only for prof washington

The Religious Dynamics of Decision Making on Gay Rights Issues in the U.S. House of Representatives, 1993–2002

ELIZABETH ANNE OLDMIXON BRIAN ROBERT CALFANO

In recent decades, homosexuality has emerged as a truly national political issue. As a result, the U.S. Congress is increasingly called upon to consider and set policy on an array of issues related to the status of gay men and lesbians. This article investigates legislator decision making pertaining to gay issues in the U.S. House of Representatives. Specifically, we examine the effect of several indicators on legislator support for progressive gay policies, including ideology, partisanship, and the characteristics of district religious affiliation. We use additive indices of legislator support for progressive gay policies, and use logit and ordered logit to derive estimates of influence. The findings indicate that even while legislator partisanship and ideology largely structure decision making, legislators are also highly responsive to the presence of conservative Protestants and (to a lesser extent) Roman Catholics in their constituency.

INTRODUCTION

Economic-class-based political alignments have traditionally characterized the dominant po- litical conflicts in Western democracies, and yet, in recent decades, the salience of these alignments has waned. Increasingly, cultural disagreement provides the subtext to politics. By “culture” we refer to a “moral order” (Wuthnow 1988)—a shared sense of how people ought to live. As a con- sequence of cultural disagreement, there “has been an upsurge in demands and political causes linked less to economics and class than to the lifestyle and social ‘identity’ in which different groups seek to legitimize the cultural values they hold dear” (Smith and Tatalovich 2003:50).

Postmaterialism has given way to a multiplicity of culturally-based political conflicts. Differ- ent values and issues animate these conflicts, but what cultural activists have in common is “status anxiety,” which is the fear that one’s cultural values are not held in esteem by society (Smith and Tatalovich 2003:25). Therefore, cultures compete for status and legitimacy in the public sphere (2003:17). As competing cultural groups demand that government institutionalize their values and preferred social relationships, cultural conflicts become political conflicts (Wildavsky 1987).

As cultural conflict unfolds, one group generally presses for recognition of its status, while another group rallies to uphold the status quo (Smith and Tatalovich 2003:25). The conflict over gay rights clearly fits this pattern. Beginning in earnest after the 1969 Stonewall Riots, gays and lesbians continually engage the political system in increasing numbers in an effort to legitimate their status, and they meet firm resistance. The conflict over gay rights represents what is arguably one of the dominant cultural cleavages of the postmaterial era in the United States (Inglehart 1990)—the conflict between religious traditionalism and progressivism (Layman and Carmines 1997).1 We note that cultural conflicts are not necessarily animated by religion (Smith and Tatalovich 2003:240). Religion, for example, tells us little about the political debates

Elizabeth Anne Oldmixon, Ph.D., Assistant Professor, Department of Political Science, University of North Texas, PO Box 305340, Denton, TX 76203. E-mail: [email protected] Brian Robert Calfano is a Lecturer in the Department of Political Science at Texas A&M University. E-mail: briacal@ polisci.tamu.edu

Journal for the Scientific Study of Religion (2007) 46(1):55–70

56 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

concerning environmentalism, gun control, or “English only” ordinances. And yet religion is “[o]ne of the strongest forces in society that affirms traditional values” (Smith and Tatalovich 2003:38), and it is directly relevant to the conflict over gay rights.

It should come as no surprise, then, that when the highest court in Massachusetts ruled in 2003 that same-sex marriages are legal under the commonwealth’s constitution, the commonwealth’s religious leaders were sharply divided. The Roman Catholic Church was only the most high profile critic of the ruling. It was joined in its opposition by the leaders of other traditional faiths, including Orthodox Jewish, Eastern Orthodox, evangelical, and African-American Protestant leaders. On the other side of the theological divide, Rev. Nancy Taylor of the United Church Christ hailed the ruling, noting that: “The court has affirmed that the quest for civil rights for all citizens will not be denied, either by prejudice or by religious doctrine.” Reform Jewish, Episcopal, and Unitarian Universalist leaders also supported the decision.2

This state-level political conflict promised to have national-level implications. In subsequent months, as legislators in Massachusetts struggled to comply with the court’s ruling, President Bush suggested the possibility of amending the U.S. Constitution to codify marriage as between a man and a woman. In his 2004 State of the Union address he said: “If judges insist on forcing their arbitrary will upon the people, the only alternative left to the people would be the constitutional process. Our nation must defend the sanctity of marriage.”3 While the “constitutional process” has stalled, proponents and opponents of gay rights continue to square off over the issue of marriage, with both camps trying to legitimate, in law, their own “moral tradition.”4

The U.S. Congress provides the arena in which factions are brought together to make policy for the nation.5 As such, it is increasingly called upon to make policy related to gay issues. To be sure, the institution has resisted wading into this policy conflict.6 Campbell and Davidson (2000: 350) estimate that “[o]f the ten thousand or so bills introduced in each Congress, a preliminary search of workload for the years 1975 through 1999 (94th through 105th Congresses) suggests that in each Congress only a very small proportion of bills—ten to twenty—deal overtly with gay rights.” However, in the 1990s, the Congress considered a series of very high profile gay- related issues, such as the Defense of Marriage Act, the Employment Non-Discrimination Act, and President Clinton’s effort to lift the ban on gays in the military. Legislators may be reluctant to legislate on this issue, but it is nevertheless a national issue.

Therefore, this article explores legislative decision making in the U.S. House of Representa- tives pertaining to gay issues. To the extent that gay issues represent a cultural cleavage between religious traditionalism and progressivism, we expect legislator decision making to be strongly influenced by phenomena that inform culture. Jelen (1997:55) notes that “the economically based New Deal party system is giving way to a new alignment based primarily on cultural issues.” Party, religion, and ideology all swing on the axis of that alignment. Therefore, we expect them to influence congressional decision making. To investigate this possibility, we use logit and or- dered logit regression to model the effect of religion, party, ideology, and various controls on legislator support for progressive gay policies. The findings indicate that while partisanship and ideology largely structure decision making, legislators are also highly responsive to the presence of conservative Protestants and (to a lesser extent) Catholics in their constituency.

CULTURE, CONGRESS, AND GAY ISSUES

Since cultural regulation often takes place at the state and local levels, much of the existing literature focuses on those levels and tends to explore policy making in the aggregate—examining referenda voting and attempting to determine the kinds of communities that have permissive moral regulations (see, e.g., Button, Rienzo, and Wald 1997; Fairbanks 1977; Haeberle 1996; Meier 1994; Meier and Johnson 1990; Meier and McFarlane 1993; Mooney and Lee 2000, 1995; Morgan and Meier 1980; Wald, Button, and Rienzo 2001, 1996). However, the national government is

RELIGIOUS DYNAMICS OF DECISION MAKING ON GAY RIGHTS 57

increasingly an arena of conflict on gay issues. Therefore, this research focuses on the U.S. House of Representatives, and it builds on previous cultural/morality policy research also conducted at this level (Adams 1997; Haider-Markel 1999, 2001; Oldmixon 2002; Steiner 1983; Tatalovich and Schier 1993; Wattier and Tatalovich 1995).

