The Yale Law Journal


Uniformity and Integrity in Immigration Law: Lessons from the Decisions of Justice (and Judge) Sotomayor

24 Mar 2014

Though courts and scholars emphasize the importance of uniformity in the interpretation and application of federal immigration law, systemic complexity makes its achievement elusive. In the immigration opinions she has drafted to date on the Supreme Court, as well as in her extensive work reviewing asylum adjudications on the Second Circuit, Justice Sotomayor has invoked uniformity as a means of promoting fairness and accountability. But she also has demonstrated how these values can be advanced even in uniformity’s absence, when the system produces conflict and divergent enforcement outcomes. Her opinions highlight how courts can meaningfully, albeit imperfectly, constrain administrative actors through consistent legal interpretation, while still accepting the diversity and discretion built into immigration law itself.

When President Obama nominated Judge Sonia Sotomayor to the Supreme Court, commentators and senators had an extensive record of judicial decisions to scrutinize for evidence of the sort of Justice she might become, given her many years of service on the Southern District of New York and the Second Circuit. Her immigration law jurisprudence, which constituted a significant portion of the caseload she managed while on the court of appeals, attracted considerable attention during the public vetting of her nomination. Because of the political salience of immigration law, her opinions offered a potential window into one of the debates that came to frame her nomination—the compatibility between the “empathy” the President said he sought in a nominee and the “fidelity” to law expected of judges.

As part of that debate, Senator Charles Schumer’s office conducted an analysis of her immigration decisions during her time on the Second Circuit. According to their analysis, in 848 asylum cases, she voted in favor of the government 83% of the time, putting her in line with the 17.1% remand rate of the Second Circuit as a whole.1 Schumer touted these numbers as a sign that, “[e]ven in immigration cases, which would most test the so-called ‘empathy factor,’ Judge Sotomayor’s record is well within the judicial mainstream.”2 On the surface, then, her immigration law record arguably sufficed to conclude that fidelity to law, and not empathy for litigants, would drive Justice Sotomayor’s decision-making.

But rarely in public discussions over judicial nominations is the weighty concept of fidelity meaningfully defined. In the asylum setting, is fidelity to law consistent with giving asylum claimants generous consideration and working within the bounds of the law to advance a principle of protection?3 Did Schumer’s framing of the issue reflect a presumption that faithful interpretation generally results in victory for the government?4 In general, does fidelity to law require uniformity or consistency in its interpretation and application, regardless of underlying circumstances?

In the immigration setting, the belief that judicial interpretation should be geared toward promoting a systemic objective of uniformity has exerted powerful influence over judges and commentators alike. This occasion marking Justice Sotomayor’s first five years on the Supreme Court provides us with an opportunity to interrogate this assumption, because her immigration opinions simultaneously highlight the theoretical value of uniformity and the far messier reality of the system’s operation. Though her immigration-related work on the Court has been limited to date, she has written two notable immigration opinions—one dissent and one majority. Together with her work on the Second Circuit helping to superintend the Board of Immigration Appeals (BIA), these opinions illuminate the tension between aspiration and reality at the heart of immigration law.

But rather than lament the tension, I suggest that Justice (and Judge) Sotomayor’s opinions demonstrate how federal courts can play a limited but important role in promoting consistency and transparency, if not uniformity, in the law’s implementation. As Justice Sotomayor has sought to do, courts can meaningfully, albeit imperfectly, constrain administrative actors by engaging in legal interpretation with uniformity concerns in mind. Importantly, however, courts can play this role while still accepting and perhaps even embracing certain forms of diversity in the administration of federal law.

I. the limits of uniformity as a systemic objective

The aspiration to uniformity in the interpretation and administration of the law cuts across legal domains. Uniform interpretation and application of the law advances numerous values that an ideal version of our legal system should serve, including fairness, transparency, accountability, and efficiency. Uniformity can advance the elemental principle of fairness by providing parties bound by the law with notice of its content and how it will be applied. By promoting transparency as to the law’s meaning, uniformity helps constrain the discretion of executive officials and ensure that the system’s actors are held accountable by creating a standard against which to judge their actions. These constraints, in turn, help promote pragmatic values such as efficient administration by bringing clarity to the law and its implementation. Uniformity also facilitates equal treatment by subjecting the same conduct to the same rules or sanctions, regardless of where or when the conduct occurs.

