Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Explanations for Inequality and Partisan Polarization in the U.S., 1980–2020

Explanations for Inequality and Partisan Polarization in the U.S., 1980–2020 It is well-known that American partisans have grown more polarized over the past several decades. This is primarily due to growing ideological uniformity within each of the two parties among both elites and lay people, leading to greater on-average attitudinal differences (McCarty 2019). Increasing within-party uniformity stems from two main sources: people switching parties so that their partisanship better aligns with preexisting political stances (Levendusky 2009)In some cases, people may simply leave a political party that no longer reflects their policy preferences; this is especially relevant to elected officials (who may be forced to “leave” a party by citizens voting them from office).and persuasion processes within parties, led primarily by activists (Layman, Carsey, and Horowitz 2006). Partisan polarization is not only due to increasing homogeneity within the parties; some partisans have also become more extreme over time (Layman, Carsey, and Horowitz 2006). Given these changes, it is not surprising that enmity between the parties has also grown (Mason 2018).The literature on partisan polarization is voluminous, and, although some dissenters (Fiorina 2017) and debates (Krupnikov and Ryan 2022) remain, scholarly conventional wisdom has coalesced around the above conclusions. With this in mind, is there anything left to say about partisan polarization? We believe there is. Scholars who research polarization have almost exclusively focused on the relationship between Americans’ policy opinions and their partisanship. In this article, we discuss a different type of partisan polarization underappreciated by scholars: belief polarization, or disagreements over what people perceive to be true (see, e.g. Lee et al. 2021; O’Connor and Weatherall 2018).Politics is a democratic society’s main venue for normative debates—arguments over what a polity ought to do. Normative proposals that receive enough support from elected officials are enshrined into law, organizing resources and placing limits on behavior for the presumed benefit of citizens. However, these debates over “ought” inevitably involve debates over “is” as well. To win a policy debate, a person proposing a specific law must persuade others that it removes a harm or brings a benefit to people. Such arguments rely heavily on factual assertions about the nature of problems and opportunities as well as on a proposed law’s ability to address them (Suhay 2017).We argue that, as political actors’ policy proposals drift apart substantively—as is the case when polarization grows—so will their accompanying arguments. The debate over the extent of climate change is the most familiar current example of a politicized debate over facts. Democratic politicians have argued that climate change threatens humanity in specific ways and that their proposed policy solutions would remove the threat. Republicans have argued both that Democrats exaggerate the threat of climate change and that their proposed policy solutions threaten to harm the economy (Brewer 2012). Many other debates are laden with factual claims, even deeply moral ones. For example, those defending abortion rights not only champion women’s liberty but also argue that women are harmed when access to abortion is restricted (e.g. they have unsafe abortions); those challenging abortion not only champion the sanctity of life but also argue that women are harmed when they have abortions (e.g. they experience psychological damage) (Siegel 2012). Regardless of the specific argument, differing factual assertions represent an important normative problem in that partisans often cherry-pick evidence that bolsters their point of view or interpret evidence in a biased way. The tendency to transgress norms regarding accurate and unbiased reporting of evidence likely increases as policy debates become more polarized (Parkhurst 2017).In this article, we discuss a set of contested facts that rest at the center of debates over governmental efforts to regulate business, redistribute resources, and provide social welfare. These perceived facts are causal attributions for economic inequality—people’s explanations for why some flourish in the American economy while others struggle to get by (Kluegel and Smith 1986; Suhay, Klašnja, and Rivero 2021).To a significant extent, politics is about “who gets what, when, and how” (Lasswell 1936). While there is always jockeying over who ought to get what (when and how), the stakes of the competition have increased in recent decades as economic inequality has grown. Since the 1980s, incomes among the top 1% have skyrocketed; incomes among the top 10% have risen somewhat; and incomes among the bottom 90% have barely budged (Mishel, Gould, and Bivens 2015; Saez 2018). At the same time, middle- and lower-income Americans receive fewer employer-provided benefits, such as health care and pensions, and face greater instability in their employment (Hacker 2019). These struggles have been exacerbated by two great economic shocks: the Great Recession of 2008 and, more recently, the COVID-19 pandemic.After some decades tacking to the political center on the economy (Hacker and Pierson 2010), the Democratic Party began to shift leftward on the economy in response to these events. Under Obama, the party passed the Affordable Care Act, raised taxes on the affluent, eased student loan burdens, tackled gender discrimination in employment, and tightened financial regulations (Rodriguez 2016). As is increasingly the case, the party would have liked to attain much more than it could accomplish given Republican opposition (Hacker and Pierson 2010). The 2020 Democratic platform under Biden was especially economically progressive (Yglesias 2020)—including proposals to greatly expand social welfare spending (e.g. education, housing, transit), raise taxes on the affluent, and expand tax credits for low- and middle-income Americans.Most political scientists and pundits begin and end their analysis of partisan politics with such proposals and policy changes. We argue that the rhetoric politicians use to persuade people to support such proposals is also worth study. The Democratic Party has long justified its left-leaning economic policies with two central claims: significant economic inequality exists between individuals and social groups, and these great inequalities are unfair because society, not individuals, are to blame for them (Brewer and Stonecash 2015; Gerring 1998). The latter proposition is especially important. It is difficult to deny that many harsh inequalities exist in the United States. Exorbitant wealth as well as homelessness are plain to see. However, such inequalities might be tolerated if they are viewed as the outcome of a meritocratic system. Democrats argue instead that “the American Dream”—success via hard work—is not a reality for many. Thus, low-income people deserve government assistance. As we discuss in the next section, these arguments have intensified in recent years as the Democratic Party has moved left on economic policy.What about the Republican Party? After a period of post-war centrism, the party moved right on the economy under Ronald Reagan and then continued its rightward march for at least three decades. Republican officeholders have loosened regulations, especially over the financial sector, and repeatedly cut taxes for the affluent (Hacker and Pierson 2010). Although their aims with respect to social welfare have often been just as conservative, Republicans have been more successful resisting new social welfare reforms proposed by Democrats than rolling back existing reforms (Grossmann and Hopkins 2016). Despite running a somewhat populist campaign, Trump largely governed in a fiscally conservative manner (Hacker and Pierson 2020).Republicans justify their economically conservative stances with a very different set of factual assertions. First, they tend to emphasize aggregate economic growth and downplay the extent of inequality. Second, Republicans argue that existing inequalities are fair—successful people have achieved success via hard work or ingenuity, and those facing difficult economic circumstances are to blame for them. Third, in response to Democrats’ instinct to use government to combat inequality, Republicans argue government efforts to intervene in business affairs, redistribute wealth, and assist those in need often do more harm than good, depressing the economic output of both firms and individuals (Brewer and Stonecash 2015; Gerring 1998). These narratives justify Republicans’ conservative economic agenda by insisting that the status quo is fine: inequality is minimal; inequalities that do exist are “just deserts”; and, even if one wished to help, government intervention in fact undermines individual and aggregate prosperity.These well-publicized factual narratives are likely influential. If they were not, parties most likely would not bother with them. Members of the public aligned with one of the political parties are enormously responsive to “party cues” (Lenz 2012; Zaller 1992). Even when exposed to information from both sides of the aisle, partisans will gravitate toward views endorsed by co-partisans, including elites and peers (Toff and Suhay 2019). Furthermore, in a fragmented and polarized media environment, those with partisan preferences are more likely than others to receive information from, and in turn be influenced by, those who share their ideological perspectives (Levendusky 2013; Stroud 2011). Most scholars interested in the public’s beliefs about inequality pay little attention to how political communication might affect them. However, Iyengar (1991) finds that Americans’ information environments shape their attributions for poverty. Jones (2020) draws on survey data from over a sixty-year period to demonstrate that partisan polarization over evaluations of the economy mirrors patterns in elite communication.These narratives are likely crucial in keeping the party base “on board” with the parties’ economic agendas and attracting swing voters to the parties. In their comprehensive study of Americans’ beliefs about inequality, Kluegel and Smith (1986) demonstrate a link between blaming society for economic inequality and support for generous social welfare and affirmative action, and vice versa for those who blame individuals (also see Hunt 2007; Hunt and Bullock 2016). Similar relationships have been found internationally (e.g. Piff et al. 2020).Experimental studies demonstrate these attributions can play a causal role in shaping attitudes. Appelbaum (2001) finds that people experimentally assigned to read vignettes about individuals who were poor through no fault of their own were more likely than other study participants to support liberal social policy views at the end of the experiment. Piff et al. (2020) find experimental subjects who played a game that taught them about the various challenges of experiencing poverty increased their situational explanations for poverty and consequently became more supportive of redistribution. In a series of articles based on experimental data, Petersen and colleagues provide evidence for an automatic “deservingness heuristic”—triggered when subjects learn about individuals who are (or are not) trying to help themselves and leading to a desire for more (or less) government assistance for similar individuals (Petersen 2012; Petersen et al. 2010). Many other scholars have converged on the notion that perceived deservingness is the key to generating a motivation to assist an individual or group suffering some type of difficulty (Brewer and Stonecash 2015; Cook and Barrett 1992; Gilens 1999; McCall 2013; Van Oorschot 2006; Weiner, Osborne, and Rudolph 2011).In the sections that follow, we describe an increasing