Don't talk to me about Superprecedents

Jeff Rosen's essay in the New York Times recounts the ongoing debate about whether Roe v. Wade is a "superprecedent." The term seems to come from a Michael Luttig opinion, referring to Roe as "super stare decisis," which I have argued here is part of a emerging conservative take on Roe. The basic idea is that mainstream conservatives will accept that Roe is settled (as modified by Casey) but begin to chip away at it over the long haul.

The expression "superprecedent" is highly misleading, because it confuses several different things we could say about a precedent, and several different reasons why the precedent is (or should be) safe from overruling.

(1) People have tried repeatedly to reverse a particular precedent, and fought long and hard for many years, but were rebuffed repeatedly in the courts. As a result, although they think the precedent is wrongly decided, they accept it as settled law. They will attempt in future litigation to get the courts to read it narrowly but will cease trying to overturn it directly. This conception of "superprecedent" means nothing more than that a struggle over constitutional meaning has been settled in practice. It does not mean that everyone accepts the precedent wholeheartedly or unreservedly or believes that it has "legs," i.e, that it should be extended generously to other situations or expanded in the same way as precedents that are clearly correct expressions of the Constitution. Former opponents will continue to read it narrowly; they will simply cease to demand that it be overruled.

(2) A particular precedent is foundational in the sense that an enormous amount of subsequent law depends on its correctness, and if the precedent were overturned, this would have significant ripple effects in doctrine. It would be very difficult to preserve large amounts of existing doctrine and this would be a bad thing. It would cause a great deal of useful law to be greatly unsettled, undermining a wide range of doctrines that people have relied on for many years. This conception of "superprecedent" analogizes the precedent to a support beam in a house that, even if not installed correctly or in the right place, cannot now be removed without seriously endangering the safety of the occupants. The first conception, by contrast, might apply to a precedent that isn't all that crucial to lots of other areas of doctrine, and that is one reason why its opponents are willing (grudingly, to be sure) to accept it as settled.

(3) A particular precedent which was at one point highly controversial is now widely agreed to be a foundational and highly admirable articulation of constitutional values. Someone who doubts its correctness has taken themselves out of the mainstream of constitutional discourse. This reflects the fact that constitutional arguments proceed over long periods of time and sometimes, on reflection, people at a later time come to form a consensus that a particular interpretation of the Constitution is correct, so correct, that it has come to form a basic postulate of constitutional argument. Because the precedent is important in precisely this way, few people hesitate to read it broadly (from their perspective). Indeed, people build large theoretical edifices around the precedent and fight continously over who best understands its true meaning and deepest ramifications. Indeed, people tend to think that an argument is bolstered if it is consistent with or builds on this precedent, and, conversely, they tend to think that their opponents' arguments are undermined or refuted if they cannot explain or justify the precedent.

These three possible cases describe precedents that Sanford Levinson and I would call part of the constitutional canon, but they are canonical for different reasons, because they are (1) the result of a protracted struggle where one side has given up (at least for the time being); (2) essential to preserving a large body of law that is (at least in part) valuable for other reasons; and (3) a widely recognized exemplar of how to do constitutional interpretation correctly, which helps define who is in the mainstream and who is "off the wall."

It should be obvious that Brown v. Board of Education falls into all three categories. Roe v. Wade, on the other hand, falls into at most categories (1) and (2), and probably only category (1). One way to see this (surely not the only way) is by thinking about Roe's place in contemporary legal culture. Ask yourself whether a substantial number of lawyers, judges, and legal academics continue to think Roe was wrongly decided, and whether a substantial number believe that it should be read as narrowly as possible in future cases. If the answer to these questions is yes, then Roe does not fall into category (3), as a paragon or exemplar of good constitutional interpretation. And if it can be read very narrowly without doing much harm to the fabric of the law, that suggests that probably does not fit into category (2) either. Lawrence v. Texas, for example, might be justified on the basis of Griswold and Eisenstadt, without using Roe or Casey.

If Roe falls into category (1), that is, it is a precedent whose opponents have tried for many years without success to overturn, and who therefore have conceded that it will not be overruled, there is very little that is "super" about it. Luttig's expression "super stare decisis" is a better (although also inaccurate) expression, because it focuses on the fact that stare decisis (and not Roe's underlying correctness) is at issue. (Luttig's expression is incorrect because it suggests that the rules of stare decisis are different for this precedent. They are not. The precedent is simply settled, like many other precedents where opponents have given up trying to get rid of it).

Of course, things can always change. For many years Roe was not even in category (1); lots of people were trying very hard to get rid of it, and many still are. In time, if Roe proves useful to large number of people both on the left and the right, Roe may eventually fall into category (2) or even (mirabile dictu) category (3). But that is not the case at present. Roe's status as a precedent is not currently the same as (for example) Brown's. Talking about Roe as a "superprecedent" obscures far more than it clarifies.

UPDATE: Kevin Drum thinks the real superprecedent is Griswold v. Connecticut:

If my understanding of Roe is correct, it's based on a generalized right of privacy as decided in Griswold v. Connecticut, which in turn was based on our current understanding of the doctrine of substantive due process. I suspect you can't overturn Roe without also substantially overturning Griswold and significantly weakening the modern application of substantive due process at the same time. Rosen mentions this, and it seems like it's really the key issue: not whether Roe is a superprecedent, but whether Griswold's interpretation of substantive due process is a superprecedent.

Two points. First, Kevin's argument shows why the concept of superprecedents is inherently confusing, and why the distinctions I've drawn help sharpen the analysis. If Griswold is a "superprecedent," it's a different type than Roe would be. Griswold like Brown, probably falls into categories (1) (2), and (possibly)(3). People have given up trying to overrule it. Lots of later cases depend on it. Finally, it's widely (but not universally!) accepted and admired as articulating important basic rights of all Americans. The key bit of evidence for this is the Bork hearings, which demonstrated widespread public support for Griswold; Bork's inability to justify the decision became a reason to think that he was out of the mainstream. Indeed, the Democrats' strategy in the Bork hearings was to get Bork on record as equivocating about Griswold (and Bolling v. Sharpe and constitutional protections for women) rather than focus on his opposition to Roe v. Wade. That strategy made the most sense if by 1987 Griswold had started to turn into a hallowed and foundational precedent akin to Brown.

My hesitation about whether Griswold really fits into category (3) is that there are several constitutional scholars I respect who still believe that Griswold is wrongly decided and should be read as narrowly as possible. So I suspect that Griswold does not have quite the same elevated status as Brown (or Marbury v. Madison or Gibbons v. Ogden), although it seems far closer to that status than Roe v. Wade.

Second, if one was determined to do so, nothing could be easier than jettisoning Roe and retaining Griswold, and doing so would not endanger the web of cases outside of the abortion area that depend on Griswold: Eisenstadt, Carey, and Lawrence. Roe is distinguishable because in the case of abortion, but not contraception or same sex sexual relations, the state has an interest in the protection of embryonic or fetal life that has come into being following conception. Sometimes it might be difficult to determine whether a particular method of contraception actually prevented conception or operated as an abortifacent, so courts would have to devise rules to guide states who wanted to ban abortion but not contraception, but this would not necessitate overruling Griswold.