Gender and the U.S. Supreme Court: An Analysis of Voting Behavior in Gender-Based Claims and Civil-Rights and Economic-Activity Cases

Article excerpt

Although several studies examine the influence of gender in the U.S. Supreme Court, research is still unsettled regarding the overall influence of gender in the voting behavior of Supreme Court justices. Employing data from Harold Spaeth's U.S. Supreme Court Databases, I systematically examine the extent to which gender influences individual Supreme Court justices' voting decisions across four issue areas. I find that although gender does not appear to influence the voting behavior of Supreme Court justices in civil-rights cases, male and female justices vote significantly differently in cases involving sex discrimination, reproductive rights, and economic activity.

The vacancies left by Justice David Souter and Justice John Paul Stevens presented President Obama with the rare opportunity to appoint two new members to the U.S. Supreme Court. President Obama's nomination, and the subsequent appointments, of Sonia Sotomayor and Elena Kagan has drawn attention to the criteria presidents use when selecting candidates to the Supreme Court, particularly whether gender, race, and ethnicity should be taken into account. After learning of President G. W. Bush's nomination of now Chief Justice John Roberts as her replacement, retired Justice O'Connor commented, "He's good in every way, except he's not a woman" (Landers, 2005:A1). Some believe it is important to have a "representative" Supreme Court, while others feel that background characteristics such as race, gender, and ethnicity are of little significance, and the sole criterion for selecting a nominee should be judicial qualifications.1 A question this discussion raises is whether personal characteristics of judges, such as gender, influence the decision-making process of the Supreme Court. Is gender irrelevant to judging, or have Justice Sandra Day O'Connor and Justice Ruth Bader Ginsburg brought a different judicial perspective to the bench? Does it make a difference whether both male and female justices serve on the U.S. Supreme Court?

In recent years, judicial scholars have begun to devote attention to exploring whether behavioral differences are apparent in the decisions male and female judges make (see, e.g., O'Connor and Segal, 1990; Songer, Davis, and Haire, 1994; Martin and Pyle, 2000). Much of this research, however, focuses on examining state supreme courts, U.S. District Courts, and the U.S. Courts of Appeals, while few studies look at the U.S. Supreme Court. Surprisingly, we really have little evidence about whether or not the behavior of male and female Supreme Court justices, while controlling for ideology, differs over a range of issue areas. This article assesses whether differences persist in the final voting decisions of male and female Supreme Court justices across four issue areas: sex discrimination, reproductive rights, civil rights, and economic activity.

THEORY

Some legal scholars conclude that as the number of female attorneys and judges continues to grow, we are much more likely to observe differences between male and female judges, as well as see women having a profound effect on law (see, e.g., Menkel-Meadow, 1985). The major goal of this research is to assess whether the voting patterns of the first two female justices to serve on the U.S. Supreme Court differ from that of their male colleagues. That is, are Justice O'Connor and Justice Ginsburg more likely to vote liberally than the male justices serving on the Court in cases involving gender-based claims (sex discrimination and reproductive rights), civil rights, and economic activity. There are two primary explanations that provide the basis for expecting that Justice O'Connor and Justice Ginsburg will cast more liberal votes than their male counterparts-"different voice" theory and the belief that the role orientations of male and female judges differ and, thus, female judges substantively represent and act on the behalf of women.

"Different Voice" Theory. Research has shown that a gender gap exists between men and women in regards to compassion issues, attitudes toward violence and war, the economy, and partisanship. Women are more likely to report being Democrats than men; to support government spending on social issues, such as welfare and homelessness; and have more liberal attitudes toward violence and war (see, e.g., Shapiro and Mahajan, 1986; Conover and Sapiro, 1993; Kaufmann and Petrocik, 1999).

Many relate the differences that occur between men and women, especially on compassion issues and their attitudes toward violence and war, to Carol Gilligan's "different voice" theory. This theory suggests that men and women use different approaches to moral reasoning (Gilligan, 1982). Men are more likely to use an "ethic of justice"; they approach moral dilemmas by concentrating on autonomy. In contrast, women tend to use an "ethic of care"; they resolve moral dilemmas by focusing on a responsibility to others.

When applied to law, the "different voice" theory suggests that male and female judges will differ in conceptualizing the question the court is addressing, as well as how to resolve the dispute (see, e.g., Menkel-Meadow, 1985; Sherry, 1986; Davis, 1992-93). Thus, female judges may bring a "feminine perspective" toward law, while men may use a "masculine perspective" toward law. A "feminine perspective" maintains that individuals have a responsibility to others (including society), recognize the facts and elements of the specific case, and take into account the experiences each participant brings to the situation. In contrast, a "masculine perspective" toward law emphasizes competition, autonomy, rules, and individual rights. These different perspectives may, at times, be reflected in the behavior of male and female judges. Thus, male and female judges may vote differently.

Role Orientations and Substantive Representation. A stream of judicial politics research examines the influence of judges' role orientations on decision making (see, e.g., Gibson, 1979; Scheb, Bowen, and Anderson, 1991). Role orientation is commonly defined as "norms that constrain the behavior of jurists" (Epstein and Walker, 2001:34). Scholars often contend that each judge has a view on what he or she considers the proper role of judges and the courts. Moreover, it often is thought that judges act in accordance with these role orientations. Some evidence suggests that male and female judges may differ in their judicial role orientations. Allen and Wall (1987), examining whether women judges adopt a "token" or "outsider" role, find that women state supreme court justices tend to adopt the role of outsider. In all three issue areas (women's issues, criminal rights, and economic liberties), the female justices tend to take the most independent position on the court. Further, the female justices are more likely to support the most extreme pro-woman position in cases concerning "redressing gender-based societal inequalities" (p. 242).

Martin (1993), also interested in understanding the role orientations of women judges, conducts a survey about their roles in the judiciary. Her findings suggest that women judges most consistently disagree with the notion that women judges must be reluctant to ignore precedent because of their "token" status. In addition, women judges tend to believe they have a unique role to play in the judicial process and that female judges should represent their different life experiences on the bench. She also finds that a majority of female judges surveyed believe that one of the major roles of a female judge is to influence the attitudes of her male colleagues in cases involving women's rights issues, such as sex-discrimination claims. Since it appears that the role orientations of female judges partly focus on the representation of women's issues, we should expect behavioral differences to be apparent between male and female judges, at least on some issues. Thus, female judges may in fact "act for women" in addition to "representing women."

