A Privileged Point of View: Effects of Subjective Socioeconomic Status on Naïve Realism and Political Division

Abstract

In the United States, both economic inequality and political conflict are on the rise. We investigated whether subjective socioeconomic status (SSS) may help explain why these dual patterns emerge. We hypothesized that higher SSS may increase naïve realism—the belief that one perceives the world as it is, rather than as interpreted through one’s own knowledge and beliefs—regarding political issues. Using a representative sample of the American electorate, we found that higher SSS predicted more political naïve realism toward those from a different political party (Study 1). The remaining experiments examined the causal relationship between SSS and political naïve realism (Studies 2–5). We extended these findings by investigating whether SSS influenced participants’ willingness to exclude those with contrary views from a vote (Studies 4 and 5). Together, these studies demonstrate that SSS enhances political naïve realism and can lead to the exclusion of others with contrary opinions.

Keywords

socioeconomic status naïve realism social cognition attitudes political psychology

Economic inequality has increased in the United States to historically high levels (Saez & Zucman, 2014). In 1963, the average income among the richest 10% of Americans was approximately 7 times the average income of the poorest 10%. By 2016, that gap grew to a factor of thirteen (Smith, 2017). Political disagreement is also growing. Since 1994, both elected officials and citizens have become more divided along party lines, less willing to compromise, and more likely to view the opposition not merely as mistaken, but as a threat to the nation (Pew Research Center, 2014, 2016). Recent research suggests that these two trends are related: More inequality may cause more political polarization (Voorheis et al., 2015).

Why might socioeconomic status (SES) differences create political polarization? Much research indicates that people have a range of self-serving biases (Kunda, 1990; Miller & Ross, 1975). Consistent with this self-interest perspective, wealthier individuals are more likely than poorer individuals to support laissez-faire economic policies and oppose redistribution policies (Gelman et al., 2010). Even being made to feel momentarily higher in SES can cause people to endorse more economically conservative opinions (Brown-Iannuzzi et al., 2015). Thus, as the distance between the wealthy and the poor grows, their difference of opinion on economic matters is also likely to grow.

Differences of opinion alone, however, are not sufficient to explain the increased animosity and decreased willingness to compromise. In this article, we suggest that a second factor is also at work: People who occupy higher positions in the economic hierarchy may experience greater naïve realism. Naïve realism is the assumption people have that they see the world objectively, without recognizing that their perception of the world may be shaped by sensory processes, desires, and prior knowledge (see Ross & Ward, 1995, 1996). Wealthier people may attribute their higher economic standing to themselves—creating a sense of competence and expertise (e.g., Butler, 2014; Lerner, 1965). As a result, wealthier people may believe their opinions—particularly political opinions related to economic issues—are objectively correct, thus fueling the experience of political naïve realism. This stance of political naïve realism may in turn lead wealthier individuals to feel more entitled to maintain uncompromising positions and to react unfavorably toward those who disagree with them.

Critically, the enhanced stance of political naïve realism may be due to feeling higher status, as opposed to having more financial resources. This is because financial resources do not fully capture the social comparisons people make, and thus may be inconsistently related to where people place themselves in the socioeconomic hierarchy and, in turn, how competent they feel relative to others. As a result, people who feel higher status may believe their political opinions are objectively correct because they feel smarter than others. The current research, therefore, investigates whether feeling higher status increases perceptions of competence and one’s stance of political naïve realism. Furthermore, we investigate whether this enhanced stance of political naïve realism leads to political conflict. In the following sections, we review each of these ideas, in turn.

Subjective Socioeconomic Status Enhances Perceived Competence

One of the most cross-culturally prevalent hierarchies is based on socioeconomic resources. However, because people rarely know their objective socioeconomic worth, let alone the objective worth of others, they often make subjective comparisons to determine their SES and the SES of others—which we call subjective socioeconomic status (SSS; e.g., Brown-Iannuzzi et al., 2015). A number of potentially related factors are relevant to SSS, including one’s economic resources (income and wealth; e.g., Adler et al., 2000), sociocultural experiences (e.g., Stephens et al., 2014), and social comparisons (e.g., Brown-Iannuzzi et al., 2015; Kraus, 2018). This research focuses on the social comparative aspects of SSS.

At any given level of objective resources, people may feel relatively higher or lower status depending on their current social comparisons (e.g., Brown-Iannuzzi et al., 2015; Piff et al., 2010). Social comparisons based on SES may be particularly important for determining our abilities relative to others (e.g., Zell & Alicke, 2010; see also Festinger, 1954). This is because people tend to justify status hierarchies by accepting the stereotype that higher status groups are more competent than lower status groups (e.g., Fast & Chen, 2009; Fiske et al., 2002; Jost & Burgess, 2000). In addition, when people are successful, such as when they feel higher status, they tend to believe they are responsible for their success (e.g., Miller & Ross, 1975). Together, these findings suggest that higher SSS may signal to a person that they are highly competent.

Due to these cultural narratives of attributing success to the person, it may be that people make these internal attributions regardless of how high SSS is obtained. Lending evidence to this possibility, one study asked participants to complete logic problems and were told that their payment—a relatively larger versus smaller amount—was determined randomly (Butler, 2014). After participants completed this task and received their payment, they were asked to report on their ability to reason and their work ethic relative to others. Despite knowing payment was randomly assigned, participants who were highly paid rated themselves as better reasoners and harder workers than their lower paid counterparts (see also Lerner, 1965). These findings suggest that even when status is randomly assigned, people may attribute their standing to internal factors such as competence. A heightened sense of perceived competence may, in turn, fuel the experience of political naïve realism.

SSS Enhances Political Naïve Realism

Bias refers to a systematic deviation in judgments from a normative standard (e.g., Kahneman et al., 1982). Although there are many psychological biases that can contribute to political decisions, we focus here on one type of bias—naïve realism—because a consequence of this bias is interpersonal conflict (Kennedy & Pronin, 2008).

Naïve realism is the fundamental assumption that one sees the world objectively, without recognizing the power of our own sensory processes, beliefs, and knowledge to shape our perceptions (see Ross & Ward, 1995, 1996). Naïve realism is comprised of three interrelated principles: (1) We see the world objectively and our resulting attitudes are, therefore, unbiased (Frantz, 2006; Pronin et al., 2002); (2) if others are provided with the same (presumed) objective information, their resulting attitudes will be the same as ours (e.g., Gilovich, 1990; Ross et al., 1977); and (3) the failure of others to share our views is assumed to be because (a) they were not exposed to the same information, (b) they must be lacking cognitively, or (c) they are biased by distorting personal influences (for reviews see Pronin, 2008; Pronin et al., 2004). The current research focuses on the third principle as it is most related to political conflict.

Higher SSS individuals might be more likely to exhibit political naïve realism than lower SSS individuals. Because higher SSS individuals may attribute success to their own competence (e.g., Lerner, 1965), they may be more likely to assume their opinions—particularly opinions related to financial decisions—are objectively correct, thus fueling one’s experience of naïve realism (see also Kahan et al., 2017). Critically, based on the third principle of naïve realism (e.g., Pronin, 2008), the mere existence of differences in opinions can lead to biased evaluations of those who hold opposing beliefs.

Naïve realism resulting from exposure to differences in opinions may take many forms. For example, participants who were made aware of political differences, real or imagined, viewed others with opposing political views less favorably than did people who were unaware of political differences (Keltner & Robinson, 1993). Individuals also tend to exaggerate the difference between their own views and the views of their opponents (Keltner & Robinson, 1997). Furthermore, people tend to believe others who disagree with them have biased attitudes and will behave in a biased, self-serving manner (Robinson & Keltner, 1996; Yair & Sulitzeanu-Kenan, 2015). Thus, these biased judgments may take the form of expressing negativity toward the disagreeing other and/or an asymmetry in the soundness of one’s own attitude relative to the soundness of the disagreeing other’s attitude.

