Dismantling the Wall


Between 2017 and 2021, the Trump Administration waged an unprecedented battle on U.S. asylum structure, procedure, and substantive law. Seeking to alter long-standing legal principles and practices in a host of areas, the former administration’s efforts to demolish asylum protections were systematic and comprehensive. The Immigration Policy Tracking Project cataloged no fewer than ninety-six discrete policy and regulatory changes that the former administration implemented to curtail access to asylum.1  97 Policies, Immigr. Pol’y Tracking Project, https://immpolicytracking.org/policies/?subject_matter=asylum-withholding-and-cat [perma.cc/7WLW-D4A2]. The Immigration Policy Tracking Project (IPTP) catalogs the known immigration policies of the Trump Administration. Each entry contains underlying source documents, relevant predecessor policies, and the current status of each policy. IPTP was created by Professor Lucas Guttentag and is maintained by students at Stanford and Yale law schools supported by a team of immigration experts. See Immigr. Pol’y Tracking Project, https://immpolicytracking.org/home [perma.cc/C7SG-HTPT]. While some of the administration’s actions, such as the decision to separate children from their parents at the border, were carried out in the open, many other actions were largely hidden from public view. In their totality, scholars have characterized those changes without much hyperbole as the end of asylum in the U.S.,2See generally Andrew I. Schoenholtz , Jaya Ramji-Nogales & Philip S. Schrag, The End of Asylum (2021). a veritable administrative wall to refugees.

Despite widespread initial optimism upon the election of a new president and some incremental steps, the Biden Administration has yet to roll back the majority of these changes, let alone take steps to expand access to asylum or increase fairness in the system. Within his first month in office, President Biden promised to undertake a comprehensive review of the U.S. asylum system and promulgate regulations consistent with our international legal obligations within 270 days­­­­;3  See Exec. Order No. 14010, 86 Fed. Reg. 8,267, 8,271 (Feb. 2, 2021). however, that deadline has come and gone without any proposed regulations or an explanation for their absence.4Ctr. for Gender & Refugee Stud., Deadly Inertia: Needless Delay of “Particular Social Group” Regulations Puts Asylum Seekers at Risk 1 (2022), https://cgrs.uchastings.edu/sites/default/files/Deadly%20Inertia%20-%20PSG%20Regs%20Guide‌_Feb.%202022.pdf [perma.cc/MHQ6-72L5].

The contemporary U.S. asylum system was born through the robustly bipartisan 1980 Refugee Act.5Pub. L. No. 96-212, 94 Stat. 102 (1980) (codified as amended in scattered sections of 8 U.S.C.) From that moment to the present, the nation has not witnessed such unmitigated antipathy towards refugees and asylum seekers as during the Trump era. The toll paid by these changes, measured both in human lives and the erosion of our national values, is staggering. Yet, the fissures revealed by this unparalleled period of restriction of access to asylum can guide us both in understanding the extent of the present asylum crises and in knowing how best to move forward.

Through all its bluster about building a physical wall along the southern border to keep refugees and other immigrants out, the Trump Administration succeeded in erecting an administrative wall, preventing countless bona fide refugees from seeking or obtaining the protection for which they are eligible. To date, that wall has not been taken down. At best, the Biden Administration has tinkered with this barrier to refugees, and at worst, it has deliberately left some sections standing.

In this Essay, we will summarize the status quo of this crisis. We will highlight warning signs that began to appear even before the Trump Administration to understand how we reached this point. We will then propose solutions to chart a pathway forward, exploring strategies for implementing lasting reforms aimed at tearing down this administrative wall and replacing it with a more fair and welcoming system.

I.        Where Do We Stand Now?

After four years of sustained efforts by the Trump Administration to erode asylum protections in the U.S., the nation is still at a crossroads. A palpable tension persists between the long-standing humanitarian ethos that beckons to our shores those “huddled masses yearning to breathe free”6  Emma Lazarus, The New Colossus, Poetry Found., https://www.poetryfoundation.org/poems/46550/the-new-colossus [perma.cc/CQU7-R7RU]. and a xenophobic impulse bent on dramatically reducing immigration. As one Trump Administration official confessed, their “mantra [had] persistently been presenting aliens with multiple unsolvable dilemmas to impact their calculus for choosing to make the arduous journey to [the U.S.].”7Julia Ainsley, Stephen Miller Wants Border Patrol, Not Asylum Officers, to Determine Migrant Asylum Claims, NBC News (July 29, 2019, 7:31 PM), https://www.nbcnews.com/politics‌/immigration/stephen-miller-wants-use-border-agents-screen-migrants-cut-number-n1035831 [perma.cc/8YEC-8D86] (emphasis added). But, as the Somali refugee and poet Warsan Shire has poignantly observed, “no one leaves home unless home is the mouth of a shark.”8  “Home” by Warsan Shire, Facing Hist. & Ourselves, https://www.facinghistory.org‌/standing-up-hatred-intolerance/warsan-shire-homeh [perma.cc/WC9E-9BJH]. Thus, the Trump Administration’s efforts would have to be quite cruel to alter the outcome of the archetypal refugee’s tragic risk calculus.

In pursuit of their goal to create “unsolvable dilemmas” for refugees and asylum seekers, in 2018, the former administration shocked the country through its “zero-tolerance” program that intentionally separated thousands of children from their parents in the hopes of creating a “deterrent effect.”9Philip Bump, Here Are the Administration Officials Who Have Said That Family Separation Is Meant as a Deterrent, Wash. Post (June 19, 2018), https://www.washingtonpost.com/‌news/politics/wp/2018/06/19/here-are-the-administration-officials-who-have-said-that-family-separation-is-meant-as-a-deterrent [perma.cc/V8MH-E6D6]. Although the formal policy was rescinded in the wake of widespread public opprobrium, efforts to erect hurdles for refugees and asylum seekers persisted, often out of the public eye.

By the final years of the Trump Administration, through a combination of a lowered cap on refugee admissions and “increased vetting” processes for refugee admissions, the U.S. reached its lowest ebb of resettlement numbers in the forty-year history of the Refugee Act, descending beneath even the years that followed 9/11.10Kira Monin, Jeanne Batalova & Tianjian Lai, Refugees and Asylees in the United States, Migration Pol’y Inst. (May 13, 2021), https://www.migrationpolicy.org/article/refugees-and-asylees-united-states-2021 [perma.cc/6QWU-VBPY]. Yet, the true impact of this reduction went beyond just the denial of resettlement opportunities during one of the worst global refugee crises since WWII. The dramatic reductions in refugee admission also ensured long-term damage to the U.S. resettlement apparatus writ large by causing many refugee resettlement organizations—whose funding streams derive in part from per capita payments from the U.S. government for each refugee resettled—to close shop. By the end of the fiscal year 2019, more than one hundred resettlement offices in the U.S. were shuttered due to plummeting refugee admissions.11Tania Karas, US Refugee Agencies Wither as Trump Administration Cuts Numbers to Historic Lows,World (Sept. 27, 2019, 12:00 PM), https://www.pri.org/stories/2019-09-27/us-refugee-agencies-wither-trump-administration-cuts-numbers-historic-lows [perma.cc‌/N7TJ-HB4N].

