Belief in Meritocracy Reexamined: Scrutinizing the Role of Subjective Social Mobility

Despite decreasing intergenerational mobility, strengthening the ties between family background and children’s economic outcomes, Western citizens continue to believe in meritocracy. We study how meritocratic beliefs about success relate to individuals’ social mobility experiences: Is subjective upward mobility associated with meritocratic attributions of success and downward mobility with structuralist views? Whereas previous studies addressed the relevance of individuals’ current position or objective mobility, we leverage diagonal reference models to disentangle the role of subjective mobility, origin, and destination. Surveying a representative Dutch sample (n = 1,507), we find, echoing the Thomas theorem, that if people experience social mobility as real, it is real in its consequences: subjective upward mobility is associated with stronger meritocratic beliefs, and downward mobility is associated with stronger structuralist beliefs—but has no bearing on people’s meritocracy beliefs. This helps understand the muted political response to growing inequality: a small share of upwardly mobile individuals may suffice to uphold public faith in meritocracy.

believes that societal success reflects the meritocratic rewards of hard work (Mijs 2018b).
Past work has studied the discrepancy between fact (e.g., declining intergenerational mobility) and perception (e.g., belief in meritocracy) through the lens of misinformation (Kuklinski et al. 2000), motivated reasoning (Kuziemko et al. 2015), and reward expectations theory (Heiserman and Simpson 2017). In-depth interview studies suggest that people's social mobility experience may be a neglected source of their inequality beliefs (Edmiston 2018;Folkes 2021;Miles, Savage, and Bü hlmann 2011). Having experienced upward mobility may strengthen belief in a meritocratic society where everyone, in principle, can climb the social ladder. Having experienced downward mobility, or no mobility at all, may instead depress those beliefs. Perceptions play a crucial role here.
That is, what matters is whether people consider themselves to be socially mobile or not. Indeed, subjective mobility, rather than objective mobility, most likely affects meritocratic beliefs.
To contribute to this literature at the intersection of social psychology and social stratification, we build on these insights from recent qualitative scholarship to examine how subjectively experienced mobility relates to individuals' beliefs on the meritocratic nature of economic success. Whereas objective mobility is only a proxy for social mobility experiences (ascribing to individuals what scholars determine constitutes upward or downward mobility), our focus on subjective mobility allows for a more direct assessment of the experiential drivers of beliefs about the meritocratic nature of success. 1 In doing so, our study adds novel insights to a long tradition of quantitative scholarship that has focused on the association between objective social mobility and political preferences (Breen 2001;De Graaf, Nieuwbeerta, and Heath 1995), political distrust (Daenekindt, van der Waal, and de Koster 2018), social disorientation (Daenekindt 2017), and support for redistribution (Wilson et al. 2022).
Our study on attributions of economic success complements recent research documenting the association between subjective and objective social position and legitimations of inequality (Buchel, Luijkx, and Achterberg 2021) and the link between subjective mobility and beliefs about poverty (Gugushvili 2016). In a methodological innovation to this scholarship, we draw on diagonal reference models (DRMs) to untangle the role of subjective social origin, social destination, and social mobility (see ''Methods'') in an original high-quality data set on a representative sample of the Dutch population. The Netherlands constitutes an especially suitable case to examine the relationship between experienced social mobility and belief in meritocracy given the documented contrast between 1 To empirically substantiate this claim, we reestimate all models using measures of objective mobility, based on respondents' and their parents' highest completed level of education (cf. Daenekindt et al. 2018:272-73). Online Appendix S6 provides full results. We find no significant association (p \ .05) between downward or upward objective mobility and belief in meritocracy or evidence of a gender interaction. We find a negative association between downward mobility and structuralist beliefs about inequality (-0.28, p \ .05) and a gender interaction indicating that upwardly mobile women-but not men-are more likely, by about .15 points (p \ .05), to attribute success to structural factors. In contrast to our findings on subjective social mobility, these results for objective social mobility do not lend clear theoretical interpretation, emphasizing that it is valuable to address the experiential drivers of beliefs about the meritocratic nature of success, which resonates with the Thomas theorem: (only) if social mobility is experienced as real, it is real in its consequences. perception and reality. Although the Dutch public's perception is of an economically egalitarian nation (Lechner 2012), scholarly research describes high levels of wealth inequality-comparable to the United States (Balestra 2018)-and declining overall social mobility (RMO 2011).
The next section gives a concise theoretical background to our research and introduces our hypotheses before we proceed to discuss results. We reflect on implications for theory, research, and practice in the discussion.

