Turning to the Courts: A Quantitative Analysis of the Gay and Lesbian Movement’s Use of Legal Mobilization

Abstract

We quantitatively examine the factors influencing the lesbian and gay movement’s involvement in state-level appellate cases through an analysis of Lambda Legal’s judicial agenda from 1981 through 2000. We combine a list of state appellate cases involving gay and lesbian rights during this period with data from Lambda’s “Docket Update” to create a data set that identifies which cases Lambda was involved in. We theorize that Lambda’s involvement will be shaped by the potential policy implications of the case, the organization’s ability to influence the outcome, and organizational membership concerns. Using logistic regression, we examine the effect of case characteristics, the political and cultural climate in the state, and variations in the state’s lesbian and gay community on the likelihood of Lambda participating in a particular state-level case. We add to the understanding of movement tactical decisions and the factors influencing the likelihood of legal mobilization.

ACKNOWLEDGMENTS

The authors would like to thank Kathryn Lively and Tommy Wornick for their assistance on the project.

Notes

¹ While Lambda Legal currently identifies as an inclusive organization working for lesbian, gay, bisexual, and transgender (LGBT) rights (Lambda Legal 2013a), we use the term “gay and lesbian movement” throughout the paper to better reflect the name and agenda of the organization during the period under study.

² In her analysis of Lambda and its agenda, Andersen (2005) argued that Lambda, throughout much of its early history, purposefully joined cases once they reached the appellate stage in order to conserve resources by skipping the expensive trial court stage.

³ Because there are so few cases in which Lambda is counsel (only 20 out of 286), the only alternative to defining participation broadly is to limit our analyses to Lambda’s participation as amicus curiae. We decided against this option for several reasons. First, Lambda itself defines its participation broadly. In its most recent annual reports, Lambda identified the number of cases in which it was involved without distinguishing between those in which it was acting as counsel and those in which it filed amicus briefs (Lambda Legal 2013b). In addition, many of the variables in this study have the potential to influence both forms of participation. Therefore, we conclude that operationalizing Lambda’s participation to include both acting as amicus curiae and acting as counsel is a reasonable alternative.

⁴ In addition, Lambda was criticized early on for emphasizing gay male rights at the expense of lesbians and their interests, including family and custody rights (Andersen 2005; Pinello 2003).

⁵ We also considered the total number of cases on Lambda’s docket the previous year to see if the size of the organization’s caseload shaped their involvement in future cases. It was not significant so we do not include the measure here.

⁶ Lambda’s own resources, such as staff and income, should have a substantial influence on its level of legal mobilization (Hoover and den Dulk 2004). However, the only available measure of Lambda’s organizational resources is its annual income (Andersen 2005), which is so highly correlated with year that it cannot be included in the analyses. In addition, when we used the budget measure, rather than our measure of historical time, it was not significant.

⁷ Pinello (2003) identified gay rights cases through a series of Westlaw searches using key terms such as homosexual, sexual orientation, and lesbian. He then supplemented his Westlaw searches with cases found through secondary sources like the Lesbian/Gay Law Notes and through discussions with major legal organizations. Any case that involved “… (1) the presence of a lesbian or gay litigant (self-identified or alleged) advancing or defending a gay civil right or liberty that was adjudicated on the merits, or (2) an action otherwise directly affecting lesbians and gay men as a class” was included in his list (Pinello 2003:163) . See Appendix 1.1 in Pinello (2003) for further details.

⁸ Nebraska’s nonpartisan legislature is coded zero.

⁹ These data were compiled by Stefanie A. Lindquist and are available at http://academic.udayton.edu/sppq-TPR/tpr_data_sets.html.

¹⁰ We use state supreme court measures for all cases, even those at the intermediate appellate level, because they are the only state-level judicial ideology scores available.

¹¹ Though The Gayellow Pages has been published since 1973, there are seven years in which no volume was published. For these years—1979, 1981, 1984, 1989, 1991, 1993, 1996, 1998—the average number of organizations from the previous and following years was used.

¹² We also considered a broader community measure that counted all bars and businesses listed in the directory, but it was highly correlated with citizen ideology, one of our public opinion measures. Models using the broader measure had substantively similar results to those presented below.

¹³ Since data are only available every ten years, we linearly interpolate the intervening years to create a measure from 1981 to 1998.

¹⁴ While religiosity overall is associated with lower support for lesbian and gay rights (Loftus 2001), there are dramatic variations by denomination in people’s views of homosexuality. We considered using the proportion of conservative Protestants in each state, rather than the proportion of adherents overall, to better capture religious resistance to lesbian and gay rights, but the measure was too highly correlated with region.

¹⁵ We do not include a separate measure capturing ally participation in our analyses because it is too highly correlated with the total number of briefs. In addition, many of the court decisions had incomplete information on the organizational affiliation of participating counsel and brief filers. For example, the written decisions in several states would list the first organization on the brief and then just list “and others” making the identification of specific organizations impossible. Therefore, we use the count of total briefs since it was the most straightforward to code with the lowest possibility of error.

¹⁶ We also examined whether the case was in the court of last resort and whether the state was a participant in the case. Neither measure was significant so we do not report them here.

¹⁷ One could argue that the state-level measures are better measures, both theoretically and empirically, than the general regional indicators, in that they more specifically capture the state environment in which the case is being adjudicated. It is difficult to say, in that the models explanatory power is almost identical, regardless if one uses the region or state measures.

Additional information

Notes on contributors

Melinda D. Kane

Melinda D. Kane is an associate professor of sociology and an affiliate of the Women’s and Gender Studies program and the Center for Diversity and Inequality Research at East Carolina University. Her research focuses on the intersection of inequality and mobilization through studies of the LGBT movement in the United States. Her recent work has examined the neighborhood-level distribution of same-sex couples in Columbus, Ohio, and geographic variations in the presence of LGBT organizations and counter-organizations, including Exodus chapters, LGBT campus groups, and Metropolitan Community Churches. Recent publications have appeared in Research in Social Movements, Conflicts and Change, City & Community, Sociological Forum, and the Journal of Homosexuality.

Thomas Alan Elliott

Thomas Alan Elliott is a PhD candidate in sociology at the University of California-Irvine. His research interests include sexuality, social movements, and culture. His recent work has focused on newspaper coverage of social movements as a cultural outcome of mobilization. His dissertation examines newspaper coverage of homosexuality since 1950 and the reasons why the discourse around homosexuality has undergone such significant changes during that time.