Republicans are more optimistic about economic mobility,but no less accurate

Abstract

Do Americans overestimate economic mobility? Using representative surveys of the public and local government officials, we assess claims of widespread misperceptions about economic mobility by measuring the accuracy of participants’ perceptions of both relative and absolute mobility. Republican members of the public and government officials are more optimistic than are Democrats about poor children’s chances of reaching the highest income quintile (relative mobility) and earning more than their parents (absolute mobility). Democrats also rate race and family wealth as more important to children’s chances than do Republicans. However, partisan tendencies to overestimate or underestimate mobility are roughly symmetric despite differences in optimism; we only observe small and inconsistent differences in belief accuracy by party for both the public and local officials. Finally, accuracy is no greater for perceptions of state and local mobility than at the national level.

1. Introduction

Belief in the American Dream—the ideal that economic mobility is uniquely achievable in the United States via hard work—often exceeds reality. How much do Americans overestimate economic mobility and which groups are most likely to do so? Previous research shows that Americans overestimate the probability of moving up in the income distribution and that conservatives are most vulnerable to this tendency (e.g., Alesina et al., 2018; Manza and Brooks 2020). Correspondingly, Republicans are more likely to express belief in people’s ability to get ahead by working hard than are Democrats, who instead emphasize the role of structural factors such as the education system and racial discrimination in limiting mobility (Newport 2018; Horowitz et al., 2020).

Prior studies focus on public perceptions of relative mobility at the national level. However, these partisan belief differences about mobility might differ for perceptions of absolute mobility—earning more than your parents—or for contexts that are closer to people’s lived experiences. In addition, past studies have not investigated the accuracy of beliefs about mobility among local officials, who are more polarized than the public but also more sophisticated and knowledgeable about policy issues.

Leveraging new data on relative and absolute mobility in the United States and across states and local areas from Chetty et al. (2014a), Chetty et al. (2014b), and Chetty et al. (2017), we compare Americans’ perceptions of economic mobility across income quintiles and generations at both the national and state/local level in paired surveys of local officials and the general public.

Our findings confirm that Republicans have more optimistic perceptions of mobility in both relative and absolute terms than Democrats and correspondingly rate race and parental wealth as less important for a person’s likelihood of getting ahead. These partisan differences hold among both the public and local officials. However, the partisan differences in optimism that we observe do not translate to consistent differences in belief accuracy—Republicans overestimate mobility and Democrats underestimate it in roughly equal measure. Finally, the inaccuracy we observe among the public in estimating mobility was not mitigated by expertise (local officials are not consistently better at estimating mobility) or lived experience (respondents did not have more accurate perceptions of mobility in their state or local area compared to the United States as a whole).

2. Prior work and expectations

Americans are polarized in their views about social and political issues. This polarization often extends from values and public policy opinions to disagreement over facts (e.g., Frankovic 2016, 2018). Scholars attribute belief polarization and associated misperceptions to factors ranging from elite cues (e.g., Zaller 1992) to directionally motivated reasoning (e.g., Taber and Lodge 2006).

Political elites may not be immune to these pressures. Local officials tend to be more knowledgeable, which is generally associated with greater belief accuracy (Gottfried et al., 2013), but are also likely to have more polarized opinions. The net effect of these competing pressures on belief accuracy and polarization is not clear. One prior study (Lee et al., 2020) found that local officials hold more accurate beliefs than the public across several issues but that this greater belief accuracy is not associated with reduced partisan belief polarization.

Public beliefs may also differ at the local level where knowledge is less dependent on elite cues and media coverage. For instance, Rapeli (2014) finds that people’s knowledge of local-level politics often differs from their knowledge of national-level politics in Europe.

We examine these questions in the context of economic mobility, where prior research indicates misperceptions are common. Public opinion data shows that Americans overestimate relative economic mobility. Alesina, Stantcheva, and Teso (2018) measure Americans’ perceptions of upward national relative mobility and compare them to true values. They find that Americans (especially conservatives) overestimate mobility, a tendency that could affect Americans’ attitudes toward redistributive policies (Manza and Brooks 2020; Alesina et al., 2018).

This study contributes to prior research in four ways. First, we investigate perceptions of relative mobility along with absolute mobility within the same population. Second, we measure and evaluate accuracy in addition to optimism to determine whether Democrats or Republicans are more accurate or whether they are equally (in)accurate but in different directions. Third, we use the same instrument to survey both members of the public and local officials, allowing us to compare their perceptions of both relative and absolute mobility and their accuracy. Fourth, we asked respondents for their perceptions of economic mobility in both their local area or state and the United States to compare these beliefs.

