Female Judges Evaluate Law Differently

With the retirement of Souter from the SCOTUS the search for the replacement is on. I have proposed in the past that the replacement should be a female minority in order to get a more diverse life experience onto the court. A life experience that is representative of a great swath of the country. The idea that the next justice be female is typically met with scorn and hostility based on the premise that "we should simply get the most qualified person" or the "smartest" or the "best" or by claims that it shouldn't or doesnt matter whether a woman is evaluating the law. I want to address that topic and assert that it does matter and that there is evidence to support a material distinction between female and male justices.

There is a difference between female and male justices as distinct classes. In a groundbreaking, and award winning, paper (pdf) "Untangling the Causal Effects of Sex on Judging," Washington University's Christina L. Boyd and Andrew D. Martin and Northwestern School of Law's Lee Epstein use a statistical analysis of decisions in various categories of cases to arrive at the conclusion that,

The likelihood of a judge deciding in favor of the party alleging discrimination decreases by about 10 percentage points when the judge is a male. Likewise, we find that men are significantly more likely to rule in favor of the rights litigant when a woman serves on the panel. Both effects are so persistent and consistent that they may come as a surprise even to those scholars who have long posited the existence of gendered judging.

i dont want to say that this is the only study on the subject or that every study comes out the same. The Boyd et al paper makes that clear in a brief survey of the prior research in the field,

On the other hand, the resulting research findings are so mixed that they practically defy characterization. By our count, social scientists and legal academics have produced nearly 30 systematic,multivariate analysis of the extent to which female judges make decisions distinct from their male colleagues (individual effects) or cause male judges to behave differently than they otherwise would(panel effects). Of those, roughly one-third purport to demonstrate clear panel or individual effects, a third report mixed results, and the final third found no sex-based differences whatsoever.

Because some of the existing studies examine areas of the law for which even Justice Wilson and others in her theoretical camp would have difficulty sustaining claims of gender difference(e.g., disputes involving the Internal Revenue Service), uneven results are not surprising. But even those investigations focusing on areas where a link between sex and judging seems quite plausible (e.g., sex discrimination) are notable for their incongruous findings. Three recent examples serve to make the point. In her analysis of statutory sex discrimination disputes in the U.S. circuits between 1999-2001, Peresie (2005) reports non-trivial sex eects at both the individual and panel level: female judges were more likely to nd for the plaintiff, as were panels that included a female. Westergren's (2004) examination of a similar set of cases, however, reveals neither individual nor panel effects based on gender. Crowe's (1999) study of the same courts splits the difference. Like Peresie, she finds that female judges were more favorable toward plaintiffs in sex-discrimination suits; and like Westergren, she unearths no evidence that the presence of a female judge affected the decision making of males on the panel.

After reading the Boyd paper i find their argument about the methodological short comings of regression analysis in these situations compelling enough that i favor their propensity score matching methodology.

They deal primarily in sex discrimination cases. This is an important point because it is where a common sense observer would say the biggest difference would be if females are deciding cases differently than men. The difference in probability that a woman alleging sex discrimination will receive the vote of a female justice instead of a male justice is twice as great at the highest degree of liberalism.

Observe that the estimated probability of a female judge casting a pro-plaintiff is close to 0.40 at the highest levels of liberalism;for even the most left-of-center male, that figure just barely exceeds 0.20.

This is the individual effect finding. This finding is signifigant because it shows how even the most liberal, enlightened of men is less likely to reach the same conclusion as a liberal woman. Part of the argument against using female as a selection criteria for the next vacancy has been that a man is just as capable of evaluating the case as a woman. That may still be true, however, the two genders appear to reach different conclusions in these types of cases. The panel effect, the effect of having a woman on a panel versus an all male panel is also striking.

If our descriptive analysis of the differences between male and female judges heartens proponents of gender-difference theories, our assessment of panel effects brings even more encouraging news. As we show in Figure 6, for not one model does the 95% confidence interval come near the zero line(indicating no difference between all-male and mixed-sex panels). Rather, we observe large causal effects, ranging from 0.12 to 0.16 meaning that the likelihood of a male judge ruling in favor of the plaintiff increases by 12% to 16% when a female sits on the panel.

On its face, this causal effect of panel composition is quite substantial, perhaps surprisingly so. Think about it this way. Because panels with female judges are significantly more plaintiff friendly than all-male panels, defendants should be more likely to settle after they observe assignment to a mixed-sex panel. To the extent that this form of selection bias exists, it ought mitigate against a finding of a strong causal panel effect. As a result, our findings, however substantial, may actually underestimate the impact of panel composition on outcomes.

According to the Boyd analysis in cases where discrimination is alleged having a woman on the panel makes it much more likely that the allegations will be sustained. This is a causal relationship meaning that the change in ruling comes from the change in gender. Without a woman on the panel the plaintiff would have lost in more than 1 in 10 instances.

In many ways the findings of the study affirm the previous theories about the way women will treat cases of gender discrimination and possibly discrimination in general. Justice Jane Matthews of the Federal Court of Australia had this to say (Doc) in response to earlier findings in the same vein,

To me this makes a great deal of sense. We are, all of us, the product of our life experiences. It is now openly acknowledged that judges, particularly appellate judges, are in the daily process of making the law. In doing so, they clearly impose their own values which are in turn molded by their life experiences. In general, women’s life experiences are very different from that of men. As a result, they are probably more sensitive to gender issues, and perhaps also to other aspects of the human condition.

The proposition that women do find differently from their male counterparts in gender related cases is borne out by a number of United States studies, which have shown no difference between the outcomes of cases of cases heard by men and women in economic rights or criminal cases, but a significant difference in discrimination case, with women being more sensitive to the discrimination issues involved. Whether this necessarily leads to better decision making might be a matter of debate. But I do not think that anyone would dispute that he public perception of the judiciary would be enhanced if it were seen to be more representative of the community at large.

If you are willing to accept that there is a difference between the way men and women adjudicate cases, if only in cases with alleged discrimination, the next question is whether or not we want this difference to be affirmed as good. It is possible that the women who tend to rule in favor of plaintiffs in discrimination cases go too far. for example we might assume that the 40% likelihood figure for the most liberal women finding in favor of a plaintiff is too high and constitutes discrimination in favor of the plaintiff. At the same time it is also possible that the 20% figure for the male justice it too low. This would still indicate that a woman is an important component in adjudicating these decisions.

Of course not all women are the same and not every woman is going to have such a high individual factor. However, if the mere presence of a woman is such an important factor in these cases then it might indicate that having women on the supreme court makes a very big difference in the out come of cases involving discrimination. Again whether that outcome is desirable is up for debate.