Asylum processing at the border likewise ground to a halt as a result of iterative and culminating procedural changes. The litany of changes included: altering the internal guidelines for asylum officers to drive down positive credible fear findings; creating tortuously long wait times to seek asylum at the border (resulting in a waitlist that eventually climbed to 26,000 individuals); enlisting Customs and Border Protection officials—who had made headlines for coercing asylum seekers to withdraw their requests for asylum—to conduct asylum screenings; and creating the “Remain in Mexico” policy (deceptively called Migration Protection Protocols or MPP) that forced nearly 71,000 people to live in perilous conditions along the U.S.-Mexico border while waiting for a hearing, at which virtually no one would have access to counsel and where nearly all decisions resulted in denial.12  See Schoenholtz et al., supra note 2, at 47–67; Adolfo Flores & Hamed Aleaziz, The Supreme Court Ruled That Biden Must Restart Trump’s “Remain in Mexico” Program, BuzzFeed News (Aug. 25, 2021, 1:49 AM), https://www.buzzfeednews.com/article/adolfoflores/the-supreme-court-ruled-that-biden-must-restart-trumps?utm_source=dailybrief&utm_medium=email&utm_campaign=DailyBrief2021Aug25&utm_term=DailyNewsBrief [perma.cc/‌QTX5-P6QE]. Layered on top of these changes were additional grounds for denial for asylum seekers who sought to avoid this labyrinth of “unsolvable dilemmas” by entering the U.S. surreptitiously outside of a designated point of entry, or who transited through another country en route to the U.S. without first seeking asylum there.13  See Schoenholtz et al., supra note 2, at 47. The coup de grâce of the Trump Administration’s war on asylum at the border was the imposition of Title 42 expulsions, which used the COVID-19 pandemic as an opportunity to shut off what little trickle remained of asylum seekers pursuing protection along the southern border.14  See id. at 79–86.

The former administration’s ire was not cabined to just border policy, however; asylum processing within the interior likewise suffered as a result of significant substantive changes to asylum law. One category of asylum claims given particular attention during the Trump presidency involved putative refugees fleeing persecution inflicted by nonstate persecutors. Threatened by transnational criminal organizations and terrorist groups to domestic abusers and rebel factions, a significant number of asylum seekers flee their countries due to persecution committed at the hands of nongovernmental actors.15  See generally Charles Shane Ellison & Anjum Gupta, Unwilling or Unable? The Failure to Conform the Nonstate Actor Standard in Asylum Claims to the Refugee Act, 52 Colum. Hum. Rts. L. Rev. 441 (2021) (setting forth a framework for the nonstate actor standard under the Refugee Act). Indeed, the lion’s share of asylum claims brought by applicants fleeing Mexico and the Northern Triangle (i.e., El Salvador, Guatemala, and Honduras) are nonstate persecutor claims.16  See United Nations High Comm’r for Refugees, Women on the Run: First-Hand Accounts of Refugees Fleeing El Salvador, Guatemala, Honduras, and Mexico, 15 (2015) https://www.unhcr.org/56fc31864.html [perma.cc/HP34-2H8U] (stating that women fled due to violence at the hands of criminal armed groups and due to “brutal domestic violence”); United Nations High Comm’r for Refugees, Children on the Run: Unaccompanied Children Leaving Central America and Mexico and the Need for International Protection, 44–45 (2014) https://www.unhcr.org/56fc266f4.html [perma.cc/T3JY-ZFS6] (citing gang violence and domestic violence as reasons children flee).

The former administration characterized these claims as illegitimate and made significant efforts to greatly limit their probability of success. Specifically, former Attorney General Jeff Sessions wrote in a 2018 precedent decision that “[g]enerally,” claims based upon harms “perpetrated by non-governmental actors will not qualify for asylum.”17A-B- (Matter of A-B- I), 27 I. & N. Dec. 316, 320 (A.G. 2018) (emphasis added). This conclusion stemmed in part from the attorney general’s holding that when a persecutor is a nonstate actor, an applicant for protection must establish that her government either “condone[s]” her persecution or is “complete[ly] helpless[]” to protect her from such persecution,18  Id. at 337 (quoting Galina v. Immigr. & Naturalization Serv., 213 F.3d 955, 958 (7th Cir. 2000)). a standard vastly more difficult to satisfy than the preexisting one.19Ellison & Gupta, supra note 15, at 463–66.

His conclusion also stemmed from his assessment that “[w]hen private actors inflict violence based on a personal relationship with a victim, then the victim’s membership in a large[] group may well not be ‘one central reason’ for the abuse.”20  Matter of A-B- I, 27 I. & N. Dec. at 338. In other words, Sessions casted doubt upon whether such victims could prove the required causal link, or nexus, between the persecution they suffered or fear and a statutorily protected ground.

Finally, Sessions sought to further delegitimize claims involving domestic violence in particular by overturning an Obama-era precedent decision that had recognized that women who were persecuted by their abusers because of their inability to leave their domestic relationship were members of a valid “social group” under the Refugee Act.21  Id. at 346 (overruling A-R-C-G-, 26 I. & N. Dec. 388 (B.I.A. 2014)). In upending that settled precedent, Sessions argued that a social group so defined was impermissibly circular, insufficiently particular, and inadequately distinct in the eyes of the society in question.22  Id. at 334–36.

The following year, under Attorney General Barr, the administration continued the onslaught by overturning another Obama-era precedent that had recognized that persecution based upon one’s family membership could qualify one for asylum protection—another blow to a large category of claims.23L-E-A- (Matter of L-E-A- II), 27 I. & N. Dec. 581, 596 (A.G. 2019) (overruling in part L-E-A- (Matter of L-E-A- I), 27 I. & N. Dec. 40 (B.I.A. 2017)). Even in the final days of the Trump Administration, the acting attorney general continued working to restrict access to asylum by doubling down on the already heightened nonstate actor and nexus tests the administration had created in prior decisions.24A-B- (Matter of A-B- II), 28 I. & N. Dec. 199, 201–02 (A.G. 2021).

The foregoing summary is nonexhaustive, but the enumerated examples listed here illustrate the extent to which the Trump Administration could implement distortions to long-standing procedural and substantive requirements related to asylum eligibility, all without any assistance from Congress. In their combination, these changes amounted to an administrative wall of ineligibility, barring from protection those “huddled masses yearning to breathe free.”

For its part, the Biden Administration has recognized that violence “perpetrated by criminal gangs, trafficking networks, and other organized criminal organizations” along with “sexual, gender-based, and domestic violence” are among some of the root causes of migration from the Northern Triangle, and can give rise to viable claims for refugee protection.25  See Exec. Order No. 14010, 86 Fed. Reg. 8,267, 8,268 (Feb. 2, 2021). As noted above, President Biden has ordered a “comprehensive examination of current rules, regulations, precedential decisions, and internal guidelines” with the goal of evaluating “whether the United States provides protection for those fleeing domestic or gang violence . . . consistent with international standards.”26  Id. at 8,271. And the new administration has rolled back a small number of the harmful practices and precedents from the Trump era.27Alejandro N. Mayorkas, Termination of the Migrant Protection Protocols Program, Dep’t of Homeland Sec. 2 (June 1, 2021), https://www.dhs.gov/sites/default/files/publications/21_0601_termination_of_mpp_program.pdf [perma.cc/74SE-77FZ]; A-B- (Matter of A-B- III), 28 I. & N. Dec. 307 (A.G. 2021); L-E-A- (Matter of L-E-A- III), 28 I. & N. Dec. 304 (A.G. 2021); A-C-A-A- (Matter of A-C-A-A- II), 28 I. & N. Dec. 351 (A.G. 2021).

