Abstract
Economic inequality is a pervasive and growing source of social problems such as poor health, crime, psychological disorders, and lack of trust in others. U.S. citizens across the political spectrum both underperceive the extent of economic inequality and would prefer to live in a society with much less inequality than both exist in reality and in their subjective estimations. Across multiple studies, we examined the ability of “moral foundations” to predict people’s desire to reduce economic inequality (while also replicating research showing widespread desire for a more equal society). Moral foundations endorsements consistently predicted desire to reduce inequality even when controlling for other relevant factors (e.g., political orientation). In addition, requests for donations to an organization focused on reducing economic inequality were able to elicit more money when the requests largely appealed to the type of moral foundations endorsed by participants.
Severe economic inequality is the cause of much societal strife (e.g., Wilkinson & Pickett, 2009). In the United States, the richest 1% of individuals own over 40% of the nation’s wealth, and the top wealth quintile owns over 84% (Davies, Sandstrom, Shorrocks, & Wolff, 2009; Keister, 2000; Wolff, 2002). Inequality also contributes to deficits in subjective well-being (Alesina, Di Tella, & MacCulloch, 2004) and higher homicide rates (Daly, Wilson, & Vasdey, 2001). Research has demonstrated that respondents in the United States massively underestimate the degree of wealth inequality in the United States and that, regardless of their political persuasion or socioeconomic status, they desire an even more equal society than the one that they estimate exists (Norton & Ariely, 2011).
Despite agreement that an ideal society would be much more equal than the reality in the United States, conservatives and liberals disagree on how to deal with inequality due to differences in perceptions of procedural justice (Kay et al., 2009). Accordingly, the current research seeks to explain differences in desire to reduce inequality using moral foundations theory, which comprises factors related to minimizing harm, loyalty to the in-group, obedience to authority, maintaining purity, and, perhaps most importantly in terms of economic issues, perceptions of justice/fairness (Graham, Haidt, & Nosek, 2009).
(Mis)Perceptions of Inequality
Regardless of income or political orientation, respondents living in modern Western democracies with relatively high levels of wealth inequality (e.g., United States and Australia) underestimate inequality in their countries and report preferences for wealth gaps that are much smaller than even their much-lower-than reality estimates (Norton & Ariely, 2011; Norton, Neal, Govan, Ariely, & Holland, 2014). Most respondents prefer to live in a country with a modest wealth gap like Sweden, or even a completely equal society, than one with a massive gap like the United States (Norton & Ariely, 2011).
Although different measures have been used to assess estimates of wealth inequality (e.g., Eriksson & Simpson, 2012; Norton & Ariely, 2011), Norton and Ariely (2013) emphasize all assessments consistently reveal a desire for much less inequality in comparison to respondents’ estimates and to the real-world amount. Americans prefer a reduced wealth gap, in general, and also narrower pay gaps between CEOs and typical workers. Indeed, U.S. respondents estimated a 30:1 gap, ideally want a 7:1 gap, and actually live with a 354:1 gap (Kiatpongsan & Norton, 2016).
Negative Effects of Inequality
Wilkinson and Pickett (2009) outline myriad negative effects of wealth inequality including reductions in physical and mental health, increased substance abuse, increased violence, lower educational achievement, and mistrust. Moreover, income inequality predicts increased homicide rates (Daly et al., 2001). Data from the 1999 European Values Survey show that solidarity among both the rich and the poor is higher in countries with less income inequality (Paskov & Dewilde, 2012), while data from large U.S. and European surveys indicate that happiness is reduced when inequality increases (Alesina et al., 2004).
Perceived fairness may play a particularly important role in the relationship between inequality and negative outcomes. Indeed, perceived fairness mediates the relationships between inequality and negative outcomes (Alesina & La Ferrara, 2005). Policies addressing fairness that reduce inequality predict greater overall happiness across nations (Oishi, Schimmack, & Diener, 2011). Experimental research also demonstrates that people will sacrifice some of their own wealth in order to decrease inequality between those with very high and those with very low randomized payouts (Dawes, Fowler, Johnson, McElreath, & Smirnov, 2007). Hence, fairness seems critical to inequality and its negative effects, suggesting that other moral concerns could also underlie evaluations of inequality.
