Elizabeth Warren, Hillary Clinton, and Democratic Voters' "Palmore Problem"

by Michael C. Dorf

Elizabeth Warren should be a formidable candidate for the Democratic nomination for president in 2020. As noted in a recent Fivethirtyeight.com article, she is ideologically left/center-left, which is just about smack in the middle of the Democratic primary electorate. Moreover, she has a message and policy track record that should make her an attractive general election candidate who could win back enough of the Rust Belt voters who went for Trump in 2016 to (re)turn Pennsylvania, Ohio, Michigan, and Wisconsin blue. Given her successful championing of the Consumer Financial Protection Bureau, Warren could run as a genuine progressive populist while accurately painting Trump's purported populism as a thin veneer over standard-issue Republican stroke-the-rich policies.

That is not to say that Warren is a perfect candidate. She lacks foreign policy depth (although she undoubtedly understands much more about the world than Trump does); she risks following in the footsteps of each of the last three major-party nominees from Massachusetts (Dukakis, Kerry, and Romney), who all lost after running lackluster campaigns that failed to make an emotional connection with voters; and she badly mishandled Trump's "Pocahontas" attack with her DNA test stunt.

But no candidate is perfect, and Warren's mix of pluses and minuses ought to put her at or near the top of the Democratic field--at least on paper. And yet, analysts are already downgrading her prospects on the ground that she is not sufficiently "likable." At first blush, the claim seems absurd. Donald Trump is one of the least likable people ever to hold public office. How can likability matter? "Not likable" appears to be code for "a powerful woman." A recent Politico story discusses comparisons of Warren and Hillary Clinton. It includes a photo of the two women side by side. The caption states: "Like former Secretary of State Hillary Clinton, gendered terms like 'shrill' or 'scoldy' are already ascribed to Sen. Elizabeth Warren, as people dismiss her as a viable 2020 contender."

And thus we come to a question for Democratic primary voters (or, more immediately, for Democratic activists trying to decide whom to support in the pre-primary process): If you otherwise would support Warren for the nomination but think that sexism would undercut her prospects in the general election, should you support a different candidate instead, or would doing so implicate you in the very sexism you decry in prospective general election voters?

I refer to the foregoing question as the "Palmore problem" in light of a unanimous 1984 SCOTUS case, Palmore v. Sidoti, in which the Court invalidated a trial court judge's custody decision on equal protection grounds. A white child's white mother remarried an African American man, prompting the child's father to petition for custody. The trial court granted the petition on the ground that the child would encounter prejudice living in a mixed-race household. The SCOTUS reversed that ruling even though there was no evidence that the trial judge shared the racial prejudice of the people who would discriminate against the child. Taking account of that prejudice implicitly and impermissibly gave it the force of law. CJ Burger wrote: "Private biases may be outside the reach of the law, but the law cannot, directly or indirectly, give them effect."

Assuming one thinks the Palmore principle is sound (and I do), should it apply in deciding whom to vote for?
In asking that question, of course I do not mean to ask whether the Palmore principle literally has force in the context of an individual's decision about supporting a primary candidate. It does not. But if the principle is sound, then in order for it not to apply in the private realm, there must be some reason why it has force in one context but not the other.

That could be true. For example, the Establishment Clause limits religious displays on public property (at least until the SCOTUS says it doesn't, possibly in a pending case). But not only doesn't the Establishment Clause apply to private actors; its logic doesn't apply either; thus, a person of faith can strongly believe that the government should not display religious symbols on public property even while displaying religious symbols on her own property.

However, the Palmore principle doesn't seem to turn on anything uniquely governmental. It is a principle about complicity. It says that if it is wrong for A to take account of some factor, then it is also wrong for B to take account of the fact that A takes account of that factor where B's decision ends up reinforcing A's prejudices. That principle, if sound, recommends itself to private actors as well.

What would application of the Palmore principle mean here? One might think that it wouldn't apply because the "likability" objections aren't sexist. Sure, this objection would go, some people who made the point about Clinton and now make it about Warren used gendered language, but at its root lie genuine personality or character issues.

That objection strikes me as wildly implausible. Of course Hillary Clinton said and did many questionable things that might lead particular people to dislike her. And there are undoubtedly some things that Warren has said and done (such as the DNA test episode) that dent her likability. But the evidence that sex plays a role is overwhelming. Consider the right's demonization of Nancy Pelosi, whose personal style is quite different from that of Clinton. As brilliantly explored in this satirical piece by Alexandra Petri, the fact that voters who claim not to be sexist find fault with just about every female candidate is telling.

Moreover, even if there were a female candidate whose electoral prospects would not be as burdened by sexism as one expects that Warren's may be--Amy Klobuchar or Kirsten Gillibrand, say--voting for her over Warren because of fear that Warren's prospects would be undercut by sexism still violates the Palmore principle. It would be like a firm defending against a charge of sex discrimination by saying that it would promote "the right kind of women," when, in the particular case, it denied a promotion to a woman based on sex-stereotyped views. That's still sex discrimination.

What to do?

If Warren is not your top candidate on the merits, then there is no problem. You're not violating the Palmore principle by supporting someone else. However, if, absent "likability" concerns that are code for sexism, Warren would be your top choice, should you support her in the primaries even though you think that her nomination would reduce the likelihood of defeating Trump? If you would otherwise support a different female candidate who, you project, would also do worse than a male candidate, should you stick with that female candidate to avoid violating the Palmore principle? Not necessarily.

Although the Palmore principle is sound, it is not absolute. Consider an analogy. Suppose you are a high school principal deciding which of two countries a class should visit as part of an educational language immersion program. Country A would be better than Country B in most respects, but there is a substantial risk to the students' safety in Country A due to racism among some of its inhabitants. Although it would violate the Palmore principle to choose Country B for that reason, you would nonetheless be justified in sending the class to Country B, because safety is a higher priority than strict compliance with the Palmore principle.

Consider Palmore itself. The trial judge thought that raising a child in a biracial family would subject her to "stigmatization." The SCOTUS did not think that diffuse harm was a sufficient reason for tacitly endorsing racial prejudice. Suppose, however, that there had been evidence that permitting the mother to retain custody would subject the child to a substantial risk of violence or death at the hands of prejudiced neighbors. Presumably at that point the Court would have asked whether awarding the father custody would be narrowly tailored to the government's interest in the child's safety, i.e., it would have applied strict judicial scrutiny rather than finding that the custody decision was per se invalid. And presumably there would be some cases in which a prima facie violation of the Palmore principle would withstand such strict scrutiny.

Strict judicial scrutiny is a tool for courts applying the Constitution, but we can find a rough analogy to it in individual decisions about voting (and other matters): the Palmore principle is valid but defeasible in the face of a good reason to override it. Defeating Donald Trump is an excellent such reason. Whether defeating a generic Republican presidential candidate would be a sufficiently good reason to violate the Palmore principle could be a somewhat harder question. One's answer would depend on the extent to which one thinks Trump's uniquely Trumpian qualities make him more awful than the policy awfulness of the GOP (in which Trump also partakes). I'll leave that question for another day. For now, it suffices to observe that, whatever one thinks of other Republicans, Trump is an existential threat to American democracy, world peace, and the habitability of planet Earth. Even a small diminution in the likelihood of defeating him in 2020 is too high a price to pay for compliance with the Palmore principle.