Abstract
Scholars have long criticized the notion of meritocracy, in part because many achievements that are ostensibly earned stem from the intergenerational transmission of advantage. Although much research has demonstrated this link, fewer studies have considered public attitudes toward these constructs, including whether perceptions of privilege mitigate the symbolic power that educational accomplishments hold. In this article, I use data from an original, nationally representative survey experiment (N=1,800) that focuses on public perceptions of college degree holders. I find that, if anything, college graduates who are perceived as wealthy are perceived as more intelligent than they otherwise would be. Yet I also find evidence that less-privileged respondents are more likely than their more-privileged peers to convey status on those who may have faced obstacles in completing college.
Generations of social scientists have criticized the notion of meritocracy because it is imbued with inequality (McNamee and Miller 2009). Although outcomes such as status, wealth, and power purportedly reflect objective notions of merit, these outcomes often capture privileges that can be passed down through generations (see, for example, Blau and Duncan 1967; Sewell, Haller, and Portes 1969).
Educational outcomes are no exception. A degree from a highly selective college (or any college, for that matter) is considered a great accomplishment, and for many people a college degree conveys a great deal of skill and competence (Quadlin and Powell 2022). Research has shown repeatedly, however, that socioeconomic status (SES) is a better predictor of college attendance and completion than academic performance measures per se (see, for example, Dynarski 2015; Pfeffer 2008; Reardon 2011). This is true especially at highly selective colleges, where applicants often receive a premium in the admissions process for engaging in high-status activities and hobbies that are only accessible to a select few (Arcidiacono, Kinsler, and Ransom 2022; Stevens 2009), and students disproportionately come from families in the top 1 percent of the national income distribution (Aisch et al. 2017). Children raised in high-SES families have substantial advantages when it comes to education, suggesting that the more economic privilege one has, the easier it is to achieve success in education.
Despite many studies that have demonstrated this relationship between economic privilege and educational success, less scholarly research has assessed public perceptions of privilege, including the extent to which these perceptions affect the symbolic power of a college degree in the United States. Research suggests contrasting perspectives on this question. On the one hand, Americans as a whole believe deeply in the power of meritocracy and the idea that individual accomplishments reflect individual effort (Kluegel and Smith 1986; McCall 2013). When a person encounters someone with a high-status college degree, their first impression may be that this person is highly accomplished, rather than that the person may have had economic advantages that made the degree more attainable. Thus we might expect college graduates to be highly regarded in terms of their intelligence, even if (and perhaps especially if) they are perceived as economically privileged. At the same time, recent events (for example, the college admissions “scandal” involving multiple celebrity parents and their children) imply that Americans may not be surprised at the extent to which educational success can be purchased.1 Members of the public may question whether educational credentials indeed indicate a person’s intelligence—or if they, instead, are merely indicators of intergenerational wealth.
In this article, I ask whether perceptions of privilege enhance—or impede—perceptions of intelligence. In other words, when a person is perceived as economically privileged, to what extent do these perceptions affect the status that their educational credentials convey? I assess these questions using data from an original online survey experiment with a large sample (N = 1,800) of respondents who are nationally representative of the U.S. adult English-speaking population. The experiment is designed to capture how Americans think about college degrees of varying selectivity levels, and the extent to which these degrees signal intelligence, economic privilege, and likability—a construct that is often considered alongside measures of intelligence and competence, as I discuss in the sections that follow.
BACKGROUND
Educational Credentials and Americans’ Belief in Meritocracy
The American Dream is premised on the belief that anyone in the United States can achieve success as long as they work hard. This is a dominant ideology in American society—one that has simultaneously comforted and motivated generations of Americans, and one that has had an outsize influence on political discourse (Cullen 2003; Hochschild 1995). A related, but ultimately separate, construct is meritocracy, the idea that people achieve success and power on the basis of their merit, and that people who are successful were able to advance because they worked hard and proved themselves to be superior to others in one or more relevant areas (McNamee and Miller 2009). Research shows that Americans have largely bought into these belief structures and rely on them in their everyday lives. For example, in a recent iteration of the International Social Survey Programme (ISSP), approximately 95 percent of U.S. adults indicated that advancement in society is determined on the basis of hard work (Mijs 2021; see also Kluegel and Smith 1986; McCall 2013; McNamee and Miller 2009; Young 1958).2 This pattern suggests that when a person achieves success in the United States, most Americans assume that this person earned their success through hard work, superior intellect, outstanding skills, or some other indication of merit.
But despite Americans’ strong belief in the American Dream and meritocracy, much research across the social sciences has shown that intergenerational privileges—including, but not limited to, parental occupation, education, income, and wealth—have just as much (if not more) predictive power in explaining people’s outcomes relative to individual-level measures of ability or achievement (see Blau and Duncan 1967; Conwell 2021; Conwell and Ye 2021; Erikson and Goldthorpe 2002; Ermish, Jäntti, and Smeeding 2012; Hout 2018; Pfeffer and Killewald 2018; Quadlin and Conwell 2021; Sewell, Haller, and Portes 1969). Predictors of academic performance (for example, grades, test scores) and educational attainment (for example, college completion, years of education completed) have been widely studied in this regard. This is partly because schooling is compulsory in the United States up to a certain point, and thus virtually everyone has some data to report on their academic performance and educational experiences, which is convenient for quantitative studies of these relationships. At the same time, educational outcomes are highly salient to the intergenerational transmission of advantage because the grades, test scores, and credentials earned through schooling are purported to measure and reflect traits such as intelligence and competence.