However, there remains much work to be done. The Biden Administration continues to wade through the morass of tangled policies, precedents, and regulations it inherited. It has failed to meet its own deadline for promulgating regulations that would correct some long-standing deficiencies in asylum law.28  See Ctr. for Gender & Refugee Stud., supra note 4. Hundreds of children have yet to be reunited with their parents.29Priscilla Alvarez, Parents of 391 Migrant Children Separated at Border Under Trump Still Have Not Been Found, Court Filing Says, CNN (May 19, 2021, 5:45 PM), https://www.cnn.com/‌2021/05/19/politics/ms-l-children-reunited/index.html [perma.cc/Y3CA-56SQ]. And the Remain in Mexico policy has been forcibly reinstated by court order.30Joseph Choi, Judge Orders Trump’s ‘Remain in Mexico’ Policy to Be Reinstated, Hill (Aug. 15, 2021, 8:13 AM), https://thehill.com/policy/national-security/567918-judge-orders-trumps-remain-in-mexico-policy-to-be-reinstated [perma.cc/YYH9-AVZF]; Flores & Aleaziz, supra note 12.

Yet, inaction is not the only problem at present. Deeply troubling developments are flowing out of the Biden Administration as well. The Biden Administration defied international and U.S. legal obligations through its mass expulsion of Haitians along the border in the fall of 2021.31  United Nations, UN Rights Experts Condemn US Expulsion of Haitian Migrants and Refugees, UN News (Oct. 25, 2021), https://news.un.org/en/story/2021/10/1103962 [perma.cc/RR66-9WWD]. On August 20, 2021, regulations were proposed to streamline the credible fear screening process that would also curtail a fulsome hearing before an immigration judge.32Rebecca Beitsch, Biden Administration Seeks to Speed Review of Asylum Cases, Hill (Aug. 18, 2021, 10:09 AM), https://thehill.com/policy/national-security/568351-biden-administration-seeks-to-speed-review-of-asylum-cases [perma.cc/67J9-CCNH]. While the proposed regulations helpfully provide for nonadversarial adjudication of applications for asylum, withholding of removal, and protection under the Convention Against Torture by asylum officers following positive credible fear screenings, they also make review of errant negative decisions more difficult to reverse.33  See Ted Hesson, U.S. Plans to Double Number of Asylum Officers in Biden Border Overhaul, Reuters, https://www.reuters.com/world/us/us-plans-double-number-asylum-officers-biden-border-overhaul-2021-08-18 [perma.cc/FB6Z-6TSN] (last updated Aug. 18, 2021, 11:31 AM). Despite a number of comments criticizing a deficiency of due process in the proposal, the interim final rule published on March 29, 2022—while modifying the review procedures—continues to prioritize rapid processing and tight deadlines to the exclusion of full and fair adjudications.34Procedures for Credible Fear Screening and Consideration of Asylum, Withholding of Removal, and CAT Protection Claims by Asylum Officers, 87 Fed. Reg. 18,078 (Mar. 29, 2022) (to be codified at 8 C.F.R. pts. 208, 212, 235, 1003, 1208, 1235, 1240). And perhaps most concerning, the administration not only continued the use of harmful Title 42 public health expulsions at the southern border, but it vigorously defended them even in the face of court rulings that held the expulsions unlawful and pronouncements by public health officials that they are unnecessary.35Claire Hansen, Biden Administration Digs in Heels on Controversial Title 42 Border Order, U.S. News (Aug. 3, 2021, 12:02 PM), https://www.usnews.com/news/national-news/articles/2021-08-03/biden-administration-digs-in-heels-on-controversial-title-42-border-order [perma.cc/A6LN-K9PX]; Sarah Sherman-Stokes & Lindsay M. Harris, Despite Promises, Biden Looks a Lot Like Trump on Border Issues, Bloomberg L. (Oct. 20, 2021, 4:00 AM), https://news.bloomberglaw.com/us-law-week/despite-promises-biden-looks-a-lot-like-trump-on-border-issues [perma.cc/74TJ-AR58]; Letter to CDC Director Walensky, HHS Secretary Becerra, and DHS Secretary Mayorkas on the August 2021 Title 42 Order, Colum. Pub. Health (Sept. 1, 2021), https://www.publichealth.columbia.edu/sites/default/files/sept_1_2021_title_42‌_letter.pdf [perma.cc/KEK2-UVKB] [hereinafter Letter to CDC Director]; Huisha-Huisha v. Mayorkas, No. 21-5200, 2022 WL 628061, at *4 (D.C. Cir. Mar. 4, 2022) (holding that Title 42 expulsions without screening for relief under 8 U.S.C. § 1231 is contrary to law); Title 42 Termination with Respect to Unaccompanied Noncitizen Children, CDC (Mar. 12, 2022), https://www.cdc.gov/media/releases/2022/s0311-title-42-termination.html [perma.cc/FR7W-PEUZ] (exempting unaccompanied children but leaving intact existing Title 42 orders for adult asylum seekers). Although the administration recently announced that the order calling for the expulsions will be terminated as of May 23, 2022,36  CDC Public Health Determination and Termination of Title 42 Order, CDC (Apr. 1, 2022), https://www.cdc.gov/media/releases/2022/s0401-title-42.html [perma.cc/SM28-QR7L]. the order could be reinstated at any time, and there is some risk that the termination could be enjoined, as it has already been challenged by a group of states in federal court.37Rebecca Beitsch, Three GOP States Sue over End of Title 42, Hill (April 4, 2022, 11:47 AM), https://thehill.com/latino/3258301-three-gop-states-sue-over-end-of-title-42 [perma.‌cc/38CH-765D]. In short, the U.S. asylum system remains in a state of crisis. And this crisis coincides with the large influx of Afghan evacuees arriving in the U.S., thousands of whom must seek asylum.38Michelle Hackman, Afghan Refugees in the U.S.: How They’re Vetted, Where They’re Going, and How to Help, Wall St. J. (Jan. 21, 2022, 7:26 AM), https://www.wsj.com/articles/afghan-refugees-in-the-u-s-how-theyre-vetted-where-theyre-going-and-how-to-help-11630677004 [perma.cc/MV2R-43SB].

Significant continued efforts will be required just to reverse Trump era changes and revert to the status quo ante. Yet, simply undoing the damage wrought by the former administration is not enough, because any advances made now can be just as easily reversed by a future neo-Trumpian president. If the U.S. asylum system is to be securely reconstructed, additional lasting reforms are required. To lay the groundwork for exploring such durable solutions, we will proceed to analyze how we got here.

II.      How Did We Reach This Point?

In the wake of the horrors of the Holocaust, the global community of nations convened to create the 1950 U.N. Convention Relating to the Status of Refugees (popularly known as the Refugee Convention). This document, as amended by the 1967 Protocol Relating to the Status of Refugees, continues to serve as the historical foundation of the contemporary refugee framework globally and within the U.S.39The United States acceded to Articles 2 through 34 of the Refugee Convention when it signed on to the Protocol in 1968. See Immigr. & Naturalization Serv. v. Stevic, 467 U.S. 407, 416 (1984); Immigr. & Naturalization Serv. v. Cardoza-Fonseca, 480 U.S. 421, 429 (1987). However, it was not until the passage of the 1980 Refugee Act that the U.S. brought itself into full conformity with its international obligations under the Refugee Convention and Protocol.

The Refugee Act defines a refugee as:

[A]ny person who is outside any country of such person’s nationality or, in the case of a person having no nationality, is outside any country in which such person last habitually resided, and who is unable or unwilling to return to, and is unable or unwilling to avail himself or herself of the protection of, that country because of persecution or a well-founded fear of persecution on account of race, religion, nationality, membership in a particular social group, or political opinion.408 U.S.C. § 1101(a)(42).

