an open-question approach can be used, letting the applicant volunteer knowledge rather than testing an ability to recall specific citations by rote. One should not assume that people who attend intensive Bible study necessarily recall every topic perfectly, just as lawyers do not always recall every case and rule that they read in law school.

Every question in an RSD interview should have an underlying purpose and logic. Narrative questions should produce an assertion by the applicant that she participated in an activity like Bible study--an observable, concrete behavior that could trigger persecution by the Eritrean government. The interviewer should test the credibility of this assertion mainly through further narrative questions, pressing for additional details about the way the Bible- study program operated. When and where did it meet? How were the meetings organized? It might be useful to ask, briefly, about the knowledge learned during such study groups, but it is essential to see this line of questioning in its context. The only purpose for inquiring about religious knowledge in such a case is to determine if the applicant actually engaged in the religious activity. Refugee status determination should not ultimately be concerned about whether the applicant was sincere in her Bible study. (273) It is possible to test the credibility of such a claim without testing religious knowledge at a11. (274)

Traditions of judicial deference toward first-instance decision makers have slowed the development of more systematic approaches to credibility assessment in RSD, and it is essential that the UNHCR and scholarly analysis continue to work to fill this gap. It would be useful to expand this research to analyze the subsequent analyses and judgments and test whether the types of questions correlate with the results of RSD proceedings.

Even where the legal basis of an asylum claim is simple, and perhaps especially where it is simple, the evidentiary challenges of RSD are complex. It is easy to say "I am a believer in a persecuted faith," but, so long as refugee status remains a precious exception to the general norms of migration law, it is essential that there be a higher bar for winning refugee protection. However, assessing the credibility of religious assertions in a way that will be fair, reliable, and not offensive to religious freedom is a difficult balancing act. Adjudicators must remember that human rights norms aim ultimately to shield religious faith from official scrutiny.

(1.) A refugee is any person who, "owing to well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable, or owing to such fear, is unwilling to avail himself of the protection of that country." Geneva Convention Relating to the Status of Refugees art. 1, July 28, 1951, 189 U.N.T.S. 150 [hereinafter Refugee Convention]. Major interpretive disputes have developed as to the definition's applicability in civil war cases, gender-based persecution and family violence, victims of people trafficking, draft evaders, and military deserters, among other issues.

(2.) Yan v. Gonzales, 438 F.3d 1249, 1252 (10th Cir. 2006).

(3.) Tuan N. Samahon, The Religion Clauses and Political Asylum: Religious Persecution Claims and the Religious Membership-Conversion Imposter Problem, 88 GEO. L.J. 2211, 2211-12 (2000).

(4.) See id. at 2213 (noting the purpose of the Immigration and Naturalization Service's strict interpretation of the Immigration and Nationality Act's "well-founded fear of persecution" requirement as deterrence of massive numbers of applicants).

(5.) See id. at 2211-12 ("[A]n adjudicator faces the dilemma of determining whether a religious convert is an imposter or a legitimate member of a religious group that is persecuted in the applicant's home country.").

(6.) Refugee Appeal No. 74611/03, [paragraph] 41, at 10 (June 12, 2003) (N.Z. Refugee Status Appeals Auth.),

(7.) Guy Coffey, The Credibility of Credibility Evidence at the Refugee Review Tribunal, 15 INT'L J. REFUGEE L. 377, 378 (2003); Michael Kagan, Is Truth in the Eye of the Beholder? Objective Credibility Assessment in Refugee Status Determination, 17 GEO. IMMIGR. L.J. 367, 369 (2003).

(8.) Kagan, supra note 7, at 371-74.

(9.) Id. at 413.

(10.) Id. at 373 n.24.

(11.) Id. at 413.

(12.) See Steve Norman, Assessing the Credibility of Refugee Applicants: A Judicial Perspective, 15 INT'L J. REFUGEE L. 273, 282-86 (2007) (issuing a summary of the types of evidence considered in refugee status determinations); see also James A. Sweeney, Credibility, Proof and Refugee Law, 21 INT'L J. REFUGEE L. 700, 700-01 (2009) (discussing the necessity of credibility assessments since applicants will generally have little tangible or documented evidence).

(13.) Asylum Policy Instructions: Assessing Credibility in Asylum and Human Rights Claims, UK BORDER AGENCY (Dec. 7, 2009), asylumpolicyinstructions/apis&credibility.pdf?view=Binary [hereinafter Asylum Policy Instructions].

(14.) Kagan, supra note 7, at 384.

(15.) Id. at 384-97.

(16.) See generally id. at 399-403 (explaining the previously proposed analytical method by which to assess the countervailing factors that are relevant in assessing internal credibility); cf. Asylum Policy Instructions, supra note 13 (describing how the internal credibility assessments evaluate level of detail, inconsistencies, and mitigating circumstances).

(17.) Norman, supra note 12, at 275-76.

(18.) Mariham Iskander Wahba, Reading Between the Lines: The Issue of Credibility in Refugee Status Determination Interviews and Service Provision for Victims of Sexual and Gender Based Violence in Cairo (2009) (unpublished student paper, American University Cairo) (on file with author).

(19.) See id. at 7-8 (describing concerns about refugees' credibility in claiming to have been raped).

(20.) Id.

(21.) Id.

(22.) Kagan, supra note 7, at 369.

(23.) See generally Brian Gorlick, Common Burdens and Standards: Legal Elements in Assessing Claims to Refugee Status, 15 INT'L J. REFUGEE L. 357, 357-61 (2003) (concluding "that the absence of consensus amongst states on common standards for assessing evidence in refugee status determinations remains an obstacle").