Moral Foundations Theory
People use several evolved psychological mechanisms that factor into their moral decision-making. Along with a natural disposition to be concerned with fairness (e.g., Brosnan & de Waal, 2014), at least four other inherent moral foundations have been identified: concerns with harm/care, in-group loyalty, obedience to authority, and maintaining purity.
Harm and fairness, the “individualizing foundations,” are relatively more important to liberals than to conservatives, while the other “binding foundations” are relatively more important to conservatives than to liberals (Graham et al., 2009). More than just a predictor of general political attitudes, moral foundations endorsements also predict charitable donations, with individuals of each political orientation being more likely to support charities whose mission statements are consistent with the individual’s moral foundations (Winterich, Zhang, & Mittal, 2012). Importantly, in regard to the current research, moral foundations endorsements are also predictive of political behavior (Franks & Scherr, 2015; Johnson et al., 2014). Accordingly, we should also expect moral foundations to predict differences in desire to reduce inequality.
Desire to Reduce Inequality
Although Americans agree that there is too much inequality in the United States, various factors predict differences in support for redistributive public policies. People sometimes justify inequality when they feel trapped in and dependent on their economic system (Kay et al., 2009). People tolerate inequality more when they perceive it to be caused by fair and just processes (Tyler, 2011). Adopting such a procedural justice perspective assumes that one could easily improve their socioeconomic status through hard work and talent. Americans, especially poorer individuals and conservatives, overestimate upward social mobility (Davidai & Gilovich, 2015), and this prospect of social mobility (POSM) leads individuals to engage in system justification (Day & Fiske, 2017) and is associated with an increased tolerance for income inequality (Shariff, Wiwad, & Aknin, 2016). Anticipation of promotions and pay increases also predicts decreased support for redistributing wealth to reduce inequality (Rainer & Seidler, 2008).
The primary purpose of the current research is to determine the extent to which moral foundations (e.g., Graham et al., 2009) predict desires to reduce wealth inequality (operationalized as the difference between their subjective estimates of inequality and their subjective preferences for an ideal level of inequality). Specifically, the research examined whether the liberal individualizing foundations (harm and fairness) and the conservative binding foundations (loyalty, authority, and purity) uniquely predicted desires to reduce wealth inequality. In Study 1a, we attempted to replicate the general pattern of results from Norton and Ariely (2011) and used the Moral Foundations Questionnaire (MFQ) to determine which of the various moral foundations have the ability to predict desire to reduce inequality. In Study 1b, we sought to confirm the importance of specific moral foundations as unique predictors of inequality attitudes.
Studies 1a and 1b
In Study 1a, we sought to support prior findings regarding American’s perceptions of wealth inequality. Additionally, we predicted that endorsement of the individualizing and binding foundations would predict desire to reduce inequality. We also explored the unique ability of each moral foundation to predict desire to reduce inequality.
In Study 1b, we sought to replicate the results of Study 1a. After our exploratory examination of the unique contributions of each separate moral foundation, we were able to make a priori hypotheses that the moral foundations of purity and fairness would uniquely predict desire to reduce inequality.
Method
Participants
A total of 275 undergraduates (125 in 1a; 150 in 1b) participated for course extra credit. The samples were predominantly female (75%; 73%) and Democratic leaning (65%; 61%). The income distribution was relatively even between those from households making below US$50,000 (30%; 33%), those making between US$50,000 and US$100,000 (40%; 39%), and those making above US$100,000 (30%; 28%). We focus on these demographic groupings because they are consistent with prior research in this area (Norton & Ariel, 2011). For Studies 1a, 1b, and 2, data collection was maximized during a single week in order to minimize the number of politically relevant events that occurred while conducting a particular study.
Materials and Procedure
All Studies
Participants completed informed consent documents and demographic questionnaires. Sample characteristics are discussed prior to results for each study. Participants reported estimates of the percentage of wealth owned by each wealth quintile (i.e., estimated wealth distribution) and their ideal quintile-by-quintile wealth distribution (see Supplemental Figure 1 for comparisons with Norton & Ariely, 2011), with the ideal top quintile amount being subtracted from the estimated top quintile amount to create a measure of adjusted wealth inequality. Participants in all studies also completed the MFQ (Graham et al., 2009). Combined zero-order correlations for variables appearing in all studies appear in Supplemental Table S1.