Importantly, because SSS shapes ideological beliefs and political attitudes (for a review see Brown-Iannuzzi et al., 2017), we anticipate that higher SSS may fuel political naïve realism as this domain is often related to financial and economic decisions.

Political Naïve Realism May Contribute to Political Conflict

The failure of others to share our views can lead to interpersonal conflict. For example, perceiving oneself as more objective relative to the “biased” opposition is associated with taking longer to allocate hypothetical funds in a negotiation task (Keltner & Robinson, 1993), underweighting a partner’s input on subsequent tasks (Liberman et al., 2012), and choosing to compete rather than cooperate (Kennedy & Pronin, 2008). Therefore, naïve realism can lead to justifying one’s world view by denigrating contradictory views as biased, which can ultimately lead to interpersonal or intergroup conflict (Ross & Ward, 1995).

To the extent that higher SSS individuals are more likely to experience political naïve realism, they may also be more likely to engage in political conflict. Although we are unaware of any published research that directly tests the link between SSS and politically divisive behavior, our own previous work is suggestive (Brown-Iannuzzi et al., 2015). We found that manipulating people’s perceptions of their SSS shaped their attitudes toward economic redistribution: Participants who momentarily felt higher SSS were less supportive of redistributive policies than were people who felt momentarily lower SSS. Furthermore, we investigated how participants might react to other’s redistribution recommendations (Study 4). Higher SSS individuals thought that others with contrary opinions about redistribution were more biased than those who agreed with them, but lower SSS individuals did not assume that those who disagreed were more biased than those who agreed. The research on naïve realism reviewed above can explain why people would assume others who disagree with them are biased. It does not, however, explain why it occurred only among higher SSS individuals. If high- (vs. low-) SSS individuals experience higher levels of political naïve realism, they may be more likely to react negatively toward and less likely to seek compromise with those who disagree with them.

Overview of the Current Research

The current research investigated whether SSS influenced perceived competence, political naïve realism, and political conflict. Using a representative sample of the American electorate, Study 1 investigated whether SSS was associated with more naïve realism toward members of the opposing political party. After establishing the relationship between SSS and political naïve realism using an ecologically valid sample, we turned to highly controlled experiments to understand causality. Studies 2 and 3 investigated whether SSS influenced one’s own perceived competence and political naïve realism. In Study 3, however, one’s SSS was described as being a result of random chance, as opposed to a mixture of skill and chance. Thus, we investigated whether regardless of how high-SSS is obtained, people would attribute success to themselves. Studies 4 and 5 extended the findings by investigating whether high-SSS influenced participants’ willingness to exclude those with contrary views from a vote.

Naïve realism toward another who disagrees with one’s opinion is often measured across three dimensions: uninformed/incompetent, irrational, and biased (Ross & Ward, 1995, 1996; for reviews see Pronin, 2007, 2008; Pronin et al., 2004). We have no hypotheses whether SSS will influence all dimensions of naïve realism equally. In fact, there may be reason to suspect a trade-off between dimensions. That is, the disagreeing other may be perceived as incompetent or biased or self-interested. But, if the disagreeing other is deemed incompetent, then there may be no reason to infer the other is biased as well. Thus, effects on individual items may be inconsistent. As a result, we remain agnostic whether SSS will influence all dimensions of political naïve realism.¹ Finally, naïve realism can be conceptualized as the asymmetry between how someone sees themselves versus the disagreeing other on a given dimension of naïve realism. Thus, when we ask participants to report about themselves and the disagreeing other on any of these dimensions of political naïve realism, we also calculate and analyze the self–other asymmetry.

Across all studies, sample sizes were determined a priori based on obtaining adequate power (1 − β ≥ 80) to detect a small-to-medium effect size (d = .35; G*Power v.3.1.9.4, Faul et al., 2009). We also report sensitivity analyses for each study (Faul et al., 2009). We report all relevant measures, manipulations, and exclusions, if any.² Across all analyses, when the assumption of equal variances was violated, degrees of freedom were adjusted to provide a more conservative test of our hypotheses.

Study 1

Method

We utilized data from the American National Election Survey (ANES) 2016 Time Series Study. This survey included open-ended responses asking respondents what they disliked about both major political parties and the parties’ presidential candidates. We coded these responses for words representing naïve realism. Then, we predicted naïve realism from the respondent’s SSS, political ideology, and the interaction between SSS and political ideology. We hypothesized that higher SSS participants would display more naïve realism in their open-ended comments about the opposing party.

Sensitivity analyses and respondents

In total, 4,271 respondents completed the ANES 2016 Time Series Study. Of those, 2,374 completed our measures of interest. Sensitivity analyses revealed that given our sample size, we would be able to detect a small effect size (f ² = 0.003) with adequate power (1 − β = .80) using a two-tailed test. For comprehensive details on how respondents were selected for the survey, please see the ANES website (https://electionstudies.org/). To ensure results are representative of the American electorate, we utilized the pre-election sample weights calculated by ANES.

Measures

SSS

Participants rated their SSS from “lower class” to “upper class.” Participants were randomly assigned to complete either a 6-point or 8-point scale. The 8-point scale included these response options: 1 = lower class, 2 = lower working class, 3 = working class, 4 = upper working class, 5 = lower middle class, 6 = middle class, 7 = upper middle class, and 8 = upper class. The 6-point scale included these response options: 1 = lower class, 2 = working class, 3 = lower middle class, 4 = middle class, 5 = upper middle class, and 6 = upper class. Because participants did not complete the same version of the SSS question, we standardized scores for each version separately and then merged them into a single variable.

Political ideology

Two variables assessed political ideology: (a) conservative versus liberal leaning (1 = extremely liberal; 7 = extremely conservative), and (b) political party affiliation (1 = strongly Democrat; 7 = strongly Republican). These variables were averaged together (r = .70, p < .001) to create one index of political ideology, with higher numbers indicating more conservative ideology. This index was standardized prior to analysis.

Political naïve realism

Although respondents were asked to complete several open-ended responses, a priori we identified four questions which could be used to assess naïve realism. These were as follows: “What does the respondent dislike about the (1) Democratic Party, (2) Democratic Presidential nominee, (3) Republican Party, and (4) Republican Presidential nominee?” The ANES separates responses to these open-ended questions from the rest of the data. The only identifiable information in the open-ended response data files is assigned participant ID number, used for linking data sets. Because there were many spelling errors, research assistants who were blind to the hypothesis corrected these prior to analyzing the open-ended responses and prior to linking the open-ended response data with the rest of the ANES data.

To measure political naïve realism in the open-ended responses, we coded responses using the Linguistic Inquiry and Word Count (LIWC) 2015 program (Pennebaker et al., 2015). We created a new “dictionary” to represent words related to three dimensions of naïve realism: incompetent/uninformed, irrational, and biased (Ross & Ward, 1995, 1996; for reviews see Pronin, 2008). Prior to analyzing the open-ended responses, the research team chose the words using a thesaurus and consensus among the research team. Words chosen to represent incompetent/uninformed were ignorant, stupid, dumb, uneducated, not smart, and uninformed. The words chosen to represent irrational were crazy, extreme, insane, nut/nuts, ridiculous, and irrational. The words chosen to represent biased were dishonest/not honest, lie/lies/lying/liar/liars, lean/leans/leaning, con artist/con man/con woman, crooked, corrupt, shady, sneaky, no integrity/lacking integrity, not trustworthy/untrustworthy, and bias/biased. For each participant, the LIWC analysis coded each of their open-ended responses and, for each response, produced the percentage of words related to each dimension of naïve realism. After responses were coded for political naïve realism, the data file was merged with the full ANES file that contained SSS and political ideology measures.