Legion are the opinions and decisions—issued by courts and the administrative agency entrusted with the interpretation and application of this statute—describing, defining, and refining the various facets to the refugee definition. While Congress has passed numerous laws affecting asylum eligibility and procedure (including statutory bars to filing and relief, screening mechanisms for claims asserted at the border, relevant burdens of proof, and guidance related to making credibility determinations), it has left the core refugee definition largely untouched. Instead, Congress has preferred to grant wide latitude to administrative agencies to fill in gaps through the regulatory process, case-by-case adjudication, and interagency policy guidance. And courts have fashioned judicial doctrines of deference—dubbed Chevron deference when the agency is construing an ambiguous statute, and Auer deference when it is interpreting an ambiguous regulation—that provides additional latitude to the agency to shape asylum policy.

This legislative flexibility has been both implied and explicit. In addition to the regular functioning of administrative agencies by virtue of the Administrative Procedures Act, Congress specifically provided that asylum is a discretionary form of relief and that the Attorney General has been granted the authority to establish by regulation “additional limitations and conditions, consistent with [the statute], under which [a noncitizen] shall be ineligible for asylum.”41  Id. § 1158(b)(2)(c).

The Trump Administration unequivocally demonstrated that the inherent and explicit discretionary powers Congress conferred upon the executive could be easily weaponized and used, if not to defeat the purpose of the Refugee Act, then at least to undermine it. Although many of President Trump’s policy and regulatory proposals aimed at limiting procedural access to asylum were quickly enjoined by the courts,42  See e.g., E. Bay Sanctuary Covenant v. Barr, 391 F. Supp. 3d 974 (N.D. Cal. 2019) (enjoining the third country transit bar); Pangea Legal Servs. v. U.S. Dep’t of Homeland Sec., 512 F. Supp. 3d 966 (N.D. Cal. 2019) (enjoining Procedures for Asylum and Withholding of Removal; Credible Fear and Reasonable Fear Review, 85 Fed. Reg. 80,274 (Dec. 11, 2020) (to be codified at 8 C.F.R. pts. 208, 235, 1003, 1208, and 1235), dubbed the “death to asylum” rule, see infra note 83); E. Bay Sanctuary Covenant v. Trump, 354 F. Supp. 3d 1094 (N.D. Cal. 2018) (enjoining President Trump’s entry ban proclamation and rule for asylum seekers at the southern border). some of the more pernicious substantive changes to asylum eligibility, such as those relating to the nonstate actor test, nexus, and social group law received widespread deference from the courts of appeals.

For example, in the nonstate-actor context, the Trump Administration, as noted above, attempted to heighten the standard for demonstrating state involvement in persecution. Long before the passage of the 1980 Refugee Act, U.S. adjudicators recognized that a refugee can be one fleeing state-perpetrated persecution or nonstate persecution from which the state is either “unable or unwilling” to provide effective protection.43  See, e.g., Rosa v. Immigr. & Naturalization Serv., 440 F.2d 100, 102 (1st Cir. 1971); Tan, 12 I. & N. Dec. 564, 568 (B.I.A. 1967); Pierre, 15 I. & N. Dec. 461, 462 (B.I.A. 1975). In the decades that followed, the unable-or-unwilling test was recognized and accepted by the immigration agency and every federal court of appeals.44Ellison & Gupta, supra note 15, at 467–91. But in 2018, Attorney General Jeff Sessions wrote in Matter of A-B- I that to satisfy the nonstate-actor requirement, an applicant must show that their government either “condone[s]” the persecution or is “complete[ly] helpless[]” to protect them from such persecution.45  Matter of A-B- I, 27 I. & N. Dec. 316, 337 (A.G. 2018). Despite the linguistic difference between the words “unwilling” and “condone” or the words “unable” and “completely helpless,” Sessions did not acknowledge that he was heightening the standard, let alone provide a rational explanation for the departure. Sessions’s omission was even more striking given evidence showing that asylum seekers are twice as likely to lose their case when the condone-or-complete-helplessness language is cited than when the unwilling-or-unable test is used.46Ellison & Gupta, supra note 15, at 485.

Nevertheless, courts of appeals in the Second,47Scarlett v. Barr, 957 F.3d 316, 332–33 (2d Cir. 2020). Third,48Galeas Figueroa v. Att’y Gen., 998 F.3d 77, 91 (3d Cir. 2021). and Fifth49Gonzales-Veliz v. Barr, 938 F.3d 219, 232–34 (5th Cir. 2019). Circuits have deferred to Session’s condone-or-complete-helplessness formulation. While some courts of appeals have questioned the validity of the condone-or-complete-helplessness test,50  See, e.g., Galloso v. Barr, 954 F.3d 1189, 1192 (8th Cir. 2020) (holding that because the familiar unable-or-unwilling test came first, it must control); Rosales Justo v. Sessions, 895 F.3d 154, 164 (1st Cir. 2018) (citing Matter of A-B- I but applying the familiar unwilling-or-unable test); Juan Antonio v. Barr, 959 F.3d 778, 790 n.3, 795 (6th Cir. 2020) (finding that Matter of A-B- I, including the condone-or-complete-helplessness standard, had been abrogated). only the Court of Appeals for the D.C. Circuit has clearly held that the condone-or-complete-helplessness articulation is a new and heightened standard, inconsistent with the traditional test.51Grace v. Barr, 965 F.3d 883, 898–900 (D.C. Cir. 2020) (concluding that Matter of A-B- I adopted a “new, more demanding standard ‘without acknowledging or explaining the change’”) (quoting Appellee’s Brief at 52, Grace v. Barr, 965 F.3d 883 (D.C. Cir. 2020) (No. 19-5013)). Even then, the court held that the agency failed to provide any explanation for the change; it did not foreclose the possibility of the agency providing a reasonable explanation to which the court would later defer.52  Id. at 900–03.

An equally salient example of the executive branch’s ability to unilaterally undermine the purposes of the Refugee Act exists in the area of social group law, where the vast majority of courts have afforded deference to the agency’s three-part test for analyzing the cognizability of a particular social group.53  See, e.g., Granada-Rubio v. Lynch, 814 F.3d 35, 38 (1st Cir. 2016); Ordonez Azmen v. Barr, 965 F.3d 128, 134 (2d Cir. 2020); S.E.R.L. v. Att’y Gen., 894 F.3d 535, 550–55 (3d Cir. 2018); Nolasco v. Garland, 7 F. 4th 180, 187 (4th Cir. 2021); Gonzales-Veliz v. Barr, 938 F.3d 219, 230 (5th Cir. 2019); Zaldana Menijar v. Lynch, 812 F.3d 491, 498 (6th Cir. 2015); Fuentes v. Barr, 969 F.3d 865, 871 (8th Cir. 2020); Villegas Sanchez v. Garland, 990 F.3d 1173, 1180 (9th Cir. 2021); Rodas-Orellana v. Holder, 780 F.3d 982, 992–93 (10th Cir. 2015); Alvarado v. U.S. Att’y Gen., 984 F.3d 982, 989 (11th Cir. 2020). While the Board of Immigration Appeals (BIA) had originally adopted a common sense approach to the statutory term “particular social group,” it later sought to narrow eligibility by further defining that term.

Initially, the BIA employed the canon of statutory construction ejusdem generis (meaning of the same kind) to conclude that the ambiguous term should be construed consistently with the other protected characteristics listed in the refugee definition: race, religion, nationality, and political opinion.54Acosta, 19 I. & N. Dec. 211, 233 (B.I.A. 1985). In a carefully reasoned decision in 1985, the BIA held that the common thread that united each of these protected characteristics is that they all were grounded in an immutable characteristic, that is, a characteristic that was so fundamental to one’s identity that one either could not change it or should not be required to change it.55  Id. Thus, the very sensible immutability test was born for evaluating social groups. A particular social group consisted of a group of individuals who shared a common immutable characteristic.

