Elite Mobilization: A Theory Explaining Opposition to Gay Rights

Abstract

Media and scholastic accounts describe a strong backlash against attempts to advance gay rights. Academic research, however, increasingly raises questions about the sharply negative and enduring opinion change that characterizes backlash among the mass public. How can we reconcile the widespread backlash described by the media with the growing body of academic research that finds no evidence of the opinion change thought to be its hallmark trait? We argue that rather than widespread opinion change, what appears to be backlash against gay rights is more consistent with elite-led mobilization—a reaction by elites seeking to prevent gays and lesbians from achieving full incorporation in the polity. We present evidence from what is widely considered to be a classic case of anti-gay backlash, the 2010 Iowa Judicial Retention Election. Analysis of campaign contribution data in Iowa versus other states between 2010 and 2014, and voter roll-off data exploiting a unique feature of the 2010 retention election supports this argument. The results simultaneously explain how reports of backlash might occur despite increased support for gay rights, and an academic literature that finds no evidence of backlash.

American political history is characterized by an ongoing tension between those pursuing civil rights and those defending the status quo. While the barriers to political incorporation take many forms, laws preventing acceptance of and participation by traditionally marginalized groups like gays and lesbians gradually continue to fall. One explanation for advances in gay rights is the increase in public acceptance that typically precedes policy changes (e.g., Baunach 2012; Bishin and Smith 2013; Flores 2014; Garretson 2018).¹

The public plays a large role in advancing or impeding minority rights by supporting referenda and rewarding candidates who support their views. As a result, those defending the status quo often appeal to public opinion to prevent advances in minority rights. Of particular concern for minority rights advocates is the possibility of mass opinion backlash (MOB)—that pressing for policy gains will boomerang by mobilizing the public against them and make it more difficult for the group to achieve its objectives. If pushing for policy angers the masses, then the public might not just oppose the policy in question, or increase dislike or distrust of the group, but also help create lasting barriers to achieving policy gains (e.g., Rosenberg 1993; Klarman 2013).

The potential costs of MOB are substantial. Backlash can embolden opponents and dissuade supporters of gay rights. It may encourage candidates to advocate policies opposing the group, facilitate their election, and increase their desire to service the angry masses, while simultaneously dissuading activism among the undecided or even the sympathetic but electorally vulnerable. It may also disillusion activists and make future mobilization more difficult (e.g., Adam 2017). Symbolically, it can signal that the group and its concerns are neither accepted nor legitimate (e.g., Tate 2003). By inciting deep-seated opposition, backlash may effectively foreclose obtaining policy for the foreseeable future. In sum, by arousing negative opinion in a democratic society backlash poses a potentially significant threat to a groups inclusion and perceived legitimacy (see Pierceson 2013 for a review of these arguments).

Despite extensive accounts of backlash in legal and political studies of political science, and numerous journalistic accounts of anti-gay opinion backlash, the evidence of opinion backlash is more mixed than is widely appreciated.² While numerous studies find increased opposition to gay rights among at least some groups, recent scholarship questions the extent and even existence of anti-gay backlash with respect to attitudes about gay marriage (e.g., Barclay and Flores 2015; Kreitzer et al. 2014).

How can we reconcile the widespread anti-gay backlash described in media and academic accounts with the growing body of research that finds scant evidence of the opinion change thought to be its hallmark sign? We argue that what appears to be MOB is actually a byproduct of elite-led mobilization. Claims of backlash result from ambiguity in the use of the term “backlash” by scholars and journalists, and attempts by anti-gay elites and their organizations to contest public acceptance of, and policy advances in, gay rights. Instead of exclusively reflecting a shift in opinion against the group, journalists and scholars also use the term “backlash” to describe a reaction or counter-mobilization to challenges by those already opposed to gay rights.

Building on the important implications that different meanings of the term backlash confer, we develop a theory of elite-led mobilization, which argues that much of what is described as mass backlash is more consistent with elite driven opposition to the acceptance of gays and lesbians in society. While there are substantial differences between our theory and mass driven accounts of backlash as opinion change, crucial tests differentiating between mass and elite driven processes are hard to identify. Nonetheless, examining evidence surrounding a “most difficult” case, the Iowa judicial retention elections of 2010, we present substantial evidence to support our theory. The election is described as perhaps the classic contemporary case of backlash by news outlets ranging from CNN, the New York Times, and National Public Radio, and scholars widely attribute the judges' defeat to their ruling (e.g., Clopton and Peters 2013; Harris 2019).

1. The Importance and Definition of Backlash

Backlash is commonly used to describe the reaction to a marginalized group's challenge to a dominant group's status, power, or values (e.g., Lipsett and Raub 1970). Over time, the concept of backlash has become something of a conventional wisdom in understanding opposition to gay rights in the United States (e.g., Rosenberg 1993; Ball 2006; Klarman 2013). The development of the backlash narrative to explain opposition to gay rights parallels the backlash used to explain growth in public opposition to the black civil rights movement (e.g., Ball 2006; Weaver 2007). In both cases, backlash is defined consistent with Vesla Weaver's description of racial backlash: “…the politically and electorally expressed resentment that arises from racial advance, intervention, or excess” (Weaver 2007: 237).

While research has just begun to define and theorize about anti-gay backlash and to articulate what sorts of events are most likely to foster backlash, questions about what instigates the reaction or resentment and what form it takes are largely overlooked. While scholars of social movements, law and society, historians, and journalists documenting the history of the gay rights movement commonly recognize the different and important roles that elites and masses play, studies focusing on anti-gay backlash tend to focus on mass behavior (e.g., Bull and Gallagher 1996; Eskridge Jr, 2002; Fetner 2008; Fejes 2008; Burack 2008). Synthesizing research on psychological threat, Bishin et al. (2016) identifies the circumstances most likely to trigger backlash, the groups that should lash back, and the form it should take. Specifically, challenges to the status quo that violate peoples' sense of importance, safety, symbolic values, or status are most likely to foster backlash. Simply put, backlash is driven by an individual's response to social change.