To create a variable representing political naïve realism toward each party, we averaged these three dimension scores together across responses to the party and the party nominee. Because these indices of naïve realism were positively skewed (skewness = 3.55 and 8.88, respectively), we log transformed each. Then, we created a difference score (naïve realism toward the Democratic Party and nominee − naïve realism toward the Republican Party and nominee), which reflects the amount of naïve realism toward Democrats relative to Republicans.³

Objective SES

To measure objective SES, we standardized income (1 = under $5,000 annually; 28 = $250,000 or more) and education (1 = less than first grade; 16 = doctorate degree). Then, we averaged these scores together to create one index of SES where higher numbers reflect higher SES.

Results and Discussion

To investigate whether SSS was associated with higher levels of naïve realism, we regressed the naïve realism difference score onto SSS, political ideology, and their interaction. Because we hypothesized the enhanced stance of political naïve realism may be due to feeling higher status, as opposed to having more financial resources, we also controlled for SES.⁴

As expected, the association between ideology and the naïve realism difference score was significant and positive, B = 0.110, SE = 0.01, p < .001, 95% CI [0.09, 0.13] (see Table 1). This finding reflects the tendency to describe the opposing political party with greater political naïve realism.

Table 1.

ANES Survey: Results of Multiple Linear Regression Analysis Predicting Naïve Realism.

Predictors	B	SE	p	95% CI
Constant	0.14	0.01	<.001	[0.12, 0.15]
SES	−0.04	0.01	.002	[−0.06, −0.02]
SSS	−0.01	0.01	.419	[−0.03, 0.01]
Political ideology	0.11	0.01	<.001	[0.09, 0.13]
SSS × Ideology	0.03	0.01	<.001	[0.01, 0.05]

Note. ANES = American National Election Survey; CI = confidence interval; SES = socioeconomic status; SSS = subjective socioeconomic status.

The main hypothesis, that political naïve realism would be greater for higher SSS individuals, was tested by the interaction between political ideology and SSS. Consistent with our hypothesis, the interaction between SSS and political ideology significantly predicted political naïve realism, B = 0.03, SE = 0.01, p < .001, 95% CI [0.01, 0.05].

We probed the interaction in two ways (Preacher et al., 2006; see Figure 1 and Table 2). First, when SSS was the moderator, we estimated separate regression lines at low (−1 SD), moderate, and high (+1 SD) levels of SSS. All simple slopes were positive and significantly different from zero. Critically, higher SSS individuals exhibited the most political naïve realism toward the opposition. Second, when political ideology was the moderator, we estimated separate regression lines when participants were politically more liberal (−1 SD), moderate, and more conservative (+1 SD). Among more politically liberal participants, higher SSS individuals exhibited significantly more naïve realism toward Republicans relative to Democrats. Conversely, among more politically conservative participants, higher SSS individuals exhibited more naïve realism toward Democrats relative to Republicans.

Figure 1.

Top graph: The interaction between political ideology and SSS on the naïve realism difference score. The three lines represent three levels of SSS: 1 standard deviation below the mean, 1 standard deviation above the mean, and at the mean of SSS. Bottom graph: The interaction between political ideology and SSS on the naïve realism difference score. The three lines represent three levels of political ideology: 1 standard deviation below the mean, 1 standard deviation above the mean, and at the mean of political ideology. SSS = subjective socioeconomic status.

Table 2.

ANES Survey: Simple Slopes Predicting Naïve Realism.

Moderator	B	SE	t	p
Political ideology predicting naïve realism at
Lower SSS (−1 SD)	0.08	0.01	5.26	<.001
M	0.11	0.01	11.71	<.001
Higher SSS (+1 SD)	0.14	0.01	12.26	<.001
SSS predicting naïve realism at
More liberal (−1 SD)	−0.04	0.01	−2.83	.005
M	−0.01	0.01	−0.81	.420
More conservative (+1 SD)	0.02	0.01	1.86	.062

Note. ANES = American National Election Survey; SSS = subjective socioeconomic status.

We also investigated the interactive effect between SSS and political ideology among each dimension of political naïve realism. The interaction only reached traditional levels of statistical significance when predicting perceived bias, B = 0.05, SE = 0.01, p < .001, 95% CI: [0.02, 0.07] (see Table 3). We probed the interaction by estimating separate regression lines at low (−1 SD), moderate, and high (+1 SD) levels of SSS. Again, all simple slopes were positive and significantly different from zero. Importantly, higher SSS individuals thought their political opposition was most biased.

Table 3.

ANES Survey: Results of Multiple Linear Regression Analysis Predicting Perceived Bias.

Predictors	B	SE	t	p
Constant	0.24	0.01	19.54	<.001
SES	−0.05	0.02	−3.10	.002
SSS	−0.02	0.01	−1.25	.23
Political ideology	0.16	0.01	12.54	<.001
SSS × Ideology	0.05	0.01	3.75	<.001
Simple slopes
Lower SSS (−1 SD)	0.11	0.02	5.63	<.001
M	0.16	0.01	12.52	<.001
Higher SSS (+1 SD)	0.20	0.02	13.07	<.001

Note. ANES = American National Election Survey; SES = socioeconomic status; SSS = subjective socioeconomic status.

The results from this nationally representative sample provide evidence consistent with our hypothesis: Higher SSS individuals express more naïve realism toward the opposing political party and its presidential nominee. Advantageously, these data are generalizable to a representative sample of the American electorate and to attitudes toward political parties. Disadvantageously, these data cannot determine causality.⁵ Thus, for the remaining studies we turn to a highly controlled experimental approach using an investment task.

Study 2

Method

Sensitivity analyses and participants

Participants (N = 265) were recruited through Amazon MTurk. Thirteen participants failed the attention check and were excluded from the analyses. The final sample was 252 participants (147 males, 105 females). Sensitivity analyses revealed that given our sample size we would be able to detect a small-to-medium effect size (d = 0.35) with adequate power (1 − β = .80) using a two-tailed test. The average age of the sample was 35.62 years (SD = 11.69), the median income was US$40,000 to US$49,999 (range from less than US$5,000 to more than US$175,000 annually), and the median education was a 4-year college degree.

Procedure

To manipulate SSS, participants were asked to play an investment game (see Brown-Iannuzzi et al., 2015, Study 3). In this game participants were given US$0.40 of seed money to “invest” among six investment options. These options were portrayed as companies (named Company A through Company F) and were based on real corporations. A brief paragraph described each business, the company’s past stock performance, and the company’s price-to-earnings ratio. Participants were given a guide to interpret the stock and a price-to-earnings ratio. Participants were told that outcomes in the game depended on sound decision-making and chance fluctuations in the market. That is, success was described as resulting from an unspecified combination of skill and chance. The game was also described as having a system of redistribution. Participants were told that the top third of all earners would be assessed 20% of their assets to offset the losses of the bottom third of earners, while the bottom third of all earners would be given a credit equal to 20% of their assets. After participants invested their money, the computer ostensibly simulated the next 6 months of stock-market activity to determine their outcomes.

After participants invested their seed money, they were randomly assigned to one of two conditions. In the high-SSS condition, participants were told they had performed “better than 89% of all players to date.” In the low-SSS condition, participants were told they had performed “worse than 89% of all players to date.” In both conditions, participants earned US$0.12 from their investment decisions; thus, objective outcomes were the same across conditions.

After learning their outcomes in the game, participants answered four questions related to their perceived competence in the game. These questions were, “Relative to other players, how talented [skilled, competent, good] are you at investing?” (e.g., 1 = much less talented than others, 7 = much more talented than others). These items were averaged together to create one index of perceived competence in the game (Cronbach’s α = .98).

Next, participants were given the opportunity to recommend assessment and credit rates for future iterations of the game.⁶ They were instructed that the policy recommendations would not affect their current earnings.