(24.) See Coffey, supra note 7, at 404-05 ("The Tribunal's decisions are not frequently set aside on the basis of its treatment of credibility."); Jens Vedsted-Hansen, The Borderline Between Questions of Fact and Questions of Law, in PROOF, EVIDENTIARY ASSESSMENT AND CREDIBILITY IN ASYLUM PROCEDURES 57, 57 (Gregor Noll ed., 2005) ("[D]ecisions made by the [Danish] Refugee Appeals Board cannot be subject to judicial review."); Sweeney, supra note 12, at 724-25 ("[S]ince 'credibility' findings are about the facts, they also tend to be nearly immune from appeal.").

(25.) See Kagan, supra note 7, at 409.
 In a system where most appeals are by rejected asylum-seekers and
 where incorrect rejections expose people to serious human rights
 abuses, a standard designed to affirm first instance decisions ...
 should be of great concern.... Unfortunately in the asylum context,
 deferential review will likely undermine this goal by restraining
 the BIA's capacity to develop precedents about how credibility
 assessment ... should be conducted.

Id.; cf. Norman, supra note 12, at 291-92.
 Concern has been expressed that, particularly with respect to RSD
 decision makers from whom the only avenue of appeal is on narrow
 points of law, they might on occasion 'be tempted to seek to
 insulate their decisions from judicial review by disposing of all
 issues by adverse findings on credibility'. However, the assessment
 of evidence does not lend itself to anything like scientific
 precision; and findings of fact are clearly something about which
 reasonable minds may differ.

Norman, supra note 12, at 291-92 (footnotes omitted).

(26.) Sweeney, supra note 12, at 725.

(27.) Id. at 725-26.

(28.) See id. at 705-06 ("The Introduction to the API states that decision makers will often have to decide whether 'they believe' the applicant's evidence (emphasis added). Later, the API seems resigned to the fact that assessing a claim's credibility 'inevitably involves an element of subjectivity on the decision maker's part'...." (footnotes omitted)).

(29.) See Norman, supra note 12, at 287-88 ("There may be numerous reasons why an applicant will not, or cannot, respond to questions put to them in the refugee assessment process.... It is important to remember, however, that what constitutes a 'meaningful response' to questions put may differ between different cultures and individuals."); see generally Zachary Steel, Naomi Frommer & Derrick Silove, The Mental Health Impacts of Migration: The Law and Its Effects; Failing to Understand: Refugee Determination and the Traumatized Applicant, 27 INT'L J.L. & PSYCHIATRY 511, 512-25 (2004) ("Complex traumatic presentations can easily be misunderstood by decision-makers, leading to adverse findings about the protection needs of an applicant.").

(30.) See Fassehaye v. Gonzales, 414 F.3d 746, 756 (7th Cir. 2005) ("We have never held that an applicant's affidavit alone is insufficient to establish membership in a certain group.").

(31.) Najafi v. INS, 104 F.3d 943, 949 (7th Cir. 1997).

(32.) Id.

(33.) Yan v. Gonzales, 438 F.3d 1249, 1252 (10th Cir. 2006); see also Guo v. Ashcroft, 361 F.3d 1194, 1200 (9th Cir. 2004) (criticizing the immigration judge for interpreting ambiguities against the asylum seeker).

(34.) See, e.g., X v. Canada, CA0-00119, [2001] CanLII 26881 (Immigration and Refugee Bd., Refugee Div., Nov. 22, 2001), (highlighting that in a case concerning an alleged Jehovah's Witness, the decision maker considered it a negative factor that he could not provide specific Bible references for his beliefs given that in general Jehovah's Witness theology places great stress on scripture).

(35.) X v. Canada, VA0-00941, [2001] CanLII 26996, at 5 (Immigration & Refugee Bd., Refugee Div., June 27, 2001), 2001canlii26996/2001canlii26996.pdf.

(36.) Ahmadshah v. Ashcroft, 396 F.3d 917, 920 n.2 (8th Cir. 2005) ("Even if [petitioner] did not have a clear understanding of Christian doctrine, this is not relevant to his fear of persecution."); see Yan, 438 F. 3d at 1252-55 ("We agree with the Eighth Circuit that a detailed knowledge of Christian doctrine may be irrelevant to the sincerity of an applicant's belief."); Refugee Appeal No. 74611/03, [paragraph] 10, at 3 (June 12, 2003) (N.Z. Refugee Status Appeals Auth.) ("[A] comprehensive knowledge of Christianity is no indicator of the strength of an individual's conviction.").

(37.) Yan, 438 F.3d at 1255.

(38.) See, e.g., T v. Sec'y of State for the Home Dep't, [2006] EWCA (Civ) 483, [8] (Eng.) (finding negative credibility in a case before a British adjudicator concerning an asserted Eritrean Pentecostal--which was ultimately upheld by the Immigration Appeal Tribunal and Court of Appeal--despite an acknowledgement that the applicant's testimony was internally consistent and also consistent with independent information about conditions in Eritrea); id. at [5] (revealing that the adjudicator doubted the plausibility of evangelical parents not passing their religion onto their young children).

(39.) Refugee Appeal Nos. 74862-74865, [paragraph] 85, at 19 (Feb. 19, 2004) (N.Z. Refugee Status Appeals Auth.),

(40.) See Refugee Appeal No. 74611/03, [paragraph] 11, at 3 ("Rather than relying on a claimant's own description of his internal state of mind, the Authority sought to establish the genuineness of his convictions by considering other evidence--namely, his activities since arriving in New Zealand, when for the first time he was free to practice his Christianity."); Refugee Appeal No. 72356/2000, [paragraph] 6, at 2 (Feb. 28, 2001) (N.Z. Refugee Status Appeals Auth.), Ref_20010228_72356.pdf (citing regular attendance at church in New Zealand as a positive credibility factor); Refugee Appeal No. 72323/2000, [paragraph][paragraph] 33-34, at 8 (Sept. 25, 2001) (N.Z. Refugee Status Appeals Auth.), 72323.pdf (citing testimony of a pastor in New Zealand as highly persuasive in accepting appellant's sincerity); X v. Canada, AA0-00442, [2001] CanLII 26828 (Immigration and Refugee Bd., Refugee Div., Apr. 5, 2001), (reaching a negative credibility finding in an application by a Chinese man who claimed to be a member of the Falun Gong in part because the claimant had barely made any effort to participate in the movement in Canada).