Arguably the clearest definition builds on this work by defining anti-gay backlash as “…a large, negative, and enduring shift in opinion against a policy or group that occurs in response to some event that threatens the status quo.” (Bishin et al. 2016). This definition, which we adopt in this paper, makes clear that backlash is not just any reaction, resentment, or counter-mobilization to a challenge to the status quo on gay rights. Instead, only those initiated by the masses rather than elites, and that are expressed through some form of lasting attitudinal or behavioral change qualify as backlash. Attempts to mobilize segments of the public or attempts to develop or implement strategies to oppose the group in other ways (e.g., litigation, legislation) by elites which are often referred to as counter-mobilization, or events that cause those who already hold strong views to speak out may be highly consequential but are not backlash.

MOB has become a conventional wisdom used to explain the politics of gay rights, and especially against court rulings that advance them (e.g., Stoutenborough et al. 2006). However, a robust debate around the extent to which shifting public attitudes lead gays and lesbians to face attitudinal, legal, legislative, or electoral backlash following favorable judicial rulings has emerged (e.g., D'Emilio 2006; Keck 2009). Surveying outcomes across a wide range of outcomes from court rulings, to legislation, to referendum results, Thomas Keck finds little evidence of a consistent backlash against advances in gay rights (2009). Similarly, studies that examine how opinion changes following major advances on gay rights also raise questions about the extent to which backlash occurs. (e.g., Barclay and Flores 2015; Bishin et al. 2016; Tankard and Paluck 2017).

2. Understanding Opposition to Gay Rights

How can we understand the conflicting explanations for opposition to gay rights? We argue that substantial ambiguity about the term backlash impedes the development and evaluation of alternative explanations of opposition to gay rights. One important consequence is that its conditions and expectations are seldom clearly articulated (e.g., Weaver 2007). Seldom are distinctions made between whether a reaction occurs among the masses or among political elites, whether the reaction takes the form of opinion change, campaign contributions, voting behavior, or something else, or whether it occurs only among specific groups or among the broader public (e.g., Weaver 2007: 237).

The term backlash is a catch-all used to reference a variety of different concepts, often interchangeably, even within subfields (e.g., Rosenberg 1993; Keck 2009). What some call backlash others refer to as a reaction, resentment, or counter-mobilization. While some law and society scholars emphasize that backlash is typically the immediate and adverse reaction by the public as a consequence of losing in court (e.g., Rosenberg 1993; Klarman 2013), others describe a more nuanced counter-mobilization in which elites play a significant role (McCann 1994; Adam 2017). Legal Mobilization Theory, for instance, holds that the law helps shape society by influencing social meaning, creating norms, and shifting the perception of rights and group legitimacy, in ways that may not be immediately evident in the days following a ruling (e.g., McCann 1994; Goldberg-Hiller 2004).

These complex and indirect effects make it difficult to evaluate imprecise accounts of backlash. If we are not clear about what should stimulate such a reaction, who should react, and what form that reaction should take, it is extraordinarily difficult to identify and evaluate alternative explanations for these anti-gay reactions. Vesla Weaver's observation that “…we are not sure where backlash begins and ends and what the unique characteristics of backlash are…” applies to the study of gay rights equally well (2007: 238). Given this, mixed findings about the extent to which mass backlash occurs are unsurprising.

3. Alternative Explanations for Opposition to Gay Rights

One consequence of this ambiguity is that the opposition to social change often attributed to mass backlash, might actually be driven by elites, either directly through elite action, or in combination with cue giving to mobilize the masses (e.g., Weaver 2007; Dorf and Tarrow 2014). Studies of the religious right, for example, document sophisticated media driven efforts to mobilize existing members and grow their membership (Berlet 1998; Frankl 1998; Lesage 1998; Fetner 2008). Among the most common explanations for political conflict are pluralist accounts of politics in which organized interests battle over policy (e.g., Truman 1951). In pluralist theory, the reaction against those challenging the status quo is led by elites who oppose the policy, rather than a bottom-up reaction in which the public changes their opinions to oppose the policy or group.

While scholars describe important differences between whether majoritarian interest group pluralism or biased pluralism led by economic elites are most influential, they nonetheless share a common theme (e.g., Schattschneider 1960; Gilens and Page 2014). The counter-mobilization hypothesis described in pluralist theories reflects a “top-down” process in which groups organize and mobilize in favor of policies they support and against those they oppose (e.g., Truman 1951).

With respect to gay rights, this process is consistent with Zein Murib's description of backlash as “…resistance and opposition, that is mobilized in response to increased visibility for minority groups, and motivates the formation of interest group coalitions” (2017: 18). While facing organized opposition makes achieving policy more difficult, it does not present the same impediments as does mass opinion change. In fact, to the extent that elite opposition leads to counter-mobilization, it may help to “…develop a shared agenda, and present a strong united front to defend against opponents” (Murib 2017: 19).

Elites may also be more strategic in the use of issues to achieve their goals than is widely appreciated (e.g., Haider Markel 2001; Dorf and Tarrow 2014). Vesla Weaver's (2007) incisive theory of frontlash holds that losers in a conflict “…propose new programs of action…” and “…manipulate the agenda in a way favorable to…” their interests on other related issues (236). With frontlash, strategic and proactive “elites aim to control the agenda and resist changes through the development of a new issue…” in order to overcome “…defeat of a longstanding political discourse or elite program” (238).