BACKGROUND
Our assessment of the relationship between subjective mobility and the way people explain economic success finds its starting point in the notion of attributional styles. As Weiner (1985:549) observed, ''[W]ithin the achievement domain . . . the most dominant [causal attributions] are ability and effort. That is, success is ascribed to high ability and hard work, and failure is attributed to low ability and the absence of trying.'' In line with previous studies, we focus on meritocratic and structuralist attributions of success. Whereas the perceived importance of hard work best reflects a meritocratic view of success as determined by effort and actions within a person's control, the lottery of birth reflects a distinctly structural cause of life outcomes that lies beyond a person's control (Kluegel and Smith 1986). By focusing on lay attributions of success, we aim to reveal how popular ideas constitutive of general inequality beliefs relate to individual experiences.
Although the public's general overestimation of the meritocratic nature of their society is well documented (Hunt 2007;Mijs 2016Mijs , 2018bReynolds and Xian 2014), there are two reasons to expect that individuals who experienced upward social mobility are especially likely to attribute success to ability and effort. A motivated reasoning perspective produces the expectation that upwardly mobile individuals would be particularly motivated to justify their economic success as meritocratically deserved (Hypothesis 1a; Buchel et al. 2021;Olivos 2021;Reynolds and Xian 2014). Looking at attributions through an experiential inference perspective yields the same expectation: if lay attributions reflect lessons learned through experience, we would expect upwardly mobile individuals to take their own (successful) experience of climbing the social ladder as indication of their society's meritocratic nature (Hypothesis 1b; Cech and Blair-Loy 2010;Hunt 2007;Mijs 2018a).
The motivated reasoning and experiential inference perspectives, however, produce different expectations regarding the consequences of downward mobility. From a motivated reasoning perspective, downwardly mobile individuals should attribute success to structural causes to externalize blame for falling down the social ladder (Hypothesis 2a). From an experiential inference perspective, it is less clear what downward mobility might do to a person's understanding of meritocracy. We may expect some downwardly mobile individuals to lose faith in the meritocratic nature of success, having experienced a fall down the social ladder despite their efforts. Yet the very fact of their downward mobility from a position of relative advantage may strengthen belief in the notion that economic success in society is far from structurally determined. These contrasting expectations mean that there is no clear relationship between downward mobility and attributions of success (Hypothesis 2b).

Data
Our study is based on an original survey fielded with a representative sample of the Dutch population (Mijs, De Koster, and Van der Waal 2021). We contracted CentERdata to recruit a sample of 1,500 adult respondents from the Longitudinal Internet Studies for the Social Sciences (LISS) panel, which is a true probability sample drawn from the population register by Statistics Netherlands. 2 Response rates for our study were higher than anticipated (89 percent), yielding a sample of 1,636 respondents. Our data set for this study consists of 1,541 respondents (94 percent) for whom a match could be found with information from previous LISS surveys (see ''Measures''). We further restrict our sample to adults who are head of the households or their spouse to exclude live-in parents and children with a dependent economic status. Our final sample (n = 1,507) matches population statistics on gender but skews to an older demographic, reflecting our focus on heads of household. 3 We took several steps to secure data quality. First, we designed the survey to be short: the median time of completion was 11 minutes. Second, to minimize selection bias, we gave our survey a nondescript name (''Social Topics in the Netherlands''). Third, we ran checks for survey straightlining but found no concerning patterns.

