Women Lawyers before the Supreme Court of Canada

Women Lawyers before the Supreme Court of Canada ERIN B. KAHENY University of Wisconsin-Milwaukee JOHN J. SZMER University of North Carolina at Charlotte TAMMY A. SARVER Benedictine University American and Canadian scholars have found significant evidence of gender bias in the legal system, including the legal profession ~ABA, 2007; CBA, 1993; Foot and Stager, 1990; Kay and Brockman, 2000; Lopez, 2008; Noonan et al., 2005; Rhode, 2002: 1001!. Interestingly, scholars studying lawyer gender bias in the US have recently inquired as to whether gender bias against women lawyers has crept into the decisional processes of the US Supreme Court ~USSCT!. Specifically, a study conducted by Szmer and colleagues ~2010! indicates that women lawyers are not disadvantaged in finding success on case outcomes but does suggest that conservative justices are less likely to side with parties represented by women attorneys. On the other hand, the study also concluded that the justices are more likely to side with the position advocated by women lawyers in cases raising “women’s issues” ~Szmer et al., 2010: 27!. Acknowledgments: We wish to thank Frank Collucci and the anonymous reviewers for CJPS for their comments and suggestions for improving the manuscript and Donald Songer for providing much of the data employed in the study. We also thank Jacqueline Baldwin for providing a French translation of the abstract. Erin B. Kaheny, Department of Political Science, University of Wisconsin, Milwaukee, PO Box 413, Milwaukee, WI 53201, phone: ~414! 229-6560, fax: ~414! 2295021, ekaheny@uwm.edu John J. Szmer, Department of Political Science, University of North Carolina at Charlotte, 9201 University City Blvd., Charlotte, NC 28223, phone: ~704! 687-3941, fax: ~704! 687-3497, jjszmer@uncc.edu Tammy A. Sarver, Department of Political Science, Benedictine University, 5700 College Road, Lisle, IL 60532, phone: ~630! 829-6473, fax: ~630! 829-6231, tsarver@ ben.edu Canadian Journal of Political Science / Revue canadienne de science politique 44:1 (March/mars 2011) 83–109 doi:10.10170S000842391000106X © 2011 Canadian Political Science Association ~l’Association canadienne de science politique! and0et la Société québécoise de science politique 84 E.B. KAHENY, J.J. SZMER, AND T.A. SARVER Given the similarities between the USSCT and the Supreme Court of Canada ~SCC! and the similar history of gender inequities in the legal professions of both countries, we test whether lawyer gender has the same influence on the SCC. We examine the impact of female lawyers on SCC decision making for several reasons. First, it is important to assess whether the gender bias in the USSCT extends to a similarly situated common law high court. As Hausegger and Haynie ~2003: 635! note, comparative analyses of this sort will help test the breadth of judicial decision-making theories. Second, due to the historically greater proportion of women justices on the SCC, it becomes important to assess whether a more gender-balanced decision-making body collectively views women lawyers in a more favourable manner. Finally, it is crucial to compare how women, as advocates, shape ~or fail to shape! the policies made by both courts. Gender Bias as Sex Discrimination in the Legal Profession In approaching this topic, it is helpful to examine feminist scholarship that has sought to explain the persistence of gender discrimination in the legal profession and elsewhere. One school of thought suggests that disparate treatment of women in the political and legal realms results from the application of “gender schemas” ~Valian, 2005: 198!. “A schema” is “a cognitive structure that enables an individual to make sense of and impose structure on the many stimuli with which he or she is bombarded in the course of daily living” ~Matland, 1994: 278, citing Nisbett and Ross, 1980!. Specifically, a schema serves as “a mental representation of a category of objects, events, or persons and their typical characteristics” ~Haslett et al., 1992: 57!. Thus, an individual’s “gender schema” would involve those things deemed to be “‘the typical attributes’ of men and women” and it is through perceptions of such traits, that gender schemas can perpetuate “gender stereotypes” ~Haslett et al., 1992: 58, citing Taylor and Crocker, 1981 and Fiske, 1982!. As Valian ~1999! explains, individuals naturally assign different attributes, physical and mental, to men and women. Men are viewed as independent, strong, and rational; women are seen as communal, emotional, and nurturing. It is not hard to understand, therefore, how the application of gender schemas might ultimately lead to the “overrating” of men and the “underrating” of women ~Valian, 1999: 1045; Valian, 2005: 198!. Indeed, because individuals tend to form impressions about men and women from extreme cases, an individual’s view of the sexes can be quite polarized, super-intensifying schematic applications of gender attributes to the benefit of one and the detriment of the other ~Valian, 1999!. Abstract. Recent work by Szmer, Sarver, and Kaheny ~2010! exploring US Supreme Court decision making has suggested that lawyer gender might play a role in influencing judicial voting behaviour. Specifically, while women lawyers were not revealed to have a more difficult time winning cases before the US Supreme Court, the study did suggest they face a tougher challenge in gaining support from the more conservative justices on that bench. Here, we test whether women lawyers face similar challenges before the SCC. Our findings do not reveal any disadvantage for litigation teams with larger proportions of women and, in most instances, such teams have an advantage. Specifically, in our model of civil rights and liberties votes, litigation team gender had no bearing on individual SCC justice decisions. However, in a pooled model of all issues combined and in separate models of criminal and economic votes, SCC justices were more likely to side with litigation teams with larger proportions of women lawyers. Résumé. Une étude récente de Szmer, Sarver et Kaheny ~2010! explore la manière dont la Cour suprême des États-Unis prend ses décisions, suggérant que le sexe des avocats pourrait avoir une influence sur le comportement décisionnel des juges. Plus spécifiquement, bien que les avocates n’aient pas plus de difficulté que leurs collègues masculins à gagner leurs procès à la Cour suprême des États-Unis, l’étude suggère que leur plus grand défi est d’obtenir le soutien des juges plus traditionnels de cette cour. Dans le présent article, nous cherchons à déterminer si les avocates canadiennes font face à un défi semblable à la Cour suprême du Canada. Les résultats de notre étude ne révèlent aucun désavantage pour les équipes d’avocats comprenant plus de femmes et dans la plupart des cas, ces équipes bénéficient même d’un avantage. Plus précisement, dans notre modèle décisionnel en matière de droits et libertés civiles, le sexe des membres des équipes d’avocats n’avait aucune incidence sur les décisions individuelles des juges de la Cour suprême du Canada. Cependant, dans un modèle commun réunissant tous les types de dossiers et dans des modèles séparés pour les décisions sur des dossiers criminels et financiers, les juges de la Cour suprême du Canada étaient plus enclins à prendre parti pour des équipes comportant une plus grande proportion d’avocates. Gender schema theory has received empirical support in a variety of political contexts, including studies of voting behaviour ~Sanbonmatsu, 2002! and the perceptions of women politicians ~Kahn, 1996!. Most recently, the previously noted study of attorney effects on the USSCT ~Szmer et al., 2010: 28–29! suggested that one potential reason for its findings is that justices apply “gender schemas” when assessing the credibility of the information presented by male and female attorneys. Aside from this study, however, few scholars have examined whether appellate judge perceptions of lawyer arguments are affected by the application of gender schemas. Instead, prior studies of gender and the American and Canadian judiciaries have focused either on the sex of judges ~for example, Brockman, 1993; McCormick and Job, 1993; Peresie, 2005; Songer et al., 1994; Songer and Johnson, 2007; Wetstein and Ostberg, 2007!, or on trial attorney gender ~for example, Nelson, 2004!. In fact, we could not identify a study to date that has assessed the potential influence of lawyer gender on judicial decision making in Canada at any level. Building from prior scholarship on the USSCT ~Szmer et al., 2010!, we are unique in providing an analysis of the effect of female lawyers on SCC decision making. 