4. Whither Backlash on Gay Rights?

Given the stakes, concerns about anti-gay backlash, particularly in response to court rulings, are understandable. After all, court rulings are the primary mechanism by which gay rights have advanced. Judicial backlash reflects the idea that a favorable court ruling will serve to incite anti-gay opinion, lead to creation of a precedent that both defeats the policy in question and serves as a barrier in future cases, or cause elites to counter-mobilize (e.g., Ball 2006; D'Emilio 2006; Keck 2009). With respect to MOB, the fear is that a negative reaction may lead to further attempts to curtail gay rights in the legislature, via referendum, or through the election of politicians who oppose gay rights and promise to pass anti-gay rights legislation. While much of the conjecture about mass backlash focuses on courts' power to influence public opinion on gay rights, evidence supporting this conjecture is decidedly mixed.

A number of studies find evidence consistent with MOB. Research shows that the Supreme Court can move aggregate public opinion on salient issues (Johnson and Martin 1998). Others find that the Court can influence opinion but that rulings may influence different groups differently, causing opinions among groups to diverge (Franklin and Kosaki 1989). On gay rights issues, research suggests that Court decisions on Bowers v. Hardwick and Lawrence v. Texas both led the public to be less supportive of the proposition that “homosexual relations” “should be legal” (Stoutenborough, Haider-Markel, and Allen 2006; Persily et al. 2006). Public opinion following the Massachusetts State Supreme Court ruling in Goodridge v. Department of Public Health, a broad opinion that legalized gay marriage in Massachusetts exhibits similar characteristics (Persily et al. 2006). Others describe these cases as specific examples of a broad pattern of backlash (e.g., Rosenberg 1993).

Evidence of backlash is not limited to judicial rulings. Mass reaction to the election of gay or lesbian legislators is associated with the introduction of anti-gay legislation in state legislatures (Haider Markel 2010). Similarly, the defeat of the three judges in the 2010 Iowa judicial retention elections was driven by their ruling in Varnum v. Brien (2009) which legalized gay marriage (Harris 2019; Clopton and Peters 2013).

While some studies show support for MOB there are reasons to question whether enduring mass opinion change central to backlash actually occurs. One reason for this lies in the psychology of opinion formation and attitude change which finds that once formed, opinions are difficult to change (e.g., Zaller 1992). People tend to only receive information they see as relevant, and go to great lengths to avoid changing their opinions even when they accept such information (Taber and Lodge 2006). Absent a shift in the importance or relevance people ascribe to an issue, any opinion change is unlikely to be long-lasting (e.g., Converse 2006). Moreover, claims of MOB increasingly conflict with the persistent increases in public support for gay rights over time (Flores 2014; Garretson 2018).

Studies documenting how people accept information also help explain the absence of backlash among mass publics. Finding no evidence of backfire in a study of 52 issues, Wood and Porter (2016) speculate that subjects are willing to accept factual corrections, even when they contradict positions of those people or groups with whom they strongly identify. Moreover, on topics that aren't direct questions of fact, like whether one supports gay marriage, attitudes are primarily cued by elites. For example, attitudes toward transgender individuals changed as party elites polarized on the issue (Jones and Brewer 2018). Research on backfire effects suggests that mass backlash is much less likely than widely believed but also that individuals predisposed to support a position can be mobilized by elites on an issue.

Research on attitude change on gay rights produces similar findings. Conducting experimental tests designed to stimulate backlash, and employing historical observational studies, Bishin et al. (2016) finds no evidence of opinion change on gay rights. Barclay and Flores (2015) find that the little attitude change that occurs is among those opponents who became more accepting of gay marriage. Similarly, Kreitzer et al. (2014) finds no evidence of opinion backlash following the State Supreme Court's finding the marriage ban unconstitutional. Further building on these findings, Tankard and Paluck (2017) finds that the Supreme Court ruling in Obergefell did not change opinions about gay marriage but led to increased acceptance of social norms typically supportive of gay marriage.

Taken together, this research suggests that there are both theoretical and empirical reasons to question the existence of backlash. The changes in opinion that do occur often seem to be fleeting, or occur in different directions for different groups, but generally lack a consistent pattern. By raising questions about the extent to which backlash occurs, this research heightens the possibility that some other factor might better explain opposition to gay rights.

5. Explaining the Paradox of Backlash without Opinion Change

The rapidly growing literature on backlash presents an interesting puzzle. On the one hand, scholars and journalists frequently describe anti-gay backlash in the form of opinion change and political mobilization, numerous attempts to overturn policy, attempts to ban gay marriage, and to punish those who work to advance gay rights. On the other hand, scholars in a variety of contexts, studying a wide range of marginalized groups, employing a wide variety of methods, repeatedly find scant evidence of the large, negative, and enduring opinion change indicative of backlash.

How can we explain this paradox? Ambiguity and inconsistency in the way pundits and journalists use the term backlash provide only part of the answer. We argue that what appears to be anti-gay backlash is a product of elites espousing anti-gay views and policies pursuing their agenda. Rather than describing a widespread change in policy opinion or intensity, backlash is often used to describe a reaction by groups and individuals that oppose gay rights. The behavior of these elites is described by elite mobilization theory (EMT). The important implication is that much of what is described as backlash is instead a “top-down” reaction instigated by representatives and leaders of organizations seeking to prevent advances in gay rights and public acceptance of gays and lesbians as legitimate members of society. Here mass mobilization also occurs but it is initiated from above and generally targeted at those who are sympathetic to their cause. The resulting reactions to challenges to the status quo thus explain the paradox of “backlash without opinion change.”