partisan divide in beliefs about whether the American economy rewards merit or, rather, distributes rewards in a biased way. We analyze party platforms from 1980 to 2020 and survey evidence from approximately the same period. Over the past approximately two decades, Democratic party leaders and members of the public have become more pessimistic about the relationship between hard work and mobility and the existence of equal opportunity; Republican party leaders and members of the public have simultaneously become more optimistic, although these shifts are not as large as those that have occurred among Democrats. The size and timing of changes within the public suggest they are caused not only by party switching but also by within-party persuasion. Finally, we provide suggestive evidence that the public’s factual beliefs are politically consequential. Even controlling for partisanship and various demographic characteristics, perceptions of whether the U.S. economy is meritocratic are strongly associated with people’s views on taxation, social welfare policy, affirmative action, and presidential candidates. To the extent that political elites defend their economic policy positions with factual claims, those claims appear to have the intended downstream consequence of attitude change.1Explanations for Inequality in Party Platforms, 1980–2020To begin, we examine how discussions of economic inequality found in the Democratic and Republican party platforms have evolved over time. We focus on party platforms, given both their ability to predict parties’ actual policy initiatives (Gerring 1998) and their encapsulation of party messaging at a given point in time.We started by gathering the text of all party platforms from 1980 to 2020Note that the Republican platforms from 2016 and 2020 are identical. Nevertheless, we treat them as distinct in the dataset, as the Republican Party affirmed its commitment to the 2016 language at its 2020 national convention.from the website of the American Presidency Project at the University of California, Santa Barbara (www.presidency.ucsb.edu). For each party platform, all explanations for economic and related inequalities, such as education or health, were identified and categorized. The categorization scheme sought to distinguish common individual-level or group-level attributions from common societal or institutional attributions. The three individual or group categories are: (1) individual ability or character; (2) individual effort or responsibility; and (3) socialization in the family or community culture. These types of statements tended to “blame” individuals or subsets of the population for inequality. The three societal or institutional categories are: (4) access to educational or job opportunities; (5) discrimination; and, finally, (6) problematic policies or excessive government.After an initial period during which the first two authors read a sample of platforms and then created and revised the coding scheme, the second author read each platform and manually coded attributions for inequality. The first author reviewed all coded sections and suggested changes as necessary.These types of statements tended to “excuse” individuals and groups of responsibility for inequality by blaming society at large or its institutions, although note that statements in the final category ironically point the finger at the very policies intended to help the disadvantaged. See Table 1 for a summary of each coded category.The small proportion (13%) of attributions that did not fit into any of these categories are not included in our analyses.Table 1:Party platform coding scheme.Description of codeIndividual or group blameCharacter & abilityAbility and skill (or other characteristics) lead to success/positive life outcomes or lack thereofEffortEffort and personal responsibility lead to success/positive life outcomes, or lack of effort or personal responsibility leads to lack of success/negative life outcomesCulture & familyFamily or community culture leads to success/positive life outcomes (if good) or inhibits success or leads to negative life outcomes (if bad)Societal or institutional blameAccess to opportunitySociety, its structure, or its institutions are biased and thus directly cause inequality; access to opportunity and/or barriers to success are unequally distributedDiscriminationExplicit mention of inter-individual or systemic discrimination or bias against specific social groups (e.g. race, sex, disability) as a cause of inequalityExcess governmentExcess government or badly designed policies indirectly cause inequalityThe 22 platforms we reviewed contained an average of 28,387 words, ranging from 4838 to 42,092. The platforms averaged 49 attributions, from a low of 6 to a high of 97. However, the number and nature of these attributions varied dramatically by political party and over time.In Figure 1, we visualize the aggregate number of attributions for economic and related inequalities per party, per year.Note that these counts somewhat underrepresent the number of times each party mentions inequality, as platforms sometimes note inequality without providing an accompanying explanation.More often than not, Democrats provided more reasons for inequality than Republicans, reflecting greater concern over inequality. This said, the pattern is not consistent until 2008. Thereafter, Democrats discussed the reasons for inequality more frequently, year over year. Prior to 2008, the parties were more equally matched. (In two unusual years, 1988 and 1992, Democrats’ attention to inequality dropped far below Republicans’; however, this is mainly an artifact of two unusually short Democratic platforms.)Figure 1:Number of times parties offered explanation for inequality in each platform.Next, we examine the specific causal attributions in the platforms. Figure 2 presents the proportion of each coding category per party, per year, out of the total number of coded items. Two general trends are