Although the number of female judges in the federal judiciary has been increasing over time, women are still underrepresented in the federal, state, and local court systems. Scholars have posited that this is an inherent weakness in our democratic system. Indeed, some suggest that it is problematic to have a political system that is dominated by men while a majority of the population is women because it, in effect, "suggests to women citizens that the political system is not fully open to them" (Lawless 2004:81). As a result, research has become concerned with female representation in our political system and, in particular, symbolic and substantive representation. Symbolic representation refers to the effects women public officials can have on female citizens and their perceptions of the political system. In contrast, substantive representation signifies that women officeholders will take into account the interests of their female constituents when making decisions. Thus, substantive representation suggests that the behavior of female officeholders will differ from their male colleagues, such as having different policy priorities. As a result women in public office can have a significant effect on political processes and policy outcomes.

If female judges differ in how they view their role as a judge, it is reasonable to expect that female judges may "act for women" rather than just "represent women" numerically in the political system. Peresie's (2005) study of the U.S. Courts of Appeals provides an example of female judges "acting for women" or substantively representing women. Her findings suggest that female judges are more likely than their male colleagues to vote liberally in cases involving sex discrimination, regardless of their ideology. Thus, it is possible that female judges' substantive representation may emerge in their final voting decisions.

Two Separate Theories? In the context of judicial decision making, "different voice" theory and the literature on substantive representation suggest that gender may influence the final voting decisions judges reach. However, they differ in terms of developing a framework by which to expect behavioral differences to emerge between male and female judges. "Different voice" theory posits that female judges may use a "feminine perspective." An important factor that female judges tend to take into account more than their male peers when reaching a decision is guaranteeing that all individuals are given the right of full membership of a community. This includes focusing on a responsibility to others and considering the relevance of contextual elements of the case.

When applied to law, substantive representation argues that female judges will be concerned about representing the interests of women. They will be inclined to "act for women." The major difference between these two theories is that "different voice" theory does not incorporate that judges may represent the interests of a female constituency. Although "different voice" theory and substantive representation develop separate frameworks for the expected role of gender in judicial behavior, it is possible that both can influence how a judge reaches a decision. Female judges may be more likely to "act for women" when the court is addressing an issue of concern to women. "Different voice" theory may become a stronger influence on a judge's behavior when deciding a case involving a "genderless" issue.

Sex'Discrimination and Reproductive-Rights Cases. Much of the work on symbolic and substantive representative is at the legislative level. These studies continually find that women public officials tend to be more liberal and more supportive of issues of concern to women than their male counterparts (see, e.g., Swers 2002). In terms of the judiciary, some female judges perceive their role as a judge differently than their male counterparts, believing that one of the major roles of a female judge is to influence the attitudes of her male colleagues in cases involving women's rights issues, such as sex discrimination (Martin, 1993). If female judges differ in how they view their role as a judge, it is reasonable to expect that female judges may be substantively representing, in addition to numerically representing, women in the political system. Feminist scholarship proposes that when substantively representing women, female public officeholders will be more likely than their male counterparts to be cognizant of or more responsive to the concerns of women. Moreover, it is thought that a primary goal of women public officials is to represent the concerns of women and thus strive to ensure that public policy is a reflection of such interests (Carroll, 2001). I expect that Justice O'Connor and Justice Ginsburg will engage in substantive representation. Thus, they will be more likely to vote liberally in sex-discrimination and reproductive-rights cases than their male counterparts.

The Equal Protection Clause provides that "no state shall . . . deny to any person within its jurisdiction the equal protection of its laws." Cases coming before the courts involving the Equal Protection Clause often evaluate the actions, policies, or both concerning the development and implementation of discriminatory practices, especially those targeted at political minorities, such as women. Under equal-protection jurisprudence, female judges are often considered a political minority. Some argue that as a political minority female judges will support the position of the minority because they themselves are a member of a similar group. Applying this logic to Supreme Court decision making, I expect that the female justices will be more likely to vote in favor of the female litigant in sex-discrimination and reproductive-rights cases than their male colleagues because they may tend to protect political minorities, and they belong to this particular group.

Civil Rights. Some suggest that the "different voice" theory helps to explain the different opinions men and women have on compassion issues, such as civil rights. Civil-rights laws attempt to ensure full and equal citizenship and to protect individuals from arbitrary and capricious decisions. The civil-rights cases that come before the Supreme Court tend to involve questions concerning issues such as discriminatory practices and equality. Applying the "different voice" theory to the civil-rights-voting decisions of the Supreme Court, I argue that the female justices will place greater emphasis on not tolerating discriminatory practices and on ensuring equality for all members within the community. Thus, Justice O'Connor and Justice Ginsburg will cast more liberal votes than their male colleagues in cases involving civil rights.

Economic Activity. Cases that come before the Supreme Court involving economic activity often include issues such as anti-trust, regulation of federal transportation, and regulation of public utilities (i.e., oil producers, cable television, etc.). Empirical studies that examine the political behavior of men and women find that women are more likely to report being Democrats than men, tend to take the liberal position on compassion issues, and often have a more pessimistic view of the economy (Shapiro and Mahajan, 1986; Chaney, Alvarez, and Nagler, 1998; Kaufmann and Petrocik, 1999). In economic-activity cases, a liberal vote opposes business, the employer, and arbitration, while supporting competition, liability, indigents, small business vis-à-vis big business, debtors, bankruptcy, consumers, the environment, and accountability (Segal and Spaeth, 1993). Thus, the liberal position for economic activity cases tends to oppose what is often referred to as "big business." In his analysis of policy preferences of men and women, Gainous (2002) finds that women are more likely to oppose "big business" than men. Based on the above studies, I maintain female justices will be more likely to cast a liberal vote in economic-activity cases than their male counterparts.

JUDICIAL DECISION MAKING AND THE ATTITUDINAL MODEL

The attitudinal model maintains that judges decide cases by considering the facts of the case vis-à-vis their ideological attitudes and values (Segal and Spaeth, 2002). This model argues that voting decisions often reflect a judge's personal policy preferences. The attitudinal model is most relevant to decision making in the Supreme Court, since justices can vote according to their personal policy preferences because of the lack of institutional constraints placed on the Court.2 However, as noted, several judicial studies find significant differences between male and female judges (see, e.g., Davis, Haire, and Songer, 1993; Martin and Pyle, 2000; Peresie, 2005). Thus, some research suggests that although political ideology may affect judicial decision making, a judge's gender may also influence this process.

