Superheroes for Change: Physical Safety and Socially Progressive Attitudes PDF

EJSP RESEARCH ARTICLE Superheroes for change: Physical safety promotes socially (but not economically) progressive attitudes among conservatives Jaime L. Napier* , Julie Huang†, Andrew J. Vonasch‡ & John A. Bargh§ * Department of Psychology, New York University Abu Dhabi, Abu Dhabi,...

EJSP RESEARCH ARTICLE Superheroes for change: Physical safety promotes socially (but not economically) progressive attitudes among conservatives Jaime L. Napier* , Julie Huang†, Andrew J. Vonasch‡ & John A. Bargh§ * Department of Psychology, New York University Abu Dhabi, Abu Dhabi, UAE † Stony Brook University, Stony Brook, New York, USA ‡ Department of Psychology and Neuroscience, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA § Department of Psychology, Yale University, New Haven, Connecticut, USA Correspondence Abstract Jaime Napier, Department of Psychology, New York University Abu Dhabi, P.O. Box Across two studies, we find evidence for our prediction that experimentally 129188, Abu Dhabi, UAE. increasing feelings of physical safety increases conservatives’ socially E-mail: [email protected] progressive attitudes. Specifically, Republican and conservative participants who imagined being endowed with a superpower that made them Received: 10 August 2016 invulnerable to physical harm (vs. the ability to fly) were more Accepted: 13 March 2017 socially (but not economically) liberal (Study 1) and less resistant to social change (Study 2). Results suggest that socially (but not economically) https://doi.org/10.1002/ejsp.2315 conservative attitudes are driven, at least in part, by needs for safety Keywords: political ideology, threat, and security. safety, social change ‘Let me assert my firm belief that the only thing we have to Threat and Political Attitudes fear is fear itself—nameless, unreasoning, unjustified terror which paralyzes needed efforts to convert retreat into The theory of ideology as motivated social cognition advance.’—Franklin D. Roosevelt, Inaugural Address, holds that there is an “elective affinity” between 1933 psychological needs for certainty and safety on the one hand, and the structure and content of politically In the first inaugural address of Franklin D. conservative ideology on the other hand (Jost et al., Roosevelt (1938), given amidst the widespread disquiet 2013; Jost, Glaser, Kruglanski, & Sulloway, 2003), of the Great Depression, the president famously especially regarding social issues (Malka & Soto, warned Americans that their fear could serve 2015). Supporting this perspective, a surge of evidence as a psychological impediment to much needed connects individual differences in threat sensitivity social change. Decades later, research bears out to conservative attitudes. Block and Block (2006) Roosevelt’s supposition: Across several disciplines and found that preschool children characterized as methodologies, research consistently demonstrates an fearful and inhibited were statistically more likely association between threat, broadly defined, and than their more emotionally resilient classmates political conservatism. Such work has shown that: (i) to report conservative attitudes at age 23. As adults, political conservatives are, on average, more likely to conservatives (vs. liberals) perceive the world as a perceive threat than their liberal counterparts; and (ii) more dangerous place (Duckitt, Wagner, du Plessis, & the existence of threat, in myriad forms, is associated Birum, 2002) and appear to be more perceptually with increased endorsement of conservative attitudes vigilant to potentially threatening stimuli (Carraro, that resist efforts toward social change (for reviews, Castelli, & Macchiella, 2011; Castelli & Carraro, 2011; see Hibbing, Smith, & Alford, 2014; Jost, Federico, & Shook & Clay, 2011; Vigil, 2010). Napier, 2013; Jost, Gaucher, & Stern, 2015). Here, Evidence for the connection between threat and we test the novel hypothesis that the opposite of political attitudes has also emerged at the biological threat—that is, heightened feelings of safety—will level. For instance, self-identified political conservatism increase socially progressive beliefs, especially among is positively correlated with the size of the right conservatives. Specifically, we test the prediction that amygdala (a region of the brain implicated in processing experimentally inducing feelings of safety will increase emotion, including fear; Kanai, Feilden, Firth, & Rees, social liberalism among Republicans (Study 1) and 2011), and Republicans (vs. Democrats) displayed more acceptance of social change among conservatives activation in the right amygdala during a risk-taking (Study 2). task (Schreiber et al., 2013), suggesting that European Journal of Social Psychology 48 (2018) 187–195 Copyright ª 2017 John Wiley & Sons, Ltd. 187 Physical safety and liberalism J. L. Napier et al. conservatives may experience stronger emotional perception of a dangerous world is correlated with social reactions than liberals when engaging in risky conservatism, but unrelated to economic conservatism. endeavors. Oxley et al. (2008) documented ideological It should be noted that the evidence in favor of threat differences in physiological responses to threat, such being differentially related to social versus economic that participants who endorsed more socially ideology has been almost exclusively correlational at conservative (vs. liberal) beliefs showed heightened this point. One noteworthy exception to this is research startle reflexes in response to unexpected loud noises on disgust, which has revealed opposite effects on social and elevated skin conductance specifically in response versus economic political positions, such that priming to fear-inducing images (see also Dodd et al., 2012). disgust increases social conservatism, especially on There is also evidence that the existence of threat issues regarding sexuality (Terrizzi, Shook, & McDaniel, increases conservatism, in general, and seems to make 2013), presumably because it increases concerns about “liberals think like conservatives” (Nail, McGregor, contamination, but also leads people to become more Drinkwater, Steele, & Thompson, 2009). For instance, liberal on economic issues, presumably because it researchers have documented a “conservative shift” in increases