Unequal protection.

• Author: Robinson, Russell K.
• Position: IV. Diverging Votes through Conclusion, with footnotes and appendix, p. 198-233

4. Diverging Votes

   The analysis above suggests that Justice Kennedy's jurisprudence is skeptical of equal protection claims brought by women and people of color but generally supportive of sexual-orientation-based claims. This Part provides empirical evidence to support that argument. This is the first empirical study that systematically analyzes Justice Kennedy's votes regarding constitutional claims based on race, sex, and sexual orientation from the time Justice Kennedy joined the Court in 1988 to 2015. (286)

   A description of the study's methodology follows. In order to define the population of relevant cases, my research assistant and I collected all of the cases in which Justice Kennedy participated through the U.S. Supreme Court Database, a frequently cited source for research on the Court. (287) For the race category of cases, we relied on the Database's list of cases that included some element of race that is frequently litigated, for example, voting rights, school desegregation, employment discrimination, and affirmative action. For the sex category, we relied on those cases that ruled on abortion, sex discrimination, and employment discrimination based on sex. The Database lacks an LGBT category, which required conducting an independent Westlaw search of Supreme Court cases with the following parameters: "gay," "lesbian," "homosexual," "bisexual," "transgender," or "sexual orientation" in the synopsis of the Court's opinions. This search turned up eleven cases. (288) A list of cases in each category appears in the Appendix.

   Once satisfied with the set of cases, we tracked how many times Justice Kennedy cast a liberal or conservative vote in the race, sex, and sexual orientation cases. We adopted the ideological methodology of the Database, which defines as liberal those decisions that result in an outcome that is pro-affirmative action, prochoice in abortion, pro-gay rights, and pro-plaintiff in a discrimination case. The cases deemed conservative are the opposite. The Database's ideological coding was generally accurate. The central concern of this project is whether Justice Kennedy voted to protect the rights of people of color, women, and LGBT people, and in a handful of cases, the Database's liberal/conservative categorization was not a good proxy for this question. (289) Accordingly, in four cases, we overrode the Database's coding to fit with this study's narrow focus. (290)

   To demonstrate differences injustice Kennedy's voting record in race, sex, and LGBT rights contexts, we calculated the rate at which Justice Kennedy cast liberal votes in each context. We further divided the data to see how often Justice Kennedy cast a liberal vote in closer cases, such as those where the Court voted nonunanimously or where the majority could summon only the five-vote minimum.

   In focusing on constitutional cases, (291) we found a clear divergence between race and sex cases, on the one hand, and sexual orientation cases on the other. As depicted in the figures below, this divergence is strongest in the most contested and divisive--and likely the most important--cases. In nonunanimous cases, Justice Kennedy cast a liberal vote in 33% of race cases and 15% of sex cases. In other words, in the majority of the constitutional cases involving issues of race, and the vast majority of those involving sex equality, Justice Kennedy cast conservative votes. In sexual orientation cases, by contrast, Justice Kennedy cast liberal votes 83% of the time. Focusing on a subset of nonunanimous cases, those decided by a 5-4 vote, we see even starker disparities. In this category, Justice Kennedy cast just one liberal vote out of twelve votes in the race cases (8.3%) and none in the sex cases. By contrast, in the sexual orientation cases Justice Kennedy cast liberal votes in 75% of the cases.

   These findings are consistent with my effort to disrupt the conventional explanation of equal protection tiers in which race triggers the most demanding level of judicial protection, sex receives a little less protection than race, and sexual orientation receives less protection than race and sex because it warrants, at most, "rational basis with bite." (292) The key divide in equal protection law is not between those groups deemed suspect or quasi-suspect (i.e., race and sex) and those denied such status (i.e., sexual orientation and disability). (293) Rather, we might consider a revised, three-tiered model: one tier uniting race and sex (in most respects), a distinct tier for sexual orientation, and a minimum rationality track for everything else. Even this description does not suffice in that it suggests a hierarchy in which race and sex are on top. (294) As suggested by the empirical study, sexual orientation seems to be emerging as a site of greater protection than is currently available to people of color and women.

