The Impact of Judicial Selection on LGBT Rights Cases

Find Your State

Know the laws in your state that protect LGBT people and people living with HIV.

In 2015, Lambda Legal commissioned a series of statistical analyses on an expansive new dataset on state high court decisions adjudicating LGBT rights claims. The study was conducted by a team of independent researchers led by Anthony Michael Kreis with support from Ryan Krog and Allison Trochesset. The research compared the outcomes in LGBT rights cases in states with different judicial selection methods, finding that processes through which different states select judges can play a role in how state high courts rule in LGBT cases. Below is a summary of the study. For a complete analysis of the dataset, variables, and findings, please visit Lambda Legal’s Fair Courts Project at

Briefly, the study found that:

  1. State high courts whose judges stand for election are less supportive of LGBT rights claims.
  2. Results suggest that lack of support for LGBT rights among state high courts with elected judges can be attributed to ideological factors playing a larger role in shaping judges’ decisions on these courts.

A. Data: A Look at State High Court Cases

To examine the implications of judicial independence for state courts’ treatment of LGBT rights claims, the study’s dataset included all cases involving LGBT claims decided by state high courts starting in 2003, after the U.S. Supreme Court handed down its ruling in Lambda Legal’s case Lawrence v. Texas, through 2015. The search recovered a total of 127 relevant cases.1 Although the data contain decisions from 43 different states of the 50 states and the District of Columbia, there is some variation in the sample, with some states having issued more relevant decisions than others. California had the most LGBT rights cases during this time handing down 12 decisions, followed by Massachusetts with 10 cases.

After identifying a set of relevant cases, rulings were then classified as either favorable or unfavorable to LGBT rights. Cases that either directly upheld rights for LGBT persons, e.g. in favor of marriage rights for same-sex couples, or decisions that the parties would reasonably foresee yielding results that which could particularly benefit LGBT persons, e.g. second-parent adoption, were coded as “pro-LGBT.” Those cases where courts denied LGBT rights claims or restricted the legal rights that LGBT persons could avail themselves of were coded as “anti-LGBT.” Cases that were not decided on the merits were excluded.

B. Key Variables

The notable variable throughout the analysis is the method by which states select their judges. Although the exact system varies from state to state, most select and retain judges via one of four broad schemes: (1) Partisan Elections; (2) Nonpartisan Elections; (3) Uncontested Retention Elections; (4) Lifetime tenure or reappointment. Thirty-eight states have some type of judicial elections; the remaining twelve grant life tenure or use reappointment of some form.

Though the central focus of this research is investigating the extent to which outcomes in LGBT rights cases are influenced by judicial selection methods, it is important to account for additional factors that might also influence judges’ decisions on these issues. To summarize, the study controls for four sets of factors that can influence the state high courts’ rulings on LGBT rights: (1) the institutional design of a state’s judicial selection mechanism; (2) characteristics of the panel of judges hearing a case, such as their judicial ideology2; (3) the nature of the legal questions3 being adjudicated in a case; (4) the political context of the state4 in which a court operates. Next, using statistical models, the study generates predicted probabilities of the likelihood a court would rule in favor of LGBT rights, given these factors.

C. Findings

FINDING: State high courts whose judges stand for election are less supportive of LGBT rights claims.

1. The Impact of Judicial Selection of LGBT Rights Claims

The study first examines how the judicial selection mechanism employed for high state judges effects LGBT rights claims. Results of the study show that courts whose judges face either partisan or nonpartisan elections are less supportive of LGBT rights claims. Figure 1 shows that high court judges elected through partisan elections are the least supportive of LGBT rights claims according to the data, supporting a pro- LGBT claim in only 53 percent of cases. Slightly more supportive are courts where high court judges are elected through nonpartisan elections (70 percent of cases), followed closely by high courts where judges are appointed and run in uncontested retention elections (76 percent of cases.) The high courts that are most supportive of LGBT rights are those where the judges are granted lifetime tenure or reappointed, supporting a position favorable to LGBT rights in 82 percent of cases in the data.5

Figure 1.

FINDING: Results suggest that this lack of support among state high courts with elected judges can be attributed to ideological factors playing a larger role in shaping judges’ decisions on these courts.

2. The Impact of Judicial Ideology of LGBT Rights Claims

Judicial selection is not the only factor driving LGBT litigation outcomes. A primary factor in state high courts’ willingness to rule against or in favor of upholding LGBT rights is the ideological disposition of the sitting justices. That fact notwithstanding, the role of ideology is noticeably amplified in courts subject to elections, as compared to those whose members are not. In other words, ideology plays a larger role in the decision-making process in less independent courts, where judges are subject to competitive elections. This amplification effect is not uniform for all judges. Judges sitting on ideologically conservative courts are far more sensitive to the appointment mechanism than their counterparts sitting on ideologically liberal courts. This sensitivity helps explain why conservative courts where judges face partisan elections are the least supportive of LGBT claims.

Figure 2.

Figure 2 demonstrates the effect of the ideological disposition of sitting justices on support for LGBT rights claims. Figure 2 shows more ideologically conservative courts are less supportive of LGBT claims. The empirical results lend strong support for an ideological account for state court decision-making in cases involving LGBT legal claims. For instance, the probability of an ideologically moderate court ruling in a pro-LGBT position is 62 percent, holding all other variables at their mean or modal value. Contrast this with a highly conservative court, where the predicted probability of a pro-LGBT ruling drops precipitously to 32 percent; or for a very liberal court where the probability jumps to 90 percent.

