Midnight Agency Adjudication: Attorney General Review of Board of Immigration Appeals Decisions

In Advancing Executive Branch Immigration Policy Through the Attorney General’s Review Authority, former Attorney General Alberto Gonzales and Patrick Glen provide a rich history and detailed study of the power of the Attorney General to refer and review decisions of the Board of Immigration Appeals (“BIA”). Their account is both descriptive and normative. After a review of contemporary decisions where the Attorney General has exercised review authority, Gonzales and Glen conclude that “[i]n many cases, the Attorney General was able, through referral and review, to provide a clear, cogent, and definitive legal or policy prescription for immigration officials on the issue resolved.”1 They also argue that Attorney General review should happen more often, contrasting “the robust exercise of [review] authority in the George W. Bush Administration to the near-absence of usage in the Obama Administration.”2 Finally, the authors dismiss objections to the fairness and transparency of Attorney General review. Gonzales and Glen conclude that procedural due process objections, which have been voiced most prominently regarding Attorney General Mukasey’s decision in Matter of Silva-Trevino,3 are ill-founded, and that reform proposals to provide notice and an opportunity to be heard prior to Attorney General review “are premised mostly on superficial gains in the optics of referral” and not worth pursuing.4

Attorney General review of BIA decisions is controversial for a number of reasons. There is some truth to the adage that “[w]here you stand depends on where you sit.”5 So it is not surprising that a former Attorney General and a current Department of Justice (“DOJ”) attorney seek “to advance the referral authority as an important arrow in the executive branch’s policy quiver,”6 while asserting that criticism of its use simply reflects the institutional biases of the critics.7 Certainly the question of who wins and who loses, or as Gonzales and Glen assert, “whether certain ideological inclinations are being well served,”8 comes into play whether one is extolling or criticizing Attorney General review authority. That does not, however, negate other legitimate concerns about the exercise of this power.

Attorney General review might also be seen as objectionable because it conflicts with a core value of our legal system: that disputes are resolved by an impartial adjudicator who has no interest in the outcome. Adjudication within executive branch agencies, however, has long been a controversial exception to this model. In a number of administrative contexts including removal proceedings, adjudicators who decide contested cases “are employees of the very agency whose caseload they adjudicate. . . . [And thus] potentially subject to the supervision and control of one of the interested parties.”9 And it is typical that their decisions can be referred for review by the agency head.10 As Christina Boyd and Amanda Driscoll explain, “adjudication’s potential to be an effective policymaking tool depends largely on the institutionalized oversight role held by agency leaders over the [administrative judges’] rulings, since the agency head (or his direct delegee), serves as the agency’s final appellate judge.”11 There are, of course, long-standing concerns about this oversight of agency adjudication, but it is not anomalous—as Gonzales and Glen establish, the BIA has always exercised adjudicatory authority that is delegated from the Attorney General and subject to Attorney General review.

My Essay will explore one key aspect of contemporary Attorney General review authority that is not examined in the otherwise comprehensive account provided by Gonzales and Glen: timing. Both Matter of Compean and Matter of Silva-Trevino, which were issued by Attorney General Mukasey and later vacated by Attorney General Holder, were decided after President Obama was elected and before he was inaugurated. And Matter of R-A- was referred by Attorney General Ashcroft to himself just as the functions of the former Immigration and Naturalization Service (“INS”), an agency within the Department of Justice, were transferred to the newly-created Department of Homeland Security (“DHS”). These are what administrative law scholars might term “midnight agency adjudications,” although the phrase has not been used because this practice has never been examined. Indeed, the rich literature on the topic of regulations promulgated in times of presidential transitions has not yet considered the adjudication analog of an agency head who refers a controversial issue to himself and renders a decision upending agency precedent on his way out the door.12

From an administrative law perspective, this is an important phenomenon. In this Essay, I highlight and contextualize examples of midnight agency adjudication by the Attorney General and identify some concerns that arise from this practice, setting the stage for potential further exploration of the issue in immigration law or any other administrative context. In addition, I contend that timing is a central part of the story of Attorney General review of BIA decisions. The failure of Gonzales and Glen to acknowledge the prevalence of midnight agency adjudication in their otherwise comprehensive account undermines their conclusions about the efficacy and procedural fairness of the practice.

I.     Midnight Adjudication By the Attorney General

“Not all administrative work done during the waning days of a presidential administration is objectionable as a form of midnight mischief.”13 Indeed, the Administrative Conference of the United States (“ACUS”) began its recommendation on “Midnight Rules” by observing that while a “spurt in late-term regulatory activity has been criticized by politicians, academics, and the media during the last several presidential transitions. . . . [T]he perception of midnight rulemaking as an unseemly practice is worse than the reality.”14 In this Essay, I do not argue that midnight adjudication is always improper or necessarily “unseemly.” Instead, I assert that the same concerns voiced about midnight rules arise when an agency head in the incumbent administration refers a case to himself to decide on the eve of a presidential transition, particularly when the decision seems rushed and the outcome upends settled agency precedent and is likely to conflict with the policy preferences of the incoming administration.

In addition, two key features distinguish midnight agency adjudication from midnight rulemaking—the comparative ease with which a decision can be reversed, and the absence of procedures that require notice and an opportunity to provide input for those who will be impacted by the agency decision. These features create remarkable flexibility for agency policy formulation via referral and review of agency adjudication, but they also raise concerns when an agency head aims to use midnight adjudication to entrench a policy choice in a time of transition.

A.     The Midnight Rulemaking Analogy

Scholarly literature and policy studies of midnight rulemaking are voluminous; in this Essay I will only briefly recount common observations about the practice. First, there is a well-documented phenomenon of an increase in the volume of regulatory activity in the three months preceding a presidential administration.15 To some degree, this can be explained by the human tendency to procrastinate and the typical bureaucratic obstacles that impede all agency rulemaking,16 which together may produce a rush to “get as much done as possible at the end of the term.”17 But midnight rulemaking can also reflect what Professor Nina Mendelson has termed “agency burrowing”: an attempt to entrench a particular policy choice in anticipation of a presidential transition.18 Midnight rulemaking is criticized for its potential to decrease political accountability, truncate opportunities for public participation, produce regulations of lesser quality, and either tie the hands of the incoming administration or force them to expend significant time and political capital evaluating, and possibly undoing, midnight rules.19 At its worst, it reflects a deliberate effort by an outgoing administration to “hurry through last-minute rulemakings” in order to enact policy they know will be an anathema to the incoming administration.20

To address these concerns, incoming presidential administrations routinely issue a memorandum on the first day of office imposing a moratorium on issuing new regulations until they have been approved by an agency head appointed by the new President, and seeking to withdraw regulations that have been published but have not yet taken effect.21 In addition, ACUS and various commenters recommend that incumbent administrations should do the following: manage the rulemaking process to avoid “an actual or perceived rush of the final stages of the process;”22 explain the timing when a rule is issued by an outgoing administration after a presidential election; be hesitant to issue a midnight rule that is significant or especially controversial without providing the incoming administration an opportunity to weigh in; and use transparent procedures to safeguard the opportunity for public comment on midnight rules.23

Existing studies of midnight rules recognize that agencies have other policymaking tools at their disposal that can be used in the waning months of a presidential administration.24 Nevertheless, midnight adjudication by an agency head has received scant scholarly attention.25 Professor Jack Beermann, who served as the consultant to the ACUS study of midnight rules, observed that there is “noticeable increase” in “non-notice and comment [activity] . . . during the midnight period,” but concludes these actions are not as troubling as midnight rules.26 A regulation promulgated via notice-and-comment rulemaking can be revised or rescinded only by promulgating a new regulation using the same procedures.27 In contrast, a policy announced in an agency memorandum or by agency adjudication can be withdrawn with relative alacrity in a new presidential administration. For that reason, Professor Beermann concludes that these other agency actions are “less problematic [than midnight rules], because the incoming administration can revoke or alter them without notice and comment, and less likely to be done, because given easy revision, it may not be worth the effort to issue them at the end of the term.”28