6. Toward a Theory of Elite Mobilization

Our central argument is that attempts to make policy restricting gay rights, and to oppose policies supporting gay rights, are driven by organized interests rather than an organic response by the mass public to challenges to the status quo. These interests are run and represented by elites opposing gay rights who hold policy goals and instrumental goals (see Bishin 2009 for a general discussion). Their primary policy goal is to prevent gays and lesbians from achieving full inclusion and legitimacy in the polity. They oppose advances in gay rights as they arise, and look to find new areas to impose restrictions on gay rights where possible. These elites also pursue instrumental goals to obtain the resources needed to pursue their policy goals, primarily in terms of finances, popular support, and access to politicians (Fetner 2008). Consequently, elites opposing gay rights work to enhance their visibility and legitimacy to obtain resources for their organizations, and in turn enhance their ability to achieve their policy goals.

By anti-gay elites, we refer to the broad set of individuals who hold positions of power or influence in the polity (e.g., Eldersveld 1989). Elites occupying different positions hold different types of roles. Generally, we see three sets of elites who design and implement strategies, and mobilize opposition to gay rights. One group consists of the founders, leaders, organization heads, and major contributors dedicated to opposing gay rights nationally by funding and coordinating strategy across and within states and in some cases countries. These include billionaire funders like Phillip Anschutz, Richard DeVoss, and the Mayer Family Foundation, and leaders of organizations like James Dobson, founder of Alliance Defending Freedom, the Family Research Council, and Focus on the Family, as well as people like Anita Bryant, who founded Save Our Children, and Tony Perkins, the President of the Family Research Council (e.g., Bull and Gallagher 1996; Capehart 2016). In many cases the positions held overlap with leaders of religious groups, like Jerry Falwell and Pat Robertson who, along with others, founded the Moral Majority.

A second group of elites includes the heads of affiliated state and local organizations with ties to the national organization, the spokespeople for these organizations at all levels, as well as the political strategists and lobbyists who develop strategies and write legislation and referenda that opposes gay rights, and who work to mobilize still other elites. This tier includes people like David Lane, a born-again Christian who travels the country to organize and mobilize pastors to lead from the pulpit by building local level political coalitions (Eckholm 2011).

A third group includes local influencers, like religious leaders such as pastors who preach from the pulpit, local political candidates and their volunteers, and activist church members who work to mobilize the like-minded by educating people about the message, organizing them politically, and mobilizing them to get their family and friends to participate.

The most forceful opposition to gay rights emanates from the religious right (e.g., Burack 2008; Fejes 2008; Fetner 2008). While the modern movement known as the religious right initially developed out of a general concern about social change, especially increased societal permissiveness and laws banning school segregation, it politically crystallized in the campaign to repeal Miami Dade's nondiscrimination ordinance in 1978 (Fejes 2008). The political organization created in Miami was nationalized and exported to cities across the country to repeal local ordinances protecting gay rights (Fejes 2008). Building on well-developed evangelical Christian organizations, the religious character of this opposition fueled an agenda that opposed gay rights and feminism, and was staunchly pro-life (Fetner 2008). The evangelical community exploited these issues to increase their power and leveraged their well-organized and professional institutions to mobilize supporters through their connections to churches and the media empire, which had television and radio stations reaching millions of adherents (Burack 2008; Fetner 2008).

While these appeals often went unseen by the broader public, opposition to gay rights was among the most effective issues for mobilizing and raising money from evangelicals (Fetner 2008; FitzGerald 2017). Importantly, while the movement takes strong stands on several issues, beginning in the 1990's they shifted toward emphasizing opposition to gay rights in order to tap into popular opposition on an issue that was financially lucrative (FitzGerald 2017). Today, they remain heavily committed to opposing gay rights (Smith 2007). Jonathan Capehart's (2016) insightful social network analysis, for instance, demonstrates that financial backing to oppose gay rights disproportionately comes from a small number of wealthy donors through religiously affiliated socially conservative political organizations.

Examination of the broader struggle for gay rights depicts coordinated campaigns by elites to cultivate and mobilize those who either already oppose gay rights or those who are predisposed to do so (e.g., Bull and Gallagher 1996; Burack 2008; FitzGerald 2017). In much the same way that evangelicals came to support the Catholic Church's opposition to abortion, there is a long history of elite mobilization and coordination against gay rights by various religious advocacy groups (e.g., Lewis 2017). As just one example, elites mobilized to create the National Campaign to Protect Marriage in January of 1996, in which major conservative organizations worked to write, introduce, and lobby for passage of state level Defense of Marriage laws resulting in 41 states considering such legislation by the end of that year (Haider Markel 2001). Some of these advocacy organizations, like the Family Research Council are still active today and were among those that 24 years later worked to defeat the judges in Iowa (Paulson 2004).

Where possible, these elites organize campaigns, build broad coalitions, and coordinate closely with candidates and parties. In June of 2003, the most prominent leaders of the Christian right met and formed what would become known as “the Arlington Group” to address their recent policy losses (FitzGerald 2017). Their central goal was to refocus the movement around the issue of gay marriage, by passing the Federal Marriage Amendment and placing constitutional amendments banning gay marriage on the ballot in 13 states (FitzGerald 2017). The group, which would eventually grow to include 70 distinct religious advocacy groups, included Muslims, Catholics, Mormons, and African American groups that opposed gay marriage (FitzGerald 2017). Working in close coordination with the Bush-Cheney campaign, they used gay marriage to mobilize religious conservatives and especially evangelical voters by placing prominent evangelicals like Ralph Reed, former Executive Director of the Christian Coalition, directly in the campaign and holding weekly strategy calls with top campaign officials including Karl Rove (FitzGerald 2017).