The collective wisdom of this literature is that legislator decision making is structured by partisanship, ideology, and religion. These phenomena inform individual affect toward competing cultural groups operating in the public space, and they inform our sense of what is morally correct and incorrect. We expand on the existing literature in important ways. First, our analysis includes an explicit mix of cultural and socioeconomic variables, which is essential if we are to establish the distinctiveness of cultural issues (Wald, Button, and Rienzo 2001). To be sure, Haider-Markel’s (1999, 2001) analysis of Congress’s treatment of gay issues includes a mix of cultural and socioeconomic variables. However, he combines them into a single measure of constituency opinion. While that is perfectly appropriate (and quite sophisticated) given the puzzle addressed by his research, it does not allow us to discriminate between cultural and socioeconomic indicators. In particular, it does not allow us to measure the effect of constituency-level religious dynamics.

Second, we distinguish between different kinds of congressional decision making. The roll call vote is just one (albeit important) aspect of the legislative process. A more holistic approach allows us to consider legislator decision making in areas not controlled by leadership, or influenced by rules and committees. Therefore, we consider voting and co-sponsorship decisions separately. Finally, this research explores a relatively new cultural issue. The House only started to address gay issues regularly, and in nonsymbolic ways, in the 1990s. Having been on the national political radar for a shorter period of time than, for example, abortion, gay issues present legislators with more uncertainties. Public attitudes on this issue are changing rapidly (Brewer 2003a; Wilcox and Norrander 2002), and legislators may not be certain of the electoral repercussions of the positions they take.

They might, therefore, be more attentive to district-level cultural indicators. Legislators cannot assume, as they can with abortion, that their attitudes are consistent with the modal attitudes of their constituents. This forces legislators to respond to district-level cues to the extent that they are available. With gay rights, then, the expectation is that district-level cultural variables will emerge as strong predictors of legislator decision making.

MODELING CONGRESSIONAL SUPPORT FOR GAY RIGHTS

Gay issues make their way onto the legislative agenda in the same way that every other policy issue does—legislators introduce policy initiatives in an effort to credit claim, position-take, or engage in symbolic posturing or legislative advocacy (Mayhew 1974). Legislators view gay issues as salient and needing a federal response, and they draft legislation accordingly. With that in mind, we proceed in an attempt to address the question: How do legislators decide? Given that these issues represent a cultural cleavage drawing on status concerns (Smith and Tatalovich 2003; Tatalovich and Smith 2001), what sorts of considerations affect their decision making? We expect that among legislators, cultural traditionalism7 will diminish support for pro-gay policies, while progressivism will enhance support for pro-gay policies. Partisanship, ideology, and religious identification are used to approximate cultural affect.

The Dependent Variables

Our dependent variables are additive indices that reflect the level of pro-gay decision making by each legislator in the areas of voting and co-sponsorship in each Congress from the 103rd through the 107th. Relevant legislation was selected after consultation with the Human Rights Campaign (HRC) and by conducting an independent search of the Congressional Record.8 To

58 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

calculate the voting index, each legislator was given a score of +1 for each pro-gay vote, −1 for each anti-gay vote, and a 0 for each “present” vote and missed vote.9 High scores indicate high levels of support for progressive gay policies. To calculate the sponsorship/co-sponsorship index,10 each legislator was given a +1 for each time he/she sponsored or co-sponsored a gay- friendly bill, and a 0 for every time he/she did not. Again, high scores indicate high levels of support for progressive gay policies.

The Independent Variables

Partisanship, ideology, and religion are used to approximate culture. Legislator partisanship is a dummy variable (Democrats = 1, Republicans = 0). District-level partisanship is operationalized using the percentage of the vote for the Democratic candidate in each congressional district in the most recent presidential election (which is the 1992 election for the 103rd and 104th Congresses, the 1996 election for the 105th and 106th Congresses, and the 2000 election for the 107th Congress). We expect Democratic partisanship in the House to produce higher levels of support for pro-gay policies (Campbell and Davidson 2000; Haider-Markel 2001, 1999). At the district level, areas of higher Democratic partisanship tend to be more supportive of progressivism (Haider-Markel and Meier 1996; Wald, Button, and Rienzo 1996). Thus, we expect that legislators representing districts with higher levels of Democratic partisanship will be more supportive of pro-gay policies.

Poole and Rosenthal’s DW-NOMINATE scores measure legislator ideology (Poole and Rosenthal 1991). These scores range from −1 to +1, with −1 being the most liberal and +1 being the most conservative. However, there were high levels of collinearity between elite parti- sanship and ideology. Rather than suffer unstable coefficients, we reduce collinearity by regressing ideology on elite partisanship and using the residuals from that equation in place of the DW- NOMINATE scores (Haider-Markel 2001; Wattier and Tatalovich 1995). This allows for a more confident interpretation of the direction and strength of the relationship between ideology and support for gay rights, although it makes substantive interpretations of the coefficient more diffi- cult. Because it is associated with traditionalism and the stability of social norms, conservatism is expected to depress support for pro-gay policies.

Religion has been shown to structure attitudes across an array of issues (Jelen 1997), and it is particularly relevant to the policy-making dynamic investigated here, given the importance of gay issues to religious communities and churches. Legislator religion is measured using cat- egorical variables for Roman Catholic (Catholic = 1) and Protestant conservative (Protestant conservative = 1).11 In order to isolate African-American elites from white Catholics and Protes- tants, the models also include a dummy for African-American Protestants (African-American Protestants = 1). We use the percentage of each congressional district classified as Catholic and Protestant conservative to measure district-level religion. Because of the difficulty in deriving an accurate count of African-American Protestants, the percentage of African Americans in each district is used as a proxy. District-level data are taken from Bradley et al. (1992).12

To be sure, ideal measures of religion would ascertain the level of religious commitment on the part of individual adherents. Although Guth and Kellstedt (2001) and Yamane and Oldmixon (2006) have made good strides in quantifying religious “salience” at the elite level, such measures often remain profoundly difficult to quantify (Benson and Williams 1982). It is also worth noting that Fastnow, Grant, and Rudolf (1999) have demonstrated the usefulness of categorical religious measures in the legislative arena. This makes sense, given that denominational identification is a central component of an individual’s religious identity, and it “remains a strong source of po- litical attitudes and orientations” (Wilcox, Jelen, and Leege 1993:89). Indeed, denominational identification encapsulates “differences in belief, practice, and commitment, even for individu- als with minimal religiosity. Thus, [one] would expect denominational preference to influence political attitudes and behavior” (Kellstedt and Green 1993:55). Denominational identification

RELIGIOUS DYNAMICS OF DECISION MAKING ON GAY RIGHTS 59

provides individuals with a group identity that is attached to “broader cultural traditions,” and it is politically important for that reason (1993:65).

Opponents of gay rights often see progressive sexuality as a threat to traditionalism and “proper” social values (Wilcox 2000). They are guided by a moral vision that considers traditional families and sexual relations as preferred by God and, therefore, correct and worthy of honor. These people are often affiliated with religious denominations that espouse high levels of theological and social conservatism (Button, Rienzo, and Wald 1997: ch. 6; Wald 1997: ch. 6). Thus, we expect that legislators who are conservative Protestants and legislators with greater proportions of conservative Protestants in their districts will be less supportive of gay rights.