Professor Mendelson does consider other forms of agency policy burrowing, most notably the phenomenon of “personnel entrenchment” where political appointees, “whose jobs within an executive branch agency typically would be terminated with the departure of the President, can be moved to a civil service position with tenure in the same agency.”29 She recognizes that “agencies may ‘entrench’ policies by resolving individual cases,” and posits that a policy that can be reversed by a new administration without notice or public input nevertheless may be “close to irreversible” as a practical matter.30 Professor Mendelson also concludes that other forms of policy entrenchment may be more problematic than midnight rules because they lack the procedural discipline, transparency, and political accountability of rulemaking procedures, and thus “may present a greater risk of abuse.”31

B.     A Tale of Two Decisions

In the last three months of the Bush Administration, Attorney General Mukasey issued two decisions setting aside BIA rulings via a process that was closely analogous to midnight rulemaking. Both Matter of Compean and Matter of Silva-Trevino were decided by the Attorney General after President Obama was elected and before he was inaugurated. Notably, in anticipation of the presidential transition the Bush Administration had issued a memorandum that counseled against promulgating midnight rules, announcing that absent “extraordinary” circumstances executive branch agencies should finalize any new regulations prior to the presidential election in November 2008.32 The policy did not address last-minute agency adjudications, however. Compean and Silva-Trevino each overturned settled BIA precedent and were classic illustrations of the most “troubling” feature of midnight agency action:33an attempt to enshrine the policy preferences of the outgoing administration “given the known or possibly different policy preferences of the incoming administration.”34 Indeed, both decisions were ultimately vacated by Attorney General Holder: Matter of Compean after just four months; Matter of Silva-Trevino after six years.35

Despite these similarities, the Attorney General used markedly different procedures when he referred these decisions to himself to decide. The legal issue at stake in Compean was highly visible, and the referral and review process was hurried but still transparent. Compean expressly considered the continued validity of Matter of Assaad, a 2003 opinion where the BIA, in a rare en banc decision, upheld its longstanding precedent establishing procedural requirements for motions to reopen removal proceedings based on ineffective assistance of counsel.36 The government had argued in Assaad that intervening Supreme Court precedent required the BIA to reconsider the constitutional basis for that decision.37 The BIA accepted amicus curiae briefs from advocacy organizations arguing to the contrary, and concluded that its existing precedent remained valid.38

Matter of Compean was one of three BIA cases decided in the spring of 2008 where the BIA applied the precedent it had reaffirmed in Assaad, but concluded that the respondents did not meet the procedural requirements to pursue a motion to reopen based on ineffective assistance of counsel.39 In August 2008, Attorney General Mukasey certified all three decisions to himself for review, and in a public notice stated that he would consider anew whether there was a constitutional right to effective assistance of counsel in immigration proceedings and what standards should apply to adjudicate claims that counsel’s performance was deficient.40 The notice requested briefs from the parties and amici curiae to address a list of questions that the Attorney General would consider, and required briefs to be filed in just over a month.41 As was true when the BIA decided Assaad, advocacy organizations filed briefs arguing that the Attorney General should affirm a constitutional basis to reopen removal proceedings based on ineffective assistance of counsel claims and retain the standards established by the Board to adjudicate these claims.42

On January 7, 2009—less than two weeks before President Obama’s inauguration—Attorney General Mukasey issued a decision in Matter of Compean overturning the BIA’s precedent on ineffective assistance of counsel.43 The Attorney General concluded “that the Constitution does not confer a constitutional right to effective assistance of counsel in removal proceedings,”44 but that the BIA and immigration courts could, as a matter of administrative discretion, reopen removal proceedings based on egregious deficiencies of counsel, with exacting procedural requirements.45

Although Attorney General Mukasey’s decision in Matter of Compean was not reported widely in the media,46 it generated significant controversy among legal cognoscenti focused on immigration law during the presidential transition. Incoming Attorney General Eric Holder was specifically asked about his views on Compean in written questions propounded by Senators during the confirmation process; he replied “I intend to reexamine the decision.”47 Three days after he took office, a number of immigration advocacy organizations, law firms, and attorneys submitted a letter supporting respondents’ motion to reexamine Attorney General Mukasey’s decision “in the spirit of the White House Memorandum . . . direct[ing] that all new regulations issued by the departing administration be submitted for review and approval by new agency heads.”48 The letter objected that the “process by which Attorney General Mukasey reviewed these cases and issued Compean was too sudden and hurried,”49 and asserted that the decision, which departed dramatically from longstanding agency precedent, had undermined the fairness of removal proceedings and created confusion.50

On June 3, 2009, Attorney General Holder issued a new decision in Compean, vacating Attorney General Mukasey’s decision in its entirety.51 Attorney General Holder concluded, “I do not believe that the process used in Compean resulted in a thorough consideration of the issues involved, particularly for a decision that implemented a new, complex framework in place of a well-established and longstanding practice that had been reaffirmed by the Board in 2003 after careful consideration.”52 Announcing that “[t]he preferable administrative process . . . is one that affords all interested parties a full and fair opportunity to participate and ensures that the relevant facts and analysis are collected and evaluated,” Attorney General Holder ordered the Executive Office for Immigration Review to initiate rulemaking procedures to address the proper standards and procedural framework for reopening cases based on claimed ineffective assistance of counsel.53 He also directed that the BIA and immigration judges should apply the standards delineated in pre-Compean precedent to all pending and future cases until such rules are issued.54 Almost six years later, proposed rules have not been published, although the matter is still pending on the DOJ’s regulatory agenda.55

Criticisms of Attorney General Mukasey’s late-term decision in Matter of Compean echo common criticisms of midnight rules: that the process was rushed, truncating the opportunity for public input in an effort to enshrine the policy views of an outgoing presidential administration. The fact that the decision was vacated by the incoming Attorney General in just over four months supports Professor Beermann’s view that because midnight adjudications by agency heads can be revoked without notice and comment procedures, “it may not be worth the effort to issue them at the end of the term.”56 But Compean also arguably illustrates what Professor Mendelson views as a “positive consequence” of late-term policy entrenchment—that it can “raise public awareness regarding an issue . . . or create a sharpened dialogue on a policy question.”57 Although the general public was not engaged in the debate, Compean did prompt brief attention from Congress in the context of a confirmation hearing, and sustained attention from the DOJ on an issue of significant import to the fairness and efficiency of immigration court proceedings.58 Unfortunately, since regulations have not yet been promulgated we are back at square one on the policy issue (a situation that, as it turns out, is not unusual when the Attorney General intervenes to review a BIA decision—a point I develop later on). Nevertheless, consistent with Professor Mendelson’s view, one prominent immigration attorney, writing after Attorney General Holder vacated the original decision, opined that “[i]n the end, former Attorney General Mukasey appears to have done us a favor. By provoking a reevaluation . . . he (perhaps inadvertently) brought us to the point of notice-and-comment rulemaking on the ineffective assistance of counsel issue by a decidedly less antagonistic Department of Justice.”59

This so-called “silver lining” of policy entrenchment in the midst of a presidential transition—that it might “prompt heightened public discussion and interbranch dialogue, and, in doing so, encourage a more deliberative and more democratic decision-making process”60—cannot happen when referral and review by an agency head happens in secret. Indeed, Professor Mendelson acknowledges that any positive consequences of midnight rules flow from the fact that agency rulemaking is “a comparatively disciplined and transparent process for setting policy.”61 Without procedures that provide notice and an opportunity to be heard, there is a greater risk that late-term policymaking can be employed to undercut political accountability, foreclose opportunities for public input, and force an incoming presidential administration to accept or expend considerable resources and political capital undoing new interpretations of law or policy enshrined in precedent decisions issued by the outgoing agency head. Matter of Silva-Trevino, decided by Attorney General Mukasey on November 7, 2008, illustrates such a scenario.