THE BEHAVIOR OF FEMALES JUDGES: PAST RESEARCH

In developing empirically testable expectations concerning whether gender influences the voting decisions of Supreme Court justices, it is useful to consider the existing literature on female judges. Interestingly, prior work on the extent to which gender affects the voting decisions of judges is unsettled. While some find significant differences between male and female judges, others find more similarities than differences.

Similarities Between Male and Female Judges. Several studies reveal that although the voting behavior of male and female judges does not significantly differ in certain issue areas, in other areas some differences emerge. The earliest studies analyzing the influence of gender on voting decisions find little support for the expectation that the behavior of male and female trial-court judges differ (Kritzer and Ulman, 1977; Gruhl, Spohn, and Welch, 1981).3

In their study of the Michigan Supreme Court, Martin and Pyle (2000) conclude that while gender does not appear to influence voting decisions in women's issues or discrimination cases, male and female judges differ in their decisions regarding divorce law. In a subsequent study, Martin and Pyle (2005), examining divorce-law cases in all fifty states, find that female justices are more likely to support the female litigant than their male colleagues. McCall's (2003) findings comparing the decisions of male and female judges on state supreme courts between 1980 and 1998 are twofold. While before 1992 female state supreme court justices are more likely to vote pro-woman in sexual-harassment cases than their male peers, after 1992 the votes of male and female state supreme court justices are remarkably similar.

By analyzing decisions of district court judges, Walker and Barrow (1985) discover that the voting patterns of male and female district court judges do not vary in "women's rights" and criminal-procedure cases. However, their results suggest that female judges are more likely to be "pro government" and that male judges are more likely to support personal-liberties claims than female judges. Songer, Davis, and Haire's (1994) study of the U.S. Courts of Appeals reveal that while male and female judges significantly vote differently in employment-discrimination cases; no differences surface in obscenity or criminal search-and-seizure cases. By analyzing the voting patterns of male and female courts of appeals judges, Westergren's (2004) results lead her to conclude that male and female judges do not vote differently in sex-discrimination cases.

Differences Between Male and Female Judges. Several studies reveal significant differences between male and female judges. Prior work suggests state supreme courts are more likely to hand down pro-woman decisions in sex-discrimination cases when there is at least one woman serving on the court and that female state supreme court justices tend to be the most pro-women members of the court (Gryski, Main, and Dixon, 1986; Allen and Wall, 1987). Songer and Crews-Meyer's (2000) study indicates that female state supreme court judges are more likely to vote liberally than their male colleagues in obscenity and death-penalty cases. In addition, the presence of at least one female member on the court increases the probability that male justices vote liberally. McCall's (2008) findings support her expectation that female state supreme court judges vote more liberally than their male peers in search-and-seizure, juvenile, and domestic-violence cases. McCall and McCall's (2007) results suggest that female state supreme court judges are more likely to vote liberally in Fourth Amendment cases than their male colleagues after 1991. Furthermore, female judges serving with at least one other female judge are more likely to vote liberally than their male colleagues. Peresie's (2005) analysis of the influence of gender on the voting behavior of U.S. Courts of Appeals judges in sexual-harassment and sex-discrimination cases confirms that female judges are more likely to vote "pro-woman" than their male colleagues. Similar to the findings of Songer and Crews-Meyer (2000), her results suggest that the presence of a female on the panel significantly increases the likelihood that a male judge will vote liberally.

The Supreme Court. Much of the research on female judges examines differences between male and female judges in courts other than the Supreme Court. In fact, there are only a handful of studies that address gender differences in the Supreme Court (see, e.g., O'Connor and Segal, 1990; Davis, 1993; Aliotta, 1995; Palmer, 2002). However, consistent with much of the research on women judges, the existing literature on the influence of gender on judicial decision making in the Supreme Court is also mixed.

Many hoped that as a result of the appointment of the first female Supreme Court justice, the Supreme Court would become more responsive to the concerns of women. In light of this, Sherry (1986), applying Carol Gilligan's (1982) "different voice" theory to the Supreme Court, examines the opinion writing of Justice O'Connor and Justice Rehnquist. Her findings suggest that Justice O'Connor's use of a "feminine perspective" toward law explains several of the cases in which Justice Rehnquist and Justice O'Connor disagree.

Building upon Sherry's (1986) work, Davis (1993) and Aliotta (1995) analyze whether gender affects the voting decisions of the Supreme Court. Davis's findings reveal few significant differences in the voting decisions of Justice O'Connor and Justice Rehnquist in civil rights, the establishment clause, and criminal-procedure cases. Davis contends that it is questionable whether Justice O'Connor uses a "feminine perspective" toward law or rather whether she is simply not as conservative as her rítale counterpart. Aliotta (1995), assessing whether Justice O'Connor votes differently than her male counterparts by examining equal-protection cases, finds that Justice O'Connor is not more likely to vote liberally. O'Connor and Segal (1990) find that the Supreme Court's level of support for women's issues increased after the appointment of Justice O'Connor. In her analysis, Palmer (2002) suggests that although ideology may be the driving force behind levels of support for sex-discrimination claims, Justice O'Connor's support for such claims increased after the appointment of Justice Ginsburg.

While these studies begin to uncover the influence of gender in the Supreme Court, there are several limitations of this previous work. The first limitation involves the methodological approaches used in past research. While several studies quantitatively examine gender differences in the final voting decisions of Supreme Court justices, all but one analyze final vote outcomes without taking into account other factors that may potentially influence final voting decisions, such as ideology (but see, Aliotta, 1995).4 Judicial scholarship continually finds that the ideology of Supreme Court justices is a significant predictor of their voting decisions. To determine whether gender has an independent effect on voting decisions in the Supreme Court, one must control for the personal policy preferences of the individual justices.

Second, nearly all of the studies of the U.S. Supreme Court examine one issue area. To assess the extent to which male and female justices differ in terms of their voting patterns, it is important to study the voting outcomes of Supreme Court justices across a number of different issue areas, while controlling for significant factors that affect judicial voting decisions. This allows for a more extensive examination of what issue areas (if any) male and female justices differ in their voting patterns.