Just as income and wealth are socially heritable, research shows that success in education is socially heritable. SES often is a better predictor of college attendance and completion than academic performance measures per se (Pfeffer 2008; Reardon 2011). One particularly striking analysis by Susan Dynarski (2015) in the New York Times Upshot examines rates of bachelor’s degree completion by SES and math achievement in the Education Longitudinal Study of 2002 (ELS-02).3 In these nationally representative data, high-SES students in the top quartile of math achievement had a very high likelihood of completing a bachelor’s degree (74 percent), but this outcome was far less certain for low-SES students with comparable achievement (41 percent). In fact, low-SES students in the top quartile of math achievement were equally likely to earn a bachelor’s degree as high-SES students in only the second quartile of math achievement (that is, those in the 25th to 49th percentiles; 41 percent likelihood for both groups). This contrast in educational outcomes between students at the top and bottom of the family income hierarchy clearly demonstrates the power of socioeconomic privilege in predicting educational credentials, even conditional on ability or achievement measures.
College Degrees as Signals of Privilege and/or Merit
Although social scientists have frequently pointed out the strong relationship between social origins and educational outcomes, less research has considered how the public thinks about these dynamics. In particular, I focus on perceptions of socioeconomic privilege and the extent to which these perceptions either enhance or impede the positive attributes that a college degree conveys. Although few studies speak to this topic directly, existing theory and research provide two contrasting perspectives that can help guide our thinking on this question.
The first perspective, which I call an enhancement perspective, suggests that perceptions of socioeconomic privilege further enhance the merit that a college degree conveys. In this perspective, perceptions of privilege and merit are considered additive, and a college degree-holder who is perceived as coming from a high-SES family will be perceived as demonstrating even more merit than they otherwise would. Theories of cumulative advantage (as well as empirical research in this line of work) posit that those who are raised in more-privileged families will typically, though not always (Streib 2020), become privileged adults (Blau and Duncan 1967; DiPrete and Eirich 2006; Merton 1968; Sewell, Haller, and Portes 1969). This process happens through a variety of mechanisms, including, but not limited to, educational opportunities that are restricted to privileged children starting at an early age and continuing through postsecondary education (Massey et al. 2011; Owens 2018); exposure to dominant forms of cultural capital (Bourdieu 1986; DiMaggio 1982); and a sense of confidence (or perhaps entitlement) that equips children to advocate for themselves more effectively than their less-privileged peers (Calarco 2011).
For the intergenerational transmission of advantage to occur, people would not necessarily perceive privileged individuals as any less meritorious than less-privileged individuals with the same credentials—or, alternatively, gatekeepers and those in positions of power would need to perceive privileged individuals as meritorious (a possibility that I consider in this study). Some research suggests that the most privileged students who attend elite colleges tend to secure the highest-paying post-college opportunities, in part because they have the cultural capital and other skills that appeal to elite gatekeepers, such as those who conduct hiring at top consulting firms (Binder, Davis, and Bloom 2016; Rivera 2015). This pattern generally supports the notion that privileged college students are perceived as more competent, or at least equally competent, relative to their less-privileged peers. Yet these studies were not conducted in an experimental framework, and thus more research is needed to establish the causal relationships behind perceptions of privilege and merit.
In a similar vein, insights from sociological social psychology support the idea that people work to confirm, rather than challenge, their expectations of others. A key principle of status characteristics theory is that people form expectations for others’ task performance on the basis of consensually held status beliefs. For example, women are often expected to be less effective than men in task groups, in part because women are broadly considered less competent and less worthy than men in society (Ridgeway 2011; Ridgeway and Correll 2004). Status beliefs and performance expectations work in tandem, and they also work to generate self-fulfilling prophecies (Berger, Cohen, and Zelditch 1972). When individuals are expected to perform well, they “are offered more opportunities to contribute to the group task, are more likely to accept (or create) opportunities to contribute, and their contributions are more likely to be evaluated positively and accepted by the group” (Dippong and Kalkhoff 2015, 3). Accordingly, we might expect individuals to work much harder to confirm their expectations of others’ competence than they would to refute these expectations. When they encounter a person who has achieved academic success, they may not readily challenge that person’s intelligence, even if they suspect (or know of) that person’s economic privilege, because a person can rationally be both privileged and intelligent simultaneously.
The second perspective, which I call an impediment perspective, is the idea that perceptions of socioeconomic privilege impede perceptions of merit. This perspective posits that when a person is perceived as coming from a wealthy family, others may be skeptical of that person’s individual abilities, and therefore their accomplishments are viewed in a less favorable light than they otherwise would be.