Silva-Trevino is startling for its “processless Attorney General review”62—a point that has been developed in law review literature,63 was noted by the Third Circuit Court of Appeals,64 and was argued by amici curiae in support of a motion for the Attorney General to reconsider his Silva-Trevino decision.65 I will not repeat those criticisms here, but rather will highlight the differences between the hurried-but-transparent process in Compean and the behind-the-scenes referral of Silva-Trevino.

In contrast to the highly visible legal issue in Compean, which previously had been considered by the BIA in an en banc decision, the outcome in Silva-Trevino turned on the application of well-settled precedent that had not been challenged before the BIA.66 At his removal hearing, Mr. Silva-Trevino was denied the opportunity to petition for relief from deportation based on the immigration judge’s conclusion that his conviction was a “crime involving moral turpitude” under immigration law.67 When he appealed to the BIA, the government filed only a three-paragraph memorandum in opposition.68 The BIA sustained Mr. Silva-Trevino’s appeal in the summer of 2006 in an unpublished opinion, one of over 30,000 decisions issued by the Board via a letter mailed to the respondent that do not operate as precedent.69 The case was remanded to the immigration judge to permit Mr. Silva-Trevino to apply for relief, and it languished there for a year until Attorney General Gonzales, without notice, referred the case to himself for decision.70

The Attorney General’s referral in Silva-Trevino was closed to public input and shielded from public view.71 Mr. Silva-Trevino’s counsel was not notified until a month later, when he received a letter dated August 8, 2007, informing him that the BIA decision was no longer final and the case was now pending before the Attorney General.72 Attorney General Gonzales, who had instigated the referral, resigned three weeks later and the matter was left pending in the DOJ for another year.73 Letters from Mr. Silva-Trevino’s counsel requesting information regarding the reason for the referral and the procedures for filing a brief before the Attorney General went unanswered.74

On November 7, 2008—three days after President Obama was elected—Attorney General Mukasey issued a decision which, like Compean, reversed longstanding BIA precedent.75 The decision was not made public until November 19, 2008.76Silva-Trevino was a blockbuster decision that “disregard[ed] a century of jurisprudence”77 in favor of a very government-friendly analysis to ascertain whether a respondent had been convicted of a crime involving moral turpitude.78 Attorney General Mukasey justified his intervention by stating that even though each of the federal courts of appeals applied the same legal test to resolve this issue the law was in disarray, necessitating a “new, standardized approach—one that accords with the statutory text, is administratively workable, and furthers the policy goals underlying the Act.”79 The most notable feature of the Silva-Trevino approach was a third step in the analysis that, in direct conflict with the rationale of existing precedent, permitted an adjudicator to look beyond the language of the statute of conviction and, in certain cases, the narrow record of conviction to consider “any additional evidence that the adjudicator determines is necessary or appropriate to resolve accurately the moral turpitude question.”80 Thus, upon referring to himself, without notice, an unpublished BIA decision that applied an uncontested framework for analysis, the Attorney General “completely rewrote” the approach used by the BIA and federal courts of appeals to assess whether a conviction constitutes a crime involving moral turpitude.81

After Attorney General Mukasey issued his surprise decision, advocates filed a motion to reconsider on December 8, 2008, supported by an amicus brief.82 Consistent with the Bush Administration memorandum that counseled against promulgating midnight rules,83 the Attorney General could have declined to rule on the motion to reconsider, leaving the issue to be resolved by the incoming Attorney General. Instead, Attorney General Mukasey issued a one-paragraph order on January 15, 2009—two business days before he left office—stating:

Having reviewed the motion and supporting materials, including briefs submitted by various nonprofit organizations as amici curiae, I find no basis for reconsideration of the decision. Among other things, this matter was properly certified and decided in accordance with settled Department of Justice procedures, and there is no entitlement to briefing when a matter is certified for Attorney General review.84

As Gonzales and Glen explain, the DOJ then fought a losing battle in the federal courts to defend the new approach announced in Attorney General Mukasey’s Silva-Trevino decision.85 While two federal circuit courts deferred to the Attorney General’s interpretation, five federal circuits courts concluded that the Attorney General’s Silva-Trevino decision was inconsistent with the unambiguous language of the statute.86 For example, the Fifth Circuit, upon petition for review after the BIA applied this new analysis in Mr. Silva-Trevino’s own case, declined to defer to the Attorney General’s interpretation, adhering instead to its long-standing precedent that Congress did not intend “immigration judges to consider relevant extrinsic evidence in order to classify a conviction as a crime of moral turpitude.”87 The court expressly noted that it was, “if anything, a little ironic” that the Attorney General claimed deference for his interpretation by citing his duty to “ensur[e] uniform application of the law,” when in fact Silva-Trevino undermined the “broad consensus” that existed on the legal question before the Attorney General’s decision.88 The Fifth Circuit did not reach the argument that the referral and review of Mr. Silva-Trevino’s cases violated procedural due process.89

In April 2015—almost six and a half years after Attorney General Mukasey decided Silva-Trevino—Attorney General Holder vacated the decision and remanded the case to the BIA to decide anew.90 Attorney General Holder concluded that because five federal circuit courts had rejected its analysis, Silva-Trevino had not achieved its stated goal of “establishing a uniform framework for ensuring that the [Immigration and Nationality] Act’s moral turpitude provisions are fairly and accurately applied.”91 In addition, the Attorney General noted that intervening Supreme Court precedent “cast doubt on the continued validity of the third step of the framework set out by Attorney General Mukasey’s opinion, which directs immigration judges and the Board to . . . inquire into facts outside the formal record of conviction.”92 Attorney General Holder directed that “[t]he Board should solicit and consider briefs from the parties and interested amici as it deems appropriate to ensure that its conclusions on these issues are reached after full and fair consideration of all relevant arguments.”93 The case is still pending before the BIA.

C.     Additional Examples

As it turns out, other notable Attorney General decisions were issued in times of transition. It is beyond the scope of this Essay to analyze each one, but I will briefly note some additional examples. Matter of R-A- is well-known for its unusual procedural history: three different Attorneys General issued referred decisions that declined to decide the underlying question of whether and under what legal framework domestic violence can form the basis of an asylum claim.94 Two of these referrals can be characterized as midnight agency adjudications.