Participants in Studies 1a and 1b reported preferences for living in societies with varying levels of wealth inequality with choices between living in societies where the top quintile owned (1) 84% (U.S. distribution) versus 36% (Swedish distribution), (2) 84% versus 20% (completely equal distribution), (3) 36% versus 20%, and (4) 51% (slight top quintile hegemony distribution) versus 36% of the wealth (see Supplemental Figure 2 for depiction of choices).
Results
Americans Prefer Sweden Again
Consistent with prior research (Norton & Ariely, 2011), a vast majority of participants in both studies reported that they would prefer to live in a country with a wealth distribution mirroring Sweden’s (where those in the top quintile own 36% of the wealth) rather than one with a wealth distribution mirroring the United States (top quintile owns 84% of the wealth). A large, though somewhat smaller majority also said they would prefer to live in a completely equal country rather than one as unequal as the United States. The Swedish wealth distribution was also preferred to the completely equal distribution and to a slight hegemony of the top quintile. These results are summarized in Table 1.
Studies 1a and 1b Choices for Preferred Wealth Distributions.
Underestimates of Inequality
Participants in Studies 1a and 1b vastly underestimated the wealth of the top 20% at only 63% and 62.8%, respectively. One-sample t tests revealed that the difference between reality (84%) and our respondent’s estimates was significant in both studies, t(124) = −12.28, 95% CI [−24.35, −17.59], p < .001, and t(149) = −13.79, 95% CI [−24.19, −18.12], p < .001. This was true regardless of participant demographics (see Table 2). Comparisons of groups by party affiliation, sex, and household income found only a significant difference for sex whereby men’s estimates were significantly more accurate than women’s, t(123) = −4.28, p < .001 and t(148) = −4.71, p < .001. Overall, these results are similar to those found by Norton and Ariely (2011).
Studies 1a and 1b Estimates of the Percentage of Wealth Owned by the Top 20% by Demographic.
Note. 95% CI for Sex 1a [−23.42, −8.60]; 95% CI for Sex 1b [−22.5, −9.0].
aSignificant difference between demographic groups in the same row.
Ideal Wealth Distributions
The amount of wealth that would be owned by the top quintile in participants’ ideal societies in Studies 1a (M = 34.96, SD = 13.0) and 1b (M = 35.3, SD = 13.1) was significantly lower than the real-world amount, t(124) = −42.18, 95% CI [−51.34, −46.74], p < .001, and t(149) = −45.38, 95% CI [−50.8, −46.7], p < .001, as well as their own underestimates, t(125) = −24.12, 95% CI [−30.34, −25.74], p < .001, and t(149) = −25.81, 95% CI [−29.8, −25.6], p < .001. These patterns held among all demographics. However, there were differences between men and women and between Republicans and Democrats (see Table 3).
Study 1 Ideal Amount of Wealth Owned by the Top 20% by Demographic.
Note. 95% CI for Party [2.29, 11.64]; 95% CI for Sex [−15.57, −1.55].
aSignificant difference between demographic groups in the same row.
Regression Analyses
Regression analyses were conducted on a variable computed by subtracting participants’ ideal amount of wealth owned by the top quintile from their estimates. The dependent variable was regressed on each of the demographic variables (i.e., party affiliation, sex, and income) and each of the five moral foundations.
The combination of predictor variables predicted 16% of the variance in adjusted wealth inequality in Study 1a, F(8, 116) = 2.75, p = .01, 18% of the variance in adjusted wealth inequality in Study 1b, F(8, 141) = 3.87, p < .001, and 14% of the variance in the combined data set, F(8, 266) = 5.50, p < .001. With all demographic variables held constant, two moral foundations—fairness and purity—uniquely predicted adjusted inequality in both studies.
Individuals who more strongly endorsed the fairness foundation desired a greater reduction in inequality. Individuals who more strongly endorsed the purity foundation showed less of a difference between their predicted and ideal distributions, that is, they desired less of a reduction in inequality. Additional variables became significant in the combined data set: sex (males made smaller adjustments), loyalty, and authority. The full model and individual predictor statistics are summarized in Table 4.
Studies 1a and 1b Linear Regression Analysis.
Note. DV = adjusted inequality (top 20% estimate minus top 20% ideal).