Conservative religious organizations use a variety of tools to mobilize supporters. Initially the strength of the religious right was cultivated through targeted direct mail campaigns and radio and television shows to publicize their message and raise money (e.g., Bull and Gallagher 1996). With time, religious officials who had been reluctant to get involved in politics were encouraged to preach on core political issues (FitzGerald 2017). Groups identify potential supporters through petition drives, voter registration drives, by sharing information with other conservative Christian organizations and, in some controversial cases, by obtaining church directories (FitzGerald 2017). Lower level elites contact and mobilize potential supporters often at the urging of activist pastors and fellow parishioners (e.g., Schulte 2010a; Eckholm 2011). They also use grassroots mobilizing techniques like microtargeted phone calls and Web advertisements during political campaigns (Pettys 2011; FitzGerald 2017).

These organizations pursue their policy goals in three primary ways. First, they identify key issues and work to pass policy. They do this directly, by sponsoring or supporting policy referenda that limit gay rights, by writing state and federal laws, and by lobbying sympathetic federal, state, and local officials. They also pursue political objectives indirectly, as when they recommend or support political appointees charged with creating, deciding, or enforcing laws.

Second, anti-gay elites work to identify, recruit, and support allies, and to defeat opponents of their agenda. In particular, they work to ensure the election of like-minded officials and to defeat those who will oppose them (Lesage 1998). Since the late 1990's, for instance, as public opinion became increasingly supportive of gay rights, the religious right increasingly prioritized court appointments to help neutralize the laws written by state legislatures they saw becoming increasingly hostile to their positions (FitzGerald 2017).

Third, in order to achieve their instrumental goals, they seek to sway opinion, mobilize supporters, and build public opposition to gay rights. They constantly look for opportunities to make both their organizations and their positions relevant (Lesage 1998). One reason for this is that, as Fred Fejes (2008) notes, historically few Americans had been exposed to gays and lesbians as they were only publicly out in large numbers in a handful of cities. Absent their visible presence as a threat, anti-gay positions have to be cultivated and widely publicized to build support for their socially conservative mission (Frankl 1998). As a consequence, elites look for opportunities to generate positive coverage for their positions (Fetner 2008).

The history of the gay rights movement also helps explain when and why anti-gay groups mobilize against them. Anti-gay elites act any time they can advance their goals. While much of their activism is of relatively low salience they tend to be visible when they mobilize in response to gay rights advocates' challenges to the status quo. Beginning in the mid 1970's, laws extending and protecting gays and lesbians from employment discrimination passed in a small number of local governments. In response to the challenges to the status quo that nondiscrimination laws represented, elites opposing gay rights allied with conservative religious organizations, especially evangelicals, to mobilize against gay rights (Fejes 2008; Fetner 2008). Evangelical leaders discovered the issue could help them raise money and mobilize their supporters, and thereby increase their political power (e.g., Fetner 2008; FitzGerald 2017).

Today, restrictions on gay rights are widely contested. Anti-gay elites and their organizations constantly look for opportunities to make policy advances, either by promulgating policy (e.g., bathroom laws) or opposing attempts by pro-gay rights groups to advance policy (e.g., gay marriage). Importantly, they also look for opportunities to gain visibility, popular support, and enhance perceptions of their relevance and legitimacy. One way they do this is by exploiting current events to achieve these instrumental goals. When a city or state considers legislation advancing gay rights, for instance, these organizations can use that policy as an opportunity to remind supporters that an immediate threat to their values exists, to fundraise among their supporters, and to have them mobilize and reach out to the appropriate government officials.

7. A Strategy for Evaluating Elite Mobilization

Despite significant differences between elite mobilization and MOB, definitive tests between the two theories are hard to identify as evidence of the strategic choices that lead to the elite behavior at the core of EMT is difficult to obtain. Because identifying the universe of potential cases of anti-gay backlash is difficult, our approach is to identify a “difficult” case, one widely accepted to be a hallmark example of MOB and see whether EMT better explains the outcome. Specifically, we examine the 2010 Iowa Judicial Retention Elections that followed the Iowa Supreme Court's 2009 ruling that overturned the ban on gay marriage.

Perhaps more than any other recent event, the 2010 defeat of three Iowa jurists who ruled to legalize gay marriage is widely hailed as a classic case of anti-gay backlash. Scholars claim that the justices lost because of their ruling on gay marriage (Harris 2019). Moreover, the press described the election as one characterized by backlash. The Des Moines Register described the race as “backlash fueled” while the New York Times, Time magazine, and numerous others described the events in Iowa as examples of backlash, generally describing a context in which the public rose up to defeat judges who contravened its will.

On the other hand, news reports describing the campaign to oppose the judges raise questions about whether the events were consistent with MOB. The campaign neither came about as an immediate response to the ruling nor did it focus on issues beyond gay marriage and judicial activism. While marriage opponents immediately pressed for a constitutional amendment following the ruling, the issue was not publicly tied to the judicial retention election until August 6, 2010, well after the April 2009 Varnum ruling, when losing primary Republican gubernatorial candidate Bob Vander Plaats launched a campaign against the judges retention (Buller 2011: 8). Moreover, the television ads run during the campaign focused on the gay marriage ruling, and the judicial activism it represented (Brennan Center 2010). While both issues are consistent with what one might expect of a campaign designed to target religious conservatives, it is important to assess the competing theories more systematically.

Judicial retention elections provide an almost ideal setting for theory testing as those angered by a ruling or a series of rulings can relatively easily hold a judge accountable by opposing their retention. Importantly, the elections provide a relatively low salience context in which elite cues should be influential, and on which both masses and elites might be active on an issue that provides a threat sufficient to trigger opinion change. Further, we can test these implications for mass and elite behavior in two contexts: campaign fundraising and voter behavior.

8. Iowa's 2010 Judicial Retention Elections

In 2010, three Iowa State Supreme Court justices were voted out of office. Their defeat followed the 2009 state Supreme Court ruling to overturn the state's gay marriage ban. Not only did the Supreme Court create a new policy—gay marriage was not legal prior to the Court's ruling—but they did so on an issue on which only about 27 percent of the public agreed with them.³ The Supreme Court justices were the first in Iowa's 48-year history of retention elections not to be retained (Sulzberger 2010).