Under this test, one would expect that family, gender and sexual orientation, and shared past experiences that place one at risk of persecution would pass muster. Indeed, a number of groups were readily identified using this immutability test. In 1988, for example, former police and military were recognized as possessing a shared past experience that could constitute a cognizable social group.56Fuentes, 19 I. & N. Dec. 658 (B.I.A. 1988). In a 1990 decision, members of the LGBT community were acknowledged as possessing an immutable characteristic.57Toboso-Alfonso, 20 I. & N. Dec. 819 (B.I.A. 1990). In 1996, clan membership and women fearing female genital mutilation were likewise recognized as valid social groups.58H-, 21 I. & N. Dec. 337 (B.I.A. 1996); Kasinga, 21 I. & N. Dec. 357 (B.I.A. 1996).

However, as alluded to above, the agency apparently began to harbor concerns that this social group test—though entirely workable—opened access to protection to just too many people. Thus, beginning in 2006 under the Bush Administration, the BIA began adding to the social group analysis in an effort to narrow the pool of potential applicants.59  See C-A-, 23 I. & N. Dec. 951 (B.I.A. 2006). The BIA announced that in addition to possessing an immutable characteristic, a viable social group would now also need to possess social visibility.60  Id. at 959–61. The BIA claimed that this addition was consistent with its prior decisions, but questions arose almost immediately regarding how some groups—such as past former association and sexual orientation—were visible to the extent that those characteristics were often hidden.61  See, e.g., Gatimi v. Holder, 578 F.3d 611, 615 (7th Cir. 2009); Valdiviezo-Galdamez v. Att’y Gen., 663 F.3d 582, 604 (3d Cir. 2011).

Such questions notwithstanding, the BIA continued down this path by delineating a particularity requirement the following year. There, the BIA held that a viable social group must be clearly defined and not amorphous, subjective, inchoate, or diffuse.62A-M-E-, 24 I. & N. Dec. 69, 76 (B.I.A. 2007). In applying this test, the BIA reasoned that wealth status was simply too subjective and amorphous to provide a measurable benchmark for group membership,63  Id. and thus the particularity prong was born.

The next year, in 2008, the BIA employed the newly minted social visibility and particularity requirements to hold that Honduran youth perceived to be affiliated with gangs and Salvadoran youth subjected to gang recruitment efforts likewise failed under the two additional social group prongs.64E-A-G-, 24 I. & N. Dec. 591 (B.I.A. 2008); S-E-G-, 24 I. & N. Dec. 579 (B.I.A. 2008). In the years that followed, courts began to question the addition of these two nascent social group requirements and to ask whether the original collection of social groups recognized using the immutability test would survive under the newly formulated rules. In particular, courts reacted to some statements from the BIA suggesting that groups had to be visible in the sense that one could discern group membership simply by looking at a person. If this is what social visibility meant, then the new requirement would unquestionably rule out most of the groups recognized under the immutability test (for example, LGBTQ individuals or women who had not yet undergone female genital cutting), particularly given that refugees go to great lengths to hide the characteristic that gives rise to their persecution.65  See, e.g., Gatimi, 578 F.3d at 615; Valdiviezo-Galdamez, 663 F.3d at 604. Moreover, returning to the principle of ejusdem generis that first guided the BIA in articulating the immutability test, it would seem that the statutorily recognized characteristics of religion, nationality, and political opinion could likewise fail under any requirement that one’s protected characteristic be visible to the eye.

To address these criticisms, the BIA issued a pair of decisions in 2014, clarifying that it was not requiring “ocular visibility,” but “social distinction”; that is, the society in question should recognize the group as being set apart in some way, distinct from other members of the country of feared persecution.66W-G-R-, 26 I. & N. Dec. 208, 211–12 (B.I.A. 2014); M-E-V-G-, 26 I. & N. Dec. 227, 236 (B.I.A. 2014). Since that time, almost every court of appeals in the country has deferred to this refined three-part test, though it has continued to garner scholarly criticisms that the formulation is not workable and unfairly applied.67  See Benjamin Casper, Katherine Evans, Julia DiBartolomeo Decker & Haley Steptoe, Matter of M-E-V-G- and the BIA’s Confounding Legal Standard for “Membership in a Particular Social Group,” Immigr. Briefings, June 2014, at 11–13, 19–22.

Nevertheless, the three-prong social group test is as firmly entrenched now as it has even been. And it continues to provide significant flexibility to the executive branch to define the scope of protection to social groups, both new and old. As mentioned above, the Trump Administration used this flexibility to hold that previously recognized gender-based and family-based groups were no longer cognizable under the test. Consequently, at the end of the Trump Administration, looking back to when the BIA first departed from the immutability test in 2006, there was not a single precedent BIA decision left standing recognizing a viable social group that satisfied the three-pronged test.68Following the creation of the three-prong test, A-R-C-G- and Matter of L-E-A- I were the only two precedent decisions that had recognized examples of cognizable social groups that could satisfy that test by the start of the Trump Administration. See supra notes 21–23 and accompanying text. Since both decisions were vacated during his presidency, not a single example of a cognizable group recognized in a published BIA decision was left by the time he left office.

Similarly, courts have taken a very deferential posture in relation to the agency’s elaborations upon the nexus requirement, which mandates applicants to prove that their persecution is “on account of” a statutorily protected characteristic.69  See Deborah E. Anker, Law of Asylum in the United States § 5:8 (2021) (noting widespread deference to the BIA’s nexus decision in S-P-, 21 I. & N. Dec. 486 (B.I.A. 1996)). Long before the Trump Administration, the lack of statutory or regulatory guidance as to the proper formulation of the nexus requirement resulted in unequal application of the nexus rule. With respect to nonstate persecutor claims in particular, immigration judges and the BIA often weighed more heavily nonprotected reasons for the persecution, even when protected reasons were present. In domestic violence claims, for example, the agency denied asylum despite ample country conditions evidence showing that the majority of victims of abuse are women, reasoning that the abuse occurred on account of the abuser’s desire to control the victim or simply because the abuser is a “despicable person,” rather than on account of the victim’s gender or membership in a particular social group.70  See R-A-, 22 I. & N. Dec. 906, 927 (B.I.A. 1999) (“In sum, we find that the respondent has been the victim of tragic and severe spouse abuse. We further find that her husband’s motivation, to the extent it can be ascertained, has varied; some abuse occurred because of his warped perception of and reaction to her behavior, while some likely arose out of psychological disorder, pure meanness, or no apparent reason at all.”); Karen Musalo & Stephen Knight, Gender-Based Asylum: An Analysis of Recent Trends, 77 No. 42 Interpreter Releases 1533, 1535 (2000) (stating that in D-K- (B.I.A. Jan. 20, 2000), the immigration judge “denied asylum, ruling that Ms. Kuna had not been persecuted on account of her membership in either group, or for any political reason, but solely because her husband was ‘a despicable person’”). The agency made similarly problematic nexus findings in other contexts, including in cases involving forced sterilization, human trafficking, forced marriage, religion, gang violence, sexual orientation, and membership in a family.71Anjum Gupta, The New Nexus, 85 U. Colo. L. Rev. 377, 389–422 (2014).