Measures
Our dependent variables concern belief in meritocracy as indicated by lay attributions of economic success. To allow for a direct comparison with international scholarship, we use the same questions to measure beliefs about the meritocratic and nonmeritocratic causes of success as used in the widely used International Social Survey Programme's (ISSP) Social Inequality module. Respondents were presented with a set of items, for each of which they are asked to assess its importance for economic success on a 5-point scale ranging from not important at all to essential. Our dependent variables are measured using two single items. We focus on the perceived importance of ''hard work'' as a meritocratic attribution of success (M = 3.69, SD = .79), whereas ''coming from a wealthy family'' (M = 2.87, SD = .87) best captures a structuralist understanding of life outcomes (Kluegel and Smith 1986). In recognition of the nonsymmetrical nature of inequality beliefs, we study meritocratic and structuralist attributions separately rather than collapsing the two into a single scale (cf. Hunt 2007;Mijs 2021;Reynolds and Xian 2014). 4 Our key independent variable, subjective social mobility, is based on two social ladder questions (cf. Buchel et al. 2021;Heiserman and Simpson 2017). As in the ISSP (ISSP Research Group 2012), respondents were presented with an image of a ladder with seven rungs (''In our society there are groups which tend to be towards the top and groups which tend to be towards the bottom'') and asked to, first, place themselves on the ladder and second, to think back of their childhood and place the family in which they grew up on the ladder. Based on their responses, we constructed a mobility table that distinguishes between subjectively low (Rungs 1 and 2), medium (Rungs 3-5), and high (Rungs 6 and 7) social origin and destination (Table 1). From the mobility table, we derive our measurement of upward mobility-the sum of cells above the diagonal (n = 201)-and downward mobility-the sum of cells below the diagonal (n = 122). This measurement approach sets up for a conservative estimation of the impact of subjective mobility by focusing on distinctive experiences of upward or downward mobility experienced by a relatively small group of our respondents, which we contrast to the overwhelming majority of our respondents who did not report any meaningful mobility (n = 1,184). To further contextualize our analysis, we note that the prevalence of subjectively experienced upward or downward mobility in the Netherlands is markedly lower than the average of Western and postsocialist European countries reported in Gugushvili (2016).
Our analyses include a minimal set of control variables that extant research (cf. Cech and Blair-Loy 2010;Hunt 2007;Kluegel and Smith 1986;Reynolds and Xian 2014) suggests may impact both the social mobility experience and respondents' attributions of success: gender (52 percent of the sample is female), age (centered around the mean of 59.53, SD = 15.04), immigration background (16 percent of the sample), and cohabitation (through marriage or another form of romantic relationship-about equally common in the Netherlands; 68 percent). 5,6

Analytical Strategy
To test our hypotheses, we use diagonal reference models (DRMs). DRMs were specifically designed to model consequences of social mobility net of the effects of the social position of origin and destination (Sobel 1981; van der Waal, Daenekindt, and de Koster 2017). Because of the linear dependency of mobility on origin and destination, it is not possible to model these three simultaneously using conventional methods (e.g., ordinary least squares) because it would result in identification problems. DRMs circumvent this issue by combining the influence of origin and destination into an intercept for each cell in the mobility table. The point of departure is to look at immobile individuals to capture the origin and destination effects. We specify the baseline model as: where Y is the dependent variable and subscripts i and j refer to the social positions of origin and destination of respondent k . Based on the estimated means for each diagonal cell (containing immobile individuals; m ii and m jj ), a weighted mean is created for each off-diagonal cell (containing mobile individuals). This is done through the p and the q parameters, which are constrained to sum to 1. The p parameter indicates the relative importance of the social position of origin relative to the social position of destination. The q parameter, on the other hand, Note: Cells in bold give the number of upwardly mobile respondents, cells in italics denote downwardly mobile respondents.

