Disgust sensitivity predicts political ideology and policy attitudes in the Netherlands

Individual differences in disgust sensitivity have been linked to social attitudes and ideology, but the generalizability of this effect and the nature of the political issues implicated remain unclear. In two studies using large Dutch samples, we find that disgust sensitivity predicts political attitudes for issues in several domains related to physical/spiritual purity and pathogen risk. Sensitivity to disgust was significantly associated with attitudes for a general ‘physical and spiritual purity’ factor, as well as specific issue factors regarding sex and sexual minorities, immigration, and foreign outgroups. Additionally, disgust sensitivity was associated with greater likelihood of voting for the socially conservative “Freedom Party” (Partij Voor de Vrijheid). These results suggest that the tendency to experience disgust influences a specific subset of social and political attitudes across cultures. Copyright © 2014 John Wiley & Sons, Ltd. The emotion of disgust likely evolved to discourage us from ingesting noxious or dangerous substances (Rozin, Haidt, & McCauley, 2008) and from coming into contact with dangerous pathogens (Oaten, Stevenson, & Case, 2009). However, it also seems to play an important role in our political, social, and moral beliefs (Bloom, 2004; Nussbaum, 2001). People who are more readily disgusted tend to score higher on broad measures of right-wing political ideology such as right-wing authoritarianism (Altemeyer, 1981) and social dominance orientation (Sidanius & Pratto, 2001) (Hodson & Costello, 2007; Terrizzi, Shook, & Ventis, 2010). Disgust sensitivity (DS) is also predictive of negative attitudes toward groups seen as threatening traditional social values—most consistently, gays and lesbians (Inbar, Pizarro, & Bloom, 2009; Inbar, Pizarro, Knobe, & Bloom, 2009; Olatunji, Haidt, McKay, & David, 2008; Terrizzi et al., 2010), but also immigrants and foreigners (Hodson & Costello, 2007). In line with this pattern of attitudes, people who are easily disgusted are more likely to describe themselves as politically conservative and especially as socially conservative. Higher chronic sensitivity to disgust has been shown to correlate with more conservative self-reported political ideology (i.e., selfplacement on a left–right continuum) in multiple samples from multiple labs (Inbar, Pizarro, & Bloom, 2009; Inbar, Pizarro, Iyer, & Haidt, 2012; Terrizzi et al., 2010; but see Tybur, Merriman, Hooper, McDonald, & Navarrete, 2010, for a failure to replicate). Meta-analytically combining the data from the relevant studies shows the relationship between DS *Correspondence to: Corinne Brenner, Harmony Institute, 54 West 21 Street, N E-mail: [email protected] Disgust sensitivity was also associated with approval of lowering income tax rates, b Key message: In the Netherlands, political attitudes in specific domains are related Copyright © 2014 John Wiley & Sons, Ltd. and ideology to be moderately sized but highly statistically reliable (Terrizzi, Shook, & McDaniel, 2013). However, much less is known about how DS relates to attitudes on specific political issues or policies. Inbar, Pizarro, and Bloom (2009) asked participants about their views on 10 political issues related to economic, social, and foreign policy, and found that DS was consistently associated with more conservative views only on gay marriage and abortion. Smith, Oxley, Hibbing, Alford, and Hibbing (2011) asked participants if they agreed or disagreed with statements about 16 political issues and found that greater self-reported DS predicted agreement with the more conservative positions on gay marriage, premarital sex, and abortion (whereas physiological reactivity to disgusting photographs predicted more conservative attitudes only for gay marriage and premarital sex). Although these studies paint a remarkably consistent picture, they also have significant limitations. Participants were asked about a relatively limited number of issues, and both studies used sample sizes that afforded limited power to detect smaller effects (Inbar et al. N = 84; Smith et al. N = 46; note that even with N= 100, one’s power to detect a correlation of .2 is .52, well below the conventional minimum of .80). Finally, both used American samples (Inbar et al. sampled university undergraduates and Smith et al. older adults). In fact, of all the studies showing a link between DS and political ideology, all but one sampled only North Americans (the exception is Inbar, Pizarro, Iyer et al., 2012). Especially when ew York, NY 10010, USA. ut the authors stated that this relationship failed to replicate in subsequent studies. to an individual’s sensitivity to disgust. Received 8 June 2014, Accepted 3 September 2014 Corinne J. Brenner and Yoel Inbar investigating how disgust relates to specific political issues, this obviously raises the possibility that any relationships found might be unique to the American political system, which