Settled Law, Precedent on Precedent, and Abortion: What We Learned and Didn't Learn About Judge Kavanaugh's Views on Abortion

by Michael C. Dorf

Let's begin with the obvious: The Republican-controlled Senate is extremely likely to confirm Judge Kavanaugh to the SCOTUS. Disruptions of the proceedings by protesters and complaints about the process by Senate Democrats have failed if their aim was to block his confirmation. However, that cannot have been their sole aim. Cynics will say that those Senators on the Judiciary Committee considering presidential runs have used the hearings to demonstrate their commitment to the resistance. My own view is that it is perfectly legitimate to engage in constitutional politics in order to build solidarity to fight and win another day. In that respect, I regard the strategy of Senate Democrats and the more sophisticated of the protesters as much akin to impact litigation brought in the expectation of losing a particular case but serving a larger organizing purpose.

That said, the Kavanaugh hearing has nonetheless been instructive in a number of respects. In today's essay I'll focus on abortion. My bottom line will be that despite Judge Kavanaugh's superficial efforts to conceal his views, his position is pretty obvious to anyone with any sophistication.

A document first leaked to the NY Times and then also included in a batch of material that NJ Senator Cory Booker released to the public yesterday casts doubt on Judge Kavanaugh's veracity when he said that he believes Roe v. Wade is "settled law." That declaration, even taken at face value, should never have provided any reassurance, because, as Garrett Epps observed in The Atlantic, "settled law" is not some term of art; it apparently means no more than a precedent that the Court has not yet chosen to overrule.

Nonetheless, the newly public material suggests prevarication, or at least extreme slipperiness. In 2003, Kavanaugh raised an objection to the use of the term "settled law" to refer to Roe in a draft op-ed that another Bush staffer had ghostwritten to help support the confirmation of Priscilla Owen, whose nomination to the US Court of Appeals for the Fifth Circuit was then languishing. Kavanaugh wrote:
I am not sure that all legal scholars refer to Roe as the settled law of the land at the Supreme Court level since [the] Court can always overrule its precedent, and three current Justices on the Court would do so.
Admittedly, there are ways that one can read the foregoing that are consistent with Judge Kavanaugh's recent claims that he thinks Roe is settled law. Maybe he meant in 2003 that not "all legal scholars" would call Roe settled law but Kavanaugh would call it that? That's quite a stretch. It's pretty clear from the context that 2003 Kavanaugh thought that those legal scholars who would not regard Roe as settled law would be right, given the real possibility of overruling.

Okay, but then maybe what Kavanaugh was saying in 2003 was that "settled law" means case law that is not realistically subject to being overruled and that while he thought Roe was realistically subject to being overruled in 2003 it no longer is, so that it has since become settled? That seems wholly unrealistic too.

For one thing, Kavanaugh in 2003 said that the Supreme Court can always overrule its own precedents. In the next sentence of the leaked memo, he argued that it would be better to emphasize that as a lower court judge, Owen would not be in a position to overrule Roe. That makes sense of his qualification "at the Supreme Court level" in the language from the memo I block-quoted above: Kavanaugh appeared to think in 2003 that it was either never appropriate to call SCOTUS precedent "settled law," because the Court can overrule any precedent at any time, or he thought that it's at least not appropriate to call a SCOTUS decision a matter of settled law where some sitting justices would be inclined to overrule it.

And of course, it looks like some sitting justices would still be inclined to overrule Roe. In a 2016 case, Justice Thomas expressly said that he would overrule it if given the chance. True, CJ Roberts and Justice Alito did not join in Justice Thomas's dissent in that case, but they also dissented from the majority opinion. Neither of them has ever voted to invalidate any restriction on abortion. There is good reason to think that, in addition to Justice Thomas, any or all of CJ Roberts and Justices Alito and Gorsuch would overrule Roe if presented with the opportunity. Thus, applying the standard that Kavanaugh applied in 2003, Roe is not settled law.

Accordingly, I think the most generous way to read Kavanaugh is as having been misleading albeit not downright dishonest in saying he now sees Roe as "settled law." In his own mind in 2003, "settled law" meant law that is not at all likely to be overruled; he had reason to think that Maine Senator Susan Collins would take him to be saying that about Roe now when he called it settled law; and yet, even though he didn't like the use of the term "settled law" to mean law that is only settled until the Supreme Court unsettles it, that was the definition he secretly had in mind when he used the term to describe Roe to Senator Collins. If so, I guess that's not technically a lie, but it's hardly the whole truth.