Discussion
Consistent with prior research (Norton & Ariely, 2011), respondents in Studies 1a and 1b greatly underestimated the amount of wealth inequality in the United States, even though these data were collected during an election year where a popular contender for the presidency made wealth inequality a primary focus of his campaign. At the same time, respondents of all backgrounds wanted a society with only a modest amount of inequality.
Studies 1a and 1b also advanced our understanding of the moral underpinnings motivating the desire to adjust wealth distributions to make them more equal. Purity, a binding moral foundation strongly associated with conservatism and religiosity (Franks & Scherr, 2015), uniquely predicted preference for less adjustment of wealth distributions in both studies. Fairness, an individualizing moral foundation associated with liberalism, predicted a desire to adjust the wealth distribution to make it more equal. When the samples were combined, four of the five foundations became unique predictors. The findings are consistent with the general hypothesis regarding the effects of individualizing (desired greater reduction in inequality) and binding (desired less reduction in inequality) foundations on desire to reduce wealth inequality. Because moral foundations uniquely predicted desires to reduce inequality using an adjusted wealth inequality metric, we thought it prudent to examine additional metrics measuring desires for increased equality. We also attempted to further establish that moral foundations uniquely predicted outcomes over-and-above other germane factors by including other-related potential predictors of economic desires.
Study 2
Personally adopting a procedural justice perspective assumes that one could easily improve their socioeconomic status through hard work and talent. Americans, especially poorer individuals and conservatives, overestimate upward social mobility (Davidai & Gilovich, 2015), and this POSM leads individuals to engage in system justification (Day & Fiske, 2017) and is associated with more tolerance for income inequality (Shariff et al., 2016). Accordingly, Study 2 tested whether moral foundations endorsements could predict adjustments to inequality over-and-above other important predictors related to moral foundations and desires to change income inequalities such as POSM and religiosity (which may be related to purity concerns).
In addition, based on the findings that the moral foundations predict the degree to which participants desire to reduce inequality in both Studies 1a and 1b, we sought to include a second inequality metric in Study 2. The second metric we chose to include in Study 2 involved estimating the ratio of CEO-to-worker pay and reporting an ideal CEO-to-worker pay ratio (e.g., Kiatpongsan & Norton, 2016, but in a way that made it more consistent our wealth inequality metric).
Method
Participants
Undergraduates (N = 178) participated for course extra credit. The sample was mostly female (60%) and slightly Republican leaning (53%). The income distribution was lower than that in Studies 1a and 1b. Households making below US$50,000 represented 40% of the sample, while those making between US$50,000 and US$100,000 represented 38% and those making above US$100,000 represented 22%.
Materials and Procedure
Participants completed all measured outlined in the All Studies section of Studies 1a and 1b. A second measure of inequality was assessed in Study 2: estimated and ideal CEO-to-worker income ratios. The estimated ratio was divided by the ideal ratio to create an adjusted income inequality variable. A POSM measure was included by having participants predict what percentage of people born into the lowest wealth quintile actually end up in the top 50% of the wealth distribution. Religiosity was measured using a 9-point Likert-type scale asking participants to rate themselves on a scale of 1 (not at all religious) to 9 (very religious).
Results
Estimates of Inequality
Wealth inequality
Participants underestimated the wealth of the top 20% at only 55.8%. A one-sample t test revealed that the difference between reality and respondent’s estimates was significant, t(176) = −20.16, p < .001. This was true regardless of participant demographics (see Table 5). Comparisons of groups by party affiliation, sex, and household income found only a significant difference for sex whereby men’s estimates of the wealth owned by the top quintile were higher than women’s, t(174) = 5.99, p < .001. Overall, these results are very similar to those found in Studies 1a and 1b.
Study 2 Estimates of the Amount of Wealth Owned by the Top 20% by Demographic.
Note. 95% CI for Sex [−21.3, −10.0].
aSignificant difference between demographic groups in the same row.
CEO-to-worker income inequality
The typical CEO makes 354 times what the typical worker makes. Our participants’ estimates were not significantly different from reality (M = 284.5, SD = 573.4), p = .11. No significant differences were found between demographic groups (see Table 6). The finding that participants are on average better at estimating inequality in this matter than using the Norton and Ariely’s (2011) method is consistent with prior research (e.g., Eriksson & Simpson, 2012). However, one drawback of this measure is extreme variability in responding.
Study 2 Estimates of CEO-to-Worker Pay Ratios.