86 E.B. KAHENY, J.J. SZMER, AND T.A. SARVER Lawyer Capability on the USSCT and the SCC An analysis of the effect of lawyer gender on SCC decision making must begin with a review of scholarship on the role of lawyer capability on the American and Canadian high courts. In the US, it is evident that USSCT justices rely on lawyers for information ~Corley, 2008; Johnson, et al., 2006; McGuire, 1995!. Presumably, the ability to persuasively convey information to justices varies across lawyers. Moreover, the credibility justices attach to lawyers also varies. Prior studies have identified characteristics that reflect both lawyer credibility and persuasiveness. Most prominently, USSCT studies ~for example, McGuire, 1995! consistently find that lawyers with more prior litigation experience are more likely to influence the justices’ decisions. Similarly, there is evidence that former USSCT law clerks are more successful litigators ~McGuire, 2000!. Research has also shown that an array of lawyer characteristics including experience, clerkship and level of government employment influenced Justice Blackmun’s informal assessments of the lawyers’ performances during oral arguments and, in turn, those assessments partially explain the other justices’ votes ~Johnson et al., 2006!. This line of research has been extended to the SCC. Flemming ~2004! models lawyer leave to appeal activity, finding that honorary Queen’s Counsel ~QC! designation, appellate expertise and prior success on the merits in SCC litigation all help explain the degree to which lawyers participate in leaves to appeal. Two SCC studies of the impact of lawyer capability on decisions to grant leaves found little evidence of the impact of lawyer experience or QC designation on the Court’s decision ~ Flemming, 2004; Flemming and Krutz, 2002!. However, more recently, Szmer and colleagues ~2007! demonstrate that both experience and litigation team size affect the merits decisions of SCC justices, though QC designation does not. Thus, it appears as if the lawyer also matters in SCC litigation. What is less clear is whether the gender of the lawyer matters. As noted above, a recent USSCT study found that, although the sex of the orally arguing lawyer or the gender composition of the larger litigation team was not related to case outcomes, conservative justices were less likely to vote for litigants represented by women lawyers at oral arguments ~Szmer et al., 2010!. Moreover, justices were more likely to vote in favour of women advocates ~in male versus female contests! in cases raising women’s issues. These relationships subsisted after controlling for an array of factors including attorney capability, the ideological direction of the attorney’s arguments and justice gender. While the study could not identify the precise causal mechanism, the authors suggested the use of gender schemas by political elites in a manner that could either hinder or help women advocates. Women Lawyers before the Supreme Court of Canada 87 Lawyer Gender Hypotheses Women in the US and Canada have repeatedly experienced gender inequities in their efforts to fully participate in the legal profession. This might also include women barristers litigating before the SCC. It is possible that the justices apply gender schemas that affect how they weigh the credibility of arguments presented by women. This theory, combined with the finding that arguments by female attorneys are devalued by at least a subset of the USSCT, a court similar in a number of ways to the SCC, suggests that SCC justices might be less likely to vote in favour of parties represented by women. However, there is at least one significant difference between the USSCT and the SCC: the latter’s gender diversity. While only four women have served on the USSCT, seven have served on the SCC. Presently, of the nine members of the SCC, four are women, including Chief Justice Beverley McLachlin. Perhaps the presence of more women on the SCC bench indicates that gender bias may not affect judicial decision making to the same extent there as it does in the American context. Consequently, it is possible that this enhanced gender diversity has mitigated any bias that might otherwise affect the evaluations of the legal arguments of female lawyers. As Powell observes, “over time, being supervised by a woman or simply working with women seems to affect attitudes toward them” ~1993: 110!. This was seen, Powell notes, in a study of firefighters. As men in the study gained more working experience with females, not only did they hold their female peers in higher regard, but they also held “less traditional ... attitudes toward women in general” ~Powell, 1993: 110, reviewing Craig and Jacobs, 1985!. Moreover, Powell ~1993! mentions the importance of Kanter’s ~1977a; 1977b! work in understanding how the gender composition of a given group might influence the use of gender stereotypes within that group. Kanter ~1977a, 1977b: 966!, of course, discussed a number of key problems confronted by “token” members in groups. However, she also stressed how the proportion of token members in a group can shape how such individuals are perceived. “It is easier,” Kanter argues, “to retain the generalization and distort the perception of the token” when group membership is highly unbalanced ~Kanter, 1977b: 972!. This suggests that familiarity with individuals who fall into a token group, along with the presence of a larger proportion of tokens in a given group ~that is, greater diversity!, might reduce negative stereotyping ~Powell, 1993: 112!. In addition, although scholars have not directly tested whether greater gender diversity on a court influences a judge’s perception of the arguments presented by female lawyers, some studies suggest that diversity on the bench can contribute to differences in a judge’s voting behaviour. Recent studies have found that male appellate judicial behaviour is affected 88 E.B. KAHENY, J.J. SZMER, AND T.A. SARVER by the increased presence of female colleagues ~for example, Boyd et al., 2010; Peresie, 2005!. Thus, depending on whether SCC voting trends follow those which describe more conservative members of the USSCT or whether greater gender diversity has successfully mitigated negative stereotypes, we expect to see one of two possibilities in the present study: H1: A greater proportion of women lawyers on the plaintiff/prosecution’s team relative to the defendant’s team will have a negative or no effect on the likelihood that a SCC justice will vote in favour of the plaintiff/ prosecution’s team. In addition, one should weigh the likelihood that the influence of lawyer gender will be contingent upon the gender of the justice assessing the lawyer’s argument. It is possible, for example, that women are less likely to apply gender schemas in a negative fashion toward other women. Therefore, if female lawyers do confront negative gender bias when litigating before the SCC, this bias might be less evident or not evident in the voting decisions of its female justices. Thus, we also test a second hypothesis: H2: A greater proportion of women lawyers on the plaintiff’s/prosecution’s team relative to the defendant’s team will not negatively affect the likelihood that a female SCC justice will vote in favour of the plaintiff/ prosecution. In their analysis of lawyer gender bias before the USSCT, Szmer and colleagues ~2010! argued ~and found! a female lawyer’s prospects of securing a favourable vote by a justice might be conditioned on the justice’s ideology. For example, liberal justices might also tend to apply gender schemas in ways that are not as disadvantageous toward women and, thus, might also be more receptive to the nature of their legal arguments relative to their conservative colleagues. This possibility leads to the generation of a third hypothesis: H3: Conservative justices will be less likely to vote in favour of the plaintiff/prosecution if it is represented by more women lawyers relative to the defendant, whereas liberal justices will be inclined to offer further support or, at the very minimum, will not be less likely to vote in favour of the plaintiff/prosecution. Beyond these issues, the impact of attorney gender might be conditioned upon the type of issue presented in the case by way of “issue stereotypes” ~Dolan, 2004: 78!. Kahn ~1996!, for example, notes that individuals hold stereotypes regarding the competency of women in certain issues, often called “women’s” or “female” issues. Presumably, to the extent a judge shares these stereotypes, the judge will attach greater credibility to information presented by women in cases involving women’s Women Lawyers before the Supreme Court of Canada 89 issues. Since we believe lawyer impact is a function of credibility, we expect to find that women lawyers are more successful in cases involving women’s issues. H4: Representation by proportionally more women lawyers than their opponents in cases involving women’s issues will positively affect the likelihood that a SCC justice will vote in favour of the plaintiff/prosecution. How a justice responds to arguments presented by women lawyers might also be moderated by panel gender diversity. Specifically, as the proportion of women sitting with the judge on the panel increases, we might expect the judge to be more receptive to arguments made by women. Indeed, Carli reports “both men and women show a variety of less sex-stereotyped behaviors when interacting with those of the opposite gender” ~1989: 573!. If the gender composition of a group can affect differences among its participants, might a mixed-sex panel mitigate negative gender-based perceptions of those women lawyers who argue before it? In the judicial scholarship, there is evidence that male jurists on panels with women adjust their behaviour. Gryski and colleagues ~1986!, for example, find that the presence of just one female judge on a court influences the male judges’ behaviour. More recently, Farhang and Wawro ~2004! find that the presence of one female judge on a US Courts of Appeals panel can affect the decision making of male judges in employment discrimination cases. If judges’ voting behaviour is being influenced by the presence of