(41.) See T v. Sec'y of State for the Home Dep't, [2006] EWCA (Civ) 483, [7] ("This is a matter which supports the appellant, but since it is entirely at his will, it of necessity carries limited weight.").

(42.) See Refugee Appeal No. 74611/03, [paragraph] 11, at 3 (determining claimant's sincerity with testimony from his pastor about claimant's church attendance and involvement).

(43.) See, e.g., Mezvrishvili v. U.S. Att'y Gen., 467 F.3d 1292, 1296 (11th Cir. 2006) (noting that the immigration judge found Mezvrishvili's evidence insufficient because he lacked knowledge of the distinct doctrine of his proclaimed faith); Rizal v. Gonzales, 442 F.3d 84, 90 n.7 (2d Cir. 2006) (asserting that although actual membership may be questionable, the overarching determination is whether one can be perceived as belonging to a persecuted class); Abbas v. Ashcroft, 126 F. App'x 38, 39 (2d Cir. 2005) (holding that lack of credibility regarding Christian conversion did not defeat an asylum application based on imputed religious conversion because one must consider whether the fatwa resulted in a realistic fear).

(44.) See generally Samahon, supra note 3 (discussing the issues of proving conversion from Islam to Christianity against the strict guidelines of the Immigration and Naturalization Service (INS)); see also Craig B. Mousin, Standing with the Persecuted: Adjudicating Religious Asylum Claims After the Enactment of the International Religious Freedom Act of 1998, 2003 BYU L. REV. 541, 544 (2003) ("This paper proposes that IRFA calls for adjudicating asylum claims on account of religion through the lens of international agreement, with greater sensitivity to the scope of religious persecution in the world."); Brigette L. Frantz, Note, Proving Persecution: The Burdens of Establishing a Nexus in Religious Asylum Claims and the Dangers of New Reforms, 5 AVE MARIA L. REV. 499, 502 (2007) ("This Note contends that the United States has become one of the most difficult nations in which to win asylum from religious persecution, and thus the original intention of asylum has become skewed.").

(45.) U.N. High Comm'r for Refugees [UNHCR], Religion-Based Refugee Claims Under Article 1A(2) of the 1951 Convention and/or the 1967 Protocol Relating to the Status of Refugees, [paragraph][paragraph] 28-33, U.N. Doc. HCR/GIP/04/06 (Apr. 28, 2004).

(46.) See Kagan, supra note 7, at 381-82 (revealing that an adjudicator need not decide that the asserted statement is actually true, but only that it is believable and should be accepted as true for the purposes of RSD); UNHCR, Note on Burden and Standard of Proof in Refugee Claims, [paragraph] 11 (Dec. 16, 1998), ("Credibility is established where the applicant has presented a claim which is ... on balance, capable of being believed.").

(47.) Arye Edrei, Law, Interpretation, and Ideology: The Renewal of the Jewish Laws of War in the State of Israel, 28 CARDOZO L. REV. 187, 189 n.3 (2006) (discussing the tension between the Zionist movement and implications of the movement on traditional views regarding conversion to Judaism).

(48.) Samahon, supra note 3, at 2217.

(49.) Id. at 2228-29.

(50.) U.N. High Comm'r for Human Rights, General Comment No. 22: The Right to Freedom of Thought, Conscience and Religion (Art. 18), [paragraph] 2, U.N. Doc. CCPR/C/21/Rev.1/Add.4 (July 30, 1993).

(51.) Id. [paragraph] 3.

(52.) See generally Catherine Dauvergne & Jenni Millbank, Burdened by Proof: How the Australian Refugee Review Tribunal Has Failed Lesbian and Gay AsylumSeekers, 31 FED. L. REV. 299, 299 (2003) ("Our argument in this paper is that evidentiary practices and procedures that have been developed by the Australian Refugee Review Tribunal are operating at a routinely low standard."); Jenni Millbank, 'The Ring of Truth': A Case Study of Credibility Assessment in Particular Social Group Refugee Determinations, 21 INT'L J. REFUGEE L. 1, 2 (2009) ("This article explores credibility assessment in lower level tribunals using a case study of particular social group (PSG) ground decisions made on the basis of sexual orientation."); Cecile Rousseau et al., The Complexity of Determining Refugeehood: A Multidisciplinary Analysis of the Decision-Making Process of the Canadian Immigration and Refugee Board, 15 J. REFUGEE STUD. 43, 46 (2002) ("This paper describes the methodological approach taken in creating the intellectual tools for the analysis, as well as the results obtained through their use in analyzing some forty cases decided by the IRB in Montreal.").

(53.) See generally Norman, supra note 12 (focusing, in part, on the "expressed views of Australian courts after examining credibility findings in decisions of the Refugee Review Tribunal").

(54.) Juliet Cohen, Questions of Credibility: Omissions, Discrepancies and Errors of Recall in the Testimony of Asylum Seekers, 13 INT'L J. REFUGEE L. 293, 308 (2001); Jane Herlihy, Peter Scragg & Stuart Turner, Discrepancies in Autobiographical Memories--Implications for the Assessment of Asylum Seekers: Repeated Interviews Study, 324 BMJ 324, 324 (2002).

(55.) See Kagan, supra note 7, at 393-95.

(56.) In conducting this research, I had the opportunity to converse with Dr. Herlihy and discuss her parallel research project that analyzes the way British asylum adjudicators determine credibility. I also benefited from reviewing an unpublished draft of an article by Dr. Herlihy, Kate Gleeson, and Stuart Turner. The International Journal of Refugee Law published the article, Jane Herlihy, Kate Gleeson & Stuart Turner, What Assumptions About Human Behaviour Underlie Asylum Judgments?, 22 INT'L J. REFUGEE L. 351 (2010), in July 2010.