We expected that Republicans would have more optimistic assessments of economic mobility than Democrats. As the party less in favor of government intervention to ameliorate inequality, Republicans might need to overestimate economic mobility to justify their policy positions (with Democrats needing to see low mobility to justify their positions). ¹ We also expected that government officials would have more accurate beliefs than the American public per Lee et al. (2020), who suggest that officials have incentives to be more accurately informed. Additionally, we expected both the public and government officials to have more accurate perceptions of economic mobility in their local areas or states than the country as a whole. Finally, we asked respondents which factors they believe are most important for getting ahead. Given each party’s traditional concerns, we expected Democrats to place greater emphasis on the role of circumstances outside of people’s control such as race and parental wealth than Republicans.

3. Methods and design

We conducted parallel surveys of the American public (referred to as Americans or members of the public) and local government officials (referred to as local officials). ² We collected data from 2956 respondents—556 local officials surveyed by CivicPulse from March 31–May 18, 2020, and 2400 public respondents surveyed by YouGov from March 12–April 1, 2020. ³ The YouGov survey was a non-probability online sample representative of United States adults. The CivicPulse survey targeted local officials across the United States. ⁴

Our survey measured perceptions of two types of economic mobility. Relative mobility was measured using a graphic illustration of a ladder (see Supplementary Appendix A, which provides the full survey instrument). Respondents estimated the chances that children born in one of the poorest 100 families out of 500 hypothetical families would grow up to be in each income quintile at both the local and national level. We measured perceptions of absolute mobility by asking respondents to estimate the percentage of children born around 1980, in either their state or the United States, who earned a higher income at age 30 than their parents did at the same age. Lastly, in an extension of a pilot study conducted in Michigan (Institute for Public Policy and Social Research 2018), respondents were asked to rate the importance of various factors in getting ahead in life: race, parent wealth, education level, hard work, and delaying having children until after marriage.

For each type of mobility, we consider two outcome measures: optimism and accuracy. We first calculate directional error measures that we label optimism, which is the difference between perceived mobility and true mobility. Higher values thus indicate greater optimism about mobility. We also calculate an absolute measure of error that we label inaccuracy, which is the absolute value of the difference between the estimate and the true value. Higher values of this measure represent more inaccurate beliefs about mobility (which could be either underestimated or overestimated).

Our measures of true relative mobility at the local and national level are derived from Chetty et al. (2014b) and Chetty et al. (2014a) (we match respondents to commuting zone by zip code). Levels of absolute mobility at the state and national level are derived from Chetty et al. (2017). ⁵

4. Results

We divide our results into three sections: optimism about economic mobility, inaccuracy in perceptions of economic mobility, and perceptions about which factors are important for getting ahead. Within the optimism and inaccuracy sections, we use three sets of comparisons: by partisanship, by government officials versus the public, and by local/state area versus national. We define relative mobility as the respondent’s estimated likelihood that someone born into the lowest income quintile will move into the highest income quintile (Q1 → Q5). ⁶ Respondent perceptions are compared against the true values. ⁷

4.1. Optimism

Consistent with prior research, we find that the American public is overly optimistic about relative economic mobility (though not by a large margin). The public estimates the rate of national Q1 → Q5 mobility as 10.6%, an overestimate relative to the true value of 7.5%. By contrast, the public is slightly too pessimistic about absolute economic mobility at the national level. The public expects 46.6% of children to earn more than their parents, which is a bit less than the true value of 50%. ⁸

To understand these results better, we examine how optimism in perceptions of economic mobility differs by partisanship and whether the respondent is a local official or a member of the public at the national and local/state level. In these analyses, we calculate optimism as the (signed) difference between perceived and true values. Higher values thus represent greater perceived relative or absolute mobility. Values above zero indicate perceptions exceeding the true value while those below zero indicate perceptions below the true value.

Figure 1(a) shows perceptions of the probability of moving from the bottom to the top income quintile (Q1 → Q5) compared to the true value (i.e., relative mobility). Figure 1(b) shows perceptions of the likelihood of a person exceeding their parents’ income at the same age compared to the true value (i.e., absolute mobility). Local refers to commuting zone. Data for true values are sourced from Chetty et al. (2014a) and Chetty et al. (2017).

Figure 1.

Optimism in perceptions of economic mobility. (a) Relative mobility. (b) Absolute mobility.

Our findings are summarized in Figure 1(a) (partisan optimism about relative mobility in respondents’ local area and the United States) and Figure 1(b) (partisan optimism in perceptions of state and national absolute mobility). The four points in each row of both graphs correspond to the perceived levels of mobility relative to the true values among local officials and the public: Democratic local officials (blue squares), Republican local officials (red squares), Democratic members of the public (blue triangles), and Republican members of the public (red triangles).