5
Online Appendix S2 provides descriptive statistics. 6 Data for this study were collected as part of an unrelated survey experiment that included three experimental conditions. These were controlled for by means of dummy variables in all analyses reported in this article (Online Appendix S5 provides full results).
indicates the relative importance of the social position of destination relative to the social position of origin. Because of this approach (i.e., incorporating origin and destination effects in cell-specific intercepts), DRMs circumvent the identification problem, and we can extend the model with measures of social mobility as simple covariates. 7 In focusing on mobility experiences and belief in meritocracy, we are sensitive to their gendered nature (Newman 1999). This may be at play given different social mobility rates for men and women  and varying views about the drivers of economic success because of gendered experiences of discrimination and structural barriers in the labor market (Cech and Blair-Loy 2010; Cech, Blair-Loy, and Rogers 2018). 8 Table 2 presents the results of our model with meritocratic attributions of success (''hard work'') as the dependent variable. The diagonal intercepts, which are the estimated means for socially immobile individuals, indicate that belief in hard work is not stratified: socially immobile individuals with a low subjective social position of origin and destination have similar expected values as immobile individuals with a high subjective social position of origin and destination (3.71 and 3.73, respectively). The estimate for the p parameter reflecting the importance of social position of origin relative to the social position of destination does not differ significantly from .5; hence, we cannot reject the null hypothesis that social position of origin is as important as social position of destination.

Meritocratic Attributions of Success
We now turn to the variables most relevant for testing our hypotheses, namely, upward and downward subjective social mobility. The positive coefficient for upward mobility (.16; p \ .05) suggests that individuals who experienced upward mobility more strongly believe in the meritocratic nature of success net of their social origin and their current social position-corresponding to a standard deviation difference of about one fifth. This finding is in line with both the motivated reasoning and the experiential inference perspectives (Hypotheses 1a and 1b). We do not find a significant negative association between experienced downward mobility and belief in meritocracy. Hence, we find evidence against the expectations of motivated reasoning (Hypothesis 2a) and in support of the experiential inference perspective (Hypothesis 2b). Table 3 presents the results of our model estimating structuralist attributions of success. As indicated by the diagonal intercepts, this belief is clearly stratified: among individuals not experiencing social mobility, we find a negative relationship between social position and the perceived importance of family wealth. Individuals with a low subjective social position of origin and destination believe more in the importance of family wealth (3.08) than individuals with a high social position (2.76)-a difference of two fifths of a standard deviation. Indeed, in higher social positions, people are significantly less likely to make structuralist attributions of success. In addition, the p parameter indicates that this belief is associated with the social position of origin. Turning to our focal variables, we find a statistically significant positive association between downward social mobility and structuralist attributions of success (.19; p \ .05)-corresponding to a difference of about one fourth of a standard deviation. This suggests that the experience of falling down the social ladder may make people more likely to see economic success in a structural light, which is in line with expectations from motivated reasoning (Hypothesis 2a) but runs counter to the experiential inference perspective (Hypothesis 2b). At the same time, we do not find evidence of a statistically significant negative association between upward mobility and structuralist attributions of success predicted by Hypotheses 1a and 1b.