The research here is an attempt to overcome both of the limitations of past research. By examining gender differences in the Supreme Court across a number of different issue areas, while controlling for other potentially important factors, such as ideology, we can more fully address the question of whether gender affects the final voting decisions of the Supreme Court. A significant limitation of this study, however, is that it examines the extent to which gender influences the voting decisions of the U.S. Supreme Court by analyzing the votes of two women. Thus, this analysis has little generaliziablity. Rather, this research assesses whether the voting behavior of the first two female justices to serve on the U.S. Supreme Court significantly differs from that of their male counterparts. It also presents a model to analyze the extent to which gender influences the final voting decisions of Supreme Court justices. This is particularly relevant since the Supreme Court currently has three female members.

This research also contributes to the existing literature on female judges by providing a better understanding of whether Justice O'Connor and Justice Ginsburg substantively represent women. Examining the influence of gender on Supreme Court decisions essentially breaks down to analyzing the voting decisions of two female judges. However, by being a Supreme Court justice these two women have the ability to affect public-policy decisions even after they are no longer on the Court. Further, Justice O'Connor and Justice Ginsburg may make an impact on law beyond their vote. That is, it is possible if one or both justices substantively represent women their voting decisions will represent the concerns of women and influence the laws and practices in place. Thus, studying whether gender influences the voting decisions of the U.S. Supreme Court provides an initial glimpse at the potential influence female justices may have on law.

HYPOTHESES

I have formulated five hypotheses based on legal studies, feminist legal theory, congressional studies, and empirical analyses of political behavior:

HI: Female justices will be more likely to cast a liberal vote than their male counterparts in a sex-discrimination case.

H2: Female justices will be more likely to cast a liberal vote than their male counterparts in a reproductive-rights case.

H3: Female justices will be more likely to cast a liberal vote when one of the litigants is female than their male counterparts in sex-discrimination and reproductive-rights cases.

H4: Female justices will be more likely to cast a liberal vote than their male counterparts in a civil-rights case.

H5: Female justices will be more likely to cast a liberal vote than their male counterparts in an economic-activity case.

DATA AND METHODS

In the analysis that follows, I examine the influence of gender in the voting decisions of Supreme Court justices in sex-discrimination, reproductive-rights, civil-rights, and economic-activity cases.51 include all of the voting decisions in these four issue areas that came before the Court since the appointment of Justice O'Connor through 2000. This study uses data from Harold Spaeth's U.S. Supreme Court Database. More specifically, I employ data from the Justice-Centered Burger Court and the Justice-Centered Rehnquist Court databases.^ The unit of analysis is the individual justice's vote.

The dependent variable in this analysis is the direction of the individual justice's vote as identified in the Spaeth database (conservative vote = 0 and liberal vote = l).7 I construct a separate model for each issue area I examine. To determine the extent to which gender affects the voting behavior of individual justices, it is essential to examine an array of explanatory factors that past research identifies as predictors of judicial voting behavior. First, I create a dummy variable that measures the sex of the justice (0 = male and 1 = female).

Second, a wide body of research continually finds that attitudes, beliefs, or policy preferences (ideology) best predict and explain the votes of Supreme Court justices. Thus, I measure the ideology of each justice by employing the Judicial Common Space scores constructed by Epstein et al. (2007). The Judicial Common Space scores for each justice were derived by transforming ideology scores developed by Martin and Quinn (2002) so that Supreme Court justices can be placed in the Judicial Common Space.8 A criticism of measuring the policy preferences of Supreme Court justices by using the Judicial Common Space is that this measurement creates circularity because the prediction of final vote choice of justices is based on the voting decisions of the justices. However, Martin and Quinn (2005:3) argue that when modeling votes in one issue area it is appropriate to use the Judicial Common space scores as "circularity is not a practical concern."9 An advantage of the Judicial Common Space scores is that they allow for Supreme Court justices' ideal points to change over time (i.e., a justice's policy preferences), and previous research suggests that the policy preferences of several members of the court of last resort have changed over time (Esptein et al., 1998).

In the subsequent analysis I model the voting decisions of Supreme Court justices across four issue areas; however, since I construct a separate model for each issue area, I argue that the Judicial Common Space scores are an appropriate measurement of the ideology of Supreme Court justices. Note, however, that the Judicial Common Space scores are lagged by one year to control for the concern of circularity. This creates a measurement of ideology that takes into account the actual preferences of each Supreme Court justice; however, the year in which the votes of the justices are being predicted are not included in the operationalization of ideology. The Judicial Common Space scores range on a left-right continuum from -1 (extremely liberal) to +1 (extremely conservative). The expected sign of this variable is negative, indicating that justices with liberal policy preferences are more likely than conservative justices to cast a liberal vote.

The existing literature on the influence of gender on the policy priorities of elected officials finds that female officials tend to be more aware of, or responsive to, issues that are traditionally thought of as particularly important to women (see, e.g., Carroll, 2001; Swers, 2002). Applying this belief to the judiciary, some studies find that female judges tend to be the members of the court who are most supportive of women's issues (see, e.g., Allen and Wall, 1987). Although female judges may be more supportive of women's rights issues, it may also be the case that the sex of the plaintiff may affect male and female judges in different ways. That is, female judges may be more likely to cast a liberal vote in a sex-discrimination case if the plaintiff is a female litigant. To determine whether the sex of one of the litigants affects the final voting decisions of male and female judges, I create two dummy variables-one for female litigants seeking a pro-woman outcome and one for female litigants seeking an anti-woman outcome. The expectation is that the sex of the litigants will influence male and female judges in different ways; thus, I create two interactive variables (Sex of Judge * Female Litigant Seeking a Positive Outcome, Sex of Justice* Female Litigant Seeking a Negative Outcome).10

Studies continually find that the federal government enjoys high levels of success when appearing before the Supreme Court (Handberg, 1979; Crowley, 1987). To control for this, I create a variable that identifies whether one of the parties is a federal governmental entity (U.S. Government). The expectation is that if one of the parties is a governmental entity and is seeking a liberal court outcome, justices will be more likely to vote in their favor. To test this hypothesis, this variable is coded -1 if one of the parties is an administrative agency and is requesting a conservative court outcome and +1 if one of the parties is an administrative agency that is requesting a liberal court outcome. If neither party is a governmental entity, this variable is coded 0.

