Trailblazers and Those That Followed: Personal Experiences, Gender, and Judicial Empathy

Abstract

This article investigates one causal mechanism that may explain why female judges on the federal appellate courts are more likely than men to side with plaintiffs in sex discrimination cases. To test whether personal experiences with inequality are related to empathetic responses to the claims of female plaintiffs, we focus on the first wave of female judges, who attended law school during a time of severe gender inequality. We find that female judges are more likely than their male colleagues to support plaintiffs in sex discrimination cases, but that this difference is seen only in judges who graduated law school between 1954 and 1975 and disappears when more recent law school cohorts of men and women judges are compared. These results suggest that the effect of gender as a trait is tied to the role of formative experiences with discrimination.

You know, the young women today can't possibly … understand the pressures of being first … Nothing was good enough, and it took me an awful lot of years to realize how good many of the women really were in relationship to the men's talents. I mean, when I think of it, men that were hundreds of places below us in class were getting great jobs and there were no jobs for us. -Judge Ilana Rovner ¹

Every time I wanted to do something, I had to invent a way to do it because there was no path for me. -Judge Shirley Hufstedler ²

When U.S. Courts of Appeals’ judges Ilana Rovner and Shirley Hufstedler graduated from law school in 1966 and 1949, respectively, nationwide, women comprised less than 5 percent of all law students (www.americanbar.org). In contrast, in 2011, women made up 46 percent of all students enrolled in law school. These judges’ comments about the obstacles they faced as trailblazers when entering the legal profession are quite similar to the recollections of former U.S. Supreme Court justices Sandra Day O'Connor and Ruth Bader Ginsburg about overtly discriminatory treatment in law school and the legal profession in the 1950s and 1960s (Ginsburg and O'Connor 2010). Oral histories of the first women appointed to the lower federal judiciary repeatedly touch upon the themes of blatant discrimination against women by law professors, fellow law students, and employers. When this trailblazer generation of female judges joined formerly all-male courts, many of them also came face-to-face with exclusionary traditions and practices in their new circuits (Haire and Moyer 2015). For instance, when Florence Allen joined the Sixth Circuit as the first woman, she dined alone at lunchtime because her male colleagues frequently lunched at private clubs that did not admit women.³ Given how widespread discriminatory treatment was for early female appointees, did their personal experiences with discrimination affect how these judges confronted the issue of sex discrimination in their cases?

Recent research on the U.S. Courts of Appeals has established that female judges are more likely than men to side with plaintiffs in sex discrimination cases (Boyd et al. 2010; Peresie 2005) and other types of employment discrimination cases (Moyer and Tankersley 2012; Songer et al. 1994).⁴ Scholars have speculated that women's personal experiences with sex discrimination (and men's lack of such experiences) could be driving this effect (Martin et al. 2002). For instance, in a survey of Carter's female appointees, many of whom attended law schools during periods marked by stark gender inequality, 81 percent described some form of sex discrimination as the primary challenge facing a woman or man in law (Martin 1990: 207). How might these women judges’ personal experiences affect their perspective on cases dealing with sex discrimination? Building on research by others who have examined men's and women's attitudes toward gender equality, we test the premise that the propensity to support sex discrimination plaintiffs should be highest among the group of “trailblazer” women who entered the legal profession during a time when overt discrimination against women was severe and prevalent. We posit that this increased likelihood to side with sex discrimination plaintiffs is not attributable merely to being a woman, per se, but rather about the experiences that shaped this particular class of women defined by when they attended law school. As such, our argument fits into a large body of work by sociologists that has focused on the formative experiences of generations and cohorts (Mannheim 1952 1952; Kertzer 1983; Schuman and Scott 1989).

The linkage between traits, empathy, and experiences is an important one to understand because this connection has been emphasized by those involved in the selection of those who will sit on the federal bench. When President Obama referenced the term “empathy” in his nomination of Sonia Sotomayor to the Supreme Court, it was in the context that an individual's personal experiences would allow them to better understand the plight of those bringing their claims to court (Weisman 2009). Indeed, recent scholarship has uncovered other ways that personal experiences can affect judging (e.g., Haire and Moyer 2015). For instance, one study found that conservative male judges became more supportive of plaintiffs in sex discrimination cases after having daughters, leading the authors to conclude that empathy could have a crosscutting effect with respect to ideology (Glynn and Sen 2015).

