INTRODUCTION
How does a candidate’s religion affect voting behavior? Religion is an important social identity in many countries, and emphasizing one’s faith during the campaign is likely to win votes from fellow believers. In societies with politically salient interdenominational cleavages, communicating one’s religion to voters might also lead to a loss of support among members of the out-group. Yet a candidate’s religion can also influence voting behavior in a more indirect fashion, via its implications for nonreligious divides. To the extent that religious affiliation is linked to other politically salient stereotypes, it may influence voting behavior even in political systems where divisions among the faithful are not a major fault line.
This article examines the case of Chile, where the division between the two leading religious identities, Catholicism and Evangelicalism, has had few repercussions in electoral politics, but where attitudes toward the military government of Augusto Pinochet (1973–1990) constitute a major political cleavage (Bonilla, Carlin, Love and Silva Méndez, Reference Bonilla, Carlin, Love and Méndez2011). During Pinochet’s rule, a well-organized and highly visible faction of the evangelical community expressed enthusiastic support for his government. For the non-evangelical majority of voters in Chile, who are unlikely to be directly affected by a candidate’s Evangelicalism, the stereotype that “evangelical candidates are pinochetista” could plausibly affect voting decisions. For their part, evangelical voters should favor fellow believers, regardless of their community’s historical associations with Pinochet.
The analysis draws on a survey experiment, conducted prior to Chile’s 2013 congressional election, that identifies a hypothetical candidate as evangelical, primes evangelicals’ historical support for Pinochet, and examines the effect of these separate factors on vote intention. Evangelical voters are more likely to support a fellow believer but are unaffected by the Pinochet cue. Non-evangelical voters have no significant reaction to a candidate’s Evangelicalism on its own. However, among right-wing, non-evangelical respondents, the Pinochet prime increases vote intention for an evangelical candidate. The treatment has no significant effect for non-evangelicals on the center and left. These results suggest that pinochetismo remains salient for a new generation of right-wing voters, even as its relevance fades for Chileans more broadly.
THEORY AND HYPOTHESES
A candidate’s group membership potentially influences voting based on social identity as well as stereotypes. If the candidate belongs to social group whose members share a strong sense of attachment to one another, he or she is likely to win votes from other members of the in-group (Boas, Reference Boas2014; McDermott, Reference McDermott2009b). In the context of intergroup political competition, a candidate’s group membership might also alienate voters who do not share this social identity, causing a loss of support among out-group members (Boas, Reference Boas2014; Brewer, Reference Brewer1979; Nicholson, Reference Nicholson2012; Samuels and Zucco, Reference Samuels and Zucco2014). Yet candidates whose group membership is not salient to existing political competition are less likely to lose votes among the out-group solely on the basis of this identity. Rather, effects among out-group voters will tend to be indirect, driven by stereotypes. Stereotyping allows voters to take assumptions about particular types of candidates—e.g., Jews are more liberal—and substitute them for knowledge of the candidate’s ideological tendencies, issue positions, or personal traits (Berinksy and Mendelberg, Reference Berinksy and Mendelberg2005; Boas, Reference Boas2014; Campbell, Green and Layman, Reference Campbell, Green and Layman2011; McDermott, Reference McDermott2007, Reference McDermott2009a; Sigelman, Sigelman, Walkosz and Nitz, Reference Sigelman, Sigelman, Walkosz and Nitz1995).
In Chile, evangelical Christians constitute an important identity group, a factor that is likely to win votes from fellow believers. Evangelicals make up 16.6% of Chile’s population, the second largest share in South America. I use the term “evangelical” as it is generally used in Latin America—to denote all Protestants, including mainline/historical denominations—though a large majority of Chile’s Protestants are Pentecostals and would easily qualify as evangelical in the traditional English-language sense. Chilean evangelicals of all stripes often see themselves as second-class citizens, legally and socially disadvantaged in comparison to the Catholic Church. Shared interests and grievances have helped to generate a common evangelical identity despite denominational divides. In the 2012 AmericasBarometer survey, Chilean evangelicals had a significantly higher level of confidence in “the evangelical church”—which does not exist as a unified entity—than Catholics had in the Catholic Church (5.7 versus 5.1 on a 1–7 scale; p < 0.001). Given the importance of Evangelicalism as a social identity in Chile, I hypothesize that evangelical voters will be more likely to vote for a candidate who is identified as evangelical, as has been demonstrated elsewhere in the region (Boas, Reference Boas2014; Boas and Smith, Reference Boas, Smith, Carlin, Singer and Zechmeister2015).
A candidate’s Evangelicalism is likely to win votes among fellow believers but not, on its own, to alienate non-evangelicals. Unlike in Brazil—with South America’s largest evangelical population, and with a powerful evangelical caucus that holds 13% of seats in Congress—Chile’s evangelicals have little presence in electoral politics. Since redemocratization in 1990, there has been only one evangelical, on average, in the 158-seat legislature. Few run for office, and those that do meet with limited success. Because of the limited political competition between Chilean evangelicals and those of other religious beliefs, I hypothesize that a candidate’s Evangelicalism will not directly influence the voting behavior of non-evangelicals.