RESULTS

To understand whether gender influences the voting behavior of Supreme Court justices, I begin by analyzing the voting patterns of the individual female justices across the four issue areas and compare each to two ideologically similar male justices.11 Table 1 displays the number of conservative and liberal votes cast by Justice O'Connor, Justice Kennedy, and Justice Souter in civil-liberties, sex-discrimination, reproductive-rights, and economic-activity cases. This initial glimpse at the voting decisions of Justice O'Connor provides evidence to suggest that she may not be a "traditional" conservative member of the Court. Moreover, although she votes significantly more liberally than Justice Kennedy, a moderate conservative, in reproductive-rights cases, Justice O'Connor votes more conservatively than Justice Souter in civil-rights cases. Interestingly, Justice O'Connor votes more conservatively than both Justice Kennedy and Justice Souter in cases involving economic activity. The findings suggest that Justice O'Connor votes more conservatively in civil-rights and economic-activity cases and more liberally in sex-discrimination and reproductive-rights cases. Thus, she appears to be more liberal on "women's issues."

Table 2 presents the number of conservative and liberal votes cast by Justice Ginsburg, Justice Breyer, and Justice Stevens. The results reveal that Justice Ginsburg votes significantly more liberally in sex-discrimination cases than Justice Stevens. However, the two justices appear to vote fairly similarly in the other three issue areas, and any differences that emerge do not reach statistical significance. In addition, even in cases involving women's issues Justice Ginsburg and Justice Breyer do not significantly differ in their voting decisions. To further evaluate the influence of gender in the voting patterns of the Supreme Court, logistic regression is employed, it being the appropriate methodology when one wishes to regress a dichotomous dependent variable (in this case, whether a justice casts a liberal or conservative vote) on a series of independent variables.

Sex-Discrimination and Reproductive-Rights Cases. The models for justices' votes in sex-discrimination and reproductive-rights cases are presented in Table 3.12 Despite controlling for several potentially important intervening variables, including ideology, gender continues to significantly influence the voting behavior of Supreme Court justices. Given the nature of logistic regression, it is difficult to interpret the substantive effects of the individual variables. Thus, to provide more intuitive results, I calculate the predicted probabilities of observing a liberal vote in sex-discrimination and reproductive-rights cases (based on different values of the independent variables). The predicted probabilities are displayed in Table 4.

When all of the independent variables are held at their mean, there is a slightly greater probability of observing a liberal vote in sex-discrimination claims than in reproductive-rights cases (see Table 4). The probability of a female justice casting a liberal vote in a sex-discrimination case is 0.797, while the probability of a male justice voting liberally in this issue area is 0.622. In contrast, there is a 0.798 probability of observing a liberal vote from the female justices and a 0.556 probability from male justices in cases involving reproductive rights. This finding supports the belief that women officeholders, specifically female judges, may be more cognizant, more responsive, or both to issues of concern to women than their male colleagues. In addition, this suggests that Justice O'Connor and Justice Ginsburg may be substantively representing women.

Gender, of course, is not the only variable that significantly influences the voting behavior of the Supreme Court in both sex-discrimination and reproductive-rights cases. The results indicate that ideology affects the final voting decisions of Supreme Court justices, regardless of their gender. As predicted, the more liberal a justice's policy preferences, the more likely the justice is to cast a liberal vote in both issue areas.

Whether one of the parties is female and if one of the litigants is a government entity also appears to influence voting decisions in sex-discrimination cases. We are more likely to observe a justice voting liberally when there is a female litigant seeking a pro-woman court outcome. In addition, a justice is less likely to cast a liberal vote when there is a female litigant seeking an anti-woman court outcome. This suggests that the presence of a female litigant plays a role in the voting decisions of the Supreme Court in cases relating to sex discrimination. However, the coefficient for the multiplicative term measuring the interaction between the sex of the justice and whether there is a female party seeking a pro-woman or an anti-woman decision is not statistically significant. Thus, the sex of the litigant does not appear to influence Justice O'Connor and Justice Ginsburg significantly differently than it does their male colleagues. Providing support for the literature on how the federal government tends to be fairly successful when appearing before the high court, the results suggest that a justice is more likely to agree with a governmental entity that is a party to a case involving a sex discrimination-claim (see, e.g., Handberg, 1979).

Civil-Rights and Economic-Activity Cases. Table 5 presents the results of the logit model for civil-rights and economic-activity cases, while Table 6 displays the predicted probabilities. Ideology, and whether one of the parties is a governmental entity requesting a liberal decision, significantly influences justices' votes in civil-rights cases. The coefficients for each of these variables are in the expected direction. Not surprisingly, the likelihood of observing a liberal vote increases when one of the parties is a governmental entity and is seeking a liberal decision. However, gender appears to have no effect on voting decisions of Supreme Court justices in civil-rights claims when other factors are taken into account.

While gender does not appear to influence the voting decisions of Supreme Court justices in civil-rights cases, the results indicate that gender does affect the voting behavior of justices in economic-activity cases.13 However, the coefficient for sex of the justice is not in the expected direction. Contrary to my expectation, female justices are less likely to cast a liberal vote in economic-activity cases than their male counterparts. The probability of a female justice casting a liberal vote is 46.5 percent, while the probability of a male justice voting liberally is 48.9 percent in economic-activity cases (see Table 6). This finding is difficult to explain in part because much of the existing literature that addresses gender differences in public opinion suggests that women are more likely to oppose "big business" and tend to be more liberal on economic issues then men (see, e.g., Gainous, 2002; Shapiro and Mahajan, 1986). Thus, one would expect female justices to be more likely to cast a liberal vote in economic-activity cases. The economic issues included in the analysis are largely related to "big business" issues. Unfortunately, few studies examine whether a gender gap exists in this realm of economic activity. Much of the research focuses on men's and women's evaluations of the economy and governmental activism, rather than their views on specific issues involving economic activity. It is possible that men's and women's views on "big business" have changed over time.

In addition to gender, ideology and whether one of the parties is a governmental entity seeking a liberal decision significantly influence voting decisions in economic-activity cases. The more liberal a justice's policy preferences, the more likely the justice is to cast a liberal vote in an economic-activity case. There is also a greater probability of observing a liberal vote when one of the parties is a governmental entity and is seeking a liberal outcome than when neither party is a governmental entity.