a female on the bench, it is also possible that such judges’ perceptions of female counsel may be altered in a positive fashion. This behaviour, moreover, might not be restricted to male judges. Since the SCC often hears cases in smaller five- and seven-judge panels, there is significant variation in panel gender composition, thus providing a good context to test the following hypothesis: H5: As the proportion of female justices sitting with a justice on a panel increases, a justice will be more likely to side with plaintiff/prosecutions who are represented by proportionally more women lawyers than their opponents. Finally, it is possible that justices who are more likely to hire women law clerks are more likely to respond favourably to arguments made by women barristers. First, the decision to hire a woman clerk might reflect the judge’s perception of the competency of women lawyers. However, even if such hiring decisions are a function of other factors ~for example, political reasons!, experience over time with many highly competent female clerks could help alleviate gender stereotypes. As such, we hypothesize: 90 E.B. KAHENY, J.J. SZMER, AND T.A. SARVER H6: SCC justices with a higher proportion of former women law clerks will be more likely to vote in favour of the plaintiff/prosecution when the plaintiff/prosecution is represented by a larger proportion of women barristers than the defendant. Data and Methods The data for the analysis include all SCC appeals from 1988–2000.1 The case information comes from the Canada Supreme Court Reports (CSCR).2 We incorporated in the model all non-reference cases from 1988–2000 in which an ideological direction could be determined. The unit of analysis is the justice’s vote in the case. In all, 1,393 of the 8,340 justices’ votes in non-reference cases were excluded from the analysis because of missing data. Since we are concerned with the influence of lawyer gender on the behaviour of individual justices, the dependent variable is the justice’s decision for or against the plaintiff0prosecution. Additionally, while many prior studies of lawyer capability model a judge’s vote for or against the appellant ~for example, Johnson et al., 2006; McAtee and McGuire, 2007!, we believe that is inappropriate in this context because it would not allow us to control for the tendency of the SCC to affirm ~Songer, 2008!. This is confounded by the slightly greater propensity for women lawyers to represent the respondent ~on average, 15 per cent of the appellants’ lawyers were women, compared to 19 per cent of the respondents’ lawyers!. As such, we follow the lead of Haire and colleagues ~1999!, who also studied a court with a pronounced tendency to affirm ~the US Courts of Appeals!. Thus, our dependent variable is dichotomous, coded ‘1’ if the justice’s vote in the case supported the plaintiff0prosecution at trial and ‘0’ if the justice sided with the defendant.3 Because the dependent variable is dichotomous, we employed logistic regression, with standard errors clustered by the justice to account for possible interdependent error terms resulting from repeated votes by the same justices. Independent variables There are many alternative methods for measuring lawyer gender, including the gender of the orally arguing lawyers, the total number0proportion of women lawyers representing a litigant and a dichotomous measure of the presence or absence of at least one woman on the litigation team. The first measure is likely the best. When more than one lawyer represents the litigant, the orally arguing lawyer clearly plays a significant role. We do not have the same level of certainty regarding the role of the other lawyers. Unfortunately, the SCC does not maintain records of the lawyers who present oral arguments in prior cases nor is this information Women Lawyers before the Supreme Court of Canada 91 in the CSCR. Court opinions, however, do list the names of the lawyers, which we used to determine the lawyers’ genders. Thus, absent a better alternative, we employ the proportion of women lawyers representing the parties to assess lawyer gender. Given our interest in exploring the potential role of gender schemas in this analysis, our choice was to utilize a measure that would be the most reasonable surrogate for the sex of the orally arguing lawyer. While we cannot correlate the two variables, more than half of the litigation teams in our sample were represented by only one lawyer. As such, at least in those cases, the proportion of women lawyers on the team will reflect whether the orally arguing lawyer was a woman. Moreover, it is likely that this construct, more than the sex of the orally arguing barrister alone, best captures the influence of a lawyer’s sex on the factums.4 The actual variable, lawyer team gender, was constructed to account for the adversarial nature of lawyer influence, as well as the nature of the dependent variable. As such, this measure is the difference in the proportion of women lawyers representing the plaintiff0 prosecution and the defendant. As discussed above, we expect a negative or no relationship between this variable and the likelihood of a proplaintiff0prosecution vote. Our analysis also tests for the likelihood that the impact of lawyer gender on judicial decisions is conditional on the gender of the justices. To test this, we include a variable that is the product of the lawyer team gender variable and the justice gender variable, the latter of which is coded ‘1’ if the justice is a woman and ‘0’ if the justice is a man. Consequently, we expect a positive coefficient on this multiplicative term. We also posit that women have a more significant impact in cases involving women’s issues. Prior studies have described women’s issues as those that involve questions of equality and0or the private sphere—for example, family and marriage ~Kahn, 1996; Segal, 2000!. As such, the women’s issue variable is coded ‘1’ if the case involves one of the following types of issues: rape, equality, privacy, marriage, divorce or other family related issues. To examine the potential conditional relationship between lawyer gender and women’s issues, we constructed a multiplicative term, lawyer team gender * women’s issues. We expect to find a positive coefficient on this term. Since justices who hire more women law clerks might be less likely to devalue arguments by women, we include the lawyer team gender * proportion of women clerks multiplicative term. The proportion of women clerks variable was constructed using a SCC clerk list; it is the proportion of women clerks employed by the justice in prior terms. We expect a positive coefficient on the multiplicative term. We also posit that the presence of women on the panel with the justice could influence his or her support for women lawyers. As such, we include a lawyer team gender * panel gender multiplicative term. The 92 E.B. KAHENY, J.J. SZMER, AND T.A. SARVER panel gender variable is the percentage of women justices sitting on the panel with the justice casting the vote. We expect a positive coefficient for the multiplicative term. We also suggest the impact of lawyer gender might be conditioned by justice ideology. As such, we incorporate a multiplicative term which is the product of lawyer team gender and justice ideology. The ideology measure was constructed by Ostberg and Wetstein ~2007! who, following the method employed by Segal and Cover ~1989! to estimate USSCT justice ideology, used nomination-period editorials in a sample of nine Canadian newspapers to estimate SCC justice preferences. Higher values of the ideology variable indicate increased liberalism and, thus, we expect a positive coefficient for the multiplicative term. Ideology is also an important control variable ~Ostberg and Wetstein, 1998, 2007!. Since the dependent variable is the justice’s vote for or against the plaintiff0prosecution, the measure ~ justice/plaintiff ideological congruence! estimates the congruence of the justice’s preferences ~using the Ostberg and Wetstein scores! with the ideological direction of the preferred outcome of the party that initiated the litigation at trial. If the plaintiff0prosecution preferred a liberal outcome, the variable is the justice’s Ostberg and Wetstein score ~where higher values indicate increasing liberalism!. If the plaintiff0prosecution preferred a conservative outcome, the variable is the negative of the justice’s ideology score; higher values indicate increased congruence. Many prior studies have used lawyer prior litigation experience to explain the behaviour of the courts and0or judges serving on them ~see, for example, Flemming, 2004; Haire et al., 1999; McGuire, 1995; Szmer et al., 2007!. Furthermore, lawyers with more experience are more likely to present quality oral arguments before the USSCT ~Johnson et al., 2006!, and the justices are more likely to incorporate into their opinions language from the more experienced attorneys’ briefs ~Corley, 2008!. To construct the litigation experience variable, we first calculated the total number of times each lawyer participated in a case litigated before the SCC in the previous 10 terms. To reflect the entire litigation team ~and because we could not identify the orally arguing lawyers!, we then took the mean value for the lawyers representing the plaintiff0prosecution and defendant. Because there is a diminishing marginal utility of prior litigation experience, we logarithmically