DISCUSSION
Building on long-standing interest from quantitative scholars in social mobility and political attitudes and insights from recent qualitative research highlighting the importance of subjective experiences of mobility, this research note examined the relationship between subjective social mobility and belief in the meritocratic basis of economic success. This issue has gained more contemporary relevance given declining intergenerational mobility and growing economic inequality. Specifically, we asked how experiencing subjective social mobility shapes people's belief in the meritocratic or structural causes of success. Drawing on highquality original data collection in the Netherlands, we leverage diagonal reference models to analyze how the social mobility experience associates with men and women's attributions of economic success net of social origin and social destination. Overall, the documented association between subjective social origin and attributions of success suggests that individuals with a higher social status downplay the importance of family wealth in a self-interested manner (cf. Buchel et al. 2021). In addition, our findings suggest that beyond their current social position, individuals' attributions of success are marked by the experience of social mobility itself. Given the importance of subjective mobility experiences and the absence of a clear link between objective social mobility and attributions of success, 9 we conclude in line with the Thomas theorem that (only) if people experience social mobility as real, it is real in its consequences.
Our analysis lends partial support to two perspectives linking subjective mobility experiences to lay attributions of success. In line with a motivated reasoning perspective, we find that upwardly mobile individuals are more likely to attribute success to meritocratic causes. Conversely, downwardly mobile individuals are more likely to attribute success to structural causes. However, we do not find a corresponding negative association between downward mobility and belief in meritocracy, nor does upward mobility negatively associate with structural explanations of success.
Our findings also provide mixed support to expectations of an experiential perspective whereby individuals draw inferences about the causes of success from their subjective experiences of mobility. In line with this perspective, we find a positive association between experiences of upward mobility and meritocratic attributions and no relationship between downward mobility and belief in meritocracy.
We caution against causal interpretation of the observed patterns given the cross-sectional nature of our data. Our theorizing presumes that individuals make attributions of success by inference from their own experience (cf. Mijs 2018a), but our data cannot provide direct evidence that this is indeed the mechanism linking subjectively experienced mobility and attributions of success. Rather, our findings suggest a promising avenue for future research. Quantitative scholarship could, for instance, use panel data to identify the causal effect of 9 See Footnote 1 and Online Appendix S6. changes in a person's subjective social status on that person's meritocratic and structuralist beliefs about success. Qualitative studies could draw on life-course interviews to analyze whether and how individuals link their own mobility experiences to broader beliefs about success (see Mijs and Savage 2020;Miles et al. 2011).
Despite these limitations, our findings confirm and complement past research documenting the nonsymmetrical nature of beliefs about the causes of success. Because a stronger belief in the relevance of meritocratic factors does not go together with a weaker belief in structural factors (and vice versa), a person may simultaneously attribute success to meritocratic and structural factors (Hunt 2007;Mijs 2018a). Our findings further suggest that popular belief in meritocracy may be more easily strengthened than weakened by experiences of mobility, reflective of its dominance as a cultural narrative about inequality. This underscores the pervasiveness of the meritocratic narrative on both sides of the Atlantic, as Weiner (1985) observed more than three decades ago. The findings also attest to the difficulty of addressing inequality (Wilson et al. 2022) and contesting meritocratic justifications of success and failure (Cech and Blair-Loy 2010;Mijs 2021;Reynolds and Xian 2014).
Finally, our study also contributes to scholarly understanding of the muted public response to growing economic inequality. We find that climbing the social ladder may reinforce popular belief in meritocracy, whereas falling down the ladder does not undermine it. As such, a minimum degree of upward mobility could suffice to uphold public faith in meritocracy, dampening the political impact of the declining rate of intergenerational mobility across Western countries.

FUNDING
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by a Veni grant awarded to Jonathan Mijs

BIOS
Jonathan J.B. Mijs is an assistant professor of sociology at Boston University and Veni fellow at Erasmus University Rotterdam. He studies how people perceive, explain, and evaluate social inequality. Previous work has been published in Social Problems, Socio-Economic Review, Sociology of Education, and the Annual Review of Sociology and has been featured in The Atlantic, The Washington Post, and Financial Times.
Stijn Daenekindt holds a PhD in sociology from Ghent University. In his research, he has focused on the way people's attitudes and behaviors in various domains-for example, culture, politics, and education-are shaped by processes of social inequality. His work has been published in journals such as European Sociological Review, Journal of Marriage and Family, Poetics, and West European Politics.
Willem de Koster is a professor of general sociology, in particular cultural sociology, in the Department of Public Administration and Sociology at Erasmus University Rotterdam. He studies how social groups give meaning to social and political issues, how this informs their actions, and how it shapes their responses to new information, policies, and the (urban) environment they live in.
Jeroen van der Waal is a professor of sociology of stratification at Erasmus University Rotterdam. His research aims to explain why social stratification is linked to health outcomes and political attitudes and behaviors and has been published in journals like Epidemiology, British Journal of Sociology, European Journal of Political Research, Journal of Health and Social Behavior, and Socio-Economic Review.