Introduction

The temporal scope of judicial decisions has long been “among the most difficult of those [issues] which have engaged the attention of courts.”1 When a court announces a new rule, does the new rule apply only to future cases or also to disputes arising in the past?

Over the past half century, the Supreme Court has addressed this temporal puzzle primarily through the lens of non-retroactivity doctrine. This doctrine, also known as “prospective overruling” of the law, defies the conventional conception of common-law adjudication whereby judicial decisions both bear upon past events and lay down the law for future cases. Non-retroactive adjudication constrains the effects of judicial changes in the law from applying to the past. Only events postdating a new precedent are treated as governed by it. Litigants’ rights to legal recourse under this doctrine are thus determined according to the timing and outcome of any relevant leading case.

Federal non-retroactivity doctrine peaked during the Warren Court era. It now seems destined for demise before the Roberts Court. While scholars continue to argue that non-retroactivity is a useful paradigm, this Article contends that it is not. To the contrary, non-retroactive adjudication is a defective and superfluous doctrine. It lacks a coherent and generally accepted rationale. There is no agreement within the judiciary or academe as to how the doctrine should be conceived. Non-retroactivity cannot even perform its basic job: to rationalize and contain the temporal scope of novel precedent.

This Article advances an alternative framework for understanding novel precedent, one that returns to conventional common-law reasoning. This framework orients judges’ focus toward the claims that come immediately before their courts—those over which they have direct jurisdiction. It embraces the retroactivity of judicial precedent. Disputes over rights, adjudicated by courts, can only be resolved from the perspective of hindsight, and they cannot feasibly be insulated from developments in precedent. Precedent today necessarily informs our understanding of past rights. That does not mean, however, that new rights of action are unlimited in temporal scope. Rather, interests of justice and fairness are embodied in long-recognized temporal limits on plaintiffs’ rights to obtain relief from a court.

Supreme Court jurisprudence on the temporal scope of novel precedent has effectively been dormant since the mid-1990s.2 But recent judgments have reignited the retroactivity debate. After the Supreme Court in Obergefell v. Hodges declared that the Fourteenth Amendment guarantees same-sex couples a right to marry,3 attention turned to the judgment’s remedial implications. Many scholars encourage “backdating” same-sex marriage to vindicate the rights of those previously denied the constitutional protection.4 They argue, for example, that where a same-sex spouse passed away before Obergefell was handed down, courts should uphold the partner’s claims to surviving-spouse pension benefits or to the primary share of the deceased’s estate by treating the couple as retroactively married in law.5 Others counsel restraining the judgment’s retroactive effects to protect prior reliance interests and the finality of past transactions.6 These scholars are concerned that, among other things, retroactive expansion of spousal property rights and liabilities under Obergefell would likely not have been anticipated and accounted for by same-sex couples—nor, indeed, by their creditors (who may gain access to newly deemed community property) or by those with whom they entered (now potentially voidable) transactions. Both of the competing “right” and “reliance” arguments are compelling.

A similar dynamic has followed the Court’s decision in Janus v. AFSCME, which held that the deduction of union agency fees from nonconsenting public-sector employees violated the First Amendment.7 Immediately following the judgment, plaintiffs sought to vindicate their newly announced rights through lawsuits filed across the country demanding recovery of fees paid before Janus was decided. Some scholars believe such plaintiffs have good constitutional and private-law grounds, given that Janus held that these fees were unconstitutional.8 Others do not, on the grounds that agency fees “were indisputably lawful” at the time of collection.9 Since fee refunds could bankrupt unions, these scholars would limit the retroactive effect of Janus to protect union coffers based on the unions’ reasonable reliance on the state of the law at the time agency fees were collected.10

This tension between newly declared rights and reliance on prior established rights arises whenever a court “changes” the law as previously understood. Claims for restitution of unlawful taxes are another prominent example. When a court strikes down a taxing statute or overrules prior precedent, its new precedent presents plaintiffs with a compelling claim to a remedy for their unlawfully impinged rights.11 Prima facie, plaintiffs are entitled to restitution of taxes improperly paid.12 There should, after all, be no taxation without (valid) legislation. But defendant states and municipalities have a compelling counterargument: that their treasuries should not be vulnerable to extensive money claims based on interpretations of law that were not known at the time of collection.13 The stakes are high. In the leading Supreme Court case on recovery of unconstitutional taxes, $1.8 billion of tax revenue hinged on the retroactivity of the Supreme Court’s new precedent.14 In the United Kingdom, £55 billion of tax revenue was thought to be at stake after tax provisions were retrospectively found to be incompatible with European Union law.15

Similar concerns arise when the public- and private-party interests are inverted. Where private entities have benefited from a precedent that is later overruled, states may be able to seek backward-looking remedies under the new precedent. For example, for half a century, federal doctrine exempted businesses that had no “physical presence” in a state from state sales tax.16 When the Supreme Court was invited in South Dakota v. Wayfair to overrule the physical-presence rule,17 the respondent businesses implored judicial restraint, warning that overruling the long-standing precedent would “expose all remote sellers that have relied on the rule to retroactive liability in dozens, if not hundreds, or even thousands of jurisdictions.”18 The temporal repercussions of adjudication are pervasive, and managing them may seem irresoluble.

The solution to this puzzle does not lie in denying the retroactive effects of novel precedent. A familiar flaw in theories of non-retroactivity is failure to appreciate adjudication as a dynamic experience.19 Part I of this Article shows that novel judgments do not simply replace “old” law with “new” law. Rules and principles are constantly subject to elaboration, challenge, and revision. Judgments do not apparate. They proceed from a background of complaint and litigation. The stakes are set long before issues reach a courtroom. In this context, novel judgments may be considered surprising, but they are never wholly a surprise. The protest that novel precedent unpredictably “changes” past law on which parties may have “relied” is thus overwrought. Since interpretation of law is not divisible into static points in time, adjudicating courts can only coherently understand past rights in light of prevailing law. Judicial precedent is by nature retroactive: there is no prospectivity puzzle at all.

This Article advances two core contributions to non-retroactivity scholarship. First, Part II challenges the prevailing rationales for non-retroactivity doctrine in federal law. It identifies in the case law and commentary several distinct frameworks. Each framework misconceives the essence of the temporal problem and relies on distinctions that are arbitrary and manipulable. Non-retroactive adjudication simply has no principled foundation.

Second, Part III offers an alternative framework grounded in conventional common-law reasoning: that judicial precedent is inherently retroactive. Since adjudication is dynamic, the essential question is not when it is that law is changed, but rather when it is that plaintiffs can “timely challenge” the validity of legal rules affecting them.20 This inquiry—the time at which “novel” rights of action become justiciable—asks when a plaintiff incurred the complained-of harm. As a general matter, when a court finds with hindsight that there has been a violation of a plaintiff’s right, the violation should be remedied according to prevailing law at the time of judgment. In most cases, that will be the end of the matter. In exceptional cases, where the extent of such relief would be unduly prejudicial to defendants, courts can invoke the equitable doctrine of laches, among other possible tools, to constrain the scope of litigation. After Janus, for example, a court might well be justified in limiting those who sued in Janus’s wake to a recovery period much shorter than that provided by the statute of limitations. Judges need not deny claims entirely by resorting to the sledgehammer of non-retroactivity doctrine. The response to “new” law advanced in this Article reflects a long-held value in our legal system: that equity aids the vigilant, not those who sleep on their rights.21