transformed each mean.5 Due to the adversarial nature of the judicial process, the actual litigation experience variable is the difference between the values for the plaintiff0prosecution and defendant. Positive scores should be associated with a greater likelihood of a pro-plaintiff0prosecution vote. We included two variables which also reflect lawyer expertise via prior clerkships with the SCC: plaintiff lawyer clerkship and defendant lawyer clerkship.6 The SCC clerks, like their counterparts at the USSCT, Women Lawyers before the Supreme Court of Canada 93 provide research assistance for a justice, where they develop an insider’s understanding of the SCC’s processes. Of course, a prior clerkship likely reflects more than expertise; it also reflects the scholastic achievements of the clerk while in law school ~Western Law Career Services Office, 2005!. Given the former clerks’ expertise and superiour law school achievements, we expect that lawyers with prior clerkships are more likely to influence the justices. For the purposes of our analysis, the actual concept of prior clerkship experience is operationalized as two separate dummy variables, plaintiff lawyer clerkship and defendant lawyer clerkship. We assert the former will have a positive relationship with the dependent variable while the latter will exhibit a negative relationship. Contrary to conventional wisdom, it is possible that more lawyers are a good thing—at least for the party they are representing. Larger litigation teams are advantageous because they can increase the time spent on research, attack problems from additional perspectives and present and critique mock oral arguments. In this respect, we might argue that litigation team size is an important indicator of litigation advantage ~see Galanter, 1974!. Indeed, prior studies ~see, for example, Szmer et al., 2007! found significant relationships between litigation team size and judicial decisions. The litigation team size variable is the difference in the number of lawyers representing the plaintiff0prosecution and the defendant. We expect a positive coefficient. All parties are not treated equally by the courts, and there is evidence that this inequity extends to the SCC ~McCormick, 1993; Szmer et al., 2007!. As such, we include measures of party capability, versions of which have been incorporated into several prior studies of common law appellate courts ~for example, McCormick, 1993; Songer and Sheehan, 1992!. In the absence of direct measures ~for example, financial resources; prior litigation experience!, these studies used the organizational characteristics of the parties to indirectly assess capability. Although they utilized scales which order litigant categories based on assumptions regarding experience and resources, since such scales impose a structure that might not adequately capture litigant advantages before the Canadian high court, we include dummy variables to describe the category of the plaintiff0prosecution and defendant types in our study: associations,7 businesses, local governments, provincial governments and the national government. The reference ~excluded! dummy variables are individual plaintiff and individual defendant. ~However, in our model examining only criminal cases, where a natural person typically faces off against the government, we examine only such cases and, hence, do not include party capability variables!. Here, it is on safe ground to posit that plaintiffs0 prosecutions who are not individuals are more likely to have greater resources than plaintiffs0prosecutions who are individuals and thus we expect positive and statistically significant coefficients on all of the 94 E.B. KAHENY, J.J. SZMER, AND T.A. SARVER included plaintiff litigant dummy variables. On the flip side, we can safely assume that defendants who are groups, businesses or governmental entities tend to have greater litigant resources than individual defendants. Greater resources in the hands of defendants should translate into a lower likelihood of plaintiff0prosecution success ~the dependent variable! and, therefore, we expect negative coefficients on the included defendant party capability variables. Since SCC justices tend to vote to affirm the decision of the lower court ~Songer, 2008!, all things being equal, a justice is more likely to side with the plaintiff0prosecution if it is the respondent on appeal. To account for this, we include the plaintiff appellant dummy variable ~‘1’ if the party that initiated the litigation at the trial level also initiated the appeal, ‘0’ if they were the respondent in the case!. Given our theory, we expect to find a negative coefficient. Results Because the findings may be conditional on the subject of the litigation, we present our models across all issue areas combined ~with and without hypothesized conditional relationships! and separately ~with hypothesized conditional relationships, where applicable! across criminal, economic0 labour and civil rights0liberties cases. The results are presented in Table 1. Clearly, the most interesting result across all models is the positively signed coefficient on the lawyer team gender variable; the variable is statistically significant across the two models involving all issue areas combined ~models 1 and 2! and in the separate model of votes cast in criminal cases ~model 3!. It is also significant, two-tailed, at the 0.10 level in the economic issues model ~model 4!. Therefore, plaintiffs who employ a larger proportion of women on their litigation teams relative to the proportion employed by defendants are more likely to receive a favourable vote by a Canadian justice, though this tendency is more pronounced in the criminal case context. In the civil rights and liberties issue area, on the other hand, the coefficient on the lawyer team gender variable does not approach conventional levels of significance ~see model 5!. Thus, in this area, we see neither an advantage nor a disadvantage for litigation teams with a higher proportion of women lawyers. As noted above, although we did not expect to find a strong, positive result with respect to this variable, the lack of a negative relationship is not necessarily a surprising finding and, in fact, is in line with expectations concerning the role of gender diversity on this Court. Indeed, one possibility is that the greater presence of female jurists on the SCC has minimized the negative application of gender schemas in interactions with female lawyers and0or has mitigated other forms of gender bias in the assessment of Women Lawyers before the Supreme Court of Canada 95 their legal arguments. In terms of the criminal issue area, however, we are tapping something other than the mitigation of negative gender stereotypes, since teams with a greater proportion of women criminal lawyers relative to their opponents tend to have the upper hand. The results also suggest that the influence of lawyer team gender is not conditional on a justice’s gender. None of the models including the relevant multiplicative term ~models 2–5! suggest a statistically significant relationship. Thus, women SCC justices are not more likely to support women counsel than their male counterparts. Moreover, there is little evidence to suggest the influence of lawyer gender is conditioned on the ideology of a SCC justice. Although the coefficient of the relevant multiplicative term is positive where included, it is statistically significant in criminal cases ~model 3! only. In this context, the results suggest liberal justices might be more likely than their conservative colleagues to support the position of the prosecution on appeal, if the prosecution consists of a greater proportion of women lawyers. Unlike previous USSCT studies, our results provide mixed support for the notion that justices are more likely to side with women lawyers in “women’s issue” cases. One area where this type of relationship does seem apparent, however, is in criminal cases; parties with greater proportions of female lawyers have an even more pronounced advantage in criminal cases raising a women’s issue. On the other hand, in civil rights0 liberties cases, the coefficient on the lawyer team gender * women’s issues multiplicative term is statistically significant and negative; thus, greater numbers of female counsel do not translate into more success in persuading individual justices. Rather, we see the opposite tendency.8 It is important to note that these findings do not appear to be a function of the tendency on the part of women counsel to represent the liberal position before the SCC. Indeed, in the data used for this study, women lawyers, overall, are more likely to represent the conservative side ~see appendix at http:00pantherfile.uwm.edu0ekaheny0www0! and, in the women’s issues cases we model, the difference is rather striking. On average, 21 per cent of the lawyers on the liberal side are women, compared to 34 per cent on the conservative side. Moreover, we found no support for our hypothesis that a justice’s inclination to hire female clerks positively affects the justice’s decision to support teams with a higher proportion of female lawyers. As seen across all models in which this conditional relationship was assessed ~models 2–5!, the coefficient on the associated multiplicative term is negative and, thus, is not in the hypothesized direction. Of course, the proportion of women sitting on a given panel in this dataset varies over time, and this proportion, we argued above, could also influence an individual SCC justice’s decision to support litigation teams with higher proportions of female counsel. In the Canadian context, however, there is mixed 96 TABLE 1 Logistic Regression Models of the Likelihood of a Pro-Plaintiff Vote by a SCC Justice, 1988–2000 ~Robust Standard Errors Clustered on Judge in Parenthesis! 