This perspective is informed largely by recent events and the sense of a possible growing backlash toward the wealthiest and most privileged Americans. For example, what the media referred to as a recent college admissions “scandal” involving numerous wealthy or celebrity parents and their children shows that families can buy their way into some of the nation’s top universities (Medina, Benner, and Taylor 2019). This reality was difficult for many members of the public to swallow because educational success, including admission to top universities, is theoretically supposed to be based on academic performance and skill—not the prominence of one’s family. These events may have led some members of the public to realize that the wealthy have outsized advantages in college admissions. Indeed, long-standing practices such as legacy preferences in admissions have institutionalized educational advantages for those who are already advantaged (Espenshade and Chung 2005)—something that many members of the public are deeply opposed to, as evidenced by recent successful pushes to ban legacy preferences at Johns Hopkins University (Castro 2020) and public colleges in Colorado (Jaschik 2021).
Recent research using public documents from the Students for Fair Admissions v. Harvard case relatedly shows that 43 percent of Harvard’s White students are recruited athletes, legacies, children of faculty and staff, or children of those who have made sizable donations.4 What is more, three-quarters of these White students would not have qualified for admission without these advantages (Arcidiacono, Kinsler, and Ransom 2022). Although it is unclear whether members of the public are familiar with the details of this case, many people would likely disapprove if such details became widely known, considering public attitudes toward legacy preferences in admissions. Given these events and others like them, it may be that some members of the public are deeply skeptical of the cognitive abilities of those who they perceive as privileged. If it is easier for high-SES students to achieve educational success, then the status tied to their success may be muted relative to those from humbler backgrounds.
Similarly, some literature questions whether rising inequality has affected people’s beliefs about meritocracy. Although economic inequality is on the rise and has been for several decades, some research suggests that Americans are not concerned about such inequalities, and that they may double down in their beliefs about meritocracy in order to convince themselves that opportunity is readily available (Bénabou and Tirole 2006; Jost, Banaji, and Nosek 2004). These studies broadly support the enhancement perspective, described earlier, because they imply that people tend to overlook matters of economic inequality in favor of meritocratic beliefs. Recent experimental research, however, calls this into question, finding that rising inequality makes respondents skeptical about the degree of economic opportunity in society and encourages support for policies that improve equity (McCall et al. 2017). If people are skeptical of economic opportunity, they may also be skeptical of those who have achieved success, especially in light of the many studies and popular accounts that have demonstrated a link between socioeconomic privilege and high-status educational credentials.
Additional Considerations: Likability as a Potential Mediating Variable, and Variation Across Respondent Education
In addition to adjudicating between these enhancement and impediment perspectives, I also assess two other questions: the extent to which perceptions of likability mediate the relationship between perceptions of privilege and intelligence, and whether the relationship between perceived privilege and perceived intelligence varies across social groups. As to the first question, perceptions of likability are important because they are intimately related to perceptions of competence. Much social psychological research has shown that perceptions of competence and likability are interwoven, such that people who are perceived as likable often tend to be perceived as competent, and vice-versa (Cuddy, Fiske, and Glick 2008; although this relationship may be weaker for targets in certain social groups, as I outline in the methods section). As discussed earlier, members of the public may be becoming increasingly knowledgeable about the advantages that wealth buys in college admissions, and research points to something of a growing backlash against the most privileged Americans (McCall et al. 2017). This evidence suggests that, to the extent that people who are perceived as privileged are perceived as less intelligent than they otherwise would be, this may be because these individuals are penalized in terms of their likability. I incorporate this idea into the experimental design by capturing perceptions of likability alongside measures of perceived intelligence and privilege, as discussed in the methods section.
Second, I consider the extent to which the enhancement or impediment perspectives predominate among certain social groups: specifically, among those who have and do not have bachelor’s degrees. College-educated individuals tend to come from higher-SES backgrounds than those who have not completed college (Reardon 2011). They also are familiar with, and serve to benefit from, the status advantages associated with college completion. Thus we might expect people with college degrees to take these educational credentials at face value more so than those with less education. Even if, and perhaps especially if, they perceive a given college degree-holder as coming from socioeconomic privilege, they may be more willing to perceive that person as having demonstrated merit. Conversely, individuals without college degrees may be skeptical of those whom they perceive as benefiting from an unequal, or even rigged, playing field. For this reason, we might expect perceptions of privilege to impede perceptions of merit, but only for those who do not have a college degree.
DATA AND METHODS
This study uses data from an original survey experiment fielded through the survey company YouGov (Quadlin 2019a). In recent years, scholars have increasingly relied on survey companies, such as YouGov, Qualtrics, and the AmeriSpeak panel through NORC, to collect high-quality data with established panels of respondents (see, for example, Doan, Quadlin, and Powell 2019; Galperin et al. 2020; Pedulla 2014; Ray 2017; Schachter 2016; Wildeman et al. 2017). The sample used in this study (N = 1,800) is nationally representative of the adult noninstitutionalized population when using survey weights, which are included in all analyses. The survey uses a conjoint experiment design, such that respondents are presented with and asked to assess two hypothetical people at the same time. Conjoint designs are effective for reducing social desirability bias (Schachter 2016), which is important in this study because I asked respondents to make judgments related to gender, race-ethnicity, and social class, which may invoke social desirability for some respondents. That said, capturing people’s genuine beliefs about race and gender remains a perennial challenge in surveys because Americans are ideologically committed to color and gender blindness (see, for example, Bonilla-Silva 2006; Risman and Ferree 1995). As a result, survey respondents may not make as big of a distinction between race and gender groups as they normally would in social life. This is an important topic that I return to in the discussion because it has implications for this research as well as survey methodology more broadly.