The campaign to unseat the judges centered on their role in overturning the state's marriage ban (Buller 2011). The justices refused to raise money or mount campaigns on their own behalf, though an outside group organized relatively late and campaigned on their behalf. Led by religious conservatives, retention opponents spent almost $1 million to defeat them, running hundreds of television ads in Iowa's largest media markets (Sample et al. 2010). Activists organized pastors against retention, encouraged them to emphasize the issue to their parishioners, arranged a campaign of over 100 pastors to speak against retention on three Sundays preceding the election, got 834 of them to sign a petition opposing the justices retention, and provided voter guides to churches (Eckholm 2011; Schulte 2010b). Web commercials also microtargeted those most likely to be sympathetic (Pettys 2011).

Of particular importance, these ads along with robocalls, and other materials circulated to religious conservatives emphasized the need for voters to turn over the ballot in order to vote against the judges.⁴ This instruction typifies an important cue consistent with the type described in EMT, as the judicial retention election race was listed on the back of the ballot. In contrast, retention supporters initially organized Iowans for Fair and Impartial Courts as a 501 (c) 3 which precluded direct election advocacy, (Desmoinedem 2010a). It wasn't until October 14th, less than 3 weeks before the election, that Fair Courts for Us began a radio and direct mail campaign advocating retention. Their ad also instructed retention supporters to turn over their ballots (Desmoinedem 2010b).

While it did not see the heaviest spending, the retention election was the most closely contested statewide race. In the Gubernatorial race Terry Branstad defeated unpopular incumbent Democrat Chet Culver by 9.7 points. Similarly, the senate election saw five-term incumbent Chuck Grassley defeat Democrat Roxanne Conlin by nearly 31 points.

9. Campaign Spending in Judicial Retention Elections: 2010–2014

While elites and masses participate directly in elections, they differ with respect to how they financially contribute. As a widespread “bottom up” reaction, MOB implies that the source of financial contributions should be relatively diffuse, reflecting broad-based opposition. In particular, we should see contributions from a large number of people. Further, reflecting the financial resources of the average citizen, we also expect the amount of these contributions to be small in size. Finally, as backlash results from policy that threatens (or changes) the status quo, contributions should be largest among those who most directly experience this threat. As the judges' ruling applied only to Iowa, the contributions should disproportionately come from individuals and organizations within Iowa.

Elite mobilization, conversely, is predicated on the idea that individuals in positions of power, use salient events to push for policy and to gain credibility and resources. As these individuals are well resourced but few in number, patterns of financial contributions should reflect an outsized role for elites in terms of the size, number, and source of the contributions. In particular, elite mobilization is characterized by relatively few contributions but in large amounts. Moreover, because these elites often represent national organizations that coordinate strategy, contributions are more likely to originate out of state, where the organizations are headquartered. By comparing contribution patterns across judicial retention elections during a period contemporaneous with the Iowa elections, we can examine whether the contribution patterns are more indicative of backlash or elite mobilization.

We collect campaign contribution data from all contested judicial retention elections that occurred between 2010 and 2014.⁵ While some groups announced challenges to judges but did not collect contributions or spend money, we define a race as contested only if more than $1000 was spent to oppose the judge. In 2010 in Florida, for instance, a Tea Party group announced its opposition to two justices standing for retention, but raised no money and the justices were retained with about 60 percent of the vote (May 2013).

Collecting contribution data for the retention races is challenging. We began by downloading data from the National Institute on Money in State Politics Follow the Money Web site where we identified the races and then searched for contributions to groups opposing the judges.⁶ In some cases data were ambiguous or no contributions were found.⁷ A summary of the races is seen below.

Table 1 summarizes the contribution patterns from 2010 to 2014. While nonpartisan retention elections are held in 16 states, relatively few (18 of 107) judges are opposed, and even among those targeted, the campaigns against them are not typically very well funded. In 2008, for instance, there were no well-funded races.⁸ Moreover, justices are seldom defeated, as only the three Iowa justices lost during this period (Kreitzer et al. 2014).

Table 1. List of Judicial Retention Races with Funded Opposition in 2010–2014

Notes:

Races meeting the criteria for backlash are listed in italics.

In retention elections, individuals that oppose retention typically contribute to a group that coordinates strategy, collects money and spends on its own. In practice, this simplifies the data collection process as we can look for contributions to these independent groups that oppose retention. An important complication is that in several states one or more groups coordinated a campaign against more than one judge. How do we compare, for example, the 27 contributions totaling $636,210 to defeat Thomas Kilbride in 2010 with the six contributions totaling $990,851 given jointly to defeat justices Ternus, Streit and Baker? In order to conduct conservative tests that should be more likely to find backlash than mobilization, we assume that the total money raised was divided equally to oppose each candidate.⁹

In order to test our hypotheses, we collect data on the number, amount, and geographic origin for all of the contributions given to the opposition group(s). We then calculate the per-candidate contribution average for the 13 races in the nonbacklash group and compare those results with the results from Iowa. The results are seen in Figure 1.

Figure 1. Average Number of Donors, Percentage of Out-of-State Donations, and Amount of Donation in Iowa v. Other States, 2010–2014.

The difference in donation patterns, seen in Figure 1, is striking. The left panel shows that Iowa saw far less money come from within state—only about 18 percent. In race outside of Iowa, however, almost 62 percent of the money raised came from in-state donations.

The middle panel shows that while the average number of donors is small across states, it is dramatically smaller in Iowa than in the other states. The right panel depicts the average amount contributed. Perhaps most striking is that the average contribution is large across all states, but at almost $100,000, as compared with $42,557 it is much larger in Iowa.