differs substantially even from that of many other Western democracies (Lijphart, 1994). The two studies reported here aimed to address some of these limitations. First, we asked participants about more political issues (39 in Study 1; 38 in Study 2) so that we could assess whether DS consistently predicts attitudes in some domains but not others. In order to make our estimates of these relationships more precise (and especially to be able to make credible claims about null or small relationships between disgust and attitudes in some domains), we recruited larger samples than have been previously employed (N = 237 in Study 1; N= 304 in Study 2; with sample sizes and an alpha level of .05, the resulting power to detect a correlation of .2 is above .92) and actively recruited participants from a range of ages and locations, rather than recruiting only university students. Finally, we ran both studies in the Netherlands, which allowed us to test whether the relationships observed in previous US-only research would generalize to a different culture with different political and social concerns. The Netherlands and the USA are similar in many ways, especially by the standards of cross-cultural research comparing, say, East Asian and Western cultures. Both the Netherlands and the USA are advanced Western democracies with (to varying degrees) free-market economies. Citizens of both would be considered individualistic on the individualistic–collectivistic dimension often employed in cross-cultural psychology (Hofstede, Hofstede, & Minkov, 1997; Oyserman, Coon, & Kemmelmeier, 2002) and value self-expression to a similar degree (Inglehart & Welzel, 2010). Despite these similarities, however, there are significant political and social differences between the two countries. Two major parties dominate the USA’s political landscape, whereas the Dutch electoral system comprises a diverse roster of parties. Dutch parliamentary seats are allotted in proportion to the number of votes a party captured during the election, and a majority coalition must be formed among the parties for the government to function. Legislative progress is marked by extensive discussion and compromise (Hendriks & Toonen, 2001). There are also substantial differences in social norms and attitudes between the two countries. Americans are more religious than the Dutch and less likely to believe that homosexuality, prostitution, abortion, or euthanasia are justifiable (World Values Survey, 2005). Attitudes on abortion and gay marriage specifically were the most strongly related to DS in Inbar et al.’s (2009) American sample, but Dutch social norms around the same topics are more permissive than in the USA. Same-sex marriage has been legal in the Netherlands since 2001, and abortion has been framed as a “medical affair,” with support for a government ban fading over time (Dutch Parliamentary Election Study, 1971–1989; Outshoorn, 2000). In contrast to the USA, large-scale immigration to the Netherlands is a relatively recent phenomenon, and concerns about integration, job displacement, and stress on social services are widespread (Citrin & Sides, 2008; Coenders, Lubbers, Copyright © 2014 John Wiley & Sons, Ltd. Scheepers, & Verkuyten, 2008; “Dutch Immigration: Overflow,” 2013; Mayda, 2006). In designing the political attitude items used in the current studies, we had two equally important goals. The first was to maximize ecological validity by asking about a broad swath of currently relevant political questions. To this end, in Study 1, we asked respondents about a wide range of topics including sexuality, immigration, religion, foreign policy, European integration, social traditionalism, and the environment. These 39 items were adapted from voter-information websites created to inform the public about how their political views aligned with the platforms of the major political parties (e.g., www.stemwijzer.nl). Our second goal was to test theoretical predictions about the specificity of the relationship between disgust and political attitudes. Disgust is linked to the maintenance of physical and spiritual purity, and is often evoked by behaviors seen as degrading, defiling, or “unnatural” (Rozin et al., 2008; Schweder, Much, Mahapatra, & Park, 1997). Based on this research, we expected that DS would most strongly predict attitudes for political and social issues touching on purity, such as sexuality, drug use (which can be seen as “contaminating” one’s body with foreign substances), and euthanasia (which can be seen as interfering with the “natural” course of life and death). We therefore chose items in Study 1 with an eye toward broadly representing a variety of possible purity-related attitudes. In Study 2, we took a somewhat different but complementary approach. In choosing disgust-relevant items for this study, we primarily drew on a theoretical framework known as the “behavioral immune system” (BIS; Schaller & Duncan, 2007; Schaller & Park, 2011). BIS theory proposes that over the course of human