Experimental Design
The experiment is outlined in figure 1. After viewing an instruction screen, respondents were randomly assigned to view profiles for two recent college graduates. I chose recent college graduates as the targets because I wanted to capture the extent to which people’s undergraduate institutions and college experiences signal intelligence, privilege, and likability, and these signals are perhaps most salient among this specific population. If I were to use older targets with more work experience, their occupations or graduate institutions may have confounded the primary signals used in the study. Other scholars have examined, for example, the extent to which occupations signal prestige and other personal traits (Valentino 2020), and this question is certainly of interest to many scholars of social inequality, but is not the core focus here.
The targets’ characteristics were randomized using a 2 (gender) × 4 (race) × 6 (college selectivity) × 6 (field of study) × 6 (grades received) factorial design, with each possible combination of characteristics being represented in the data. Although gender and race-ethnicity are not the core interests in this study, I vary these characteristics nonetheless because respondents’ perceptions may well vary depending on the targets’ social groups. Gender was manipulated using gendered first names. Race was manipulated using racialized first names and/or last names. For White, Black, and Latinx targets, both the first and last names are racialized and derived from prior research (Gaddis 2017b; Quadlin 2018; Weisshaar, Chavez, and Cabello-Hutt 2020). For Asian targets, I chose names from among the most common first and last names for Chinese Americans (Bartz 2009), who are the largest Asian ethnic group in the United States. The Asian names are notable because Chinese Americans frequently have White-sounding first names, making their last names most important for signaling race-ethnicity (Crabtree and Chykina 2018). This represents a slight deviation across experimental conditions, but one that is consistent with naming patterns in the United States. In addition, all eight names are intended to be perceived as middle class. This is a key consideration especially for Black and Latinx names because, for example, if a name is readily perceived as both Black and lower class, Black targets may be penalized not only because they are perceived as Black (which studies are attempting to capture), but also because the name evokes socioeconomic disadvantage (which would be a confound from the researcher’s name selection; Gaddis 2017a).
I chose undergraduate institutions representing six levels of college selectivity, according to their ranking in U.S. News and World Report (high, moderate, and low) and their sector (private and public). Notions of selectivity are based on several factors, such as academic qualifications of the entering classes and the percent of applicants rejected, and selectivity is often considered akin to prestige or college quality in the United States (see, for example, Conwell and Quadlin 2022; Stevens 2009). The targets’ majors are either female dominated or male dominated, and are intended to span diverse content areas, that is, STEM (science, technology, engineering, math), social science, and the humanities. Finally, the targets’ academic performance spans six levels, ranging from high to low. The low achievement condition (that is, mostly Cs and Ds) is intended to be just high enough that it would be conceivable for a person to earn a college degree with these grades; any lower, and it would be unlikely that the target would have been allowed to stay enrolled and graduate. Readers who are intimately familiar with higher education may quibble about whether, for example, students can graduate from Harvard while earning mostly C and D grades. Such an academic record may be improbable, or sometimes even impossible, but most members of the public are unlikely to know about minimum grade requirements at specific institutions.
Directly beneath the two profiles, I asked respondents three questions that gauged their impressions of the targets. Specifically, I asked respondents to report their perceptions of the targets in terms of how intelligent they are, how kind they are, and the extent to which they are perceived as coming from a wealthy family.5 The first two items are drawn from research on perceptions of competence and social warmth, which are considered fundamental aspects of social perception that are often complementary (Fiske et al. 2002). Yet perceptions of competence and warmth may be in conflict when assessing members of some social groups. For example, Whites and Asians are frequently viewed as possessing high competence, but low warmth (Cuddy, Fiske, and Glick 2008). Further, research shows that women who exhibit agentic traits (for example, professional women, high-achieving women) risk being perceived as lacking social warmth more so than their men counterparts (Eagly and Carli 2007; Glick and Fiske 1996; Quadlin 2018). The third item is intended to capture perceptions of intergenerational privilege, which are theoretically distinct from, but certainly help support, achieved statuses such as educational and occupational attainment (Sewell, Haller, and Portes 1969).
After respondents completed their first set of ratings, they then repeated the task two more times. I dropped twenty-two cases with missing data on one or more outcome variables (.2 percent of the sample), leaving a final sample size of 10,778 profile ratings.