In immigration law, in particular, courts, advocates, and scholars often tout uniformity as crucial to the integrity of the system. The Supreme Court historically has framed the value of uniformity in terms of one voice-ism, or the idea that, on certain subjects, the nation must speak in unison. Uniformity under this view can serve systemic and expressive interests by conveying a unified sense of how the United States will interact with the people of the world. Uniformity in theory also provides a foundation for the protection of immigrants’ rights by advancing a clear conception of how the country conceptualizes the value of immigration and the status of immigrants within the polity.5 The defense of uniformity arises most visibly in federalism debates and contexts that require the assertion of federal power to control state and local efforts to engage in immigration regulation.6 But the aspiration to uniformity hangs over the implementation of the federal immigration code itself, too.

Promoting uniform interpretation and application of federal law preoccupies the federal courts.7 As Justice Scalia has written, the Court’s “principal responsibility . . . is to ensure the integrity and uniformity of federal law.”8 The pursuit of uniformity requires constant vigilance by the Supreme Court9—a vigilance manifest most obviously in the Court’s efforts to resolve disagreements among the courts of appeals,10 but also in efforts to monitor state court interpretation of federal law.11 It also demands discipline by the courts of appeals, requiring them to take seriously the precedent of other courts and consider the value of harmonization across circuits when interpreting federal law.12

But even if the presumption in favor of uniformity is warranted, uniformity itself proves elusive in both interpretation and administration.13 Federal statutes can be complex as well as poorly drafted, and in the modern era of delegation, Congress often leaves broad concepts undefined or relies on open-ended standards to set the terms of regulation. For example, the definition of “refugee,” on which grants of asylum turn, is made up of at least eight concepts whose meanings are far from clear.14 Legal questions must percolate, often for years, before the Supreme Court elects to use its limited resources to decide a question.15 And even when the Court does weigh in to resolve circuit splits, its pronouncements often leave a great deal of space for subsequent divergence in interpretation based on the particular facts of different cases and ambiguities in the Court’s reasoning.

In administration and law enforcement, uniformity also will be difficult to achieve. Factors such as the politics of different administrations, divergence in enforcement priorities between agencies’ central offices in Washington and field offices spread out across the federal bureaucracy, and disagreements between political leadership and the civil service all ensure that the meaning and content given to the law, and the extent of the law’s enforcement, will vary depending on context. Perhaps more important, in immigration law, as in many other regulatory domains, Congress actually has erected a regime whose design thwarts uniformity, the courts’ best efforts notwithstanding. Not only does the law incorporate state and local decision-makers into the system,16 it also makes the system of removal and relief dependent on the reach of state law and the fact of state law convictions, as well as the discretion of administrative actors.

Of course, percolation and diversity can be of affirmative value in legal interpretation.17 Each might permit experimentation and the testing of competing legal theories. Elsewhere I have defended the value of diversity in enforcement and administration, in explorations of how federalism helps determine the character of various domains of social policy, including immigration policy.18 In this Essay, rather than revisit these debates, I present an account of why uniformity may not be essential to advancing all of the values uniformity ostensibly serves. I use the setting of immigration law—a domain where uniformity has enjoyed rhetorical pride of place—to make the point, highlighting Justice Sotomayor’s contributions to our understanding of that field and the role of uniformity within it.

II. uniformity and integrity in justice sotomayor’s immigration opinions

Perhaps the most publicly salient move Justice Sotomayor has made on the subject of immigration since joining the Supreme Court occurred in a case that had nothing to do with immigration law itself. In Mohawk Industries v. Carpenter, a case exploring procedural questions related to the attorney-client privilege, she used the term “undocumented immigrant” for what might have been the first time in the pages of the U.S. Reports.19 The mere choice of vocabulary, unaccompanied by explanation or context, drew media attention, because it marked a departure from the Court’s traditional and often unthinking use of terms such as “illegal immigrant” and “illegal alien.”