concerns about justice (Petrescu & Parkinson, attitudes after threatening events, such as terror alerts 2014). Although disgust might be a special type of threat or attacks (Nail et al., 2009; Willer, 2004), and following with a particular relationship to moral judgments (e.g., mortality salience manipulations (Cohen, Ogilvie, Feinberg, Antonenko, Willer, Horberg, & John, 2014; Solomon, Greenberg, & Pyszczynski, 2005; Jost, Haidt, 2001), this work nonetheless is further evidence Fitzsimons, & Kay, 2004; Landau et al., 2004). When that economic and social political positions likely have the system is threatened, people bolster their support different motivational underpinnings. for the “way things are” (Kay et al., 2009; Ullrich & The dual process model of ideology (e.g., Duckitt & Cohrs, 2009), are more likely to endorse conservative Sibley, 2010; Duckitt et al., 2002) offers a theoretical policies and Republican political candidates (Craig & framework for understanding why economic and social Richeson, 2014; Thorisdottir & Jost, 2011), exhibit political attitudes might be differentially determined. increased ingroup favoritism (Nail et al., 2009), and This model focuses on related socio-political constructs derogate those who challenge tradition—for instance, —namely, right-wing authoritarianism and social by engaging in backlash against a woman who dominance orientation—and posits that these two contradicts gender stereotypes (Rudman, Moss- worldviews have different motivational underpinnings. Racusin, Phelan, & Nauts, 2012). What is not known, Specifically, right-wing authoritarianism (i.e., a socially however, is whether people’s political attitudes are conservative belief system characterized by a preference affected by the reduction of that threat. Could making for upholding tradition and resisting social change) is conservatives feel physically safe cause them to “think motivated by needs for security and safety, whereas like liberals” and be more open to social change? social dominance orientation (i.e., a preference for hierarchy and inequality) is motivated by needs for Social versus Economic Political Attitudes power, dominance, and superiority (Duckitt et al., 2002). In line with this, research has shown that the In this work, we test the relationship between feelings perception of a dangerous world (both chronic and of safety and political attitudes. In doing so, we also experimentally induced) is associated with right-wing aim to address some of the limitations of the extant authoritarianism, but not with social dominance literature. First, we make a distinction between social orientation (Duckitt et al., 2002; Duckitt & Fisher, and economic ideology (or, attitudes about social 2003; Duckitt & Sibley, 2010). More generally, although change and inequality, Jost et al., 2003). Although threat does sometimes influence socially dominant much past research tended to rely on a unidimensional attitudes, the relationship between threat and right- measure of ideology, a growing body of evidence wing authoritarianism is consistently stronger (Onraet, suggests that social and economic attitudes are van Hiel, Dhont, & Pattyn, 2013). Similarly, work on differentially determined (Duckitt & Sibley, 2010; risk perception has shown that right-wing Feldman & Johnston, 2014; Federico, Weber, Ergun, & authoritarianism is positively associated with perceived Hunt, 2013; Malka, Soto, Inzlicht, & Lelkes, 2014; riskiness of personal danger hazards, whereas social Weber & Federico, 2013) and that threat, in particular, dominance orientation was either unrelated or is primarily associated with ideological positioning in negatively associated with perceived riskiness in these regard to social/cultural issues, and not economic issues domains (Choma, Hanoch, Gummerum, & Hodson, (Federico, Johnston, & Lavine, 2014; Malka & Soto, 2013). 2014, 2015; Malka et al., 2014). For instance, Malka In one experimental study, participants increased in et al. (2014) found that those who valued security and right-wing authoritarianism (but not social dominance stability over creativity and self-expression were more orientation) when they were told to imagine that society likely to endorse anti-progressive stances on social will deteriorate in the future (i.e., become economically issues (e.g., opposition to homosexuality, abortion, unstable, with higher crime and an ineffective feminism, and immigration), but these values were government) as compared to when the future society unrelated to opposition to income inequality. Similarly, was described as unchanged or improved (Duckitt & analyses by Wright and Baril (2013) revealed that the Fisher, 2003). However, participants in that study did 188 European Journal of Social Psychology 48 (2018) 187–195 Copyright ª 2017 John Wiley & Sons, Ltd. J. L. Napier et al. Physical safety and liberalism not become less authoritarian when society was would be more socially (but not economically) liberal described as becoming safer—that is, more economically (Study 1) and less resistant to social change (but not prosperous and with lower crime—as compared to more egalitarian; Study 2). Because conservatives remaining unchanged. It is hard to conclude from this perceive threat more readily at baseline, when study—which surveyed undergraduates in New compared to liberals, we also predicted that the effect Zealand—whether this result indicates that feelings of of safety on political attitudes would be more pro- safety do not influence political attitudes, or whether nounced for Republicans (vs. Democrats; Study 1) and the participants’ default assumption is that society is conservatives (vs. liberals; Study 2). generally safe, and thus, the notion that society is getting Both studies were reviewed and approved by the even safer and more stable is not impactful. university institutional review board before data Relatedly, a second concern about the previous work collection, and we adhered to all APA ethical guidelines examining the link between threat and political in conducting this research. Participants were informed attitudes is that the construct of “threat” has been that their participation was voluntary and confidential, operationalized in many different ways, oftentimes in and they could discontinue the study at any time ways that are not completely separate from ideological without