Legislator identification as a Catholic and the presence of Catholics at the district level are also expected to produce lower levels of support for pro-gay policies. However, the relationship might be weaker for a number of reasons.13 First, Catholic opinion tends to be more “permissive” than evangelicals with regard to sexual morality and homosexuality (Jelen 1997:61). Second, while opposed to gay marriage, the Church calls for the humane treatment of homosexuals.14

Perhaps it is not surprisingly then, that recent public opinion research from the Pew Forum on Religion and Public Life finds majorities of Catholics from all theological persuasions favor gay rights (as distinct from gay marriage) (Green 2004). The Pew findings help to place in focus the question of intra-denominational conflict on cultural and theological issues (Wuthnow 1988). To the extent that such conflict may occur in some denominations along certain issue dimensions, it appears not to exist within the American Catholic public regarding gay rights.

Three additional district-level sociodemographic variables are included in the analysis.15

These variables are generally employed when studying economic policies because they approxi- mate community-level demand for government action (Wald, Button, and Rienzo 2001). However, it is likely that they will also affect cultural policy making. Percentages of the district classified as Latino and African American are included to control for the attitudinal and behavioral racial patterns that have been documented by various scholars (Conway 2000; Glynn et al. 1999; Erikson and Tedin 1995). The scholarship assessing the relationship between race and support for gay rights finds mixed results. For example Button, Rienzo, and Wald (1997: ch. 3) note that while African Americans tend to support the notion of equality opportunities for homosexuals, they are less supportive of civil rights protections. Because higher levels of education are usually associated with higher levels of tolerance, we expect this to be associated with support for pro-gay policies (Button, Rienzo, and Wald 1997: ch. 3; Haider-Markel and Meier 1996; Haider-Markel et al. 2000; Wald, Button, and Rienzo 1996; Wilcox and Wolpert 2000). This variable is operationalized as the percentage of each district with a college degree.

Haider-Markel (1999) demonstrates the importance of taking interest group activity into account when considering the conditions under which legislators will be supportive of gay rights. In a prima facie sense, it seems that interest groups should not have much influence over legislative behavior when dealing with high profile moral issues. However, they may have some success at the margins by convincing members that a given piece of legislation makes only incremental changes from existing policy. Interest group activity is operationalized as the amount of money contributed to the primary and general election campaigns of representatives in the election cycle prior to the start of each Congress.16

ANALYSIS

We derive estimates using logit and ordered logit. Table 1 reports the results of the roll call analysis from all Congresses under consideration, and Table 2 reports the results of the sponsorship/co-sponsorship analysis. The statistically significant chi-square statistic for each model in each Congress indicates that the model performs robustly across time, and that the independent variables significantly increase our ability to explain both voting behavior and sponsorship/co-sponsorship decision making on this issue. The coefficients indicate the direction

60 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

TABLE 1 ANALYSIS OF SUPPORT FOR PROGRESSIVE GAY POLICIES,

ROLL CALL INDICES

103rd 104th 105th 106th 107th Congress Congress Congress Congress Congress

Latino (%) 0.017/(0.121) 0.019/(0.068) 0.042/(0.001) 0.019/(0.119) 0.024/(0.025) African American (%) −0.029/(0.114) −0.031/(0.099) −0.020/(0.239) −0.004/(0.848) −0.025/(0.192) College educated (%) 0.088/(0.000) 0.069/(0.000) 0.041/(0.052) 0.043/(0.032) 0.047/(0.010) Democrat (%) 0.039/(0.130) 0.090/(0.000) 0.046/(0.042) −0.014/(0.566) 0.014/(0.513) Catholic (%) −0.027/(0.069) −.005/(.709) −0.044/(0.008) −0.003/(0.844) −0.028/(0.055) Conservative −0.066/(0.001) −0.069/(0.001) −0.078/(0.000) −0.056/(0.014) −0.064/(0.001)

Protestant (%) HRC 0.000/(0.000) 0.000/(0.005) 0.001/(0.002) 0.000/(0.001) 0.000/(.000) Party (Dem = 1) 4.031/(.000) 2.865/(0.000) 4.687/(0.000) 6.918/(0.000) 6.473/(0.000) Ideology −8.311/(0.000) −4.722/(0.000) −9.007/(0.000) −15.254/(0.000) −8.378/(0.000) White Catholic −0.717/(0.038) −0.570/(0.075) −0.500/(0.154) −0.212/(0.582) −0.468/(0.196) White Conservative −1.040/(0.082) −0.519/(0.372) −0.296/(0.548) −0.140/(0.802) 0.490/(0.316)

Protestant African-American −0.209/(0.823) 0.118/(0.898) −0.767/(0.521) −123/(0.920) −0.184/(0.856)

Protestant

N 421 417 424 421 420 Pseudo R2 0.5266 0.5145 0.5855 0.6454 0.6122 Log likelihood −197.44252 −195.23647 −183.41976 −154.27447 −173.70176 LR χ 2 (12) 439.32 413.80 518.14 561.53 548.46

Notes: Entries are ordered logit coefficients. χ 2 significant across all models. P values in parentheses. Two-tailed tests.

of the relationship between each independent variable and dependent variable. The p values in parentheses indicate the level of statistical significance that each relationship achieves. Tables 3 and 4 report the predicted probabilities of supporting gay issues at the highest levels.

Note that across all models, ideology, district level of education, legislator partisanship, and the level of HRC contributions emerge as strong and significant predictors of legislator decision making. The negative coefficients indicate that ideological conservatism produces lowers levels of support for pro-gay policies, regardless of whether legislators are voting or making sponsorship/co-sponsorship decisions. The coefficients are significant at the 0.000 level in all models. The substantive effect of ideology is considerable. Clearly, this is an issue that divides political elites along ideological lines.

This is also an issue that divides individuals on the basis of socioeconomic status, using educational attainment as an indicator thereof. District level of education is strongly associated with pro-gay legislator decision making, and the relationship is highly and reliably significant. When voting, legislators from the most highly educated districts are at least 350 percent more likely to offer strong support for pro-gay policies than legislators from minimally educated districts. Table 4 suggests that the substantive effect of educational attainment is even more impressive regarding sponsorship/co-sponsorship decisions. This is an important point, and it suggests that education may serve as a useful indicator of (or proxy for) cultural traditionalism versus cultural progressivism (Oldmixon 2002). Legislators from highly educated districts do not just support gay issues when called to the floor for a roll call vote, they are also likely to go out of their way to support gay issues before they get to the floor. That is, they are highly motivated to take positions on this issue at every opportunity.

RELIGIOUS DYNAMICS OF DECISION MAKING ON GAY RIGHTS 61

TABLE 2 ANALYSIS OF SUPPORT FOR PROGRESSIVE GAY POLICIES,

SPONSORSHIP INDICES

103rd 104th 105th 106th 107th Congress Congress Congress Congress Congress

Latino (%) −0.004/(0.703) 0.030/(0.064) −0.003/(0.808) −0.002/(0.855) 0.023/(0.048) African American (%) −0.024/(0.218) −021/(0.473) −0.028/(0.297) −0.057/(0.010) −0.024/(0.288) College educated (%) 0.089/(0.000) 0.138/(0.000) 0.085/(0.002) 0.092/(0.000) 0.077/(0.000) Democrat (%) 0.076/(0.004) 0.087/(0.032) 0.113/(0.002) 0.143/(0.000) 0.096/(0.000) Catholic (%) 0.016/(0.303) −0.029/(0.167) 0.006/(0.747) −0.017/(0.290) −0.017/(0.263) Conservative −0.047/(0.039) −0.124/(0.000) −0.044/(0.118) −0.074/(0.010) −0.053/(0.016)