While Congress amended the asylum statute in 2005 through passage of the REAL ID Act, it largely codified what had already been widely recognized by courts72A notable exception, however, existed with respect to Ninth Circuit case law that had held that the nexus requirement was satisfied where the persecutor was motivated at least in part—rather than in central part—by a protected ground. H.R. Rep. No. 109-72, at 162–63 (2005).—namely, that a refugee need establish that her protected characteristic was or would be at least one central reason for her past or feared persecution.738 U.S.C. § 1158(b)(1)(B)(i). The BIA would later interpret this statutory language to signify that in mixed-motive asylum cases “the protected ground cannot play a minor role[, and] . . . cannot be incidental, tangential, superficial, or subordinate to another reason for harm.”74J-B-N-, 24 I. & N. Dec. 208, 214 (B.I.A. 2007). Courts mostly deferred to this interpretation as well.75  See Anker, supra note 69, at § 5:12.

Arguing that the REAL ID Act did little to solve the problems resulting from the lack of standards in the nexus analysis, some legal scholars argued that a simple “but-for” test would suffice in most claims.76  See, e.g., Stephen H. Legomsky & Cristina M. Rodríguez, Immigration and Refugee Law and Policy 985–90 (5th ed. 2009); Anjum Gupta, The New Nexus, 85 U. Colo. L. Rev. 377 (2014) [hereinafter Gupta, New Nexus]; Anjum Gupta, Nexus Redux, 90 Ind. L.J. 465 (2015) [hereinafter Gupta, New Redux]. Again leveraging this lack of statutory guidance and the judicial deference afforded to the agency, the Trump Administration, through Acting Attorney General Rosen, held that “[t]o establish the necessary nexus, the protected ground: (1) must be a but-for cause of the wrongdoer’s act; and (2) must play more than a minor role[, and] . . . cannot be incidental or tangential to another reason for the act.”77  Matter of A-B- II, 28 I. & N. Dec. 199, 208, 212 (A.G. 2021) (emphasis added). Elaborating further, the attorney general explained that where an individual is targeted as a means to another end—for example, targeting a son in order to force a father to comply with the persecutor’s demand—there is no nexus because the persecutor has no particular animus against the family.78  Id. at 209. So though the “applicant’s status as a member of his father’s immediate family may have been a but-for cause of the harm he suffered,” it was “not a ‘central’ reason” according to Rosen.79  Id. Rosen’s application of his own rule makes clear that the second part of the test, in effect, swallows the first and makes the test more, not less, burdensome than it had been, even after passage of the REAL ID Act. The Trump Administration used the opportunity created by agency discretion to interpret and apply the law to fashion a policy that would provide the least amount of protection possible.

Finally, and perhaps most tellingly, just prior to leaving office, the Trump Administration tried to implement a cadre of new discretionary bars to asylum by invoking the statutory authority given to the attorney general to establish by regulation “additional limitations and conditions . . . under which [a noncitizen] shall be ineligible for asylum.”808 U.S.C. § 1158(b)(2)(c). The regulation represented the most radical and breathtaking restriction on the scope of asylum ever proposed.81Procedures for Asylum and Withholding of Removal; Credible Fear and Reasonable Fear Review, 85 Fed. Reg. 80,274 (Dec. 11, 2020) (to be codified at 8 C.F.R. pts. 208, 235, 1003, 1208, and 1235). It designated three factors as “significantly adverse.”82  Id. at 80, 282. Those factors barred asylum for individuals who: (1) “unlawful[ly] ent[er]” or attempt to enter the United States, with just a narrow exception for those fleeing from “a contiguous country;” (2) fail to “seek asylum or refugee protection in at least one country” through which they “transited before entering the” U.S.; and (3) use “fraudulent documents to enter the United States, unless the [noncitizen] arrived . . . by air, sea, or land directly from the applicant’s home country without transiting through any other country.”83Id. The rule also created nine other “adverse factors” that all but guaranteed “the denial of asylum as a matter of discretion.”84Id. Such factors included any individual who has: (1) “spent more than 14 days in any one country that permitted application for refugee, asylee, or similar protections prior to entering or arriving in the United States”; (2) traveled through “more than one country prior to arrival in the” U.S.; (3) incurred certain criminal convictions that remain valid for immigration purposes; (4) accrued “more than one year of unlawful presence” prior to filing an application for asylum; (5) failed to file a required tax return; (6) “had two or more prior asylum applications denied for any reason”; (7) “withdr[ew] an asylum application with prejudice or . . . abandoned an asylum application”; (8) missed an asylum interview; or (9) failed to “file a motion to reopen within one year” of a change in circumstances.85  Id. Only where such applicants could establish “extraordinary circumstances . . . involving national security or foreign policy considerations,” or demonstrate “by clear and convincing evidence, that the denial . . . would result in an exceptional and extremely unusual hardship” can they possibly overcome these new asylum bars.86  Id.

The rule appeared to be a calculated attempt to guarantee the denial of as many claims as possible. Given the urgency surrounding the need to flee persecution, many bona fide asylum seekers are unable to wait in their country long enough to obtain a visa to enter the U.S. For those few who have the resources and ability to seek a U.S. visa from a third country, the requirement of nonimmigrant intent ensures that if an applicant discloses her fear of persecution, and thus her intent to abandon her foreign residence, she will not be granted a U.S. visa to enter. If she obtains a visa by misrepresenting her true intentions to seek asylum in the U.S. or by presenting fraudulent documents, she would have faced denial under the new rule. If she remained for more than two weeks in that third country while waiting for her visa decision, she would have faced an additional ground of denial under the rule. If she accurately represented her intentions during her visa interview—resulting in a denial of her visa—and was thus forced to travel through other countries to present herself at a U.S. port of entry, she would have added yet another reason for denial under the rule. Once she arrived at the U.S. border, because of arbitrarily created wait times and the Remain in Mexico policy, she would have been denied entry. If, out of fear and desperation, she sought to enter the U.S. outside of a port of entry to seek asylum from within the U.S., she would have incurred yet another reason for denial under the rule.

This final regulation was aptly dubbed the “death to asylum” regulation, and it represented the apotheosis of the Trump Administration’s efforts to create “unsolvable dilemmas” for refugees and asylum seekers.87Bill Frelick, The Trump Administration’s Final Insult and Injury to Refugees, Hum. Rts. Watch (Dec. 11, 2020, 6:00 AM), https://www.hrw.org/news/2020/12/11/trump-administrations-final-insult-and-injury-refugees# [perma.cc/5QFE-BZT5]. In fact, White House Senior Policy Adviser, Stephen Miller, candidly admitted that stopping asylum seekers was “all [he] care[d] about”88Molly Olmstead, Stephen Miller: Stopping Asylum-Seekers Is “All I Care About,” Slate (Feb. 21, 2020, 1:21 PM, https://slate.com/news-and-politics/2020/02/stephen-miller-immigration-this-is-my-life.html [perma.cc/3MTV-L4EH]. and that he would have been “happy if not a single refugee foot ever again touched America’s soil.”89  See Ellen Cranley, Stephen Miller Said He ‘Would be Happy If Not a Single Refugee’ Came to the US, According to Ex Trump Aide, Bus. Insider (Jan. 28, 2019, 10:42 AM), https://www.businessinsider.com/stephen-miller-said-he-would-be-happy-if-not-a-single-refugee-came-to-us-2019-1 [perma.cc/ZQ4C-948V]; see also Michael D. Shear & Maggie Habberman, Trump’s Temporary Halt to Immigration Is Part of Broader Plan, Stephen Miller Says, N.Y. Times (Aug. 26, 2020), https://www.nytimes.com/2020/04/24/us/politics/coronavirus-trump-immigration-stephen-miller.html [perma.cc/Z34A-7QT6]. While this “death to asylum” regulation never went into effect—having been enjoined by a federal court because Chad Wolf was not lawfully serving as Acting Secretary of DHS at the time the rules were promulgated—the statutory authority employed by the Trump Administration to create this rule (i.e., 8 U.S.C. § 1158(5)(B)) remains on the books.