Yet an evangelical candidate’s religion might influence the voting behavior of non-evangelicals indirectly, if it is connected to stereotypes that do have a strong effect on voting decisions in Chile. One such stereotype concerns pinochetismo—support for or sympathy with the military regime of Augusto Pinochet (1973–1990). Public assessments of Chile’s military regime remained profoundly divided through the 2000s, and both personal and family attitudes toward the regime correlate quite strongly with party preference and vote intention for the left- and right-wing coalitions (Huneeus, Reference Huneeus2003; Huneeus and Maldonado, Reference Huneeus and Maldonado2003; López and Morales, Reference López and Morales2005; Tironi, Agüero and Valenzuela, Reference Tironi, Agüero and Valenzuela2001). If pinochetista sentiment at the individual level affects how Chileans vote, perceptions of a candidate as pinochetista are likely to do so as well.
Given the prominent embrace of the Pinochet regime by a group of evangelical pastors, it is likely that, during the Pinochet years, Chileans stereotyped evangelicals as pinochetista. In 1975, the leaders of a variety of evangelical churches established an explicitly pro-Pinochet organization, the Council of Pastors. Though it represented only a minority of Chilean evangelicals, the Council sought a hegemonic position, and state authorities routinely portrayed it as the official voice of the evangelical community (Lagos Schuffeneger, Reference Lagos Schuffeneger1988). Starting that year, the Council sponsored an annual inter-denominational service, the Evangelical Te Deum, that was regularly attended by Pinochet and other government officials. Noting the heavy media coverage of Pinochet’s appearances at this event, a 1986 magazine article concluded that “public opinion maintains the idea that evangelicals are—in general—addicted to the regime” (Maldonado, Reference Maldonado2012, 66).
A stereotype that predominated in the 1980s would not necessarily persist in the electorate three decades later. Yet Chile’s 2013 election was an unusually propitious time for memories of the Pinochet regime, and evangelicals’ support for it, to influence voting behavior. During the campaign season, Chile observed both the 40th anniversary of the coup that overthrew Allende and the 25th anniversary of the plebiscite in which Pinochet was defeated. At the Evangelical Te Deum two months before the election, Bishop Eduardo Durán Castro, president of the National Council of Evangelical Churches of Chile, publicly begged forgiveness on behalf of all evangelicals “for not doing enough when our brothers were deprived of their rights or harassed for thinking differently” (La Nación, 2013). The attendance of Chile’s president and most presidential candidates at this event ensured that Durán Castro’s statement got a fair amount of attention in the news cycle.
Given the strong and consistent relationship between support for Pinochet and present-day behavior and attitudes, whether an evangelical candidate is perceived as pinochetista should affect non-evangelicals’ voting decisions. The direction of the effect should vary based on ideological tendency. Pinochetista attitudes in the electorate correlate strongly with support for the right-wing versus center-left electoral coalitions. Hence, I hypothesize that priming evangelicals’ historical ties to Pinochet will make right-wing, non-evangelical voters more likely to support an evangelical candidate and will make center-left, non-evangelical voters less likely to do so.
Finally, I hypothesize that the Pinochet prime will not affect evangelicals’ support for an evangelical candidate, regardless of voter ideology. Evangelicals have less need to rely on stereotypes when evaluating a candidate from their own community. They are more likely to know that staunch supporters of Pinochet were a minority faction of Chile’s evangelicals and to discount the information provided by the treatment. Moreover, pinochetista associations should not affect interest-driven voting for co-religionists. Many of the key policy issues of concern to evangelicals, such as rights and benefits on par with those of the Catholic Church, are likely to be pursued by evangelical politicians of all stripes.
RESEARCH DESIGN
To test the effect of candidate Evangelicalism and pinochetista stereotypes on voting behavior in Chile, I conducted an online survey experiment involving 1,035 Chileans during two and a half weeks prior to the November 17, 2013 general election. To recruit respondents, I used advertisements on Facebook, following an approach that has been used previously in Brazil (Boas, Reference Boas2014; Samuels and Zucco, Reference Samuels and Zucco2014). The Appendix offers additional details on the recruitment process.
As is often the case with Facebook-based recruitment, the online sample was substantially younger and better educated than Chile’s population, though it was similar in most other respects. Percent evangelical (14.8) was comparable to that of the 2012 census (16.6), though there were fewer Catholics and a much larger share with no religion. On gender, party identification, ideology, and region of the country, most categories differ from census or national probability sample figures by only a few percentage points. Full tables are in the Appendix.
The survey experiment consisted of a 2 × 2 factorial design in which a first treatment primed evangelicals’ historical association with Pinochet and a second, orthogonal treatment identified a candidate as evangelical. The first factor involved the following question, in which the control group read only the text in Roman type and the treatment group also read the text in italics:
Evangelicals have adopted a variety of political positions in Chile. For example, in the 1990s, the National Christian Alliance pushed for the Religion Law, which guarantees to all religious groups the same rights as the Catholic Church. In addition, in the 1970s and 1980s, the Council of Pastors lent their support to the government of Gen. Augusto Pinochet.