Meanwhile, at various points in Judge Kavanaugh's testimony, he referred to Planned Parenthood v. Casey--the 1992 SCOTUS case that reaffirmed what the lead opinion called the "central holding" of Roe--as not just a precedent entitled to the respect due any other precedent, but a "precedent on precedent." Like the two different definitions of settled law, this was another piece of confirmation legerdemain.

Calling a case a "precedent on precedent" makes it sound like a kind of double precedent with double the strength of an ordinary precedent. But it isn't. We can contrast a precedent on precedent with how Pennsylvania Senator Arlen Specter referred to Roe in the confirmation hearing of CJ Roberts. Specter called Roe a "super precedent" or even a "super-duper precedent." The super/super-duper theory is not entirely clear, but I take it to have a kind of Nietzschean what-doesn't-kill-me-makes-me-stronger quality: If the Court has considered whether to overrule a precedent and decided against doing so, then to overrule it in a third case requires not only that the Court have been wrong enough in the first case to warrant overruling, but so wrong in its judgment not to overrule in the second case as to warrant overruling in the third. By this theory, each time a case is considered for overruling but survives, it becomes harder to overrule, because its precedential status goes from ordinary to super to super-duper to super-duper-duper, etc.

Whatever the merits of Specter's Nietzschean theory, when Judge Kavanaugh refers to Casey as a "precedent on precedent," he may mislead some listeners into thinking he is saying that he believes the argument needed to overrule Casey and thus Roe's central holding is much stronger than would be needed to overrule any other precedent, but in fact that's almost certainly not what he means. What he actually appears to mean is simply that Casey is both a precedent on abortion and a precedent on precedent. It's a precedent on precedent, because in most cases in which the Court is asked to overrule a precedent, it quickly cites the standard for overruling and then either overrules or does not; by contrast, Casey contains an extensive discussion of when the Court should and should not overrule its past precedents. It is thus a "precedent on precedent."

But here's the thing. Judge Kavanaugh did not say, nor is there any reason to think he believes, that a precedent on precedent should be harder to overrule than any other kind of precedent. I can illustrate by asking the reader to suppose that Casey were not a precedent on precedent. Suppose instead that in some other case--let's call it X v. Y, the Court had included a substantial discussion of when to overrule its precedents and then decided that some old case, call it A v. B, shouldn't be overruled. Then imagine that the Court had been asked to overturn Roe in a case called Shmasey. Suppose that applying the principles laid down in X v. Y regarding overruling precedent, the Court in Shmasey had concluded that Roe should not be overturned. In this hypothetical world, Shmasey would not be a "precedent on precedent." It would just be a precedent on abortion. And yet Shmasey would be just as easy or difficult to overrule as Casey. To overrule Shmasey, the Court would have to either say that the Shmasey Court misapplied X v. Y's rules for when to overrule, or the Court would have to first say that X v. Y misstated the standard for overruling precedents, so that X v. Y is overruled, and then say that applying the correct standard for overruling precedents results in overruling Roe and Shmasey. The fact that the actual precedent on precedent happens to be part of the same case as the precedent that reaffirms Roe is interesting, I suppose, but under the logic that I take Judge Kavanaugh to be invoking, that does not make Casey or Roe harder to overrule than any other precedent. "Precedent on precedent" is misdirection.

What, then, are Judge Kavanaugh's actual views about abortion? There's no real secret here. I am willing to stipulate that Judge Kavanaugh does not have a concrete plan to overrule Roe and that he would, if faced with the question, sincerely try to consider all the arguments. I also think that given his values and jurisprudential views, it is extraordinarily unlikely that, after doing so, he would not end up either voting to overturn Roe outright or cutting it back in a way that makes it effectively a dead letter.

