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 The most recent comprehensive analysis is an unpublished paper by three political scientists, Christina Boyd, Lee Epstein, and Andrew Martin.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.