(57.) Id. See also Millbank, supra note 52 ("A benefit of drawing upon a subset of cases as part of a broader inquiry into credibility determination is that it can provide a 'complete' set of cases on a particular issue to offer both comparative perspectives and information on longitudinal trends.").

(58.) Jane Herlihy, Kate Gleeson & Stuart Turner, What Assumptions About Human Behaviour Underlie Asylum Judgments? 8-9 (unpublished manuscript) (on file with author).

(59.) See generally id.

(60.) Id. at 20-21.

(61.) Hilary Evans Cameron, Risk Theory and 'Subjective Fear': The Role of Risk Perception, Assessment, and Management in Refugee Status Determination, 20 INT'L J. REFUGEE L. 567, 568 (2008).

(62.) Id.

(63.) Herlihy, Gleeson & Turner, supra note 58, at 22.

(64.) I was the director of AMERA's operations in Egypt in 2007 and 2008, and I continue to work as a consulting attorney for the organization. Additionally, I have practiced refugee law in Egypt dating back to 1998. This personal involvement provides some of the factual background for the context of refugee status determination that I relate in this Article.

(65.) See NAIROBI CODE: MODEL RULES OF ETHICS FOR LEGAL ADVISORS IN REFUGEE CASES R. 6.7 (2007), available at ("A legal advisor or organization employing a legal advisor may use information collected from clients' cases in publications and writings without the consent of affected clients only if the publication is sanitized of any unique details that would allow an interested person to identify the person involved.").

(66.) Where UNHCR-RSD Happens, RSD WATCH, where-unhcr-rsd-happens/(last updated June 20, 2010).

(67.) My description of the refugee protection system is based on my own knowledge of it as a practitioner. See also Kagan, supra note 7, at 368-69 & nn.6, 8; Michael Kagan, Assessment of Refugee Status Determination Procedure at UNHCR's Cairo Office 2001-2002 (Am. Univ. Cairo, Forced Migration and Refugee Studies, Working Paper No. 1, 2002),

(68.) In the period studied, different nationalities of refugees were subject to different registration and RSD systems. In general, most Iraqis were recognized as refugees on a prima facie basis. RSD for Sudanese was suspended indefinitely, and all Sudanese registering at UNHCR were given temporary protection. The system described in this Article applied only to Eritreans who were not in detention. For many Eritreans, especially for those arrested at Egypt's borders and those who lacked documentation, serious protection problems occurred in the period studied. Many of these problems are summarized in HUMAN RIGHTS WATCH, SINAI PERILS: RISKS TO MIGRANTS, REFUGEES, AND ASYLUM SEEKERS IN EGYPT AND ISRAEL (2008), available at

(69.) Description of the RSD procedure in Egypt is based on my knowledge as a practitioner and supervisor of refugee legal aid in Egypt.


(71.) Id. [section] 6.2.

(72.) UNCHR, supra note 70, [section] 4.3.3.

(73.) Id. [section] 4.3.4.

(74.) Description of RSD procedure in Egypt is based on my knowledge as a practitioner and supervisor of refugee legal aid in Egypt.

(75.) 12 NGOs Seek 'Incremental Steps' from UNHCR on Evidence Disclosure, RSD WATCH (Oct. 15, 2008), 12-ngos-seekincremental-steps-from-unhcr-on-evidence-disclosure/.

(76.) I was the director of AMERA's operations in Egypt when this process began. See supra note 64.

(77.) See generally Kagan, supra note 67 (discussing UNHCR-Cairo's failure to implement "many of the United Nation's [sic] own standards and precedents of procedural fairness").

(78.) See generally RSD WATCH, (last visited Oct. 15, 2010) (tracking UNHCR reform progress).

(79.) UNHCR, STATISTICAL YEARBOOK 2008, annex, tbl. 12 (2009) [hereinafter 2008 Yearbook],; UNHCR, STATISTICAL YEARBOOK 2007, annex, tbl. 12 (2008) [hereinafter 2007 Yearbook],

(80.) This chart includes the countries for which UNHCR reported at least 100 RSD decisions in 2007 or 2008, but it excludes some countries due to apparent anomalies in the data that appear to make the reported recognition rate artificially high or low. On the high end, the chart excludes countries where the reported recognition rate was over 95 percent, suggesting that the jurisdiction may have been practicing a form of prima facie or group-based recognition rather than RSD based on the individual merits of the cases. Other countries were excluded because more than 30 percent of the decisions were listed by UNHCR as "otherwise closed," meaning that the overall recognition may be artificially deflated by an unusually large number of cases that were not decided on the merits.

(81.) First, although statistics published by the UNHCR make it possible to compare Eritrean applications in different RSD systems, this does not mean that all systems receive the same types of refugee claims or even claims from the same nationality. Second, a number of states, especially in Europe, tend to be very limited in their willingness to recognize Convention refugee status, but are relatively generous in their willingness to extend complimentary forms of protection (defined here as any affirmative grant of protected status other than recognition as a refugee under the Refugee Convention, supra note 1) to asylum seekers. UNHCR and North American jurisdictions tend to do the reverse; they generally do not grant complimentary forms of protection to Eritreans but recognize Convention refugee status at a relatively high rate. This study does not focus on different legal interpretations of the Refugee Convention; it instead focuses on the practical question of whether an asylum seeker ultimately received some form of protection after RSD. For this reason, the chart includes a composite recognition rate, including both the Convention refugee recognitions and grants of complimentary protection.

(82.) Email from Anonymous to Michael Kagan, Senior Fellow American University in Cairo (November 15, 2009) (on file with author) [hereinafter Anonymous Email].

(83.) Id.


(85.) Id.

(86.) Id. at 1.