Below, we review the state of the literature on attitudes toward and perceptions of sex discrimination, and how this research might be applicable to the study of federal judges. We then set forth our theoretical account that links psychological research to explain how the personal experiences of judges lead them to draw different inferences from case fact patterns involving claims of sex discrimination. We test our hypotheses using a sample of sex discrimination cases from the U.S. Courts of Appeals (1995–2008) and find evidence that the earliest wave of female judges who confronted overt discrimination in law school and the legal profession were more likely to side with plaintiffs in cases of sex discrimination. We conclude by examining the implications of our finding for accounts of judging more generally and for the judicial selection process.

Attitudes About Gender Inequality and Discrimination

To understand how a judge's personal experiences could translate into greater judicial support for sex discrimination plaintiffs, we must first know something about men and women's attitudes toward gender inequality and discrimination. Overall, researchers are in agreement that, on most issues, attitudes about women in the workplace and roles within the family have liberalized over time (Bolzendahl and Myers 2004; Carter et al. 2009; Simon and Landis 1989). For instance, surveys from the early 1970s showed that a majority of all respondents rated the women's movement unfavorably, but that by 1974, a majority rated the movement favorably (Huddy et al. 2000).⁵ Researchers attribute this liberalization to a combination of individual attitude change and population replacement (Brewster and Padavic 2000; Ciabattari 2001). In addition to period effects, the literature also reports small or no differences at all in gender role beliefs when men and women are compared over time (Bolzendahl and Myers 2004; Crosby 1982).⁶

However, there is some indication that men and women in the mass public differ in their perceptions of sex discrimination specifically. In a comparison of surveys from 1975 to 1987, Simon and Landis (1989) report that the percentage of female respondents who perceived discrimination against women increased by 10 points, while the percentage of men with this view actually dropped slightly.⁷ In more recent surveys, women express significantly higher levels of support for gender-based affirmative action than men, but attitudes about the existence of gender discrimination are also an important predictor (Kane and Whipkey 2009). Brewster and Padavic (2000) compare 1977 and 1996 GSS data, finding that men are significantly more conservative than women in their attitudes about working women in both time periods, and also that men's attitudes have been slower to change than women's (despite a liberalizing trend over time).

Employed women in particular appear to be more aware of sex discrimination than other groups (Negowetti 2014; Simon and Landis 1989). Surveys of lawyers (Coontz 1995; Epstein 2004) and doctors (Carr et al. 2000) also reveal this difference in awareness of gender bias as well, with women seeing both discrimination and sexual harassment in employment contexts more often than men.

Experimental research by social psychologists also supports the general conclusion from survey data that women perceive sex discrimination and sexual harassment more often than men (O'Connor et al. 2004; Rotundo, Nguyen, and Sackett 2001; Wiener et al. 1995; Wiener et al. 1997). One study speculates that “women are more likely to find evidence of harassment because they are more sensitive to sexual misconduct than are men” (Wiener et al. 2004: 62).

In assessing the applicability of these strands of research to understanding judicial decision making processes, research that uses experimental methods or asks subjects to evaluate vignettes is most similar to the judicial setting, where judges carefully consider the detailed facts of a case in making their ruling. Judges are confronted not with discrimination in the abstract (as a question on a survey), but with “the empirical realities of women's lives” (MacKinnon 2002: 832), which they must evaluate in light of controlling legal doctrines. For instance, a plaintiff claiming disparate treatment under Title VII must show that “but for” her sex, she would not have been subject to an adverse employment action. O'Connor et al. (2004: 91) point out that being able to “put oneself in the target's shoes will affect the extent to which one believes the target's story about the complex facts at issue.”

Judges in discrimination cases must be able to put themselves in the “shoes” of both the plaintiff and the defendant because of legal doctrines that structure their decision making. For example, in cases of employment discrimination brought under Title VII, judges employ shifting burdens analyses in disparate treatment claims. As another illustration, in a sexual harassment case, a judge will evaluate whether a “reasonable person” would concur with the plaintiff's assessment that she faced a hostile work environment. A theoretical construct that relates to this process of understanding is the concept of empathy, which scholars have leveraged to explain why legal decision makers side with particular parties (Glynn and Sen 2015; Negowetti 2014). In the next sections, we draw on this line of research to develop a framework for our analysis that connects empathy with gender and judging.