Ideal Levels of Inequality
Ideal wealth distributions
The amount of wealth that would be owned by the top quintile in participants’ ideal societies (M = 33.0, SD = 15.4) was significantly lower than both the real-world amount, t(175) = −43.9, 95% CI [−53.3, −48.7], p < .001, and their own underestimates, t(175) = −19.8, 95% CI [−25.3, −20.7], p < .001. This was true across demographics (see Table 7) and demonstrates widespread desire for reduced inequality.
Study 2 Ideal Amount of Wealth Owned by the Top 20% by Demographic.
Note. 95% CI for Sex [−14.2, −4.5].
aSignificant difference between demographic groups in the same row.
Ideal pay ratios
The ratio of CEO income to worker income in participants’ ideal societies (M = 76.2, SD = 177.8) was significantly lower than both the real-world amount, t(175) = −20.1, 95% CI [−304.2, −251.4], p < .001, and their own underestimates, t(175) = −15.6, 95% CI [−235.2, −182.4], p < .001. This finding was true across demographics (see Table 8) and provides further evidence that desire for decreased inequality is robust across metrics.
Study 2 Ideal CEO-to-Worker Pay Ratios.
Regression Analyses
Regression analyses were conducted on the same wealth inequality measure as Studies 1a and 1b (adjusted wealth inequality) as well as on a variable computed by dividing the estimated CEO-to-worker income ratio by the ideal ratio (adjusted income inequality). The second outcome variable was computed by division instead of subtraction because the open-ended nature of the CEO pay items resulted in far greater variability of responding for both the estimated and ideal values. Outcomes were regressed on every demographic variable (i.e., party affiliation, sex, and household income), moral foundations (i.e., harm, fairness, loyalty, authority, and purity), and POSM and religiosity.
Adjusted wealth inequality
Background predictors, POSM, religiosity, and the five moral foundations predicted 20% of the variance in adjusted wealth inequality, F(10, 166) = 4.12, p < .001. Several variables uniquely predicted participants’ adjusted inequality: party affiliation, income, POSM, fairness, and purity. Democrats, those who strongly endorsed fairness, and individuals from higher income households made larger adjustments to the wealth distribution. Individuals with higher POSM and endorsement of purity made smaller adjustments to their wealth distributions. These results demonstrate that, although the POSM meaningfully predicts people’s desire to reduce inequality, two moral foundations—fairness and purity—are still able to predict wealth inequality desires over and above other relevant variables. A full summary of the model and all individual predictors is provided in Table 9.
Study 2 Linear Regression Analysis for Adjusted Wealth Inequality.
Note. DV = adjusted wealth inequality (top 20% estimate minus top 20% ideal)
Adjusted income inequality
The combination of predictors explained 15% of the variance in the desire to reduce income inequality, F(10, 165) = 2.87, p = .003. However, only two variables predicted participants’ estimated adjusted pay ratios: fairness and loyalty. Those who more strongly endorsed the fairness foundation made larger adjustments. Conversely, those who more strongly endorsed loyalty made smaller adjustments. Across Studies 1a, 1b, and 2, a concern for fairness consistently predicted a desire for inequality. A full summary of the model and all individual predictors is provided in Table 10.
Study 2 Linear Regression Analysis for Adjusted Income Inequality.
Note. DV = adjusted income inequality (CEO: worker pay estimate/CEO: worker pay ideal).
Discussion
Study 2 further illustrated that respondents in the United States substantially underperceive wealth inequality as operationalized by Norton and Ariely (2011) and want a much more equal society. Participants were on average more accurate in their estimates of CEO-to-worker income inequality. However, these estimates varied wildly, and the ideal CEO-to-worker pay gap was still much lower than both participant estimates and reality. This supports the claim that Americans of all demographics generally want less inequality regardless of the way that question is asked (e.g., Norton & Ariely, 2013).
In addition, moral foundations, particularly fairness and purity, once again predicted desire to reduce wealth inequality. The observed effects further establish that these two moral concerns are central to making moral appraisals of wealth inequality. In addition, although the perceptions of upward mobility were inversely related to desire to reduce inequality, the POSM effects did not suppress fairness’ ability to predict outcomes.
Loyalty, a binding foundation that was not a particularly meaningful predictor in Studies 1a and 1b also predicted people’s desire to reduce income inequality. However, it is unclear why a different binding foundation would emerge as a unique predictor of this variable.