When asked about her choice of language during a public appearance at Yale Law School on February 3, 2014—the lecture that followed the symposium for which I prepared this Essay—she articulated a substantive theory of the legal significance of illegal status. She also displayed a belief in the sociological implications and political meaning of the language of Supreme Court opinions.20 She likened immigration status offenses to regulatory or administrative offenses and offered that every one of us is capable of violating (and probably has violated) rules of this kind.21 She observed that defining a person by his or her illegal acts degrades the individual by eliding this reality, which in turn enables us to ignore the fact that the status violator is entitled to respect as a person despite his or her immigration status.

Justice Sotomayor’s explanation for her choice of terminology highlights the hybrid legal/political role of the Supreme Court as an institution, and it could well have an impact on the construction of the unauthorized immigrant in public discourse.22 Other Justices have employed similar language in recent cases.23 But the actual immigration law opinions Justice Sotomayor has written in her first five years on the Court have not been injected with the same sort of expressive political statements, though each position she has taken could be characterized as more pro-immigrant than the majority or dissent against which she was writing.24 In both style and substance, she has been the careful and measured interpreter of reputation, hewing closely to traditional modes of statutory interpretation and self-consciously attempting to effectuate the schemes of Congress as written.

These tendencies in her opinion writing ultimately help answer the crucial question with which I began this Essay: How can courts promote the values served by uniformity within a system that ultimately has an ambivalent relationship to uniformity? I thus turn to the two major immigration opinions she has written since joining the Court—her dissent in Chamber of Commerce v. Whiting and her opinion for the Court in Moncrieffe v. Holder—as well as some of her work reviewing asylum adjudications while on the Second Circuit, to address this dilemma in immigration law. As her opinions suggest, the best courts can do in their oversight of the sprawling arms of the regime is to promote a kind of consistency in legal interpretation to guide law enforcement and administration. By consistency, I mean a predictable approach to resolving immigration law questions that pushes administrative actors to adhere to consistent legal standards and articulate their reasons for acting, but that nonetheless accepts divergent outcomes as non-threatening, particularly when they are the product of the system Congress has designed. Sometimes this divergence will stem from the federal government’s necessary dependence on, or desire to incorporate, state and local actors in its enforcement regime. Sometimes it will result from Congress’s interest in writing a code that permits equities to be taken into account through the device of administrative discretion. But regardless of its source, divergence need not be unfair or threatening to good order, particularly if it results from a clearly defined and justified design.

A. Reconciling Uniformity and Immigration Federalism

Since Justice Sotomayor’s confirmation, the Court has confronted two significant cases concerning the scope of states’ authority to enact laws that effectively function as immigration regulations—the first time the Court has dealt directly with the federal-state balance of power in the immigration arena in nearly three decades.25 In both cases, the Justices’ opinions highlight different strategies for reconciling the statutory schemes created by Congress and the built-up practice of state and local involvement in immigration enforcement with the concept of federal exclusivity and the corresponding interest in a uniform immigration policy.26 In the first of the two cases, Chamber of Commerce v. Whiting,27 Justice Sotomayor’s dissent points the way toward an approach that keeps federal law and administration supreme while accepting the diversity in enforcement that might come from having an integrated regulatory regime—an approach echoed in Justice Kennedy’s far more high-profile majority opinion in Arizona v. United States.28

Whiting required the Court to interpret the meaning of a discrete provision of the Immigration Reform and Control Act of 1986. When Congress adopted its scheme to sanction employers who hired unauthorized workers, it responded to the various state laws then on the books that did the same by expressly preempting “any State or local law imposing civil or criminal sanctions (other than through licensing and similar laws) upon those who employ . . . unauthorized aliens.”29 Nearly thirty years after the fact, states seeking to re-enter the immigration enforcement game began taking advantage of the parenthetical savings clause. Arizona enacted the Legal Arizona Workers Act and provided that employers who commit a second violation of the law’s prohibition on intentionally or knowingly employing unauthorized aliens would have their licenses to do business in the state revoked. Advocates’ critiques of Arizona’s interpretation of IRCA’s savings clause drew heavily on the negative metaphor of the patchwork, emphasizing that laws such as Arizona’s would create multiple and therefore confusing immigration law regimes. By attaching far harsher penalties than federal law to the hiring of unauthorized workers and giving rise to differential enforcement patterns depending on geography, the laws would undermine values such as notice, predictability, and fairness.