In the original R-A- decision, the BIA ruled against the respondent in June 1999, reversing a grant of asylum by an immigration judge to a Guatemalan woman who had been severely abused by her husband.95 In December 2000, the DOJ published a proposed rule to address when persecution based on “membership in a particular social group” would form the basis for asylum.96 The proposed rule was specifically intended to address the “novel” issue of “the extent to which victims of domestic violence may be considered to have been persecuted under the asylum laws.”97 The Acting Commissioner of the INS then referred the R-A- decision to “the Attorney General on January 8, 2001,” asking her “to vacate the decision ‘immediately’ and remand it to the Board” pending promulgation of final regulations.98 Attorney General Reno complied on the eve of the inauguration of President George W. Bush, issuing a two-paragraph decision “direct[ing] the Board to stay reconsideration of the [R-A-] decision until the proposed rule[s] [are] published in final form.”99

Attorney General Ashcroft was apparently motivated by a different type of transition—the imminent abolition of the former INS, which was part of the Department of Justice, and creation of the DHS—when he “re-certified [R-A-] to himself” in February 2003.100 This action sparked concern that the Attorney General intended to propose new regulations to supplant those still pending before the DOJ or to reinstate the BIA decision denying asylum to R‑A-,101 in “a last ditch effort” to retain DOJ control over the issue before the DHS took over the functions of the INS on March 1.102

As would happen later on with Attorney General Gonzales’ referral of Silva-Trevino, Attorney General Ashcroft initially referred the R-A- decision to himself without notice or an opportunity to be heard.103 Almost a month passed before the respondent’s attorneys were informed that the case was again pending before the Attorney General.104 And despite substantial development in the case law since the BIA first decided R-A- six years earlier, Ashcroft originally refused a request to “allow the parties to re-brief the case.”105

In November 2003, 62 members of Congress sent a letter to the Attorney General urging him to accept additional briefs from both parties in light of “a wealth of significant relevant legal developments,” which apparently prompted him to reverse course.106 And then in a surprising move, the DHS filed a brief retracting the government’s prior opposition to a grant of asylum for R-A-.107 In its brief before the Attorney General, DHS argued that “[a] final rule is the best vehicle for providing much needed guidance on the adjudication”108 of domestic violence claims, “rather than by issuing a precedent decision analyzing the facts of a specific case.”109 DHS recommended that the Attorney General postpone any decision pending publication of the final social group regulations or in the alternative to order affirmance of the immigration judge without issuing an opinion, so as not to supplant the rulemaking process.110 On January 19, 2005—one day prior to President Bush’s second inauguration—Attorney General Ashcroft again remanded to the BIA, reinstating the order that the BIA should reconsider the issue following final publication of the proposed social group rules.111 Final rules have yet to be published, but after yet another referral by Attorney General Mukasey and remand to the BIA, this time with instructions to rule without waiting for the regulations to become final,112 R-A- received asylum in 2009.113

R-A- was one of six immigration decisions that Attorney General Ashcroft decided in the three-month period between the date he announced his resignation shortly after President Bush was re-elected and the date his successor, Alberto Gonzales, was confirmed. These cases had been pending before the Attorney General for five to nine years; most had been referred during the Clinton Administration.114 Gonzales and Glen identify these cases as notable and analyze them substantively, but do not discuss the significant delays between referral and decision or the impetus for this docket-clearing effort before Attorney General Ashcroft left office.115 But these cases too can be described as decisions in a time of transition, as the timing indicates they were issued in anticipation of the incumbent Attorney General’s departure.

II.     Acknowledging and Assessing Midnight Adjudication

In their Article, Gonzales and Glen note with approval that “Attorneys General during the George W. Bush administration used the [referral] authority with significantly more frequency than any administration since that of John Kennedy.”116 Just over half of these decisions were decided in the midst of important transitions, as detailed above, but Gonzales and Glen do not incorporate this timing into their description or analysis. They note in passing that criticism of Attorney General review authority has focused on decisions issued “in the waning days of the Bush Administration,”117 but they do not acknowledge that policy entrenchment akin to midnight rulemaking was the impetus for the Compean and Silva-Trevino decisions. Nor do they mention the notable timing of the Reno and Ashcroft referrals in R-A-, or the spate of decisions in long-pending cases issued after Ashcroft had announced his resignation.

This oversight or omission is puzzling. Contemporaneous observers were well-aware that the timing of these decisions was not coincidental or inconsequential.118 Moreover, the midnight adjudication aspect of Attorney General review is central to any effort to assess its efficacy. Gonzales and Glen assert that “[i]n many cases, the Attorney General was able, through referral and review, to provide a clear, cogent, and definitive legal or policy prescription for immigration officials on the issue resolved.”119 In reality, several of the most significant Attorney General decisions of the Bush administration—Compean, Silva-Trevino, and R-A- -—ultimately did not resolve the underlying legal or policy issue, as detailed above. In each of these important cases, Attorney General referral and review might be described as a “duck, bob, and weave” route to policy development. These cases suggest that midnight agency adjudication can result in chaotic development of law and policy across presidential administrations, and that, as Professor Beermann predicted, efforts to entrench the preferences of an outgoing administration via agency adjudication may not be durable.

In addition, as studies of midnight rulemaking establish, transparency is especially important when an agency attempts to set policy or establish new decisional frameworks on the eve of a presidential transition.120 When the Silva-Trevino case was referred to the Attorney General, however, no one outside of the Justice Department—including Mr. Silva-Trevino and his attorney—was informed of what issues were under consideration or given an opportunity to present arguments to the Attorney General. Instead, Attorney General Mukasey was advised only by attorneys within the Department of Justice before issuing his surprise decision.121

In their Article, Gonzales and Glen respond to critics who contend that this process was deficient.122 Professor Mary Holper has argued that the secretive process employed by the Attorney General in Silva-Trevino excluded important input from a number of important constituencies123 and “should command less deference from the courts because of the due process violations to the individual and the lack of deliberation by the agency.”124 Laura Trice has developed the procedural due process argument in a comprehensive, and I think convincing, analysis.125 Procedural due process applies “to all persons within the United States, including aliens, whether their presence here is lawful, unlawful, temporary, or permanent.”126 Trice notes that “the right [of noncitizens facing deportation] to meaningful notice and opportunity to be heard in appeals before the BIA is generally quite robust.”127 She concedes that there is no precedent on whether Attorney General review demands the same procedural rights as an appeal to the BIA,128 and then employs the Mathews v. Eldridge calculus to argue that “preventing meaningful participation by the parties upon certification [to the Attorney General] raises serious due process concerns.”129 Gonzales and Glen respond with their own detailed explication of the Mathews v. Eldridge factors, focusing their analysis on arguments that the Attorney General has a strong interest in maintaining flexible procedures and that any risk of erroneous deprivation is obviated by the availability of judicial review.130

I am not convinced that the Mathews v. Eldridge calculus is the proper framework to assess whether referral to the Attorney General without notice and an opportunity to be heard violates procedural due process.131 Fundamental fairness is the touchtone of procedural due process,132 and the central meaning of this protection is that a party to an adjudicatory proceeding who may be deprived of a protected interest has a constitutional right to notice and a meaningful opportunity to be heard.133 Moreover, “when the absence of procedural due process is egregious—when, for example, the government affords no notice . . . and the claimant’s interest is momentous,” a court can find a procedural due process violation without weighing the Mathews v. Eldridge factors or comparing the procedures afforded in other contexts.134 In appellate proceedings focused on legal interpretations and policy issues, an “adequate opportunity for argument” is provided by filing briefs.135 Thus, the argument that a party facing deportation “does not necessarily have a role to play”136 and can be precluded from offering written argument when the Attorney General reviews and then overturns a BIA decision in his favor, even as the Attorney General receives counsel from the DOJ, is so at odds with procedural due process jurisprudence that a reviewing court may conclude without further balancing that referral and review under these circumstances is fundamentally unfair.