Discussion

The results of this study shed some light on the role gender plays in the final voting decisions of Supreme Court justices. By empirically examining whether gender affects judicial vote choice, I find that gender differences exist in the voting patterns of male and female justices in sex-discrimination, reproductive-rights and economic-activity cases. However, it is important to note there are two major problems this study is unable to overcome. First, although Justice O'Connor was appointed by President Reagan, she is not a traditionally conservative justice; she is much more liberal on social issues, such as reproductive rights, than her conservative colleagues. Thus, this analysis is unable to compare and contrast the voting decisions of a conservative female justice and a liberal female justice. Second, this analysis lacks generalizability. It is not possible to definitively determine whether gender influences the voting decisions of the Supreme Court or that female justices substantively represent women, since the sample comprises just two women. However, it provides an initial examination of the ways in which gender may influence the voting decisions of the Supreme Court.

This research contributes to the literature on women judges, as well as judicial decision making, in several ways. First, unlike past studies of the influence of gender in the voting behavior of the Supreme Court, I systematically explore gender differences across a number of different issue areas. While female justices do not appear to vote differently than their male colleagues in every issue area that comes before the Supreme Court, gender does appear to be playing a role in decisions regarding sex discrimination, reproductive rights, and economic activity. Second, this study employs the same methodological approach in examining gender differences across four different issue areas. This provides a more extensive evaluation of the role gender plays in the voting decisions of Supreme Court justices than past research. Third, it provides a starting point to examine whether women serving on the court of last resort may have a profound effect on law, one beyond their vote.

The finding that Justice O'Connor and Justice Ginsburg are more likely to vote liberally in sex-discrimination and reproductive-rights cases than their male colleagues provides tentative support that Justice O'Connor and Justice Ginsburg, in addition to symbolically representing women, substantively represent women and are at times more cognizant, responsive, or both to the concerns of women than their male peers. This analysis also suggests that ideology consistently influences the voting behavior of the Supreme Court. A question this raises is whether the female justices are voting in favor of the "pro-woman" position more than their male colleagues because they are more liberal. That is, are Justice O'Connor and Justice Ginsburg more cognizant or responsive to the concerns of women than their male counterparts because they are more liberal, or are their positions on such cases a function of their gender? According to "different voice" theory, these female justices may be conceptualizing the questions coming before the Court differently than their male colleagues, focusing on a responsibility to others, placing a greater value on the individual experiences the participants bring to the case, or doing both (see, e.g., Behuniak-Long, 1992). I argue that their use of a "feminine perspective" is reflected by the female justices being more likely to cast a liberal vote. In addition, the female justices are more likely to support the liberal position in cases disproportionally affecting women. Thus, this initial evidence suggests that they may be substantively representing women. These explanations also provide support for the argument that Justice O'Connor and Justice Ginsburg's support of the "pro-woman" positions in sex-discrimination and reproductive-rights cases are a function of their gender.

The results of this study have implications for the possible effect on law, its practices, and traditions that female judges and justices, specifically Justice O'Connor and Justice Ginsburg, may have beyond their vote. In recent years, research has begun exploring the ways in which political institutions are gendered or rather the ways in which an institution displays masculine and/or feminine features (Duerst-Lahti, 2002). A stream of feminist legal scholarship argues that law suffers from a masculine bias and that the judiciary is a gendered institution that emphasizes masculine characteristics (see, e.g., West, 1991; Aliotta, 2003.) That is, the norms of the courts are masculine in nature and to be able to function within the institution it is necessary to conform to the informal rules, such as valuing autonomy. Further, some scholars argue that when women began entering the federal judiciary system, because it was necessary for them to conform to the masculine norms of the institution, the behavior of male and female judges were more similar than different. However, studies showing that gender influences the voting decisions of judges, at least in some issue areas and in certain courts, indicate that the gendered character of the judiciary may be changing. In terms of the Supreme Court, the current study suggests that Justice O'Connor and Justice Ginsburg are willing to deviate from a masculine norm that may be in place and substantively represent women, regardless of the decisions their male colleagues make. Thus, they may be playing a role in changing the gendered nature of the Supreme Court.

The findings reveal that Justice O'Connor and Justice Ginsburg have a greater probability of supporting the pro-woman position than their male counterparts in terms of voting. If gender plays a role in the voting decisions of Supreme Court justices, it is possible that it influences other elements of judicial decision making. In terms of deciding which cases to review the Supreme Court uses the "rule of four" to decide whether or not to grant certiorari. It is possible that female justices, especially now that there are three serving on the Supreme Court, can have a substantial effect on which cases are granted certiorari. Further, if the female members continue to be the most pro-woman members of the Court, it may be likely that more gender-based claims are granted certiorari than when there were fewer female justices serving on the Court or when it was composed of all male members.

As Rohde and Spaeth (1976:172) note, the written opinions "are the core of the policy-making power of the Supreme Court." If gender is influencing the voting decisions of Supreme Court justices, as it appears it is in some issue areas, it is probable that at least some differences exist between how male and female justices approach the opinion-writing process or the opinions themselves. In addition, if gender influences opinion writing, we may see the female justices having a distinct influence on the precedent that the Court upholds or establishes. Finally, having a more diverse Supreme Court, at least in terms of gender, may affect the final outcome of a case.

Consistent with much of the judicial literature on women judges, this analysis suggests that gender influences judicial decision making, but only in some issue areas. This phenomenon is difficult to explain, but deserves further attention. Although there is not a definitive answer that explains why gender appears to influence judicial decision making in only some areas of the law and not others, additional studies on the role orientations of judges could provide greater insight concerning this topic. Much of the work on the role orientations of female judges tends to focus on female judges' perceptions of their role in the judiciary, as well as their role in representing women's concerns (see, e.g., Martin, 1993). Examining whether the role orientations, the policy positions of male and female judges, or both vary would provide a more comprehensive picture of whether a judge's gender influences role orientation and, thus, the way he or she approaches law. Thus, a potential goal for future research involves analyzing the role orientations of male and female judges. This could contribute to a greater understanding of why female judges do not consistently behave differently than their male colleagues.

* Katherine Felix Scheurer (kate.scheurer@mail.business.und.edu) is an assistant professor in the Department of Political Science and Public Administration at the University of North Dakota in Grand Forks.