Methods
I begin by showing descriptive statistics for perceptions of intelligence, likability, and privilege, including how these perceptions vary across the main experimental manipulations. I then use linear regressions, first to assess how the main experimental manipulations are associated with perceptions of intelligence, and then to examine the extent to which perceptions of privilege and likability also factor into these perceptions. Finally, I consider whether these perceptions vary according to respondents’ educational attainment—that is, whether respondents have not attained a bachelor’s degree (n = 7,665), or whether they have attained a bachelor’s degree or more (n = 3,113). For this component of the analysis, I use both structural equation models (SEM) and linear regressions to quantify each of the pathways of interest and compare them across education groups. For all analyses, standard errors are clustered by respondent.
RESULTS
Perceptions of Intelligence, Privilege, and Warmth
Table 1 shows descriptive statistics for the three perceptions of interest in this study: intelligence, privilege, and kindness. These perceptions are shown across each of the target characteristics (that is, gender, race-ethnicity, college selectivity, field of study, academic performance), along with test statistics to indicate which groups are perceived as significantly different from each other.
Perceptions of Intelligence, Privilege, and Kindness by Target’s Characteristics
One point becomes immediately apparent: college characteristics and experiences—including college selectivity, field of study, and especially academic performance—look to be much more determinative of individual perceptions than a person’s race and gender. Broadly speaking, among targets who attended different types of colleges, majored in different fields of study, and earned different grades, respondents tended to view these groups quite differently for each of the perceptions under consideration. As an illustration of this point, consider the targets with the lowest grades (mostly Cs and Ds in college) versus those with the highest grades (mostly As). Respondents perceived the highest-achieving targets as considerably more intelligent than their lowest-achieving peers, with a 2.27-point differential between these groups (p < .001). The highest achievers were also perceived as more likable (p < .001) and as coming from wealthier families than the lowest achievers (p < .001), although the point differentials for perceived intelligence are by far the largest of the three.
Similar patterns are observed among targets who attended colleges of varying selectivity levels, and among those who majored in different fields of study. College selectivity looks to have the largest effects on perceptions of wealth. Targets who attended highly selective private universities (such as Harvard) are perceived as wealthier than those who attended less selective public universities (such as Montclair State; p < .001). The biggest disparities across fields of study are in terms of intelligence. The mean for the highest-rated field, biology (6.76), is only slightly higher than that for the lowest-rated field, psychology (6.50), although this point differential is statistically significant (p < .001). That STEM majors are rated as more intelligent than other majors underscores the growing importance of STEM fields in the national discourse, even at the same time that much of the country expresses anti-science beliefs (O’Brien and Noy 2020). This is a potential source of tension that may be a fruitful area for future research—that is, the extent to which individuals who espouse anti-science attitudes see the value of scientific careers for themselves, their children, or other members of the public, especially given the prestige and strong economic returns that are tied to many scientific careers (Kim, Tamborini, and Sakamoto 2015; VanHeuvelen and Quadlin 2021). Although some individuals may be decidedly anti-science, they may nonetheless endorse STEM majors and careers for instrumental reasons, but this question is ultimately empirical.
By comparison, race and gender are much less determinative of perceptions of intelligence, wealth, and kindness. Men and women have comparable mean scores across each of these perceptions. In addition, all four race-ethnicity groups are perceived similarly in terms of their intelligence and kindness. In an exception, both Black targets (p < .01) and Hispanic targets (p < .001) are perceived as coming from less-privileged families than their White counterparts, which is consistent with research on the demography of race, income, and wealth in the United States (Hamilton and Darity 2010). It may be surprising that gender and race-ethnicity groups are not rated more disparately, considering that research has pointed to wide gulfs in perceptions of these groups (Berger, Cohen, and Zelditch 1972; Ridgeway 2011). Part of this equalization may be due to the fact that all the targets in the experiment were described as college graduates. I suspect that this standardization of educational attainment is mitigating some of the group-based disparities that have been observed in research. That said, we might still expect larger race and gender effects in this experiment, given the large racial and gender disparities we observe in social life; this is an important topic that I return to in the discussion.
A final point is that respondents’ mean ratings of privilege typically were lower than their ratings of intelligence and kindness. Broadly, this speaks to Americans’ general distaste for the concept of privilege, even as Americans seem to be becoming more aware of how powerful it is and how it operates. As other articles in this volume mention (see, for example, Koenig 2022), the recognition of undeserved status may be one of the key mechanisms fueling the pro-Trump movement as well as larger forces of political polarization in the United States. Although many of the vignette characters could have reasonably been interpreted as quite privileged on the basis of their race and undergraduate institution, among other attributes, respondents may have been more likely to gravitate toward perceptions of intelligence and kindness than perceptions of privilege per se.
