By Samuel Beswick, Assistant Professor, Peter A. Allard School of Law, The University of British Columbia
I recently suggested on Balkinization that a storm seems to be brewing concerning the place of non-retroactivity doctrine (also called the doctrine of “prospective overruling”) in federal law. Non-retroactivity doctrine attempts to define the temporal scope of novel judgments based on whether they embody “old” law or establish “new” law. In short: if a court’s decision is based on pre-existing law, it will govern disputes retroactively in the ordinary way. If, on the other hand, a judgment makes “new” law, it may apply only prospectively to future cases. Prospectivity tends to be favored when the potential fiscal or administrative fallout of a new precedent exceeds what a court thinks the defendant should bear.
Seemingly abandoned by the Supreme Court in the early 1990s, recent judgments raise questions about whether the doctrine should be revived.
Consider two examples. Bostock v. Clayton County held that firing someone for being homosexual or transgender amounts to sex discrimination under Title VII of the Civil Rights Act of 1964. Janus v. AFSCME held that the deduction of union agency fees from nonconsenting public-sector employees violates the First Amendment. Were these holdings premised on the constitutional law that existed before the dates they were handed down? Or did these judgments make “new” laws that determine rights for the future but that do not apply to past cases?
In a forthcoming article in the Yale Law Journal, I argue that non-retroactivity doctrine’s “old” law versus “new” law framing rests on a false dichotomy. Judicial precedent is inherently retroactive. Those who suffered rights-violations before the dates Bostock and Janus were, respectively, handed down seemingly have good legal grounds for pursuing remedies from the courts.
My positive argument advances in two stages. The first stage invokes John Goldberg and Ben Zipursky’s theory of rights of action to show that novel rights of action will always precede their elucidation in novel judgments. This is inevitable: a novel judgment can only follow the filing of a (novel) claim. As the Supreme Court put it in Danforth v. Minnesota, “the underlying right necessarily pre-exists [the Court’s] articulation of the new rule.” I build on my previous work on the discoverability of mistakes of law to explain when it is that novel rights of action can be taken to have ripened or accrued, opening the door for plaintiffs’ legal action.
The second stage considers the limits on retroactive adjudication. Retroactivity does not mean perpetually reviving old disputes to be decided under more favorable novel precedent. Rights of action face many temporal constraints. In McGirt v. Oklahoma, the final judgment of the 2019/20 term, Justice Gorsuch observed that important doctrinal limits include “procedural bars, res judicata, statutes of repose, and laches, to name a few.” Laches, I argue, has untapped potential as a safety valve to ensure justice in the face of novel precedent. For example, while retroactivity might justify the claims in the Janus II litigation for restitution of past-paid union agency fees (as Will Baude contends), the doctrine of laches might justify limiting such plaintiffs to a recovery period much shorter than that provided by statutes of limitation (thereby partially protecting union coffers).
In light of the limits on rights of action already recognized in law, non-retroactivity doctrine is superfluous. That is a sufficient reason to abandon it. But there is further reason to denounce it: the doctrine is, I contend, unprincipled on its merits and a source of confusion over the precedential authority of judgments. Retroactivity is an inescapable aspect of adjudication that should be embraced.