Protestant (%) HRC 0.000/(0.000) 0.000/(0.038) 0.001/(0.000) 0.000/(0.000) 0.000/(0.000) Party (Dem = 1) 3.603/(0.000) 3.591/(0.000) 4.005/(0.000) 3.690/(0.000) 6.091/(0.000) Ideology −5.914/(0.000) −5.978/(0.000) −7.591/(0.000) −5.284/(0.000) −9.912/(0.000) White Catholic −0.678/(0.061) −0.572/(0.235) −0.504/(0.241) −0.390/(0.305) −0.179/(0.627) White Conservative −1.647/(0.050) −476/(0.615) 0.185/(0.799) −0.842/(0.277) −0.645/(0.307)

Protestant African-American −1.650/(0.069) −0.993/(0.525) 1.094/(0.431) −0.432/(0.669) −1.275/(0.262)

Protestant (%) Constant – −7.860/(0.000) – – – N 423 428 425 423 430 Pseudo R2 0.4801 0.6722 0.6487 0.6131 0.6602 Log likelihood −185.12677 −87.713584 −129.64984 −142.3947 −149.76346 LR χ 2 (12) 341.86 359.72 478.92 451.30 581.94

Notes: Entries are logit or ordered logit coefficients. χ 2 significant across all models. P values in parentheses. Two-tailed tests.

As expected, Democratic legislators are more likely to support pro-gay policies than their Republican colleagues. The relationship is strongly significant and substantively impressive. This polarization between partisan elites makes complete sense, given the growing religious and cul- tural polarization among party activists over the last three decades (Layman 1999). As Democratic activists have become increasingly secular and supportive of cultural progressivism, and as Re- publican activists have become increasingly evangelical and supportive of cultural traditionalism, congressional candidates are also more likely to reflect those respective cultural groups. Indeed, candidates are themselves party activists. When voting, House Democrats are at least 14 times more likely to support pro-gay policies at the highest levels than their Republican colleagues, and we find a similar pattern in the sponsorship/co-sponsorship models.

It is worth noting, however, that the influence of elite partisanship in these models is sub- ject to more than one interpretation. Party voting on gay issues may reflect the ability of party leaders to enforce discipline more than it does solidarity with cultural groups and norms. To be sure, roll call voting is done under the watchful eyes of leaders, and leaders have considerable powers at their disposal to encourage party discipline (Cox and McCubbins 1993). The effect of congressional parties on decision making with respect to cultural issues, however, is probably minimal. These are often considered votes of conscience (Snyder and Groseclose 2000). What is more, we find that elite partisanship influences decision making in the roll call voting and sponsorship/co-sponsorship models. Sponsorship/co-sponsorship decisions allow legislators to take public positions, but they have no direct effect on policy.

62 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

TABLE 3 PREDICTED PROBABILITIES OF SUPPORTING PROGRESSIVE GAY POLICIES,

ROLL CALL INDICES

103rd 104th 105th 106th 107th Value Congress Congress Congress Congress Congress

Baseline 0.019 0.005 0.106 0.005 0.004 Latino (%) Min 0.017 0.005 0.079 0.004 0.003

Max 0.081 0.027 0.662 0.025 0.027 % � 382.80 498.53 740.21 481.93 836.28

African American (%) Min 0.027 0.007 0.130 0.005 0.005 Max 0.006 0.002 0.052 0.009 0.002 % � −76.93 −79.18 −60.01 70.22 −63.12

College educated (%) Min 0.006 0.002 0.065 0.003 0.002 Max 0.237 0.046 0.295 0.023 0.019 % � 4061.39 2201.78 354.62 736.17 964.55

Democrat (%) Min 0.009 0.001 0.021 0.013 0.003 Max 0.130 0.232 0.448 0.006 0.010 % � 1377.93 30,095.57 2,039.79 −58.04 280.91

Catholic (%) Min 0.036 0.006 0.233 0.006 0.007 Max 0.007 0.005 0.021 0.006 0.001 % � −80.12 −14.03 −90.79 1.59 −83.30

Cons. Protestant (%) Min 0.050 0.015 0.269 0.012 0.010 Max 0.001 0.000 0.002 0.000 0.000 % � −98.84 −98.95 −99.42 −97.08 −98.83

HRC Min 0.013 0.005 0.042 0.004 0.002 Max 0.931 0.037 0.965 0.327 0.164 % � 7171.44 708.82 2,173.80 8.757.39 6,992.26

Ideology Min 0.931 0.378 0.998 0.996 0.929 Max 0.000 0.000 0.000 0.000 0.000 % � −99.99 −99.99 −99.97 −100.00 −100.00

Party (Dem = 1) 1 0.502 0.079 0.918 0.810 0.674 % � 2508.49 1417.35 769.37 15,972.38 18,729.21

White Catholic 1 0.010 0.003 0.069 0.004 0.002 % � −47.88 −40.64 −34.95 −15.95 −34.84

White Cons. 1 0.008 0.004 0.084 0.005 0.006 Protestant % � −58.17 −32.25 −20.30 3.68 76.09

African-American 1 0.022 0.008 0.081 0.008 0.004 Protestant % � 14.11 60.99 −23.05 56.48 23.99

Note: Variables were set at their mean value for continuous variables and at 0 for dichotomous variables. Probabilities were estimated with the CLARIFY for Stata 8. Probabilities are rounded.

Therefore, while party leaders have the ability and the incentive to whip votes toward the party position, there is very little incentive to exert influence over sponsorship/co-sponsorship decisions—particularly among majority party leaders given that they ultimately control access to the floor. Thus, at a stage in the legislative process not controlled by leadership or influenced by rules and committees, partisanship still emerges as significant. At least in this policy domain, party voting likely indicates affinity toward a set of values on the part of legislators.

District-level Democratic partisanship is also associated with higher levels of support for gay issues in both the roll call and sponsorship/co-sponsorship models. To the extent that Democratic

RELIGIOUS DYNAMICS OF DECISION MAKING ON GAY RIGHTS 63

TABLE 4 PREDICTED PROBABILITIES OF SUPPORTING PROGRESSIVE GAY POLICIES,

SPONSORSHIP INDICES

103rd 104th 105th 106th 107th Value Congress Congress Congress Congress Congress

Baseline 0.005 0.031 0.019 0.000 0.000 Latino (%) Min 0.006 0.025 0.020 0.000 0.000

Max 0.006 0.251 0.025 0.000 0.003 % � 1.05 914.73 23.61 18.44 810.40

African-American (%) Min 0.007 0.039 0.027 0.001 0.001 Max 0.003 0.038 0.013 0.000 0.000 % � −63.12 −3.40 −54.00 −95.90 −46.77

College educated (%) Min 0.002 0.005 0.006 0.000 0.000 Max 0.082 0.625 0.225 0.008 0.006 % � 5,245.32 11,387.86 3,431.73 8,261.95 3,899.93

Democrat (%) Min 0.001 0.006 0.000 0.000 0.000 Max 0.162 0.508 0.637 0.197 0.036 % � 15,141.06 8,753.12 211,626.95 15,494,333.07 117,489.90

Catholic (%) Min 0.004 0.057 0.018 0.001 0.001 Max 0.013 0.015 0.035 0.000 0.000 % � 223.53 −73.31 89.85 −57.58 −57.06