apparent. First, societal/institutional attributions were much more common than individual/group ones. Second, as expected, Republicans mentioned individual/group attributions more often than Democrats, and Democrats mentioned societal/institutional attributions more often than Republicans.Figure 2:Six explanations for economic inequality in the party platforms, 1980–2020.These general trends overlook meaningful nuance. On the Democratic side, one sees heightened attention to “culture and family” precisely as the party tacked to the center (1988, 1992, 1996, 2000). Throughout the period under study, Democrats said relatively little about either effort or individual character; by the 2020 platform, all traces of such individual (and family) blame had disappeared. Moving to societal blame, the Democratic Party has been surprisingly consistent in its focus on disparities in access to opportunity among Americans. Noteworthy declines occurred between 1992 and 2000, as the party shifted somewhat to individual and cultural explanations, and again in 2008 and especially 2020. Interestingly, these last two dips are mostly accounted for by increased attention to discrimination. 2020 is the only year in which Democrats mentioned discrimination more than barriers to opportunity. Throughout the entire period, Democrats rarely criticized government as a cause of inequality, although there was a perceptible increase in the 1996 Clinton platform.Turning to Republicans, one sees a rise and then persistent interest in cultural explanations for inequality (with a notable spike in 2008, when John McCain ran against Barack Obama). Despite Republicans’ reputation as a party that emphasizes “hard work,” this attribution appeared at only a slightly higher rate than in Democratic platforms, although it did increase post-2000. Finally, the old-fashioned emphasis on individual character and ability has played a minimal role in Republican platforms; however, it is of note that the highest levels of attention to this theme were under Trump’s leadership. With respect to blame placed on society and its institutions, the most popular attribution in this category among Republicans is government blame. From 1980 through 1996, over half of Republicans’ attributions for inequality located the cause of inequality in government. Such interest fell sharply under George W. Bush and then increased somewhat under Romney and Trump. Despite their conservatism, Republicans have sometimes discussed uneven access to opportunity—especially under George W. Bush, reflecting his “compassionate conservatism.” Beginning in 2012, this fell sharply. Finally, Republicans did not talk much about discrimination throughout the period we study. Upon examination, the slight increase under Trump was almost entirely due to references to “reverse” discrimination.2Explanations for Inequality in the PublicWe now shift our focus to the American public. To identify trends in how Americans understand economic inequality, we exploit questions from three national survey providers: the American National Election Study (ANES), the General Social Survey (GSS), and Pew Research Center (Pew). To identify questions relevant to our inquiry, we conducted keyword searches of terms associated with causal attributions for inequality on each provider’s website. Note that we only use questions asked in at least three surveys and covering at least approximately a decade. From these searches, we identified the questions below.Beginning with the ANES, we draw on a question asked consistently over three decades. The question asks respondents whether they agree that “one of the big problems in this country is that we don’t give everyone an equal chance” (1984–2012). Respondents were assigned a 1 if they agreed with each statement and a 0 if they disagreed or selected the neutral option.Next, we rely on three questions periodically asked by Pew. The first question asks respondents if “most people who want to get ahead can make it if they’re willing to work hard” (or hard work and determination are no guarantee of success for most people) (1999–2019). The second question asks, “which is generally more often to blame if a person is poor?” Lack of effort on his or her own part (or circumstances beyond his or her control) (2010–2018). Finally, the third question asks whether respondents think “poor people today have it easy because they can get government benefits without doing anything in return” (or poor people have hard lives because government benefits don’t go far enough to help them live decently) (1997–2017).Turning to the GSS, we utilize three items from a battery of questions on Black-White economic disparities (1977–2021). The question battery begins by stating: “On the average, African-Americans [or Blacks] have worse jobs, income, and housing than white people.” This is followed by the question stem “Do you think these differences are … ” and then a series of explanations: (1) Because most African-Americans just don’t have the motivation or will power to pull themselves up out of poverty? (2) Mainly due to discrimination? (3) Because most African-Americans don’t have the chance for education that it takes to rise out of poverty? Respondents were assigned a 1 if they agreed with a given explanation and a 0 if they disagreed.The GSS carries another question on perceptions of fairness in the U.S. economy. The question reads: “Some people say that people get ahead by their own hard work; others say that lucky breaks or help from other people are more important. Which do you think is most important?” We don’t use this question because “luck”—emphasized in the alternative to “hard work”—is an unusually unpopular explanation for inequality in the U.S. and, thus, an inappropriate contrast (McCall 2013; Suhay n.d.).In each of the above cases, our outcomes are constructed as the share of respondents in each cross section who chose, or agreed with, each italicized statement. Additional information about the coding of all