0.20~0.03!⫹ — — — — 0.38~0.11!⫹ ⫺0.08~0.06! 0.03~0.05! ⫺0.01~0.23! ⫺0.34~0.27! — 0.12~0.07!* 0.28~0.13!* 0.11~0.04!** 0.03~0.06! 0.12~0.11! 0.63~0.13!⫹ 0.03~0.02! 0.12~0.02!*** — 0.16~0.14! ⫺0.34~0.07!⫹ ⫺1.31~0.21!⫹ 0.38~0.13!** 0.04~0.09! ⫺1.48~0.19!*** ⫺0.37~0.14!** ⫺0.49~0.11!*** 0.02~0.06! 0.16~0.11! 0.64~0.13!⫹ 0.03~0.02! 0.12~0.02!*** 0.16~0.11! 0.16~0.14! ⫺0.35~0.07!⫹ ⫺1.29~0.21!⫹ 0.42~0.13!*** 0.03~0.09! ⫺1.43~0.18!*** ⫺0.36~0.15!** ⫺0.51~0.11!*** Model 3 CRIM Model 4 ECON Model 5 CIVRTS 0.59~0.34! ⫺0.28~0.25! 0.14~0.15! — ⫺1.38~0.61!⫹ 0.63~0.94! 0.34~0.39! 0.08~0.23! ⫺0.67~0.32!⫹ ⫺0.56~1.85! 1.54~0.25!*** 0.04~0.22! ⫺1.77~0.35!⫹ 0.21~0.05!*** 0.67~0.85! ⫺0.08~0.13! ⫺0.24~0.06!*** 0.00~0.12! 0.91~0.19!⫹ 0.05~0.02!* 0.18~0.03!*** 0.20~0.19! — — — — — — — — 0.16~0.05!⫹ 0.37~0.16!** 0.59~0.14!⫹ 0.10~0.04!* 0.02~0.05! ⫺0.12~0.13! 0.10~0.23! ⫺0.49~0.08!⫹ ⫺0.14~0.48! 0.93~0.19!*** ⫺0.39~0.14!⫹ ⫺1.47~0.31!*** ⫺0.23~0.12!* ⫺0.08~0.20! 0.63~0.12!⫹ 0.02~0.06! 0.10~0.02!*** 2.23~0.92!** ⫺0.26~0.29! ⫺0.06~0.18! 0.12~0.19! 1.35~0.35!⫹ ⫺0.12~0.07! 0.14~0.07!* 0.37~0.21! 0.63~0.21!** 0.74~0.16!*** ⫺16.19~0.40!⫹ ⫺0.18~0.20! 0.59~0.18!*** 0.27~0.61! 1.71~0.53!⫹ 1.15~0.46!⫹ T.A. SARVER Lawyer Team Gender Lawyer Team Gender * Justice Gender Lawyer Team Gender * Justice Ideology Lawyer Team Gender * Women’s Issue Lawyer Team Gender * Proportion Women Clerks Lawyer Team Gender * Panel Gender Justice0Plaintiff or Pros. Ideological Congruence Plaintiff0Prosecution Appellant Plaintiff0Prosecution Lawyer Clerkship Defendant Lawyer Clerkship Litigation Team Size Lawyer Experience Justice Gender Plaintiff Association Plaintiff Business Plaintiff Local Govt. Plaintiff Prov’l Govt. Plaintiff Nat’l Govt. Defendant Association Defendant Business Defendant Local Govt. Model 2 AND Model 1 E.B. KAHENY, J.J. SZMER, Independent Variable ⫺0.84~0.09!*** ⫺0.99~0.09!*** — — — — 0.65~0.09!*** 6,958 64.6 0.082 ⫺0.83~0.09!*** ⫺0.98~0.09!*** ⫺0.00~0.06! ⫺0.28~0.12!⫹ ⫺0.04~0.38! 1.27~0.30!*** 0.33~0.21! 6,958 65.0 0.093 — — ⫺0.16~0.21! ⫺0.19~0.29! ⫺0.02~0.59! 1.50~0.37!*** 0.31~0.32! 4,302 67.6 0.006 ⫺0.97~0.18!*** ⫺0.68~0.11!*** 0.09~0.04!* — ⫺0.43~0.41! 0.53~0.72! 0.49~0.27! 1,665 61.7 0.188 0.93~0.44!⫹ 0.88~0.44!⫹ 0.06~0.07!⫹ ⫺0.37~0.22! 0.88~0.40!* 3.80~1.19!* ⫺2.58~0.56!*** 664 64.5 0.269 p ⱕ 0.05 ⫽ *; p , 0.01 ⫽ **; p ⱕ 0.001 ⫽ ***; ⫹significant at p ⱕ 0.05 but sign of coefficient is opposite to that hypothesized. ~Tests are one-tailed except of those performed on the Justice Gender, Justice Ideology, Proportion of Women Clerks, and Women’s Issue variables as well as the constant!. Women Lawyers before the Supreme Court of Canada Defendant Prov’l Govt. Defendant Nat’l Govt. Justice Ideology Women’s Issue Proportion Women Clerks Panel Gender Constant Number of Observations % Predicted Correctly PRE 97 98 E.B. KAHENY, J.J. SZMER, AND T.A. SARVER support for the proposition. In model 2 ~all issues combined! and model 3 ~votes cast in criminal cases!, the coefficient on the multiplicative term, lawyer team gender * panel gender, suggests the possibility of this type of conditional relationship. On the other hand, the multiplicative term fails to reach significance in the civil rights model ~model 5! and, if anything, suggests the opposite relationship in economics cases. Next, we briefly discuss our controls, as they too reveal interesting findings. First, we find a positive, statistically significant relationship between the ideological congruence term and the dependent variable when examining votes cast across all issue types as well as economic cases. In these contexts, justices are more likely to side with the plaintiff if their ideological preferences favour the plaintiff ’s preferred position. In criminal cases, the coefficient is positive but is not statistically significant, suggesting that ideological considerations weigh less in this area. Indeed, in criminal cases, the results suggest that lawyer characteristics, including a litigation team’s size, experience and gender composition play more of a role than ideology. On some levels, this is surprising given previous research demonstrating significant effects of ideology on SCC decisions ~Ostberg and Wetstein, 2007!. However, the role of lawyer gender was not examined in previous models of SCC justice votes and, as noted here, its influence is particularly pronounced in criminal cases. As its influence is less pronounced in civil rights and liberties cases, our findings suggesting a nonsignificant influence for judicial ideology in this issue area are reasonable considering previous work ~see, for example, Ostberg and Wetstein, 2007!. Litigant resources also appear to influence SCC justice votes, but the results suggest caution when assuming which litigant types tend to be more advantaged. As seen across the models, the relative advantages of a given litigant type are conditional on the issue area. For example, we hypothesized that individual plaintiffs would be at a greater disadvantage when confronting governments and, thus, expected negative and statistically significant coefficients for all defendant government variables. However, while the provincial and national governmental defendant variables performed as expected in most instances, they did not perform as hypothesized in the civil rights0liberties model ~model 5!. Here, the coefficients on both of these variables were positive and significant. The local government defendant variable, moreover, was positively signed in the civil rights model as well. These results suggest that scholars should be careful in utilizing the traditional ~US! party capability scale in studies of the SCC.9 Indeed, certain litigant categories may not be advantaged or disadvantaged in the same way in Canada, and the extent to which party resources translate into judicial success appears contingent on the issue area.10 The results also suggest that controls for both lawyer capability and ideological congruence could affect Women Lawyers before the Supreme Court of Canada 99 the observed impact of party capability. Indeed, this is not surprising given that many ~see, for example Galanter, 1974; McCormick, 1993! posit that the party capability advantage at least in part stems from the ability to use superiour resources to hire expert legal counsel. This is also supported by McGuire’s ~1998! finding that the apparent advantage of the US national government in USSCT litigation disappears after controlling for lawyer experience. Finally, Sheehan and colleagues ~1992! suggest, at least in the context of the USSCT, observed party capability advantages are a function of the ideological congruence of the justices with typically advocated legal positions by certain classes of litigants. Our results with respect to our lawyer controls are also notable. First, contrary to our hypothesis, the data do not suggest a decisive advantage for plaintiffs represented before the SCC by a former clerk. Although the coefficient is positive as hypothesized, it reached significance only in the model of votes cast in economic cases. Similarly, if this particular experience should give defendants represented by former clerks an advantage, we should see plaintiffs more likely to lose when facing a former clerk. In fact, that is not what happens, according to the results in Table 1. A plaintiff ’s chances of securing a favourable vote by a SCC justice increases in all models if they confront a former clerk among the defense team, save in the context of civil liberties0rights disputes ~model 5!. Such results are at odds with the notable advantages of clerks who litigate before the USSCT ~see, for example, Johnson et al., 2006, finding an indirect relationship between clerkships and oral argument quality; McGuire, 2000, who finds that former clerks working near the USSCT are more successful! and, thus, warrant further investigation. Finally, like Szmer and colleagues ~2007!, our results lend some support to the notion that litigation team size is an influential factor in predicting litigant success before the SCC; however, in the present analysis, support for this proposition is mixed. Across criminal and economic cases ~model 4!, the litigation team size variable is positively signed and statistically significant. However, no support for our hypothesis was found in the context of civil liberties0 rights cases or in the models of votes cast across all issue areas ~models 1 and 2!.11 To assist in providing more meaningful interpretations of our logistic regression coefficients, we present a graph of the predicted probabilities of a justice’s vote for the plaintiff0prosecution across different values of lawyer team gender. Given our interest in assessing the role of lawyer gender, the probabilities here and those presented below are based on the results produced in model 1. As seen in Figure 1, there is a noticeable increase in the probability of a SCC justice casting a proplaintiff0prosecution vote as the proportion of women lawyers on the plaintiff ’s side increases relative to the defendant’s. 100 E.B. KAHENY, J.J. SZMER, AND