Finally, participants were asked to imagine another person who disagreed with their recommended redistribution rates. As a measure of naïve realism, participants were asked to determine why this other person may disagree with their recommendations. These questions were as follows: (a) “To what extent do you think the person who disagrees with you is biased?” (b) “To what extent do you think the person who disagrees with you is self-interested?” (c) “To what extent do you think the person who disagrees with you is smart?” and (d) “To what extent do you think the person who disagrees with you is competent?” (1 = not at all, 5 = extremely). Responses to all four items were averaged into an index of naïve realism with higher numbers indicating more naïve realism (Cronbach’s α = .59). In addition, we created a perceived bias dimension by averaging together the first two items (Spearman–Brown = .53), and a perceived competence of the disagreeing other dimension by averaging together the second two items (Spearman–Brown = .88).

Participants reported other demographic information including political party affiliation and political ideology as descriptive data.

Results and Discussion

We hypothesized that high-SSS individuals would attribute their success to their competence more than low-SSS individuals. In addition, we hypothesized that high-SSS individuals would report more political naïve realism toward the disagreeing other than low-SSS individuals. Finally, we hypothesized that high- (vs. low-) SSS individuals would have a larger self–other asymmetry in perceived competence. For comprehensive results, see Table 4.

Table 4.

Study 2: Means, Standard Deviations, Test Statistics, 95% Confidence Intervals, and Effect Sizes for Dependent Variables of Interest.

Dependent variable	High-SSS M (SD)	Low-SSS M (SD)	t	p	95% CI_{Mean Dif}	d
Perceived competence	4.68 (1.39)	2.79 (1.64)	9.80	<.001	[1.51, 2.27]	1.24
Political naïve realism	3.40 (0.75)	3.02 (0.65)	4.35	<.001	[0.21, 0.56]	0.55
Bias dimension	3.51 (1.01)	3.15 (0.91)	2.95	.004	[0.12, 0.60]	0.37
Competence dimension	2.71 (0.90)	3.12 (0.89)	−3.60	<.001	[−0.63, −0.19]	0.45
Competence asymmetry (self–other)	0.72 (1.45)	−0.79 (1.43)	8.27	<.001	[1.14, 1.86]	1.05

Note. SSS = subjective socioeconomic status; CI = confidence interval.

As predicted, high-SSS participants believed they were significantly more competent in the game than did low-SSS participants. In addition, participants in the high-SSS condition displayed significantly more political naïve realism than did participants in the low-SSS condition. We further investigated whether SSS condition would influence both dimensions of naïve realism: perceived bias and competence of the disagreeing other. Results revealed that high-SSS participants thought the disagreeing other was more biased and less competent than low-SSS participants.

Finally, we hypothesized that high- (vs. low-) SSS participants would have a larger difference in perceived self-competence relative to perceived competence of the disagreeing other. To calculate this difference, we standardized perceived competence of the disagreeing other and subtracted this score from the standardized perceived competence of oneself. High-SSS participants had a larger difference in perceived self–other competence than low-SSS participants. Together, these findings supported our hypotheses.

Attributing success to the self, however, may be due to the specific instructions of the game. Similar to the real-world, the rules of the game suggest that outcomes are a mixture of skilled investing and chance. However, given previous evidence that people make this inference even in games of chance, we elected to examine the strength of this effect. Therefore, to provide a conservative test of our hypothesis, in Study 3 participants were told that outcomes were determined entirely by random chance. One of the important implications of naïve realism is that it can lead to conflict in people's behavior. In the final two studies, we investigate politically divisive behavior.

Study 3

Method

Sensitivity analyses and participants

The sample was 523 participants (247 males, 266 females, 5 non-binary/third gender, 3 preferred not to say, 2 did not report), recruited through Amazon MTurk. Sensitivity analyses revealed that given our sample size, we would be able to detect a small-to-medium effect size (d = 0.25) with adequate power (1 − β = .80) using a two-tailed test. The average age of the sample was 35.64 years (SD = 11.27), the median income was US$50,000 to US$59,999 (range from less than US$5,000 to more than US$175,000 annually), and the median education was a 2-year college degree.

Procedure

Participants were again asked to play an investment game, except this time outcomes in the game were described as random chance. In particular, participants were told at three separate points in the instructions: “Your outcomes will be determined by random chance, similar to a coin-flip” (bolded in the instructions). In addition, when participants learned about their investment results they were told, “Based on random chance, you did better/worse than 89% of all players to date.” Then, participants completed the same perceived skill items as in Study 2 (Cronbach’s α = .98). Then, participants recommended redistribution rates for a future version of the game. Finally, participants were again asked to imagine another person who disagreed with their recommended redistribution rates. To assess naïve realism, we asked, “To what extent do you think the person who disagrees with you is biased?” (1 = completely unbiased, 6 = completely biased), and “To what extent do you think the person who disagrees with you is self-interested?” (1 = completely not self-interested, 6 = completely self-interested), “To what extent would you think this person understands or does not understand the investment game?” (1 = completely does not understand, 6 = completely understands), and “To what extent would you think that this person was fully paying attention or not paying attention to the game and the questions?” (1 = completely inattentive, 6 = completely attentive; Cronbach’s α = .46). In addition, we created a perceived bias dimension by averaging together the first two items (Spearman–Brown = .60), and a perceived competence of the disagreeing other dimension by averaging together the second two items (Spearman–Brown = .72).

Participants reported other demographic information including political party affiliation and political ideology as descriptive data.

Results and Discussion

We again hypothesized that high- (vs. low-) SSS individuals would attribute their success to their competence, would report more political naïve realism, and would have a larger self–other asymmetry in perceived competence. For comprehensive results, see Table 5.

Table 5.

Study 3: Means, Standard Deviations, Test Statistics, 95% Confidence Intervals, and Effect Sizes for Dependent Variables of Interest.

Predictors	High-SSS M (SD)	Low-SSS M (SD)	t	p	95% CI_{Mean Dif}	d
Perceived competence	4.85 (1.31)	2.66 (1.63)	16.89	<.001	[1.93, 2.44]	1.48
Political naïve realism	3.65 (0.79)	3.52 (0.68)	1.93	.054	[−0.002, 0.25]	0.17
Bias dimension	4.34 (0.94)	4.28 (0.97)	0.73	.467	[−0.10, 0.22]	0.06
Competence dimension	4.04 (1.18)	4.23 (1.03)	−1.96	.050	[−0.38, 0.0002]	0.17
Competence asymmetry (self–other)	0.67 (1.31)	−0.69 (1.25)	12.16	<.001	[1.14, 1.59]	1.07

Note. SSS = subjective socioeconomic status; CI = confidence interval.

Similar to Study 2, high-SSS participants believed they were significantly more competent in the game than did low-SSS participants. In addition, participants in the high-SSS condition displayed more political naïve realism than did participants in the low-SSS condition. Again, we further investigated whether SSS condition would influence both dimensions of naïve realism. There was no difference in perceived bias by SSS condition. However, similar to Study 2, participants in the high-SSS condition thought the disagreeing other was less competent than participants in the low-SSS condition.

Finally, we investigated whether high- (vs. low-) SSS participants would have a larger difference in perceived self-competence relative to perceived competence of the disagreeing other. Similar to Study 2, high-SSS participants had a larger difference in perceived self–other competence than low-SSS participants. Overall, these findings replicate Study 2. Critically, these patterns of results emerged even when participants were told that outcomes were determined entirely by random chance.

One of the important implications of naïve realism is that it can lead to conflict in people’s behavior. In the final two studies, we investigate politically divisive behavior.