three surveys can be found in the appendix. Throughout the analyses that follow, we separate respondents by partisan affiliation (Democrat or Republican). We also display trends for all participants, regardless of how they answered the party affiliation question.We begin with the ANES and Pew questions on economic inequality broadly. Figure 3 displays the percentage of Americans who agreed that “one of the big problems in this country is that we don’t give everyone an equal chance.” Agreement with this question indicates that inequalities in outcomes are the result of some sort of bias and that this problem is important to the respondent. It is worth noting, first, that Democrats and Republicans have forcefully disagreed about the existence (and importance) of a lack of equal opportunity since at least the 1980s—partisans were separated by 23 percentage points in 1984. Over time, Democrats’ endorsement is relatively steady at 60% (except for a brief rise in 1992); Republicans drifted upward initially and then downward after 1992, to 26% in 2012, creating a 35-point difference.Figure 3:Americans’ agreement we don’t give everyone an equal chance.Figure 4 shows the percentage of Americans who chose “most people who want to get ahead can make it if they’re willing to work hard” (versus hard work and determination are no guarantee of success for most people). On this more explicitly factual question, Democrats and Republicans nearly agreed with one another in 1999. At that time, 73% of Democrats agreed as did 81% of Republicans. Soon after, differences emerged. By 2019, Republicans remained at 77% agreement, but Democrats’ agreement fell to 46%, creating a 31-point difference.Figure 4:Americans’ belief most can make it if they work hard.These questions are a useful means to assess the public’s views on the fairness of the American economy in general; however, in each case, the wording leaves uncertain the extent to which people blame or excuse low-income Americans for their economic circumstances. For example, a person might agree “most people” can get ahead via hard work while simultaneously believing some people face tremendous obstacles. Figure 5 provides trends for the question “which is generally more often to blame if a person is poor … lack of effort on his or her part“ (or circumstances beyond his or her control)? In 2010, 50% of Republicans chose effort, compared to 31% of Democrats. Just two years later, Republicans climbed to 64%; after 2014, Democrats dropped to 21%. The result is an approximately 40-percentage point difference. We also point out that, on average, only 37% of Americans chose “lack of effort,” suggesting more sympathy for the poor among Americans than conventional wisdom often suggests.Figure 5:Americans’ belief the poor are poor due to lack of effort.Finally, Figure 6 presents data from a 20-year period, from 1997 to 2017, in response to the question “poor people today have it easy because they can get government benefits without doing anything in return” (or poor people have hard lives because government benefits don’t go far enough to help them live decently). This question mixes both claims about poor people’s work ethic (not “doing anything in return”) with claims about government programs (“benefits do not go far enough”). Likely for this reason, it generates some of the largest partisan differences we see. In 1997, 68% of Republicans chose “have it easy”; 43% of Democrats chose this answer. In the early 2000s, the share of respondents in all groups who agreed trended downward, with these differences remaining. But, beginning in 2004, belief polarization increased steadily through 2017. By the end of the series, 73% of Republicans said the poor “have it easy,” while only 19% of Democrats chose this answer—a difference of 54 percentage points.Figure 6:Americans’ belief poor people who receive government benefits have it easy.Thus far, we have examined questions about inequality and opportunity in the U.S. in general. However, inequality does not impact all groups equally. Black Americans have especially low levels of income and wealth relative to White Americans. While this difference is not new, the reasons offered by the parties for this difference have become more polarized in recent years—a fact evidenced by our platform analysis. Democratic elites tend to argue that discrimination against racial minorities is a major problem; Republican elites often deny this, in some cases arguing that Black Americans receive unfair advantages and implying that White Americans are the less privileged group.The GSS has been asking Americans to explain why “African-Americans [or Blacks] have worse jobs, income, and housing than White people” since the late 1970s. In Figures 7–9, we present the percentage of partisans who endorsed lack of will, lack of education, and discrimination. Across the board, partisan differences in explanations for racial inequality were less than 10 percentage points in 1977 but 3 to 6 times that difference by 2020. With respect to “lack of will,” all partisans have trended downward over time. These trends were relatively parallel through 2008; it is only after Obama’s election that polarization grew, mainly because Democrats declined by 31 points between 2008 and 2020. The resulting partisan difference is over 25 points. On education (where partisans did not differ at all in 1977), polarization gradually grew over time; however, it accelerated around 2012—Democrats’ endorsement of this reason climbed from 50 to 73% in 2020, creating an approximately 40 percentage-point partisan difference. Finally, with respect to discrimination, partisan differences were large (more than 20 points) by the 1990s; but, again, differences took off in the Obama era due to Democratic shifts—here, from 45% in 2012 to 81% in 2020. Republicans, on the other hand, have remained at 20–25% endorsement since 2002. The final partisan difference is nearly 60 points.Figure 7:Belief Black Americans have worse