Applying a plain meaning approach to the word “license,”30 the Court held that Arizona’s law was not preempted, but rather fell “well within the confines of the authority Congress chose to leave to the States.”31 The Court brushed aside the possibility of divergence in enforcement in light of congressional intent,32 noting that the licensing provision, much like “our federal system in general . . . necessarily entails the prospect of some departure from homogeneity.”33 The possibility that state laws might lead to greater enforcement was of no moment; why would Congress intend to preserve only those state sanctions that had no effect?34

Justices Breyer and Sotomayor each wrote dissenting opinions emphasizing uniformity concerns, but based on significantly different theories of the statute’s meaning. Justice Breyer, joined by Justice Ginsburg, narrowly defined the meaning of the word “licensing” in order to limit the impact of the enforcement discretion Congress left with the states. By contrast, Justice Sotomayor appeared to accept Congress’s decision to involve state and local actors in the enforcement of immigration law in a fashion that might lead to significant divergence in the scale of penalties imposed on employers in different parts of the country. But she rejected the state scheme nonetheless, on the ground that it authorized state rather than federal officials to determine when IRCA’s prohibitions had been violated. She thus advanced a theory of federal control designed to promote consistency in interpretation, if not uniformity in outcome.

In his dissent, Justice Breyer emphasized that Congress could not have intended to give states the broad authority the majority opinion allowed them, because it would “eviscerate” the preemption provision itself and undermine Congress’s efforts to protect workers from national origin discrimination and employers from erroneous prosecution.35 Breyer rejected the majority’s turn to dictionary definitions of licensing and concluded that Congress meant the clause to have a much narrower, historically specific meaning. He found that the clause applied to “the licensing of firms in the business of recruiting or referring workers for employment, such as the state agricultural labor contractor licensing schemes in existence when the federal Act was created.”36 He reconstructed the savings clause’s statutory context using the House Committee Report,37 post-enactment evidence that the scheme itself gave rise to discrimination,38 his own judgments concerning the likely disruptive effects of Arizona’s law on the reticulated federal scheme,39 and the history of federal and state regulation of agricultural labor contractors.40 In his judgment, the majority’s interpretation subverted the Act by upsetting numerous delicate compromises Congress had made. The Court’s interpretation gave states the power to enact sanctions far greater than those the Act authorized the federal government to impose—the “business death penalty,” so to speak. The majority’s interpretation thus would lead the federal government to lose control of the system of employer sanctions.

Justice Sotomayor framed her disagreement with the majority by emphasizing that Congress expressly wanted the immigration laws to be “enforced vigorously and uniformly.41 But she did not seek to narrow the definition of “licensing laws,” implicitly accepting instead the straightforward reading of the text given to it by the majority and Arizona. The problem with the state law, Justice Sotomayor emphasized, was that it substituted state officials’ judgment as to when the law had been violated for a federal determination. Such state authority was not contained in the text of the savings clause. In fact, the complex structure of adjudication erected by IRCA as a whole belied the majority’s interpretation, because the statute reflected “Congress’ intent to build a centralized, exclusively federal scheme.”42 By setting out a clear scheme for what constituted a violation of the law and erecting an administrative apparatus and particular procedures to make that determination, Congress delegated to federal administrative officials the authority to determine when the law had been violated.43 The delegation to state and local officials contained in the savings clause contemplated nothing more than state and local officials piggybacking on federal determinations that a violation had occurred, not the authority to decide complicated questions of immigration law involving whether a non-citizen was authorized to work.44

Despite rejecting the majority’s reading on the ground that it “subjects employers to a patchwork of enforcement schemes,”45 this approach still accepts a potentially broad form of state involvement in the system. It implicitly acknowledges the possibility that employers in some states will face more severe sanctions than employers in other states subject only to federal sanctions. In other words, Justice Sotomayor’s reading accepts certain departures from federal baselines, recognizing Congress’s expressed interest in having a full range of sanctions available to it, including sanctions that could only be imposed by state officials.46