Study 4

Method

Sensitivity analyses and participants

Participants (N = 437) were recruited through Amazon MTurk. Twenty-three participants failed the attention check and were excluded from the analyses. The final sample was 414 participants (209 males, 204 females, 1 did not report). Sensitivity analyses revealed that given our sample size, we would be able to detect a small-to-medium effect size (d = 0.28) with adequate power (1 − β = .80) using a two-tailed test. The average age of the sample was 40.24 years (SD = 12.19), the median income was US$50,000 to US$64,999 (range from less than US$35,000 to more than US$125,000 annually), and the median education was a 4-year college degree.

Procedure

This study utilized the same procedure from Study 2. After receiving their investment outcomes, participants were asked the following political naïve realism questions about themselves: (a) “To what extent do you think your recommendations are biased?” (1 = not at all biased, 5 = extremely biased), (b) “To what extent do you think your recommendations are based on self-interest?” (1 = not at all self-interested, 5 = extremely self-interested), (c) “Relative to other players, how competent are you?” (1 = much less competent than others; 7 = much more competent than others), and (d) “Relative to other players, how smart are you?” (1 = much less smart than others; 7 = much smarter than others; α = .09). The low internal consistency is consistent with the expected trade-off between dimensions. Perceiving opponents as either biased or incompetent will accomplish the goal of dismissing their views. Because of the low internal consistency, we again computed separate subscales. We created a perceived bias dimension by averaging together the first two items (Spearman–Brown = .77), and a perceived competence of the disagreeing other dimension by averaging together the second two items (Spearman–Brown = .84).

Then, participants were asked to imagine another person who disagreed with their redistribution recommendations. Keeping this other person in mind, participants were asked the same political naïve realism questions as above (e.g., “To what extent do you think the person who disagrees with you is biased?” α = .35). Again, we created a perceived bias dimension by averaging together the first two items (Spearman–Brown = .74), and a perceived competence of the disagreeing other dimension by averaging together the second two items (Spearman–Brown = .90).

Finally, for the weighted vote participants were told,

Each participant who plays the game has the opportunity to recommend future assessment and credit rules. We will take all of the recommendations from every participant (∼800 total) and create an average assessment and an average credit recommendation to serve as our rules in the next version of the game. However, some participants give good recommendations whereas others give poor recommendations. Because we want to make sure the next version of the game is better, we are interested in whether recommendations of other participants should be given more or less weight in the average. Remember, the weighted average will serve as the assessment and credit rules in the next version of the investment game. More weight in the average means the next version of the game will more closely reflect this participant’s recommendations. Less weight in the average means the next version of the game will not closely reflect this participant’s recommendations.

Then participants are asked, “How much weight should be given to the recommendations of the person who disagrees with your recommendations?” (1 = significantly less weight; 9 = significantly more weight).

Participants reported other demographic information including political party affiliation and political ideology as descriptive data.

Results

Political naïve realism

For comprehensive results, see Table 6.

Table 6.

Study 4: Naïve Realism Means and Standard Deviations for High- and Low-SSS Participants.

Predictors	High-SSS M (SD)	Low-SSS M (SD)	t	p	95% CI_{Mean Dif}	d
Political naïve realism toward the self	−0.10 (0.51)	0.10 (0.51)	−4.09	<.001	[−0.30, −0.11]	0.40
Bias dimension (self)	2.81 (1.28)	2.70 (1.27)	0.82	.412	[−0.14, 0.35]	0.08
Competence dimension (self)	4.99 (1.09)	4.29 (1.45)	5.59	<.001	[0.46, 0.95]	0.55
Political naïve realism toward the other	0.07 (0.59)	−0.07 (0.56)	2.46	.014	[0.03, 0.25]	0.24
Bias dimension (other)	3.31 (1.08)	3.26 (1.02)	0.48	.634	[−0.15, 0.25]	0.05
Competence dimension (other)	4.13 (1.13)	4.41 (1.11)	−2.54	.012	[−0.50, −0.06]	0.25
Naïve realism asymmetry (other–self)	0.17 (0.76)	−0.17 (0.72)	4.74	<.001	[0.20, 0.49]	0.47
Bias dimension asymmetry (other–self)	0.51 (1.16)	0.56 (1.25)	−0.45	.653	[−0.29, 0.18]	0.04
Competence dimension asymmetry (self–other)	0.86 (1.38)	−0.12 (1.30)	7.47	<.001	[0.72, 1.24]	0.74

Note. SSS = subjective socioeconomic status; CI = confidence interval.

First, we investigated whether SSS influenced political naïve realism toward the self. Participants in the high-SSS condition displayed less naïve realism toward themselves than participants in the low-SSS condition. Again, we investigated both dimensions of naïve realism: perceived bias and competence. There was no difference in perceived bias by condition. Consistent with Studies 2 and 3, participants in the high-SSS condition thought they were more competent than participants in the low-SSS condition.

Next, we investigated whether participants in the high-SSS condition displayed more political naïve realism toward the imagined disagreeing other than did participants in the low-SSS condition. Replicating the previous studies, participants in the high-SSS condition displayed significantly more naïve realism than participants in the low-SSS condition. Again, there was no difference in perceived bias toward the imagined other by condition. Participants in the high-SSS condition thought the other, however, was less competent than participants in the low-SSS condition.

Finally, we investigated whether high-SSS participants reported more naïve realism toward the other relative to themselves. As expected, high-SSS participants displayed more naïve realism toward the other relative to themselves than low-SSS participants. There was no difference in the amount of bias expressed toward the other relative to themselves by condition. High-SSS participants had a larger difference in perceived self–other competence than low-SSS participants.

Weighted vote

Next, we investigated whether SSS influenced participants’ weighting of the disagreeing other’s vote. Consistent with our hypothesis, high-SSS participants wanted the vote of the imagined contrarian to have significantly less weight (M = 4.87, SD = 2.03) than did low-SSS participants (M = 5.31, SD = 1.98), t(411) = 2.21, p = .028, 95% CI_{Mean Dif} [0.05, 0.82], d = 0.22.

The mediating role of political naïve realism

Finally, we investigated whether SSS predicted naïve realism asymmetry (other–self), which in turn predicted the weighted vote. To test this pattern, we used the PROCESS macro (model 4) with 5,000 bootstrapped resamples (Hayes, 2013). There was a significant indirect effect of SSS on the weighted vote measure through naïve realism, indirect effect b = −0.39, 95% CI [−0.57, −0.23] (see Figure 2 for specific path results). This pattern suggests that high- (vs. low-) SSS caused more political naïve realism which, in turn, predicted the weighted vote.

Figure 2.

Study 4: Investigating whether the asymmetry of political naïve realism mediated the relationship between SSS and weighted vote.

Discussion

These findings suggest that high- (vs. low-) SSS increased naïve realism which, in turn, led to suppressing another’s contrary opinion. In a final study, we replicate and extend upon these findings. Specifically, instead of having participants imagine another person, we randomly assigned participants to see the recommendations of (ostensibly) another person. The other person either recommended significantly more or less redistribution. Therefore, the design for Study 5 was a 2 (SSS: High vs. Low) × 2 (Recommendation of the Other Person: Increase vs. Decrease Redistribution) between-subjects design.

Study 5

Method

Sensitivity analyses and participants

Participants (N = 261) were recruited through Amazon MTurk. Thirty-two participants failed the attention check, and four additional participants failed to complete the primary measures of interest. After excluding these participants from analyses, the final sample was 229 participants (118 males, 109 females, 2 did not report gender or any of the following demographic information). Sensitivity analyses revealed that given our sample size, we would be able to detect a small-to-medium effect size (f = 0.19; η² = 0.035) with adequate power. The average age of the sample was 33.24 years (SD = 11.05). The median income of the sample was US$35,000 to US$49,999 (range from under US$35,000 to over US$125,000 annually), and the median education was a 4-year college degree.