3. The Interaction between Judicial Selection and Judicial Ideology on LGBT Rights Claims

Although the effects of variables for judicial selection mechanisms and ideology of judges are interesting in their own right, the interaction between these variables can provide several key insights into how the institutional design of a court can condition judicial behavior. It is plausible that certain selection systems encourage judges to behave more ideologically than others on cases dealing with LGBT rights. For instance, with political constituencies that must be appeased, elected judges may respond to their constituents by voting in ways that reflect their constituents’ views.

Figure 3.

Figure 3 presents the interaction between the judicial selection mechanism variables and the judicial ideological positioning on the high court. Figure 3 shows that as courts become more ideologically conservative, they become less supportive or LGBT rights; however, this effect is strongest for courts where judges are selected through partisan elections.

Figure 4.

Figure 4 presents the probability of a state high court’s supporting LGBT rights across different methods of selection. Judges sitting on ideologically conservative courts are far more sensitive to the appointment mechanism than their counterparts sitting on ideologically liberal courts. As the graph displays, liberal courts are more likely to vote in favor of LGBT rights claims, regardless of how the judges are seated. Conservative courts where judges face partisan elections are the least supportive of LGBT rights claims. The probability of a conservative panel of judges who face partisan elections is only 20 percent, holding all other variables at their mean or modal value. That probability increases to 37 percent for nonpartisan elections, and up to 42 percent when facing uncontested retention elections, and more than doubles up to a probability of 57 percent for courts that have lifetime tenure or reappointment systems.

4. The Interaction between Political Context of the State and Judicial Selection Mechanisms on LGBT Rights Claims

Overall, if judges are responsive to the electorates of their states, then the clearest way to observe the implications of the effects observed in Figures 3 and 4 would be to examine courts’ treatment of LGBT issues in liberal versus conservative states. This is explored by examining the interaction between judicial selection mechanism and the political context of the state as determines by the State Citizen Ideology score.

Figure 5:

Figure 5 presents the predicted likelihood of support for LGBT claims across selection mechanisms for an average state designated “conservative,” compared to an average state designated “liberal” (drawn from the 25th and 75th percentile of the State Citizen Ideology variable, respectively). As demonstrated by the graph, there is a considerable gap between judges elected in partisan elections in states with liberal versus conservative populations. Courts whose members are selected through partisan elections in more liberal states are considerably more supportive of LGBT positions than their counterparts in liberal states with other appointment designs. Conversely, judges on courts selected through partisan elections in states with more conservative electorates are considerably more hostile to LGBT legal claims. To illustrate, for states with partisan elections, the predicted probability of a state high court ruling in a pro-LGBT position in state with a liberal citizenry is extremely high at 94 percent; in stark contrast, the probability of a pro-LGBT position drops to 22 percent for courts housed in conservative states with partisan elections. For every other type of selection mechanism, there is no statistical distinction to between courts with conservative versus liberal citizenries.



1. Specially, we searched cases that included one or several of the following terms: “homosexual,” “same-sex ,” “gender identity," “transgender,” “gay,” “lesbian,” “bisexual,” “sexual orientation,” “sexual preference,” “LGBT” or “queer.”

2. Although political scientists have developed a number of different measures for judicial ideology, this study relies on the innovative scores developed by political science professors Adam Bonica and Michael Woodruff. Their campaign finance scores (CFscores) for state high court judges are based on campaign contributions and the ideologies associated with various contributors. See, Bonica, Adam. 2013. Database on Ideology, Money in Politics, and Elections: Public version 1.0. Bonica, Adam. 2014. “Mapping the Ideological Marketplace.” American Journal of Political Science, 58 (2): 367-387, available at http:// Contributions made by the judges themselves, as well as contributions towards the judges’ campaigns are also factored in. In the absence of elections, the ideology of the appointing body (governor or legislature) was factored in, based on their campaign contributions data. A score above 0 indicates a more conservative leaning ideology, while scores below 0 are more liberal. Thus, in the context of our regression models, we expect the coefficients for this variable to be negative if conservative courts are less supportive of LGBT issues.

3. Two legal issues of particular interest are constitutional issues and family law. Cases centering on constitutional questions may be more subject to strategic litigation. Family law cases include all actions related to marriage, divorce, civil unions, child custody and adoption that were not constitutional challenges. This study includes two variables capturing whether a case involved a constitutional issue or a family law (coded 1, if yes; 0, otherwise). These variables are labeled Constitutional Law Case and Family Law Case, respectively.

4. This study employs commonly used measures developed by political scientists to capture the ideological character of a state’s citizenry and governmental 32 officials. These measures—State Citizen Ideology and State Government Ideology—were originally developed by a the team of political scientists in Berry et al. (1998) and are imputed from ADA interest group ratings of each state’s congressional delegation. See, William D. Berry, Evan J. Ringquist, Richard C. Fording and Russell L. Hanson. “Measuring Citizen and Government Ideology in the American States, 1960-93.” American Journal of Political Science, Vol. 42, No. 1 (Jan.), pp. 327-348. Both of these measures were gathered online from: A control was added for whether the deciding court was the focus of a highly competitive electoral cycle within two years prior to the decision. The relative metric for assessing competitiveness will likely be based on both primary election vote share and general election vote share, where applicable. This variable is labeled as Competitive Elections.

5. These data do not allow firm conclusions to be drawn about the impact of judicial selection mechanisms on LGBT rights claims based on the selection mechanism variable alone. There are far fewer cases in our data that were decided by courts in partisan or non-partisan elections. This may be due, in part, to strategic litigants who forum shop for courts that plaintiffs perceive as having greater independence and, therefore, are more likely to rule in favor of LGBT plaintiffs. While this explanation works well for impact litigation, it less satisfying for other types of cases like non-constitutional family law cases that are not deliberative actions intended to advance a broader end.

6. The graph demonstrates that more ideologically conservative courts are less support of LGBT rights.


A Social Justice Agenda for Achieving Fair and Impartial Courts »