Judging and Empathy

Plumm and Terrance (2009: 191) define empathy as “the ability of one person (observer) to take on the perspective of another (actor).” They go on to distinguish between trait empathy and situational empathy. Trait empathy is triggered by a similarity between the observer and actor, such as gender or race; situational empathy occurs when imagining oneself in the situation of the actor. The latter type of empathy can be induced in experimental manipulations where the subject is explicitly tasked with placing him or herself in the “shoes” of another person. However, group membership (e.g., race or gender), when salient, can mediate the effects of situational empathy that might be induced in an experiment or in a case; for instance, white mock jurors who were induced to feel empathy toward black or white criminal defendants still routinely “sentenced” the black defendants more harshly than their white counterparts (Johnson et al. 2002). Other research on the role of empathy in juror decision making has found that female mock jurors are much more likely to side with battered women who kill, regardless of empathetic induction strategies (Plumm and Terrance 2009). While men who were induced to feel more empathy for the battered woman in this study did express more positive views of the victim than men not in the empathy induction condition, overall men were still less supportive of the defendant than women were. Finally, compared with men, female mock jurors showed higher levels of empathy toward a rape victim in a mock trial, particularly those mock jurors who had personally experienced rape (Dietz et al. 1982).

These studies suggest the utility of drawing an analogy between empathetic induction as presented in experimental research (“situational empathy”) and decision making that requires a judge to understand the actions of a plaintiff and defendant in a case alleging sex discrimination (recall that situation empathy refers to the ability to imagine oneself in another person's situation). As the studies above indicate, situational empathy may vary depending on one's ability to observe a similarity between oneself and the other person (“trait empathy”). In the context of sex discrimination, trait empathy suggests that women should identify with an alleged victim who was also a woman. Empirical findings also support this approach, pointing to persistent differences between men and women's understanding of sex discrimination that is linked to the degree to which each group can relate to claims of sex discrimination. We argue, however, that it is not simply a trait that affects the type of empathetic response but how well members of that group (i.e., women) can relate to the position taken by the plaintiff. Women who experienced severe, first-hand discrimination will be particularly sensitive to claims of sex discrimination alleged by other women. Our argument considers next how these empirical findings can be extrapolated to help understand the behavior of federal judges in cases of discrimination.

Judge Gender and Discrimination

Existing surveys provide support for the contention that women and men on the federal bench will vary in their views about the prevalence, causes, and consequences of gender inequality in society. One study, conducted by Elaine Martin, surveyed President Carter's judicial appointees to the lower federal courts (30 women and 92 men). Responses to several questions point to major differences between men and women in areas relevant to sex discrimination. Fifty-five percent of all women reported “frequently” experiencing conflicts between career and family when their children were younger, compared with only 28 percent of male respondents. This cohort of female appointees also expressed higher levels of support for the women's movement (86 percent) relative to the male appointees (56 percent). Perhaps the most striking difference came in response to a question about one's major problems as a woman or man in the law. Female judges overwhelmingly (81 percent) made explicit reference to experiencing some form of sex discrimination, like “bias against women” or the “belief that a woman's place is in the home” (Martin 1990: 207). In contrast, male respondents referenced professional challenges or time management challenges as a major problem in the law (Martin 1990: 207). Martin concludes, “Women judges in this study, perhaps as a consequence of these personal experiences, evidence greater attitudinal feminism than men” (1990: 208).

A second study (Martin et al. 2002) draws from a 1988 survey on gender bias conducted by the Florida Supreme Court Task Force. (The respondents included Florida attorneys and judges of both sexes.) The authors argue that women's greater experiences with gender bias will sensitize women, more than men, to these issues (667). Consistent with these expectations, women lawyers and judges reported observing more gender harassment and sexual harassment than their male counterparts. Interestingly, male judges reported the fewest observations of gender bias of any group (even compared with male attorneys). The analysis also compared men and women's responses in several areas that tapped into a feminist consciousness: rape myths, maintaining the traditional division of labor in the home (“separate spheres”), divorce property rights, stereotypes of women, and domestic violence.⁸ Among judges, women expressed significantly more feminist answers than men in every area except for “separate spheres” (in which there was no significant differences between the sexes). Martin et al. (2002) conclude that, overall, there is a strong relationship between observations and feminist consciousness for the women in the survey, but not for the men.