Procedure

This study utilized the same procedure for the investment game as was used in Study 4 except for one major change. This time, ostensibly to ensure that the game makers would create the “best version of the game,” participants were then randomly presented with one of two versions of purported recommendations from another “randomly selected” player. Importantly, in this experiment, participants were led to believe the other person was a real player, not asked to simply imagine a hypothetical other. In one condition, the other player recommended significantly more redistribution than the status quo: raising the rates for both the assessment and credit from 20% to 35%. In the other condition, the player recommended significantly less redistribution than the status quo: decreasing the rates for both the assessment and credit from 20% to 5%. No additional information was provided about the other player. Participants were then asked six political naïve realism questions, assessing beliefs about the other person: “Do you think the recommendations this participant made above are biased/based on self-interest/fair/based on principles of what is fair?” (e.g., 1 = not at all, 5 = completely), and “To what extent do you think this participant is smart/competent?” (e.g., 1 = not at all, 5 = completely). We averaged these items together to create an index of political naïve realism toward the other player (α = .84). In addition, we created a perceived bias dimension by averaging together the first four items (α = .85), and a perceived competence dimension by averaging together the second two items (Spearman–Brown = .80).

To demonstrate specificity of our findings, participants also completed measures of the other player’s perceived warmth. These items were as follows: “Based on the recommendations this participant made above, to what extent do you think this participant is kind/warm?” (1 = not at all; 5 = extremely; Spearman–Brown = .83). If the results are due to naïve realism, as opposed to a general valence effect, we would not anticipate higher SSS to report less warmth toward the other player based on the other players’ recommendation.

Next, participants were asked the weighted vote question in reference to the recommendations they saw of the (ostensible) other player.

Finally, to investigate a different research question, we measured perceived fairness of inequality, system justification beliefs in the game, and perceived fairness of inequality. These findings are not reported here as they are not germane. Participants also reported other demographic information including political party affiliation and political ideology as descriptive data.

Results

For the following analyses, we investigated simple comparisons using a Bonferroni adjustment. See Table 7 for main and interactive effect F values. See Table 8 for simple contrasts.

Table 7.

Study 5: Main and Interactive Effects for Dependent Variables of Interest.

Predictors	Political naïve realism			Bias dimension			Competence dimension			Perceived warmth			Weighted vote
Predictors	F	p	$η_{p}^{2}$	F	p	$η_{p}^{2}$	F	p	$η_{p}^{2}$	F	p	$η_{p}^{2}$	F	p	$η_{p}^{2}$
SSS condition	0.51	.477	0.00	0.10	.758	0.00	1.76	.186	0.01	2.18	.141	0.01	0.15	.698	0.00
Other player recommendation	20.16	<.001	0.08	10.67	.001	0.05	27.28	<.001	0.11	16.96	<.001	0.07	20.24	<.001	0.08
Interaction	26.05	<.001	0.10	21.49	<.001	0.09	15.10	<.001	0.06	18.21	<.001	0.08	32.97	<.001	0.13

Note. SSS = subjective socioeconomic status.

Table 8.

Study 5: Estimated Marginal Means, Standard Errors, and Contrasts.

Predictors	Other player recommends		Mean difference	p	95% CI
	More redistribution	Less redistribution
	M (SE)	M (SE)
Political naïve realism
Low-SSS	2.94 (0.09)	2.99 (0.09)	−0.06	.669	[−0.31, 0.20]
High-SSS	3.46 (0.09)	2.60 (0.09)	0.86	<.001	[0.61, 1.10]
Bias dimension
Low-SSS	3.03 (0.11)	3.18 (0.11)	−0.15	.341	[−0.45, 0.16]
High-SSS	3.56 (0.10)	2.72 (0.11)	0.85	<.001	[0.55, 1.44]
Competence dimension
Low-SSS	3.26 (0.10)	3.38 (0.10)	−0.13	.352	[−0.40, 0.14]
High-SSS	2.75 (0.09)	3.63 (0.10)	−0.88	<.001	[−1.15, −0.62]
Perceived warmth
Low-SSS	3.36 (0.10)	2.57 (0.10)	0.79	<.001	[0.53, 1.06]
High-SSS	3.10 (0.09)	3.11 (0.09)	−0.01	.915	[−0.27, 0.25]
Weighted vote
Low-SSS	4.71 (0.26)	4.39 (0.26)	0.32	.386	[−0.40, 1.03]
High-SSS	3.15 (0.25)	5.75 (0.25)	−2.60	<.001	[−3.30, −1.91]

Note. The p values are Bonferroni adjusted for multiple comparisons. CI = confidence interval; SSS = subjective socioeconomic status.

Political naïve realism

We hypothesized that high-SSS participants would think the other player was more biased when the other player recommended more (vs. less) redistribution. We tested our hypothesis by conducting a 2 (SSS: High vs. Low) × 2 (Recommendation of the Other Person: Increase vs. Decrease Redistribution) analysis of variance (ANOVA). Critically, the results revealed a significant interaction (see Figure 3). Consistent with our hypothesis, high-SSS participants thought the other player was significantly more biased and incompetent when the other player recommended more (vs. less) redistribution. However, in the low-SSS condition, there was no difference in naïve realism regardless of whether the other player recommended more or less redistribution.⁷

Figure 3.

Study 5: Mean naïve realism ratings as a function of SSS (high or low) and the other player’s recommendation (more or less redistribution).

We also investigated the two dimensions of naïve realism: bias and competence. In the high-SSS condition, participants thought the other player was more biased when the other player recommended more versus less redistribution. However, in the low-SSS condition, there was no difference in bias regardless of the other player’s recommendations. Furthermore, in the high-SSS condition, participants thought the other player was less competent when the other player recommended more versus less redistribution. However, in the low-SSS condition, there was no difference in bias regardless of the other player’s recommendations.

Finally, we investigated whether higher SSS felt less warmth to disagreeing others. If so, this would suggest that SSS led to feeling more negatively toward disagreeing others, as opposed to specifically enhancing one’s stance of naïve realism. Critically, the results revealed a significant interaction. For participants in the high-SSS condition, there was no difference in perceived warmth toward the other regardless of whether the other recommended more or less redistribution. For participants in the low-SSS condition, however, participants thought the other player was warmer when the other player recommended more versus less redistribution. This suggests that high-SSS uniquely increased political naïve realism, as opposed to generally feeling more negative, toward others who recommended more redistribution.

Weighted vote

Next, we investigated whether the interaction between SSS and the other player’s recommendations would predict the weighted vote. We hypothesized that high-SSS individuals would give the other player’s vote significantly less weight in the average when that other player recommended an increase (vs. decrease) in redistribution. Critically, the results revealed a significant interaction (see Figure 4). In the high-SSS condition, participants thought the other player’s vote should receive significantly less weight when the other player recommended an increase in redistribution than when the other player recommended a decrease in redistribution. However, in the low-SSS condition, there was no difference in the weighting of the other player’s vote regardless of whether the other player recommended an increase or a decrease in redistribution.

Figure 4.

Study 5: Weighted vote as a function of SSS (high or low) and the player’s recommendation (more or less redistribution).

The mediating role of political naïve realism

Finally, we investigated whether naïve realism mediated the moderated relationship between SSS, other’s recommendations, and the weighted vote (see Figure 5 for conceptual model).⁸

Figure 5.

Study 5: Conceptual model tested.

To investigate this hypothesis, we used the PROCESS macro (model 8) with 5,000 bootstrapped resamples (Hayes, 2013). All continuous variables were standardized prior to analysis. Consistent with our hypothesis, there was a significant conditional indirect effect of the other player’s recommendations (more vs. less redistribution) on the weighted vote measure through naïve realism when the participant was in the high-SSS condition (B = .68, 95% CI [0.44, 0.94]), but not when the participant was in the low-SSS condition (B = −.04, 95% CI [ −0.22, 0.13]; see Figure 6 for specific path results).

Figure 6.

Study 5: Mediated moderation model in which the indirect effect of the other player’s recommendations on the weighted vote measure through naïve realism differed as a function of the low-SSS (top panel) versus high-SSS condition (bottom panel).