T.A. SARVER FIGURE 1 Graph of the Predicted Probabilities of a Justice’s Vote for the Plaintiff0Prosecution Given a Change in the Difference in the Proportion of Women Lawyers Representing the Plaintiff0Prosecution and Defendant, with Confidence Intervals To illustrate the substantive effects of all independent variables, we report changes in the predicted probability that a SCC justice will vote for the plaintiff in Table 2. Specifically, we assess the impact of individual variables as they change from their minimum to maximum values while holding the other independent variables constant at select values ~interval level variables are set to their mean; clerkship and party capability variables are held to their modal values!. In terms of these estimated probabilities, although litigation team experience appears to have a particularly pronounced effect on a SCC justice’s decision to support the plaintiff, lawyer team gender also has a pronounced effect. The change in the probability of a justice casting a proplaintiff vote is approximately 0.10 as one moves from a situation in which an all-male plaintiff team squares off against an all-female defense team to a situation in which an all-female plaintiff team confronts an all-male defense team. This positive change in the estimated probability of a proplaintiff vote is attributed to the relative gain in female representation for the plaintiff ’s team. Importantly, these changes in the estimated probabilities that a SCC justice votes for the plaintiff indicate that the change produced when altering the lawyer team gender variable is comparable to that produced when varying the justice/plaintiff ideological congruence measure from its minimum to maximum value. Here, one can see about a 0.12 difference in the probability of a pro-plaintiff vote when the congruence Women Lawyers before the Supreme Court of Canada 101 TABLE 2 Change in the Predicted Probability that a Justice Votes for the Plaintiff0Prosecution when the Independent Variable is Set to Maximum and Minimum Values, Holding Other Variables Constant ~Using Model 1! Independent Variable Lawyer Team Gender Justice0Plaintiff Ideological Congruence Plaintiff0Prosecution Lawyer Clerkship Defendant Lawyer Clerkship Lawyer Experience Litigation Team Size Plaintiff0Prosecution Appellant Plaintiff Association Plaintiff Business Plaintiff Local Govt. Plaintiff0Prosecution Prov’l Govt. Plaintiff0Prosecution Nat’l Govt. Defendant Association Defendant Business Defendant Local Govt. Defendant Prov’l Govt. Defendant Nat’l Govt. Change in the Predicted Probabilities 0.0950 0.1211 0.0310 0.1560 0.1728 0.0842 0.0060 0.0327 ⫺0.0819 ⫺0.2570 0.0937 0.0100 ⫺0.2779 ⫺0.0851 ⫺0.1125 ⫺0.1810 ⫺0.2423 Interval level variables are held at their mean values and the clerkship dummy variables along with the party capability variables are held to their modal categories ~‘0’ for the plaintiff0defendant clerkship variables, plaintiff national government v. individual person defendant!. measure is varied from its minimum to maximum values. Thus, a plaintiff will certainly be advantaged when trying to persuade a given justice if the plaintiff ’s position in a case ~liberal0conservative! is in line with the policy preferences of the justice. This attitudinal influence, however, is not the dominating force in this model. The impact of ideology competes with key characteristics of the litigation team, including experience, size and, yes, its gender composition. Not only does litigation team gender influence the likelihood of a given SCC justice voting to support the plaintiff0prosecution; its influence is on par with judicial ideology. We also estimated changes in the predicted probabilities ~Table 3! that specific SCC justices will vote for the plaintiff when altering the lawyer team gender variable from its minimum to its maximum value. The probabilities reflect situations in which the justice’s ideological preferences are consistent with the outcome preferred by the plaintiff and other variables are held constant at the select levels noted above. As seen 102 E.B. KAHENY, J.J. SZMER, AND T.A. SARVER TABLE 3 Difference in the Predicted Probabilities for Minimum ~all-female defense team v. all-male plaintiff0prosecution team! and Maximum ~all-female plaintiff0prosecution team v. all male-defense team! Values of Lawyer Team Gender, by Justice Justice Dickson Beetz Estey McIntyre Le Dain Wilson Lamer La Forest L’Heureux-Dubé Sopinka Gonthier Cory McLachlin Stevenson Iacobucci Major Bastarache Binnie Arbour Le Bel Change in the Predicted Probabilities 0.1358 0.1180 0.1437 0.1335 0.1347 0.1359 0.1197 0.1609 0.0997 0.1350 0.1374 0.1155 0.0926 0.1083 0.0844 0.1314 0.0875 0.1329 0.0458 0.1489 These numbers only reflect the impact of lawyer team gender on predicted probabilities of a justice siding with the plaintiff when the justice has the same characteristics ~i.e., values of ideology, gender and mean proportion of women clerks hired! as the justice listed. They also reflect the impact of lawyer team gender on the predicted probabilities of a justice vote for the plaintiff0prosecution when the justice’s ideology is consistent with the ideological direction of the outcome preferred by the plaintiff0prosecution. Otherwise, interval level variables are held at their means and the clerkship dummy variables along with the party capability variables are held to their modal categories ~‘0’ for the plaintiff0defendant clerkship variables, plaintiff0prosecution national government v. individual person defendant!. in the table, there is some variation among the individual justices in terms of the assessed impact of lawyer gender. While all justices show an increased probability in siding with the plaintiff’s team when this variable’s value is altered in the manner noted above, the impact of this variable is not uniform among the justices, nor is it ~as discussed above! more pro- Women Lawyers before the Supreme Court of Canada 103 nounced in the votes of the female justices. Indeed, the estimated probability of the late Justice Bertha Wilson voting for the plaintiff is in line with many of her male colleagues and, if anything, the gender composition of the litigation team had a less significant influence on the changes in the reported probabilities of the remaining female justices. The reported change in the probability for voting for the plaintiff, when stacking the plaintiff ’s team with female lawyers in a contest with a defendant with no female representation, was less than or just nearly 0.10 for Justices L’Heureux-Dubé and McLachlin. The reported change in the probability for Justice Arbour, moreover, was only 0.05. Meanwhile, Justice La Forest is the most sensitive to the hypothetical change in the lawyer gender variable ~in terms of producing a greater chance of siding with the plaintiff!. As alluded to above, moreover, the results in Table 3 do not line up neatly with a justice’s ideology, nor do we see those with notable conservative ideology scores ~assessed via scores from Ostberg and Wetstein, 2007! more likely to vote against teams with a greater proportion of women, a result that was more clearly seen in the USSCT ~see Szmer et al., 2010!. Discussion While scholars have taken note of the important role of legal counsel in appellate litigation, few studies have analyzed the influence of lawyer gender. Those which have ~Szmer et al., 2010! suggest that pursuing this line of inquiry is crucial to achieving a more complete understanding of judicial decision making and, specifically, of the role of lawyer characteristics. Unfortunately, that study of USSCT decision making was hindered by that court’s limited gender diversity. For instance, an obvious question is whether women justices treat arguments by women lawyers differently. With only four female justices ever serving on the USSCT, and two of these recent appointments, the ability to generalize is limited. Thus, to assess this question, we turned to another institutional context: the SCC. In this paper, we asked whether parties who employ a larger proportion of women lawyers before the SCC face disadvantages as a result. Our results suggest otherwise. In the model of all cases as well as those limited to criminal and economics issues, justices were more likely to side with parties represented by a larger proportion of female lawyers. In the civil rights0liberties model, the justices were neither more nor less likely to support the position represented by a greater number of women lawyers. Interestingly, the positive relationship between litigation team gender and justice decision making holds after controlling for several factors. First, it is possible that women are more likely to represent liberal 104 E.B. KAHENY, J.J. SZMER, AND T.A. SARVER parties and, therefore, any observed successes of women barristers are really a function of ideology. However, we control for ideology. It is also possible that women lawyers are systematically more or less capable than their male counterparts, such that any observed relationship between lawyer gender and the justices’ decisions is really a function of lawyer expertise. However, we offer several controls for litigation team expertise. We also found little evidence that the impact of lawyer gender is conditional. We found no moderating relationship between justice and lawyer gender. Moreover, unlike previous work conducted on the USSCT, we find that Canadian liberal justices are