1 Normatively speaking, it is thought that judges should objectively apply law without their personal policy preferences influencing their decisions. However, in reality "judges are political, and their politics seep into their decisions" (Epstein and Segal, 2005:144). In addition to politics influencing the decisions of judges, the federal nomination process is highly political. Judicial appointments become an important means by which a president (and senators) can influence public policy. When a president appoints a member to the Supreme Court, this allows him or her to have an indirect influence on shaping public policy for many years to come. Thus, ideology and politics play a significant role in the process of judicial appointments, particularly at the Supreme Court level (Epstein and Segal, 2005).

2 Justices, fot example, are rarely concerned with seeking higher office, the Court has complete discretion over its docket, and Supreme Court decisions are rarely overturned by Congress (see, e.g., Segal and Spaeth, 2002).

5 Note that in their study, Gruhl, Spohn, and Welch (1981) find some differences between male and female judges' treatment of male and female defendants.

4 Aliotta's (1995) study of the voting patterns of Supreme Court justices in equal-protection cases that came before the Court between 1981 and 1985 suggests that Justice O'Connor does not appear to use a different perspective toward law than her male counterparts. Although Aliotta (1995) examines gender differences in the voting patterns of Supreme Court justices in equal-protection cases by controlling for factors previous research finds to affect final voting decisions (i.e., party affiliation and case characteristics), as even she notes, her findings are far from definitive.

5 The Appendix includes a description of the specific issues that fall into each of these four issue areas.

6 The proceeding analysis is constructed from a selected set of variables from the databases as well as additional variables created from the existing variables in the data sets.

7 Specifically, the direction of the justices' votes in the cases are coded "liberal" if the justices' vote is in favor of the claimant in sex-discrimination claims; if the vote supports the female in reproductive-rights cases; and if the vote supports an individual's claim of a violation of their civil rights. In economic-activity cases, a liberal vote opposes business, the employer, and arbitration, while supporting competition, liability, indigents, small business vis-à-vis big ones, debtors, bankruptcy, consumers, the environment, and accountability (Segal and Spaeth 1993).

8 Martin and Quinn (2002) construct their measure of Supreme Court ideology from the voting behavior of the justices.

9 To assess whether it is appropriate to use the Judicial Common Space scores when modeling voting decisions of Supreme Court justices, Martin and Quinn (2005) estimate the ideal points excluding cases/votes in which the model is attempting to predict. Their findings suggest that excluding the case type of research interest changes the Court-specific measures only slightly.

10 The majority of parties that were female in the subsequent analysis were primarily parties in cases involving sex discrimination. Thus, the variables, "Female Litigant Seeking a Positive Outcome," "Female Litigant Seeking a Negative Outcome," "Sex of Judge * Female Litigant Seeking a Positive Outcome," and "Sex of Justice * Female Litigant Seeking a Negative Outcome" are only included in the model for voting in sex-discrimination cases.

11 To provide a preliminary analysis of the voting patterns of the male and female justices, I pair each of the female members of the Supreme Court with two male justices of similar political ideology. Justice O'Connor is paired with Justice Kennedy because he is the most similar to her in terms of political ideology. Although Justice O'Connor is often thought as the median justice, she became more liberal over time. Thus, I also pair her with Justice Souter. Accordingly, Justice Ginsburg is paired with Justice Breyer and Justice Stevens.

12 Since the individual justice's votes tend to be correlated with one another, 1 estimate Robust Standard Errors and cluster by the individual justice. In addition, I run two-tailed tests for each of the models included in the analysis.

13 To further examine whether gender influences the voting decisions of Supreme Court justices in sex-discrimination, reproductive-rights, and economic-activity cases, I ran three additional models. Rather than including one variable that measures the sex of the justice, I create a dummy variable for Justice O'Connor, Justice Ginsburg, and the male justices for each of these models. The findings remain the same.

[Reference]

REFERENCES

Allen, D., and D. Wall (1987). "The Behavior of Women State Supreme Court Justices: Are They Tokens or Outsiders?" 12 Justice System Journal 232.

Aliotta, J. M. (2003). "Gender and Judging: Some Thoughts Toward a Theory." Paper prepared for the Annual Meeting of the Midwest Political Science Association, Chicago.

- (1995). "Justice O'Connor and the Equal Protection Clause: A Feminine Voice?" 78 Judicature 232.

Behuniak-Long, S. (1992). "Justice Sandra Day O'Connor and the Power of Maternal Legal Thinking," 54 Review of Politics 417.

Carroll, Susan J., ed. (2001). The Impact of Women in Public Office. Bloomington: Indiana University Press.

Chaney, C. K., R. M. Alvarez, and J. Nagler (1998). "Explaining the Gender Gap in U. S. Presidential Elections, 1980-1992," 51 Political Research Quarterly 311.

Conover, ? J., and V Sapiro (1993). "Gender, Feminist Consciousness, and War," 37 American Journal of Political Science 1079.

Crowley, D. W. (1987). "Judicial Review of Administrative Agencies: Does the Type of Agency Matter?" 31 Western Political Quarterly 265.

Davis, S. (1993). "The Voice of Sandra Day O'Connor," 77 Judicature 134.

- (1992-93). "Do Women Judges Speak 'In a Different Voice?' Carol Gilligan, Feminist Legal Theory, and the Ninth Circuit," 8 Wisconsin Women's Law Journal 143.

Davis, S., S. Haire, and D. Songer (1993). "Voting Behavior and Gender on the U.S. Courts of Appeals," 77 Judicature 129.

Duerst-Lahti, G. (2002). "Knowing Congress as a Gendered Institution: Manliness and the Implication of Women in Congress." In C. Rosenthal (ed.), Women Transforming Congress, pp. 20-49. Norman: University of Oklahoma Press.

Epstein, L., and J. A. Segal (2005). Advice and Consent: The Politics of Judicial Appointments. New York: Oxford University Press.

Epstein, L., A. D. Martin, J. A. Segal and C. Westerland (2007). "The Judicial Common Space," 23 Journal of Law, Economics and Organization 303.

Epstein, L., and T. G. Walker (2001). Constitutional Law for a Changing America: Rights, Liberties, and Justice, 4th ed. Washington DC: CQ Press.