(87.) Paul Freston, Byker Chair, Calvin Coll., Remarks at Moved by the Spirit: Pentecostal Power & Politics After 100 Years (Apr. 24, 2006), available at -Pentecostal-Power-and-Politics-after-100-Years%282%29.aspx.

(88.) PEW RESEARCH CTR., supra note 84, at 1.

(89.) Id.

(90.) Id. at 1-2.

(91.) Acts of the Apostles 2:1-2, 4, 17-18 (NIV).

(92.) PEW RESEARCH CTR., supra note 84, at 1-2.

(93.) Id. at 15.

(94.) Anthea Butler identifies three thematic foundations of the Pentecostal faith: (1) an apocalyptic Messianic orientation (i.e., the expectation that Jesus will return soon); (2) a belief in the restoration of the apostolic age, meaning that religious and social transformations that occurred in Jesus's time can occur now; and (3) an egalitarian commitment stemming from a belief that the Spirit is "poured out on all flesh," without discrimination. Anthea Butler, Assistant Professor of Religion, Univ. of Rochester, Remarks at Moved by the Spirit: Pentecostal Power & Politics After 100 Years (Apr. 24, 2006), available at Moved-by-the-Spirit-Pentecostal-Power- and-Politics-after-100-Years%282%29.aspx.

(95.) PEW RESEARCH CTR., supra note 84, at 1.

(96.) Id. at 26.

(97.) Id. at 27.

(98.) Id. at 28.

(99.) Id. at 6, 15.

(100.) Id. at 2-3.

(101.) Abbebe Kileyesus, Cosmologies in Collision: Pentecostal Conversion and Christian Cults in Asmara, 49 AFR. STUD. REV. 75, 77 (2006).

(102.) Id.

(103.) PEW RESEARCH CTR., supra note 84, at 2.

(104.) Charlotte Spinks, Panacea or Painkiller? The Impact of Pentecostal Christianity on Women in Africa, 1 CRITICAL HALF 21, 21 (2003).

(105.) PEW RESEARCH CTR., supra note 84, at 32.

(106.) Id. at 1.

(107.) UNHCR, UNHCR Eligibility Guidelines for Assessing the International Protection Needs of Asylum-Seekers from Eritrea, at 9 (April 2009),

(108.) 2008 Yearbook, supra note 79, at 52; 2007 Yearbook, supra note 79, at 47.


(110.) Amnesty Int'l, Eritrea: Religious Persecution, at 3, AI Index AFR 64/013/2005 (Dec. 7, 2005).

(111.) According to the U.S. Department of State, the largest religious affiliation in Eritrea is Sunni Islam, accounting for an estimated 50 percent of the population. DEP'T OF STATE, ERITREA: 2008 RELIGIOUS FREEDOM REPORT sec. I (2008), available at (estimating that Orthodox Christians account for around 30 percent, followed by Roman Catholics at 13 percent, with the remainder including an assortment of Christian affiliations, such as various types of Protestants, Seventh-Day Adventists, and Jehovah's Witnesses, and small populations of Buddhists, Hindus, and Baha'is). But see Amnesty Int'l, supra note 110, at 3 (estimating the Eritrean Orthodox Church and Sunni Islam together account for around 90 percent of the population, with Roman Catholics at 5 percent, Lutheran at 1 percent, and I percent belonging to other Christian movements, including Pentecostals, various Evangelical churches, and Jehovah's Witnesses).


(113.) Id.

(114.) Id.

(115.) Id. [paragraph] 19.06.

(116.) Id.

(117.) Id.

(118.) CONST. OF ERITREA art. 19(4) ("Every person shall have the freedom to practice any religion and to manifest such practice.").

(119.) DEP'T OF STATE, supra note 111, sec. II.

(120.) U.K HOME OFFICE, supra note 112, [paragraph] 19.05.

(121.) See generally id. [paragraph][paragraph] 19.16-48 (describing the treatment of various religious groups).

(122.) DEP'T OF STATE, supra note 111, sec. II.

(123.) Id.

(124.) See id. ("NGO reports indicated there were more than 3,225 Christians from unregistered groups detained in prisons."); Amnesty Int'l, supra note 110, at 3 ("Eritrea has a highly religious population, with some 98% of its 3.7 million people belonging to a long-established branch of a major world religion.").

(125.) Id. at 5.

(126.) Id. at 1; see also Amnesty Int'l, Urgent Action: Eritrea: Torture, AI Index AFR 64/005/2008 (September 5, 2008) (urging the Eritrean government to release information on the Muslim dissidents who were arrested and detained).

(127.) Amnesty Int'l, supra note 110, at 9.

(128.) Tanya Datta, Eritrean Christians Tell of Torture, BBC NEWS, June 6, 2008,

(129.) Amnesty Int'l, supra note 110, at 2.

(130.) Id. at 2.

(131.) Datta, supra note 128.

(132.) Amnesty Int'l, supra note 110, at 2.

(133.) Id. at 1, 8.

(134.) Id. at 8-9.

(135.) Id. at 4-5.

(136.) DEP'T OF STATE, supra note 111, sec. II.

(137.) Id. secs. I-II; see also Amnesty Int'l, supra note 110, at 6-8 (describing the particularly intense persecution endured by Jehovah's Witnesses because of their conscientious objection to the war).

(138.) Amnesty Int'l, supra note 110, at 9.

(139.) Id. at 6.

(140.) Id. at 9.

(141.) See UNHCR, supra note 107, at 10 (explaining that Eritrean claimants of this type should have an assumption of eligibility for asylum).

(142.) See Universal Declaration of Human Rights, G.A. Res. 217 (III) A, art. 18, U.N. Doc. A/RES/217(III) (Dec. 10, 1948) (stating that everyone has a right to freedom of religion); see also International Covenant on Civil and Political Rights art. 18(1), Dec. 16, 1966, 999 U.N.T.S. 171 (entered into force Mar. 23, 1976).