Public statements by female judges themselves support our central argument that many experienced sex discrimination and that such experiences informed their work on the bench. As part of the American Bar Association oral history project “Women Trailblazers in the Law,” many of the first wave of female appellate judges were interviewed about their experiences in law school and the legal profession. While these interviews represent only anecdotes as opposed to systematic data, the comments within do shed some light on the challenges faced by women who were among the first to be appointed to the federal appellate bench.

First, in terms of experiences in law school, many of these judges attended law schools in which the number of women in their class was in single digits. Both Shirley Hufstedler (a Johnson appointee) and Cynthia Hall (a Reagan appointee) discussed separately in their oral histories that they were one of only two women in their Stanford law school classes. Ruth Bader Ginsburg (JD 1959) told The New York Times that the low number of women often meant additional scrutiny and pressure: “[M]ost sections had just 2 women, and you felt that every eye was on you. Every time you went to answer a question, you were answering for your entire sex … You were different and the object of curiosity” (Bazelon 2009).⁹ Hall attended law school at a time when many veterans were using the GI Bill to earn law degrees and observed that her fellow students were critical of her for “taking a good man's place.”¹⁰ Law faculty could also be hostile toward the few female students in their classes. For instance, Stephanie Seymour (JD 1965) describes how both professors and male classmates were “overt and obvious” in expressing their opinion that “I didn't belong there simply because I was female.”¹¹ Two female judges (Patricia Wald and Cynthia Hall) who attended different law schools, Yale and Stanford, both recalled that law professors liked to call on women more often than men when discussing rape cases (Wald 1994: 980).¹²

Once in the legal profession, many of these women encountered blatant sex discrimination from employers. Judge Betty Fletcher (JD 1956) recalled that she was blindsided by the discrimination she experienced when she was looking for a job right after law school, describing the prejudice as hitting her “like a ton of bricks.”¹³ In their oral histories, both Mary Schroeder (JD 1965) and Carolyn King (JD 1962) discuss experiences with flagrant discrimination during their time in private practice. When Schroeder learned she was pregnant, she was advised to keep her pregnancy a secret; similarly, King was informed that she was denied promotion to partner because the other partners believed that she should be at home with her children. A somewhat later appointee, Rosemary Barkett (JD 1970) told an interviewer that she frequently heard from other women lawyers that “they could not maintain their positions in the law firms and have children” and that she believed this was a wrong that needed to be remedied.¹⁴

Some women also faced hostility when they became judges. For instance, the first woman appointed to the U.S. Courts of Appeals, Florence Allen (JD 1914) was not welcomed to the Sixth Circuit by her new male colleagues. Kenney (2013: 141) writes that the other judges “had in fact opposed her appointment. Three judges failed to write a customary letter of congratulation.” Another colleague “was so distressed he reportedly took to his sickbed for two days following her appointment” (Ginsburg and Brill 1995: 283).

A few of these “trailblazer” judges have gone farther than simply cataloguing their experiences with discrimination and noted how their past has informed their actions as judges. For instance, Judge Patricia Wald (JD 1951) writes, “a judge is the sum of her experiences and if she has suffered disadvantages or discrimination as a woman, she is apt to be sensitive to its subtle expressions” (2005: 989). Similarly, when she was the only woman on the Ninth Circuit, Hufstedler (JD 1949) noted that many of her male colleagues “had a hard time seeing the world as it really is” in sex discrimination cases, and that she would tease them about things like “irrelevant” weight limitations for women carrying packages.¹⁵

To summarize, survey results reported in the literature demonstrate that women judges are generally more aware of gender bias and sex discrimination than men. Oral histories and public statements by women in the first cohort of federal appellate judges document experiences with blatant sex discrimination in law school and the profession. And comments by two judges in particular suggest that their behavior in sex discrimination cases was influenced by their own experiences as and observations about women.