As Figure 1 indicates, Republicans are consistently more optimistic than are Democrats both within the public and among officials. Consider the first row of Figure 1(a). On average, Democratic members of the public estimated that 9.2 children born into the poorest 100 families would make it into the richest 100 families (out of 500) while Republican members of the public estimated that 13 of those children would make it into that group. Both exceeded the true value of 7.5%, but Republican members of the public were more optimistic. The corresponding blue triangle in the graph for Democrats in the public appears at 0.017 (the signed difference of 0.092–0.075), while the red triangle for Republicans in the public appears at 0.065 (the signed difference of 0.130–0.075). Democratic and Republican officials showed a similar divide, though both underestimated mobility. Democratic officials estimated it at 3.2% and with Republicans estimating 6.5%, so they are plotted as blue and red squares, respectively, at −0.043 and −0.010.

Similar patterns of greater Republican optimism are evident for estimates of local relative mobility and for absolute mobility at both the state and national level.

These differences are confirmed in statistical tests. As Supplementary Appendix Table B1 indicates, Democratic members of the public were less optimistic about relative mobility than their Republican counterparts at the national (−0.042, p < 0.005) and local (−0.047, p < 0.005) level. Partisan differences in perceptions of absolute mobility are also significant. Per Supplementary Appendix Table B1, Democrats are less optimistic about national (−0.110, p < 0.005) and state absolute mobility (−0.113, p < 0.005) than are Republicans. These partisan differences are mirrored for local officials. ⁹

We observe a significant difference between local officials and the public in perceptions of Americans’ chance of moving up from the first to the fifth quintile. Members of the public from both parties were more optimistic about relative mobility than were local officials. For absolute mobility, however, Republican officials were the most optimistic. ¹⁰

Finally, we find no evidence that people hold more or less optimistic beliefs about economic mobility in their local area/state compared to national economic mobility. ¹¹

4.2. Inaccuracy

We also estimated models of inaccuracy for both relative and absolute economic mobility, which measure how far Democrats and Republicans are from the true values (i.e., the absolute value of the difference). We illustrate these results in Figure 2(a), which shows partisan inaccuracy for perceptions of relative mobility, and Figure 2(b), which shows partisan inaccuracy for perceptions of absolute mobility. ¹² Supplementary Appendix Table B2 models predictors of inaccurate perceptions of relative and absolute mobility at the national or local/state level.

Figure 2.

Inaccuracy of perceptions of economic mobility. (a) Relative mobility. (b) Absolute mobility.

In contrast with the conventional narrative that inaccuracy of American perceptions of economic mobility is driven by Republican overestimation of economic mobility, we find generally similar levels of inaccuracy for perceptions of economic mobility between parties. In fact, we find some evidence that Republicans may actually be more accurate in their perceptions of absolute mobility.

Figure 2(a) and Supplementary Appendix Table B2 demonstrate that Democratic and Republican officials and members of the public did not significantly differ in the accuracy of their perceptions of relative mobility (Q1 → Q5). When we consider estimates of relative mobility across all five income quintiles using Brier and logarithmic (log) scores, we also find no systematic partisan difference in accuracy (see Supplementary Appendix Table B4). ¹³

For absolute mobility, a consistent pattern does not emerge. Supplementary Appendix Table B2 indicates that Republican members of the public were slightly more accurate for national (0.018, p < 0.05) and state absolute mobility (0.019, p < 0.05) than Democratic members of the public. As for local officials, the difference in accuracy between Democratic local officials’ and Republican local officials’ estimates of absolute mobility was not significant. As such, partisans were approximately equally inaccurate in their estimates of mobility; in the instances in which they were not, Republican estimates were slightly more accurate.

We also find relatively similar levels of inaccuracy between local officials and members of the public for their perceptions of absolute mobility. However, elites were more accurate in their perceptions of relative mobility across parties at the local and national levels (p < 0.005 in both cases).

Overall, we find no evidence that people hold more accurate beliefs about relative or absolute mobility in their local area or state versus the national level. (Indeed, some evidence even suggests greater inaccuracy in relative mobility at the local level; see Supplementary Appendix Tables B5 and B6.)

Figure 2(a) shows the inaccuracy of local officials and members of the public’s perceptions of the probability of moving from the bottom to the top income quintile (Q1 → Q5). Figure 2(b) shows the inaccuracy of local officials and members of the public’s perceptions of the likelihood of a person exceeding their parents’ income at the same age compared to the true value. Local refers to commuting zone. Data for true values are sourced from Chetty et al. (2014a) and Chetty et al. (2017).