Cons. Protestant (%) Min 0.011 0.161 0.039 0.001 0.001 Max 0.001 0.000 0.006 0.000 0.000 % � −92.21 −99.87 −85.93 −98.21 −95.37

HRC Min 0.004 0.025 0.011 0.000 0.000 Max 0.328 0.309 0.723 0.559 0.010 % � 7,678.84 1,117.61 6,463.16 219,611.76 3,307.31

Ideology Min 0.540 0.892 0.945 0.023 0.870 Max 0.000 0.000 0.000 0.000 0.000 % � −99.99 −99.99 −99.99 −99.92 −100.00

Party (Dem = 1) 1 0.146 0.488 0.494 0.012 0.123 % � 2,643.38 1.496.28 2,437.60 3,271.12 29,650.17

White Catholic 1 0.003 0.019 0.013 0.000 0.000 % � −45.14 −38.39 −34.82 −29.13 −13.93

White Cons. 1 0.002 0.027 0.028 0.000 0.000 Protestant % � −71.36 −11.88 44.75 −39.49 −37.82

African-American 1 0.002 0.027 0.094 0.000 0.000 Protestant % � −70.84 −12.45 384.42 3.15 −54.44

Note: Variables were set at their mean value for continuous variables and at 0 for dichotomous variables. Probabilities were estimated with the CLARIFY for Stata 8. Probabilities are rounded.

partisanship is associated with tolerance for out-groups and alternative lifestyles, strongly Demo- cratic districts are likely to be more progressive across an array of issues. Decision making on the part of legislators reflects responsiveness to those district-level patterns. The relationship between district partisanship and sponsorship/co-sponsorship decision making is consistently significant. Legislators from strongly Democratic districts are at least 34 times more likely to support gay is- sues at the highest levels. However, the significance of this relationship falls off in three of the vote models. One possible interpretation of this finding is that while congressional elites are polarized

64 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

on gay issues on the basis of partisanship, constituents are not (Lindaman and Haider-Markel 2002). As a result, the relationship between legislator partisanship and legislator decision making is strongly and consistently significant, while the relationship between district partisanship and roll call voting is not.

The coefficients for HRC contributions indicate that there is a positive and statistically sig- nificant relationship between interest group electoral support and support for gay rights. That is, higher contribution levels are associated with higher levels of legislator support for gay issues. In the 104th Congress roll call model, candidates receiving the greatest contributions for HRC were 709 percent more likely to offer the highest level of support for gay issues—but in the other models, the substantive effect is much greater. Of course, one should be cautious in the interpre- tation of these results (Smith 1995). This is not to suggest that campaign donations cause higher levels of support for gay issues. Perhaps lobbyists do buy votes and access with their donations. We simply do not know, based on the empirical evidence presented here. Even if that is the case sometimes or generally, it is probably not the case with a high salience, nontechnical issue such as gay rights. However, lagged campaign support and legislator decision making are clearly related. What these findings do suggest is that HRC works very hard to identify and elect representatives whose stated preference is for gay rights.

The relationship between decision making and district-level conservative Protestantism achieves significance in nine of the ten models, and is nearly significant in the tenth. The negative coefficients indicate that the proportion of conservative Protestants is negatively associated with support for pro-gay policies. In the models where this relationship achieves two-tailed signif- icance, legislators representing the largest conservative Protestant communities are at least 90 percent less likely to strongly support pro-gay policies. The proportion of Catholics at the district level is also negatively associated with support for pro-gay policies (for the most part); however, the relationship only achieves significance in the 103rd, 105th, and 107th Congresses roll call models.

Even so, the substantive effect of Catholicism in those three Congresses is impressive. As Table 3 indicates, representing strongly Catholic districts reduces the likelihood of offering strong support for gay issues by at least 80 percent. This is likely the result of two causes. First is the Church’s aforementioned nuance on gay issues. Second is the lack of consistent institutional guidance on gay issues from Catholic elites. Many Catholics may not be receiving clear messages from their priests and bishops (especially if gay marriage is not an issue separated from gay rights in general), which means constituents may not be sending clear messages to their legislators on gay rights policies. What is more, the most politically active Catholics are stratified on partisan, ideological, and cultural lines, based on the salience of their religious identification (see, e.g., Layman 1999). Unlike conservative Protestants, there is wide variation among Catholic laity with regard to cultural and political preferences, which likely contributes to the lackluster performance of the Catholic district indicator.

Most striking about these results, however, is that representative religion does not have a consistently significant relationship with support for gay issues. Catholic legislators are less likely to support gay issues, but this relationship is only significant in one of the sponsorship/ co-sponsorship models. In the vote models, the relationship is significant in the 103rd and 104th Congresses, and it approaches significance in the 105th and 107th Congresses. When the rela- tionship between Catholic identification and decision making is significant, Catholic legislators are at least 40 percent less likely to support gay issues than the reference group—mainline and liberal Protestants. Conservative Protestants are also less likely to champion gay issues, but the relationship is only significant in the 103rd Congress.

Clearly, the effects of religion present an interesting puzzle, as there seems to be a disparity in the influence of district versus legislator religion. Catholic and conservative Protestant com- munities at the district levels depress legislator support for pro-gay policies. In contrast, legislator religion has little measurable effect on decision making, with an exception for roll call voting

RELIGIOUS DYNAMICS OF DECISION MAKING ON GAY RIGHTS 65

behavior by Catholics in a subset of the Congresses. Opposition to gay rights has been a major agenda item for many Christian Right groups for about 30 years (Wilcox 2000). It has been their rallying cry. So it comes as no surprise that as the salience of this issue increases in the 1990s (with the gays in the military and Defense of Marriage Act battles), risk-adverse legislators respond to grassroots religious conservatism. What is surprising is that the effect of elite conservative Protestant identification is so muted when the other variables are controlled.

One interpretation of this finding is that the religious imperative to legitimate a particular moral order comes from the district—not from within. That is, legislators position-take on roll calls and sponsorships/co-sponsorships in a way that is sensitive to district concerns; they are less interested in affirming their own moral order. However, the direct effects of religious identification on decision making may be obscured by the effect of partisanship and ideology.

For example, in an analysis of roll call voting in the Wisconsin legislature, Yamane and Oldmixon (2006) find that while the direct effects on abortion roll calls are dominated by parti- sanship, conservative Protestant identification has an indirect effect through partisanship. Thus, partisanship mediates the influence of religion. While it goes beyond the findings for this article to draw definitive conclusions in this regard, the same dynamic may be at work in this analysis. This makes sense in light of Layman’s (1999) findings on the religious and cultural partitioning of party activists.

Variables tapping the influence of race on legislator decision making perform inconsistently. Legislator identification as an African-American Protestant seems unrelated to decision making on this issue. The relationship is for the most part negative, but it is only significant in the 103rd Congress sponsorship/co-sponsorship model. The proportion of the African-American population at the district level is negatively associated with support for gay issues, but the relationship is only significant in the 104th Congress roll call model and the 106th Congress sponsorship/ co-sponsorship model.