Linking Gendered Experiences and Judicial Behavior

As we alluded to in the introduction, there is a growing body of evidence in the U.S. Courts of Appeals literature showing that female judges are more likely to support the plaintiff in sex discrimination cases (Boyd et al. 2010; Peresie 2005), as well as in other types of employment discrimination cases (Songer et al. 1994). Their presence alongside male judges on a panel is also associated with favorable outcomes for female plaintiffs in sexual harassment cases (Moyer and Tankersley 2012) and in causing male colleagues to support plaintiffs in sex discrimination cases (Boyd et al. 2010) and other employment discrimination cases (Farhang and Wawro 2004). These scholars speculate that women's greater sensitivity to gender bias because of their personal experiences may be driving the differences between men and women. Our argument builds on this contention and suggests that first-hand experience with severe, pervasive discrimination was commonplace for the first wave of female judges because of the unique set of circumstances these women faced.

The oral histories of the “trailblazer” women appointed to the federal bench describe a vastly different environment for those who attended law school in the 1950s, 1960s, and early 1970s when compared with those who received their law degrees in later eras. To illustrate, Figure 1 displays the percentage of enrolled women in law school from 1947 to 2011. It is important to note that the numbers from the 1940s and 1950s are suppressed in part because many law schools openly refused to admit women until the 1960s (Mossman 2006). As the judges’ comments from the previous section indicate, women who attended law school in earlier eras often felt isolated and under tremendous pressure because of their small numbers. They also experienced harassment and outright hostility from male classmates and faculty.¹⁶ One such judge, Ruth Bader Ginsburg has recounted that during her time at Harvard Law School, the dean invited all the female students to his house for dinner, only to ask them why they were taking places that could have been occupied by deserving men (www.oyez.org).

Figure 1. Trends in Women's Enrollment in Law School.

Another shared set of experiences among the earliest female judges is the employment environment that they faced. Prior to the 1964 passage of Title VII, employers could legally refuse to hire qualified women. Because employers continued to engage in discriminatory practices into the 1970s (Epstein 1983), the passage of the Equal Employment Opportunity Act of 1972 provided additional enforcement powers to the federal government. However, other measures, including the Pregnancy Discrimination Act (1978), were needed to promote equal employment opportunities for women in the workplace.

As summarized above, the passage of civil rights laws and societal expectations about working women shifted substantially over time. But beyond effects that signify societal shifts, we argue that, for the first wave of female judges, their personal experiences with sex discrimination and gender-based hostility were qualitatively unique, even compared with the sexism experienced by women who became lawyers in later eras. In particular, their small numbers in law schools heightened their isolation and encouraged scrutiny of them in ways that men never experienced as part of their professional training. We thus contend that the first “trailblazer” wave of female judges should have distinctive decision-making patterns that reflect these personal experiences with discrimination.¹⁷ However, as women's representation in the law increased and the legal culture shifted to reflect this, differences between men and women judges in their responses to sex discrimination should diminish over time.

The empathy literature in social psychology reviewed above suggests the utility of drawing on this concept in building a theoretical account of judicial decision making. In an analysis focused on race, Weinberg and Nielsen (2012) use an empathetic perspective to examine district court decisions on motions for summary judgment in civil rights cases. They argue that when evaluating whether a case should go forward at this stage, a trial judge's decision depends largely on his or her perception of an employer's actions against a plaintiff (2012: 323). At the appellate level, circuit judges also may review whether the facts of the case (in the light most favorable to the appealing party) supported a summary judgment decision. Appeals involving sex discrimination claims under McDonnell Douglas vs. Green and Texas Department of Community Affairs vs. Burdine also require appellate judges to assess whether the district court committed reversible error in their assessment of whether the plaintiff showed, by a preponderance of evidence, that her employer discriminated against her. As part of the analysis for Title VII disparate treatment claims, judges must assess the actions taken against the plaintiff and whether the defendant's explanation for those actions are mere pretext.

In this respect, all appellate judges are potentially induced by the very nature of their task to put themselves in the situation of the plaintiffs (i.e., situational empathy). However, we expect that there will be variation in their responses linked to whether the judge can relate personally to the plaintiff's experiences or whether they are inclined to be more deferential to the defendant's arguments. Past experiences with sex discrimination should, in effect, prime those women who experienced severe discrimination during their professionally formative years to be more responsive to female plaintiffs alleging discriminatory treatment because of their sex. However, as experience with overt discrimination diminishes over time, so too should gender differences. Thus, we test the argument that it is a gendered experience with discrimination, rather than gender alone, which enhances the likelihood of a response in support of the plaintiff.

