At the end of the 2017 term, the Supreme Court decided not to stop time. Nonpermanent residents who have been placed in removal proceedings may apply for a discretionary form of relief from the Attorney General known as “cancellation of removal.” To be eligible, an applicant must show (in addition to meeting other requirements) that she has been in the United States for at least ten consecutive years. The period of continuous physical presence is interrupted when the government serves the noncitizen with a notice to appear at a removal hearing. However, in Pereira v. Sessions, the Court held that if the notice does not list the time and place of the hearing, which a separate provision—8 U.S.C. § 1229(a)—requires be included, the continuous presence clock continues to run.
While Pereira has generated immediate upheaval for cancellation of removal eligibility, the case’s aftershocks are already proving to be far greater. Since the decision was announced, there has been a feverous debate about whether the case applies not just when cancellation of removal is sought, but to all immigration removal proceedings, since notices to appear are issued in every judicial proceeding. It is a debate that reaches beyond immigration law circles. Coverage in the popular press has been notable.
Less than a year later, hundreds of federal court opinions, inclusive of both published and unpublished decisions, have wrestled with Pereira’s application. Although it is harder to track decisions by immigration judges, the number of rulings on Pereira challenges brought by noncitizens is surely in the thousands.
The volume of cases has not helped to settle the debate. To the contrary, in Pereira’s aftermath the courts have taken wildly divergent positions on the case’s meaning and are struggling to understand the difficult issues raised by the decision. With Pereira challenges being routinely raised by immigration lawyers across the country, and the Supreme Court unlikely to weigh in again anytime soon, the need for clarity is pressing and consequential.
The government maintains that Pereira bears no relevance beyond the narrow context in which the case arose. By contrast, immigration advocates insist not only that Pereira applies to all adversarial removal proceedings but that the Court’s decision necessarily means that immigration courts lack subject matter jurisdiction in any case commenced with a notice to appear that did not include the time and place of the proceedings. That argument is hinged to this regulatory language that governs removal proceedings: “Jurisdiction vests, and proceedings before an Immigration Judge commence, when a charging document is filed with the Immigration Court by the [Immigration and Naturalization] Service.” Since 1996, the notice to appear has been the primary charging document that the government serves on noncitizens who face removal in adversarial proceedings.
It follows, according to this view, that proceedings commenced with a defective notice to appear—defective, that is, according to Pereira, for not including the information required by § 1229(a)— must be treated as void ab initio and retried because they were commenced without subject matter jurisdiction. To put it mildly, that interpretation of Pereira would implicate a whole lot of cases. As of February 2019, there were more than 850,000 active proceedings in the nation’s immigration courts. What is more, the government admits that for at least the last three years, “almost 100 percent” of the notices to appear that it served did not include the statutorily required time-and-place information. If the courts were to reopen even finalized proceedings because they were tried without subject matter jurisdiction, the number of cases affected could be in the millions.
Beyond its impact on immigration law, what is also remarkable about the debate over Pereira is that it more broadly invites us to consider a number of difficult jurisprudential questions. For starters, in deciding the threshold inquiry into whether Pereira applies only to cancellation of removal cases, we confront a fundamental problem about judicial decision-making. How do lower courts determine the controlling scope of precedent? More precisely, the debate over how broadly or narrowly to read Pereira helps shed light on an important, but underexamined, aspect of the problem: what is the proper treatment of limiting language in a judicial decision? Appellate courts sometimes say that their judgments are “limited by the facts” or should be “narrowly drawn.” In Pereira, Justice Sotomayor twice repeated that the Court’s decision was “narrow.” What are lower courts to do with this kind of limiting language? How do they figure out the precedential reach of a decision that a higher court consciously drew narrowly?
Even if we conclude that Pereira applies to all notices to appear at judicial removal proceedings (and I will show that the case must be read to apply to all such notices), it is still necessary to consider how this reading of the decision matters. That is, in what cases does it matter, and under what circumstances? Unfortunately, the predominant approach among courts and advocates has been to examine Pereira challenges solely in jurisdictional terms. This is unfortunate because courts have ended up conflating the conclusion (a correct one, as I will explain) that a Pereira defect does not have any jurisdictional significance with whether it has any consequences at all. There may still be consequences if the government serves a notice that is defective under Pereira, even if those consequences are not jurisdictional. As I will show, whether a Pereira challenge can be successfully made, and how, turns on three doctrinal considerations that, to date, courts and advocates have largely ignored.
The first consideration is whether the Court’s decision in Pereira applies retroactively and, if so, to what cases. When an appellate court decides a civil case, the starting presumption is that the decision applies to all pending cases; correspondingly, retroactivity usually does not extend to fully resolved cases. Notwithstanding these general presumptions, legislatures sometimes enact laws that permit parties to seek relief far into the future. In immigration law there are several statutes that permit noncitizens to ask a court to reopen closed proceedings. Pereira provides us with a valuable opportunity to consider how a broad statutory grant of authority to reopen a case after it has been decided can be reconciled with presumptive retroactivity norms.
This initial inquiry into retroactivity links to a second question that also must be confronted whenever a Pereira challenge is made. Even when a decision is applied in a backward-looking way, what if the legal issue was not previously raised in the prior cases? Forfeiture rules normally require that parties make arguments to preserve them, but should an argument be deemed forfeited if it was not viable until after the law-changing decision was announced?
Finally, beyond these initial issues, there is a third doctrinal problem—probably the most significant of the three—that a Pereira challenge poses. Not every legal error is equally consequential. In other contexts, the law excuses some errors as harmless. But how do we distinguish between the kinds of errors that require correction and those that we acknowledge to be errors but, for prudential reasons, allow to remain uncorrected?
The Article proceeds as follows. Part I frames the issues in dispute, providing historical context leading up to the Court’s decision in Pereira. Part II then considers Pereira’s reach and meaning. I argue that the Court’s construction of §1229(a) is necessarily applicable to all notices to appear. Part III then considers the consequences of service of a defective notice. Part III.A begins by taking up the argument that a Pereira defect has jurisdictional implications. After concluding that defective notice has no bearing on an immigration court’s jurisdiction, Part III.B argues that Pereira challenges should turn on the three factors, outlined above, of retroactivity, forfeiture, and harmless error. Finally, in Part III.C, I consider all of the different case types in which a Pereira challenge might arise. I demonstrate that the three critical factors that control whether a Pereira challenge can be made vary in each of these case types.