Predicting Perceptions of Intelligence
Table 2 builds on these descriptive statistics by presenting how respondent perceptions of wealth and kindness, along with the target’s characteristics, are associated with perceptions of intelligence. The first model shows the effects of target characteristics before incorporating the other respondent perceptions. Men and women are perceived as about equally intelligent, as are all four race-ethnicity groups. The effects of college selectivity also are relatively sparse. Targets who attended moderately selective private institutions (p < .05) and moderately selective public institutions (p < .05) are perceived as less intelligent than their peers who attended the highest-status private colleges, but other categories of selectivity are not significant. This finding may be surprising given the vast differences in selectivity between, for example, Harvard and the University of Wyoming; this is a point that I return to later in the analyses. As expected given what we saw in the descriptive statistics, biology majors are perceived as more intelligent than those who majored in several other fields, including psychology (p < .001), economics (p < .01), English literature (p < .001), and history (p < .001). Finally, academic performance has a significant and substantively large effect on perceived intelligence: each step increase in a target’s grades (for example, from “mostly Cs and Ds” to “mostly Cs”) is associated with a .45-point increase in perceived intelligence on a 10-point scale (p < .001), which is the largest effect here in substantive terms.
Predictors of Perceptions of Intelligence, N = 10,778
Models 2 and 3 incorporate perceptions of privilege and kindness. Here I start by discussing these respondent perceptions before turning to how the effects of target characteristics (such as college selectivity) change across models. In model 2, we see that perceptions of wealth are positively associated with perceptions of intelligence (p < .001). When a recent college graduate is perceived as coming from a wealthy family, they are perceived as more intelligent than they otherwise would be, net of their sex, race, and college information. After incorporating perceptions of kindness in model 3, the effect of perceived wealth remains significant (p < .001), although the size of the coefficient declines between models 2 and 3. This pattern suggests that perceptions of kindness and wealth are positively correlated, such that people who are perceived as coming from a wealthy family are also perceived as relatively kind. In addition, perceptions of kindness are positively associated with perceptions of intelligence (p < .001), which is consistent with research on the positive relationship between perceptions of competence and social warmth (Cuddy, Fiske, and Glick 2008).
Turning to the effects of target characteristics, many patterns are observable throughout the table, but the effects of college selectivity are perhaps most instructive. In model 1, we observe that students who attended low-status private and public institutions are perceived as about equally intelligent as those who attended high-status private institutions (the reference category). After accounting for perceptions of privilege in model 2, the coefficients for low-status private and public institutions are now significant and positive (p < .05 and p < .001, respectively). This pattern suggests that when respondents are presented with students from low-status institutions, their perceptions of intelligence are suppressed partly because they perceive these students as lacking privilege. Yet, once perceptions of privilege are controlled for, we see that these students are perceived as more intelligent than students who attended high-status private colleges. Finally, in model 3, the coefficients for low-status institutions are again not significant when controlling for perceptions of kindness. Thus, perceptions of intelligence among low-status college graduates look to be driven by perceptions of kindness. This explanation is consistent with what we saw in the descriptive statistics in table 1—that is, students who attended low-status institutions are often rated relatively low in terms of intelligence and wealth, but relatively high in terms of kindness.
Despite these variations, it is worth underscoring just how large the effects of academic performance are relative to that of college selectivity in guiding perceptions of intelligence. Figure 2 illustrates this point well. Both panels show mean perceptions of intelligence for White men targets. The top panel captures biology majors who received mostly As in college, across all six levels of college selectivity; the bottom panel captures psychology majors who received mostly Cs and Ds, across all six levels. From left to right in both of these panels, we see how perceptions of intelligence vary (modestly) according to college selectivity. Here it becomes abundantly apparent that one’s grades in college are far more determinative of perceived intelligence than where one went to college. Put differently, a degree from Harvard, which some scholars and members of the public have argued is an indicator of intergenerational wealth, will not “save” a person from being perceived as lacking intelligence if they received poor grades, at least in the context of this experiment.
Perceptions of Intelligence: Comparing Variation in Academic Performance versus College Selectivity
Source: Author’s tabulation from original data collected through YouGov (Quadlin 2019a).
Note: Figures are derived from OLS regressions. Models include survey weights.
Comparing Respondents with and Without Bachelor’s Degrees
As a final component of the analysis, I consider whether respondents with and without bachelor’s degrees make distinct assessments of recent college graduates in terms of their intelligence, privilege, and kindness. Figure 3 shows structural equation models for each of these groups, which help quantify and compare each of the relationships of interest. Perhaps surprisingly, these SEM results are virtually identical across groups. In both panels, we see positive pathways between perceptions of privilege and likability, between perceptions of likability and intelligence, and between perceptions of privilege and intelligence (all p < .001). What is more, the magnitudes of these relationships are extremely consistent across education groups, suggesting that regardless of whether one has attained a bachelor’s degree, the assessments of these traits follow similar processes. In general, when a recent college graduate is perceived as economically privileged, they are also perceived as more intelligent than they otherwise would be, and this pattern holds regardless of respondents’ level of education.6
Structural Equation Model Estimates for Perceptions of Intelligence, Likability, and Privilege
Source: Author’s tabulation from original data collected through YouGov (Quadlin 2019a).
Note: Standard errors clustered by respondent. Models include survey weights.