Hypothesis

As described above, trailblazer women's recollections of their time in law school and private practice indicate that many of their male classmates, male professors, and male colleagues did not respond favorably to the novelty of working with women as peers. The intensity and overt nature of the discrimination they experienced should prime this group of women to be more receptive to claims of sex discrimination. Thus, while men and women who attended law school at the same time are both part of the same law school cohort, we expect that their formative experiences during this stage of their legal career were fundamentally different with long term effects flowing from these experiences over the course of their judicial careers.

Our review of public opinion research on attitudes toward working women indicates a clear, liberalizing trend over time (Ciabattari 2001; Brewster and Padavic 2000). One analysis of surveys about the desirability of married women working outside the home even concluded there were no significant differences between men and women after 1975 (Simon and Landis 1989). As such, we hypothesize that the trailblazer group of women will be more likely than men of the same generation to support the position of the plaintiff in sex discrimination cases, but that differences between male and female judges will decrease in more recent law school cohorts.

Data and Measures

Our study draws from the dataset of sex discrimination cases in the U.S. Courts of Appeals used by Boyd et al. (2010) and originally compiled by Sunstein et al. (2006). Updated by Epstein et al. (2013), the cases extend from 1995 to 2008 and include both sex discrimination and sexual harassment claims (Sunstein et al. 2006: 159). We supplemented these data by collecting additional biographical information from the Federal Judicial Center on a judge's legal education, including the year in which he or she received his or her law degree. Where a judge was listed as having multiple law degrees (i.e., J.D. and L.L.M), only the first degree was recorded.¹⁸ Table 1 lists all female judges appointed to the U.S. Courts of Appeals, denoting those who appear in the dataset along with their birth year, appointment cohort, and year they completed their law degree. The median male judge in our sample was born in 1937 and completed their legal education in 1963, while the median female judge was born in 1944 and finished law school in 1971.

Table 1. Female Judges in the U.S. Courts of Appeals

All Appointees Through 2008.

Notes: ^*judge is either African-American or Latina; ^{^}judge did not decide any sex discrimination cases included in the dataset. The nominating president refers to the president that nominated the judge to the U.S. Courts of Appeals.

The dependent variable in our analysis is the case outcome supported by a judge's vote (coded 1 if pro-plaintiff and 0 if pro-employer).¹⁹ Our central independent variables of interest are a judge's sex (1 = female, 0 = male) and the law school graduation year. Because we argue that the effect of gender is contingent upon when a judge attended law school, we create an interaction term between judge sex and law school graduation year.²⁰

We begin by looking at the gender composition of the student population enrolled in law schools beginning in 1947 (the first year for which aggregate data are available), shown in Figure 1. The law school enrollment data point to a distinctive shift that took place in the decade of the 1970s. In 1950, the percentage of women enrolled in law school was 3 percent, and the figure rose only slightly to 3.4 by 1960. However, the percentage of women enrolled in law school rose from 8.6 in 1970 to 34 percent in 1980. Similarly, the average of the percent change from the previous year was less than half a percentage point in the 1960s, but 2.5 percent during the decade of the 1970s.²¹

We also account for other factors that influence judicial voting. The most important competing explanation for liberal voting in sex discrimination is a judge's ideological predisposition, which has been shown to have consistent effects on judicial voting in the Courts of Appeals (Sunstein et al. 2006; Zorn and Bowie 2010). We utilize the Judicial Common Space scores (Epstein et al. 2007; Giles et al. 2001), continuous measures of ideology that range from −1 (most liberal) to +1 (most conservative). The JCS measure is used to estimate individual judge preferences, as well as circuit preferences (median circuit JCS score). Following Epstein et al. (2013), we also control for the policy preferences of the panel by including two dummy variables that indicate whether the other two judges were both Democratic appointees (1 = yes, 0 = no), or both Republican appointees (1 = yes, 0 = no) (the excluded category is a panel with one Democrat and one Republican). Because standards of appellate review encourage deference to the district court, another variable accounts for the ideological direction of the district court decision (1= liberal, 0 = conservative). In addition to including fixed effects for circuit and year, we estimate models with robust standard errors clustered on judges.

