In a 1991 essay, then-Justice Sandra Day O’Connor referred to questions about whether female judges reasoned in a “different” voice as being “dangerous and unanswerable.” The basis for her concern is clear from the debate over Judge Sonia Sotamayor’s remark, in a now famous 2001 Berkeley speech, that “a wise Latina woman with the richness of her experiences would more often than not reach a better conclusion than a white male who hasn’t lived that life.”
Sotomayor’s statement has been interpreted by commentators from the right as evidence that she is “racist” (Newt Gingrich’s term) or incapable of unbiased decision-making. Those charges raise broader questions about the evidence on the effect of gender on judging. Do female judges in fact speak in a different voice? The evidence is mixed. Taken as a whole, the most systematic research suggests that gender matters in certain kinds of cases, in particular discrimination claims, which is one of the reasons why diversity should matter for selecting judges. But these cases are not a large part of the Supreme Court’s workload. Nor will statistical correlations predict the votes of any individual judge, which is why diversity is only one of the factors that should affect appointments.
A cottage industry of empirical work has tried to disentangle the influence of gender on judging. The studies have looked at a wide range of cases: sex discrimination, harassment, gay rights, or divorce—and also more general contested issues such as criminal procedure or obscenity. Results vary. The most recent comprehensive analysis is an unpublished paper by three political scientists, Christina Boyd, Lee Epstein, and Andrew Martin. By their calculation, about one-third of some 30 previous studies find differences in the votes of male and female judges or differences in the votes of a panel of judges when one member is a woman. About one-third find no such gender differences. The remaining third report gender differences in the votes either of individual judges or of judges with a woman on the panel—one or the other but not both.
Part of the reason for the variation in outcomes is that studies of this sort are extraordinarily difficult to do well, and they vary considerably in design and quality. Researchers differ in their choice of time periods, cases, courts, and ways of controlling for relevant variables. Even the most carefully conceived efforts to isolate the influence of traits like gender confront substantial obstacles. For obvious reasons, it is not possible to run controlled experiments involving actual cases. You can’t ask litigants to be part of a study that manipulates the gender composition of their appellate panel.
Nor is there any simple or uncontested way to control adequately for ideology, a key variable. If, as much social science research suggests, gender is part of the life experience that shapes people’s political ideology, then controlling for that factor may eliminate part of what is being studied. In our culture, women are socialized to be more caring and compassionate than men, and this experience can affect whether they are Democrats or Republicans, liberals or conservatives. Thus the influences of gender and of ideology on a judge’s voting patterns are intertwined.
There are also technical difficulties in determining how to assess ideology. Most studies use the party affiliation of the judge or of the president or governor who appoints the judge, but these are imperfect proxies and may not yield results as reliable as those of more complex measures. Boyd, Epstein, and Martin rely on scores that take into account additional information, including a judge’s record and the political affiliations of the senators from the judge’s state, whose support is key in the confirmation process. The differences among the studies in measuring ideology help explain why the results are inconsistent.
A related difficulty with many of the early studies of gender and judging involves distortions in the sample. On the federal bench, women skew to the political left, while men are more evenly distributed. This makes it problematic to use simple models of statistical analysis and samples of limited size. If, for example, researchers are looking at federal sex-discrimination cases during a particular period, there may be too few conservative women to know whether their presence on a panel would have had the same effect as that of a more liberal female judge. Boyd, Epstein, and Martin avoid this problem by matching pairs of federal employment-discrimination cases that are similar in relevant respects except for the sex of one member of the panel. Their study, the best one to date, finds that the probability of a judge ruling in favor of a discrimination plaintiff decreases by about 10 percent when the judge is a man. When a woman is on the panel, the likelihood that a male colleague will rule in favor of the plaintiff increases 12 percent to 16 percent.
This finding is consistent with two commonly cited earlier studies. One, from the 2005 Yale Law Journal, concluded that women on the appellate bench were significantly likelier than their male colleagues to perceive conduct as sexual harassment or discrimination, even when controlling for ideology and other background factors. Moreover, the presence of a woman on an appellate panel more than doubled the likelihood that a male judge would rule for the plaintiff in a sex harassment case, and tripled that likelihood for a sex-discrimination claim. A similar 1993 study in Judicature, which Sotomayor invoked in her speech, found that women on the federal appeals court, whether Democrats or Republicans, were more likely than their male colleagues to vote for plaintiffs in employment-discrimination cases.
Yet even if such research cumulatively suggests that the sex of a judge does influence the outcome of certain cases, sex is by no means a reliable predictor of the voting behavior of any particular nominee. Much depends on other aspects of a judge’s background and how they influence his or her world view. Consider Justices Thurgood Marshall and Clarence Thomas. One of Marshall’s greatest contributions to the court was his firsthand experience of racial injustice. As his colleague Justice Byron White recalled, Marshall “would tell us things that we knew but would rather forget; and he told us much that we did not know due to the limitations of our own experiences.” Clarence Thomas, too, experienced poverty and racism as a child, but he does not draw the same lessons from them. For similar reasons, simply putting another woman on the court would not necessarily be good for a feminist legal agenda. It matters greatly which woman.
In short, the importance of diversity in judicial appointments should be neither overlooked nor overstated. Equally important are the other qualities that Obama mentioned as criteria for his selection: a “sharp and independent mind,” a record of “excellence and integrity,” and a capacity for “empathy.” Concern for all of those qualities should guide the confirmation process.
This article also appears in Double X.