penalty. content. While it is important to document how political attitudes are affected by political events, including the Study 1 salience of terrorism (Jost et al., 2007; Ullrich & Cohrs, 2009; Willer, 2004), immigration and racial demo- We experimentally induce feelings of physical safety by graphic shifts (Craig & Richeson, 2014), or reminders having participants imagine that they are endowed with of system deterioration (Jost et al., 2015), it is also a superpower that rendered them invulnerable to important to understand if political attitudes are physical harm (vs. the ability to fly, which served as affected by seemingly unrelated events or perceptions. the control condition). Previous work has shown that Indeed, a meta-analytic review of the relationship these two superpowers are seen as equally desirable, between threat and political attitudes found that threats but that people feel significantly safer when they stemming from the external world (e.g., intergroup imagine having physical invulnerability (vs. flying) anxiety, terrorism, and economic instability) were more superpower (Huang, Ackerman, & Bargh, 2013), and strongly related to right-wing attitudes than internal this is verified in the current study. After participants threats (e.g., trait level anxiety and fear of death; Onraet were randomly assigned to superpower condition, they et al., 2013). Because societal-level threats are indicated the extent to which they were conservative inevitably related to politics, it could be that the changes (vs. liberal) on both social and economic issues. We in political attitudes observed are not solely psy- hypothesized that the invulnerability prime would chological, but at least partly due to a deliberate increase social (but not economic) liberalism as preference for more conservative policies in times of compared to the flying prime, especially among threat. Thus, one aim throughout our studies is to Republicans. separate, as much as possible, the threat from the ideological content, thereby offering a particularly strict Method test of the theory of conservatism as motivated social cognition. Specifically, we simply make people imagine Participants. One hundred and fifty-eight that they are physically safe from harm, and test participants were recruited from a University-hosted whether this reduction in perceived vulnerability affects online subject pool in exchange for a chance to win a gift political attitudes. certificate. Eight participants did not respond to our In this work, we test the idea that the need for measure of partisanship (described below), and an physical protection manifests itself in a desire for social additional five did not respond to the dependent protection (i.e., maintaining social order and tradition) measures, yielding a final sample of 145 (66.9% female; —or, as Roosevelt put it, that fear can “paralyze needed 74.5% White; average age, M = 35.31; SD = 13.18). efforts to convert retreat into advance.” Across two studies, we experimentally test the novel hypothesis Procedure. Participants engaged in a visualization that making people feel physically safe will reduce their task (taken from Huang et al., 2013) in which they tendency to seek comfort in the familiar, and instead imagined being endowed with a superpower. become more open to change. Moreover, we examine Specifically, they were told to imagine the following the effects of safety on social versus economic attitudes situation: (Study 1) and attitudes about social change versus equality (Study 2). Participants were randomly assigned ‘On a shopping trip, you wander into a strange store with to imagine being endowed with a superpower, either no sign out front. Everything is dimly lit and the shopkeeper that they were invulnerable to physical harm (safety calls you by name even though you have never seen him condition) or the ability to fly (as a control condition). before. He tells you to come close and he says to you in a We predict that participants who imagine being weird voice “I have decided to give you a gift. Tomorrow, endowed with a superpower that made them you will wake to find that you have a superpower. It will invulnerable to physical harm (vs. the ability to fly) be an amazing ability, but you must keep it absolutely European Journal of Social Psychology 48 (2018) 187–195 Copyright ª 2017 John Wiley & Sons, Ltd. 189 Physical safety and liberalism J. L. Napier et al. secret. If you purposely tell anyone or show off your power, Table 1. Means (and standard deviations) of manipulation check variables you will lose it forever. That night, you have a hard time in Studies 1 and 2 sleeping, but when you wake, you find that you do indeed Superpower have a superpower.’ Flying Invulnerability t p In the safety condition, the passage follows: Study 1 Like power 6.22 (1.98) 5.68 (2.28) 1.54.13 Feeling 5.58 (1.97) 5.49 (1.89).29.77 ‘A glass falls on the floor and without meaning to you Uncertainty 4.52 (2.33) 4.29 (2.08).64.52 accidentally step on the broken glass. It doesn’t hurt you Safety 5.49 (1.69) 6.12 (2.18)
2.00 <.05 at all though, and you realize that you are completely Study 2 invulnerable to physical harm. Knives and bullets would Like power 6.41 (2.07) 6.06 (2.05).98.33 bounce off you, fire won’t burn your skin, a fall from a cliff Feeling 5.90 (1.88) 6.03 (1.86)
.41.68 wouldn’t hurt at all.’ Uncertainty 4.37 (1.94) 4.37 (2.31).00 1.00 Control 5.27 (2.00) 5.49 (2.12)
.61.54 Safety 5.51 (2.04) 6.59 (1.82)
3.21 <.01 In the control condition, participants read a similar passage that had them imagine they could fly, namely: ‘You miss a step going down on the stairs, but instead F(1, 141) = 4.13, p =.044, η2ρ =.03. As shown in of tumbling down, you float gently to the bottom of the Figure 1, the superpower manipulation had no effect banister. You try jumping from the top of the stairs on Democrats’ level of social conservatism (M = 3.57, again and realize that you are able to fly. You can SD = 2.11 in the flying condition; M = 3.77, SD = 2.02 propel yourself through the air as if you were a bird. in the physical invulnerability condition; MD =