Discussion

These findings suggest that high- (vs. low-) SSS increased naïve realism which, in turn, led to suppressing another’s contrary opinion. These findings suggest that in addition to material advantages, higher SSS can also lead to increased influence for psychological reasons. Higher SSS individuals appear to be more inclined to dominate the views of others, in part, because they feel more objectively justified in their own views.

General Discussion

These data provide evidence that higher SSS exacerbates naïve realism. In Study 1, we found evidence of the association between higher SSS and indicators of political naïve realism in a representative sample of the American electorate. The remaining experimental studies (Studies 2–5) provide causal evidence that higher SSS increased perception of one’s competence, increased political naïve realism toward others who disagree with one’s opinion, and led to political conflict. Finally, we investigated whether political naïve realism mediated the relationship between SSS and politically divisive behavior in the context of a weighted vote (Studies 4 and 5). Consistent with our predictions, we found that relative to low-SSS individuals, high-SSS individuals gave another player’s vote significantly less weight when they imagined a disagreeing other or when the other person recommended more (vs. less) redistribution.

In the United States, the very wealthy yield far greater political influence than poorer, and even average, citizens (Bartels, 2016; Gilens & Page, 2014). One reason is that wealthier citizens participate in political activities at higher rates than poorer citizens (Bonica et al., 2013; Verba & Nie, 1972). In fact, lower SES individuals are less willing than higher SES individuals to even express a political opinion (Laurison, 2015).

A number of frameworks have been developed to better understand the causes of this association between SES and political participation. Approaches focused on differences in economic, social, and other resources suggest that higher SES individuals tend to be more politically engaged than lower SES individuals because they have more money, more free time, and better developed civic skills (e.g., Verba & Nie, 1972). Others have argued that lower SES individuals may be less likely to engage politically because they feel less politically competent or efficacious (Laurison, 2015). Our results follow in this vein, suggesting that lower SES individuals may be less likely to speak out politically due to their perceptions of themselves as less competent and due to their relatively tempered opinions of those with whom they disagree. Higher SES individuals, in contrast, may be buffered from feelings of under confidence by the perception that those who disagree with them are biased or incompetent. Our findings also move beyond understanding these intrapersonal explanations for differences in political participation (i.e., resources, competence) to an interpersonal one, highlighting how the increased naïve realism of higher SSS individuals may lead them to actively discourage others’ political participation. Moreover, our results highlight one psychological explanation for increasing political polarization, particularly among political elites who may have an enhanced sense of their own competence relative to others and be more likely to denigrate and work to exclude the views of those with whom they disagree.

Limitations and Future Directions

It is important to note the limitations of the current research. First, the experimental approach conflates, to some extent, SSS with performance in the task. Because people often assume that one’s socioeconomic standing is a result of their work ethic and competence (e.g., Fiske et al., 2002; Pew Research Center, 2018), we tried to create feedback in the game that mimicked the real-world. Thus, people may be motivated to attribute their standing to themselves, even when this standing was randomly determined (as was the case in Study 3). A separate question is whether one’s standing relative to others on any given dimension may influence naïve realism. For example, imagine participants are randomly assigned numbers and told these are rankings (where lower numbers mean higher rankings). Would participants who were given numbers 1, 2, or 3 display more naïve realism than participants given numbers 98, 99, or 100? If so, this would suggest that the simple act of thinking of oneself as higher ranked than others may enhance one’s stance of naïve realism. A future study should investigate this possibility.

In addition, Study 5 revealed an interesting pattern of results with respect to political naïve realism and perceived warmth of the other player. Namely, high-SSS participants expressed more political naïve realism toward the other player who recommended more (vs. less) redistribution, but political naïve realism did not differ depending on the other players’ recommendation among low-SSS participants. However, unexpectedly, we did find that low- (but not high-) SSS participants thought the other player was less warm when the other player recommended less redistribution. Furthermore, perceived warmth indirectly mediated the relationship between other player’s recommendation and weighted vote among low- (but not high-) SSS participants (see Supplemental Analyses). This pattern of results suggests that perceived warmth of the other may be particularly important. Although some research suggests that low-SSS individuals may be particularly attuned to the emotions of others (Kraus et al., 2010), future research should follow up on these unexpected results.

Future research should also investigate mental images of the disagreeing other. It may be that participants are imagining others from a certain SSS when they are confronted with opinions that disagree with their own opinion. Regardless of who the participant is imagining, we anticipate the pattern of findings would be similar to what was reported here. This is because those who have strong opinions, such as higher SSS individuals (Laurison, 2015), tend to easily accept evidence which is congruent with their opinion, but critically evaluate and reject evidence which is incongruent with their opinion (e.g., Kahan et al., 2017; Lord, Ross, & Lepper, 1979). Therefore, higher SSS individuals, who feel quite competent, may discount contrary opinions regardless of the source.

Importantly, naïve realism may be only one of many mechanisms through which SSS exacerbates political divisions. For example, individual differences on personality traits such as narcissism may be correlated with SSS (Piff, 2014) and may also contribute to political divisions. Future research should consider the role of individual difference variables both explaining and influencing the current findings.

Conclusion

Political disagreement in the United States is widespread and growing. We suggest that SSS may help explain this increasing political divisiveness. Higher SSS individuals may experience more naïve realism and react more unfavorably toward those who disagree with them compared with lower SSS individuals. Moreover, higher SSS may contribute to the maintenance of uncompromising positions and an increased willingness to suppress the opinions of others.

Supplemental Material

Supplemental Material, Brown-Iannuzzi_Online_Appendix - A Privileged Point of View: Effects of Subjective Socioeconomic Status on Naïve Realism and Political Division

Supplemental Material, Brown-Iannuzzi_Online_Appendix for A Privileged Point of View: Effects of Subjective Socioeconomic Status on Naïve Realism and Political Division by Jazmin L. Brown-Iannuzzi, Kristjen B. Lundberg, Aaron C. Kay and B. Keith Payne in Personality and Social Psychology Bulletin

Supplemental Material

Supplemental_Analyses_5.2.2020 - A Privileged Point of View: Effects of Subjective Socioeconomic Status on Naïve Realism and Political Division

Supplemental_Analyses_5.2.2020 for A Privileged Point of View: Effects of Subjective Socioeconomic Status on Naïve Realism and Political Division by Jazmin L. Brown-Iannuzzi, Kristjen B. Lundberg, Aaron C. Kay and B. Keith Payne in Personality and Social Psychology Bulletin

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was funded by the National Science foundation (grant: 1729446) awarded to the first and last authors.

ORCID iDs

Jazmin L. Brown-Iannuzzi

Kristjen B. Lundberg

Supplemental Material

Supplemental material is available online with this article.

Notes

References

Adler

N. E.

Epel

E. S.

Castellazzo

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.

Bartels

L. M

. (2016). Unequal democracy: The political economy of the new gilded age. Princeton University Press.

Bonica

McCarty

Poole

K. T.

Rosenthal

(2013). Why hasn’t democracy slowed rising inequality? Journal of Economic Perspectives, 27, 103–124.

Brown-Iannuzzi

J. L.

Lundberg

K. B.

Kay

A. C.

Payne

B. K.

(2015). Subjective social status shapes support for redistribution. Psychological Science, 26, 15–26.

Brown-Iannuzzi

J. L.

Lundberg

K. B.

McKee

(2017). The politics of social class: How socioeconomic status may influence political attitudes and engagement. Current Opinions in Psychology, 18, 11–14.

Butler

J. V.

(2014). Inequality and relative ability beliefs. The Economic Journal, 126, 907–948.

Fast

N. J.

Chen

S. C.

(2009). When the boss feels inadequate: Power, incompetence, and aggression. Psychological Science, 20, 1406–1413.

