On Text and Precedent
Amar, Akhil Reed, Harvard Journal of Law & Public Policy
Until recently, the conversation on originalism and the role of precedent has been dominated by two main camps, which I will call unoriginal originalists and unprecedented precedentialists. Unoriginal originalists refers to people who purport to pay close attention to text, history, and structure, but when these sources conflict with precedent, this camp basically does not have a theory at all. The theory becomes a sort of muddling through, sometimes following precedent and sometimes not. If one, however, is just going to muddle through, or be pragmatic about when to follow precedent, does that not undercut the very grounds on which one is an originalist in the first place? Why not then muddle through across the board or be pragmatic across the board?
It is tolerably clear, for example, that the exclusionary rule is completely made up from a constitutional perspective, and that no Framer ever believed that illegally seized evidence should be excluded from court; that England never had an exclusionary rule; that the Fourth Amendment definitely does not provide for an exclusionary rule; and that no state excluded evidence for the first hundred years after the Declaration of Independence, even though most of the states had Fourth Amendment counterparts. (1) If anything is clear, it is that the exclusionary rule is inconsistent with the original meaning of the Fourth Amendment, yet none of the supposedly originalist Justices on the Supreme Court reject the exclusionary rule. Even Justices Scalia and Thomas exclude evidence pretty regularly, and never quite tell us why they do so when it means abandoning the original meaning of the Fourth Amendment. (2)
What originalists ought to do is to deduce a theory of precedent from the text, history, and structure of the Constitution itself, and thus to see what are the proper metes and bounds of precedent. We have not seen a sustained effort to deduce such a theory yet, which is why we have unoriginal originalists.
The other side of the text versus precedent debate fares no better. On the other side are the unprecedented precedentialists--scholars and Justices who cannot explain why sometimes the Court ought to overrule and sometimes it ought not to overrule. Consider the following important statement from the decision in Planned Parenthood of Southeastern Pennsylvania v. Casey: (3) "[A] decision to overrule should rest on some special reason over and above the belief that a prior case was wrongly decided." (4) This point of view has recently carried the day on the modern Supreme Court, at least since the Casey decision. (5) The problem with this thesis is that it is inconsistent with both pre-Casey and post-Casey precedent. The Casey Court claims that its view of precedent--the view that a decision to overrule should rest on some special reason over and above the belief that the prior case was wrongly decided--has been "repeated in our cases." (6) To support that proposition, however, the Court cited only dissents! (7) Neither of the dissents cited was squarely on point, which leaves the careful reader with a sneaking suspicion that perhaps the Casey Court's view of when to overrule precedent was not well-established in the pre-Casey case law.
A strict count of the number of cases where the Supreme Court overruled itself on the basis of text, history, and structure alone, which excludes cases where there were overrulings because the doctrine was unworkable or because of some other pragmatic or doctrinal consideration, reveals five important cases in the twentieth century pre-Casey, and there may well be more. (8) This includes only pure, naked overrulings; that is, instances where the Court overruled itself based only on a changed view of the original meaning of the constitutional provision in question, effectively holding that the underlying case was wrong as an originalist matter. Analysis of these cases leads to the conclusion that Casey put forward a view of the sanctity of precedent that was itself unprecedented. …