The Volokh Conspiracy

Mostly law professors | Sometimes contrarian | Often libertarian | Always independent


Qualifying Qualified Immunity: When Adjudications Are Unfairly Retroactive


In a post yesterday, I introduced a series on The Fair Notice Rationale for Qualified Immunity, a paper that will be published by the Florida Law Review. The paper explores the extent to which fairness to officers supports the doctrine of qualified immunity. (For a primer on the doctrine, see my post yesterday.)

The first step of the argument requires a relatively deep dive into jurisprudence to explore the principle of prospectivity and its application to adjudication. What makes retroactive laws unfair? Are adjudications—decisions by courts—ever unfair in the same way, and if so, what makes them unfair? My discussion of this issue contributes to the literature on legal theory—it may not be right in every detail, but it poses some important questions that deserve answers.

The principle of prospectivity—that laws should ordinarily apply only prospectively, not retroactively—is a basic requirement of the ideal of the rule of law. Most people have an intuitive sense that it is unfair to be punished for something you could not have known was forbidden. The norm isn't only theoretical—its deeply ingrained in our law. The Ex Post Facto Clauses and Contract Clauses of the US Constitution prohibit some kinds of retroactive laws and statutes affecting private rights are presumed to apply prospectively.

Retroactive laws are widely considered to be unfair because they deprive people of the ability to make and follow through on their plans. Rather than treat people as self-directing agents who are worthy of participating in a regime of self-government, they treat them as means to whatever the government's ends may be. They are in tension with the ideal of the rule of law and unfair.

Yet suits against government officers do not, strictly speaking, involve retroactive laws; they involve adjudications, the application of pre-existing law to past conduct. In one sense adjudication is always retrospective – it always applies a rule to conduct that already happened. What my paper explores is whether the reasons retroactive law are unfair apply with equal force, or at least to some extent, to some adjudications, and, if so, whether some constitutional adjudications fall into that category.

Legal theorists usually consider the retroactivity of adjudication inscrutable because it depends on the answer to an underlying question, namely "When does a court 'make' law rather than merely apply it?" Theorists offer competing answers, but the question is perhaps irresolvable. To oversimplify, consider three fairly common responses. Suppose you take a hard realist view of what judges do. The professional and institutional norms might limit judicial discretion but judges generally do what they want. On this view, adjudications are little different than legislation. This clearly raises concerns about fairness to litigants. All that can be said for this perspective is that we generally tolerate that sort of unfairness, perhaps because there are no good alternatives.

By contrast, consider the classic Blackstonian view of adjudication. Courts simply enforce the law; they don't make it. There is no retroactivity problem, no unfairness. A modern version of this would be a theory that holds that there are always, or almost always, right answers, even in hard cases. Ronald Dworkin's view of adjudication and some theories of originalism are "right answer" theories. Under a right answer theory, adjudication never raises fairness concerns, at least so long as the judge is able to discern the right answer. Theorists recognize, though, that even if every legal question has a "right answer" in theory, no judge is perfect. A wrong answer might well impose a new rule (because it is not the right one) retroactively.

The most common modern view is that most judges are in fact doing something called law, some stylized, professional form of reasoning that reduces policy discretion, but that there are cases of unusual legal indeterminacy. In those cases, the law does not lead to one right answer and raises a greater risk of error.

Note that all three of these theories of judging admit some degree of judicial discretion, at least on the margins, because of legal indeterminacy, as a form of error, or as some combination of the two. Such discretion leads to legal unpredictability. Taking a cue from an article by Richard Fallon and Daniel Meltzer (in New Law, Non-Retroactivity, and Constitutional Remedies, 104 Harv. L. Rev. 1731 (1991)), my piece argues that the touchstone for unfairness ought to be the relative unpredictability of liability, not whether the court is "making law." This eliminates the metaphysical question of what counts as law and centers the inquiry on what makes retroactive law unfair: the fact that it deprives subjects of the ability to plan.


In light of this, I suggest what I think is an improvement on Fallon and Meltzer's theory of predictability. They say that we should measure the predictability of an adjudication from the standpoint of a reasonably capable lawyer. I think that is fair when the person subject to liability is a lawyer or perhaps a well-lawyered organization, but not when the person subject to liability is a layperson. Otherwise the judgment is what Jeremy Bentham called "dog law." It is no different than how we discipline dogs: the first time they do something they shouldn't, we discipline them to train them not to do it again. This may be acceptable for dogs, but it is no way to treat rational folks otherwise capable of planning their lives. Note that the predictability will depend on the defendant's relationship to the law. For constitutional cases, there is probably a spectrum, with members of the "Supreme Court bar" at one end of the spectrum and ordinary lay people at another. Most government agents are probably somewhere in the middle.

So, some adjudications do raise the same concerns about unfairness that are raised by retroactive law: those that impose liability for past conduct that the defendant could not have reasonable predicted. So far, my argument has focused only on whether some adjudications are ever unfair. The question of whether constitutional liability—or some instances of constitutional liability—are unfair to officers is the question I'll take up tomorrow.