jobs/income/housing due to lack of will.Figure 8:Belief Black Americans have worse jobs/income/housing due to education.Figure 9:Belief Black Americans have worse jobs/income/housing due to discrimination.3The Political Relevance of Explanations for InequalityHaving identified trends in both elite and mass level explanations for inequality, we conclude by examining the association between the American public’s beliefs about inequality and their political attitudes. To do this, we specified linear models in which we regressed various measures of respondents’ economic policy attitudes and presidential voting preferences on their causal attributions, controlling for party, education, income, age, sex, and race. Regressions were run using data from the most recent wave in which both relevant political attitude and causal attribution measures were available.The first question we examine is the ANES question “one of the big problems in this country is that we don’t give everyone an equal chance,” captured on a 5-point Likert scale. For this analysis, we created five outcomes from questions gauging respondents’ political attitudes. First, we created an index of respondents’ support for government spending by creating an additive scale of responses to three questions: support for educational spending, unemployment spending, and spending on the poor (CA = 0.62).“CA” denotes Cronbach’s Alpha, a measure of scale consistency that ranges from 0 to 1.Second, we constructed an index of respondents’ support for government action on health care and jobs by creating an additive scale of two questions asking whether it is the government’s responsibility to provide health care and jobs (CA = 0.67). Third, we measured respondents’ support for the government reducing levels of economic inequality with a single item capturing whether the government should take actions to reduce differences in income levels. Fourth, we measured whether respondents support affirmative action in the workplace for Black Americans using a single item. Fifth, and finally, we measured 2012 vote choice by creating a dichotomous indicator of the two-party vote share (1 = Obama, 0 = Romney).Next, we utilize the “most people can make it if they’re willing to work hard” question from Pew Research Center,We use only this item from Pew for two reasons. First, the question on poor people and welfare benefits (“…have it easy”) invites judgments about government programs, making it inappropriate for these analyses. Second, the most-recent iteration of the question on hard work among the poor (“…lack of effort”) was administered to only a subset of respondents, resulting in an unusually small sample size (N = approximately 200) and few political attitude items.operationalized as a dichotomous indicator of whether respondents agreed (versus disagreed) with this statement. To measure policy attitudes, we included single questions about whether it is the government’s responsibility to help the needy (dichotomous), if government should increase spending on the needy (3-point), if government should increase taxes on the rich (5-point), if government should increase taxes on large corporations (5-point), if government should provide Americans with health coverage (4-point), as well as the political party for which they intended to vote in the 2020 presidential election (two-party vote share).The final causal attribution in our series of analyses is taken from the GSS. It is agreement with societal explanations for racial differences, in which we constructed a count variable ranging from 0 to 3, representing the number of instances in which a respondent agreed with external (discrimination; education) or opposed internal (lack of will) explanations for racial disparities in socioeconomic status (CA = 0.51). We created six outcome variables using questions from the 2018 GSS.GSS analyses use data from 2018 due to problems with 2020 data availability in the cumulative file.First, we included four single items on government’s responsibility to help the poor (5-point), help the sick (3-point), and reduce income differentials (4-point) and on whether government should spend more on education (5-point). Second, we included a single item on respondents’ support for affirmative action in the workplace for Black Americans (4-point). Finally, we measured 2016 vote choice by creating a dichotomous indicator of the two-party presidential vote share.Details regarding how each of these items were measured are included in the appendix. Note that all of the variables in our analyses were recoded to range from 0 to 1 to ease interpretation. In addition, we scored all attribution items such that the societal-blame answers had the highest values and all outcome variables such that the more Democratic, liberal, or progressive answers had the highest values. Thus, we expect positive relationships between the key attributional predictors and the various political attitude outcomes.We begin with Table 2, which displays results for the “equal chance” variable. Across the different models, the coefficient ranges from a low of 0.17 (Obama vote) to a high of 0.28 (government should reduce inequality). All are statistically significant (p < 0.01). In short, those who strongly agreed that not giving everyone an equal chance was a “big problem” were, on average, around 20% higher on the various outcome variables than those who strongly disagreed with this sentiment. Note that the coefficients on partisanship range from −0.19 to −0.34 for the policy items, meaning that the “equal chance” variable is associated with these outcomes nearly as strongly as partisanship. This said, and not surprisingly, the association between partisanship and presidential vote preference is much stronger.Table 2:Importance of equal chance and political attitudes.Outcome variableGov spendingGov jobs/healthGov reduce inequalityAffirmative actionDem vote(1)(2)(3)(4)(5)Equal chance0.218c0.191c0.278c0.187c0.167c(0.009)(0.011)(0.013)(0.013)(0.016)Party (Republican)−0.199c−0.338c−0.263c−0.193c−0.965c(0.008)(0.010)(0.012)(0.011)(0.014)Education−0.036c0.006−0.100c0.051c0.071c(0.010)(0.012)(0.015)(0.014)(0.018)Income−0.078c−0.111c−0.092c−0.069c−0.025(0.010)(0.012)(0.014)(0.014)(0.018)Age−0.029c−0.068c−0.192c0.007−0.077c(0.011)(0.014)(0.016)(0.016)(0.021)Male−0.018c0.007−0.0110.001−0.002(0.005)(0.006)(0.008)(0.007)(0.009)White−0.018c−0.007−0.031c−0.133c−0.082c(0.006)(0.008)(0.009)(0.009)(0.012)Constant0.637c0.600c0.581c0.395c0.969c(0.011)(0.014)(0.016)(0.015)(0.021)Observations51514618517751823982R20.3230.3630.2980.2280.650Adjusted R20.3220.3620.2970.2270.649Residual std. error0.185 (df = 5143)0.216 (df = 4610)0.274 (df = 5169)0.259 (df = 5174)0.290 (df = 3974)Note that the Ns in each column in Table 2 and those that follow vary based on how many respondents took a clear stance on the outcome variable. Respondents who said “don't know” or otherwise did not choose one of the provided options were excluded from analysis. In addition, in some instances, outcome questions were assigned to a random subset of the survey sample. ANES 2012. ap < 0.1; bp < 0.05; cp < 0.01.Moving to Table 3, we regress five specific policy items and Democratic presidential voting onto the Pew item asking people to choose between affirming that most people can make it if they work hard and hard work is no guarantee of success. The coefficients here range from 0.11 (healthcare; rich pay more taxes) to 0.36 (help the needy). Each is again statistically significant (p < 0.01). The coefficients in the policy regressions are about half the strength or more of the partisanship variable.Table 3:Belief hard work no guarantee and political attitudes.Outcome variableGov help if debtGov help needyRich pay moreCorps pay moreGov healthDem vote(1)(2)(3)(4)(5)(6)Hard work no guarantee0.197c0.361c0.111c0.132c0.110c0.149c(0.021)(0.016)(0.008)(0.008)(0.008)(0.007)Party (Republican)−0.408c−0.714c−0.160c−0.222c−0.269c−0.819c(0.026)(0.020)(0.011)(0.010)(0.010)(0.009)Education−0.0450.140c0.028b0.031b0.055c0.101c(0.036)(0.026)(0.014)(0.013)(0.013)(0.012)Income−0.078b−0.472c−0.076c0.001−0.055c0.003(0.039)(0.028)(0.015)(0.015)(0.014)(0.013)Age0.0360.039a0.049c−0.009−0.088c−0.065c(0.029)(0.022)(0.012)(0.012)(0.011)(0.010)Male−0.012−0.048c−0.009−0.0050.011−0.028c(0.020)(0.015)(0.008)(0.008)(0.008)(0.007)White−0.025−0.093c0.076c0.009−0.022b−0.104c(0.023)(0.017)(0.009)(0.009)(0.009)(0.008)Constant0.693c1.596c0.646c0.786c0.788c0.968c(0.032)(0.023)(0.012)(0.012)(0.012)(0.011)Observations209285054178417240247455R20.1900.2890.1310.2060.2830.663Adjusted R20.1870.2880.1300.2050.2820.663Residual std. error0.438 (df = 2084)0.656 (df = 8497)0.244 (df = 4170)0.233 (df = 4164)0.227 (df = 4016)0.270 (df = 7447)Pew American Trends 53 (2019). ap < 0.1; bp < 0.05; cp < 0.01.Finally, in Table 4, we regress similar items as in the ANES analyses on the GSS racial inequality attribution measurement scale. The coefficients vary more here, from a low of 0.04 in the education model to highs of 0.25 in the affirmative action model and 0.29 in the voting model. But, in the main, the effect sizes are similar to what we saw in the Pew analysis.Table 4:Blame society for racial inequality and political attitudes.Outcome variableGov help poorGov healthSpending on educationGov reduce inequalityAffirmative actionDem vote(1)(2)(3)(4)(5)(6)Racial differences0.179c0.188c0.0400.109b0.250c0.288c(0.034)(0.037)(0.037)(0.046)(0.031)(0.043)Party (Republican)−0.244c−0.266c−0.123c−0.260c−0.199c−0.820c(0.041)(0.045)(0.044)(0.054)(0.036)(0.049)Education−0.110c−0.035−0.007−0.167c−0.091b0.042(0.042)(0.046)(0.045)(0.053)(0.037)(0.049)Income−0.052−0.006−0.0230.050−0.174c−0.179b(0.057)(0.063)(0.063)(0.082)(0.054)(0.086)Age−0.053−0.066−0.149c−0.209c−0.032−0.126b(0.051)(0.055)(0.054)(0.067)(0.044)(0.063)Male−0.031−0.065b−0.028−0.0110.013−0.048a(0.024)(0.026)(0.025)(0.031)(0.021)(0.029)White−0.037−0.0310.019−0.059−0.089c−0.235c(0.028)(0.031)(0.030)(0.040)(0.025)(0.039)Constant0.654c0.734c0.947c0.602c0.518c1.087c(0.061)(0.067)(0.067)(0.092)(0.058)(0.098)Observations499499509246967548R20.1830.1750.0480.2520.1670.555Adjusted R20.1710.1640.0340.2300.1600.550Residual std. error0.259 (df = 491)0.284 (df = 491)0.279 (df = 501)0.233 (df = 238)0.317 (df = 959)0.325 (df = 540)GSS 2018. ap < 0.1; bp < 0.05; cp < 0.01.4Discussion and ConclusionWe have argued that beliefs about the American economy—in particular, whether it is meritocratic and, therefore, fairAmericans tend to equate meritocracy with fairness; however, it is important to acknowledge that the two are not interchangeable. For example, Sandel (2020) argues that meritocratic systems can include excessive pay differentials and also reward those who do not provide value to society or who were lucky enough to inherit their talents.or instead biased and unfair—represent an important rift between the two parties. In national platforms since 1980, Democratic and Republican party elites have consistently disagreed as to the reasons why some people succeed and others struggle, and this disagreement has grown over the past two decades or so. The Democratic Party is especially notable for its recent emphasis on discrimination as an important cause of economic inequality and its omission in 2020 of any role individuals, families, and communities might play in inequality. The Republican Party has shifted over time from blaming inequality on government dysfunction to, more recently, a mix of government and person-focused blame. Despite the sympathies expressed by Trump for working class Americans, the Republican platform during his two runs for the presidency included surprisingly little language about barriers to advancement lower-income people may face, such as lack of job and educational opportunities.Partisan members of the public mirror their party’s platforms. Democratic citizens are pessimistic about whether “the American Dream” is a reality for most people, and they tend to avoid blaming the poor for falling