In a speech last year, Kavanaugh described William Rehnquist as his first "judicial hero" based on his opinions and dissents, presumably including his Roe dissent, although the speech does not literally contain words such as "Rehnquist was right in his Roe dissent and therefore the case should be overruled." But Kavanaugh did say something almost as revealing. He was discussing Rehnquist's opinion for the Court in Washington v. Glucksberg, which rejected a putative right of a terminally ill patient to control the timing of his or her death--pejoratively described as physician-assisted suicide. In that opinion, Rehnquist described the formula for determining whether an unenumerated right receives special judicial protection in extraordinarily narrow and history-focused terms. Kavanaugh said this:
Of course, even a first-year law student could tell you that the Glucksberg[] approach to unenumerated rights was not consistent with the approach of the abortion cases such as Roe vs. Wade in 1973, as well as the 1992 decision reaffirming Roe, known as Planned Parenthood vs. Casey. What to make of that? In this context, it’s fair to say that Justice Rehnquist was not successful in convincing a majority of the justices in the context of abortion either on Roe itself or in the later cases such as Casey, in the latter case perhaps because of stare decisis. But he was successful in stemming the general tide of free w[hee]ling judicial creation of unenumerated rights that were not rooted in the nation’s history and tradition. The Glucksberg case stands to this day as an important precedent, limiting the Court’s role in the realm of social policy and helping to ensure that the Court operates more as a court of law and less as an institution of social policy.
Notice that in this speech Judge Kavanaugh makes clear that he thinks the Glucksberg approach is correct and that it's obviously inconsistent with the approach the Court took to finding an abortion right, not only in Roe but also in Casey. In other words, he thinks that Roe and Casey are wrong. Now maybe one might think that Kavanaugh didn't say the bottom line in Casey was wrong, just that its approach to substantive due process was wrong, but that perhaps it could be justified by stare decisis. Maybe, but I'm highly dubious.

Why? Because I don't trust Judge Kavanaugh as a reader of precedent. He said just a year ago--and in response to questioning by South Carolina Senator Lindsey Graham he repeated just yesterday--that Glucksberg provides the current test for whether an unenumerated right receives constitutional protection. Yet that is an extremely contentious view that Kavanaugh simply states as a matter of fact.

Consider that just six years after Glucksberg, the Court in Lawrence v. Texas did not appear to apply the very narrow historical approach to substantive due process that Rehnquist championed in Glucksberg. Justice Scalia complained in dissent in Lawrence that Justice Kennedy's majority opinion, which did not cite Glucksberg, failed to apply its formula. And Scalia was not alone in that observation. Yale Kamisar, writing in the Michigan Law Review in 2008, pointed out the large methodological distance between Glucksberg and Lawrence.

Over time, Glucksberg has come to be the outlier. Thus, in 2015, writing for the Court in Obergefell v. Hodges, Justice Kennedy dispensed with Glucksberg this way:
The limitation of marriage to opposite-sex couples may long have seemed natural and just, but its inconsistency with the central meaning of the fundamental right to marry is now manifest. With that knowledge must come the recognition that laws excluding same-sex couples from the marriage right impose stigma and injury of the kind prohibited by our basic charter. Objecting that this does not reflect an appropriate framing of the issue, the respondents refer to Glucksberg, which called for a “ ‘careful description’ ” of fundamental rights. They assert the petitioners do not seek to exercise the right to marry but rather a new and nonexistent “right to same-sex marriage.” Glucksberg did insist that liberty under the Due Process Clause must be defined in a most circumscribed manner, with central reference to specific historical practices. Yet while that approach may have been appropriate for the asserted right there involved (physician-assisted suicide), it is inconsistent with the approach this Court has used in discussing other fundamental rights, including marriage and intimacy. Loving did not ask about a “right to interracial marriage”; Turner did not ask about a “right of inmates to marry”; and Zablocki did not ask about a “right of fathers with unpaid child support duties to marry.” Rather, each case inquired about the right to marry in its comprehensive sense, asking if there was a sufficient justification for excluding the relevant class from the right.
That passage essentially rejecting the Glucksberg methodology (though not the result of the case) appeared in a landmark decision of the Supreme Court written by the Justice that Kavanaugh would replace more than two years before Kavanaugh's speech praising the Glucksberg methodology. By reaffirming his support for the discredited Glucksberg methodology in his testimony yesterday, and going further to indicate that he thinks that discredited methodology remains the law today, Judge Kavanaugh signaled unmistakable disdain for the opinions that most strongly reject the Glucksberg methodology--those that protect a broad right to marriage and, as he himself emphasized in his speech last year, those that protect a right to abortion.

Although he may not have consciously made up his mind on the matter, as Judge Kavanaugh himself might say, "even a first-year law student could tell you" how he will very likely vote on the question whether to overrule Roe and Casey. This wolf comes as a wolf.