Analysis

Table 2 presents the results from a logit model that estimates the likelihood that a judge will cast a pro-plaintiff vote in a sex discrimination case.²² The control variables perform largely as predicted. Judicial policy preferences are partially driving voting behavior in sex discrimination cases; the variable for judicial ideology is correctly signed and significant at the 0.001 level. While being seated with two Democratic appointees is positively related to casting a liberal vote relative to voting by judges on “mixed” panels, being seated with two Republicans does not have a statistically significant effect. If the district court found in favor of the plaintiff, as expected, a judge is more likely to cast a pro-plaintiff vote, too. The control for circuit ideology fails to reach statistical significance.

Table 2. Logit Model of Pro-Plaintiff Voting in Sex Discrimination Cases

U.S. Courts of Appeals (1995–2008).

Notes: ^**p < 0.05 (two-tailed). Robust errors are clustered on the judge. Controls for year omitted for space. All models are significant at p < 0.001. The First Circuit is the excluded reference category.

Turning now to the interaction between judge sex and JD year, the coefficient does not reach conventional levels of statistical significance. However, because the coefficients and standard errors on an interaction term in a logit model are not directly interpretable, we must assess the statistical and substantive significance of the judge sex-law school year interaction by graphing the marginal effects at quantities of interest (Brambor et al. 2006). In Figure 2, we graph the marginal effect of being female conditioned by the law school graduation year, holding continuous variables at their medians and dichotomous variables at their modal values. Because there are very few observations in our data from women who attended law school in the 1940s and 1950s, we utilize 90 percent confidence intervals in our graph, though similar results obtain when 95 percent confidence intervals are used.²³ The resulting plot shows strong support for our argument about the important effects of socialization into the legal profession on gendered judicial decision making in sex discrimination cases.

Figure 2. The Diminishing Differences Between Men and Women over Time.

First, as Figure 2 shows, the marginal effect of being female is highest in the earliest cohorts and declines steadily with more recent JD cohorts. This decline is more a function of the sizeable drop in female judges’ pro-plaintiff voting (a 0.12 drop between a 1943 graduate and a 1991 graduate) than changes in male judges’ voting, which is relatively flat (only a 0.02 change over the entire time period). Second, the difference between men and women is only statistically significant during a 21 year window: law school classes of 1954 through 1975.²⁴ And while we can only speculate why the effect becomes insignificant after 1975, the law school enrollment numbers may shed some light, particularly if we focus on the rate of growth from year-to-year, rather than just the percentage of enrolled students who were female (see Figure 3). According to the ABA data, during the 1947–2011 period, 1974 was the year with the single largest year-to-year increase (4.1 percent) in the percent of law students who were female. After this peak, the rate of growth in women's enrollment experienced a steady decline from 1975 onward.²⁵

Figure 3. Rate of Growth in Women's Enrollment in Law School.

As a robustness test, we also re-estimated the model without the interaction term, instead conceptualizing a clear delineation point at which the socialization dynamics changed. We selected 1975 as the cutoff year for several reasons. First, as discussed above, 1974 was a record year for the growth in women's law school enrollment and growth slowed beginning in 1975 (though women's enrollment did continue to increase until around 2000). By this time, Title VII had been in place for a decade and had been reinforced by additional federal legislation in 1972. Additionally, surveys show that, at this point, men and women's attitudes about working women no longer diverge (Simon and Landis 1989) and that a majority of Americans had shifted their views of the women's movement to be more favorable than unfavorable (Huddy et al. 2000).

For this analysis, we created a series of dummy variables indicating whether the judge was a female who graduated law school before 1975, a female who graduated in 1975 or later, a male who graduated before 1975, or a male who graduated in 1975 or later. The results, shown in the Appendix, tell a similar story: significant differences emerge between the earliest group of women (JD before 1975) and men of the same generation, but disappear when men and women who graduated in 1975 or later are compared with each other.²⁶ There are no significant differences found between women who graduated before 1975 and women who graduated afterward. Thus, it seems that gender differences in voting in sex discrimination cases are not static, but instead vary in ways that correspond to major changes in gendered socializing experiences in the legal profession and society.