The pattern of contributions in Iowa appears more heavily elite driven than in other states. One concern is that contribution patterns in low salience nonpartisan judicial retention races in which the justices themselves do not mount campaigns do not accurately reflect the public's preferences. Alternatively, perhaps these results are less about elite mobilization than Iowans being less willing to financially contribute than are voters from other states? If true, a similar pattern should be evident across other major statewide races in Iowa.

Perhaps the ideal race in which to test for this possibility is the 2010 Iowa Senate race in which popular incumbent Chuck Grassley trounced Democrat Roxanne Conlin, as it provides the most difficult test for the EMT. The national implications accruing to winning a Senate seat along with the uncompetitiveness of the race make the contest seem especially likely to see money come from out of state, and from large contributors. If the contribution patterns seen in the judicial retention election are unique to Iowa elections, we should see few differences between the judicial retention and senate elections. To examine this hypothesis we obtain the number, amounts, and source of contributions to Roxanne Conlin, as she is the challenger to the incumbent and therefore most similar to the campaign against the justices' retention. The results are depicted in Figure 2.

Figure 2. Contributions to Democratic Senate Challenger Roxanne Conlin and Opposition to Judicial Retention in 2010.

The results in Figure 2 depict large differences between contributions in the senate and judicial retention races. Despite conditions that seem likely to portend relatively low levels of participation, the percentage of money contributed from out of state (40.4 percent) was less than half of that seen in the judicial retention election, the number of donors was substantially larger (1,623 versus 6), and at $2,251.35, the average contribution was less than 2 percent of the average contribution in the judicial race. The fact that Roxanne Conlin, a Senate challenger with little popular support shows much broader support within Iowa than did the victorious campaign to oppose retention is inconsistent with the claim that Iowans are less likely to participate financially. This finding is reinforced by examining the governor's race. The contribution data for Terry Branstad, the challenger to the incumbent Governor is seen in Figure 3.

Figure 3. Contributions to Terry Branstad and Opposition to Judicial Retention in 2010.

In Figure 3, the public's desire to remove the unpopular incumbent governor is reflected by the large number of donors (13,995), the smaller average donation ($630.46), and the fact that the vast majority of the money came from within the state (75.5 percent). If mass backlash in response to the judicial ruling were occurring we would expect a pattern of contributions that looks substantially more like those received by the Branstad campaign. These results suggest that the absence of significant citizen participation in funding the judicial retention race is not likely due to Iowans' unwillingness to participate financially in politics (National Institute for Money in State Politics, 2018).

10. Voter Participation

The advertising messages emphasizing the importance of turning the ballot over provides a powerful opportunity to test an implication of elite mobilization. To do so, we examine voter roll-off, the rate at which people choose not to vote in the retention races which were listed on the back of the ballot. ¹⁰ The difference in participation between the judicial elections and the up-ballot races is striking. For comparison, turnout in the gubernatorial race increased by about 110,355 votes between 2002 and 2010. In 2002, however, while there were 1,025,802 votes cast in the governor's race, only 620,642 votes cast in Mark Streit's judicial retention race, a gap of 405,160 missing votes. In 2010, the gap between the gubernatorial and retention race for Mark Streit dropped to 149,115, indicating substantially more interest in the judicial races (see, Iowa Secretary of State, 2003; 2010).

To assess whether patterns of voter participation are more consistent with MOB or EMT we develop and test two hypotheses. Recall that the opinion change that underlies MOB reflects a broad-based reaction to the issue or group. This reaction should be reflected by increased interest in and opposition to the justices by a broad range of citizens. Hence, our insight is that if MOB occurs and people are motivated to participate at least partly because of the justices' ruling then increased voter turnout should be negatively associated with changes in both roll-off and support for retaining each justice between 2002 and 2010.

The availability of data from the 2002 judicial retention election provides an opportunity to examine whether the gay marriage ruling led to changes in county level support for the justices between 2002 and 2010. As two of the three justices stood for retention in 2002, results from that election allow for an assessment of their popularity before the ruling.¹¹ Importantly, examining county change over time adds tremendous leverage to our analyses by controlling for all unobserved time invariant explanations (e.g., race, income, education, gender, partisanship). For these reasons, our dependent variables in these analyses are the change in roll-off and the vote share in favor of retaining Marsha Ternus and Michael Streit.

The primary targets of mobilization are religious conservatives. To account for religious conservatism and their opposition to gay rights, we employ county evangelical church membership as a proxy for those with strong anti-marriage attitudes (e.g., Burack 2008). Specifically, we employ county-level estimates from the Association of Religion Data Archives from 2000 and 2010 (e.g., Association of Statisticians of American Religious Bodies 2002; 2012).

First, we assess enthusiasm by examining voter roll-off. Our expectations are as follows: MOB implies that ballot roll-off should decrease as voter turnout increases. Alternatively, if elite mobilization occurs roll-off should decrease as the size of a county's evangelical population increases. To test these hypotheses, we regress the difference in roll-off percentage by county between 2010 and 2002 on the change in voter turnout between 2002 and 2010, and evangelical percentage (in 2010).¹² Next, to examine whether the effect of turnout or evangelical population significantly changed between the two elections, as we would expect if the court case led to mobilization or backlash, we pool the data from 2002 and 2010 and examine roll-off in each county-year as the dependent variable. We create interactions between election year (2010) and turnout and election year and evangelical population to assess these hypotheses. The results of this regression are shown in Table 2.

Table 2. Regression of Difference in Voter Roll-Off on Group Membership

Notes:

^*p < .10; ^**p < .05; ^***p < .001.