Table 3 provides something of a counterpoint to this finding, however. This table mirrors figure 2 by showing how perceptions of intelligence vary across levels of college selectivity for two distinct groups: White men who majored in biology and received mostly As, and White men who majored in psychology and received mostly Cs and Ds. These estimates are derived from separate models by respondent educational attainment. For those who have a bachelor’s degree or more, we see a distinct gradient in perceived intelligence across levels of college selectivity. Targets who attended highly selective private institutions are perceived as more intelligent than those who attended less selective public institutions, even though they majored in the same subject and received the same grades (p < .05). Put differently, respondents who have a bachelor’s degree or more tend to use college selectivity information to make assessments of others’ intelligence.
Perceptions of Intelligence: Comparing Variation in Academic Performance versus College Selectivity, by Respondent Education Level
This is not the case when we limit the sample to respondents who have less than a bachelor’s degree, however. For this group, perceptions of intelligence do not vary significantly across levels of college selectivity. This is partly driven by the fact that respondents with less than a bachelor’s degree rated students from less selective institutions relatively highly in terms of intelligence (a pattern evident in the contours in table 2). Further research is needed to fully unpack this mechanism, but I suspect that these respondents were especially likely to recognize the barriers that students from less selective colleges face, and they awarded them an intelligence “boost” as a result. Accordingly, it may not be that lower-SES respondents are skeptical of privileged people’s intelligence; instead, they may be more willing to recognize the intelligence that less-privileged individuals possess. This is a potentially important pathway in the assessment of individual merit that can be refined in future work.
DISCUSSION AND CONCLUSION
Using data from an original, nationally representative survey experiment, this article has considered the intertwined relationships between perceptions of intelligence and privilege. People make assessments of others’ intelligence every day, and these assessments matter for determining many outcomes that have an impact on people’s lives. Scholars of educational inequality are well aware that students who attend the nation’s most elite institutions are not necessarily more intelligent than similar others; instead, they may have gained admission to these institutions, at least in part, as a byproduct of the intergenerational transmission of advantage (Bero 2021). But what does the public make of this? Are graduates of the most elite institutions perceived as privileged and intelligent simultaneously? Or are members of the public skeptical of wealthy Americans to the point that people who are perceived as privileged are regarded as less intelligent than they otherwise would be?
The results from this study are very clear in this regard: when a recent college graduate is perceived as coming from a wealthy family, they are perceived as more intelligent than their peers who are perceived as less wealthy. In this sense, perceptions of privilege tend to enhance—rather than impede—perceptions of intelligence. This finding is consistent with research such as that in the status attainment tradition (Sewell, Haller, and Portes 1969), which finds a persistent positive relationship between one’s social class of origin and one’s life chances. For these relationships to persist on a large scale, privileged individuals would not face disproportionate barriers in their pursuit of status; instead, they would have an easier time demonstrating to others that they are competent and worthy. This pattern also is consistent with insights from sociological social psychology, which suggests that people work to confirm their impressions of others rather than to debunk their initial assumptions. This study likewise shows that people who are perceived as privileged are effectively given a vote of initial confidence when it comes to their intelligence.
I also considered the extent to which this relationship is consistent across social groups. Specifically, I assessed whether perceptions of recent college graduates are consistent for respondents with and without bachelor’s degrees. Considering that college graduates tend to come from high-SES families, and that college graduates likewise stand to benefit from the status advantages tied to postsecondary education, it would not necessarily be surprising if perceptions of privilege and intelligence were especially strong for bachelor’s degree–holders.
For the most part, I find that these groups have similar perceptions of college graduates, and that the direction and magnitude of these perceptions are consistent across groups. Yet I also find a deviation when it comes to the symbolic power tied to college selectivity. Whereas college graduates tend to differentiate between levels of college selectivity in assessing people’s intelligence, people without college degrees may not make such distinctions. Instead, in many cases, they may perceive recent college graduates as about equally intelligent regardless of the selectivity of their college. This is partly driven by the fact that nongraduates tend to rate students from the least selective institutions rather highly when assessing their intelligence. Perhaps these respondents are more likely than college graduates to recognize the barriers that these students face in attaining their degrees and thus are more generous in their assessments. An alternative possibility is that college graduates and nongraduates are motivated to either maintain or disrupt the status hierarchy, respectively, and that respondents’ perceptions are a reflection of these differential motivations. College graduates, for example, may be more likely to distinguish between levels of college selectivity as a way to bolster their status and promote themselves relative to those who attended less selective colleges. Those without college education, meanwhile, may minimize differences between levels of college selectivity as a way of flattening a prestige hierarchy they are not privy to. Overall, this may be a fruitful area for future research to help enhance our understanding of how and for whom educational credentials hold symbolic power.
Broadly, the findings from this study tell us much about the status implications of privilege and power in the United States. Despite much social scientific evidence that socioeconomic privilege enhances people’s chances of academic success, members of the public regard privilege as an indicator of merit in and of itself. Put differently, respondents generally perceive privileged individuals—or those who are least likely to have to “pull themselves up by their bootstraps” based on skill alone—as the most intelligent. These findings have implications for college admissions, hiring and promotion in professional organizations, and many other settings where perceptions of intelligence are used in decision-making. Perhaps the status tied to privilege would not be as strong if people were reminded about the many advantages that flow from income and wealth. This could be an effective intervention for reducing the linkages between privilege and intelligence—at least in an experimental setting, where respondents do not have to justify their choices to others or make decisions based on other competing interests. In the real world, however, these linkages are much more difficult to break. College admissions officers may continue to admit privileged students, even if they are reminded of the biases in doing so, because they are incentivized to admit entering classes that will benefit the university financially. Similarly, hiring decision-makers may continue to hire privileged applicants, even if they are reminded of the biases in doing so, because clients and coworkers are likely to regard these applicants positively. All of this is to say that bias is not the only mechanism that contributes to patterns of cumulative advantage, and any successful intervention will take into account both the cognitive biases and the social structures that reinforce inequality.