Discussion

Our analysis makes clear that recent scholarship finding differences between men and women judges in sex discrimination cases is being driven in part by the voting behavior of the cohort of trailblazer women. While it may seem unsurprising that a group that includes President Carter's nominees would be liberal in their voting patterns, it should be noted that this group also includes judges appointed by Reagan, George H.W. Bush, and Clinton. (Indeed, in our data, 36 percent of the observations of women with law degrees before 1975 were Republican appointees.) As Judges Hufstedler, a Democratic appointee, and Rovner, a Republican appointee, observed in the opening quotes, the experience of always being first was a challenging one—and formative for how these women would view the challenges faced by other women in the workplace.

Judges Hufstedler, Rovner, and others in the first wave of female judges entered the legal profession during a period of immense change in attitudes about gender and opportunities for women. From 1954 to 1975, the percentage of women enrolled in law school skyrocketed from just under 4 percent to nearly a quarter of all students. Before 1965, employment discrimination against women was not banned by federal law, and even after the passage of Title VII, the implementation of its protections did not happen overnight, particularly in male-dominated professions like the law. The shared sense of struggle among the trailblazer women bonded them together as law students (Ward 1994: 980) and later when they blazed a new trail as federal judges. Indeed, both Mary Schroeder and Dolores Sloviter told separate interviewers that the 10 women that Carter appointed to the U.S. Courts of Appeals always remained close, despite being scattered across the country (Haire and Moyer 2015: 105).²⁷

Beyond this particular issue area, our findings have important implications for judicial selection. Our results challenge the assumption held by many involved in the nomination and confirmation process that particular traits (gender, race, and ethnicity among others) predict future voting behavior consistent with stereotypes about empathy in which women judges are expected to support women in cases of sex discrimination or African Americans are more likely to support minorities in cases of race discrimination. We find evidence that judicial empathy for a plaintiff who alleges discriminatory treatment is not borne from a trait, but instead appears to form from experiences with discrimination. While we focus on first-hand personal experiences with sex discrimination and their connection to empathy, other recent work suggests that personal experiences in the context of parent-child relationships can change judicial behavior as well (Glynn and Sen 2015). Both of these perspectives emphasize the role of personal experiences in affecting behavior in ways consistent with empathetic responses.

Can male judges be sensitized to the claims of sex discrimination plaintiffs? While we find no evidence that male judges from more recent cohorts were liberalized by their experiences attending law school with large numbers of women, the literature does show that men change their voting behavior when seated with women in discrimination cases (Boyd et al. 2010; Farhang and Wawro 2004; Peresie 2005). Moreover, experimental research suggests that through empathetic induction, men are able to better take the perspective of female criminal defendants (Plumm and Terrance 2009: 202–03). Taken together with our findings, we think this suggests that male judges can be made aware of the realities of discrimination through working closely with female colleagues, particularly those whose own perspective on discrimination comes from their personal experiences. Future research should explore whether the addition of new judges who attended law school during the post-2000 era (when men and women's enrollment were close to parity) is connected to other behavioral or attitudinal impacts that can be traced back to that formative experience.

Although scholars generally theorize that the effect of gender on judicial decision making operates differently than the effect of race, our findings suggest that, in the context of judging civil rights claims, there may be similar causal mechanisms at work. The oral histories of women judges emphasize that trailblazers personally contended with gender inequality in law school (and society) and hinted at how this shapes their views on the bench. Here, we argue that it is this sense of shared struggle with discrimination that accounts for their response to plaintiffs when evaluating their claims. This is similar to the account advanced to explain why African Americans, including judges, are more likely to support the claims of minorities in affirmative action cases and Voting Rights Act cases (Cox and Miles 2008, 2009; Kastellec 2013). Oral histories, public opinion scholarship and biographical accounts emphasize that more recent African Americans—including legislators and judges—continue to feel a strong sense of shared fate that fosters a perspective where a black judge can empathize with a plaintiff who has faced potentially similar circumstances (Haire and Moyer 2015). Moreover, as with sex discrimination cases (Boyd et al. 2010), African-American judges appear to be able to change the voting behavior of white colleagues on a panel so that they are more supportive of the position of the minority (Kastellec 2013).

These findings underscore the importance of scholars continuing to investigate the role that empathy may play in judging, including its role in shaping interactions on appellate panels. Informed by literature that identifies both trait and situational empathy (Plumm and Terrance 2009), the results from our study are consistent with the interpretation that empathy is not triggered simply by a demographic trait, but may be better conceptualized as having been primed by one's life experiences and induced through the act of judging, similar to how it is induced in experimental settings.