But the vision in her dissent is still of federal monopoly in the definition of the law’s meaning.47 Even when Congress expressly authorizes states to enforce federal law, control over the meaning of federal law remains in the hands of the federal administrators to whom that authority has been clearly delegated. The centralization of adjudication helps ensure consistency in the interpretation of the law, even as the full scope of the law’s enforcement may diverge depending on geography.48

Echoes of this understanding of how state law enforcement can be incorporated into the implementation of federal law, while maintaining the latter’s integrity, appear in the majority’s opinion in Arizona v. United States.49 In Arizona, the Court confronted an omnibus state law designed in numerous ways to sanction and deter illegal immigration. Arizona framed S.B. 1070 as simply enforcing federal standards,50 but the Court held that most of its provisions were inconsistent with congressional design. The Court clearly had in mind a more complex conception of said design than the mirror-image theory advanced by Arizona.

The design, in fact, paralleled the approach Justice Sotomayor detailed in her Whiting dissent. Much as Justice Sotomayor argued that Congress intended federal adjudicators to determine the meaning of and identify violations of federal law through the procedures Congress had erected, the Court in Arizona assumed that Congress wanted its law’s prohibitions to be enforced according to the priorities determined by the federal administration.51 The meaning of federal law, though it might evolve over time, was to be determined by federal officials. Arizona’s parroting of the language of the Immigration and Nationality Act (INA) was thus insufficient to save a scheme that could have led state officials to make divergent enforcement choices. Though I would not claim that Justice Sotomayor’s dissent in Whiting served as any sort of concrete template for the majority’s reasoning in Arizona, each opinion reflects a similar theory of federal lawmaking. Congress delegates to executive officials implementation and enforcement authority, through which officials construct the meaning of federal law, leaving only as much enforcement authority to state officials as is consistent with the federal administration’s construction of the law.

These efforts to assert federal control over the operation of the system notwithstanding, however, the Arizona Court kept alive the possibility of some divergence in enforcement. It rejected the facial challenge to section 2(B) of the state law, which directs state and local police to make a “reasonable attempt . . . to determine the immigration status” of any person stopped, detained or arrested, if “reasonable suspicion exists that the person is an alien and is unlawfully present in the United States.”52 The Court accepted the long-standing practice of cooperation between state and federal law enforcement in immigration policing—interaction reflected in the design of the INA itself—noting that “[c]onsultation between federal and state officials is an important feature of the immigration system.”53 Consistent with the theory of enforcement expressed above, federal control over enforcement priorities did not necessarily preclude some state involvement in immigration policing. Even though this consultation could result in state and local officials making inquiries into immigration status where it “seems unlikely that the Attorney General would have the alien removed,”54 the Court could not conclude that these (probably minor) incursions on federal authority necessarily undermined the integrity of federal law enforcement.55

B. Uniformity and the Interpretation of the INA

The gap between theory and practice when it comes to uniformity also appears in the interpretation and enforcement of the INA itself, even when only federal actors are directly involved. In this section, I turn to some important examples of Justice Sotomayor’s jurisprudence to highlight two sources of that gap: the reliance on state-law convictions as predicates for removal and the incorporation of significant levels of administrative discretion into the code itself. When confronted with these challenges to uniformity, Justice Sotomayor has adopted positions that drive toward consistency in court interpretation without necessarily insisting on uniformity of outcome. These efforts may be the best the judiciary can do to promote the integrity of federal law by constraining the administrators who apply it.

1. Harmonizing State Law with the INA

Justice Sotomayor’s most significant immigration opinion to date as a member of the Court involved a recurring immigration law question—how to treat certain state law convictions for the purposes of federal immigration law. In Moncrieffe v. Holder, the Court considered whether a state drug conviction constituted illicit drug trafficking under federal law, such that the conviction amounted to an “aggravated felony” for the purposes of the INA. Classification of the offense of conviction as an aggravated felony would not only have made the petitioner removable, but also would have precluded the Attorney General from granting him discretionary relief.56