Faul

Erdfelder

Buchner

Lang

A.-G.

(2009). Statistical power analyses using G*Power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41, 1149–1160.

Festinger

(1954). A Theory of Social Comparison Processes. Human Relations, 7, 117–140.

10.

Fiske

S. T.

Cuddy

A. J. C.

Glick

(2002). A model of (often mixed) stereotype content: Competence and warmth respectively follow from perceived status and competition. Journal of Personality and Social Psychology, 82, 878–902.

11.

Frantz

C. M.

(2006). I AM being fair: The bias blind spot as a stumbling block to seeing both sides. Basic and Applied Social Psychology, 28, 157–167.

12.

Gelman

Kenworthy

Y. S.

(2010). Income inequality and partisan voting in the United States. Social Science Quarterly, 91, 1203–1219.

13.

Gilens

Page

B. I.

(2014). Testing theories of American politics: Elites, interest groups, and average citizens. Perspectives on Politics, 12, 564–581.

14.

Gilovich

(1990). Differential construal and the false consensus effect. Journal of Personality and Social Psychology, 59, 623–634.

15.

Hayes

A. F

. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. The Guilford Press.

16.

Jost

J. T.

Burgess

(2000). Attitudinal ambivalence and the conflict between group and system justification motives in low status groups. Personality and Social Psychology Bulletin, 26, 293–305.

17.

Kahan

D. M.

Peters

Dawson

E. C.

Slovic

(2017). Motivated numeracy and enlightened self-government. Behavioral Public Policy, 1, 54–86. https://doi.org/10.1017/bpp.2016.2

18.

Kahneman

Slovic

Tversky

(Eds.). (1982). Judgment under uncertainty: Heuristics and biases. Cambridge University Press.

19.

Keltner

Robinson

R. J.

(1993). Imagined ideological differences in conflict escalation and resolution. International Journal of Conflict Management, 4, 249–262.

20.

Keltner

Robinson

R. J.

(1997). Defending the status quo: Power and bias in social conflict. Personality and Social Psychology Bulletin, 23, 1066–1077.

21.

Kennedy

K. A.

Pronin

(2008). When disagreement gets ugly: Perceptions of bias and the escalation of conflict. Personality and Social Psychology Bulletin, 35, 833–849.

22.

Kraus

M. W

. (2018). Beggars do not envy millionaires: Social comparison, socioeconomic status, and subjective well-being. In Diener

Oishi

Tay

(Eds.), Handbook of well-being. Noba Scholar Handbook Series: Subjective Well-Being (pp. 1–18). DEF Publishers.

23.

Kraus

M. W.

Côté

Keltner

(2010). Social class, contextualism, and empathic accuracy. Psychological Science, 21, 1716–1723.

24.

Kunda

(1990). The case for motivated reasoning. Psychological Bulletin, 108, 480–498.

25.

Laurison

(2015). The willingness to state an opinion: Inequality, don’t know responses, and political participation. Sociological Forum, 30, 925–948.

26.

Lerner

M. J.

(1965). Evaluation of performance as a function of performer’s reward and attractiveness. Journal of Personality and Social Psychology, 1, 355–360.

27.

Liberman

Minson

J. A.

Bryan

C. J.

Ross

(2012). Naïve realism and capturing the “wisdom of dyads.” Journal of Experimental Social Psychology, 48, 507–512.

28.

Lord

C. G.

Ross

Lepper

M. R.

(1979). Biased assimilation and attitude polarization: The effects of prior theories on subsequently considered evidence. Journal of Personality and Social Psychology, 37, 2098–2109.

29.

Miller

D. T.

Ross

(1975). Self-serving biases in the attribution of causality: Fact or fiction? Psychological Bulletin, 82, 213–225.

30.

Pennebaker

J. W.

Boyd

R. L.

Jordan

Blackburn

. (2015). The development and psychometric properties of LIWC2015. The University of Texas at Austin.

31.

Pew Research Center. (2014). Political polarization in the American public. U.S. Politics & Policy. http://www.people-press.org/2014/06/12/political-polarization-in-the-american-public/

32.

Pew Research Center. (2016). Partisanship and political animosity in 2016: Highly negative views of the opposing party—And it’s members. U.S. Politics & Policy. http://www.people-press.org/2016/06/22/partisanship-and-political-animosity-in-2016/

33.

Pew Research Center. (2018). Latinos are more likely to believe in the American dream, but most say it is hard to achieve. https://www.pewresearch.org/fact-tank/2018/09/11/latinos-are-more-likely-to-believe-in-the-american-dream-but-most-say-it-is-hard-to-achieve/

34.

Piff

P. K.

(2014). Wealth and the inflated self: Class, entitlement, and narcissism. Personality and Social Psychology Bulletin, 40, 34–43.

35.

Piff

P. K.

Kraus

M. W.

Côté

Cheng

B. H.

Keltner

(2010). Having less, giving more: The influence of social class on prosocial behavior. Journal of Personality and Social Psychology, 99, 771–784.

36.

Preacher

K. J.

Curran

P. J.

Bauer

D. J.

(2006). Computational tools for probing interaction effects in multiple linear regression, multilevel modeling, and latent curve analysis. Journal of Educational and Behavioral Statistics, 31, 437–448.

37.

Pronin

(2008). How we see ourselves and how we see others. Science, 320, 1177–1180.

38.

Pronin

Gilovich

Ross

(2004). Objectivity in the eye of the beholder: Divergent perceptions of bias in self versus other. Psychological Review, 111, 781–799.

39.

Pronin

(2007). Perception and misperception of bias in human judgment. Trends in Cognitive Sciences, 11, 37–43.

40.

Pronin

Lin

D. Y.

Ross

(2002). The bias blind spot: Perceptions of bias in self versus others. Personality and Social Psychology Bulletin, 28, 369–381.

41.

Robinson

R. J.

Keltner

(1996). Much ado about nothing? Revisionists and traditionalists choose an introductory English syllabus. Psychological Science, 7, 18–24.

42.

Ross

Greene

House

(1977). The false consensus effect: An egocentric bias in social perception and attribution processes. Journal of Experimental Social Psychology, 13, 279–301.

43.

Ross

Ward

. (1995). Psychological barriers to dispute resolution. In Zanna

(Eds.), Advances in experimental social psychology (Vol. 27, pp. 255–304). Academic Press.

44.

Ross

Ward

. (1996). Naive realism in everyday life: Implications for social conflict and misunderstanding. In Brown

Reed

E. S.

Turiel

(Eds.), The Jean Piaget Symposium Series: Values and knowledge (pp. 103–135). Lawrence Erlbaum.

45.

Saez

Zucman

. (2014). Wealth inequality in the United States since 1913: Evidence from capitalized income tax data (National Bureau of Economic Research, Working Paper No. 20625). http://www.nber.org/papers/w20625

46.

Smith

(2017). Urban Institute’s tabulations from the current population survey 1963-2017. http://apps.urban.org/features/wealth-inequality-charts/

47.

Stephens

N. M.

Markus

H. R.

Phillips

L. T.

(2014). Social class culture cycles: How three gateway contexts shape selves and fuel inequality. Annual Review of Psychology, 65, 611–634.

48.

Verba

Nie

N. H

. (1972). Participation in America. The University Press.

49.

Voorheis

McCarty

Shor

(2015). Unequal incomes, ideology and gridlock: How rising inequality increases political polarization. https://ssrn.com/abstract=2649215

50.

Yair

Sulitzeanu-Kenan

(2015). Biased judgment of political bias: Perceived ideological distance increases perceptions of political bias. Political Behavior, 37, 487–507.

51.

Zell

Alicke

M. D.

(2010). The local dominance effect in self-evaluation: Evidence and explanations. Personality and Social Psychology Review, 14, 368–384.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.56 MB

0.17 MB