.20, You can travel entire distances without even touching SE =.44, p =.647). Republicans, however, reported the ground.’ being significantly less socially conservative in the physical invulnerability condition, M = 5.09, SD = 2.71, In both conditions, the paragraph concluded with the as compared to the flying condition, M = 6.48, SD = 2.06; following: “You don’t have any other super-powers MD = 1.39, SE =.65, p =.034. Looking at it the other though (for example, no super-strength). Everything way, there was approximately a 3-point difference else is exactly the same as it was yesterday.” between Democrats and Republicans on social Participants wrote a few sentences about how they conservatism in the flying condition, MD =
2.91, felt and responded to several questions assessing their SE =.54, p <.001. In the physical invulnerability reactions to their superpower, including mood, liking, condition, this difference, although still significant, was and feelings of safety and uncertainty. Participants then reduced by more than half, MD =
1.33, SE =.56, reported their political views on (i) social and (ii) p =.020. economic issues (1 = Very liberal; 9 = Very conservative) For economic conservatism, only a main effect for and completed a demographic questionnaire, which partisanship emerged, F(1, 141) = 40.26, p <.001, included one item asking which party (Democrat vs. η2ρ =.22: unsurprisingly, Republicans (M = 7.00, Republican) they voted (or would have voted) for in SD = 1.95) were more economically conservative than the most recent election. Forty-five participants (31%) Democrats (M = 4.62, SD = 2.12). In line with our reported a preference for the Republican (vs. expectations, safety had no effect on economic political Democratic) candidate, and this did not differ across orientation nor was there an interaction between conditions, χ 2(1) =.04, p =.833. condition and partisanship, F0 s <.17, p’s >.69, η2ρ <.001. Results In line with results found in Huang et al. (2013), participants who imagined being invulnerable to harm reported feeling significantly safer than participants who imagined having the ability to fly; no other differences emerged between conditions on reactions to the superpower (Table 1). We conducted a multi- variate ANOVA with condition (physical invulnerability vs. flying) and partisanship (Democrat vs. Republican) as fixed factors, and the two political orientation measures (social conservatism and economic conser- vatism) as the dependent variables. For social conservatism, results revealed a significant main effect of partisanship, F(1, 141) = 29.40, Fig. 1: Mean social conservatism among Democrats and Republicans p <.001, η2ρ =.17, that was further qualified by the as a function of super power condition; error bars represent standard predicted condition-by-partisanship interaction, errors of the marginal means (Study 1) 190 European Journal of Social Psychology 48 (2018) 187–195 Copyright ª 2017 John Wiley & Sons, Ltd. J. L. Napier et al. Physical safety and liberalism Discussion certificate. Nine of these participants did not respond to the political orientation measure (described below), As predicted, making Republican participants feel and an additional three did not respond to at least one physically safe increased their liberalism on social of the questions assessing each of dependent variables, issues, but their stance on economic issues was yielding a final sample of 128 (67.2% female; 73.4% unaffected. Contrary to predictions, however, Democrat White; average age, M = 34.67; SD = 13.21). participants’ attitudes (both social and economic) were unaffected by the prime. This is presumably because Democrats (and liberals) are chronically lower on threat Procedure. Among a battery of demographic perceptions than Republicans (and conservatives). This questions, participants reported their general political pattern of results mirrors that found in prior work, orientation (1 = Very liberal; 9 = Very conservative). which has shown that experimentally inducing threat Participants then engaged in the same visualization task does not affect conservatives’ responses, but causes as in Study 1. After writing a few sentences about how “liberals to think like conservatives” (Nail et al., 2009). they felt and responding to several questions assessing This study is the first evidence that we are aware of their reactions to their superpower (Table 1), that goes beyond the relationship between threat and participants responded to questions assessing their conservatism, and suggests that safety can promote resistance to change and acceptance of inequality. more liberal attitudes, at least in regard to social issues. Resistance to change was measured with two items In the next study, we aim to conceptually replicate this used in previous research (e.g., Jost et al., 2007) that finding, examining whether making conservatives feel read: “I would be reluctant to make any large-scale physically safe will increase their progressive attitudes. changes to the social order;” and “I have a preference Researchers have argued that political ideology is for maintaining stability in society, even if there seems composed of two core components—resistance to to be problems with the current system” (r =.432). change and acceptance of inequality (Jost et al., 2003). Acceptance of inequality was assessed with two items Although social and economic political issues can (and (taken from Sidanius & Pratto, 2001) that read: “It’s typically do) include elements of both social change okay if some groups have more of a chance in life than and inequality, evidence suggests that ideology is a dual others;” and “We should do what we can to equalize process, with sociocultural conservative values of conditions for different groups” (reverse scored; preserving tradition and resisting efforts toward change r =.406).1 associated with perceptions of danger, and economic conservative values that bolster hierarchy associated with a view of the world as a competitive jungle Results (Duckitt & Sibley, 2009, 2010). Thus, in Study 2, we examine whether physical invulnerability affects As in Study 1, participants in the safety condition people’s acceptance of social change, specifically, but reported feeling significantly safer than those in the not preference for hierarchy in general. flying condition; there were no differences between conditions on any other reactions to the superpower Study 2 (Table 1). We conducted a fixed-effects mixed linear model In this study, we induce feelings of physical safety (vs. a predicting the endorsement of the dimensions of control condition) as in Study 