(143.) Li v. Gonzales, 420 F.3d 500, 504-05 (5th Cir. 2005), vacated, 429 F.3d 1153 (2005).

(144.) Id.

(145.) Id. at 510.

(146.) Id.

(147.) Id. at 511.

(148.) Li v. Gonzales, 429 F.3d 1153 (5th Cir. 2005); see also Frantz, supra note 44, at 515-16 (describing the public support for Li that helped bring about the reversal).

(149.) Frantz, supra note 44, at 516-17; see also Fassehaye v. Gonzales, 414 F.3d 746, 755 (7th Cir. 2005) (ruling that Eritrean religious minorities have a well-founded fear of persecution so long as their asserted religious affiliations are credible).

(150.) See Hasan v. Bulgaria, App. No. 30985/96, 34 Eur. H.R. Rep. 1339, [paragraph][paragraph] 76-78 (2000) (declining to rule on the validity of registration requirements in the abstract); UNHCR, supra note 45, [paragraph] 19; U.N. Secretary-General, Elimination of All Forms of Religious Intolerance, [paragraph] 135, U.N. Doc. A]58/296 (August 19, 2003) (by Abdelfattah Amor).

(151.) U.N. Secretary-General, supra note 150, [paragraph] 135.

(152.) Hasan, 34 Eur. H.R. Rep. 1339, [paragraph] 62.

(153.) Id.

(154.) Id. [paragraph] 78.

(155.) Id.

(156.) Sahin v. Turkey, 2005-XI Eur. Ct. H.R. 173, [paragraph] 107.

(157.) U.N. High Comm'r for Human Rights, supra note 50, [paragraph] 2.

(158.) Id. [paragraph] 9.

(159.) Id. [paragraph] 5.

(160.) UNHCR, supra note 45, [paragraph] 29.

(161.) Seven hundred fifty-nine questions, or 66.3 percent of 1,144, the total number of questions in the sample.

(162.) Three hundred eighty-five questions, or 33.7 percent of 1,144, the total number of questions in the sample.

(163.) See United States v. Ballard, 322 U.S. 78, 81 (1944) (reasoning that the focus of the court and jury should be the sincerity instead of the validity of the defendant's teachings).

(164.) Id. at 79.

(165.) Id.

(166.) Id. at 81.

(167.) Id. at 88.

(168.) See Mousin, supra note 44, at 571 (discussing the view that some form of sincerity test is helpful in making asylum decisions).

(169.) Refugee Appeal Nos. 74862-74865, [paragraph] 19, at 5 (Feb. 19, 2004) (N.Z. Refugee Status Appeals Auth.). For another example of apparent leniency, see Refugee Appeal No. 72323/2000, [paragraph] 8, at 2 (Sept. 25, 2001) (N.Z. Refugee Status Appeals Auth.) ("[A]t the conclusion of the hearing the Authority was satisfied that this appellant has genuinely converted to Christianity and is a serious and dedicated member of an Evangelical Christian church in New Zealand.").

(170.) In twenty-one of thirty transcripts, there were between one and nine neutral questions about religious beliefs or practices. In six of thirty transcripts, no questions of this category appeared, while in three cases, ten or more appeared.

(171.) RSD Interview with Anonymous Applicant No. 13 in Cairo, Egypt (transcript on file with author).

(172.) See supra Part III.A.

(173.) For further discussion on the weaknesses of religious knowledge quizzes, see infra Part IV.A.2.

(174.) Well-Founded Fear (PBS television broadcast June 5, 2000).

(175.) Id.

(176.) Id.

(177.) RSD Interview with Anonymous Applicant No. 17 in Cairo, Egypt (transcript on file with author) [hereinafter Transcript No. 17].

(178.) RSD Interview with Anonymous Applicant No. 12 in Cairo, Egypt (transcript on file with author) [hereinafter Transcript No. 12].

(179.) One hundred twenty-one questions, or 10.57 percent of 1,144, the total number of questions in the sample.

(180.) Transcript No. 17, supra note 177.

(181.) Transcript No. 12, supra note 178.

(182.) Transcript No. 17, supra note 177. See, for example, supra text accompanying note 177.

(183.) Transcript No. 17, supra note 177.

(184.) See, e.g., RSD Interview with Anonymous Applicant No. 5 in Cairo, Egypt (transcript on file with author) [hereinafter Transcript No. 5] ("Do you have any idea about who established the Pentecostal faith in the world?").

(185.) See supra notes 177-78 and accompanying text.

(186.) Transcript No. 5, supra note 184.

(187.) RSD Interview with Anonymous Applicant No. 22 in Cairo, Egypt (transcript on file with author).

(188.) UNHCR, supra note 45, [paragraph][paragraph] 28, 30.

(189.) Id. [paragraph] 30.

(190.) See supra Part III.B (detailing the targeting of churches and bible studies by the Eritrean government).

(191.) UNHCR, supra note 45, [paragraph] 29.

(192.) See RSD Interview with Anonymous Applicant No. 3 in Cairo, Egypt (transcript on file with author).

(193.) Id.

(194.) Id.

(195.) Id.

(196.) Id.

(197.) See, for example, RSD Interview with Anonymous Applicant No. 9 in Cairo, Egypt (transcript on file with author), in which the interviewer asked for clarification regarding what the applicant meant by "salvation."

(198.) See, e.g., RSD Interview with Anonymous Applicant No. 26 in Cairo, Egypt (transcript on file with author). Id. After the interviewer asked "Why did you convert?" the applicant made reference to being saved from his sins. Id. The interviewer then asked "why can't you be saved from your sins as a Protestant?" Id. After seven more questions and answered pressing for a more precise explanation of the theological differences that led to conversion. Id. In the course of this exchange, the interviewer three times asserted that Protestant Christianity offered the same precepts the applicant claimed for Pentecostalism, saying in one case, "I understand in Protestant religion you are saved through faith in God so I don't understand what you are saying," and in another case "I don't get it." Id.

