The Right Coast

January 15, 2004
Originalism and Precedent
By Michael Rappaport

About 10 days ago, I participated in a Federalist Society Faculty Conference panel entitled the “Transition to Originalism.” The focus of the panel was the role of precedent in an originalist jurisprudence. Organized and moderated by Randy Barnett, the panel also included Steve Calabresi, Rick Kay, Larry Solum, and Keith Whittington. It was a pleasure and an honor to be included among these genuinely excellent scholars.

In general, the speakers held three different positions on precedent. The largest group defended what I believe to be the most popular view of precedent among originalists. Under this view, precedent is either entirely unconstitutional (see Gary Lawson’s works) or a practice that should be minimized to the extent possible. On the panel, Rick Kay took the Lawson position that all precedent is illegitimate, while Calabresi and Whittington sought to minimize it. Calabresi made the interesting argument that some precedent is legitimate, since precedent was accepted by the Framers, but only in the relatively rare case when all three branches had approved of a precedent over a period of time. Whittington maintained that precedent was generally problematic, but in cases when prior decisions had to be followed, as with some of the New Deal cases, they should be read narrowly and not given any generative force.

The second position was defended by Larry Solum, who argued for a system of strong precedent. Solum powerfully developed his view that the rule of law requires that precedents be followed and that decisions should be overruled only when they are inconsistent with the remainder of the law. Solum has posted his argument from the panel here and has defended his view at greater length here.

I took the third position, arguing in favor an intermediate approach to precedent. I maintained that the Constitution does not prohibit precedent, but actually requires at least a very weak version of precedent. Moreover, the best normative approach to precedent is one that generally confers precedential weight on decisions, especially a series of decisions, but allows for overruling when there are strong reasons for doing so. The remainder of this post will briefly develop this view.

The most basic question involving precedent is “what type of law are the precedent rules?” In my view, precedent rules are mainly common law rules that are discovered by the courts but can be overridden by Congress through its power to pass necessary and proper laws for carrying into execution the judicial power. There is, however, a minimal degree of precedent that derives from the vesting of judicial power in the federal courts. Given the widespread acceptance of precedent in the Anglo-American world of the late 18th and early 19th century, including by such notables as Blackstone, Madison, Jefferson, Marshall, Hamilton, Adams, Jefferson, Kent, Story, and Patterson, it is fair to conclude that the Framers’ generation would have deemed a court that completely ignored precedent not to be exercising the “judicial power.” It is important to stress, however, that the Constitution only requires an extremely weak form of precedent. While some people at the time of the Constitution embraced a strong view of precedent, many supported a weak view that conferred significant weight not on a single decision, but only on a series of decisions. Thus, there was widespread acceptance only of a weak form precedent and only that can be legitimately understood to be part of the judicial power. Although the claim that the Constitution incorporates a weak precedent rule is significant as a matter of theory, its practical effect is limited. The main practical import of the constitutional rule is to forbid the no-precedent approach and to prohibit some of the circuit court rules that forbid courts from treating unpublished opinions as precedents.

Given that the Constitution does not determine the precedent rule, but leaves it to statutory and common law, the question is what would be the most desirable precedent rule. While I did not offer a comprehensive approach at the panel, I did argue in favor of an intermediate approach that included several rules. First, there should be a presumption against following a precedent that is demonstrably erroneous. By refusing to follow such decisions, the Court can be confident that it is improving the accuracy of the law while at the same time leaving significant scope for precedent in the more common case of a decision that is erroneous, but not demonstrably so. Second, there should also be a presumption against following a decision when it was reached in violation of the precedent rules at the time and when the decision itself was erroneous. This rule would serve to deter violations of the precedent rules and help to address the situation where some judges ignore precedent rules but then expect other judges to follow their precedents. Third, decisions that have been significantly relied upon, such as a holding that Social Security is constitutional or a decision that governed the sentencing of hundreds of thousands of criminals, should generally be followed, even if they are demonstrably erroneous or otherwise eligible for overruling. (For Kyron Huigen’s discussion of the situation involving the sentencing of criminals, see here.)

The aim of an intermediate approach is to take account of both the benefits and costs of following precedent in an effort produce an optimal rule. By contrast, the leading originalist view of no-precedent ignores all of the benefits of following precedents. Not only is that view not required or even permitted by the Constitution, but it would also create enormous costs from the overturning of cases that have been relied upon and might create tremendous uncertainty as judicial decisions shift back and forth with the changing of personnel. Indeed, if originalists were to overturn popular precedents that were relied upon, the public reaction might be so great that it would doom the originalist enterprise, leading politicians of both parties to promise never to appoint judges who believe in that type of judicial behavior.

In the end, I believe the skepticism about precedent that most originalists have is based on some important insights about law and the need for accurate adjudications. And their disdain for the cynical use of precedent by unprincipled judges, as displayed in the per curiam Casey opinion, is entirely justified. But these originalist neglect the important benefits that precedent can produce. While I am convinced that a no-precedent rule is undesirable, the hard question is determining what precise precedent rules originalist should consider optimal. Since I believe this is such an important issue, I have decided to write an article developing further this intermediate view of precedent. I hope to post it in the future.