Study 2 largely supported and advanced the findings of Studies 1a and 1b, yet important questions remain. Although moral foundations predicted inequality outcomes over-and-above other politically relevant variables, more precise, continuous measures of political orientation were not used. Also, the question of whether the observed effects can extend to behavior remains. Study 3 attempted to address these issues.
Study 3
Studies 1a, 1b, and 2 were consistent in several key ways. Study 3 built on the consistent findings by including a continuous measure of political orientation. Demonstrating that moral foundations can predict desires to reduce inequality when controlling for a continuous measure of political orientation can further establish the strong influence moral foundations have on important real-world decisions and judgments. Study 3 also attempted to extend the findings of the previous studies using a behavioral outcome among a more representative, nonstudent sample to demonstrate the generalizability of the effects observed in the other studies.
Method
Participants
Workers (N = 198) from Amazon’s Mechanical Turk (mTurk) online workforce participated for US$0.50. The sample was predominantly male (56%) and predominantly Democratic (65%). Households making below US$50,000 represented 53% of the sample, while those making between US$50,000 and US$100,000 represented 36% and those making above US$100,000 represented 11%. As in the prior studies, data collection was maximized over a 1-week period.
Materials and Procedure
Participants completed all measured outlined in the All Studies section of Studies 1a and 1b. At the end of Study 3, participants were asked to donate a portion of their US$0.50 payment to an organization geared toward reducing economic inequality. Participants were randomly assigned to one of the three conditions in which the request was made (1) without moral appeals, (2) with binding moral appeals, or (3) with individualizing moral appeals (see Online Supplemental Material). Participants were then immediately debriefed that the donation request was part of the study and that they would receive their full payment.
Results
Estimated and Ideal Wealth Distributions
Participants again underestimated the wealth of the top quintile (M = 65.33, SD = 22.16), t(197) = −11.86, p < .001, and desired less inequality (M = 29.18, SD = 14.65) even than their own estimates, t(197) = −52.67, p < .001. One important demographic difference was that Democrats (M = 26.75, SD = 12.53) reported a desire for more equal ideal distributions compared to Republicans (M = 33.82, SD = 18.48), t(196) = 3.31, p = .02. The results by demographic categories are summarized in Tables 11 and 12.
Study 3 Estimates of the Amount of Wealth Owned by the Top 20% by Demographic.
Study 3 Ideal Amount of Wealth Owned by the Top 20% by Demographic.
Regression Analyses
Regression analyses were conducted for the adjusted wealth inequality variable and also for the amount of money donated. Predictor variables included sex, income, the continuous measure of political orientation, and the five moral foundations. For the donation variable, a second step was included, wherein two dummy-coded variables (one for the binding appeal condition and one for the individualizing appeal condition) were added along with interaction terms between the moral foundations and the two dummy-coded variables (10 interactions total).
The initial combination of variables predicted 13% of the variance in adjusted wealth inequality, F(8, 185) = 3.47, p = .001, and 11% of the variance in donations, F(8, 185) = 2.82, p = .006. Only purity uniquely predicted adjusted wealth inequality (see Table 13), while both fairness and purity predicted donations (see Table 14). These results again show the strength of moral foundations to predict outcomes (preferences and behaviors) related to economic inequality, even controlling for a continuous measure of political orientation.
Study 3 Hierarchical Linear Regression Analysis for Adjusted Wealth Inequality.
Note. DV = adjusted wealth inequality (top 20% estimate minus top 20% ideal).
Study 3 Hierarchical Linear Regression Analysis for Donations.
Note. DV = donation.
In Step 2 of the model for donations, the dummy variables and the interaction terms predicted an additional 19% of the variance (30% total), ΔF(12, 173) = 3.96, Δp < .001. After adding in the interactions, income became significant (higher incomes predicted lower donations) and fairness became nonsignificant. However, the Loyalty × Binding appeal, Harm × Individualizing appeal, Fairness × Individualizing appeal, and Purity × Individualizing appeal interactions were all significant. Follow-up simple slope analyses indicated that loyalty predicted increased donations when a binding appeal was made and harm, fairness, and purity predicted increased donations when an individualizing condition was made (see Table 14).