1 and examine its impact conservatism with dummy codes for dimension type on resistance to societal change and acceptance of (acceptance of inequality vs. resistance to change) inequality, the two core components of political and condition (invulnerability vs. flying), political ideology (Jost et al., 2003). We hypothesized that orientation (mean-centered), and all of the two- and participants would be less resistant to change, but not three-way interaction as predictors. Results revealed less accepting of inequality, when they imagined being the predicted significant three-way interaction between endowed with a superpower that rendered them dimension type, condition, and political orientation, invulnerable to physical harm (vs. the ability fly). b =.50, SE =.16, p =.003. Analysis of the simple slopes Because conservatives are chronically higher on threat showed that for acceptance of inequality, there was the perceptions, and based on the results from Study 1, expected main effect of political orientation, b =.26, where only Republican’s attitudes shifted in response SE =.10, p =.016, such that conservatives reported to the safety prime, we predicted that this effect would higher acceptance of inequality than liberals. There be more pronounced for (or perhaps only emerge was no significant effect of condition, b =.45, SE =.31, among) conservative participants. p =.150, and no reliable interaction between political orientation and condition, b =.14, SE =.15, p =.327. Method 1 We collected participants’ economic and social political orientation at the end of the study, with the anticipation of using them as additional Participants. One hundred and forty people, dependent measures. However, these measures were too highly recruited from an online subject pool, participated in correlated with the initial measure of political orientation (r’s both =.75, this study in exchange for a chance to win a gift p <.001) to be included in the analysis. European Journal of Social Psychology 48 (2018) 187–195 Copyright ª 2017 John Wiley & Sons, Ltd. 191 Physical safety and liberalism J. L. Napier et al. When analyzing the items measuring resistance to effect,” given the moderate levels of social liberalism change, results revealed a main effect of political and resistance to change among the liberals in our study orientation, b =.47, SE =.11, p <.001, with (Figures 1 and 2). More research is needed to fully conservatives reporting higher resistance to change understand the conditions under which liberals and than liberals. The main effect of priming condition was conservatives shift their attitudes about social issues. not significant, b =
.42, SE =.31, p =.18, but there Speculatively, the results from these studies, in was a significant condition-by-political orientation conjunction with prior work showing that threat made interaction, b =
.36, SE =.15, p =.016. Simple slopes dispositional liberals (but not conservatives) report analyses showed that there was no effect of the priming more conservative attitudes (e.g., higher ingroup condition on liberals’ (
1SD) resistance to change, favoritism, lower support for gay rights; Nail et al., b =.30, SE =.36, p =.400; among conservatives 2009), might mean that baseline differences in threat (+1SD), however, resistance to change was significantly perceptions can account for political polarization when lower in the invulnerability condition as compared to it comes to social issues. In other words, ideological the flying condition, b =
1.13, SE =.49, p =.021 differences on issues regarding social change appear to (Figure 2). be minimized (or even eliminated) to the extent that Looking at the interaction the other way, when liberals feel threatened (Nail et al., 2009) or that participants are primed with having the ability to fly, conservatives feel safe (Studies 1 and 2). we find the expected effect of political orientation, such that conservatives are more resistant to change than General Discussion liberals, b =.47, SE =.11, p <.001. Notably, when primed with invulnerability, this ideological difference Across two studies, we demonstrate a relationship was eliminated, b =.12, SE =.10, p =.244, with between safety and political attitudes. The results from conservatives equally as open to social change as these studies illustrate that it is not only the presence liberals. (vs. absence) of threat that influences ideology but also that the opposite of threat—safety and security—can Discussion foster more progressive attitudes. Specifically, we found that experimentally inducing feelings of physical safety Consistent with Study 1, results from this study can lead Republicans and conservatives to embrace confirmed that when conservatives imagined that they more progressive attitudes, including social liberalism were endowed with a superpower that made them in general and acceptance of change, in particular. invulnerable to physical harm (vs. the ability to fly), Whereas the bulk of prior work has shown that aversive and thus feeling physically safe, they were more socially conditions—including system threat, mortality salience, progressive. Specifically, results showed that the safety and even cognitive load—tend to cause a “conservative (vs. control) prime significantly reduced conservatives’ shift,” especially among liberal participants (e.g., resistance to change but did not have a reliable impact Eidelman, Crandall, Goodman, & Blanchar, 2012; Nail on attitudes about inequality. Indeed, the ideological et al., 2009), these studies offer evidence that it is, in difference on resistance to change was, remarkably, fact, possible to get conservatives to become more completely eliminated when participants were primed liberal, at least in regard to social issues. This adds to a with safety. small but growing literature on the psychological Across both studies, liberals’ attitudes were unaffected antecedents of liberalism (e.g., Van der Toorn, Napier, by the manipulation. Although we thought that & Dovidio, 2014). heightening feelings of safety might increase pro- On a theoretical level, this work is a first step in gressive attitudes across the board, the results showed uncovering the insights that can be gleaned from that only conservatives were affected by our manip- disentangling two rather broad constructs, namely, ulation. This does not appear to be due to a “floor