evolution, people developed a set of heuristics to detect the presence of parasites and pathogens in others, as well as a set of behaviors that minimized the risk of infection by avoiding individuals, groups, or behaviors that posed contagion threats. Because the risks of failing to detect a contagious individual (serious illness and possibly premature death) greatly outweigh the cost of wrongly identifying a harmless individual as contagious (the foregone benefits of a positive interaction), the BIS tends to be hypervigilant (Schaller & Duncan, 2007; Schaller & Park, 2011). Like an overly sensitive fire alarm, it “goes off” and triggers avoidance behaviors for stimuli that merely bear a resemblance to infectious agents. The emotion of disgust, with its associated “action tendencies” of withdrawal and avoidance (Haidt, 2003; Mackie, Devos, & Smith, 2000), is believed to play a crucial role in the BIS by motivating avoidance of contaminating individuals, groups, or behaviors. To be clear, the BIS account of disgust and social attitudes is complementary to, not inconsistent with, the argument that disgust is evoked by violations of physical and spiritual purity. Disgust is likely involved both in the avoidance of literal pathogen threats and in the enforcement of socially defined norms of spiritual purity—in fact, the latter are likely based to some degree on the former. In many cases, there is a great deal of overlap between the two approaches. For example, the BIS account also predicts a strong relationship between DS and attitudes toward sex and sexuality. Sexual contact Eur. J. Soc. Psychol. (2014) Disgust sensitivity predicts Dutch politics (specifically, exchanging bodily fluids with a sex partner) incurs a substantial risk of pathogen exposure (Oaten et al., 2009). Sexual promiscuity or unusual (i.e., nonnormative) sexual practices are indeed more likely to evoke disgust and rejection from the more disgust-sensitive (Inbar, Pizarro, & Bloom, 2009; Smith et al., 2011). However, the BIS account also makes a number of novel predictions (see Schaller & Park, 2011). The most important for our purposes is that those with greater chronic BIS activation (e.g., the disgust-sensitive) are predicted to be more xenophobic (i.e., more hostile toward foreigners and immigrants). Over the course of human evolution, strangers (members of other groups or tribes) would have been especially likely to harbor novel (and therefore particularly dangerous) infectious agents. Encountering outsiders should activate the BIS, motivating hostility, rejection, and the accompanying emotion of disgust. In modern Westerners, DS has been found to correlate with more negative attitudes toward immigrants and foreigners (Hodson & Costello, 2007; but see Smith et al., 2011), and experimentally priming disease threats increases negativity toward unfamiliar immigrant groups (Faulkner, Schaller, Park, & Duncan, 2004). We therefore expected that DS might predict views on immigrants and immigration, and for that reason, we included a range of questions on immigration policy. Because Middle Eastern and North African Muslims are currently the most visible and controversial immigrant groups in the Netherlands, we also included questions on attitudes toward Muslims and Islam. To assess differences in chronic DS, we used the Disgust Scale—Revised (DS-R; Haidt, McCauley, & Rozin, 1994, modified by Olatunji et al., 2007), which comprises three subscales: core, animal reminder, and contamination disgust. Core disgust items include “You see maggots on a piece of meat in an outdoor garbage pail,” while animal reminder items include “You see a man with his intestines exposed after an accident.” The five-item contamination disgust subscale taps the tendency to feel disgust at interpersonal contagion threats—for example, “You take a sip of soda, and then realize that you drank from the glass that an acquaintance of yours had been drinking from”—and maps most closely to the behavioral immune system’s conceptualization of social pathogen threats are indeed. Previous research has also shown that of the three subscales, contamination disgust is more predictive of political ideology and behavior than are the other two subscales, or the DS-R as a whole (Inbar, Pizarro, Iyer et al., 2012). In the current research, we therefore performed parallel analyses using the entire DS-R and the contamination subscale. In general, we expected that the contamination subscale would show the strongest relationships with political attitudes. We subjected participants’ responses to the political issue questions to principal components analysis (PCA). This allowed us to investigate whether political attitudes clustered into the theoretically predicted issue domains (i.e., purity in Study 1; sex and immigration in Study 2) and whether selfreported DS predicted factor scores for these domains. We report how we determined our sample sizes, all data exclusions, and all measures in both studies. Copyright © 2014 John Wiley & Sons, Ltd. STUDY 1