   These findings are subject to limitations, of course. Our measure of pro-LGBT/pro-people of color/pro-female is necessarily crude. Some may contest whether a particular victory for an LGBT (or black or female) litigant is truly in the best interest of that group. (295) Further, the number of sexual orientation cases is admittedly small. (296) Some readers might wonder whether the disparity that I identify is statistically significant. This question is important when a study utilizes a sample of a population. In this study, however, we captured the entire population of race, sex, and sexual orientation opinions. Thus, we did not construct a sample and assume that it represents the population. Nonetheless, my claim is restricted to cases that the Court decided through Summer 2015 (October Term 2014). I do not argue that in the long term the Court will necessarily continue to privilege sexual orientation over race and sex.

   There are at least two factors that will determine whether the pattern that I identify persists. First, the composition of the Court will eventually change. Justice Kennedy is seventy-nine years old. If he leaves the Court and the President replaces him with a more liberal Justice, the new Court majority may bring the race and sex cases in line with the sexual orientation precedents by "leveling up." For example, the Court might overrule Geduldig. (297) Further, the new majority could revise or supplement the rationales in Justice Kennedy's sexual orientation opinions. If Justice Kennedy leaves the Court during a Republican administration, the President might very well replace him with a Justice who is less empathetic for LGBT people. In that event, the new Court majority might refuse to extend cases such as Obergefell or recast their rationales ("leveling down"), much as Chief Justice Rehnquist (with Justice Kennedy's vote) rewrote Cleburne (298) Justice Kennedy's failure to acknowledge the unusual nature of his doctrinal moves in the sexual orientation cases and to name and situate that level of scrutiny may make them particularly vulnerable to rewriting. (299) Second, over time litigants will bring different claims, and this may influence the trend. Much depends on the content of the asserted rights that emerge from the courts of appeals and whether the Court votes to hear those cases. The next Part sheds some light on the reasons for uncertainty going forward by examining the intersection of sex and sexual orientation. I argue that Justice Kennedy's troubling rationales in sex cases could resurface in future sexual orientation cases. This is most likely to happen in cases involving transgender and bisexual people, because they are more likely to force judges to contend with the gender binary than gay and lesbian plaintiffs.

5. The Limits of LGBT Exceptionalism

   I have just told a story in which LGBT people are evading traditional equal protection roadblocks and quietly remaking constitutional law. However, I now provide a more pessimistic potential trajectory for LGBT rights. This Part laments what Obergefell (300) and Windsor (301) omit and shows how it could be turned against LGBT rights. The gains of today might evanesce tomorrow.

   1. The Inevitable Intersection of Sex and Sexual Orientation

      In Windsor, Justice Kennedy declined to acknowledge two principal reasons why many people oppose same-sex marriage--religion and a commitment to preserving gender roles. Obergefell says more about religion and gender, but in a rather cursory and perhaps evasive fashion. (302) In Windsor and Obergefell's equality analysis, Justice Kennedy failed to subject any state interest to sustained analysis. (303) This method follows and amplifies his approach in Romer. In Romer, Justice Kennedy at least mentioned two of the state interests in passing and then dismissed them as too narrow to support such a sweeping statute. (304) In Windsor, he offered even less legal analysis, preferring to repeat conclusory pronouncements. For the most part, Justice Kennedy refused even to name the state interests. (305) It is as if he was so offended by DOMA that he declined to dignify the asserted state interests by discussing them.

      Justice Kennedy's limited references to gender and religion occlude a critique of the reasons why many Americans continue to oppose same-sex marriage. To be clear, these rationales are rarely freestanding; religious and gendered defenses of traditional marriage are typically interwoven with pure homophobia, such as the notion that sex between men is disgusting. (306) Still, it is important to address these grounds because they may strike people as both distinct from rank homophobia and more defensible. (307) In Windsor's aftermath, religious liberty became the rallying cry for opponents of same-sex marriage, including those seeking exemptions from generally applicable laws. Obergefell notes this dynamic and expresses respect for religious people, yet fails to provide any discernable principle for resolving disputes between religious liberty and LGBT rights.

      Justice Kennedy's omission of gender is...