only more likely to side with litigation teams with higher proportions of women in criminal cases. Additionally, the results do not seem to be a function of the justices assigning female lawyers greater credibility in women’s issue cases ~except a small number of cases involving rape! and, in fact, in the area of civil rights0 liberties cases, our results suggest that teams with more women might be less successful when arguing a “women’s issue.” Finally, to the extent that possible working relationships have mitigated the application of gender stereotypes, it does not seem to be rooted in clerkship hires; justices known to have hired greater numbers of women law clerks are not more likely to side in favour of litigation teams with a greater proportion of women. What might be more important, however, is the gender composition of the panel on which the SCC justice serves in a given case but, even here, the evidence is mixed. In summary, these results suggest that, despite the persistence of gender discrimination against Canadian female lawyers, women arguing before the SCC, in some circumstances, fare much better than their male counterparts when attempting to persuade the justices. The possible reasons for this finding deserve further scholarly inquiry. One possibility is that this is a function of the composition of the SCC itself. The notable presence of women on this elite institution might not only mitigate negative gender bias but also enhance the status of women in the eyes of male justices. Throughout their tenure on the bench, the male justices we examined served on the SCC with multiple women colleagues whereas, on the USSCT, the male justices have not. Further, male-only decisional contexts are unique in the pool of cases we analyzed; in fact, of the cases examined here, nearly 96 per cent included at least one woman on the panel. We, of course, recognize that there are limitations to this study. For one thing, the time period ~1988–2000! prevents us from studying whether the all-male, pre-Wilson Court treated women lawyers differently. Of course, as noted above, few women lawyers participated in SCC appeals during that time. Furthermore, a future study could examine more recent cases to determine the impact of a woman chief justice. Of course, we are not sure whether that would change the findings dramatically, since Women Lawyers before the Supreme Court of Canada 105 we generally find that the Court already looked more favourably upon parties represented by more women lawyers under male chief justice stewardship. Additionally, we fail to test whether the SCC’s greater gender diversity was indeed the underlying causal mechanism for our findings, but recent work involving gender and judicial decision making as well as feminist theory may prove fruitful in future research seeking to shed light on whether this is the case. Perhaps the critical mass theory, which, to date, has helped explain the decisional contexts in which we might see marked differences in the voting behaviour of male and female judges ~for example, Boyd et al., 2010!, might also help explain the threshold point at which gender schemas are no longer applied in a negative fashion or the point at which negative stereotypes no longer control.12 Notes 1 We chose this period because it is longer than most lawyer studies and there was significant gender variation among the actors ~1988 represented the first full year in which more than one woman served on the SCC; prior to that point, female lawyer participation drops dramatically!. While we would like to examine the McLachlin Court in greater detail, the difficulties in data collection were prohibitive. Moreover, the case level data from Haynie et al. ~1999! and Tate et al. ~2002! was not available beyond 2000 when we began the study. Indeed, one of the co-principal investigators on the grant recently co-authored an article examining judge gender and also stopped at 2000 ~Songer and Johnson, 2007!. Additionally, we would like to examine cases prior to 1982 to see if the all-male Court treated women lawyers differently, but during that time few women participated in SCC litigation. For example, looking at the first 15 cases per year decided between 1978 and 1981, one woman participated in four in 1981, three in 1980, two in 1979 and no women at all participated in 1978. 2 These data are part of a larger project funded by the National Science Foundation, “Collaborative Research: Fitting More Pieces into the Puzzle of Judicial Behaviour: A Multi-Country Database and Program of Research,” SES-9975323; and “Collaborative Research: Extending a Multi-Country Database and Program of Research,” SES0137349, C. Neal Tate, Donald R. Songer, Stacia Haynie and Reginald S. Sheehan, principal investigators. 3 See the online appendix ~http:00pantherfile.uwm.edu0ekaheny0www0! for an alternative operationalization of the dependent variable ~predicting the likelihood of a proappellant vote!. 4 Our gender schema theory assumes justices know the lawyer’s sex. When only one lawyer prepares the factum and delivers the oral argument, the justices will obviously know the lawyer’s sex. It is also possible, however, that gender schemas can operate via name recognition since there is evidence justices read the factums ~see, for example, Hogg, 2006, discussing the materials read by a SCC justice or Greene et al., 1998: 110, quoting a SCC justice describing review of factums when deciding to grant leaves!, as well as previous research suggesting written work might be discounted on the basis of the author’s sex ~for example, Paludi and Strayer, 1985!. Additionally, Szmer and others ~2010! report high correlations among orally arguing attorney sex and litigant team gender composition, and their findings with respect to the effect of lawyer gender did not significantly differ when utilizing either measure. 106 E.B. KAHENY, J.J. SZMER, AND T.A. SARVER 5 Technically, it is ln ~1 ⫹ mean number of cases!, since mean values of zero occur in the data. 6 The SCC provided a complete list of SCC clerks. 7 Associations include business0trade associations, labour organizations, educational associations and institutions, religious0charitable associations, political organizations and associations of indigenous peoples. 8 Note that we did not include the women’s issues multiplicative term in economic cases because none of the economic cases in our data involved women’s issues. 9 It is also possible that any atheoretical results are a function of multicollinearity. Indeed, the VIFs for many of the party capability controls are above three. 10 Of course, some of our results may be due to coding conventions adopted in the voting data upon which we rely and the relatively few number of votes contained within some of these party capability variables when reporting their influence in the context of narrower issue areas. In our sample of civil rights cases, for example, local government plaintiffs lost in all of the small number of cases in which they appeared. Further, we expected to find a positive, significant coefficient on the plaintiff national government variable across all models. The data, though, do not support this proposition. One possible explanation is that cases involving “the Queen” were designated as the national government even if the case was being handled by the provincial government. Although both should be advantaged relative to individual parties and all other litigants for that matter, the assumption is that the national government would tend to have a higher success rate than provincial governments. Thus, perhaps the reason we do not find statistically significant results for the national government before the SCC in all of these issue areas is that, in fact, in a significant number of cases, the national government was not litigating but, rather, it was the provincial government. Inspection of a smaller sample of cases in which the data source indicates the national government was a party to a lawsuit did give rise to this concern. Fortunately, however, we have no reason to believe this possible issue has any influence with respect to our variables of interest pertaining to lawyer gender. 11 We also tested these models with a panel size variable, which might reflect case complexity. With the exception of model 5 ~civil rights!, increases in this variable were negatively related to a vote for the plaintiff0prosecution. Larger panels in the civil rights context, however, increased the likelihood of a justice vote for the plaintiff. We also tested a panel size0lawyer team gender interaction in our combined issues model. Lawyer team gender was significant ~and positive! in seven- and nine-justice panels; in five-justice panels, it was not significant. This is not surprising, since fivejustice panels typically involve clear-cut applications of straightforward legal principles. 12 Of course, analyses that further probe the development and application of these gender stereotypes in the thought processes of relevant actors in high court litigation might also prove fruitful. References American Bar Association ~ABA!, Commission on Women in the Profession. 2007. A Current Glance at Women in the Law. www.abanet.org0women0CurrentGlanceStatistics2007. pdf ~25 April 2008!. Boyd, Christina L., Lee Epstein and Andrew D. Martin. 