threat and political ideology. It raises the question of whether there are meaningful differences in the effects of different types of threat on people’s political attitudes. While threats to the legitimacy and stability of the sociopolitical system seem to lead people to defend all aspects of that system, including both social and economic arrangements (Jost et al., 2015), the studies presented here reveal that physical threat is especially tied to people’s attitudes concerning preserving the social order and resisting social change. Preferences for hierarchy or inequality, by contrast, may be especially enhanced under different types of threat—for instance, the threat of Whites losing power (i.e., competition) or Fig. 2: Mean resistance to change as a function of superpower symbolic threats to the national landscape or values condition and political ideology; error bars represent standard errors (i.e., national identity threat). In an era of historically of the simple slopes (Study 2) unprecedented economic inequality, understanding 192 European Journal of Social Psychology 48 (2018) 187–195 Copyright ª 2017 John Wiley & Sons, Ltd. J. L. Napier et al. Physical safety and liberalism the social conditions that lead people to accept could help explain demographic differences in political inequality, even when that inequality is the result of attitudes. For instance, people who are high (vs. low) change, is an important next step for future researchers. in social class tend to be more socially liberal, but more This research also adds to a growing body of evidence economically conservative (Feldman & Johnston, suggesting that, at least on some occasions, social and 2014; Malka et al., 2014). If high social class decreases economic political attitudes are differentially deter- feelings of vulnerability, the current work could help mined (Malka & Soto, 2015). In our studies, we found explain why this economically conservative demo- that increasing people’s sense of safety did make them graphic is more progressive on social issues than their more socially liberal, but did not influence their stance less advantaged counterparts.2 Similarly, although the on economic issues. It is important to point out, though, relationship between age and political attitudes is that many political issues contain elements of both complicated (Danigelis, Hardy, & Cutler, 2007), some social change and equality. Additional research is work suggests that attitudes on social, but not economic, needed to unravel when and how (and, perhaps, for issues become more conservative with age (Cornelis, whom) attitudes toward change versus equality will Van Hiel, Roets, & Kossowska, 2009). One fascinating take precedent in shaping political opinion. One topic for future research would be to examine whether intriguing proposition is that different types of threat feelings of physical vulnerability (because of age or will lead to conservatism, but through different sickness, for example) lead people to become less pathways. During times of physical danger, opposition socially progressive than they once were. On the flip to same-sex marriage, immigration, or gender equality side, the relative physical strength (or even perceived might be mostly driven by a desire to preserve tradition, invincibility) of young people might be causally related whereas these same political stances, during periods of to the fact that they are society’s most reliably liberal other types of threat (e.g., economic), may be driven cohort, especially on social issues (Leonhardt, 2012). primarily by a motivation to preserve status hierarchies. Indeed, if there is any conservative trend among young One further contribution of our studies is that we use people, it is libertarianism—socially liberal but eco- a manipulation that is completely devoid of ideological nomically conservative (Harvard University, 2014). content. Whereas much of the prior research has relied Decades ago, Roosevelt noted that fear can paralyze on social and political threats (e.g., Jost et al., 2015), the social change; here, we offer empirical support for his results from these studies—where participants simply observation by showing that ameliorating fear can imagined feeling safe—are particularly good evidence facilitate social change. Just as threat can turn liberals for a link between ideology and needs for safety and into conservatives, safety can turn conservatives into security more generally. It should be noted that while liberals—at least while those feelings of threat or there is substantial evidence that links conservatism safety last. and threat (Onraet et al., 2013), there is some disagreement about the underlying mechanism. Whereas some researchers argue that conservatism is References associated with a “negativity bias” (e.g., Hibbing et al., Bargh, J. A., & Shalev, I. (2012). The substitutability of 2014), others contend that it is instead associated with physical and social warmth in everyday life. Emotion, 12, arousal more generally, including both negative and 154–162. https://doi.org/10.1037/a0023527 positive arousal (Tritt, Inzlicht, & Peterson, 2013). Our Block, J., & Block, J. H. (2006). Nursery school personality studies cannot speak to this debate, as it is conceivable and political orientation two decades later. Journal of that imagining oneself as physically invulnerable to Research in Personality, 40, 734–749. https://doi.org/ harm could have been arousal reducing, as well as 10.1016/j.jrp.2005.09.005 threat reducing. Carraro, L., Castelli, L., & Macchiella, C. (2011). The Our findings also demonstrate that physical needs automatic conservative: Ideology-based attentional and desires can manifest in our sociopolitical beliefs, asymmetries in the processing of valenced information. complementing a growing catalog of intervention PloS One, 6(11), e26456. https://doi.org/10.1371/journal. pone.0026456 attempts in other domains of physical experience Castelli, L., & Carraro, L. (2011). Ideology is related to basic (Bargh & Shalev, 2012, Study 3; Huang, Sedlovskaya, cognitive processes involved in attitude formation. Journal Ackerman, & Bargh, 2011; IJzerman et al., 2012, Study of Experimental Social Psychology, 47, 1013–1016. https://doi. 2; Zhong & Liljenquist, 2006). For example, research org/10.1016/j.jesp.2011.03.016 has shown that behaviors addressing people’s needs