Petitioner had pleaded guilty to possession of marijuana with intent to distribute under Georgia law after being found with the equivalent of two or three marijuana cigarettes during a traffic stop. In subsequent removal proceedings, the immigration judge ordered his removal on the ground that his conviction constituted an aggravated felony.57 The Court granted certiorari to resolve a dispute among the courts of appeals over whether a conviction under a state statute that criminalizes conduct encompassed by both the felony and misdemeanor provisions of the Controlled Substances Act (CSA) constitutes a conviction for an aggravated felony, which requires that the conviction be for an offense that “proscribes conduct punishable as a felony under” the CSA.58

In a 7-2 decision, the Court held that Moncrieffe’s offense did not amount to an aggravated felony.59 The Court reached the conclusion that his state conviction did not constitute illicit trafficking under the CSA60 by applying the so-called categorical approach (or at least a modified version of it). A state offense constitutes a categorical match with a federal offense “only if a conviction of the state offense ‘“necessarily” involved . . . facts equating to [the] generic [federal offense].’”61 Under this methodology, the Court examines what the state offense necessarily involved, not the facts that led to conviction. It therefore presumes that the conviction “rested upon [nothing] more than the least of th[e] acts criminalized.”62 In Moncrieffe’s case, his conviction for possession with intent to distribute did not necessarily require the prosecutor to show remuneration or possession of more than a small amount of marijuana. His offense therefore could have corresponded either to the felony or misdemeanor offense of the CSA.63 Accordingly, under the categorical approach, he was not convicted of an aggravated felony.

The central conceit of the categorical approach is that it promotes uniform treatment of convictions by looking not at the underlying facts of a state-law conviction, but at the elements of the conviction, assessing whether they map onto federal law.64 Perhaps its most important virtue is that it cabins immigration judges’ discretion, preventing them from re-trying immigrants’ underlying crimes in order to determine their immigration consequences. Justice Sotomayor explicitly rejected the government’s proposal that non-citizens be given an opportunity during immigration proceedings to demonstrate that their convictions involved only a small amount of marijuana, precisely because it would have opened the door to “post hoc investigation into the facts of predicate offenses,” undermining systemic interests by leading to “minitrials conducted long after the fact.”65 Not only would such reconsideration of the circumstances of conviction be a drain on the resources of a backlogged system, it would present fairness concerns by imposing heavy burdens on those in removal proceedings, many of whom might be in detention. Relitigation could require non-citizens in civil removal proceedings to gather potentially stale evidence without the benefits of the sorts of procedural protections that govern the criminal process and might help facilitate relitigation of the underlying offense.66

In dissent, after claiming that a true categorical approach would have led to a holding that petitioner’s conviction amounted to an aggravated felony,67 Justice Alito called for setting the categorical approach aside in circumstances like the ones presented by Moncrieffe’s case.68 To buttress his interpretation, he pointed to some of the anomalies that the majority’s supposed uniform interpretation would create—namely, that the immigration consequences of a conviction for possession with intent to distribute would “vary radically” depending on the state in which the individual was convicted.69 He contended that the majority’s approach would lead to “significant disparity [in] our treatment of drug offenders,” which is precisely what the categorical approach was supposed to avoid.70 In addition, in some states, major drug trafficking crimes would be categorically excluded from the aggravated felony category just by virtue of the drafting of state law—a result he did not believe Congress could have intended.71 He thus underscored how efforts to promote uniform legal interpretation on one level could result in disparate treatment on another.

Moncrieffe, then, reflects a trade-off: a choice between a system that treats like conduct alike but enables administrative actors to determine who precisely is similarly situated, and one that applies a consistent rule to cabin the discretion of executive actors but empowers the federalism that underlies the system to produce disparate and arguably unfair outcomes. Both approaches promote a kind of uniformity, but at the expense of another.

The majority was not blind to the dilemma posed by Justice Alito and the government. Justice Sotomayor emphasized that the INA still enabled the government to smooth out the anomalies the Court’s approach might produce. She made clear that avoiding an aggravated felony designation did not mean avoiding removal.72 Moncrieffe was still removable as a “controlled substance offender.” The significance of his win before the Court was simply that he became eligible for cancellation of removal—a form of relief the INA gives the Attorney General the discretion to deny regardless of whether the non-citizen meets the eligibility criteria for it in the statute. In other words, if in reality Moncrieffe had been a significant drug trafficker, the government could still have removed him.