Epstein, L., V. Hoekstra, J. A. Segal, and H. J. Spaeth (1998). "Do Political Preferences Change? A Longitudinal Study of U.S. Supreme Court Justices," 60 Journal of Politics 801.

Gainous, J. (2002). "Is There a 'Woman's Perspective?' An Exploration of Gender Differences Along Republican and Conservative Lines." Paper presented at the Annual Meeting of the Southwestern Political Science Association, New Orleans.

Gibson, J. L. (1979). "Discriminant Functions, Role Orientations, and Judicial Behavior: Theoretical and Methodological Linkages," 39 Journal of Politics 984.

Gilligan, C. (1982). In a Different Voice: Psychological Theory and Women's Development. Cambridge, MA: Harvard University Press.

Gruhl, J., C. Spohn, and S. Welch (1981). "Women as Policymakers: The Case of Trial Judges," 25 American Journal of Political Science 308.

Gryski, G., E. Main, and W. Dixon (1986). "Models of State High Court Decision Making in Sex Discrimination Cases," 48 Journal of Politics 143.

Handberg, R. (1979). "The Supreme Court and Administrative Agencies: 1965-1978," 6 Journal of Contemporary Law 161.

Kaufmann, ?. M., and J. R. Petrocik (1999). "The Changing Politics of American Men: Understanding the Sources of the Gender Gap," 43 American Journal of Political Science 864.

Kritzer, H., and T. Uhlman (1977). "Sisterhood in the Courtroom: Sex of Judge and Defendant in Criminal Case Disposition," 14 Social Science Journal 77.

Landers, R. (2005). "O'Connor: Roberts 'Good in Every Way, Except He's Not a Woman,"' Spokesman Review [Spokane], July 20, p. Al.

Lawless, J. L. (2004). "Politics of Presence? Congresswomen and Symbolic Representation," 57 Political Research Quarterly 81.

Martin, E. (1993). "The Representative Role of Women Judges," 77 judicature 166.

Martin, A. D., and ?. M. Quinn (2005). "Can Ideal Point Estimates be Used as Explanatory Variables?" Unpublished working paper.

- (2002). "Dynamic Ideal Point Estimation via Markov Chain Monte Carlo for the U.S. Supreme Court, 1953-1999," 10 Political Analysis 134.

Martin, E., and B. Pyle (2005). "State High Courts and Divorce: The Impact of Judicial Gender," 36 University of Toledo Law Review 923.

- (2000). "Gender, Race and Partisanship on the Michigan Supreme Court," 63 Albany Law Review 1205.

McCall, M. (2008). "Structuring Gender's Impact: Judicial Voting Across Criminal Justice Cases," 36 American Politics Research 264.

- (2003). "Gender, Judicial Dissent, and Issue Salience: the Voting Behavior of State Supreme Court Justices in Sexual Harassment Cases, 1980-1998," 40 Social Science Journal 79.

McCall, M., and M. A. McCall (2007). "How Far Does the Gender Gap Extend? Decision Making on State Supreme Courts in Fourth Amendment Cases, 1980-2000," 44 Social Science Journal 67.

Menkel-Meadow, C. (1985). "Portia in a Different Voice: Speculations on a Woman's Lawyering Process," 1 Berkeley Women's Law Journal 39.

O'Connor, K., and J. A. Segal (1990). "Justice Sandra Day O'Connor and the Supreme Court's Reaction to Its First Female Member," 10 Women and Politics 95.

Palmer, B. (2002). "Justice Ruth Bader Ginsburg and the Supreme Court's Reaction to Its Second Female Member," 24 Women and Politics 1.

Peresie, J. L. (2005). "Female Judges Matter: Gender and Collégial Decision Making in the Federal Appellate Courts," 114 Yale Law Journal 1759.

Rohde, D. W., and H. J. Spaeth (1976). Supreme Court Decision Making. San Francisco: W H. Freeman.

Scheb, J. M. II, T. Bowen, and G. Anderson (1991). "Ideology, Role Orientations, and Behavior in the State Courts of Last Resort," 42 Western Political Quarterly 427.

Segal, J. ?., and H. J. Spaeth (2002). The Supreme Court and the Attitudinal Model Revisited. Cambridge: Cambridge University Press.

- (1993). The Supreme Court and the Attitudinal Model. New York: Cambridge University Press.

Shapiro, R. Y., and H. Mahajan (1986). "Gender Differences in Policy Preferences: A Summary of Trends from the 1960s to the 1970s," 50 Public Opinion Quarterly 42.

Sherry, S. (1986). "Civic Virtue and the Feminine Voice in Constitutional Adjudication," 72 Virginia Law Review 543.

Songer, D., and ?. A. Crews-Meyer (2000). "Does Judge Gender Matter? Decision Making in State Supreme Courts," 81 Social Science Quarterly 750.

Songer, D. R., S. Davis, and S. Haire (1994). "A Reappraisal of Diversification in the Federal Courts: Gender Effects in the Courts of Appeals," 56 Journal of Politics 425.

Swers, M. L. (2002). The Difference Women Make: The Policy Impact of Women in Congress. Chicago: University of Chicago Press.

Walker, T., and D. Barrow (19850. "The Diversification of the Federal Bench: Policy and Process Ramifications," 47 Journal of Politics 596.

West, R. (1991). "Jurisprudence and Gender." In ?. T. Bartlett and R. Kennedy (eds.), Feminist Legal Theory: Readings in Law and Gender, pp. 201-34. Boulder, CO: Westview Press.

Westergren, S. (2004). "Gender Effects in the Courts of Appeals Revisited: The Data Since 1994," 92 Georgetown Law Journal 689.

APPENDIX

Sex discrimination includes cases involving discrimination based on sex.

Reproductive rights includes cases involving abortion and contraceptives.

Civil rights include cases involving constitutional questions regarding the right to vote; desegregation; employment discrimination on the basis of race, age, or working conditions; and reapportionment (other than plans governed by the Voting Rights Act).

Economic activity includes cases involving anti-trust; mergers; bankruptcy; sufficiency of evidence; governmental liability; liability for nongovernmental entities; liability for punitive damages; Employee Retirement Income Security Act; federal regulation of securities; environmental protection of natural resources; corruption; zoning; arbitration; federal consumer protection; federal public utilities regulations; and miscellaneous economic regulation.