Each of the foregoing examples illustrates that the conditions that made possible many of the changes implemented by the Trump Administration long predated President Trump. The posture of extreme deference permeating the U.S. asylum system emanates both from the asylum statute itself and from long-standing principles of administrative law. The significant degree to which courts have deferred to agency constructions of the asylum statute and regulations are well-established features of Chevron and Auer deference. Under the Chevron deference doctrine, a court first determines whether Congress clearly expressed its intent with regard to “the precise question at issue” by employing the “traditional tools of statutory construction.”90Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 842, 843 n.9 (1984). Where a statute is ambiguous, however, the court is required to defer to any reasonable or permissible agency interpretation, even where that interpretation is contrary to a prior construction by that court.91  Id.; United States v. Mead Corp., 533 U.S. 218, 229, 237 (2001); Nat’l Cable & Telecomms. Ass’n v. Brand X Internet Servs., 545 U.S. 967, 984 (2005). This two-step approach also applies to an agency’s interpretation of its own regulations (referred to as Auer deference).92Auer v. Robbins, 519 U.S. 452, 461 (1997); Kisor v. Wilkie, 139 S. Ct. 2400, 2412 (2019). Taken together, these statutory and judicially created doctrines of deference grant the executive branch near hegemony in fashioning U.S. asylum policy.

III.    Where Do We Go from Here?

As noted above, the Biden Administration has begun the process of rolling back Trump-era executive orders, vacating some past attorney general opinions, issuing new policy guidelines, and embarking on the first steps to promulgate new regulations, ostensibly to realign the U.S. with its international treaty obligations.93  See, e.g., Matter of A-B- II, 28 I. & N. Dec. 199 (A.G. 2021); Matter of L-E-A- III, 28 I. & N. Dec. 304 (A.G. 2021); Exec. Order No. 14010, 86 Fed. Reg. 8,267 (Feb. 2, 2021). As laudable as the administration’s promises are, we are still waiting to see how they will be implemented. As measured by the administration’s foot-dragging on new substantive asylum regulations, use of Title 42 expulsions, and its mishandling of claims presented by Haitians at the southern border, there are serious reasons to be concerned. However, even if positive changes are on the horizon, none will be insulated from further revision or deconstruction by some future administration that wishes to revert to the cruel policies of the past. So where do we go from here?

We contend that Congress, the administration, and the courts must take action to provide an enduring solution to this crisis, one that ensures that future presidents cannot defy our international obligations under the Refugee Convention and Protocol. We do not purport to offer a comprehensive solution here, but we do endeavor to sketch out a broad array of solutions that would markedly improve the status quo.

First, we concur with the excellent policy proposals that have been advanced by scholars to improve the functioning and fairness of the asylum system.94  See e.g., Schoenholtz et al., supra note 2, at 124–29; T. Alexander Aleinikoff, Donald Kerwin, Ctr. for Migration Stud. & Zolberg Inst. on Migration & Mobility, Improving the U.S. Immigration System in the First Year of the Biden Administration (2020), https://cmsny.org/wp-content/uploads/2020/11/Improving-the-US-Immigration-System_Proposals_FINAL.pdf [perma.cc/BTT5-MU26]. Such reforms include repealing the attorney general’s authority to create additional discretionary bars; ensuring that asylum, like withholding of removal, is a mandatory form of protection; repealing the provision of law relied upon by President Trump to create the Remain in Mexico policy; granting increased independence to immigration courts by making them Article I courts; and creating mechanisms for indigent asylum seekers to be appointed counsel.95See sources cited supra note 94. Several of these changes were proposed in the Refugee Protection Act of 2019, sponsored by Senator Patrick Leahy and Representative Zoe Lofgren.96S. 2936, 116th Cong. (2019); H.R. 5210, 116th Cong. (2019). However, the bill has yet to pass.

Adding to that list of substantive and procedural remedies, we advocate for an explicit legislative prohibition on prosecuting asylum seekers who enter the U.S. without authorization, consistent with our international treaty obligations under Article 31 of the Refugee Convention. Article 31 provides that states “shall not impose penalties, on account of [the] illegal entry or presence, on refugees who . . . enter or are present . . . without authorization, provided they present themselves without delay to the authorities and show good cause for their illegal entry or presence.”97Convention Relating to the Status of Refugees art. 31(1), July 28, 1951, 189 U.N.T.S. 137 (emphasis added). Coupled with this added legislative protection, Congress should include a provision barring the separation of parents and children except where the best interest of the child requires it. Such changes would protect against any future effort to adopt a policy such as “zero-tolerance.”

Relatedly, Congress should clarify that public health-related expulsions cannot be myopically pursued at the expense of our international legal obligations to provide asylum to bona fide refugees presenting themselves at U.S. borders. While managing present and future public health crises is important, it is a false bifurcation that the U.S. can only pursue one of these objectives at a time. Alongside testing, social distancing, masking, vaccinations, and contact tracing, asylum and refugee processing can coexist.98  Letter to CDC Director, supra note 35. There may be reasonable and necessary processing delays at the border and in refugee camps that occur during a pandemic, but Congress should foreclose the use of any public health crises as a pretext to freeze our humanitarian obligations.

To strengthen the refugee resettlement system in the U.S., Congress should increase funding afforded to domestic refugee resettlement organizations, both to offset the dramatic cuts experienced during the Trump era and to provide more sustainable funding levels to ensure resettlement organizations are able to execute adroitly their mission of welcoming refugees into the U.S. As these organizations are scrambling to meet the needs of Afghan evacuees, more resources are needed, particularly to assist with the process of seeking asylum.99Marco Poggio, 83,000 Afghans Made It to the US. Now They Need Lawyers, Law360 (Feb. 6, 2022, 8:02 PM), https://www.law360.com/articles/1462197/83-000-afghans-made-it-to-the-us-now-they-need-lawyers [perma.cc/3MCV-GZ99].

Likewise, we contend that Congress should adopt the familiar but-for causation test (without the second part of Attorney General Rosen’s test) as a safe harbor provision that is sufficient to establish that the persecution occurred “on account of” the applicant’s protected characteristic. That is, where an applicant can show that but for the protected characteristic, the persecution would not have occurred (or would have been much less likely to occur), the nexus requirement will have been met.100Gupta, New Nexus, supra note 76, at 383. In the domestic violence context, for example, the “but-for” approach would provide asylum applicants with a clear benchmark for demonstrating nexus, as applicants would be able to show that but for their membership in a particular social group (defined by gender and other characteristics), the abuse would likely not have occurred. The but-for causation model recognizes that there may be many causes for abuse, each of which is necessary for the abuse to occur, but the existence of multiple necessary factors does not negate the fact that any one of those factors is an actual cause of the abuse.101  Id. In asylum law, such a model would recognize that while the persecutor’s “despicable” or “criminal” nature may have been one reason for the abuse, the applicant’s protected status was not only another reason for the abuse, but was a necessary reason. This approach, which shifts the focus from the intent of the persecutor to the status of the applicant, would also more closely conform to the BIA’s recognition that individuals should be protected from persecution that occurs on account of characteristics they are unable to change or should not be required to change.102  Id. at 390 n.45. Further, such a modification would put an end to the BIA’s flawed means-to-an-end reasoning and realign U.S. law more closely with international standards.103See Gupta, New Nexus, supra note 76; see also Gupta, Nexus Redux, supra note 76 (providing an alternative basis for proving nexus in the small number of cases where it may not be possible to show but-for causation). This alternative burden-shifting framework begins with a prima facie showing that the protected ground played a role in bringing about the persecution, similar to the “contributing cause” standard proposed by other scholars. See, e.g., Michelle Foster, Causation in Context: Interpreting the Nexus Clause in the Refugee Convention, 23 Mich. J. Int’l L. 265 (2002); James C. Hathaway & Michelle Foster, The Causal Connection (“Nexus”) to a Convention Ground, 15 Int’l J. Refugee L. 461 (2003).