Gonzales and Glen reject these procedural due process arguments without acknowledging what was really at stake when the Attorney General delayed notice to the respondent, declined to disclose what legal and policy issues were under review, and refused the permit the filing of briefs with the Attorney General in R-A- (when the case was initially pending before Attorney General Ashcroft) and Silva-Trevino. They contend the Attorney General must have flexibility to truncate procedures in this way because he may be too busy with “the myriad tasks and responsibilities that come with the position” and thus unable to “expend additional time and effort in the review of case materials.”137 In a similar vein, the authors assert that referral and review procedures should be ad hoc, with the option of closing-off input from the respondent and the public, because the Attorney General must “retain[] the maximum amount of flexibility to determine in specific cases how and to what extent he will be involved in the review.”138

As the authors explain, however, the Attorney General is advised by attorneys in several units of the DOJ when he decides cases certified from the BIA;139 certainly he can and does delegate the tasks of reading briefs and drafting opinions. Moreover, the legal issues at stake in Silva-Trevino and R-A- were not so time-sensitive that they required immediate resolution through truncated procedures that precluded the respondent and other interested persons from filing briefs, as evidenced by the fact that both cases had been pending before various components of the DOJ for years. Indeed, Compean illustrates that the Attorney General can act expeditiously while still providing notice and accepting briefs on the issues under consideration.140

These arguments for maximum flexibility to permit the Attorney General to review BIA decisions without notice and an opportunity to be heard are a smokescreen designed, I believe, to obscure the real reason that the Attorney General has sometimes employed “processless” referral and review. Unlike midnight rulemaking, midnight agency adjudication can happen largely behind the scenes. And sometimes last-minute policy entrenchment by an outgoing administration will occur only if it is implemented quietly, without notice as to the precise nature of the contemplated agency action, so as not to trigger backlash or negative publicity. In short, R-A- and Silva-Trevino echo the cautionary note sounded by Professor Mendelson: there is a greater risk of abuse when policy entrenchment happens outside the public eye, without procedural discipline or opportunity for input.

III.     Conclusion

The stalled rulemaking efforts after remand to the BIA in R-A- and Compean are emblematic of larger trends in administrative law. Promulgating regulations takes more time than ever before, involves a myriad of procedural steps, and can prompt highly visible public battles about the direction of agency policy.141 In contrast, agency adjudication can provide a quick route to a policy change via referral and review by an agency head, which until that moment when the decision is announced can develop without public input.142 “[W]e anticipate agency heads will use their adjudication oversight role to pursue agency priorities and protect the agency’s vested . . . interests” and will be “strategic in using this process to advance the political interests of [their agency] and pursue the goals of the current presidential administration.”143 Increasingly, as an expression of an agency preference for policy formulation without notice-and-comment procedures, we also might anticipate policy entrenchment in times of transition via midnight agency adjudication.

  1. [1]. Alberto R. Gonzales & Patrick Glen, Advancing Executive Branch Immigration Policy Through the Attorney General’s Review Authority, 101 Iowa L. Rev. 841, 897 (2016).

  2. [2]. Id. at 895.

  3. [3]. See Mary Holper, The New Moral Turpitude Test: Failing Chevron Step Zero, 76 Brook. L. Rev. 1241, 1291–92 (2011). See generally Laura S. Trice, Note, Adjudication by Fiat: The Need for Procedural Safeguards in Attorney General Review of Board of Immigration Appeals Decisions, 85 N.Y.U. L. Rev. 1766 (2010).

  4. [4]. Gonzales & Glen, supra note 1, at 908.

  5. [5]. See Rufus E. Miles, Jr., The Origin and Meaning of Miles’ Law, 38 Pub. Admin. Rev. 399, 399 (1978) (emphasis omitted).

  6. [6]. Gonzales & Glen, supra note 1, at 896.

  7. [7]. Gonzales and Glen dismiss as “empty” the “optics-based concerns” regarding notice and public input, stating that “[i]t is also revealing that these arguments are raised solely by academics and organizations whose business is representing the claims of the alien. Their objections are ultimately less to the lack of procedures than they are to the nature of the decision reached in certain cases.” Id. at 911–12.

  8. [8]. Id. at 912.

  9. [9]. Margaret H. Taylor, Refugee Roulette in an Administrative Law Context: The Déjà Vu of Decisional Disparities in Agency Adjudication, 60 Stan. L. Rev. 475, 480 (2007).

  10. [10]. Id. at 481.

  11. [11]. Christina L. Boyd & Amanda Driscoll, Adjudicatory Oversight and Judicial Decision Making in Executive Branch Agencies, 41 Am. Pol. Res. 569, 570 (2013).

  12. [12]. A search for the terms “midnight adjudication” or “midnight agency adjudication” in the Westlaw database of secondary sources yields only one instance where either phrase was used in a law review article. Holper, supra note 3, at 1291. Professor Holper used the phrase to describe the Silva-Trevino decision, discussed below in Part I.B. In contrast, a search for the terms “‘midnight rule’ or ‘midnight rulemaking’” in the Westlaw database of secondary sources yields 338 results as of May 28, 2016, including 77 sources in the database of law reviews and journals.

  13. [13]. William L. Andreen, Commentary on Chapter 2: Is There a Silver Lining to Midnight Mischief?, in Transitions: Legal Change, Legal Meanings 84, 85 (Austin Sarat et al. eds., 2012).

  14. [14]. Administrative Conference Recommendation 2012-2: Midnight Rules, Admin. Conference of the U.S. 1 (June 4, 2012), [hereinafter ACUS Recommendation].

  15. [15]. See Jack M. Beermann, Midnight Rules: A Reform Agenda, 2 Mich. J. Envtl. & Admin. L. 285, 286, 290–99 (2013).

  16. [16]. Id. at 300–08 (describing the related phenomena of “hurrying,” “delay,” and “waiting” when agencies issue regulations at the end of a presidential term). In its recommendation on midnight rules, the Administrative Conference of the United States concluded that “the majority of the [midnight] rules appear to be the result of finishing tasks that were initiated before the Presidential transition period or the result of deadlines outside the agency’s control....” ACUS Recommendation, supra note 14, at 2.

  17. [17]. Beermann, supra note 15, at 300.

  18. [18]. Nina A. Mendelson, Agency Burrowing: Entrenching Policies and Personnel Before a New President Arrives, 78 N.Y.U. L. Rev. 557, 560–61 (2003).

  19. [19]. See Beermann, supra note 15, at 312–16; see also Maeve P. Carey, Cong. Research Serv., R42612 Midnight Rulemaking 2 (2012).

  20. [20]. Andreen, supra note 13, at 87.

  21. [21]. See Beermann supra note 15, at 335–52; Carey, supra note 19, at 3–9; see, e.g., Memorandum from Victor S. Rezendes, Managing Dir., U.S. Gen. Acct. Off., to Hon. Henry A. Waxman, Comm. on Gov’t Reform, House of Representatives, & Hon. Joseph I. Lieberman, Chairman, Comm. on Governmental Affairs, U.S. Senate (Feb. 15, 2002),

  22. [22]. ACUS Recommendation, supra note 14, at 4.

  23. [23]. Id. at 4–5; Mendelson, supra note 18, at 652–57.

  24. [24]. See Beermann, supra note 15, at 315, 334–35, 349.

  25. [25]. See supra note 12.

  26. [26]. Beermann, supra note 15, at 292–94.

  27. [27]. Perez v. Mortgage Bankers Ass’n, 135 S. Ct. 1199, 1206 (2015) (the APA mandates that “agencies use the same procedures when they amend or repeal a rule as they used to issue the rule in the first instance”).

  28. [28]. Beermann, supra note 15, at 293–94.

  29. [29]. Mendelson, supra note 18, at 563.

  30. [30]. Id. at 596. For example, “[a]n agency may settle a litigation by signing a consent decree, issue a permit, issue funds under a grant program... or recognize an Indian tribe.” Id. (footnote omitted). These actions can be undertaken without notice-and-comment procedures, and as a practical matter bind a successor administration.