To what extent to you agree or disagree with the political positions of evangelicals in Chile?
The outcome was measured via the next question, which also administered the second treatment:
Suppose that Alejandro Pérez is running for Congress for the [Alliance for Chile/New Majority]. He is 35 years old, a businessman, and a member of an evangelical church.
How likely are you to vote for a person like this?
Vote intention was scored on a one-to-seven scale, with endpoints labeled “not at all” and “very” and intermediate points unlabeled. The fictional candidate’s electoral coalition was randomized between the major left- and right-wing alliances.
Several features of the survey are discussed further in the Appendix. Random assignment ensured balance between treatment and control groups with respect to demographic and political variables. The survey included several screeners, or attention checks (Berinsky, Margolis, and Sances, Reference Berinsky, Margolis and Sances2014); treatment effects rarely differ significantly based on screener passage. Finally, some survey respondents in districts with evangelical candidates for Congress were assigned to a “real candidate” version of the vote intention question; to maintain comparability, I exclude these observations.
RESULTS
Treatment effects on vote intention provide support for nearly all of the hypotheses advanced above. Results are summarized graphically in Figure 1, and numerically in the Appendix (Tables 11–13). For each religious/ideological group, the figure plots the effect of the Pinochet prime for respondents who are told that the candidate is evangelical (Pinochet|Evangelical) and the effect of candidate Evangelicalism for respondents who are not given the Pinochet prime (Evangelical|¬Pinochet).
Effect of Pinochet Stereotypes and Candidate Evangelicalism on Vote Intention.
For evangelical respondents, there is evidence of in-group identity voting. Mentioning that a candidate belongs to an evangelical church increases vote intention by 1.7 points on the 7-point scale, or 24 percentage points. Moreover, evangelicals’ support for an evangelical candidate is unaffected by the Pinochet prime. As shown in the Appendix, this null effect does not vary with ideology.Footnote 1
Also as expected, there is no evidence of out-group identity voting among non-evangelicals. Simply describing a candidate as evangelical does not significantly change vote intention for either right-wing or center-left respondents.
Finally, there is evidence that pinochetista stereotypes affect non-evangelicals’ support for an evangelical candidate. For non-evangelical voters on the right (ideological self-identification scores of 7–10), priming evangelicals’ Pinochet connection boosts vote intention for an evangelical candidate by 1.19 points on the 7-point scale, or 17 percentage points. However, null results are obtained for non-evangelical respondents on the center-left, which runs contrary to expectations.Footnote 2
CONCLUSION
A candidate’s status as an evangelical Christian, a significant religious minority in Chile, can affect voting behavior both directly, through in-group identity voting, and indirectly, through stereotypes that influence the out-group. Consistent with social identity theory and with findings from elsewhere in the region (Boas, Reference Boas2014; Boas and Smith, Reference Boas, Smith, Carlin, Singer and Zechmeister2015), evangelical voters respond positively to a hypothetical candidate’s Evangelicalism. Given evangelicals’ limited political ambitions in Chile, there are no out-group identity voting effects; non-evangelicals are no less likely to vote for an evangelical candidate when informed of his religion. However, priming stereotypes that link Evangelicalism to a politically salient cleavage, pinochetismo, does influence non-evangelicals’ intention to vote for an evangelical candidate. Right-wing non-evangelicals respond positively to the Pinochet prime, though the majority on the center-left are not significantly affected.
The ideologically asymmetrical effects of the Pinochet prime offer insight into the evolution of this traditionally important cleavage. Like Evangelicalism, pinochetismo is an important social identity in Chile. Right-wing voters’ reaction to the Pinochet prime can be interpreted as in-group identity voting, parallel to evangelicals’ preference for a fellow believer. Quietly defending Pinochet’s legacy is still a marker of identity for the hard-core Right, which seeks to keep his memory alive, even in the younger generations that were primarily sampled by the online survey.Footnote 3 Yet pinochetismo also poses much less of a threat to the rest of the Chilean political spectrum than it did 1–2 decades ago, which could explain the lack of out-group voting effects on the center-left. Many former supporters distanced themselves from the dictator after evidence of his involvement in embezzlement and tax evasion, and the first right-wing president after redemocratization, Sebastián Piñera, had opposed Pinochet during his rule. As personal memories of the dictator fade into the past, and unrepentant pinochetistas shrink in number and political power, this cleavage may be losing salience for most Chilean youth, while still influencing behavior and attitudes on the right.
Finally, the results of this analysis underscore that the weight of the past cannot account for Chilean evangelicals’ limited presence in electoral politics when compared to other countries like Brazil. If anything, evangelicals’ historical ties to Pinochet would seem to help present-day candidates win right-wing votes beyond their natural base of fellow believers. Hence, this article underscores the likely importance of other, non-behavioral factors, such as motivations to enter the electoral arena in the first place, as explanations for Chilean evangelicals’ limited political success.
SUPPLEMENTARY MATERIALS
To view supplementary material for this article, please visit http://dx.doi.org/10.1017/XPS.2015.17.