The results presented in Table 2 depict a large and significant role for Evangelical % across all models. Counties with large evangelical populations demonstrated much lower roll-off rates in the 2010 judicial retention election, a result consistent with EMT. In contrast, we find no evidence that turnout is negatively associated with roll-off. Importantly, the interaction terms for the Evangelical and the 2010 election are large, negative, and statistically significant, while the interactions between Turnout and the 2010 election are not significant.

Next, we examine the change in support for retention between 2002 and 2010 by pooling county-level election returns from 2002 and 2010. As with the roll-off analysis, we create interactions between the 2010 variable and Turnout to assess the effect of backlash, and create an interaction between the 2010 variable and evangelical population variable to test elite mobilization.

While change in voter turnout and the evangelical population allow us to assess the competing theories, other factors not controlled for by our design might influence support for the judges as well. One possibility is that increased opposition to the judges reflects increased conservative dissatisfaction with President Obama. To account for this possibility, we include county-level Obama vote share in 2008.¹³ As the vote share data is normally distributed, we employ ordinary least squares (OLS) regression to estimate the results, which are seen in Table 3 below.

Table 3. Comparison of Influences on Retention Votes in 2002 Versus 2010

Notes:

^*p < .10; ^**p < .05; ^***p < .001.

Note turnout variables are for 2010 and 2002 as appropriate; Evangelical% is for 2010 and 2000.

We begin by examining the change in Streit and Ternus support between 2002 and 2010. The significant and negative intercept reflects the large decrease in support for the justices between 2002 and 2010. Perhaps most striking, we see that Evangelical is associated with decreased (change in) support for retaining the justices. In contrast, we do not observe any significant result for Turnout. The models do a surprisingly good job of explaining change in support given the noisiness of the data.

The results presented in the rightmost two columns are estimated on data pooled across 2002 and 2012 and allow us to assess the influence of evangelical percentage and turnout above and beyond the effect that dissatisfaction with Obama might have. In this case, the key variables are the interaction terms that identify the effect of evangelical percentage in 2010 as compared with 2002, and voter turnout in 2010 as compared with 2002. These results are enlightening. It appears that opposition to Obama was highly (negatively) associated with support for the judges—increased support for Obama in 2008 is a large and statistically significant predictor of support for retention. Equally striking, the interaction terms for the evangelical variable are also very large, negative, and statistically significant, just as EMT predicts. Conversely, the interactions between turnout and the 2010 election are not significant, contrary to MOB.

11. Discussion

Overall, the campaign finance and voter participation results strongly support elite mobilization and are inconsistent with MOB. Our use of county level data, however, may diminish the confidence we place in the inferences we draw about individual behavior. One concern is that omitted variables that change at different rates across counties might explain the results.¹⁴

We further corroborate our findings by employing data from a panel study conducted before and after the court ruling. While these data do not allow us to assess the effects of elite cues during the campaign, they do permit assessment of MOB (Kreitzer et al. 2014). Specifically, contrary to MOB, support for gay marriage increased from 26.8 to 28.7 percent in the week following the election, and by March 2011, reached about 36 percent (Kreitzer et al. 2014). Moreover, the data also allow for assessing the key assumption of elite mobilization that elites seek to mobilize the likeminded to elicit opposition to gay rights. If true, then evangelicals should indicate that gay marriage is a particularly important issue. Indeed, results from the second wave of the study show that 55.2 percent of evangelicals but only 24.55 percent of nonevangelicals, saw gay marriage as “the most” important issue, or as “very important.” Moreover, neither these differences nor the gay marriage opinion data are a product of evangelicals being more informed as 63 percent of evangelicals, and 63.4 percent of nonevangelicals, reported that they had heard “a lot” about the ruling in the last week.

12. Conclusion

EMT provides an alternative to backlash based accounts of why advances in gay rights are frequently met with opposition and antipathy. We argue that opposition to gay rights is often the result of strategic action taken by anti-gay interest groups and the elites that represent them. Elites holding positions opposing gay rights take visible and forceful public positions in order to activate and mobilize their like-minded supporters. Despite their important implications for democratic citizenship, scholars, journalists, and pundits seldom distinguish between the massive changes in opinion that characterize backlash and the elite driven process we describe here.

Using the case of the 2010 Iowa judicial retention elections, a contest widely viewed as a classic case of MOB, we develop and test two key implications of EMT. To examine elite activism, we examine patterns of campaign contributions across all competitive state judicial retention elections between 2010 and 2014. Donation patterns are inconsistent with the predictions of MOB. The pattern of contributions in Iowa was small in number, large in amount, and disproportionately originated from out of state. To investigate who was mobilized, we examined voter roll-off and vote choice. We find decreased roll-off only among evangelical counties in 2010, a shift that represented a sharp change from 2002. These changes cannot be attributed to changes in ideological or partisan factors as the results hold even after controlling for support of President Obama. Moreover, they are consistent with polling data which depict no negative shift in attitudes toward gay marriage following the ruling despite high levels of exposure to the news about it.

These findings have important implications for academics and activists. Claims of backlash against a wide range of marginalized groups are common in American politics. EMT may, consequently, help explain recent research contradicting conventional wisdoms about mass-backlash with respect to the black civil rights movement (Rubin and Elinson 2017), and anti-immigrant sentiment (Flores 2017). In at least some cases, opposition to minority rights may be better described as a process initiated by elites rather than a “bottom up” mass movement.

In light of this evidence, gay rights advocacy groups that might otherwise hesitate to vigorously pursue equality, as the Human Rights Campaign did with respect to marriage equality, should not delay. The opposition mobilized against gays and lesbians and those pursuing fuller incorporation in the polity appears to come from those who already oppose gay rights rather than new recruits. Specifically, we find no reason to fear that the public will turn to tyrannize gays and lesbians in response to their attempts to achieve equality.