  31. [31]. Id. at 658.

  32. [32]. Carey, supra note 19, at 5. The Memorandum, from White House Chief of Staff Joshua B. Bolten, was issued to the heads of executive departments and agencies, and stated that except for “extraordinary circumstances” regulations should be proposed no later than June 1, 2008 and issued in final form before November 1, 2008. Id.

  33. [33]. Andreen, supra note 13, at 86 (“Last-minute rulemakings may also appear particularly political in nature. They may be hurried and as a result poorly considered. Even more troubling is the fact that many late-term rulemakings appear to involve policy decisions on which the incoming administration will likely disagree. Such actions will, of course, divert a new administration from pursuing its policy initiatives as it attempts to repair the damage that it perceives has just been done.” (footnote omitted)).

  34. [34]. Mendelson, supra note 18, at 599.

  35. [35]. Compean, 24 I. & N. Dec. 710 (Attorney Gen. 201o); Silva-Trevino, 26 I. & N. Dec. 550, 709 (Attorney Gen. 2015).

  36. [36]. Assaad, 23 I. & N. Dec. 553 (B.I.A. 2003) (en banc).

  37. [37]. Id. at 557.

  38. [38]. Id. at 557–58.

  39. [39]. Compean, 24 I. & N. Dec. at 714–15.

  40. [40]. AG to Review Certain BIA Decisions on Ineffective Assistance of Counsel Standards, 85 Interpreter Releases 2296, 2296 (2008).

  41. [41]. Id. In the Compean decision, the Attorney General stated that he gave notice of the referral on August 7, 2008. Compean, 24 I. & N. Dec. at 712. This notice was published in Interpreter Releases on August 25, 2008. AG to Review Certain BIA Decisions on Ineffective Assistance of Counsel Standards, supra note 40. Briefs from the parties and amici were to be submitted electronically and by mail, filed and postmarked by September 15, 2008. Id. at 2297.

  42. [42]. Compean, 24 I & N Dec. at 711 (listing as amici “Advocates for Human Rights; Massachusetts Law Reform Institute, and other organizations; National Immigrant Justice Center; American Immigration Law Foundation, and other organizations; Immigration Law Clinic at the University of Detroit Mercy School of Law; Immigrant and Refugee Appellate Center; and others”).

  43. [43]. See generally id.

  44. [44]. Id. at 714.

  45. [45]. Id. at 727–740.

  46. [46]. A search in the archives of the New York Times and the Washington Post, along with a Google search, yielded no major newspapers in the United States reporting on Attorney General Mukasey’s decision in Compean. The Guardian of the United Kingdom ran a fairly extensive story, however, reporting criticism from “[i]mmigration lawyers [who] describe [the] move by [the] attorney general as a last-minute evisceration of a constitutional right.” Daniel Nasaw, US Immigrants Facing Deportation Have ‘No Right to an Effective Attorney, The Guardian (Jan. 8, 2009, 4:50 p.m.), In an interview with the Guardian reporter, a Department of Justice spokesperson defended the decision, stating that “it looks like it was well thought-out before it was finally signed off and publicized [sic]” and “that if [President] Obama’s nominee as attorney general, Eric Holder, disagrees with the decision he may overturn it.” Id.

  47. [47]. Eric Holder Sworn in as Attorney General; Addresses Immigration-Related Questions, 86 Interpreter Releases 414, 414 (2009).

  48. [48]. Letter from Nadine Wettstein et al., Am. Immigration Law Found., to Eric Holder, Attorney Gen., U.S. Dep’t of Justice 2 (Feb. 6, 2009),

  49. [49]. Id.

  50. [50]. Id. at 3–6.

  51. [51]. Compean, 25 I. & N. Dec. 1 (Attorney Gen. 2009).

  52. [52]. Id. at 2.

  53. [53]. Id.

  54. [54]. Id. at 2–3.

  55. [55]. See Office of Regulatory Affairs, Office of Mgmt. & Budget, Motions to Reopen Removal, Deportation, or Exclusion Proceedings Based Upon a Claim of Ineffective Assistance of Counsel (2014),

  56. [56]. Beermann, supra note 15, at 294.

  57. [57]. Mendelson, supra note 18, at 620.

  58. [58]. See supra notes 41, 47.

  59. [59]. Thomas K. Ragland, The Right to Effective Assistance of Counsel in Immigration Proceedings, Am. Bar Ass’n: Section of Litig.: Immigration Litig. (last visited on June 19, 2016),

  60. [60]. Andreen, supra note 13, at 91

  61. [61]. Mendelson, supra note 18, at 657.

  62. [62]. Trice, supra note 3, at 1780.

  63. [63]. Id. at 1776–80; Holper, supra note 3, at 1252–54.

  64. [64]. Jean-Louis v. Attorney Gen., 582 F. 3d 462, 470 n.11 (3d Cir. 2009).

  65. [65]. Memorandum of Law of Amici Curiae American Immigration Lawyers Assoc. et al. in Support of Reconsideration, Silva-Trevino, 24 I. & N. Dec. 687 (Attorney Gen. 2008) [hereinafter Reconsideration Memo],

  66. [66]. Holper, supra note 3, at 1252; Trice, supra note 3, at 1790; Id. at 5.

  67. [67]. Trice, supra note 3, at 1776; see Silva-Trevino, 24 I. & N. Dec. 687, 691 (Attorney Gen. 2008) (explaining the legal framework).

  68. [68]. Trice, supra note 3, at 1778.

  69. [69]. In fiscal year 2006, the Board of Immigration Appeals decided 36,350 cases appealed from immigration judges. U.S. Dept. of Justice, Executive Office for Immigration Review, FY 2006 Statistical Yearbook S2 (2007),

  70. [70]. Holper, supra note 3, at 1252–53; Trice, supra note 3, at 1778.

  71. [71]. Holper, supra note 3, at 1275 (“No member of the public had any idea that the agency was considering a complete overhaul of the categorical approach before the decision was published.”).

  72. [72]. Reconsideration Memo, supra note 65, at 5 n.5. While the referral notice stated that the Attorney General referred the case to himself on July 7, 2007, there is some indication that the referral happened as early as April. Id.

  73. [73]. Steven Lee Myers & Philip Shenon, Embattled Attorney General Resigns, N.Y. Times (Aug. 27, 2007),

  74. [74]. Reconsideration Memo, supra note 65, at 6.

  75. [75]. See Silva-Trevino, 24 I. & N. Dec. 687 (Attorney Gen. 2008).

  76. [76]. Holper, supra note 3, at 1253; Reconsideration Memo, supra note 65, at 6.

  77. [77]. Silva-Trevino v. Holder, 742 F. 3d 197, 204 (5 Cir. 2014); Holper, supra note 3, at 1274 (“The Attorney General altered over a century of immigration law without input from members of the public or the affected party himself.”).

  78. [78]. As Gonzales and Glen describe it, “[t]he policy import of the Attorney General’s [Silva-Trevino] decision was clear: To permit a fuller examination of the circumstances of an alien’s criminal conviction would have the likely effect of permitting findings of crimes involving moral turpitude in cases where strict application of the categorical and modified categorical approaches would have proven inconclusive.” Gonzales & Glen, supra note 1, at 877.

  79. [79]. Silva-Trevino, 24 I. & N. Dec. at 688.

  80. [80]. Id. at 704.

  81. [81]. Trice, supra note 3, at 1779.

  82. [82]. See generally Reconsideration Memo, supra note 65.

  83. [83]. See supra note 32.

  84. [84]. Holper, supra note 3, at 1253.

  85. [85]. Gonzales & Glen, supra note 1, at 877.