In regard to substantive asylum eligibility, the nonstate actor test should be clarified through notice-and-comment rulemaking, such that requisite state protection is measured by whether the applicant has suffered past persecution from a nonstate actor (and thus the state failed to protect) or whether the applicant possesses a well-founded fear of future harm that the state is either unable or unwilling to prevent. Once an applicant establishes past persecution at the hands of a nonstate actor, the state’s unwillingness or inability to protect should be presumed, and the burden should shift to the Department of Homeland Security to show that the state is, in fact, willing and able to stop the persecution. The reformed state protection test should also clarify that there is no penalty for refugees who reasonably elect not to seek state protection.104See Law Scholars Urge DOJ and DHS to Adopt Protective Regulations Related to Nonstate Actor Persecution, Am. Immigr. Laws. Ass’n (May 27, 2021), https://www.aila.org/infonet/law-scholars-urge-doj-and-dhs-to-adopt-protective [perma.cc/S8MX-EGSX]; Andrew Paul Janco, ‘Unwilling’: The One-Word Revolution in Refugee Status, 23 Contemp. Eur. Hist. 429 (2014).

With respect to particular social group analysis, the new “particularity” and “social distinction” requirements should be eliminated by regulation.105See Stephen Legomsky & Karen Musalo, Asylum and the Three Little Words That Can Spell Life or Death (May 28, 2021), https://www.justsecurity.org/76671/asylum-and-the-three-little-words-that-can-spell-life-or-death. [perma.cc/W6D7-K8NK]. These additional social group requirements, in the hands of xenophobic adjudicators, are too easily utilized to deny asylum to those with legitimate fear of harm due to characteristics they are powerless to change.106Id. Accordingly, the “immutability” test should be restored as the correct and sufficient test for the cognizability of particular social groups. Under the immutability test, groups based on gender, for example, would be explicitly recognized as cognizable social groups.

Finally, informed by the lessons learned from the Trump era—and the antecedents that made that era possible—one core feature of any lasting solution to shore up asylum protections must include curbing the current degree of explicit and implicit deference afforded to the executive branch to narrow access to asylum and refugee protection. Whatever merits may exist for the Chevron and Auer deference doctrines in general, recent experience has demonstrated that rather than unique agency expertise in administering this area of law, animus can become the driving force in adopting one interpretation over another. For this reason, scholars have called for an explicit end to the practice of courts granting Chevron and Auer deference to the immigration agency’s decisions interpreting statutes and regulations, particularly as it relates to asylum and withholding.107Shoba Sivaprasad Wadhia & Christopher J. Walker, The Case Against Chevron Deference in Immigration Adjudication, 70 Duke L.J. 1197 (2021); Maureen A. Sweeney, Enforcing/Protection: The Danger of Chevron in Refugee Act Cases, 71 Admin. L. Rev. 127 (2019). These forms of protection are anchored in our international treaty obligations, which constitute the supreme law of the land, and represent an area of law where courts are best able to handle questions of statutory interpretation.108Sweeney, supra note 107.

The flexibility Congress granted to the executive branch was never intended to be used to circumscribe access so significantly to asylum as to constitute its functional demise. Congress intended for U.S. law to embody the full protections encompassed within our international treaty obligations as provided for in the Refugee Convention and Protocol. Shortly after the passage of the Refugee Act, the Supreme Court examined its legislative history and explained that Congress “believed that apparent differences between the Protocol and existing statutory law could be reconciled by the Attorney General . . . and did not require any modification of statutory language.”109Immigr. & Naturalization Servs. v. Stevic, 467 U.S. 407, 417–18 (1984). The Court noted that “to the extent that domestic law was more generous than the Protocol, the Attorney General would not alter existing practice.”110  Id. at 429 n.22. Of critical importance here, the Court explained that “to the extent that the Protocol was more generous than the bare text of [the statute,] . . . the Attorney General would honor the requirements of the Protocol and hence there was no need for modifying the language of [the statute] itself.”111Id. In other words, the liberty Congress granted to the executive to administer the asylum system was intended to be used to remain faithful to our core obligations under international law and to maximize protection, not diminish them.

Should courts continue to afford some deference to the immigration agency’s asylum and refugee decisions, courts must recognize that deference is not absolute. Noting the “[r]epeated egregious failures of the Immigration Court and the Board to exercise care commensurate with the stakes in an asylum case,” even before the Trump Administration, some judges have observed that “[d]eference is earned; it is not a birthright.”112Kadia v. Gonzalez, 501 F.3d 817, 821, 823 (7th Cir. 2007) (“Deference is accorded to the factfindings of government agencies because they know more about the activities they regulate than the courts do.”). The aim of the Refugee Act was to codify the U.S.’s obligations to provide safety to bona fide refugees. Thus, we contend that courts should view the immigration agency’s adjudicatory interpretations (whether of statutes or regulations) that have the effect of restricting asylum with some skepticism and, at an absolute minimum, strictly adhere to the Supreme Court’s instructions cabining deference.113Kisor v. Wilkie, 139 S. Ct. 2400, 2415-18 (2019). The Supreme Court has explained that deference to an agency’s interpretation should only be given after a determination that the regulation is “genuinely ambiguous,” that “the character and context of the agency interpretation entitles it to controlling weight,” that the agency’s interpretation “in some way implicate[s] its substantive expertise,” and that its “reading of a rule . . . reflect[s] its ‘fair and considered judgment.’” Id. at 2406 (quoting Auer v. Robbins, 519 U.S. 452, 462 (1997)).


Though not a panacea, the legislative, regulatory, and judicial shifts in the administration of asylum and refugee law for which we advocate here would go a long way in preventing the sort of norm-defying abuses witnessed during the Trump era. At the heart of what made those abuses possible is an excess of discretionary authority to restrict access to asylum. Such legally sanctioned discretion casts a penumbra of judicial obsequiousness. And within that shadow, as recent history has shown, an administration can take license to pursue an agenda that defies the humanitarian ethos undergirding our refugee and asylum systems. In an executive authority that lacks even a modicum of commitment to international human rights, the U.S. asylum system can cease to function. While the Biden Administration may eventually make good on its promise to strengthen and restore the U.S. asylum system—and we sincerely hope it does—those changes would last only as long as the tenure of its ideological proponent. As such, a paradigm shift is needed. If we are to take down the administrative wall of ineligibility and durably reconstruct a more just, fair, and welcoming system for asylum seekers, let us do so on a firmer foundation. Our refugee and asylum systems should be grounded upon the normative judgment that refugees deserve protection, and that core obligation must not be subject to degradation in the name of deference.

* Shane Ellison is a Senior Lecturing Fellow and Supervising Attorney in the Immigrant Rights Clinic at Duke Law School.

** Anjum Gupta is Vice Dean, Professor of Law & Judge Chester J. Straub Scholar, and Director of the Immigrant Rights Clinic at Rutgers Law School.