The weak performance of elite and district measures of African-American identification may be attributable to a number of factors. First, while African Americans express traditional (or conservative) opinions on many issues of social morality, they tend to “draw a sharp distinction between rejecting a certain type of behavior and enacting public policy to forbid it. Though clearly hostile to homosexuality,” African Americans tend more than whites to oppose “employment discrimination against gays” (Wald 2003:278). What is more, the conservative views expressed by African Americans do not seem related to cultural conservatism, per se. Gay rights has a different symbolic meaning for black and white evangelicals. For the latter group, opposition to gay rights is strongly associated with their affect toward homosexuals. For the former group, the magnitude of this association is weaker (Calhoun-Brown 1997:124–26).

In the roll call models, the size of the Latino population is positively related to support for pro-gay policies. The relationship is significant in the 104th, 105th, and 107th Congresses, and it approaches significance in the 103rd and 106th Congresses. The relationship is also significant and positive in the 104th and 107th Congress sponsorship/co-sponsorship models. The effect of the district-level race variables is interesting. Latinos and African Americans both tend to support Democrats (although this is less the case with Latino voters); both communities tend to embrace traditionalism on matters of morality; and the Latino community is strongly Catholic. Yet they affect roll call voting in different ways, with strongly Latino districts enhancing support for gay issues, and strongly African-American districts depressing support for gay issues.

One final note, a dummy variable for “out” gay or lesbian representatives was originally included in all models. The dummy was included as a way to explore Haider-Markel, Joslyn, and Kniss’s (2000) argument that while sympathetic heterosexual legislators can provide good representation for gays and lesbians, gays and lesbians tend to make better representatives of their community’s interests. Since an accurate count of gays in Congress cannot be made, the analysis included gays who were “out” at the time, or subsequently came “out.” The variable was eventually dropped from the models because in some Congresses it was a perfect predictor of

66 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

support for gay issues. While this suggests that even for gays and lesbians there is some value to “descriptive” representation beyond the symbolic element, we could not test this argument, as such.

CONCLUSIONS

Our effort examines legislator decision making on “gay” issues. As we note above, gay issues represent one of the most salient political cleavages of the postmaterialist era—the cleavage between religious traditionalism and progressive sexuality. Therefore, the analysis proceeded on the assumption that on gay issues, legislators are driven by a desire to legitimate a culture, a set of values, in law. We predicted that variables reflective of culture would affect decision making, and, in the end, that expectation is largely fulfilled. Ideology, religion, and partisanship—all structure legislative decision making in predictable ways. So, too, the interest group activity and district educational levels.

Interestingly, cultural phenomena at both the district and elite levels affect decision making. While previous research on abortion found that legislator partisanship, religion, and ideology informed legislator decision making, we found that district-level variables significantly shape de- cision making on gay issues. In a more systematic way than Oldmixon (2002) found in her abortion analyses, legislators appear to be synthesizing district-level attitudes in their own decision-making calculus. Legislators do not simply bow to the pressures of their constituents; nor do they act as brave trustees. That dichotomy probably does not make much sense in this policy domain. Rather, cultural phenomena are almost certainly mutually reinforcing, and this likely produces dyadic correspondence between legislators and the modal preferences of their districts. This is consis- tent with Rohde (1991) and Aldrich and Rohde’s (2004) argument that congressional parties are internally homogenous in the preferences and externally polarized against one another. This po- larization is rooted in electoral politics.

Thus, mainline Protestant, Democratic, highly educated districts elect members of Congress who share their values—at least on high salience issues. In the same way, strongly Catholic and evangelical districts that vote Republican are likely to elect conservative Republicans. To the extent, then, that mass-level religious, partisan, and ideological cleavages form around gay issues, those same cleavages are likely to emerge in Congress. Interestingly enough, these cleavages emerge at multiple stages of the legislative process. Sponsorship and co-sponsorship is a simple position-taking opportunity. Vulnerable members can demonstrate aggressive legislative activity, and minority members can advocate for policy when they are denied access to the floor (Kroger 2003). Theoretically, the stakes of roll call voting are higher because policy is made at this stage of the legislative process. And yet, these cultural cleavages emerge throughout.

Finally, we note that the stark polarization of elites on gay issues and cultural issues more generally likely augments the historical cleavages between House members along the economic policy dimension. This can only intensify the level of division within the institution. The rise of the cultural agenda fosters legislator differences of a more absolutist sort than has been previously known. These issues arouse the passions of party activists—the same party activists who play a key role in selecting their party’s congressional candidates, and who likely reject compromise (Layman and Carmines 1997; Smith and Tatalovich 2003:17).

As a result, legislators do not simply disagree over an instrumental policy goal; the fundamental issue frame is in dispute. What for one group is clearly a religious wrong, for another is clearly a civil right. In that context, legislators become the defenders of a way of life—possibly a divinely ordained way of life. These are just the kind of wedge issues that di- vide parties and challenge governance (Dodd and Oppenheimer 1997, 2001; Lowi 1995). To the extent that culturally controversial issues, including gay issues, remain salient, the House of

RELIGIOUS DYNAMICS OF DECISION MAKING ON GAY RIGHTS 67

Representatives will be challenged to address them in ways that mitigate their divisiveness and that can work toward deliberation and compromise.

ACKNOWLEDGMENTS

The authors thank Ted Jelen and the JSSR’s anonymous referees for their helpful comments and suggestions for revision.

NOTES

1. Leege et al. (2002) characterize the cleavage as between religious traditionalism and the growth of feminism. However, given the work of Inglehart (1990) and Lowi (1998), it is useful to construe the cleavage more broadly as between religious traditionalism and the growth of progressivism.

2. Paulson, Michael, “The gay marriage ruling reaction pro and con/religious community; strong divided opinions mark clergy response.” The Boston Globe [November 19, 2003] B6.

3. From the 2004 State of the Union Address, http://www.whitehouse.gov/news/releases/2004/01/20040120-7.html. 4. From the 2004 State of the Union Address, http://www.whitehouse.gov/news/releases/2004/01/20040120-7.html. 5. See Federalist 10, obviously. 6. When Congress has considered gay issues, it has tended to be hostile (Ellis 1998). 7. While it is outside the scope of this article, it is worth noting that at the mass level, Brewer (2003b) finds that the effect

of moral traditionalism (or, cultural traditionalism for our purposes) on opinions related to gay issues is mediated by political knowledge.

8. The legislative items used to calculate the index for each Congress are listed in the Appendix. 9. We interpret not voting and voting “present” as ambivalence. Classifying both actions as zero allows for that ambiva-

lence to be quantified. This coding rule allows for more observations, but it can also be theoretically justified. One cannot know with complete certainty why a legislator might miss a given vote. However, whether a legislator misses a vote or votes “present,” they have forgone an opportunity to go on the record as either for or against gay issues. The nonchoice lacks affect and we treat it as such.

10. Sponsorship and co-sponsorship items were included in the analyses if the legislation in question advanced gay rights, had at least 49 co-sponsors, and did not receive a floor vote. This allows us to analyze different kinds legislative decisions. The benchmark of 49 co-sponsors ensures that there is enough variation for meaningful analysis. Representatives taking the affirmative step of signing on to high profile gay rights legislation—even though that legislation never reached the floor—were credited with supporting gay rights.