  86. [86]. See Silva-Trevino v. Holder, 742 F. 3d 197, 200 n. 1 (5th Cir. 2014).

  87. [87]. Id. at 201.

  88. [88]. Id. at 205.

  89. [89]. Id.

  90. [90]. See generally Silva-Trevino, 26 I. & N. Dec. 550 (Attorney Gen. 2015).

  91. [91]. Id. at 552 (quoting Silva-Trevino, 24 I & N Dec. 687, 688 (Attorney Gen. 2008)).

  92. [92]. Id. at 553.

  93. [93]. Id. at 554.

  94. [94]. Gonzales & Glen, supra note 1, at 886–87.

  95. [95]. R-A-, 22 I. & N. Dec. 906, 194 (B.I.A. 1999)

  96. [96]. Asylum and Withholding Definitions, 65 Fed. Reg. 76589 (proposed Dec. 7, 2000) (to be codified at 8 C.F.R. pt. 208).

  97. [97]. Id.

  98. [98]. INS Seeks Review of Matter of R-A-, 78 Interpreter Releases 233 (2001).

  99. [99]. R-A-, 22 I. & N. Dec. 906, 906 (Attorney Gen. 2001); see also Attorney General Reno Vacates Matter of R-A-, 78 Interpreter Releases 256 (2001). Attorney General Reno also issued a decision in Matter of A-H- on the last day of her term. This unpublished decision held that BIA decisions that are referred to the Attorney General are not final, so that an order of the BIA cannot be executed (and need not be stayed) while the case is pending before the Attorney General. Apparently, no notice was given or briefs requested when the Attorney General decided this important procedural question. Her decision was not published until it was appended to the published decision of Attorney General Ashcroft in a later case, Matter of E-L-H-, which remanded to the Board for reconsideration, in light of Matter of A-H-, of whether a BIA decision continues to operate as precedent while a case is “pending review before the Attorney General.” E-L-H-, 23 I. & N. Dec. 700 (Attorney Gen. 2004). The Board concluded that a precedent decision of the BIA “applies to all proceedings involving the same issue unless and until it is modified or overruled by the Attorney General, the Board, Congress, or a Federal court.” E-L-H-, 23 I. & N. Dec. 814, 815 (B.I.A. 2005).

  100. [100]. See Refugee Women at Risk: Proposed Justice Department Regulations Would Eliminate Gender-Based Violence as a Basis for Obtaining Asylum in the U.S., Human Rights First: Asylum News 11 (Feb. 20, 2003),; Update: Refugee Women Still at Risk, Human Rights First: Asylum News 14 (May 7, 2003),

  101. [101]. See Refugee Women at Risk: Proposed Justice Department Regulations Would Eliminate Gender-Based Violence as a Basis for Obtaining Asylum in the U.S., supra note 100; Update: Refugee Women Still at Risk, supra note 100; see also Letter from Gregory W. Meeks et al., U.S. Congress, to John Ashcroft, Attorney Gen., Dep’t of Justice (Feb. 27, 2003), (expressing concern regarding “your possible issuance of new regulations that threaten women and girls seeking asylum in the United States”); Letter from Patrick Leahy, U.S. Senate to Hon. John Ashcroft, Attorney Gen., Dep’t of Justice (March 3, 2003), (urging the Attorney General not to reinstate the BIA’s decision in Matter of R-A-, and expressing support for the proposed social group regulations). The author is grateful to Human Rights First for restoring links to provide public access to these documents.

  102. [102]. Natalie Rodriguez, Give Us Your Weary but Not Your Battered: The Department of Homeland Security, Politics and Asylum for Victims of Domestic Violence, 18 Sw. J. Int’l L. 317, 329 (2011). The Department of Homeland Security was created by statute enacted in November, 2002 and “open[ed] its doors”—taking over the functions of the former INS—on March 1, 2003. Creation of the Department of Homeland Security, Homeland Security, (last updated September 24, 2015).

  103. [103]. Attorney General Ashcroft “re-certified [Matter of R-A-] to himself” on February 21, 2003; R-A-’s “attorneys were not notified until March 19, [2003]”. Update: Refugee Women Still at Risk, supra note 100.

  104. [104]. Id.

  105. [105]. Briefing Deadline Approaching on Leading Gender-based Asylum Case, Human Rights First: Asylum News 23 (Jan. 3, 2004),; see also Attorney General Remands Matter of R-A-, 82 Interpreter Releases 269 (Feb. 7, 2005).

  106. [106]. Letter from Lucille Roybal-Allard et al., U.S. Congress, to John Ashcroft, Attorney Gen., Dep’t of Justice (Nov. 4, 2003),

  107. [107]. Rachel L. Swarns, Ashcroft Weighs the Granting of Political Asylum to Abused Women, N.Y. Times (Mar. 11, 2004),; see DHS Advocates Asylum Grant in Matter of R-A-, 81 Interpreter Releases 245 (2004).

  108. [108]. Brief for Department of Homeland Security’s Position on Respondent’s Eligibility for Relief at 4, R-A-, 23 I. & N. Dec. 694 (Attorney Gen. 2004) (No. A-73-753-922),

  109. [109]. Id. at 3.

  110. [110]. Id. at 3.

  111. [111]. R-A-, 23 I. & N. Dec. 694 (Attorney Gen. 2005).

  112. [112]. R-A-, 24 I. & N. Dec. 629 (Attorney Gen. 2008).

  113. [113]. See Karen Musalo, A Short History of Gender Asylum in the United States: Resistance and Ambivalence May Very Slowly Be Inching Toward Recognition of Women’s Claims, 29 Refugee Surv. Q. 46, 59–60 (2010). It was not until 2014, 15 years after the original R-A- decision, that the Board issued a precedent decision granting asylum in a domestic violence claim. A-R-C-G-, 26 I. & N. Dec. 388 (B.I.A. 2014). A timeline and explanation of the history of domestic violence claims is available at the Center for Gender & Refugee Studies website,

  114. [114]. Mr. Ashcroft announced his resignation on November 9, 2004; his resignation took effect on February 5, 2005 when Mr. Gonzales was confirmed. John Ashcroft, Wikipedia, (last visited June 19, 2016). Six of the nine cases decided by Attorney General Ashcroft were decided during this period: A-H-, 23 I. &. N. Dec. 774 (Attorney Gen. 2005) (decided by the BIA in 2000 and referred by the INS to the Attorney General; decided by Attorney General Ashcroft on January 26, 2005); R-A-, 23 I. & N. 694 (Attorney Gen. 2005) (pending before the BIA after Attorney General Reno’s remand and referred by Attorney General Ashcroft to himself; remanded to the BIA on January 19, 2005); Marroquin-Garcia, 23 I. & N. Dec. 705 (Attorney Gen. 2005) (decided by the BIA in 1997 and referred by the Board to the Attorney General; decided by Attorney General Ashcroft on January 18, 2005); Luviano-Rodriguez, 23 I. & N. Dec. 718 (Attorney Gen. 2005)(decided by the BIA in 1996 and referred by the INS to the Attorney General; decided by Attorney General Ashcroft on January 18, 2005); C-Y-Z-, 23 I. & N. Dec. 693 (Attorney Gen. 2004) (decided by the BIA in 1997 and referred by the INS to the Attorney General; Attorney General Ashcroft denied the request to certify for review on December 1, 2004); E-L-H-, 23 I. & N. Dec. 700 (Attorney Gen. 2004) (decided by the BIA in 2000 and referred by the INS to the Attorney General; vacated and remanded by Attorney General Ashcroft on December 1, 2004).