(199.) See, e.g., RSD Interview with Anonymous Applicant No. 27 in Cairo, Egypt (transcript on file with author). Id. The interviewer asked "Why did you convert to Pentecostalism?" and then "What did you find in Pentecostal that was not in Orthodox?" Id. When the applicant mentioned a search for "eternal life," the interviewer asked if Orthodox followers do not get eternal life. Id. When the applicant said she had been living in sin before conversion, the interviewer asked "how is that so?" and then asked the same again after the applicant provided a theological viewpoint as an answer. Id. The applicant then made a statement about the importance of accepting Jesus as one's savior. Id. The interviewer responded by saying "In Orthodox they do accept Jesus as their personal savior, so what's the difference?" Id.

(200.) See RSD Interview with Anonymous Applicant No. 4 in Cairo, Egypt (transcript on file with author).

(201.) Id. The interviewer asked for specific Bible references for speaking in tongues, and when the applicant mentioned Acts 2, the interviewer replied, "I have the Holy Spirit." Id. The interviewer stated that the applicant's lawyer and the interpreter did as well, but that they do not necessarily speak in tongues. Id. The applicant then mentioned that a verse in Corinthians explained more, to which the interviewer said, "I don't need to read it. You are supposed to tell me." Id.

(202.) Id.

(203.) Id.

(204.) Id.

(205.) Id.

(206.) Anonymous Email, supra note 82.

(207.) See supra note 201.

(208.) See RSD Interview with Anonymous Applicant No. 11 in Cairo, Egypt (transcript on file with author).

(209.) See generally Samahon, supra note 3, at 2220-21 (discussing the difficulty in disassociating the sincerity of a person's belief and the truthfulness of the belief itself, as well as the problematic correlation between outward behavior and inward belief).

(210.) United States v. Ballard, 322 U.S. 87, 92-93 (1944) (Jackson, J. dissenting).

(211.) See RESTATEMENT (SECOND) OF TORTS [section] 526 (1977) (explaining the elements of fraud and the typical claim for fraud).

(212.) See Ballard, 322 U.S. at 93 (Jackson, J., dissenting) ("How can the Government prove these persons knew something to be false which it cannot prove to be false?").

(213.) Cf. Serbian E. Orthodox Diocese v. Milivojevich, 426 U.S. 696, 709 (1976) (stating that the resolution of religious doctrinal disputes "is for ecclesiastical and not civil tribunals" because "there is substantial danger that the State will become entangled in essentially religious controversies or intervene on behalf of groups espousing particular doctrinal beliefs").

(214.) Refugee Appeal No. 72323/2000, [paragraph] 8, at 2 (Sept. 25, 2001) (N.Z. Refugee Status Appeals Auth.),

(215.) See supra note 40.

(216.) See supra note 41.

(217.) SZAKD v Minister for Immigration [2004] FMCA 78, 2004 WL 723864 (19 Mar. 2004) (Austl.) ("Having regard to the current teachings of the Catholic Church, I am firmly of the view that a person of single sex orientation must have at least considered their position in the Church and whether they wished to continue practise Catholicism.").

(218.) See Homosexuality, CATHOLIC.COM, homosexuality.asp (last visited Oct. 15, 2010) ("[Homosexual] persons are called to fulfill God's will in their lives and, if they are Christians, to unite to the sacrifice of the Lord's cross the difficulties that they may encounter from their condition.").

(219.) Samahon, supra note 3, at 2212.

(220.) See Matthew 26:41 (NIV) ("Watch and pray so that you will not fall into temptation. The spirit is willing, but the body is weak.").

(221.) In cases that do not involve religion, credibility assessment may operate to force other types of minorities into a particular mold in an effort to discern genuine refugees from other migrants. For an example that is both colorful and cause for concern, see WAAG v Minister for Immigration [2002] FMCA 191, 2002 WL 2025758 (30 Aug. 2002) (upholding the appeal of an Iranian gay man who was rejected at first instance because the tribunal doubted his sexual orientation, in part because of his inability to cite any inspirational gay-oriented art and lack of knowledge of or interest in icons such as Oscar Wilde, Greco-Roman wrestling, Bette Midler, Madonna, or "any kind of emotion-stirring or dignity-arousing phenomena in the world around him").

(222.) Samahon, supra note 3, at 2215-16.

(223.) Toufighi v. Mukasey, 538 F.3d 988, 1000 (9th Cir. 2008).

(224.) Id. at 990-91.

(225.) Id. at 991.

(226.) Id. at 992-97.

(227.) Id. at 999-1000 (Berzon, J., dissenting).

(228.) See id. at 1000 ("Toufighi's beliefs has no direct bearing on the question whether he will be persecuted based on imputed religious beliefs.").

(229.) See INS v. Elias-Zacarias, 502 U.S. 478, 482 (1992) (holding that the persecuting group's motives for persecution, not its general motives, must determine whether asylum is proper).

(230.) Ward v. Canada, [1993] 2 S.C.R. 689, 747.

(231.) Toufighi, 538 F.3d at 1000 (Berzon, J., dissenting).

(232.) As the Seventh Circuit stated:

Certainly true conversion does matter in one sense. If one is a believer in a religious faith, one would presumably wish to practice that faith. Religious adherence could take the form of attending services, meeting with others of the same faith, personal prayer, or openly sharing one's belief, to name a few examples. If any activity necessary to a convert could trigger persecution in Iran, such a practice should be brought to the attention of the immigration judge. To evaluate the relevance of this practice to the life of the alien, the immigration judge should be satisfied with the sincerity of the alien's new religious commitment.

Najafi v. INS, 104 F.3d 943, 949 (7th Cir. 1997).

(233.) United States v. Bastanipour, 980 F.2d 1129, 1131 (7th Cir. 1992).

(234.) Id. at 1130.

(235.) Id. at 1132.