11. Denominations are classified in a way that is closely consistent with Kellstedt and Green (1993), Green and Guth (1991), and Roof and McKinney (1987). The following denominations are classified as Protestant conservative: Adventists, Christian Scientist, evangelical wings of mainline denominations (such as Missouri Synod Lutherans and Free Methodists), Mormon, Nazarene, American Baptist Convention, Assemblies of God, Baptist, Baptist Missionary Association, Brethren in Christ, Christian Missionary Alliance, Christian Reformed Church, Church of God, Church of Christ, Conservative Baptist Association, Evangelical, Independent Baptist, and Southern Baptist. It is worth noting that Mormons and traditional conservative Protestants have different sources of scripture and make different eschatological assumptions. Indeed, there are “massive doctrinal differences” between Mormons and Assemblies of God, for example. However, they are grouped together in this instance, based on each denomination’s support for traditional norms on issues of sexual morality (Wald, Button, and Rienzo 1996:1162). That said, it may not always be appropriate to group traditional and nontraditional conservatives together.

12. Religion data from the Glenmary Research Center (Bradley et al. 1992) are aggregated by state and county. John Green reaggregated these data by congressional districts and graciously provided them to the authors. See Green and Guth (1991).

13. While Catholic and conservative Protestant identification are both associated with religious traditionalism in this model, we note that the larger social agendas of both traditions diverge in many areas. On issues of sexual morality, however, they tend to complement each other.

14. See, for example, Persona Humana, available from: http://www.vatican.va/roman curia/congregations/cfaith/ documents/rc con cfaith doc 19751229 persona-humana en.html.

15. District-level socioeconomic and cultural variables are included to measure legislator responsiveness to social struc- ture. Haider-Markel (1999) suggests that these can be considered proxies for constituency opinion.

16. These data were obtained from the Center for Responsive Politics.

68 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

REFERENCES

Adams, G. D. 1997. Abortion: Evidence of an issue evolution. American Journal of Political Science 41:718–37. Aldrich, J. and D. Rohde. 2004. Congressional committees in the partisan era. In Congress reconsidered, edited by

Lawrence C. Dodd and Bruce I. Oppenheimer. Washington, DC: CQ Press. Benson, P. L. and D. L. Williams. 1982. Religion on Capitol Hill: Myths and realities. San Francisco, CA: Harper & Row. Bradley, M. B., N. M. Green, D. E. Johnson, M. Lynn, and L. McNeil. 1992. Churches and church membership in the

United States, 1990. Atlanta: Glenmary Research Center. Brewer, P. R. 2003a. The shifting foundations of public opinion about gay rights. Journal of Politics 65:1208–20. ——. 2003b. Values, political knowledge, and public opinion about gay rights. Public Opinion Quarterly 67:173–

201. Button, J. W., B. A. Rienzo, and K. D. Wald. 1997. Private lives, public controversies: Battles over gay rights in American

communities. Washington, DC: CQ Press. Calhoun-Brown, A. 1997. Still seeing in black and white: Racial challenges for the Christian right. In Sojourners in the

wilderness: The Christian right in comparative perspective, edited by Corwin E. Smidt and James M. Penning. Lanham, MD: Rowman and Littlefield Publishers.

Campbell, C. C. and R. H. Davidson. 2000. Gay and lesbian issues in the congressional arena. In The politics of gay rights, edited by C. A. Rimmerman, K. D. Wald, and C. Wilcox. Chicago: University of Chicago Press.

Conway, M. M. 2000. Political participation in the United States, 3rd ed. Washington, DC: CQ Press. Cox, G. W. and M. D. McCubbins. 1993. Legislative leviathan: Party government in the House. Berkeley: University of

California Press. Dodd, L. C. and B. I. Oppeheimer. 2001. Congress and the emerging order: Assessing the 2000 elections. In Congress

reconsidered, edited by L. C. Dodd and B. I. Oppenheimer. Washington, DC: CQ Press. ——. 1997. Congress and the emerging order: Conditional party government or constructive party government? In

Congress reconsidered, edited by Lawrence C. Dodd and Bruce I. Oppenheimer. Washington, DC: CQ Press. Ellis, M. 1998. Gay rights: Lifestyle or immorality? In Moral controversies in American politics: Cases in social regulatory

policy, edited by R. Tatalovich and B. Daynes. Armonk: M.E. Sharpe. Erikson, R. S. and K. L. Tedin. 1995. American public opinion, 5th ed. New York: Addison Wesley Longman, Inc. Fairbanks, D. 1977. Religious forces and “morality” policies in the American States. Western Political Quarterly 30:411–

17. Fastnow, C., J. T. Grant, and T. J. Rudolf. 1999. Holy roll calls: Religious tradition and voting behavior in the U.S. House.

Social Science Quarterly 80:687–701. Glynn, C. J., S. Herbst, G. J. O’Keefe, and R. Y. Shapiro. 1999. Public opinion. Boulder: Westview Press. Green, J. C. 2004. American religious landscapes and political attitudes. Pew Forum on Religion and Public Life. Available

from: http://pewforum.org/docs/index.php?DocID=55 Green, J. C. and J. L. Guth. 1991. Religion, representatives, and roll calls. Legislative Studies Quarterly XVI:571–84. Guth, J. L. and L. A. Kellstedt. 2001. Religion and Congress. In In God we trust: Religion & American political life,

edited by Corwin E. Smidt. Grand Rapids: Baker Academic. Haeberle, S. 1996. Gay men and lesbians at city hall. Social Science Quarterly 77:190–97. Haider-Markel, D. P. 1999. Redistributing values in Congress: Interest group influence under sub-optimal conditions.

Political Research Quarterly 52:113–44. ———. 2001. Morality in Congress? Legislative voting on gay issues. In The public clash of private values: The politics

of morality policy, edited by C. Z. Mooney. Chatham: Chatham House Press. Haider-Markel, D. P., M. R. Joslyn, and C. J. Kniss. 2000. Minority group interests and political representation: Gay

elected officials in the policy process. Journal of Politics 62:568–77. Haider-Markel, D. P. and K. J. Meier. 1996. The politics of gay and lesbian rights: Expanding the scope of the conflict.

Journal of Politics 58:332–49. Inglehart, R. 1990. Cultural shift in advanced industrial society. Princeton: Princeton University Press. Jelen, T. G. 1997. Religion and public opinion in the 1990s: An empirical overview. In Understanding public opinion,

edited by B. Norrander and C. Wilcox. Washington, DC: Congressional Quarterly Press. Kellstedt, L. A. and J. C. Green. 1993. Knowing God’s many people: Denominational preference and political behavior. In

Rediscovering the religious factor in American politics, edited by David C. Leege and Lyman A. Kellstedt. Armonk: M.E. Sharpe, Inc.

Kroger, G. 2003. Position taking and cosponsorship in the U.S. House. Legislative Studies Quarterly XXVIII:225–46. Layman, G. C. 1999. “Culture wars” in the American party system: Religious and cultural change among partisan activists

since 1972. American Politics Quarterly 27:89–121. Layman, G. C. and E. G. Carmines. 1997. Cultural conflict in American politics: Religious traditionalism, postmaterialism,

and U.S. political behavior. Journal of Politics 59:751–77. Leege, D. C., K. D. Wald, B. S. Krueger, and P. D. Mueller. 2002. The politics of cultural differences: Social change and

voter mobilization in the post-New Deal period. Princeton: Princeton University Press.

RELIGIOUS DYNAMICS OF DECISION MAKING ON GAY RIGHTS 69

Lindaman, K. and D. P. Haider-Markel. 2002. Issue evolution, political parties, and the culture wars. Political Research Quarterly 55:91–110.