to Choma, B. L., Hanoch, Y., Gummerum, M., & Hodson, G. avoid physical contamination (such as receiving (2013). Relations between risk perceptions and socio- vaccinations or washing one’s hands) have the added political ideology are domain- and ideology- dependent. benefit of dampening prejudice against groups that are Personality and Individual Differences, 54, 29–34. https://doi. heuristically associated with disease (Huang et al., org/10.1016/j.paid.2012.07.028 2011), and that experiences of physical warmth Cohen, F., Ogilvie, D. M., Solomon, S., Greenberg, J., & ameliorate feelings of loneliness and isolation (i.e., Pyszczynski, T. (2005). American roulette: The effect of social coldness; IJzerman et al., 2012). reminders of death on support for George W. Bush in the The current research leads to the interesting 2 possibility that aspects of the physical environment We are indebted to an anonymous reviewer for this suggestion. European Journal of Social Psychology 48 (2018) 187–195 Copyright ª 2017 John Wiley & Sons, Ltd. 193 Physical safety and liberalism J. L. Napier et al. 2004 Presidential Election. Analyses of Social Issues and Public Harvard University (2014). Institute of politics’ young Policy, 5, 177–187. https://doi.org/10.1111/j.1530-2415. americans’ attitudes toward politics and public 2005.00063.x service: 24th edition. http://www.iop.harvard.edu/ Cornelis, I., Van Hiel, A., Roets, A., & Kossowska, M. (2009). survey-young-americans%E2%80%99-attitude-toward- Age differences in conservatism: Evidence on the politics-and-public-service-24th-edition mediating effects of personality and cognitive style. Journal Hibbing, J. R., Smith, K. B., & Alford, J. R. (2014). Differences of Personality, 77, 51–87. https://doi.org/10.1111/j.1467- in negativity bias underlie variations in political ideology. 6494.2008.00538.x Behavioral and Brain Sciences, 37, 297–307. https://doi.org/ Craig, M. A., & Richeson, J. A. (2014). More diverse yet less 10.1017/S0140525X13001192 tolerant? How the increasingly diverse racial landscape Huang, J. Y., Ackerman, J. M., & Bargh, J. A. (2013). affects White Americans’ racial attitudes. Personality and Superman to the rescue: Simulation physical invulner- Social Psychology Bulletin, 40, 750–761. https://doi.org/ ability attenuates exclusion-related interpersonal biases. 10.1177/0146167214524992 Journal of Experimental Social Psychology, 49, 349–354. Danigelis, N. L., Hardy, M., & Cutler, S. J. (2007). Population https://doi.org/10.1016/j.jesp.2012.12.007 aging, intracohort aging, and sociopolitical attitudes. Huang, J. Y., Sedlovskaya, A., Ackerman, J. M., & Bargh, J. A. American Sociological Review, 72, 812–830. https://doi.org/ (2011). Immunizing against prejudice: Effects of disease 10.1177/000312240707200508 protection on attitudes toward out-groups. Psychological Dodd, M. D., Balzer, A., Jacobs, C., Gruszczynski, M., Smith, Science, 22, 1550–1556. https://doi.org/10.1177/ K. B., & Hibbing, J. R. (2012). The political left rolls with 0956797611417261 the good, the political right confronts the bad: Connecting IJzerman, H., Gallucci, M., Pouw, W. T., Weissgerber, S. C., physiology and cognition to preferences. Philosophical Van Doesum, N. J., & Williams, K. D. (2012). Cold-blooded Transactions of the Royal Society B, 367, 640–649. https:// loneliness: Social exclusion leads to lower skin doi.org/10.1098/rstb.2011.0268 temperatures. Acta Psychologica, 140, 283–288. https://doi. Duckitt, J., & Fisher, K. (2003). The impact of social threat on org/10.1016/j.actpsy.2012.05.002 worldview and ideological attitudes. Political Psychology, 24, Jost, J. T., Federico, C. M., & Napier, J. L. (2013). Political 199–222. https://doi.org/10.1111/0162-895X.00322 ideologies and their social psychological functions. In M. Duckitt, J., & Sibley, C. G. (2009). A dual-process Freeden (Ed.), Oxford handbook of political ideologies motivational model of ideology, politics, and prejudice. (pp. 232–250). New York: Oxford University Press. Psychological Inquiry, 20, 98–109. https://doi.org//10.1080/ Jost, J. T., Fitzsimons, G., & Kay, A. C. (2004). The ideological 10478400903028540 animal: A system justification view. In J. Greenberg, S. L. Duckitt, J., & Sibley, C. G. (2010). Personality, ideology, Koole, & T. Pyszczynski (Eds.), Handbook of experimental prejudice, and politics: A dual-process motivational model. existential psychology (pp. 263–283). New York: Guilford Press. Journal of Personality, 78, 1861–1894. https://doi.org/ Jost, J. T., Gaucher, D., & Stern, C. (2015). “The world isn’t 10.1111/j.1467-6494.2010.00672.x fair”: A system justification perceptive on social Duckitt, J., Wagner, C., du Plessis, I., & Birum, I. (2002). The stratification and inequality. In J. Dovidio, & J. Simpson psychological bases of ideology and prejudice: Testing a dual (Eds.), APA handbook of personality and social psychology, process model. Journal of Personality and Social Psychology, 83, volume 2 group processes (pp. 317–340). Washington, DC: 75–93. https://doi.org/10.1037/0022-3514.83.1.75 American Psychological Association. Eidelman, S., Crandall, C. S., Goodman, J. A., & Blanchar, J. C. Jost, J. T., Glaser, J., Kruglanski, A. W., & Sulloway, F. (2003). (2012). Low-effort thought promotes political Political conservatism as motivated social cognition. conservatism. Personality and Social Psychology Bulletin, 38, Psychological Bulletin, 129, 339–375. https://doi.org/ 808–820. https://doi.org/10.1177/0146167212439213 10.1037/0033-2909.129.3.339 Federico, C. M., Johnston, C. D., & Lavine, H. G. (2014). Jost, J. T., Napier, J. L., Thorisdottir, H., Gosling, S. D., Context, engagement, and the (multiple) functions of Palfai, T. P., & Ostafin, B. (2007). Do needs to reduce negativity bias. Behavior and Brain Sciences, 3, 311–312. uncertainty and threat predict political conservatism or https://doi.org/10.1017/S0140525X13002550 ideological extremity? Personality and Social Psychology Federico, C. M., Weber, C. R., Ergun, D., & Hunt, C. (2013). Bulletin, 33, 989–1007. https://doi.org/10.1177/ Mapping the connections between politics and morality: 0146167207301028 The multiple sociopolitical orientations involved in moral Kanai, R., Feilden, T., Firth, C., & Rees, G. (2011). Political intuition. Political Psychology, 34, 589–610. https://doi.org/ orientations are correlated with brain structure in young 10.1111/pops.12006 adults. Current Biology, 21, 677–680. https://doi.org/ Feinberg, M., Antonenko, O., Willer, R., Horberg, E. J., & 10.1016/j.cub.2011.03.017 John, O. P. (2014). Gut check: Reappraisal of disgust helps Kay, A. C., Gaucher, D., Peach, J. M., Laurin, K., Friesen, J., explain liberal-conservative differences on issues of purity. Zanna, M. P., & Spencer, S. J. (2009). Inequality, Emotion, 14, 513–512. https://doi.org/10.1037/a0033727 discrimination, and the power of the status quo: Direct Feldman, S., & Johnston, C. (2014). Understanding the evidence for a motivation to see the way things are as the determinants of political ideology: Implications of way they should be. Journal of Personality and Social structural complexity. Political Psychology, 35, 337–358. Psychology, 97, 421–434. https://doi.org/10.1037/a0015997 https://doi.org/10.1111/pops.12055 Landau, M. J., Solomon, S., Greenberg, J., Cohen, F., Haidt, J. (2001). The emotional dog and its rational tail: A Pyszczynski, T., Arndt, J., Miller, C. H., … Cook, A. (2004). social intuitionist approach to moral judgment. Deliver us from evil: The effects of mortality salience and Psychological Review, 108, 814–834. reminders of 9/11 on support for President George W. Bush. 194 European Journal of Social Psychology 48 (2018) 187–195 Copyright ª 2017 John Wiley & Sons, Ltd. J. L. Napier et al. Physical safety and liberalism Personality and Social Psychology Bulletin, 30, 1136–1150. Shook, N. J. & Clay, R. (2011). Valence asymmetry in attitude https://doi.org/10.1177/0146167204267988 formation: A correlate of political ideology. Social Leonhardt, D. (2012, 22 June). Old vs. young. The New York Psychological and Personality Science, 2, 650–655. https://doi. Times. org/10.1177/1948550611405219 Malka, A., & Soto, C. J. (2014). How encompassing is the Sidanius, J., & Pratto, F. (2001). Social dominance: An intergroup effect of negativity bias on political conservatism? theory of social hierarchy and oppression. Cambridge: [Commentary]. Behavioral and Brain Sciences, 37, 320–321. Cambridge University Press. https://doi.org/10.1017/S0140525X13002653 Terrizzi, J. A., Shook, N. J., & McDaniel, M. A. (2013). Malka, A., & Soto, C. J. (2015). Rigidity of the economic right? The behavioral immune system and social conservatism: Menu-independent and menu-dependent influences of A meta-analysis. Evolution and Human Behavior, 34, psychological dispositions on political attitudes. Current 99–108. https://doi.org/10.1016/j.evolhumbehav.2012. Directions in Psychological Science, 24, 137–142. https://doi. 10.003 org/10.1177/0963721414556340 Thorisdottir, H., & Jost, J. T. (2011). Motivated closed- Malka, A., Soto, C. J., Inzlicht, M., & Lelkes, Y. (2014). Do mindedness mediates the effect of threat on political needs for security and certainty predict cultural and conservatism. Political Psychology, 32, 785–811. https://doi. economic conservatism? A cross-national analysis. Journal org/10.1111/j.1467-9221.2011.00840.x of Personality and Social Psychology, 106, 1031–1051. https:// Tritt, S. M., Inzlicht, M., & Peterson, J. B. (2013). Preliminary doi.org/10.1037/a0036170 support for a generalized arousal model of political Nail, P. R., McGregor, I., Drinkwater, A., Steele, G., & conservatism. PLoS ONE, 8, e83333. https://doi.org/ Thompson, A. (2009). Threat causes liberals to think like 10.1371/journal.pone.0083333 conservatives. Journal of Experimental Social Psychology, 45, Ullrich, J., & Cohrs, J. C. (2009). Terrorism salience increases 901–907. https://doi.org/10.1016/j.jesp.2009.04.013 system justification: Experimental evidence. Social Justice Onraet, E., van Hiel, A., Dhont, K., & Pattyn, S. (2013). Research, 20, 117–139. https://doi.org/10.1007/s11211- Internal and external threat in relationship with right- 007-0035-y wing attitudes. Journal of Personality, 81, 233–248. https:// Van der Toorn, J., Napier, J. L., & Dovidio, J. F. (2014). We doi.org/10.1111/jopy.12011 the people: Intergroup interdependence breeds liberalism. Oxley, D. R., Smith, K. B., Alford, J. R., Hibbing, M. V., Miller, Social Psychological and Personality Science, 5, 616–622. J. L., Scalora, M., Hatemi, P. K., et al. (2008). Political https://doi.org/10.1177/1948550613512511 attitudes vary with physiological traits. Science, 321, Vigil, J. M. (2010). Political leanings vary with facial 1667–1670. https://doi.org/10.1126/science.1157627 expression processing and psychosocial functioning. Group Petrescu, D. C., & Parkinson, B. (2014). Incidental disgust Processes Intergroup Relations, 5, 547–558. https://doi.org/ increases adherence to left-wing economic attitudes. Social 10.1177/1368430209356930 Justice Research, 27, 464–486. https://doi.org/10.1007/ Weber, C. R. & Federico, C. M. (2013). Moral foundations s11211-014-0221-7 and heterogeneity in ideological preferences. Political Roosevelt, F. D. (1938). Inaugural Address, March 4, 1933. as Psychology, 34, 107–126. https://doi.org/10.1111/j.1467- published in S. Rosenman (Ed.), The public papers of 9221.2012.00922.x Franklin D. Roosevelt, volume two: The year of crisis, 1933 Willer, R. (2004). The effects of government-issued terror (pp. 11–16). Random House: New York. warnings on presidential approval ratings. Current Research Rudman, L. A., Moss-Racusin, C. A., Phelan, J. E., & Nauts, S. in Social Psychology, 10, 1–12. http://www.uiowa.edu/ (2012). Status incongruity and backlash effects: Defending ~grpproc/crisp/crisp10_1.pdf the gender hierarchy motivates prejudice against female Wright, J. C., & Baril, G. (2013). Understanding the role of leaders. Journal of Experimental Social Psychology, 48, dispositional and situational threat sensitivity in our moral 165–179. https://doi.org/10.1016/j.jesp.2011.10.008 judgments. Journal of Moral Education, 42, 383–397. https:// Schreiber, D., Fonzo, G., Simmons, A. N., Dawes, C. T., doi.org/10.1080/03057240.2013.820659 Flagan, T., Fowler, J. H., & Paulus, M. P. (2013). Red brain, Zhong, C.-B., & Liljenquist, K. (2006). Washing away blue brain: Evaluative processes differ in Democrats your sins: Threatened morality and physical cleansing. and Republicans. PloS One, 8, e52970. https://doi.org/ Science, 313, 1451–1452. https://doi.org/10.1126/science. 10.1371/journal.pone.0052970 1130726 European Journal of Social Psychology 48 (2018) 187–195 Copyright ª 2017 John Wiley & Sons, Ltd. 195