Egregiously Egregious? A Dobbs-Inspired Follow-Up Regarding "Precedent on Precedent"
by Michael C. Dorf
My latest Verdict column criticizes the way in which Justice Alito's majority opinion in Dobbs uses selective quotations from pro-choice/liberal scholars and others to argue that Roe was wrongly decided. I pay special attention to the opinion's reliance on Justice Ginsburg's 1993 Madison Lecture and John Hart Ely's post-Roe article The Wages of Crying Wolf. I explain that while Justice Ginsburg did think Roe went too far too fast and would have been better grounded in equal protection, she thought it was rightly decided and should not be overruled. Ely, for his part, did think Roe was wrong as an original matter but reasonably well-informed observers--presumably including the majority Justices in Dobbs and the law clerks who worked on the case for them--would know that Ely also thought and said that the Casey Court was correct to reaffirm Roe. Moreover, I explain that we must place Ely's views about Roe must in the context of his broader argument that the central function of judicial review is to strengthen democracy--a function that the Roberts Court conservative super-majority is not only failing to carry out but is actively undermining.
One can criticize the Dobbs opinion in numerous other ways, of course, and my co-bloggers--especially Prof. Colb--have offered a large number of such critiques. In today's essay, I want to pivot from critique to analysis, to ask a question posed by the Dobbs majority's views with respect to stare decisis. I expect that my co-bloggers and I will return to critique sooner or later, but for now I'm going to focus on the stare decisis issue as an abstract matter. In particular, I want to address the following issue:
Suppose a court in Case P2 is asked to overrule Precedent P1. The court weighs all of the relevant factors and concludes that P1 was a close case but it was either rightly decided or, if wrong, not so wrong as to justify overruling, in light of the other stare decisis factors of workability, effect on other areas of law, and reliance. Accordingly P2 reaffirms P1. Some time passes and now the court is asked in P3 to overrule P1 and P2. Should the fact that P2 reaffirmed P1 provide any insulation against overruling? Or does the P3 court face exactly the same question was presented in P2?
Justice Alito's Dobbs opinion doesn't really grapple with that issue. It describes Roe as "egregiously wrong" and then explains why what I'll call the "non-merits" stare decisis factors (workability, effect on other areas of law, and reliance) don't call for retaining Roe, but it does so as if it's addressing the same question that the Casey Court addressed. However, while that might be the right approach to the non-merits stare decisis factors, it's insufficient with respect to the merits. It's clear that the Dobbs majority thinks Casey was wrong in deciding not to overrule Roe, but the Court does not provide a clear account of just how wrong Casey was.
That's an important omission if one concludes that the right answer in my abstract italicized scenario above is that the reaffirmation of P1 in P2 should make it harder for the P3 court to overrule P1 than if P2 had never arisen. But is that the correct approach? Let me suggest that there's no easy answer to that question.
Simple disagreement with a precedent is not a sufficient ground for overruling. Putting aside non-merits stare decisis factors, the standard for overruling is something more than on-balance wrong but probably something less than egregiously wrong, which seems like Justice Alito's hostility to abortion rather than an attempt to articulate a legal standard. For present purposes, it's not important that we settle on a precise verbal formula or attempt to quantify how wrong a decision must be to warrant overruling, nor to attend to the various qualitative dimensions of wrongness. Accordingly, I'll stipulate that to qualify for overruling, a precedent must be clearly wrong.
But there could be agreement that P1 was wrong, even if there is disagreement about how wrong it was. Suppose then, that a Court concludes 5-4 in P2 that while P1 was wrong, it was not clearly wrong. The four dissenters in P2 think that P1 was clearly wrong. A few years later P3 comes along. It is certainly a conceptual possibility to conclude that P1 was clearly wrong but that reasonable people could disagree and think P1 was merely wrong. Judges who take that view will think that they would have had grounds to join the dissenters in P2 but that they should not now overrule P2 because the P2 court itself was not clearly wrong in its judgment that P1 was merely wrong but not clearly wrong. Put differently, the P3 court could think that P1 standing alone warrants overruling but that P2--which reaffirmed P1--did not. Thus, the stare decisis effect of P2 provides extra, albeit indirect, protection for P1.