Now, does the majority’s acknowledgment of this safety valve for the government suggest that its categorical classification of offenders like Moncrieffe amounts to drawing distinctions without a difference? If the discretion removed from the immigration judge when he determines a non-citizen’s eligibility for relief is simply revived during the discretionary phase of his adjudication, do the constraints of the categorical approach amount to much?

Justice Sotomayor offered a potential explanation for why her particular framing of the case still served an important constraining function. On the one hand, she downplayed the government’s warnings of disparate treatment of serious offenders by highlighting the government’s discretion to deny relief and therefore still remove drug offenders not found to have committed aggravated felonies. But she also expressed an affirmative preference for the underinclusiveness of the categorical approach at the threshold eligibility stage. In her view, constraining the government on the front end would better constrain the government overall by avoiding the costs of relitigating old prosecutions.73

Despite the discretionary safety valve left open to the government, the constraints of categorical analysis seem likely to enable greater numbers of non-citizens to surmount the eligibility hurdle for relief. This approach arguably will produce more uniform legal interpretation and impose fewer burdens on individuals in proceedings than the diverse approach to adjudication advocated by Justice Alito. A finding of eligibility might in turn inform and limit the subsequent discretionary determination, and it might cabin ex ante enforcement decisions, too. Congress built the back-end discretion of the relief determination into the design of the statute precisely to make discretion an exception, rather than part and parcel of the adjudicator’s job. Justice Sotomayor’s resolution of the case thus highlights how consistency in interpretation, even more than uniformity of outcome, can promote the integrity of the system and fairness toward litigants, through the constraints consistency imposes on government actors.74

2. Interpretive Consistency in Asylum Adjudications

The near-impossibility of achieving uniformity in the application of federal law has been vividly on display in the adjudication of asylum claims—a process that has become notorious for yielding disparate aggregate outcomes based sometimes on seemingly arbitrary factors, including the identity of adjudicators.75 Pursuant to the INA, the courts of appeals superintend these adjudications, and the courts’ criticism of the work of immigration judges has been pointed. Judge Richard Posner of the Seventh Circuit has drawn attention to what he regards as poorly reasoned agency judgments that in some instances have “fallen below the minimum standards of legal justice.”76 And in response to reported disparities in the remand rates in asylum cases by the courts of appeals themselves,77 he has said: “This is supposed to be a uniform body of federal law, not something turning on the luck of who you draw for an immigration judge or what circuit you happen to be in for the appeal. . . . This is a pathological picture that I am painting.”78

While Justice Sotomayor has not been so explicitly critical, her asylum opinions from her time on the Second Circuit do reflect the view that the courts should work to simultaneously respect and constrain immigration judges and the BIA. On the one hand, in most asylum cases, she and her Second Circuit colleagues follow(ed) a fairly straightforward and familiar script of deference to the administrative agency,79 and Judge Sotomayor took notice when she believed her colleagues had departed from traditions of deference and judicial restraint.80 But many of her asylum opinions reflect what I would call a BIA-forcing approach, consisting of prods to the agency to clearly articulate its interpretations of asylum law and to make them consistent over time.81 In some instances, Justice Sotomayor framed her BIA-forcing prods as part of the court’s duty to ensure uniform and therefore fair treatment of asylum applicants.82 And often they were accompanied by careful review of the factual record before the immigration judge, not to second-guess his or her factual findings, but to ensure that the judge’s legal conclusions were actually supported by the facts before him or her.83

But given what we know about the disparate outcomes at the administrative level, not to mention among the courts of appeals, does the emphasis in any given opinion on promoting uniformity in law have a naïve quality to it? As I noted at the outset of this Essay, harmonization can be challenging, especially with respect to the application of open-ended standards. For example, the process of applying the INA’s definition of “refugee” to novel circumstances, such as the evolution of China’s family planning policy, seems likely to lead to disparate outcomes whether arbitrary factors related to the adjudicators’ identities are in play or not. The open-textured nature of the definition of “refugee,” the centrality of credibility determinations to the adjudication of a claim,84 and the discretionary component of the INA’s asylum provisions85 all push against uniformity in outcome.Disparities in adjudication of the asylum laws thus may be inevitable whatever the courts do. And confusion can also be a signal of a poorly drafted statute or of an inherently difficult standard to apply to novel circumstances. In these cases, percolation and the generation of multiple theories might in the long run produce better interpretations.