Lowi, T. 1998. Foreword: New dimensions in policy and politics. In Moral controversies in American politics: Cases in social regulatory policy, edited by R. Tatalovich and B. W. Daynes. Armonk, NY: M.E. Sharpe.

——. 1995. The end of the Republican era. Norman: University of Oklahoma Press. Mayhew, D. R. 1974. Congress: The electoral connection. New Haven: Yale University Press. Meier, K. J. 1994. The politics of sin. Armonk: M.E. Sharpe. Meier, K. J. and C. Johnson. 1990. The politics of demon rum. American Politics Quarterly 18:404–29. Meier, K. J. and D. R. McFarlane. 1993. The politics of funding abortion. American Politics Quarterly 21:81–101. Mooney, C. and M. Lee. 2000. The influence of values on consensus and contentious morality policy: U.S. death penalty

reform, 1956–1982. Journal of Politics 62:223–39. ——. 1995. Legislating morality in the American states: The case of pre-Roe abortion regulation reform. American

Journal of Political Science 39:599–627. Morgan, D. and K. Meier. 1980. Politics and morality: The effect of religion on referenda voting. Social Science Quarterly

61:144–48. Oldmixon, E. 2002. Culture wars in the congressional theater: How the U.S. House of Representatives legislates morality,

1993–1998. Social Science Quarterly 83:775–88. Poole, Keith T. and Howard Rosenthal. 1991. Patterns of congressional voting. American Journal of Political Science

35:228–78. Rohde, D. W. 1991. Parties and leaders in the postreform House. Chicago: University of Chicago Press. Roof, W. C. and W. McKinney. 1987. American mainline religion. New Brunswick: Rutgers University Press. Smith, R. A. 1995. Interest group influence in the U.S. Congress. Legislative Studies Quarterly 20:89–139. Smith, T. A. and R. Tatalovich. 2003. Cultures at war: Moral conflicts in Western democracies. Orchard Park, NY:

Broadview Press. Snyder, Jr., James M., and T. Groseclose. 2000. Estimating party influence in congressional roll-call voting. American

Journal of Political Science 44:187–205. Steiner, G. Y., ed. 1983. The abortion dispute in the American system. Washington, DC: Brookings Institution. Tatalovich, R. and D. Schier. 1993. The persistence of ideological cleavage in voting on abortion legislation in the House

of Representatives, 1973. 1988. American Politics Quarterly 21:125–39. Tatalovich, R. and T. A. Smith. 2001. Status claims and cultural conflict: The genesis of morality policy. Policy Currents

10:2–8. Wald, K. D. 1997. Religion and politics in the United States, 3rd ed. Washington, DC: CQ Press. —–. 2003. Religion and politics in the United States, 4th ed. Lanham, MD: Rowman & Littlefield Publishers, Inc. Wald, K., J. Button, and B. Rienzo. 2001. Morality politics v. political economy: The case of school-based health center.

Social Science Quarterly 82:221–33. ——. 1996. The politics of gay rights in American communities: Explaining antidiscrimination ordinances and policies.

American Journal of Political Science 40:1152–78. Wattier, M. and R. Tatalovich. 1995. Senate voting on abortion legislation over two decades: Testing a reconstructed

partisanship variable. American Review of Politics 16:167–83. Wilcox, C. 2000. Onward Christian soldiers? The religious right in American politics, 2nd ed. Boulder: Westview Press. Wilcox, C., T. G. Jelen, and D. C. Leege. Religious group identifications: Toward a cognitive theory of religious mobiliza-

tion. In Rediscovering the religious factor in American politics, edited by David C. Leege and Lyman A. Kellstedt. Armonk: M.E. Sharpe, Inc.

Wilcox, C. and B. Norrander. 2002. Of moods and morals: The dynamics of opinion on abortion and gay rights. In Understanding public opinion, 2nd ed., edited by B. Norrander and C. Wilcox. Washington, DC: Congressional Quarterly Press.

Wilcox, C. and R. Wolpert. 2000. Gay rights in the public sphere: Public opinion on gay and lesbian equality. In The politics of gay rights, edited by C. A. Rimmerman, K. D. Wald, and C. Wilcox. Chicago: University of Chicago Press.

Wildavsky, A. 1987. Choosing preferences by constructing institutions: A cultural theory of preference formation. Amer- ican Political Science Review 81:3–44

Wuthnow, R. 1988. The restructuring of American religion: Society and faith since World War II. Princeton: Princeton University Press.

Yamane, D. and E. A. Oldmixon. 2006. Affiliation, salience, advocacy: Three religious factors in public policy-making. Legislative Studies Quarterly, in press.

70 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

APPENDIX

Votes

103rd Congress (1993–1994) H.Amdt. 188, H.R. 2492, roll no. 313; H.Amdt. 316, H.R. 2401, roll no. 460; H.Amdt. 317, H.R. 2401, roll no. 461; H.Amdt. 318, H.R. 2401, roll no. 462; H.Amdt. 751, H.R. 4649, roll no. 321; H.Amdt. 490, H.R. 6, roll no. 92

104th Congress (1995–1996) H.Amdt. 890, H.R. 2546, roll no. 759; H.Amdt. 1286, H.R. 3396, roll no. 314; H.R. 3396, Defense of Marriage Act on passage, roll no. 316

105th Congress (1997–1998) H.Amdt. 789, H.R. 4194, roll no. 349; H.Amdt. 855, H.R. 4276, roll no. 398; H.Amdt 871, H.R. 4380, roll no. 414

106th Congress (1999–2000) H. Amdt. 201, H.R. 1501, roll no. 223; H. Amdt. 284, H.R. 1691, roll no. 298; H. Amdt. 356, H.R. 2587, roll no. 346; Graham motion to instruct, H.R. 4205, roll no. 470; Conyers motion to instruct, H.R. 4205, roll no. 471; H.R. 4892, Scouting for All Act, motion to suspend the rules and pass the bill, roll no. 468

107th Congress (2001–2002) H.Amdt. 310, H.R. 2944, roll no. 352; H.Amdt. 311, H.R. 2944, roll no. 354; H.Amdt. 312, H.R. 2944, roll no. 353; H.R. 7, Charitable Choice Act of 2001, motion to recommit, roll no. 253

Sponsorship/Co-Sponsorship

103rd Congress (1993–1994) H.R. 431, Civil Rights Act of 1993; H.R. 4636, Employment Non-Discrimination Act of 1994

104th Congress (1995–1996) H.R. 1863, Employment Non-Discrimination Act of 1995

105th Congress (1997–1998) H.R. 1858, Employment Non-Discrimination Act of 1997; H.R. 3081, Hate Crimes Prevention Act of 1997

106th Congress (1999–2000) H.Con.Res. 259, Expressing the concern of Congress regarding human rights violations against lesbians, gay men, bisexuals, and transgendered individuals around the world; H.R. 1082, Hate Crimes Prevention Act of 1999; H.R. 2355, Employment Non-Discrimination Act of 1999; H.R. 3650, Permanent Partners Immigration Act of 2000

107th Congress (2001–2002) H.Con.Res. 173, International Human Rights Equality Resolution; H.R. 638, Domestic Partner- ship Benefits and Obligations Act of 2001; H.R. 690, Permanent Partners Immigration Act of 2001; H.R. 1343, Local Law Enforcement Hate Crimes Prevention Act of 2001; H.R. 2692, Employment Non-Discrimination Act of 2001