2010. “Untangling the Causal Effects of Sex on Judging.” American Journal of Political Science 54 ~2!: 389–411. Brockman, Joan. 1993. “A Difference without a Distinction.” Canadian Journal of Law & Society 8 ~1!: 149–64. Women Lawyers before the Supreme Court of Canada 107 Canadian Bar Association ~CBA!. 1993. Touchstones for Change: Equality, Diversity and Accountability: Report of the Canadian Bar Association Task Force on Gender Equality in the Legal Profession. Ottawa ON: Canadian Bar Association. Carli, Linda L. 1989. “Gender Differences in Interaction Style and Influence.” Journal of Personality and Social Psychology 56 ~4!: 565–76. Corley, Pamela. 2008. “The Supreme Court and Opinion Content: The Influence of Parties Briefs.” Political Research Quarterly 61 ~3!: 468–78. Craig, J.M. and R.R. Jacobs. 1985. “The Effect of Working with Women on Male Attitudes Toward Female Firefighters.” Basic and Applied Social Psychology 6: 61–74. Dolan, Kathleen. 2004. Voting for Women: How the Public Evaluates Women Candidates. Boulder CO: Westview Press. Farhang, Sean and Gregory Wawro. 2004. “Institutional Dynamics on the US Courts of Appeals: Minority Representation under Panel Decision Making.” Journal of Law, Economics and Organization 20 ~2!: 299–330. Fiske, S.T. 1982. “Schema-Triggered Affect: Applications to Social Perception.” In Affect and Cognition: The 17 th Annual Carnegie Symposium on Cognition, ed. M.S. Clark and S.T. Fiske. Hillsdale NJ: Erlbaum. Flemming, Roy B. 2004. Tournament of Appeals: Granting Judicial Review in Canada. Vancouver BC: UBC Press. Flemming, Roy B. and Glen S. Krutz. 2002. “Repeat Litigators and Agenda Setting on the Supreme Court of Canada.” Canadian Journal of Political Science 35~4!: 811–33. Foot, David K. and D.A. Stager. 1990. “Intertemporal Market Effect on Gender Earnings Differentials: Lawyers in Canada, 1970–1980.” Applied Economics 21 ~8!: 1101–28. Galanter, Marc. 1974. “Why the Haves Come Out Ahead: Speculations on the Limits of Legal Change.” Law and Society Review 9~1!: 95–160. Greene, Ian, Carl Baar, Peter McCormick, George Szablowski and Martin Thomas. 1998. Final Appeal: Decision-making in Canadian Courts of Appeal. Toronto: James Lorimer & Co. Gryski, Gerard, Eleanor C. Main and William J. Dixon. 1986. “Models of State High Court Decision Making in Sex Discrimination Cases.” Journal of Politics 48 ~1!: 143–55. Haire, Susan Brodie, Stefanie A. Lindquist and Roger Hartley. 1999. “Attorney Expertise, Litigant Success, and Judicial Decisionmaking in the US Courts of Appeals.” Law and Society Review 33 ~3!: 667–85. Haslett, Beth, Florence L. Geis and Mae R. Carter. 1992. The Organizational Woman: Power and Paradox. Norwood NJ: Ablex Publishing. Hausegger, Lori and Stacia Haynie. 2003. “Judicial Decisionmaking and the Use of Panels in the Canadian Supreme Court and the South African Appellate Division.” Law and Society Review 37~3!: 635–58. Haynie, Stacia L., Reginald Sheehan, Donald R. Songer, and C. Neal Tate. 1999. “Collaborative Research: Fitting More Pieces into the Puzzle of Judicial Behavior: a MultiCountry Data Base and Program of Research.” National Science Foundation, SES-9975323. Hogg, Peter W. 2006. “Judicial Interview Process: Opening Remarks to Ad Hoc Committee on Supreme Court Appointment.” February 27. http:00www.justice.gc.ca0eng0deptmin0pub0scc-csc0sp-dis.html ~30 October 2008!. Johnson, Timothy R., Paul J. Wahlbeck and James F. Spriggs, II. 2006. “The Influence of Oral Arguments on the US Supreme Court.” American Political Science Review 100 ~1!: 99–113. Kahn, Kim Fridkin. 1996. The Political Consequences of Being a Woman: How Stereotypes Influence the Conduct and Consequences of Political Campaigns. New York: Columbia University Press. Kanter, Rosabeth Moss. 1977a. Men and Women of the Corporation. New York: Basic Books. 108 E.B. KAHENY, J.J. SZMER, AND T.A. SARVER Kanter, Rosabeth Moss. 1977b. “Some Effects of Proportions on Group Life: Skewed Sex Ratios and Responses to Token Women.” American Journal of Sociology 82 ~5!: 965–90. Kay, Fiona M. and Joan Brockman. 2000. “Barriers to Gender Equality in the Canadian Legal Establishment.” Feminist Legal Studies 8 ~2!: 169–98. Lopez, Maria Pabon. 2008. “The Future of Women in the Legal Profession: Recognizing the Challenges Ahead by Reviewing Current Trends.” Hastings Women’s Law Journal 19 ~1!: 53–103. Matland, Richard E. 1994. “Putting Scandinavian Equality to the Test: An Experimental Evaluation of Gender Stereotyping of Political Candidates in a Sample of Norwegian Voters.” British Journal of Political Science 24 ~2!: 273–92. McAtee, Andrea and Kevin T. McGuire. 2007. “Lawyers, Justices, and Issue Salience: When and How Do Legal Arguments Affect the U.S. Supreme Court?” Law and Society Review 41: 259–78. McCormick, Peter. 1993. “Party Capability Theory and Appellate Success in the Supreme Court of Canada, 1949–1992.” Canadian Journal of Political Science 3: 535–40. McCormick, Peter and Twyla Job. 1993. “Do Women Judges Make a Difference? An Analysis by Appeal Court Data.” Canadian Journal of Law and Society 8 ~1!: 135–48. McGuire, Kevin T. 1995. “Repeat Players in the Supreme Court: The Role of Experienced Lawyers in Litigation Success.” Journal of Politics 57 ~1!: 187–96. McGuire, Kevin T. 1998. “Explaining Executive Success on the US Supreme Court.” Political Research Quarterly 51 ~2!: 505–26. McGuire, Kevin T. 2000. “Lobbyists, Revolving Doors and the U.S. Supreme Court.” Journal of Law and Politics 16 ~1!: 113–37. Nelson, Mary Stuart. 2004. “The Effect of Attorney Gender on Jury Perception and Decision-Making.” Law and Psychology Review 28 ~1!: 177–94. Nisbett, Richard E. and Lee Ross. 1980. Human Inference: Strategies and Shortcomings of Social Judgment. Englewood Cliffs NJ: Prentice-Hall. Noonan, Mary C., Mary E. Corcoran and Paul N. Courant. 2005. “Pay Differences Among the Highly Trained: Cohort Differences in the Sex Gap in Lawyers’ Earnings.” Social Forces 84 ~2!: 153–72. Ostberg, C. L., and Matthew E. Wetstein. 1998. “Dimensions of Attitudes Underlying Search and Seizure Decisions of the Supreme Court of Canada.” Canadian Journal of Political Science 31 ~4!: 767–87. Ostberg, C. L., and Matthew E. Wetstein. 2007. Attitudinal Decision Making in the Supreme Court of Canada. Vancouver: UBC Press. Paludi, M.A. and L.A. Strayer. 1985. “What’s in an Author’s Name? Differential Evaluations of Performance as a Function of Author’s Name.” Sex Roles 12 ~3–4!: 353–61. Peresie, Jennifer L. 2005. “Female Judges Matter: Gender and Collegial Decisionmaking in the Federal Appellate Courts.” Yale Law Journal 114 ~7!: 1759–90. Powell, Gary N. 1993. Women and Men in Management. Newbury Park CA: Sage Publications. Rhode, Deborah L. 2002. “Gender and the Profession: The No-Problem Problem.” Hofstra Law Review 30 ~5!: 1001–13. Sanbonmatsu, Kira. 2002. “Gender Stereotypes and Vote Choice.” American Journal of Political Science 46 ~1!: 20–34. Segal, Jeffrey A. and Albert D. Cover. 1989. “Ideological Values and the Votes of US Supreme Court Justices.” American Political Science Review 83 ~2!: 557–65. Segal, Jennifer. 2000. “Representative Decision Making on the Federal Bench: Clinton’s District Court Appointees.” Political Research Quarterly 53 ~1!: 137–50. Sheehan, Reginald S., William Mishler and Donald R. Songer. 1992. “Ideology, Status, and the Differential Success of Direct Parties before the Supreme Court.” American Political Science Review 86 ~2!: 464–71. Women Lawyers before the Supreme Court of Canada 109 Songer, Donald R. 2008. The Transformation of the Supreme Court of Canada: An Empirical Examination. Toronto: University of Toronto Press. Songer, Donald R. and Reginald S. Sheehan. 1992. “Who Wins on Appeal? Underdogs and Upperdogs on the United States Courts of Appeals.” American Journal of Political Science 36: 235–58. Songer, Donald R., Sue Davis and Susan Haire. 1994. “A Reappraisal of Diversification in the Federal Courts: Gender Effects in the Courts of Appeals.” Journal of Politics 56 ~2!: 425–39. Songer, Donald R. and Susan W. Johnson. 2007. “Judicial Decision Making in the Supreme Court of Canada: Updating the Personal Attribute Model.” Canadian Journal of Political Science 40 ~4!: 911–34. Szmer, John J., Susan W. Johnson and Tammy A. Sarver. 2007. “Does the Lawyer Matter? Influencing Outcomes on the Supreme Court of Canada.” Law and Society Review 41 ~2!: 259–78. Szmer, John J., Tammy A. Sarver and Erin B. Kaheny. 2010. “Have We Come a Long Way Baby? Female Attorneys before the United States Supreme Court.” Politics and Gender 6: 1–36. Tate, C. Neal, Stacia L. Haynie, Reginald Sheehan, and Donald R. Songer. 2002. “Collaborative Research: Extending a Multi-Country Database and Program of Research.” National Science Foundation, SES-0137349. Taylor, Shelley E. and Jennifer Crocker. 1981. “Schematic Bases of Social Information Processing.” In Social Cognition: The Ontario Symposium, vol. 1, ed. E. Tory Higgins, C. Peter Herman and Mark P. Zanna. Hillsdale NJ: Lawrence Erlbaum. Valian, Virginia. 1999. “Roundtable: The Cognitive Bases of Gender Bias.” Brooklyn Law Review 65: 1037–61. Valian, Virginia. 2005. “Beyond Gender Schemas: Improving the Advancement of Women in Academia.” Hypatia 20 ~3!: 198–213. Western Law Career Services Office. 2005. Clerkship Handbook for Western Law Students. http://www.law.uwo.ca/info-current/cso/2005Clerkship.htm ~3 March 2010!. Wetstein, Matthew W. and C.L. Ostberg. 2007. “Gender Differences in the US and Canadian Supreme Courts.” Paper presented at the 2007 Annual Meeting of the Midwest Political Science Association, Chicago IL.