(236.) Id.

(237.) Samahon, supra note 3, at 2233-34.

(238.) Ward v. Canada, [1993] 2 S.C.R. 689, 747; see INS v. Elias-Zacarias, 502 U.S. 478, 489 (1992) (Stevens, J., dissenting) ("It does not matter to the persecutors what the individual's motivation is."); Kagan, supra note 7, at 368 ("Credibility is not one of the explicit criteria for refugee protection in international law.").

(239.) Cf. Joseph M. Perillo, The Origins of the Objective Theory of Contract Formation and Interpretation, 69 FORDHAM L. REV. 427 (2000) (discussing courts' longstanding preference for objective tests in contract cases).

(240.) United States v. Bastanipour, 980 F.2d 1129, 1132 (7th Cir. 1992) ("Whether Bastanipour believes the tenets of Christianity in his heart of hearts ... would not, we imagine, matter to an Iranian religious judge.").

(241.) Id.

(242.) Id.

(243.) Id.

(244.) RSD Interview with Anonymous Applicant No. 1 in Cairo, Egypt (transcript on file with author).

(245.) Id.

(246.) Id.

(247.) The applicant was subsequently recognized as a refugee.

(248.) See supra Part IV.A.2 (discussing the weaknesses of religious knowledge questions).

(249.) Seven hundred fifty-nine questions, or 66.3 percent of 1,144, the total number of questions in the sample.

(250.) Four hundred fifty-one questions, or 59.4 percent of 749, the total number of narrative questions in the sample.

(251.) Two hundred eighty-one questions, or 37 percent of 749, the total number of narrative questions in the sample.

(252.) UNHCR, supra note 45, [paragraph] 29.

(253.) E.g., Refugee Appeal No. 72323/2000, [paragraph] 8, at 2 (Sept. 25, 2001) (N.Z. Refugee Status Appeals Auth.).

(254.) UNHCR, Interviewing Applicants for Refugee Status (RLD4) (1995), available at

(255.) Id.

(256.) RLD4--Interviewing Applicants for Refugee Status, REFWORLD, 04.html (last visited Oct. 15, 2010).

(257.) UNHCR, supra note 254.

(258.) UNHCR, supra note 45, [paragraph] 29.

(259.) Four hundred forty-nine, or 39.2 percent of 1,144, the total number of questions in the sample.

(260.) Six hundred ninety-five, or 60.8 percent of 1,144, the total number of questions in the sample.

(261.) Some variation occurred. In six of thirty cases at least half of the questions were open in form, and in one case, 73 percent of the questions were open. However, in another case, only 11 percent (five of forty-five) of the questions were open.

(262.) Cf. UNHCR, Refugee Protection: A Guide to International Refugee Law, at 50 (Dec. 1, 2001), ("Parliamentarians can promote effectiveness [of RSD] by allocating sufficient resources for refugee status determination.").

(263.) RSD Interview with Anonymous Applicant No. 16 in Cairo, Egypt (transcript on file with author).

(264.) UNHCR, supra note 254.

(265.) E.g., Kagan, supra note 7, at 394 (providing an example of an applicant who appeared dishonest to the UNCHR, but, in truth, did not understand the questions being asked).

(266.) RSD Interview with Anonymous Applicant No. 10 in Cairo, Egypt (transcript on file with author).

(267.) See Kagan, supra note 7, at 385-89.

(268.) Cohen, supra note 54, at 295.

(269.) In seven transcripts, there were five or fewer questions about religious beliefs or knowledge, but in an equal number of cases there were more than fifteen questions about religious beliefs or knowledge. It is in this group that the most contentious exchanges about matters of faith between asylum seekers and UN refugee eligibility officers occur.

(270.) See supra notes 259-61 and accompanying text.

(271.) Herlihy et al., supra note 54, at 324.

(272.) See United States v. Ballard, 322 U.S. 78, 87 ("If one could be sent to jail because a jury in a hostile environment found those teachings false, little indeed would be left of religious freedom.").

(273.) For an example of a tribunal focusing on the risk of persecution but rejecting credibility based on a supposition about how a recent Christian convert would behave, see Refugee Appeal No. 70721/97, [paragraph][paragraph] 1, 9 (July 30, 1998) (N.Z. Refugee Status Appeals Auth.), In the appeal, the court accepted credibility of conversion, but doubted risk of persecution because
 [i]t is implausible that the appellant would specifically telephone
 his sister to tell her about his conversion, knowing that she was
 the kind of person who would discuss this information with others
 even outside her family, and knowing also that in such event the
 family would suffer great embarrassment and even be placed in great


(274.) In a decision regarding a Nigerian member of a Christian organization who said he had been attacked by Muslims for opposing Islamic law, an adjudicator rejected credibility but without any direct focus on religious belief. See X (Re), MA0-02843, [2002] CanLII 52613 (Immigration & Refugee Bd., Refugee Div., Jan. 3, 2002), The reasoning focused instead on the applicant's unexplained inability to provide details about the town where he said he had lived and about the Christian community to which he said he belonged. Id.

Michael Kagan, The author (J.D. Michigan 2000) was a Senior Fellow in International Human Rights Law at the American University in Cairo (AUC). This research was supported by AUC's Center for Migration and Refugee Studies in cooperation with African Middle East Refugee Assistance (AMERA) in Egypt. The views expressed are solely my own, but I am grateful for the comments, critiques, and suggestions that I received from Amanuel Abraham, Rania El Gindy, Jane Herlihy, Ann Lesch, Rebecca Mikhail, and Natalie Young.
COPYRIGHT 2010 Vanderbilt University, School of Law
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2010 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Kagan, Michael
Publication:Vanderbilt Journal of Transnational Law
Date:Nov 1, 2010
Previous Article:Al-Bihani, not so charming.
Next Article:Catfish, shrimp, and the WTO: Vietnam loses its innocence.

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |