Subjective status and perceived legitimacy across countries

Received: 10 July 2019 | Accepted: 21 May 2020 DOI: 10.1002/ejsp.2694 RESEARCH ARTICLE Subjective status and perceived legitimacy across countries Mark J. Brandt1 | Toon Kuppens2 | Russell Spears2 | Luca Andrighetto3 | Frederique Autin4 | Peter Babincak5 | Constantina Badea6 | Jaechang Bae7 | Anatolia Batruch8 | Julia C. Becker9 | Konrad Bocian10 | Bojana Bodroža11 | David Bourguignon12 | Marcin Bukowski13 | Fabrizio Butera8 | Sarah E. Butler14 | Xenia Chryssochoou15 | Paul Conway16 | Jarret T. Crawford17 | Jean-Claude Croizet18 | Soledad de Lemus19 | Juliane Degner20 | Piotr Dragon13 | Federica Durante21 | Matthew J. Easterbrook22 | Iniobong Essien23 | Joseph P. Forgas24 | Roberto González25 | Sylvie Graf26 | Peter Halama27 | Gyuseog Han28 | Ryan Y Hong29 | Petr Houdek30 | Eric R. Igou31 | Yoel Inbar32 | Jolanda Jetten33 | William Jimenez Leal34 | Gloria Jiménez-Moya25 | Jaya Kumar Karunagharan35 | Anna Kende36 | Maria Korzh37 | Simon M. Laham38 | Joris Lammers39 | Li Lim33 | Antony S. R. Manstead40 | Janko Međedović41 | Zachary J. Melton42 | Matt Motyl43 | Spyridoula Ntani15 | Chuma Kevin Owuamalam35 | Müjde Peker44 | Michael J. Platow45 | J. P. Prims42 | Christine Reyna46 | Mark Rubin47 | Rim Saab48 | Sindhuja Sankaran49 | Lee Shepherd50 | Chris G. Sibley51 | Agata Sobkow52 | Bram Spruyt53 | Pernille Stroebaek54 | Nebi Sümer55 | Joseph Sweetman56 | Catia P. Teixeira2 | Claudia Toma57 | Adrienn Ujhelyi36 | Jojanneke van der Toorn58 | Alain van Hiel59,60 | Alejandro Vásquez-Echeverría61 | Alexandra Vazquez62 | Michelangelo Vianello63 | Marek Vranka64 | Vincent Yzerbyt65 | Jennifer L. Zimmerman46 Tilburg University, Tilburg, The Netherlands 1 2 University of Groningen 3 University of Genova 4 CeRCA, Université de Poitiers CNRS 5 University of Presov 6 Université Paris Nanterre Gwangju Welfare Foundation 7 8 University of Lausanne 9 University of Osnabrueck Sopot Faculty of Psychology, SWPS University of Social Sciences and Humanities 10 Department of Psychology, Faculty of Philosophy, University of Novi Sad, Novi Sad, Serbia 11 12 University of Lorraine 13 Jagiellonian University 14 College of DuPage 15 Panteion University of Social and Political Sciences This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2020 The Authors. European Journal of Social Psychology published by John Wiley & Sons Ltd Eur J Soc Psychol. 2020;50:921–942. wileyonlinelibrary.com/journal/ejsp | 921 922 | Florida State University 16 17 The College of New Jersey 18 Université Clermont-Auvergne et CNRS 19 University of Granada Hamburg University 20 University of Milano-Bicocca 21 22 University of Sussex 23 FernUniversität in Hagen University of New South Wales 24 25 Pontificia Universidad Católica de Chile 26 Czech Academy of Sciences 27 Slovak Academy of Sciences 28 Chonnam National University 29 National University of Singapore University of Economics in Prague 30 University of Limerick 31 University of Toronto 32 University of Queensland 33 Universidad de los Andes 34 35 The University of Nottingham Malaysia Campus 36 ELTE Eötvös Loránd University, Budapest, Hungary Ural State Law University 37 University of Melbourne 38 University of Cologne 39 Cardiff University 40 Institute of Criminological and Sociological Research, Belgrade, Serbia 41 University of Illinois, Chicago, IL, USA 42 43 Civil Politics 44 MEF University The Australian National University 45 DePaul University 46 The University of Newcastle, Australia 47 American University of Beirut 48 University of Warsaw 49 Northumbria University 50 University of Auckland 51 52 Wroclaw Faculty of Psychology, SWPS University of Social Sciences and Humanities Vrije Universiteit Brussel 53 University of Copenhagen 54 55 Sabanci University 56 University of Exeter 57 Universite libre de Bruxelles 58 Utrecht University 59 Leiden University 60 University of Gent Universidad de la República, Montevideo, Uruguay 61 62 Universidad Nacional de Educación a Distancia 63 University of Padova 64 Charles University 65 Université catholique de Louvain BRANDT eT Al. | SUBJECTIVE STATUS AND PERCEIVED LEGITIMACY Correspondence Mark J. Brandt, Tilburg University, Tilburg, The Netherlands. Email: M.J.Brandt@tilburguniversity.edu Funding information Center for Intercultural and Indigenous Research, Grant/Award Number: 15110006; H2020 European Research Council, Grant/ Award Number: 759320; Center for Social Conflict and Cohesion Studies, Grant/ Award Number: 15130009; Fondo Nacional de Desarrollo Científico y Tecnológico, Grant/Award Number: 1161371; Spanish Ministry of Economy and Competitiveness, Grant/Award Number: PSI2016-79971-P; Czech Science Foundation, Grant/ Award Number: 20-01214S; Institute of Psychology, Czech Academy of Sciences, Grant/Award Number: RVO: 68081740 923 Abstract The relationships between subjective status and perceived legitimacy are important for understanding the extent to which people with low status are complicit in their oppression. We use novel data from 66 samples and 30 countries (N = 12,788) and find that people with higher status see the social system as more legitimate than those with lower status, but there is variation across people and countries. The association between subjective status and perceived legitimacy was never negative at any levels of eight moderator variables, although the positive association was sometimes reduced. Although not always consistent with hypotheses, group identification, selfesteem, and beliefs in social mobility were all associated with perceived legitimacy among people who have low subjective status. These findings enrich our understanding of the relationship between social status and legitimacy. KEYWORDS legitimacy, social identity, status, system justification 1 | I NTRO D U C TI O N consistent with one another; however, for people with low social status who are disadvantaged by the social system, seeing the self Scholars across the social sciences have debated whether or not peo- and one's group as positive conflicts with the motivation to see the ple with low social status are complicit in their oppression. On the one social system as positive. To resolve this psychological conflict peo- hand, unequal social systems tend to be relatively stable across time, ple with low social status may, under some conditions, legitimize the suggesting some degree of passivity, if not complicity among the pop- social system more than people with high social status. ulace; however, on the other hand, social movements aimed at dis- Scholars have debated the status-legitimacy hypothesis, testing rupting or altering social systems are often spearheaded by the very the hypothesis several times with mixed results (find support: e.g., people disaffected by existing status arrangements. The status-legiti- Henry & Saul, 2006; Li, Yang, Wu, & Kou, 2020; Sengupta, Osborne, & macy hypothesis (as coined by Brandt, 2013; Jost, Banaji, & Sibley, 2015; see Jost, 2017 for a recent summary of relevant work; do Nosek, 2004; Jost & Hunyady, 2003; Jost, Pelham, Sheldon, & Ni not find support: e.g., Brandt, 2013; Caricati, 2017; Caricati & Lorenzi- Sullivan, 2003) predicts that people with lower status will, at least in Cioldi, 2012; Vargas-Salfate, Paez, Liu, Pratto, & Gil de Zúñiga, 2018; some conditions, be more likely to see the social system as legitimate Zimmerman & Reyna, 2013). This work suggests that there is still than people with higher status.1 In this article, we examine the condi- broad scholarly interest in how status and legitimacy are related. We tions under which this hypothesis finds support. We used a novel, aim to add two things to this literature. With Approach 1, we test po- 30-country study on the association between subjective status and tential individual- and societal-level moderators of the associations perceived legitimacy. We had two aims. First, we aimed to understand between different levels of subjective status and perceived legitimacy whether and when people with lower subjective status perceive (i.e., a relative focus). Under what conditions and among people with higher levels of legitimacy than people with higher subjective status. what kinds of perceptions and psychological characteristics is there Second, we aimed to understand what types of factors are associated evidence for the status-legitimacy hypothesis (cf. McGuire, 2013)? with perceived legitimacy for people with low subjective status. With Approach 2, we focus on people with low levels of social status and tested potential predictors of perceived legitimacy among 1.1 | Status-legitimacy hypothesis The logic behind the status-legitimacy hypothesis is that people have motivations to see themselves, their groups, and their larger social systems in a positive light (for full details see Jost et al., 2003). For people with high social status, all of these motivations are generally 1 A related hypothesis predicts that lower power will be associated with higher levels of perceived legitimacy (van der Toorn et al., 2015). Our data also test this hypothesis; however, to sharpen the focus of the manuscript at the request of reviewers we have moved the power related analyses and discussion to supplemental materials. this group. Unlike the first approach, Approach 2 does not focus on whether the predictors differ from people with high levels of status. 1.2 | Approach 1: Testing moderators of the association between subjective status and perceived legitimacy The status-legitimacy hypothesis follows from the idea that people are addressing the threatening feelings of aversive anxiety and arousal resulting from psychological conflict. People alter their 924 | BRANDT eT Al. behaviors, beliefs, and perceptions in order to address this threat, cognitions that could increase dissonance. The dissonance that and when the feelings of threat are assuaged they are less likely to people experience from being in a low status position might alter their behaviors, beliefs, and perceptions (for reviews see Jonas emerge “from the contradictory cognitions that (a) the system et al., 2014; Proulx, Inzlicht, & Harmon-Jones, 2012). Factors that is putting me (and my group) at a disadvantage, and (b) through make social hierarchy appear less consequential in material or psy- our acquiescence, my group and I are contributing to the stability chological terms (we discuss several factors below) should reduce of the system” (Jost et al., 2003, p. 16). Therefore, people who the effects of low subjective status on the anxious arousal that recognize that they are not doing enough to mitigate inequality results from status-based dissonance. Therefore, this reasoning should experience more dissonance. This recognition should in- suggests that when people's feelings of threat are addressed, the crease perceived legitimacy for people who are low status and is status-legitimacy link will be positive (i.e., low status people will be part of the contradictory cognitions originally predicted to cause less inclined to legitimize the system as a way to assuage feelings of lower status people to see the system as more legitimate than threat). Conversely, when people's anxious arousal is exacerbated, higher status people do. the status-legitimacy link will be negative (i.e., low status people will see the system as more legitimate). The amount of civil liberties, meritocratic culture, and inequality within a country could also affect feelings of dissonance. Countries and contexts with more civil liberties and more 1.2.1 | Moderators that reduce threat meritocratic cultures increase the amount of dissonance people low in status experience by implying they have choice and control over their outcomes (Jost et al., 2003, p. 17). Inequality There are potentially many factors that can mitigate anxious may increase the conflict between self/group motivations and arousal. We focus on factors that have appeared in the system acceptance of the system for low status group members (e.g., justification and social identity literatures because these litera- Henry & Saul, p. 376). It may also create the impression that tures have been the focus of debates on this issue (for a recent people have less control over their place within society. Both example see Jost, 2019; Owuamalam, Rubin, & Spears, 2019). The factors should increase the amount of anxious uncertainty most obvious factor is the identification with and salience of a val- and increase the necessity of rationalizing the system. No ued group, which is one way social identity theorists suggest peo- support was found for these predictions in some prior studies ple address anxious arousal (Hogg, 2014) or cope with the threat (Brandt, 2013; Caricati, 2017; Caricati & Lorenzi-Cioldi, 2012; of having a low status (McMahon & Watts, 2002; Tajfel & Trump & White, 2018; for one exception on one measure see Turner, 1979). Consistent with this, the original paper proposing Vargas-Salfate et al., 2018). the status-legitimacy hypothesis argued that if group interests are salient and accessible (something that is likely to be correlated with identification), then support for the status-legitimacy hy- 1.2.3 | Structural moderators that affect threat pothesis is unlikely to emerge (Jost et al., 2003, 2004). Similarly, people with high self-esteem and positive self-views make plans, Structural factors, including levels of inequality or the stability have high levels of personal agency, and high levels of self-cer- of the hierarchy, influence what types of options people per- tainty (e.g., Campbell, 1990; Harter, 1978), suggesting that they ceive that they have and their place within the society. These are less likely to be psychologically affected by uncontrollability perceptions may affect people's perceptions of the whole sys- and similar types of threats associated with low status (see Laurin, tem. For example, to the extent that inequality exacerbates Kay, & Landau, 2018 or Schoel, Bluemke, Mueller, & Stahlberg, feelings of hierarchy and the threatening feelings of low status, 2011 for a similar argument in different domains). This is consist- as well as making status differences more salient, this should ent with the idea that self-esteem should push against system jus- further motivate challenges to the inequality by the low status tification motivations for low status people (cf. Jost, Gaucher, & group (Tajfel & Turner, 1979). This is the opposite of the inequal- Stern, 2015, p. 330). 2 ity prediction above. When looking at the stability of the hierarchy, one of the traditional predictions of social identity theory 1.2.2 | Structural moderators that increase threat through dissonance is that low status group members are less likely to seek social change when the status hierarchy is perceived as stable, compared to when the hierarchy is perceived as unstable (Ellemers, van Knippenberg, & Wilke, 1990; Tajfel & Turner, 1979). That Other factors can exacerbate feelings of anxious arousal by in- is, when the status hierarchy is stable, there is little scope and creasing the amount of dissonance people experience. The hope for social change (i.e., no cognitive alternative to the sta- most direct prediction comes from assessing the contradictory tus quo), and people are more likely to accept the system as 2 Social identity theory and system justification theory have also both used self-esteem as an outcome variable. Although this is interesting, it is not the focus of our investigation. legitimate. This hypothesis is also consistent with work finding that system stability increases a system justification motivation (Laurin, Gaucher, & Kay, 2013). | SUBJECTIVE STATUS AND PERCEIVED LEGITIMACY TA B L E 1 Summary of the moderator hypotheses tested in this investigation (Approach 1) Variable 925 Moderator hypothesis Reduce threat Group Identification Identification-Moderation Hypothesis: The link between status and legitimacy will be negative for people low in group identification and positive for people high in group identification. Self-Esteem Self-Esteem-Moderation Hypothesis: The link between status and legitimacy will be negative for people low in self-esteem and positive for people high in self-esteem. Increase threat through dissonance Inequality contribution Inequality Contribution-Moderation Hypothesis: The link between status and legitimacy will be negative for people who see themselves as contributing to inequality and positive for people who do not see themselves as contributing to inequality. Civil Liberties Civil Liberties Hypothesis: The link between status and legitimacy will be negative in countries with high levels of civil liberties and positive for people in countries with low levels of civil liberties. Meritocracy Meritocracy Hypothesis: The link between status and legitimacy will be negative in countries with meritocratic cultures and positive for people in countries with less meritocratic cultures. Inequality System Justification Theory (SJT) Inequality Hypothesis: The link between status and legitimacy will be negative in countries with high levels of inequality and positive for people in countries with low levels of inequality. Structural Factors That Affect Threat TA B L E 2 Summary of the predictor hypotheses tested in this investigation (Approach 2) Inequality Social Identity Theory (SIT) Inequality Hypothesis: The link between status and legitimacy will be negative in countries with low levels of inequality and positive for people in countries with high levels of inequality. Stability Stability-Moderator Hypothesis: The link between status and legitimacy will be negative when people see the status hierarchy as stable and positive when people see the status hierarchy as unstable. Variable Predictor hypothesis Social mobility Social Mobility-Legitimacy Hypothesis: People with low status who see the system as having high social mobility will be more likely to see the system as legitimate. Stability Stability-Legitimacy Hypothesis: People with low status who see the status hierarchy as stable will be more likely to see the system as legitimate. Identification Identification-Legitimacy Hypothesis: People with low status who have high group identification will be less likely to see the system as legitimate. Self-esteem Self-Esteem Legitimacy Hypothesis: People with low status who have high self-esteem will be less likely to see the system as legitimate. 1.2.4 | Summary We have identified several potential moderators that might help us 1.3 | Approach 2: Finding the predictors of perceived legitimacy for people with lower levels of subjective status predict when we are more or less likely to find support for the status-legitimacy hypothesis. These moderators are expressed as indi- In addition to searching for moderators, we also ask what pre- vidual hypotheses in Table 1. In the cross-national study that follows, dicts perceived legitimacy for people with low levels of social we test these eight hypotheses. status. Put another way, assuming that people with low levels of 926 | BRANDT eT Al. subjective status vary in the extent to which they see the system allows them to eventually achieve social change as a group (e.g., due as legitimate, what predicts this variation? In contrast to the prior to longer-term status instability; Owuamalam, Rubin, & Spears, 2016, section, this line of questioning does not necessarily imply mod- 2019; Owuamalam, Rubin, Spears, & Weerabangsa, 2017; for a eration effects, as the predictors of perceived legitimacy for lower countervailing view see Jost, 2019). In short, using the system rather status groups could be the same as the predictors of perceived than rejecting it, can be seen as a viable vehicle for group interests, legitimacy for higher status groups (moderation is possible, but especially for high identifiers. This idea can also be tested with our not necessary). This approach moves away from the explicit status data and essentially predicts the opposite of the predictions in the comparisons of the status-legitimacy hypothesis and is consistent prior paragraph. with the broader questions that inspired this hypothesis: Do people see systems that oppress them as just and legitimate, and if so under what conditions? We focus on four different variables that 1.3.1 | Summary might predict perceived legitimacy for people with lower levels of social status (see Table 2). We identified several potential predictors of perceived legitimacy Some work suggests that people desire to live in social systems for people with lower social status that might help us understand in which there is upward mobility and the system is relatively stable the reasons some people with low social status perceive the social (Laurin et al., 2013; Martorana, Galinsky, & Rao, 2005). Consistent system as legitimate. Some predictors indicate that the system is ful- with this, people who perceived no possibilities to move up to a higher filling the person's goals and other predictors indicate that personal status group are less likely to justify the system and more likely to and group goals are prioritized. These hypotheses are specified indi- engage in collective action (Day & Fiske, 2017; Ellemers et al., 1990; vidually in Table 2. In the cross-national study that follows, we test Mandisodza, Jost, & Unzueta, 2006; Tajfel, 1981; Wright, Taylor, & these four hypotheses. Moghaddam, 1990). Similarly, systems with more stable social hierarchies are seen as more legitimate (Laurin et al., 2013) and are less likely to trigger efforts to change the system (Bettencourt, Charlton, 1.4 | An international, multi-lab approach Dorr, & Hume, 2001; Ellemers et al., 1990; Tajfel & Turner, 1979). This suggests that when people see the system as providing social mobil- We conducted a cross-country study on the association between ity and stability, they are more likely to see the system as legitimate. subjective status and perceived legitimacy that allowed us to test When people's group- and self-interests are prioritized, it for moderators of the status-legitimacy relationships, as well as to may be less likely that people perceive the system as legitimate understand what predicts perceived legitimacy for people with if they or their group are not benefitting from that system. That lower levels of subjective status. 3 We consider three types of is, higher levels of group identification and self-esteem may both effects: be associated with lower levels of perceived system legitimacy To assess moderators (Approach 1): among people with low social status. This follows from the idea that group- and self-interest motivations are negatively related 1. We test the interactions between status and the to system-related motivations among low status groups (Jost proposed moderator variables. If the interaction is et al., 2004). It is also consistent with the idea that the effects of significant, we test whether the effect of subjective group- and self-interests—when sufficiently strong—may be more status is negative (consistent with the status-legiti- prominent than the effects of system-interests (Jost et al., 2003, macy hypothesis) at the predicted levels of the mod- 2004, 2011) and with the finding that people with low levels of erator variables (e.g., when identification is low). group identification may be more likely to accept the current situation (Rubin & Hewstone, 2004; Spears, Jetten, & Doosje, 2001). According to these ideas, people with lower subjective status who The assessment of predictors of perceived legitimacy among people with low status will take two steps (Approach 2): have high group identification or who have high self-esteem will be less likely to see the system as legitimate. From our reading of the literature, the predictions in the prior 2a. We examine the main effects of the predictors on perceived legitimacy. Because the hypotheses about paragraph seem most consistent with a straightforward extension of social identity theory's work on group identification to research on perceived legitimacy. However, it is important to note that recent work by some scholars (and co-authors of this article) has predicted the opposite, at least as a function of an additional qualifier. A series of papers by Owuamalam and colleagues have argued that high group identifiers and people under conditions of high group salience may be likely to see the system as legitimate to the extent that it is seen as one that affords the group collective social mobility in the long term and 3 Recently, system justification theorists have proposed that a low sense of power, rather than status, is associated with greater perceived legitimacy. Whereas status indicates the amount of prestige and respect a person or group is accorded in the system, power indicates the amount of control a person has over valued resources (Magee & Galinsky, 2008). Trusting and legitimizing outside sources of control (e.g., governments) can help restore people's sense of control, something that is lacking with low feelings of power (van der Toorn et al., 2015; cf. Friesen et al., 2014; Kay et al., 2008). By perceiving the system as legitimate, people with low feelings of power can regain some feelings of control. We therefore simultaneously tested all hypotheses for interpersonal sense of power. Complete results are in the supplemental materials. | SUBJECTIVE STATUS AND PERCEIVED LEGITIMACY 927 predictors of perceived legitimacy are not specific to a list of all measures and manipulations. Analyses include partici- people with lower levels of subjective status (e.g., social pants who are over 18 and have completed at least the subjective mobility could predict perceived legitimacy for people status and the system justification, trust in government, confidence with both high and low status), this main effect analysis in societal institutions, and legitimacy of the status hierarchy mea- tells us if there is an average effect across the sample. sures. The final sample included 12,788 participants (4,252 men, 8,478 women, 58 with missing responses, Mage = 25.3, SDage = 10.7). 2b. When there is a significant interaction between the predictor and status, we examine whether the predictor still has the predicted significant effect 2.2 | Key predictor variable: subjective status (e.g., a positive effect of social mobility) for people with lower levels of status. This will tell us whether To measure subjective social status, we used the MacArthur Scale of the effect of the predictor is specific to people with Subjective Social Status (Adler, Epel, Castellazzo, & Ickovics, 2000), low subjective status. modified to capture people's sense of status within their country. We chose this measure to allow easier comparisons across countries. Participants were asked to rate themselves on a ladder that ranged 2 | M E TH O D from 1 to 10, where 10 was high status and 1 was low status. The instructions for the measure read as follows: 2.1 | Participants and procedure Think of this ladder as representing where people Sixty-six distinct samples were collected by researchers from 30 coun- who are the best off—those who have the most money, tries. We aimed for 150 participants per sample, so that we had at the most education, and the most respected jobs. At least 150 participants per country to give us approximately 80% the bottom are the people who are the worst off—who stand in [country]. At the top of the ladder are people power to detect a small to medium effect (i.e., r = .22) within each have the least money, least education, and the least re- sample. Data collection was not continued after analysis. To ensure spected jobs or no job. The higher up you are on this that respondents were part of the social and political system, we only ladder, the closer you are to the people at the very included participants who indicated they were either born in the coun- top; the lower you are, the closer you are to the people try or had lived in the country for six or more years and so not every at the very bottom. Please choose the number of the sample resulted in 150 participants.4 All exclusions are reported. rung of the ladder where you think you stand at this Countries, samples, type of sample, proportion of men, mean age, and time in your life, relatively to other people in [country]. sample size are presented in Table 3. Samples included a mix of student samples, community samples, and samples with both students This one-item measure had adequate test-retest reliability in prior and members of the community (the latter two were both considered samples (e.g., Operario, Adler, & Williams, 2004) and is correlated non-student samples for the sake of analyses). Samples are primarily with objective measures of status (e.g., income; Goodman et al., 2003; samples of university students, although some include community Sakurai, Kawakami, Yamaoka, Ishikawa, & Hashimoto, 2010). Responses samples from Mechanical Turk (USA2), community email lists (e.g., in our sample were above the midpoint (M = 6.03, SD = 1.53), but spanned USA6), or a representative sample (NLD3). Multiple samples per coun- the entire range of the measure. See supplemental materials for means try helps guard against the possibility that results for any particular across countries on this and all other individual-level variables. country are dependent on one particular sample. Research was conducted in accordance with APA and national guidelines. Data, code, and materials are available at the following link: https://osf.io/5uxc7/. 2.3 | Key outcome variable: perceived legitimacy Participating labs used materials designed by the first three authors. These labs translated the materials and adjusted them when We included four measures of perceived legitimacy and system justifi- necessary for their language and cultural context (e.g., replacing cation. Multiple research groups studying the status-legitimacy hypoth- “United States” in the system justification measure). Participants esis have used all these measures (e.g., Brandt, 2013; Henry & first completed measures of demographics, including the measure Saul, 2006; Jost et al., 2003; Li et al., 2020). The 8-item general system of subjective social status. They then completed measures about justification scale (Kay & Jost, 2003) includes items like “In general, I find either their perceptions of themselves and their group's position in society to be fair” and “Society is set up so that people usually get what society (the moderators and predictors) or their perceptions of sys- they deserve” (M α = .78, SD α = .06, α range [.66, .85]; Scale M = −0.64; tem legitimacy (the outcomes). The supplemental materials include Scale SD = 1.14; −3 = Disagree strongly to +3 = Agree strongly).5 4 Six or more years was chosen to ensure that participants who were international or exchange students were not included in the sample. 5 For each scale, alpha or the correlation coefficient (for 2-item scales) was calculated in each country separately. Mean, standard deviation, and range of alphas and correlation coefficients across countries are reported in the text. 928 | TA B L E 3 Sample BRANDT eT Al. Sample demographic information. Sorted in alphabetical order by country Country Type Proportion Male M age N AUS1 Australia Student 0.27 21.4 209 AUS2 Australia Student 0.34 19.5 114 AUS3 Australia Student 0.34 19.2 80 AUS4 Australia Student 0.29 20.1 163 AUS5 Australia Student 0.32 23.2 222 BEL1 Belgium Student 0.35 20.0 623 BEL2 Belgium Student 0.54 21.7 137 BEL3 Belgium Student 0.28 19.3 194 BEL4 Belgium Student 0.24 21.9 91 CAN1 Canada Student 0.33 18.7 180 CHL1 Chile Student 0.28 20.9 156 COL1 Colombia Student 0.27 19.9 139 CZE1 Czech Republic Non-Student 0.23 27.5 154 CZE2 Czech Republic Non-Student 0.29 23.7 241 CZE3 Czech Republic Non-Student 0.30 24.8 311 DNK1 Denmark Student 0.16 24.0 162 FRA1 France Student 0.31 21.2 303 FRA2 France Student 0.11 20.8 163 FRA3 France Student 0.46 18.6 180 DEU1 Germany Student 0.36 25.3 50 DEU2 Germany Student 0.18 22.6 133 DEU3 Germany Student 0.22 22.7 151 DEU4 Germany Non-Student 0.81 37.2 89 GBR1 Great Britain Student 0.14 19.8 213 GBR2 Great Britain Student 0.07 19.1 138 GBR3 Great Britain Student 0.18 19.6 169 GBR4 Great Britain Student 0.08 19.1 118 GRC1 Greece Non-Student 0.49 35.9 444 HUN1 Hungary Student 0.22 20.2 144 IND1 India Non-Student 0.74 31.2 449 IRL1 Ireland Student 0.46 24.6 145 ITA1 Italy Student 0.47 43.0 103 ITA2 Italy Student 0.47 44.0 103 ITA3 Italy Student 0.05 26.9 109 LBN1 Lebanon Student 0.51 18.9 204 MYS1 Malaysia Student 0.43 20.8 146 MYS2 Malaysia Non-Student 0.44 24.6 63 NLD1 Netherlands Student 0.26 19.8 184 NLD2 Netherlands Student 0.19 20.0 232 NLD3 Netherlands Non-Student 0.49 40.4 766 NLD4 Netherlands Student 0.20 21.2 176 NZL1 New Zealand Student 0.18 21.0 180 POL1 Poland Student 0.16 28.0 214 POL2 Poland Non-Student 0.20 23.2 160 (Continues) | SUBJECTIVE STATUS AND PERCEIVED LEGITIMACY TA B L E 3 929 (Continued) Sample Country Type Proportion Male M age N POL3 Poland Non-Student 0.65 26.4 166 RUS1 Russia Student 0.34 19.4 117 SRB1 Serbia Student 0.24 20.3 159 SRB2 Serbia Non-Student 0.39 30.7 173 SGP1 Singapore Student 0.27 19.9 196 SVK1 Slovakia Student 0.19 22.6 268 SVK2 Slovakia Non-Student 0.49 46.7 166 KOR1 South Korea Student 0.30 20.6 119 ESP1 Spain Student 0.24 23.1 148 ESP2 Spain Student 0.30 32.7 252 CHE1 Switzerland Student 0.51 20.3 131 TUR1 Turkey Student 0.30 20.5 122 TUR2 Turkey Student 0.23 20.0 99 USA1 United States Student 0.20 20.6 224 USA2 United States Non-Student 0.57 35.1 214 USA3 United States Student 0.42 19.8 316 USA4 United States Student 0.27 19.9 195 USA5 United States Student 0.17 19.4 181 USA6 United States Non-Student 0.23 36.1 368 USA7 United States Non-Student 0.48 41.5 116 URY1 Uruguay Student 0.28 22.2 169 URY2 Uruguay Student 0.21 20.6 184 The 4-item trust in government scale often included in the Finally, a 2-item measure of the perceived legitimacy of the status American National Election Studies (2015; cf. Brandt, 2013) in- hierarchy was used to capture perceived legitimacy in this specific cludes items like “How much of the time do you think you can trust domain (based on Mummendey, Kessler, Klink, & Mielke, 1999). The the government to do what is right?” (1 = None of the time, 2 = Some items were prefaced with “Differences in power and status between of the time, 3 = Most of the time, 4 = Just about always) and “Would groups in [country] are...” and then participants rated the stems “… you say the government is pretty much run by a few big interests illegitimate” and “…unfair” on a scale ranging from −3 = Disagree looking out for themselves or that it is run for the benefit of all the strongly to 3 = Agree strongly (M r = .52, SD r = .14, r range [.33, .80]; people?” (1 = Few big interests, 2 = Benefit of all). Items were first Scale M = −0.83; Scale SD = 1.30). These items were measured in standardized using z-scores and then averaged to form a scale (M the same block of questions as the stability of the status hierarchy α = .61, SD α = .17, α range [−.10, .78,]; Scale M = −0.02; Scale measure (see below). SD = 0.75). 6 The 7-item measure assessing confidence in societal institutions adopted from the General Social Survey (2017), and used in Brandt (2013), was used to tap into perceptions of both governmental and economic systems. For each of seven institutions, participants are asked how much confidence they have in them 2.4 | Individual level measures of moderators and predictors 2.4.1 | Group identification (0 = None at all, 3 = A great deal). Institutions include the armed forces, the police, the courts, the governments of the country, We measured group identification with a three-item scale about congress, major companies, and banks and financial institutions participants’ identification with their social class. The scale was (M α = .77, SD α = .04, α range [.63, .83,]; Scale M = 1.35; Scale prefaced with, “The following questions are about people with a SD = 0.55). similar background and social class as yourself. Social class refers to people with similar opportunities in terms of income, education, 6 The negative α (−.06) for the trust in government scale is found in Russia. The next lowest α is .20 in Lebanon. and social standing, as in the ladder measure that you completed earlier.” The items were “I identify with people from my social 930 | BRANDT eT Al. class”, “I feel solidarity with my social class”, and “My social class groups in [country] are...” and then participants rated the stems “… is an important part of how I see myself”. Participants responded difficult to change” and “…will remain stable over time” on a scale to these items on a scale ranging from −3 = Disagree strongly to ranging from −3 = Disagree strongly to 3 = Agree strongly (M r = .31, 3 = Agree strongly (M α = .67, SD α = .10, α range [.38, .78]; Scale SD r = .14, r range [−.06, .76]; Scale M = 0.82; Scale SD = 1.14).7 M = 0.51; Scale SD = 1.17). 2.4.2 | Self-esteem 2.5 | Societal level measures of moderators We included three measures to assess societal level conditions. We measured self-esteem with the validated single-item measure To capture the amount of civil liberties, we used the 2015 Civil (Robins, Hendin, & Trzesniewski, 2001). This measure reads “I have Liberties subscale of the Democracy index created by the Economist high self-esteem” with answers that ranged from 0 = Not at all to Intelligence Unit (2016). This index uses a combination of survey 6 = Very true of me (Item M = 3.50; Item SD = 1.57). data and expert ratings to estimate how democratically free individual countries are. It has been used in prior work testing similar 2.4.3 | Social mobility questions (Brandt, 2013). To assess the extent to which the culture holds meritocratic values, we combined four items from international surveys. The first two We measured perceptions of social mobility with six items. They are the importance of ambition and hard work for getting ahead in included, “In general, people can easily get ahead in society”, “In life in the ISSP (2009). The third and fourth come from the sixth wave general, people can climb the social ladder and be successful”, of the World Values Survey (2016). The third asks participants to re- “People with a similar background and social class to my own can spond to an item ranging from 1 = “Competition is good. It stimulates easily get ahead in society”, “It is easy for people with a similar people to work hard and develop new ideas” to 10 = “Competition background and social class to my own to climb the social ladder is harmful. It brings out the worst in people”. The fourth asks partic- and be successful”, “I am motivated to climb up the social ladder”, ipants to respond to an item ranging from 1 = “In the long run, hard and “I am able to climb up the social ladder”. All items were meas- work usually brings better life” to 10 = “Hard work doesn't generally ure on a scale ranging from −3 = Disagree strongly to 3 = Agree bring success—it's more a matter of luck and connections”. All four strongly and had good reliability (M α = .72, SD α = .06, α range [.52, items were standardized, the third and fourth items were reverse .81]; Scale M = 0.82; Scale SD = 0.93). scored, and all four items were combined to form a scale (α = .82). To measure the objective inequality, we used the Gini index 2.4.4 | Contribute to inequality from 2013 and obtained from the World Bank (World Bank, 2016). For some countries, the 2013 data were not available, and in these cases the most recent data were included instead. The Gini We created five items to measure whether people felt that they index assesses the amount of income inequality within a region, contributed to inequality and the stability of the system. One item, with higher scores indicating greater inequality. It is a common “I could do more to change differences in power and status between measure of inequality within a society (e.g., Oishi, Kesebir, & groups in society”, did not correlate as expected with the other items Diener, 2011). and substantially reduced the reliability of the scale (α = .58), so we omitted it. The remaining four items read “I contribute to keeping society the way it is”, “I contribute to maintaining the current social 2.6 | Covariates hierarchy”, “I don't do anything to change the current differences in power and status in society”, and “I am not trying to change the We included three covariates for analyses that focused on the current differences in power and status in society”. All items were individual level: self-reported age, gender (0 = women, 1 = men), measured on a scale ranging from −3 = Disagree strongly to 3 = Agree and type of sample (0 = student, 1 = non-student) to adjust for strongly and the four remaining items created a reliable scale (M α = potential background influences. For the models that include pre- .71, SD α = .09, α range [.46, .84]; Scale M = −0.14; Scale SD = 1.12). dictors at the societal level of analysis, we included two countrylevel covariates. We included the GDP per capita for each of the 2.4.5 | Stability of the status hierarchy countries to control for overall wealth (World Bank, 2016). To control for broader regional trends (Kuppens & Pollet, 2014), we also included contrast codes for each of the continents repre- We used a 2-item measure of the perceived stability of the status sented in our data. hierarchy. This was based on measures used in studies from a social identity perspective (Mummendey et al., 1999). The items 7 were prefaced with “Differences in power and status between next lowest correlation is .18 in Canada. The negative correlation (−.06) between the stability items is found in Malaysia. The | SUBJECTIVE STATUS AND PERCEIVED LEGITIMACY 2.7 | A note on coding 931 highest and people with the lowest levels of subjective status. This is in contrast to the prediction of the status-legitimacy hypothesis. We coded all of the variables to range from 0 to 1. Multilevel regression To test whether these results are impacted by covariates, we in- coefficients are then the proportion difference in the outcome variable as cluded age (country-mean-centered), gender (country-mean-centered), one goes from 0 (minimum) to 1 (maximum) on the predictor variable. For type of sample (grand mean-centered), and the type of perceived legiti- example, a coefficient of .05 is a 5% difference in the outcome between macy measure (contrast coded) as covariates, including the interactions people scoring the lowest and the highest on the predictor variable. between these contrast codes and status and sense of power.9 This gives us the average effect of the primary predictors across the four measures of perceived legitimacy and controls for mean differences between the 3 | R E S U LT S measures. The estimate for subjective status is nearly identical to the 3.1 | Preliminary analyses: the effects of subjective status on perceived legitimacy estimate without covariates (b = .151, SE = 0.005, 95% CI [0.141, 0.161]).10 The effect of subjective status is not the same across all people in all situations. We re-estimated the model with covariates and included random slopes at the country-level for subjective status and We use multilevel models to account for participants’ nesting within coun- sense of power. The estimated slope of subjective status on perceived tries and within samples/labs. Models were estimated using lme4 (Bates, legitimacy for each of the countries is in Figure 2. Although the effect Mächler, Bolker, & Walker, 2015) and lmerTest (Kuznetsova, Brockhoff, & of subjective status is always estimated to be positive when predict- Christensen, 2016) packages in R (R Core Team, 2016). Because we are ing perceived legitimacy, it varies in size. In the next section of the interested in the individual-level association between subjective status article, we consider moderators of the effects of subjective status. and perceived legitimacy, we centered subjective status at the countrymean (Enders & Tofighi, 2007). The correlations between the predictor variables and the measures of perceived legitimacy are in Figure 1 (created with GGally, Schloerke et al., 2016). The measures of system justification, 3.2 | Approach 1: Moderators of subjective status on perceived legitimacy trust, confidence, and legitimacy of the status hierarchy were moderately inter-correlated (M α = .70, SD α = .06, α range [.56, .82]). To combine items, We tested individual-level moderators and country-level modera- reduce the overall number of models, and facilitate generalization across tors. Individual level moderators were country-mean-centered and measures, we nested these four measures within participants, resulting in included in separate models as fixed effects in the four-level multi- a four-level multilevel model: legitimacy measures nested in persons, level model used above (see Figure 3 for correlations between nested in samples/labs, nested in countries. In short, this estimates the these measures).11 We built the country-level models using the average effect across multiple measures of system legitimacy that have same four-level multilevel models, with the addition of the coun- been used in the literature, while increasing measurement precision (see try-level moderators and country-level covariates.12 All country- for similar suggestions, Gelman, 2018; McShane, Tackett, Böckenholt, & level variables were grand-mean-centered, with the exception of Gelman, 2019). It is analogous to conducting four studies and estimating the contrast codes for continents.13 In the country-level models, the meta-analytic estimate across the four studies. 8 First, we tested the main effect of subjective status on perceived legitimacy. This test conceptually replicates many prior tests of these hypotheses (e.g., Brandt, 2013; Henry & Saul, 2006; Jost et al., 2003). We regressed perceived legitimacy on country-mean-centered subjective status. We also included country-mean-centered sense of power in this analysis and all other analyses because it was part of our original analysis plan (see Appendix S1). People with higher subjective status are more likely to see the social system as legitimate compared to people with lower subjective status (b = .152, SE = 0.005, 95% CI [0.142, 0.162]). There is a 15% difference in perceived legitimacy between people with the We also conducted the analyses with only the system justification scale because this scale is perhaps the most well-developed of the four scales and it was reliable in all of the countries. Conclusions in the main text are essentially unchanged. There are two 8 exceptions. The negative interaction between stability and status was negative and non-significant and the positive interaction between civil liberties and status was positive and non-significant when only looking at the system justification scale. Importantly, the sizes of these coefficients were very similar to the coefficients in the models including all legitimacy measures, suggesting that the reduction in error variance when using additional data is the reason for the different conclusions when using all legitimacy measures or only the system justification scale. 9 We used the simr package in R (Green & MacLeod, 2016) to see how much power our model and sample size had to detect the effects of subjective status when b = .01, .02, .05, and .10. These analyses showed that we had adequate power to detect effects of at least .02 (i.e. a 2% difference in the outcome variable between the minimum and maximum of the predictor variable; power ≈ 49%, 97%, 100%, 100% respectively). 10 The variance for the intercept at each level of the multilevel models that included covariates was also calculated (Participants σ = .049, Labs σ = .006, Countries σ = .098, Residual σ = .18). Descriptively, there is more variation between countries than there is between labs. 11 We used the simr package in R (Green & MacLeod, 2016) to see how much power our model and sample size had to detect individual-level interaction effects when b = .02, .05, .10, and .20. We used the model with the identification moderator as our base model for these analyses. These analyses showed that we had adequate power to detect moderation effects of at least .10 (i.e., a .10 difference in the unstandardized slope of status/power at the minimum and maximum of the moderator variable; power ≈12%, 46%, 98%, 100% respectively). 12 We used the simr package in R (Green & MacLeod, 2016) to see how much power our model and sample size had to detect country-level interaction effects for both subjective status and sense of power when b = .01, .02, .05, .10, and .20 We used the model with the meritocracy moderator as our base model for these analyses. These analyses showed that we had adequate power to detect country-level moderation effects of at least .20 (power ≈6%, 10%, 45%, 87% respectively). 13 To check the robustness of our results to outliers at the country-level (cf. Ullrich, & Schlüter, 2012), we visually inspected histograms of the country-level predictors. There were clear outliers for both the Gini index and the measure of civil liberties. These models were run both with and without outliers. All models are in the figures summarizing results and the primary models discussed in the text include all data. 932 | BRANDT eT Al. F I G U R E 1 Correlations (above the diagonal) and density plots (below the diagonal) between subjective status, sense of power, and the outcome variables. All variables were country-mean-centered before calculating the correlations. Density plots for each individual variable are on the diagonal. Density plots are a variant of a histogram. The area below the diagonal uses density plots between two variables. SJ = System Justification, Trust = Trust in Government, Confid = Confidence in Societal Institutions, Legit = Legitimacy of the status hierarchy [Colour figure can be viewed at wileyonlinelibrary.com] we included random slopes for subjective status and sense of power. the effect across the range of the moderator variables, rather than only focusing on points ±1 SD of the mean of the moderator variable To interpret the interactions, we used marginal effects plots cre- (cf. Aiken & West, 1991). We also included a histogram of the mod- ated with interplot (Solt & Hu, 2015). This plots the marginal effect erator variable in each plot to illustrate how much of the sample is (sometimes called the simple effect) of the predictor variable for the located at different levels of the moderator variable (cf. Hainmueller, entire range of the moderator variable. This helps us to understand Mummolo, & Xu, 2019). | SUBJECTIVE STATUS AND PERCEIVED LEGITIMACY 933 F I G U R E 2 Estimated slopes of subjective status for each country from the multilevel model. Dots are randomly jittered horizontally to help show the data. 3.2.1 | Moderators that reduce threat emerged between identification and subjective status (Figure 4a). We tested the hypotheses that are based on the assumption that levels of identification and non-significant at the lowest levels of group identification and self-esteem reduce the experience of identification (Figure 4b). It is stronger and significant at high levels. threat. If the hypotheses are supported, we should find positive in- This interaction is in the direction predicted by the hypothesis; how- teractions between the moderators and status, such that at high lev- ever, the effect of subjective status is never significantly negative. els of the moderator variables the link between status and perceived Because this effect is not negative, it is only partially consistent with legitimacy is positive and at low levels of the moderator variables the the full prediction that perceived legitimacy will be higher for low link is negative. status than for high status groups among lower identifiers. The positive marginal effect of subjective status is weaker at low The results for group identification and self-esteem are in Self-esteem, however, did not significantly interact with subjec- Figures 4 and 5, respectively. One significant positive interaction tive status (Figure 5). In short, although group identification showed 934 | BRANDT eT Al. F I G U R E 3 Correlations (above the diagonal) and density plots (below the diagonal) between primary predictor variables and the individual level variables. All variables were country-mean centered before calculating the correlations. Density plots for each individual variable are on the diagonal. ID = Group identification; SelfEst = Self-esteem; SocMob = Social mobility; ConInqu = Contribution to inequality; Stable = Stability of the status hierarchy [Colour figure can be viewed at wileyonlinelibrary.com] partial support for the hypotheses in the status domain, there was the hypotheses are supported, we should find negative interactions be- no support for the hypothesis for self-esteem. tween the moderators and status, such that at high levels of the moderator variables the link between status and perceived legitimacy is 3.2.2 | Moderators that increase threat through dissonance negative and at low levels of the moderator variables the link is positive. There were no significant negative interactions between any of the four moderator variables expected to increase the experience of threat (Figures 6–9). In the case of civil liberties, there Inequality contribution, civil liberties, meritocracy, and inequality were was a positive interaction between civil liberties and subjective all expected to increase the experience of threat through dissonance. If status. When probing the interaction, we found that the effect | SUBJECTIVE STATUS AND PERCEIVED LEGITIMACY 935 F I G U R E 4 Panel A: Effects of subjective status and group identification on perceived legitimacy. Covariates are included in the model. Error bars are 95% confidence intervals. Panel B: Marginal effect of subjective status on perceived legitimacy (y-axis) across the range of group identification (x-axis). Panel C: Marginal effect of group identification on perceived legitimacy (y-axis) across the range of subjective status (x-axis). For both Panels B and C, the grey band around the slope is the 95% confidence interval. In all panels a null effect is highlighted with the dashed line hypothesis. In short, in no cases did we find support for the hypotheses about moderators that could increase the experience of threat. 3.2.3 | Structural factors as moderators The hypotheses suggested that inequality and perceived stability of the social system were structural factors that might moderate the association between status and perceived legitimacy. In contrast to the hypothesis tested in the prior paragraph, one hypothesis is that F I G U R E 5 Effects of subjective status and self-esteem on perceived legitimacy. Covariates are included in the model. Error bars are 95% confidence intervals. A null effect is highlighted with the dashed line countries with high levels of inequality will have stronger, positive associations between status and perceived legitimacy because it makes status differences more salient. We did not find evidence for this (Figure 9). For stability, the hypothesis predicted that the link between status and perceived legitimacy would be negative when perceived stability was high compared to low. This hypothesis would be supported with a negative interaction effect. This effect did not emerge (Figure 10a). Instead, we found the opposite interaction effect (Figure 10b). Contrary to the hypothesis, the link between status and perceived legitimacy is strong and positive when stability is high and it is weaker or non-significant when stability is low. In short, we do not find evidence that inequality may exacerbate status-based conflict. We also find that stability is a significant moderator, but in the opposite direction from that expected. It indi- F I G U R E 6 Effects of subjective status and contributing to inequality on perceived legitimacy. Covariates are included in the model. Error bars are 95% confidence intervals. A null effect is highlighted with the dashed line cates that stability is interpreted differently in the context of real-life social inequality (see also Verkuyten & Reijerse, 2008), compared to stability that is manipulated in groups created in the lab (e.g., Ellemers et al., 1990). A possible interpretation of the pattern found in this study could be that the perception that social inequality will of subjective status (Figure 7b) was positive and stronger at high not change might be an additional reason for judging that inequality levels of civil liberties. At lower levels of civil liberties, the effect is unacceptable (assuming people perceive there to be inequality). At of subjective status was still positive, but weaker and non-signifi- the same time, we note that the reliability of our stability measure cant. This interaction is in the opposite direction predicted by the could be improved. 936 | BRANDT eT Al. F I G U R E 7 Panel A: Effects of subjective status and civil liberties on perceived legitimacy. Covariates are included in the model. Error bars are 95% confidence intervals. Panel B: Marginal effect of subjective status on perceived legitimacy (y-axis) across the range of civil liberties (x-axis). The grey band around the slope is the 95% confidence interval. In all panels a null effect is highlighted with the dashed line. See footnote 13 for outlier explanation [Colour figure can be viewed at wileyonlinelibrary.com] 3.3 | Approach 2: Predictors of perceived legitimacy for people with low levels of subjective status We tested the hypotheses about the predictors of perceived legitimacy for people with low levels of subjective status using the same models described and presented above. To assess whether there was support for the hypotheses, we first examined the main effects of the predictors on perceived legitimacy. This main effect analysis tells us whether there is an average effect across the sample. When there is a significant interaction between the predictor and status, we examine whether the predictor still has the expected effect (e.g., a positive effect of social mobility) for people with lower levels of status (regardless of the pattern for those higher in status). This will F I G U R E 8 Effects of subjective status and meritocracy on perceived legitimacy. Covariates are included in the model. Error bars are 95% confidence intervals. A null effect is highlighted with the dashed line tell us whether the effect of the predictor is specific to people with low social status or sense of power. The hypotheses predicted that people would see the system as more legitimate when it fulfilled a social mobility or stability related goal. For social mobility, there was clear support for the hypothesis. There was a large positive effect of social mobility, such that higher levels of perceived social mobility were associated with higher levels of perceived legitimacy (Figure 11a). This positive effect was moderated by subjective status (Figure 11a), suggesting that the positive effect is variable across levels of subjective status. At low levels of subjective status (Figure 11b), perceived social mobility remained a significant positive predictor of perceived legitimacy. This is consistent with the hypothesis. For stability, results were not consistent with the hypothesis. There was a negative effect of stability on perceived legitimacy (Figure 10a). This negative effect was moderated by subjective status (Figure 10a); however, at low levels of subjective status (Figure 10c) the effect of stability remained negative and significant. In short, when people see social mobility as a possibility, people with low subjective status see the system as more legitimate; however, F I G U R E 9 Effects of subjective status and inequality on perceived legitimacy. Covariates are included in the model. Error bars are 95% confidence intervals. A null effect is highlighted with the dashed line. See footnote 13 for outlier explanation [Colour figure can be viewed at wileyonlinelibrary.com] perceiving the system as more stable is associated with less legitimacy at lower levels of status, which is inconsistent with the hypothesis. We also made the prediction that when personal or group interests are fulfilled, perceived legitimacy of the system is less necessary. The hypotheses predicted that identification and self-esteem | SUBJECTIVE STATUS AND PERCEIVED LEGITIMACY 937 F I G U R E 1 0 Panel A: Effects of subjective status and stability of the subjective status hierarchy on perceived legitimacy. Covariates are included in the model. Error bars are 95% confidence intervals. Panel B: Marginal effect of subjective status on perceived legitimacy (y-axis) across the range of stability (x-axis). Panel C: Marginal effect of stability on perceived legitimacy (y-axis) across the range of subjective status (x-axis). For Panels B and C the grey band around the slope is the 95% confidence interval. In all panels a null effect is highlighted with the dashed line would be negatively associated with levels of perceived legitimacy this debate beyond straightforward main effects by taking two for people with low levels of subjective status. approaches. Inconsistent with hypotheses, there are significant positive effects of identification (Figure 4a) and self-esteem (Figure 5). To see whether this was consistent across people with low subjective status, we examined the interaction effects. The positive effect of self-es- 4.1 | Approach 1: Moderator of the association between subjective status and perceived legitimacy teem was not moderated by subjective status (Figure 5). The positive effect of group identification was moderated by subjective status Approach 1 examined how constructs that reduce threat, constructs (Figure 4a). Here, we find that the effect of identification is not differ- that increase threat, and perceived structural factors may be moder- ent from zero for people who are very low status (Figure 4c). Although ators of the association between status and perceived legitimacy. We this does not confirm the hypothesis, it is also not contrary to it. tested eight specific moderator hypotheses and found partial support for one: the identification-moderation hypothesis (see Table 4). 4 | GENER AL DISCUSSION That is, the most common result was no clear support for the hypotheses. Sometimes the lack of support was due to non-significant interactions. The sensitivity analyses reported in footnotes 9, 11, and Debates around status and perceptions of legitimacy in psychology 12 indicate that we have substantial power to detect main effects are characterized by mixed findings and different theoretical foci and interactions at the individual level. For country-level interaction from different research groups. Consistent with some past work effects, we only had sufficient power to detect large effects. In the (e.g., Brandt, 2013; Caricati, 2017; Caricati & Lorenzi-Cioldi, 2012; cases of null results, studies with even greater statistical power may Kraus & Callaghan, 2014; Vargas-Salfate et al., 2018; Zimmerman find evidence in support of these hypotheses, in opposition to these & Reyna, 2013), but inconsistent with the status-legitimacy hy- hypotheses, or for a null effect (time will tell). In many other cases, pothesis (e.g., Henry & Saul, 2006; Jost et al., 2003), we found that the lack of support was due to a significant interaction that was the subjective status is positively associated with perceived legitimacy. opposite of the prediction (e.g., stability-moderator hypothesis). In However, this is not our primary contribution. We sought to advance these cases (highlighted with ✗op in Table 4), it seems less plausible F I G U R E 1 1 Panel A: Effects of subjective status and social mobility on perceived legitimacy. Covariates are included in the model. Error bars are 95% confidence intervals. Panel B: Marginal effect of social mobility on perceived legitimacy (y-axis) across the range of subjective status (x-axis). For Panel B, the grey band around the slope is the 95% confidence interval. In all panels a null effect is highlighted with the dashed line 938 | BRANDT eT Al. TA B L E 4 Summary of moderation hypotheses (Approach 1) and whether they were supported used as a vehicle for improvements to group status in the long term (Owuamalam, Rubin, & Spears, 2018). Our data lendsome credence to this hypothesis. Moderator hypothesis Although almost never in a direction supportive of the hypothe- Reduce threat Identification-moderation hypothesis ✓ Self-esteem-moderation hypothesis ✗ Increase threat through dissonance ses (see Table 5), we find that group identification, self-esteem, and beliefs in social mobility are all associated with greater perceived legitimacy among people with low subjective status. In this way, we helped fulfill Jost’s (2017) call to better understand what leads peo- Inequality contribution-moderation hypothesis ✗ Civil liberties hypothesis ✗op Meritocracy hypothesis ✗ SJT inequality hypothesis ✗ 4.3 | Strengths, limitations, and future directions SIT inequality hypothesis ✗ We used a design that leverages larger samples from multiple countries Stability-moderator hypothesis ✗ to understand how subjective status is associated with perceived legiti- ple who are oppressed to uphold the system. Structural factors that affect threat op Note: ✓ = indicates support and partial support for the hypothesis, ✗ = indicates no support for the hypothesis. ✗op = indicates a significant interaction effect in the opposite direction. Abbreviations: SIT, Social identity theory; SJT, System justification theory. macy and related constructs. This helps us avoid low statistical power (Sedlmeier & Gigerenzer, 1989) and uses less-WEIRD sampling (Henrich, Heine, & Norenzayan, 2010). That said, hardly any of the samples were representative, and we were not able to recruit participants from all the regions of the globe. Most notably, we were unable to secure data from the continent of Africa, and Asia is underrepresented. However, given TA B L E 5 Summary of predictor hypotheses (Approach 2) and whether they were supported the consistency of our results with those of researchers who have used data from even more diverse and representative samples (Brandt, 2013; Caricati, 2017; Caricati & Lorenzi-Cioldi, 2012; Vargas-Salfate Predictor hypothesis Social Mobility-Legitimacy Hypothesis ✓ Stability-Legitimacy Hypothesis ✗op Identification-legitimacy hypothesis ✗ Self-esteem legitimacy hypothesis ✗op Note: ✓ = indicates support and partial support for the hypothesis, ✗ = indicates no support for the hypothesis. ✗op = indications a significant interaction effect in the opposite direction. et al., 2018), we believe that our data provide a reasonable approximation at this time for the regions that we do cover. A related limitation is that although we report the range of Cronbach's α and r for all scales across countries, we have not formally tested measurement invariance (He & van de Vijver, 2012). It is encouraging that some of the scales we use have been shown to be invariant across at least some countries (e.g., Davidov & Coromina, 2013). Clearly, more work is needed on the social psychology of perceived legitimacy and its development in understudied regions; this future work will surely have both practical and theoretical benefits. that the issue was a lack of statistical power. The predictive power of the hypotheses for moderators was essentially nil. Many of our samples are student samples. This allowed us to use our resources efficiently, but necessarily implies that the samples are younger and better educated than the general population. 4.2 | Approach 2: Predictors of perceived legitimacy for people with lower levels of subjective status Although studies on similar topics (and social psychology more broadly) have relied on student and non-representative samples, it is always possible that samples with more objectively and subjectively low status people, or that deliberately recruit people from Approach 2 examined the predictors of perceived legitimacy for disadvantaged groups (e.g., homeless shelters, soup kitchens) would people with low subjective status. We tested four specific hypoth- reveal different results. That being said, 34% of our sample self-re- eses about these predictors and found support for one: the social ported a subjective social status below the midpoint of the scale, mobility-legitimacy hypothesis (see Table 5). Most of the predictions indicating that our sample cannot easily be dismissed as a sample were not confirmed and, in most cases, results were opposite to the consisting only of people who seem themselves as high status. predicted direction. As one example, the identification-legitimacy Moreover, our samples included representative (sample NLD3) and hypothesis predicted a negative association for members of low community samples similar to those which have been used to study status/power groups. Yet, we found the opposite. Some members low social status in other work (e.g., MTurk in sample USA1 was of our team have started to think through this type of effect (and used in Plantinga, Krijnen, Zeelenberg, & Breugelmans, 2018; Shah, have published some findings consistent with the data in this manu- Shafir, & Mullainathan, 2015). Additionally, our student samples are script). As explained earlier, they have proposed that among people from a diverse array of university types, including public and private with low subjective status, identification will be positively associ- universities and community colleges serving a diverse range of stu- ated with system legitimacy, especially when the system can be dents. Lastly, the main effects of status are similar to studies using | SUBJECTIVE STATUS AND PERCEIVED LEGITIMACY representative samples (e.g., Brandt, 2013), suggesting that our find- 939 4.5 | Conclusion ings are consistent with data using other sampling techniques. We started this project because we noted that there is evidence both consistent and inconsistent with the status-legitimacy hy- 4.4 | Where to now? pothesis. Our study found very little support for moderator hypotheses across 30 countries and several moderator variables. To Our project provides scholars with additional evidence that they can the extent that the phenomenon predicted by the status-legitimacy 14 use to inform the direction of research on status and legitimacy. hypothesis exists, this appears to be quite rare. That does not mean This additional evidence comes in two forms. First, these hypothe- that people with low subjective status never see the system as rel- ses were inspired by both social identity and system justification atively legitimate. Our data also shed light on when this is most theories and primarily made inaccurate predictions in our study. likely to be the case. We find that group identification, self-esteem, This may be because the prediction was wrong, or because some and beliefs in social mobility are all associated with perceived le- auxiliary assumptions did not hold in our particular samples and gitimacy among people who are low status (as well as those with measures. For example, nearly every moderator predicted to reduce high status). We hope that our findings inspire scholars interested feelings of threat or increase feelings of threat through dissonance in this domain to pursue questions related to status and legitimacy were not supported. This may mean that these moderators are not to enrich our understanding of these constructs around the world. associated with threat as expected (auxiliary assumption did not hold), or that threat is not a key mechanism linking status to legiti- AC K N OW L E D G E M E N T S macy (theoretical prediction was wrong). Learning from theoretical All authors provided feedback on the drafts of the manuscript and failures and inaccurate predictions like these can often be more in- helped collect data. Mark J. Brandt, Toon Kuppens, and Russell formative than learning from theoretical successes (Ferguson & Spears organized the project, designed the study, and wrote Heene, 2012; Popper, 1959). the initial drafts of the manuscript. Julia C. Becker, Matthew J. It is also plausible that some of the inaccurate predictions (and Easterbrook, Simon M. Laham, Jolanda Jetten, Chuma Kevin perhaps most of them) can be addressed by increasing the specific- Owuamalam, Mark Rubin, Joseph Sweetman, and Jojanneke van ity of the theories and how theoretical concepts are translated into der Toorn provided additional feedback in the drafting of the concrete operationalizations. For example, social identity theory has manuscript. This research was supported by Fondecyt program traditionally treated legitimacy as a moderator and used outcome (1161371), the Center for Social Conflict and Cohesion Studies measures such as group evaluations, allocation, and social change (15130009), the Center for Intercultural and Indigenous Research strategies. Here, we tried to extend the theory's predictions to le- (15110006), the European Research Council (ERC) under the gitimacy, but did not find consistent support. This may suggest that European Union's Horizon 2020 research and innovation pro- social identity theory's predictions about social change strategies do gramme (grant agreement No 759320), and the Spanish Ministry not extend to legitimacy in a straightforward way, which may also of Economy and Competitiveness (PSI2016-79971-P). The con- help to explain the limited support for predictions derived from social tribution of Sylvie Graf was supported by the grant 20-01214S identity theory. Recent work has therefore investigated how a social from the Czech Science Foundation and by RVO: 68081740 of the identity perspective can be better applied to predict system legiti- Institute of Psychology, Czech Academy of Sciences. macy (Owuamalam et al., 2019). Second, our data provide descriptive information about the precisely estimated associations between a large number of socially and CONFLIC TS OF INTEREST There are no conflicts of interest to report. politically relevant variables. That is, not only did our study fail to confirm a number of hypotheses about status and legitimacy, but it also T R A N S PA R E N C Y S TAT E M E N T provides the relevant associations and non-associations that relevant Data, code, and materials are available at the following link: https:// theories need to explain. For example, despite expectations of both osf.io/5uxc7/. perspectives, a negative association between status and legitimacy was not found at low levels of identification. Theories need to account ORCID for this pattern of results. Similarly, perceived stability was associated Mark J. Brandt https://orcid.org/0000-0002-7185-7031 with lower levels of legitimacy, despite expectations. Updates will need to account for this pattern of results. Although post-hoc expla- REFERENCES nations are possible, we hope that theorists can build clearly specified Adler, N. E., Epel, E. S., Castellazzo, G., & Ickovics, J. R. (2000). Relationship of subjective and objective social status with psychological and physiological functioning: Preliminary data in healthy, White women. Health Psychology, 19, 586–592. https://doi. org/10.1037/0278-6133.19.6.586 Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage. models that can be used to explain the current data and to make new, falsifiable predictions (e.g., Muthukrishna & Henrich, 2019). 14 See Appendix S1 for data that can inform scholars interested in people's sense of power. 940 | American National Election Studies (2015). Codebook for the time series cumulative data file 1948–2012. Retrieved from http://www.elect ionst udies.org/study pages /anes_times eries_cdf/anes_times eries_ cdf_codebook_int.pdf Bates, D., Mächler, M., Bolker, B., & Walker, S. (2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67, 1–48. Bettencourt, B., Charlton, K., Dorr, N., & Hume, D. L. (2001). Status differences and in-group bias: A meta-analytic examination of the effects of status stability, status legitimacy, and group permeability. Psychological Bulletin, 127, 520–542. https://doi. org/10.1037/0033-2909.127.4.520 Brandt, M. J. (2013). Do the disadvantaged legitimize the social system? A large-scale test of the status–legitimacy hypothesis. Journal of Personality and Social Psychology, 104, 765–785. https://doi. org/10.1037/a0031751 Campbell, J. D. (1990). Self-esteem and clarity of the self-concept. Journal of Personality and Social Psychology, 59, 538–549. https://doi. org/10.1037/0022-3514.59.3.538 Caricati, L. (2017). Testing the status-legitimacy hypothesis: A multilevel modeling approach to the perception of legitimacy in income distribution in 36 nations. The Journal of Social Psychology, 157, 532–540. https://doi.org/10.1080/00224545.2016.1242472 Caricati, L., & Lorenzi-Cioldi, F. (2012). Does status matter? Testing hypotheses from strong form of system justification theory. Revue Internationale de Psychologie Sociale, 25, 67–95. Davidov, E., & Coromina, L. (2013). Evaluating measurement invariance for social and political trust in Western Europe over four measurement time points (2002–2008). Ask: Research & Methods, 22, 37–54. Day, M. V., & Fiske, S. T. (2017). Movin’ on up? How perceptions of social mobility affect our willingness to defend the system. Social Psychological and Personality Science, 8, 267–274. https://doi. org/10.1177/1948550616678454 Economist Intelligence Unit (2016). Democracy Index 2015. Democracy in an age of anxiety. London, UK: The Economist. Retrieved from https://www.eiu.com/democracy2015 Ellemers, N., Knippenberg, A., & Wilke, H. (1990). The influence of permeability of group boundaries and stability of group status on strategies of individual mobility and social change. British Journal of Social Psychology, 29, 233–246. https://doi.org/10.1111/j.2044-8309.1990.tb00902.x Enders, C. K., & Tofighi, D. (2007). Centering predictor variables in cross-sectional multilevel models: A new look at an old issue. Psychological Methods, 12, 121–138. https://doi. org/10.1037/1082-989X.12.2.121 Ferguson, C. J., & Heene, M. (2012). A vast graveyard of undead theories: Publication bias and psychological science’s aversion to the null. Perspectives on Psychological Science, 7, 555–561. https://doi. org/10.1177/1745691612459059 Friesen, J. P., Kay, A. C., Eibach, R. P., & Galinsky, A. D. (2014). Seeking structure in social organization: Compensatory control and the psychological advantages of hierarchy. Journal of Personality and Social Psychology, 106, 590–609. https://doi.org/10.1037/a0035620 Gelman, A. (2018). The failure of null hypothesis significance testing when studying incremental changes, and what to do about it. Personality and Social Psychology Bulletin, 44, 16–23. https://doi. org/10.1177/0146167217729162 General Social Survey (2017). General social surveys, 1972–2016: Cumulative Codebook. Retrieved from http://gss.norc.org/docum ents/codebook/GSS_Codebook.pdf Goodman, E., Adler, N. E., Daniels, S. R., Morrison, J. A., Slap, G. B., & Dolan, L. M. (2003). Impact of objective and subjective social status on obesity in a biracial cohort of adolescents. Obesity Research, 11, 1018–1026. https://doi.org/10.1038/oby.2003.140 Green, P., & MacLeod, C. J. (2016). SIMR: An R package for power analysis of generalized linear mixed models by simulation. Methods in Ecology and Evolution, 7, 493–498. BRANDT eT Al. Hainmueller, J., Mummolo, J., & Xu, Y. (2019). How much should we trust estimates from multiplicative interaction models? Simple tools to improve empirical practice. Political Analysis, 27, 163–192. https://doi. org/10.1017/pan.2018.46 Harter, S. (1978). Effectance motivation reconsidered. Toward a developmental model. Human Development, 21, 34–64. He, J., & van de Vijver, F. (2012). Bias and equivalence in cross-cultural research. Online Readings in Psychology and Culture, 2, 2307-919. https://doi.org/10.9707/2307-0919.1111 Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world? Behavioral and Brain Sciences, 33, 61–83. https://doi. org/10.1017/S0140525X0999152X Henry, P. J., & Saul, A. (2006). The development of system justification in the developing world. Social Justice Research, 19, 365–378. https:// doi.org/10.1007/s11211-006-0012-x Hogg, M. A. (2014). From uncertainty to extremism: Social categorization and identity processes. Current Directions in Psychological Science, 23, 338–342. https://doi.org/10.1177/0963721414 540168 International Social Survey Programme (2009). ISSP 2009 Social Inequality IV. Retrieved from https://www.gesis.org/issp/modules/issp-modul es-by-topic/social-inequality/2009/ Jonas, E., McGregor, I., Klackl, J., Agroskin, D., Fritsche, I., Holbrook, C., … Quirin, M. (2014). Threat and defense: From anxiety to approach. In J. M. Olson & M. P. Zanna (Eds.), Advances in experimental social psychology (vol. 49, pp. 219–286). Cambridge, MA: Academic Press. Jost, J. T. (2017). Working class conservatism: A system justification perspective. Current Opinion in Psychology, 18, 73–78. https://doi. org/10.1016/j.copsyc.2017.08.020 Jost, J. (2019). A quarter century of system justification theory: Questions, answers, criticisms, and societal applications. British Journal of Social Psychology, 58, 263–314. https://doi.org/10.1111/bjso.12297 Jost, J. T., Banaji, M. R., & Nosek, B. A. (2004). A decade of system justification theory: Accumulated evidence of conscious and unconscious bolstering of the status quo. Political Psychology, 25, 