4.3. Factors in getting ahead

We also asked respondents to rate the importance of five factors—race, parent wealth, education, delaying having children until marriage, and hard work—in getting ahead in life. We then segmented the results by partisanship and graphed the average ratings for both members of the public (Figure 3(a)) and local officials (Figure 3(b)). Democrats ranked uncontrollable factors, such as race and parent wealth, as more important for getting ahead than their Republican counterparts. These two factors showed the largest partisan differences for both local officials and members of the public, with Democrats rating them above fairly important on average and Republican rating them as less than fairly important on average. Furthermore, Democratic members of the public ranked more mutable factors, like delaying having children and working hard, as less important for getting ahead than did Republicans. ¹⁴

Figure 3.

Perceived importance of various factors in getting ahead. (a) Local officials. (b) Public.

Figure 3(a) shows mean local official perceptions of the importance of various factors in getting ahead in life, while Figure 3(b) shows mean public perceptions of the importance of the same factors. The factors race and parent wealth exhibit the greatest party mean differences for both the public and local officials.

5. Conclusion

Consistent with prior evidence, we find that Americans overestimate relative mobility and Republicans are more optimistic than Democrats about children’s chances, but our results challenge and expand these findings in important ways. First, though Republicans have more optimistic perceptions of both relative and absolute economic mobility than Democrats, we did not find that one party held consistently more accurate beliefs. Second, greater expertise and lived experience did not consistently improve the accuracy of people’s beliefs—we found only mixed evidence that local officials have more accurate perceptions than the public and no indication that respondents have more accurate perceptions about mobility in their state/local area. We did confirm that partisan differences extend to explanations for mobility, however: Democrats perceive the uncontrollable factors of race and parental wealth to be more important for getting ahead than do Republicans.

Overall, these results complicate scholarly narratives about inaccurate perceptions of economic mobility. The public overestimates relative mobility but underestimates absolute mobility on average; both approach the correct answers in aggregate. Moreover, partisan differences in perceptions at both the elite and public levels do not correspond to systematic differences in belief accuracy. Republicans tend to overestimate relative mobility while Democrats tend to underestimate it. Correspondingly, Republicans generally think Americans can overcome disadvantaged family backgrounds but Democrats are more likely to think immutable factors constrain poor children. Finally, notions about what might reduce partisan differences or increase belief accuracy are unsupported—neither local knowledge nor government experience is consistently associated with significantly greater accuracy.

Given that we examine cross-sectional descriptive data, we cannot be confident about causal order. Republican partisanship may lead to more optimistic perceptions of economic mobility due to differences in information exposure (e.g., more rags-to-riches narratives) or policy preferences (e.g., greater opposition to redistributive policies). Alternatively, belief in the traditional bootstrap narrative may lead citizens to identify with the more conservative party. Similar potential explanations apply for Democrats. Some may wish to justify their preference for government intervention in the economy or a strong safety net, while others may have been motivated to support the more liberal party due to their perceptions of low levels of mobility. More complex causal paths are also possible. For instance, some people may be more likely than others to believe in and need to defend a just world (Jost et al., 2004), which would lead them toward support for the Republican Party and greater perceptions of economic mobility.

The findings also have other limitations. First, we use local government officials to represent political elites; these officials may not be representative of elected officials at higher levels of government such as state legislators or members of Congress. Second, we cannot control respondent effort. It is possible that some respondents guessed randomly or put little thought into answering questions about mobility, which asked them to estimate quantities with which they were likely unfamiliar and which may be sensitive to question wording or context effects (Swan et al., 2017). Third, innumeracy may have also limited respondents’ ability to translate their perceptions of economic mobility into our specific quantities of interest. Survey fatigue could similarly have resulted in respondents simply repeating similar or identical answers for both state/local and national mobility.

Nonetheless, these results provide valuable new evidence on perceptions of economic mobility in the United States. Americans do not uniformly overestimate mobility. Instead, the seemingly ubiquitous narrative of the American Dream has become yet another issue shaped by polarized partisan differences in public and official beliefs.

Supplemental Material

sj-pdf-1-rap-10.1177_20531680211050506 – Supplemental Material for Republicans are more optimistic about economic mobility, but no less accurate

Supplemental Material, sj-pdf-1-rap-10.1177_20531680211050506 for Republicans are more optimistic about economic mobility, but no less accurate by Matt Grossmann, Kayla Hamann, Jennifer Lee, Gabrielle Levy, Brendan Nyhan and Victor Wu in Research & Politics