The data for this study come from a conjoint survey experiment using a nationally representative sample of U.S. English-speaking adults. Conjoint experiments have been shown to reduce social desirability bias in individual perceptions, which can be a concern for some types of survey experiments, especially those that ask respondents to report impressions of people from diverse racial and gender groups (for related discussion, see Quadlin 2019b). Although I have taken care to mitigate the potential for social desirability bias here, it is a perennial concern among survey experimentalists, especially given Americans’ deep-seated tendencies toward color and gender blindness. Future research could incorporate other modes of data collection, such as interviews or even surveys with open-ended components, to triangulate the perceptions reported here.
Research could also make in-depth assessments, for example, of Americans’ current narratives or theories of intelligence, privilege, and kindness. How do these narratives account for higher education credentials and experiences? Is intelligence a fixed and innate trait, or does it grow and expand when we give people the opportunity to attend an elite college—especially students from humbler backgrounds? Similarly, does college provide an environment that engenders feelings of kindness and trust in students, or does this kindness not shine through as brightly for those who attend less selective institutions where resources are scarce? I find that credentials tend to shape perceptions of intelligence, privilege, and kindness, but more work could be done to assess where these feelings come from.
Studies could also assess how perceptions of intelligence and privilege vary across other social divides, such as social class of origin and parents’ educational attainment. These are key measures of socioeconomic privilege that are likely to shape the way people think about, and differentially reward, achieved statuses such as education. Although scholars are well aware that many so-called achieved statuses are at least partly ascribed, these linkages are not well known among members of the public, which only helps reinforce existing forms of inequality.
Appendix
Predictors of Perceptions of Intelligence, by Respondent Education LevelNo Bachelor’s DegreeBachelor’s Degree or More No Bachelor’s DegreeBachelor’s Degree or More
FOOTNOTES
↵1. Although the media and some members of the public portrayed these events as a scandal, many subject-matter experts were not surprised that they took place and would not characterize them as particularly shocking.
↵2. This also looks to be the case for most Western countries, although the percentage of people who agree with this statement is largest in the United States. As Jonathan Mijs (2021) describes, using data from twenty-three countries in the 2014 ISSP, “A first thing to note is how strongly citizens, across the board, think success depends on hard work. With the exception of communist pre-1989 Poland, a majority in each country and time period believes theirs is a meritocracy society. A second thing to note is that the percentage of people who does, has gone up in almost every country since the late 1980s” (2).
↵3. The ELS-02 is a nationally representative survey conducted by the National Center for Education Statistics that captures education, family, and work experiences for a cohort of students who were high school sophomores in 2002. In Dynarski’s (2015) analysis, math achievement was measured using students’ scores on a standardized math assessment that all ELS respondents took, which is a standard measure of ability in many educational studies.
↵4. Students for Fair Admissions, Inc. v. President and Fellows of Harvard College, Students for Fair Admissions, Inc. v. University of North Carolina, et al., Docket nos. 20-1199, 21-707, U.S. Court of Appeals, 1st Circuit, Amicus Brief no. 19-2005.
↵5. Perceptions of kindness are not exactly the same thing as perceptions of social warmth, and thus this measure deviates slightly from prior research. However, I could only include one measure of social warmth in the survey, and I decided that respondents could gauge whether a person is kind in this context more easily than they could assess whether a person is warm. I ultimately expect that these measures would produce equivalent results.
↵6. A similar consistency is seen in the regression models in table A.1.
- © 2022 Russell Sage Foundation. Quadlin, Natasha. 2022. “Do Perceptions of Privilege Enhance—or Impede—Perceptions of Intelligence? Evidence from a National Survey Experiment.” RSF: The Russell Sage Foundation Journal of the Social Sciences 8(7): 48–69. DOI: 10.7758/RSF.2022.8.7.03. I thank Long Doan, Bianca Manago, Brian Powell, Vinnie Roscigno, Tom VanHeuvelen, and Elizabeth Zack for comments on previous drafts and guidance during the research process. I am also grateful to the editors of this issue, Hazel Rose Markus and Cecilia Ridgeway, for their thoughtful feedback on this work, as well as the other contributors for their suggestions. Direct correspondence to: Natasha Quadlin, at quadlin{at}soc.ucla.edu, 299 Haines Hall, 375 Portola Plaza, Los Angeles, CA 90095, United States.
Open Access Policy: RSF: The Russell Sage Foundation Journal of the Social Sciences is an open access journal. This article is published under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
REFERENCES
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.