881–919. https://doi.org/10.1111/j.1467-9221.2004.00402.x Jost, J. T., Gaucher, D., & Stern, C. (2015). “The world isn’t fair”: A system justification perspective on social stratification and inequality. In J. F. Dovidio & J. A. Simpson (Eds.), APA handbook of personality and social psychology (vol. 2, pp. 317–340). Washington, DC: APA. Jost, J., & Hunyady, O. (2003). The psychology of system justification and the palliative function of ideology. European Review of Social Psychology, 13, 111–153. https://doi.org/10.1080/10463280240000046 Jost, J. T., Liviatan, I., van der Toorn, J., Ledgerwood, A., Mandisodza, A., & Nosek, B. A. (2011). System justification: How do we know it’s motivated. In D. Ramona Bobocel, A. C. Kay, M. P. Zanna & J. M. Olson (Eds.), The psychology of justice and legitimacy (pp. 173–204). New York, NY: Taylor and Francis. Jost, J. T., Pelham, B. W., Sheldon, O., & Ni Sullivan, B. (2003). Social inequality and the reduction of ideological dissonance on behalf of the system: Evidence of enhanced system justification among the disadvantaged. European Journal of Social Psychology, 33, 13–36. https:// doi.org/10.1002/ejsp.127 Kay, A. C., Gaucher, D., Napier, J. L., Callan, M. J., & Laurin, K. (2008). God and the government: Testing a compensatory control mechanism for the support of external systems. Journal of Personality and Social Psychology, 95, 18–35. https://doi.org/10.1037/0022-3514.95.1.18 Kay, A. C., & Jost, J. T. (2003). Complementary justice: Effects of “poor but happy” and “poor but honest” stereotype exemplars on system justification and implicit activation of the justice motive. Journal of Personality and Social Psychology, 85, 823–837. https://doi. org/10.1037/0022-3514.85.5.823 Kraus, M. W., & Callaghan, B. (2014). Noblesse oblige? Social status and economic inequality maintenance among politicians. PLoS One, 9, e85293. https://doi.org/10.1371/journal.pone.0085293 SUBJECTIVE STATUS AND PERCEIVED LEGITIMACY Kuppens, T., & Pollet, T. V. (2014). Mind the level: Problems with two recent nation-level analyses in psychology. Frontiers in Psychology, 5, 1110. https://doi.org/10.3389/fpsyg.2014.01110 Kuznetsova, A., Brockhoff, P. B., & Christensen, R. H. B. (2016). lmerTest: Tests in linear mixed effects models. R package version 2.0-33. Retrieved from https://CRAN.R-project.org/package=lmerTest Laurin, K., Gaucher, D., & Kay, A. (2013). Stability and the justification of social inequality. European Journal of Social Psychology, 43, 246–254. Laurin, K., Kay, A. C., & Landau, M. J. (2018). Structure and Goal Pursuit: Individual and cultural differences. Advances in Methods and Practices in Psychological Science, 1, 491–494. https://doi.org/10.1177/25152 45918797130 Li, W., Yang, Y., Wu, J., & Kou, Y. (2020). Testing the status-legitimacy hypothesis in China: Objective and subjective socioeconomic status divergently predict system justification. Personality and Social Psychology Bulletin, 46, 1044–1058. Magee, J. C., & Galinsky, A. D. (2008). Social hierarchy: The self-reinforcing nature of power and status. Academy of Management Annals, 2, 351–398. Mandisodza, A. N., Jost, J. T., & Unzueta, M. M. (2006). “Tall poppies” and “American dreams” reactions to rich and poor in Australia and the United States. Journal of Cross-Cultural Psychology, 37, 659–668. https://doi.org/10.1177/0022022106292076 Martorana, P. V., Galinsky, A. D., & Rao, H. (2005). From system justification to system condemnation: Antecedents of attempts to change power hierarchies. In M. C. Thomas-Hunt (Ed.), Status and groups (research on managing groups and teams, Vol. 7) (pp. 283–313). Bingley, UK: Emerald Group Publishing Limited. McGuire, W. J. (2013). An additional future for psychological science. Perspectives on Psychological Science, 8, 414–423. McMahon, S. D., & Watts, R. J. (2002). Ethnic identity in urban African American youth: Exploring links with self-worth, aggression, and other psychosocial variables. Journal of Community Psychology, 30, 411–431. https://doi.org/10.1002/jcop.10013 McShane, B. B., Tackett, J. L., Böckenholt, U., & Gelman, A. (2019). Largescale replication projects in contemporary psychological research. The American Statistician, 73, 99–105. https://doi.org/10.1080/00031 305.2018.1505655 Mummendey, A., Kessler, T., Klink, A., & Mielke, R. (1999). Strategies to cope with negative social identity: Predictions by social identity theory and relative deprivation theory. Journal of Personality and Social Psychology, 76, 229–245. https://doi.org/10.1037/0022-3514.76.2.229 Muthukrishna, M., & Henrich, J. (2019). A problem in theory. Nature Human Behaviour, 3(3), 221–229. https://doi.org/10.1038/s4156 2-018-0522-1 Oishi, S., Kesebir, S., & Diener, E. (2011). Income inequality and happiness. Psychological Science, 22, 1095–1100. https://doi.org/10.1177/09567 97611417262 Operario, D., Adler, N. E., & Williams, D. R. (2004). Subjective social status: Reliability and predictive utility for global health. Psychology & Health, 19, 237–246. https://doi.org/10.1080/08870 4403100016 38098 Owuamalam, C. K., Rubin, M., & Spears, R. (2016). The system justification conundrum: Re-examining the cognitive dissonance basis for system justification. Frontiers in Psychology, 7, 1889. https://doi. org/10.3389/fpsyg.2016.01889 Owuamalam, C. K., Rubin, M., & Spears, R. (2018). Addressing evidential and theoretical inconsistencies in system-justification theory with a social identity model of system attitudes. Current Directions in Psychological Science, 27, 91–96. https://doi.org/10.1177/09637 21417737136 Owuamalam, C. K., Rubin, M., & Spears, R. (2019). Revisiting 25 years of system motivation explanation for system justification from the perspective of social identity model of system attitudes. British Journal of Social Psychology, 58, 362–381. | 941 Owuamalam, C. K., Rubin, M., Spears, R., & Weerabangsa, M. M. A. (2017). Why do people from low-status groups support class systems that disadvantage them? A test of two mainstream explanations in Malaysia and Australia. Journal of Social Issues, 73, 80–98. Plantinga, A., Krijnen, J. M., Zeelenberg, M., & Breugelmans, S. M. (2018). Evidence for opportunity cost neglect in the poor. Journal of Behavioral Decision Making, 31, 65–73. https://doi.org/10.1002/ bdm.2041 Popper, K. (1959). The logic of scientific discovery. New York, NY: Basic Books. Proulx, T., Inzlicht, M., & Harmon-Jones, E. (2012). Understanding all inconsistency compensation as a palliative response to violated expectations. Trends in Cognitive Sciences, 16, 285–291. https://doi. org/10.1016/j.tics.2012.04.002 R Core Team (2016). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Available from https://www.R-project.org/ Robins, R. W., Hendin, H. M., & Trzesniewski, K. H. (2001). Measuring global self-esteem: Construct validation of a single-item measure and the Rosenberg Self-Esteem Scale. Personality and Social Psychology Bulletin, 27, 151–161. https://doi.org/10.1177/0146167201272002 Rubin, M., & Hewstone, M. (2004). Social identity system justification and social dominance: Commentary on Reicher Jost et al., and Sidanius et al. Political Psychology, 25, 823–844. https://doi. org/10.1111/j.1467-9221.2004.00400.x Sakurai, K., Kawakami, N., Yamaoka, K., Ishikawa, H., & Hashimoto, H. (2010). The impact of subjective and objective social status on psychological distress among men and women in Japan. Social Science & Medicine, 70, 1832–1839. https://doi.org/10.1016/j.socsc imed.2010.01.019 Schloerke, B., Crowley, J., Cook, D., Briatte, F., Marbach, M., Thoen, E., … Larmarange, J. (2016). GGally: Extension to ‘ggplot2’. R package version 1.3.0. Retrieved from https://CRAN.R-proje ct.org/packa ge=GGally Schoel, C., Bluemke, M., Mueller, P., & Stahlberg, D. (2011). When autocratic leaders become an option: Uncertainty and self-esteem predict implicit leadership preferences. Journal of Personality and Social Psychology, 101, 521–540. https://doi.org/10.1037/a0023393 Sedlmeier, P., & Gigerenzer, G. (1989). Do studies of statistical power have an effect on the power of studies? Psychological Bulletin, 105, 309–316. https://doi.org/10.1037/0033-2909.105.2.309 Sengupta, N. K., Osborne, D., & Sibley, C. G. (2015). The status-legitimacy hypothesis revisited: Ethnic-group differences in general and dimension-specific legitimacy. British Journal of Social Psychology, 54, 324–340. https://doi.org/10.1111/bjso.12080 Shah, A. K., Shafir, E., & Mullainathan, S. (2015). Scarcity frames value. Psychological Science, 26, 402–412. https://doi.org/10.1177/09567 97614563958 Solt, F., & Hu, Y. (2015). Interplot: Plot the effects of variables in interactions terms. Retrieved from https://CRAN.R-project.org/package=inter plot Spears, R., Jetten, J., & Doosje, B. (2001). The (il)legitimacy of ingroup bias: From social reality to social resistance. In J. Jost & B. Major (Eds.), The psychology of legitimacy: Emerging perspectives on ideology, justice, and intergroup relations (pp. 332–362). New York, NY: Cambridge University Press. Tajfel, H. (1981). Human groups and social categories: Studies in social psychology. CUP Archive. Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33–47). Monterey, CA: Brooks/Cole. Trump, K. S., & White, A. (2018). Does inequality beget inequality? Experimental tests of the prediction that inequality increases system justification motivation. Journal of Experimental Political Science, 5, 206–216. https://doi.org/10.1017/XPS.2018.2 942 | Ullrich, J., & Schlüter, E. (2012). Detecting nasty data with simple plots of complex models: Comment on Brandt (2011). Psychological Science, 23(7), 824–825. https://doi.org/10.1177/0956797611434541 Van der Toorn, J., Feinberg, M., Jost, J. T., Kay, A. C., Tyler, T. R., Willer, R., & Wilmuth, C. (2015). A sense of powerlessness fosters system justification: Implications for the legitimation of authority, hierarchy, and government. Political Psychology, 36, 93–110. https://doi. org/10.1111/pops.12183 Vargas-Salfate, S., Paez, D., Liu, J. H., Pratto, F., & Gil de Zúñiga, H. (2018). A comparison of social dominance theory and system justification: The role of social status in 19 nations. Personality and Social Psychology Bulletin, 44, 1060–1076. https://doi.org/10.1177/01461 67218757455 Verkuyten, M., & Reijerse, A. (2008). Intergroup structure and identity management among ethnic minority and majority groups: The interactive effects of perceived stability, legitimacy, and permeability. European Journal of Social Psychology, 38, 106–127. https://doi. org/10.1002/ejsp.395 World Bank (2016). World Development Indicators. Retrieved from https:// data.worldbank.org/data-catalog/world-development-indic ators World Values Survey Association (2016). World Values Survey Wave 6 2010-2014. Retrieved from http://www.worldvaluessurvey.org/ WVSDocumentationWV6.jsp BRANDT eT Al. Wright, S. C., Taylor, D. M., & Moghaddam, F. M. (1990). Responding to membership in a disadvantaged group: From acceptance to collective protest. Journal of Personality and Social Psychology, 58, 994–1003. https://doi.org/10.1037/0022-3514.58.6.994 Zimmerman, J. L., & Reyna, C. (2013). The meaning and role of ideology in system justification and resistance for high-and low-status people. Journal of Personality and Social Psychology, 105, 1–23. https://doi. org/10.1037/a0032967 S U P P O R T I N G I N FO R M AT I O N Additional supporting information may be found online in the Supporting Information section. How to cite this article: BrandtMJ, Kuppens T, Spears R, et al. Subjective status and perceived legitimacy across countries. Eur J Soc Psychol. 2020;50:921–942. https://doi.org/10.1002/ ejsp.2694