But as reflected in Judge Sotomayor’s opinions, the courts of appeals can and do play a complementary and useful role in relation to the administrative process, namely by striving for consistency and integrity in legal interpretation, at least within their own circuits.86 In Zheng v. Gonzales, for example, Justice Sotomayor acknowledged the “array of positions” taken by the courts of appeals on the question of when forced insertion of an IUD constitutes persecution. The court remanded the single-member BIA decision nonetheless to give the Board the opportunity to articulate its position on the novel and confused question. Her opinion reflected both the court’s interest in demanding reasoned decision-making by the adjudicator and the interest in having difficult interpretive questions percolate up through the channels of review (and by extension, perhaps, across the courts).87 Such efforts can help make disparities in outcome more tolerable by ensuring that they stem from the vagaries we should expect from a complex legal system and not from the absence of guiding legal standards and independent supervision of the implementation of those standards.88

As a nominee, Judge Sotomayor seemed well aware of the limited but important role of the federal courts in maintaining the law’s integrity in a world of elusive uniformity. When asked during her confirmation hearings by Senator Richard Durbin whether she agreed with Judge Posner’s estimation of the quality of the BIA’s work, Judge Sotomayor acknowledged that the system of adjudication as managed by the Department of Justice had given rise to procedural and adjudicative challenges that led to “cooperation between the courts and the immigration officials in how to handle these cases, how to ensure that the process would be improved.”89 She described the courts as being in “dialogue” with the agency and acknowledged the Department of Justice’s efforts to better manage the process. But she also focused attention on Congress, too, emphasizing that the legislature was the branch with the power to allocate the resources necessary to improve adjudication. She was realistic about the power of the courts: all they can do is “ensure that due process is applied in each case according to the law required for the review of these cases.”90 In other words, amidst systemic complexity, uniformity of outcome may be unachievable, but by holding adjudicators to consistent standards, a system that produces divergent outcomes can still have integrity.


The rhetoric of uniformity looms large in immigration law, both as framed by courts and as debated in the political arena. But as is true across legal domains, the achievement of uniformity in the interpretation and application of federal immigration law proves elusive. Whatever one thinks about the importance of the objective, our immigration laws are drafted in a way that makes uniformity an unrealistic aspiration. In addition to being affected by the ordinary features of a complex system of federal administration made up of multiple decision-makers, several features of immigration law itself mean that uniformity will be more rhetoric than reality. The reliance on state and local actors in enforcement, the fact that state law serves as a predicate for numerous immigration offenses, and the open-ended standards and delegated discretion that give administrative actors considerable room to maneuver all ensure that conflict will often thwart harmony.

Yet several of Justice Sotomayor’s opinions demonstrate that the work of the courts in superintending the system is not futile. The courts can bring integrity to the system without necessarily ensuring uniformity in outcome or administration, particularly by holding administrative actors to transparent legal standards and forcing them to give reasons for their actions. These tendencies in the immigration opinions Justice Sotomayor has authored cohere with emerging themes in the evaluation of her jurisprudence—that she pays rigorous attention to the factual and institutional features of a case and believes in hewing to what can be readily identified as legal principles. This approach may reflect a cabined view of the courts’ role in managing any given regulatory scheme, but it also advances some of the same values as the rhetorically appealing but practically unattainable principle of uniformity, serving systemic and individual interests in justice through fidelity to law.

Cristina M. Rodríguez, Professor of Law, Yale Law School, is grateful to participants in the Yale Law Journal symposium in honor of Justice Sotomayor’s first five years on the Supreme Court for discussion of the issues explored in this essay. She also benefitted considerably from the excellent research assistance of Xiao Wang and Samuel Kleiner.

Preferred citation: Cristina M. Rodríguez, Uniformity and Integrity in Immigration Law: Lessons from the Decisions of Justice (and Judge) Sotomayor, 123 Yale L.J. F. 499 (2014),