Discussion
Study 3 showed once again the tendency of Americans to underperceive wealth inequality and to want less inequality even than their own low estimates among a more representative sample of U.S. citizens. The results also provided further evidence of the ability of moral foundations to predict the desire to reduce economic inequality and, in an important and novel way, provided evidence that moral foundations influence behavior aimed at reducing economic inequality, even when controlling for a continuous measure of political orientation.
In general, the observed interaction effects are consistent with prior research demonstrating an entrenching of moral values under conditions in which an appeal is made to an individual’s preferred foundations. A binding appeal, which included multiple references to Americans and the American Dream, elicited increased donations from individuals with a higher concern with loyalty. An individualizing appeal, which included references to fairness and communicated a desire to reduce the harms caused by inequality, elicited increased donations from individuals with more concern for harm and fairness. However, donations also increased as a function of purity, a binding foundation, in the individualizing appeal condition.
General Discussion
Across multiple studies, we assessed the ability of moral foundations to predict preferences to reduce economic inequality and whether participants’ desires to reduce inequality translated into behaviors. Overall, results suggest that fairness and purity are consistent moral foundations involved in our desire to reduce inequality. Study 2 demonstrated that prospects of upward mobility predicted desire to reduce wealth inequality. Importantly, Study 3 showed that moral foundations also predict behavior related to reducing economic inequality: donating to a charity focused on reducing inequality. The effects observed in Study 3 also illustrate that desires and actions intended to reduce inequality can be elicited with targeted moral appeals, and our research may provide some initial basis on which to base such appeals.
The consistent ability of the fairness foundation to predict outcomes related to inequality should come as no surprise, as massive inequality is on its face a problem caused by a lack of fairness. That fairness consistently and uniquely predicted desire to reduce inequality as well as behavior related to reducing inequality indicates that the moral foundations have validity as unique moral factors able to predict outcomes germane to each factor. However, the effects of purity are less easy to understand. When it came to predicting desire to reduce inequality in Studies 1a, 1b, and 2, purity was antagonistic to fairness. As purity concerns increased, there was more tolerance for inequality. But in the final study, an individualizing moral appeal elicited greater donations from people with higher endorsement of the binding foundation purity (though this effect did not exist at the zero-order level). More consistent with expectations, a binding appeal which in part emphasized helping Americans and restoring the American Dream elicited greater donations from participants who more strongly endorsed the loyalty foundation.
Although we took measures to address weaknesses in any single study by adding elements to subsequent studies, some limitations could not be fully addressed. First, the sample sizes in each study were modest. We attempted to ameliorate this issue by conducting some analyses with a combined data set. Second, some design elements may have detracted somewhat from our novel findings. The income inequality metric in Study 2 may have provided solid evidence that Americans desire reduced inequality, but it was also not entirely consistent with the wealth inequality metric in terms of the effects of moral foundations. In addition, this income inequality metric produced far greater variability in responding. Finally, none of our samples were truly demographically representative of the U.S. population. Three of the four samples were college students and the fourth was only the modestly more representative MTurk workers. However, the fact that our overall numbers in terms of estimated and desired wealth distributions matched up nearly identically with prior research somewhat alleviates the concern.
The United States is a nation with a substantial and growing wealth inequality problem, thereby making it increasingly important to understand how citizens perceive inequality as well as what factors predict a willingness to reduce pervasive inequality. Our results provide further evidence that U.S. citizens, especially those who value fairness and more accurately perceive the lack of opportunities for upward social mobility, want to reduce wealth inequality. Future research should attempt to discover interventions that take advantage of our evolved moral concern with fairness and increase accuracy of beliefs regarding inequality and upward mobility. Finding interventions that increase support for redistributive policies and outrage over inequality may have far-reaching societal benefits including reductions in violent crime, mental health disorders, substance abuse, and other problems associated with massive inequality.
Supplemental Material
Supplemental Material, SPPS772821_suppl_mat - Economic Issues Are Moral Issues: The Moral Underpinnings of the Desire to Reduce Wealth Inequality
Supplemental Material, SPPS772821_suppl_mat for Economic Issues Are Moral Issues: The Moral Underpinnings of the Desire to Reduce Wealth Inequality by Andrew S. Franks and Kyle C. Scherr in Social Psychological and Personality Science
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) received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
The supplemental material is available in the online version of the article.
References
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.