  115. [115]. Gonzales & Glen, supra note 1, at 866–70, 886–90 (discussing Matter of A-H-, Matter of Luviano-Rodriguez, Matter of Marroquin-Garcia, Matter of R-A-, and Matter of E-L-H-).

  116. [116]. Gonzales & Glen, supra note 1, at 858.

  117. [117]. Id. at 847.

  118. [118]. See, e.g., Briefing Deadline Approaching on Leading Gender-based Asylum Case, supra note 105; Reconsideration Memo, supra note 65.

  119. [119]. Gonzales & Glen, supra note 1, at 897.

  120. [120]. See, e.g., Mendelson, supra note 18, at 653­–59 (comparing policy changes made at the end of a presidential term via notice-and-comment rulemaking, which promotes public participation and agency accountability, with other forms of policy making of “lower visibility” that are less transparent and present a greater risk of abuse).

  121. [121]. Gonzales and Glen take umbrage with the assertion that it is improper for the Attorney General to hear the arguments of only one side, receiving advice on referred cases from government attorneys while refusing to accept briefs from affected parties and interested amici. They contend that DOJ attorneys in the Office of Immigration Litigation, who are responsible for defending BIA and Attorney General decisions (including Silva-Trevino) in court “are not just litigators,” but also serve policy-related functions, and that it is a “caricature” to assert that they are not objective. Gonzales & Glen, supra note 1, at 919. They cite two cases in which the government agreed in its brief with a position asserted by a noncitizen contesting deportation to suggest that government litigators “do not evidence any doctrinaire antipathy towards the legal positions offered by aliens.” Id.

    I have previously analyzed the role of Department of Justice attorneys in developing immigration policy, explaining the potential risks and benefits when litigators are enmeshed in policy development. Margaret H. Taylor,
  122. [122]. Gonzales & Glen, supra note 1, at 898–901.

  123. [123]. Holper, supra note 3, at 1277–81.

  124. [124]. Id. at 1285–86.

  125. [125]. See Trice, supra note 3, at 1781.

  126. [126]. Zadvydas v. Davis, 533 U.S. 678, 693 (2001). Supreme Court precedent suggests that procedural due process rights can vary depending on circumstances and status and historically extends less protection to applicants for admission. United States ex rel. Knauff v. Shaughnessy, 338 U.S. 537, 543 (1950); see also David A. Martin, Due Process and Membership in the National Community: Political Asylum and Beyond, 44 U. Pitt. L. Rev. 165, 173–80 (1983). But lawful permanent residents like Mr. Silva-Trevino have a significant stake in remaining in the United States, and are afforded full procedural due process rights in removal proceedings. Landon v. Plasencia, 459 U.S. 21, 32–35 (1982); see also Hiroshi Motomura, The Curious Evolution of Immigration Law: Procedural Surrogates for Substantive Constitutional Rights, 92 Colum. L. Rev. 1625, 1650–56 (1992).

  127. [127]. Trice, supra note 3, at 1781.

  128. [128]. Id. at 1782.

  129. [129]. Id. at 1794. Mathews v. Eldridge instructs that “resolution of the issue [of] whether the administrative procedures provided... are constitutionally sufficient requires analysis of the governmental and private interests that are affected.” Mathews v. Eldridge, 424 U.S. 319, 334 (1976). Generally this requires balancing three factors: “first, the private interest that will be affected by the official action; second, the risk of an erroneous deprivation of such interest through the procedures used, and the probable value, if any, of additional or substitute procedural safeguards;” and third, “the Government’s interest, including the function involved and the fiscal and administrative burdens that the additional or substitute procedural requirement would entail.” Id. at 335.

  130. [130]. Gonzales & Glen, supra note 1, at 906–12. Gonzales and Glen assert that “judicial review is an adequate protection against the erroneous deprivation of any cognizable right an alien may have.” Id. at 908. But the focus of the Mathews v. Eldridge test is on procedural protections at the agency level. The Supreme Court’s procedural due process jurisprudence has never suggested that the prospect that a reviewing court might correct the substantive outcome of an agency proceeding weighs so heavily (or even at all) when assessing the constitutional adequacy of agency procedures. Indeed, it “runs contrary to common sense” to suggest that judicial review on appeal minimizes the risk of erroneous deprivation to such an extent that it essentially wipes away any defect in agency procedures. J.E.F.M. v. Holder, 107 F. Supp. 3d 1119, 1141 (W.D. Wash. 2015). If that were the case, then even the most egregious deprivation of procedural rights at the agency hearing would not violate procedural due process so long as judicial review of the substantive outcome was available. Id. (noting that “[u]nder this theory, [appointed] counsel would be unnecessary even in a criminal proceeding because the accused, if convicted, could always appeal”).

  131. [131]. Trice recognizes that “the Mathews test is perhaps not a perfect fit for determining the process required when the Attorney General considers broad legal questions upon review,” as opposed to evaluating what procedures are required to ensure accurate fact-finding at the initial hearing. Trice, supra note 3, at 1783–84. She nevertheless centers her procedural due process analysis on Mathews v. Eldridge because it has been “used to determine the process due in administrative appeals before the BIA.” Id. at 1784. In addition, the Mathews v. Eldridge calculus, which weighs competing interests when a procedural due process challenge seeks more extensive hearing rights across a category of cases, 424 U.S. at 334–35 may not be the right test to apply when core procedural protections are omitted altogether.

  132. [132]. Gagnon v. Scarpelli, 411 U.S. 778, 790 (1973).

  133. [133]. See, e.g., LaChance v. Erickson, 522 U.S. 262, 266 (1998) (“The core of due process is the right to notice and a meaningful opportunity to be heard.”); Cleveland Bd. of Ed. v. Loudermill, 470 U.S. 532, 542 (1985) (“An essential principle of due process is that a deprivation of life, liberty, or property ‘be preceded by notice and opportunity for hearing appropriate to the nature of the case.’”); Mullane v. Cen. Hanover Bank & Tr. Co., 339 U.S. 306, 314 (1950) (holding that due process requires “notice reasonably calculated, under all circumstances, to apprise interested parties of the pendency of the action and afford them an opportunity to present their objections”); see also Dusenbery v. United States, 534 U.S. 161, 168 (2002) (stating that “we have never viewed Mathews as announcing an all-embracing test for deciding due process claims”).

  134. [134]. Motomura, supra note 126, at 1679–80.

  135. [135]. Trice, supra note 3, at 1789 (quoting Conn. Dep’t of Pub. Welfare v. H.E.W., 448 F.2d 209, 212 (2d Cir. 1971)).

  136. [136]. Gonzales & Glen, supra note 1, at 853–54.

  137. [137]. Id. at 910.

  138. [138]. Id. at 909.

  139. [139]. Id. at 917–19.

  140. [140]. See supra note 42 and accompanying text.

  141. [141]. See, e.g., M. Elizabeth Magill, Agency Choice of Policymaking Form, 71 U. Chi. L. Rev. 1383, 1390, 1391 n.17 (2004) (noting that “today, promulgating an important legislative rule is a labor-intensive enterprise” and citing extensive administrative law literature on the “ossification” of the rulemaking process).

  142. [142]. Boyd & Driscoll, supra note 11, at 589 (concluding “adjudication oversight is a powerful means by which agencies further their goals. Furthermore, unlike rule promulgation, adjudication permits agency activism without formally involving the public via notice-and-comment periods”).

  143. [143]. Id. at 571.


Professor of Law, Wake Forest University School of Law.

I am grateful to Chris Coughlin and Sid Shapiro for comments on this essay, and to Jessica Strokus, Wake Forest University School of Law class of 2016, for excellent research assistance.