Justice Kavanaugh's Dobbs concurrence uses a hypothetical example to push back against that conclusion, but as I'll explain, his analysis is incomplete. He writes:
Suppose that in 1924 this Court had expressly reaffirmed Plessy v. Ferguson and upheld the States’ authority to segregate people on the basis of race. Would the Court in Brown some 30 years later in 1954 have reaffirmed Plessy and upheld racially segregated schools simply because of that intervening 1924 precedent? Surely the answer is no.
I agree with Justice Kavanaugh's conclusion here, but note that he has gilded the lily. The question is not whether the existence of P2 (the hypothetical 1924 case in Justice Kavanaugh's example) absolutely precludes overruling P1 (Plessy). Surely it does not. The question is whether the existence of P2 makes it any more difficult to overrule P1 in P3 (Brown) than it would have been in the absence of the intervening P2 reaffirmation. My suggestion is that any standard for overturning other than simple present disagreement, as applied to a decision to adhere to stare decisis, should at least somewhat strengthen the original precedent.
Justice Kavanaugh himself appeared to exploit that logic during his confirmation hearing when he referred to Casey as "precedent on precedent." As I wrote at the time, that was highly misleading. All that then-Judge Kavanaugh really meant was that Casey was a precedent about precedent, but he almost certainly hoped and expected that his audience would misinterpret him to be saying that Roe drew added precedential strength (via the process I've identified above) in virtue of having been reaffirmed in Casey. Justice Kavanaugh's Plessy-based hypothetical example nicely illustrates why the process of reaffirmation should not completely insulate a precedent from overruling, but it does not address the question of how much strength (short of infinite strength) a precedent gains by reaffirmation.
That said, we can tweak Justice Kavanaugh's example to make his point more effective. By hypothesis, each time a case is reaffirmed, it grows stronger. One might think that P1 is clearly wrong, and even that P2 was clearly wrong to reaffirm P1, but that it's a closer question whether P3 was clearly wrong (rather than merely wrong) to reaffirm P2, so that the court in P4 thinks both P1 and P2 were clearly wrong but that P3 was merely wrong, and so the P4 court gives stare decisis effect to P3 and thus indirectly to P1 and P2. And if P4 doesn't do the trick, there's always P5, P6, etc. So long as a precedent gains any strength from reaffirmation, one can convert the most egregiously wrong decision into a binding precedent by reconsidering it and the chain of cases reaffirming it sufficiently many times.
Does that mean that the working hypothesis is wrong? Should reaffirmation add no strength to the original decision? That conclusion also seems problematic and contrary to the very point of stare decisis, which is that courts sometimes should adhere to precedents with which they disagree. Presumably the point of taking a case (P2/Casey) to decide whether to continue to adhere to an earlier decision (P1/Roe) is actually to reach some sort of binding decision, which means that the reaffirming decision should be entitled to some weight beyond whatever weight the initial decision had. But then we get the oddity that in principle a very weak precedent can be strengthened by repeated reconsideration and reaffirmation.
I don't have a good theoretical solution to what strikes me as a genuine puzzle. That said, the issue is unlikely to arise for two reasons. First, it would take a court acting in incredibly bad faith to continually reconsider a precedent for the purpose of strengthening it, and in the absence of such bad faith, the reconsideration question will not likely arise repeatedly in a short span. Abortion is a useful illustration. In just about every abortion case in the 49 years from Roe to Dobbs, the state and/or its amici defending an abortion regulation asked the Court to overrule Roe, even if only as an alternative ground for a ruling. But before Dobbs, Casey was the only case within that span in which the Court gave plenary consideration to overruling Roe. Repeated reaffirmations based on full reconsideration per the stare decisis factors seem highly unlikely in practice.
Second, and more fundamentally, at the Supreme Court, Justices say they give weight to precedent, but what they do tells a different story. Summarizing a substantial body of empirical literature in a 2018 article in the Supreme Court Review, Professor Fred Schauer remarked "that in cases in which either the Court as a whole or individual Justices are inclined, precedent aside, to make a particular decision, the presence of an opposed precedent is rarely a barrier to reaching the precedent-independent outcome." Because stare decisis carries little to no weight to begin with, multiplying it repeatedly has no substantial impact.
Accordingly, I conclude that the puzzle I have discussed here is interesting chiefly as a theoretical matter. It might also have some relevance for how the Justices write future opinions in which they explain why they believe they are justified in overruling what remains of the constitutional order.