behind. Republicans are quite optimistic and tend to blame the poor. With respect to the economic difficulties that Black Americans disproportionately face, Democrats are especially concerned that these difficulties are undeserved, whereas Republicans are skeptical of explanations that deny Black Americans are responsible. These partisan differences have grown larger over the period we examine in every case, especially so for racial inequality. This said, the timing of increasing polarization differs depending on whether survey respondents are focused on explaining economic inequality in general or between Black and White Americans. In the former case, polarization grew during the George W. Bush era or earlier; in the latter case, polarization accelerated after Obama’s election. While much scholarship has focused on increases in racial prejudice among Republicans at this juncture, our data show that increasing partisan disagreement in recent years over the causes of racial inequality has been driven mainly by Democrats, especially accelerating since Trump’s entry into presidential politics.The partisan differences across all of the questions we examine are large. In 2020, on-average differences between Democrats and Republicans range from a low of approximately 30 points to a high of nearly 60 points. To appreciate the size of these differences, consider that partisan disagreements on climate change and abortion—two of the most contentious issues in American politics today—are in the same range, depending on the specific topic and question wording (Hartig 2022; Kennedy and Johnson 2020). Note also that partisan polarization over the causes of inequality is not driven by the different racial composition of the party coalitions; in fact, if we examine the over-time trends among Whites only, between-party differences are generally larger than those in our aggregate figures.Results not shown and available from authors upon request.Why do these large differences exist? It is unwise to draw inferences about causal mechanisms from simple correlations; however, the changes in polarization that we observe over time can give us some purchase on this question. Some of the changes are likely due to party sorting—liberals moving into the Democratic Party, and conservatives moving into the Republican Party. Sorting on Racial Resentment, a measure that overlaps substantively with the GSS racial inequality items, continued through the election of Donald Trump (Abramowitz and McCoy 2018; Enders and Scott 2018). Yet, some of the changes we observe—especially between 2010 and 2020 and on beliefs about racial inequality—are too large and swift to be explained by party switching. Furthermore, it is during this period that arguments about economic and racial injustice became more common among Democratic activists and politicians, as reflected in our platform analysis. This suggests that, to some degree, partisans in the public are persuaded by political elites espousing beliefs about the nature of opportunity in the U.S. that justify their conservative or liberal policy positions. We don’t mean to suggest that most Americans read party platforms; rather, the factual claims captured in the platforms are conveyed via a variety of channels, from candidate speeches to interest group publicity to traditional and social media.As other scholars have documented, these beliefs appear to be consequential politically. Depending on the specific way in which beliefs about economic inequality are measured, their association with various economically progressive policy measures ranges from half to equal the effect of partisanship; explanations for inequality are also associated with presidential preferences, although the associations in this instance are dwarfed by that of partisanship. The policy relevance of beliefs about economic inequality in general and between Black and White Americans is similar. One might ask what would happen if we included them both on the right-hand side of the regression equations. Would explanations for generic or racial economic inequality matter more to political views? Unfortunately, we do not have the equivalent question wording in any one survey that would allow us to test this. However, in a working manuscript, we analyze comparable questions on generic and racial economic inequality (Suhay n.d.). There, we demonstrate that beliefs about the causes of economic inequality in general are more predictive of attitudes toward social welfare—contra Gilens (1995) and Nelson (1999)—whereas beliefs about the causes of racial inequality are more predictive of presidential candidate preferences.These consequential differences in factual beliefs are an underappreciated aspect of party polarization. To our knowledge, we are the first authors to provide such a broad and systematic analysis of changing attributions for inequality over time among partisans. We believe the implications are profound. Our analyses suggest there exists a “redistributive versus anti-redistributive” ideology that lies at the center of political debates over economic policy. Most likely, it originates among activists, politicians, and other political elites and is passed to broad swaths of the public. The notion that the public is “innocent of ideology” (Converse 1964; Kinder and Kalmoe 2017) simply is not the case in this domain: partisanship, policy preferences, and supporting factual narratives are all closely associated.This said, ideological thinking in this instance is nothing to celebrate. Debates over meritocracy are not merely “differences of opinion”; they reflect wildly different perceptions of ordinary Americans’ day-to-day realities and how resources are gained and lost. While our society is complex, there are truths about the extent to which the U.S. lives up to its reputation as a place where people can achieve “the American Dream” through hard work and the conditions under which the Dream is more or less likely. The political parties and their allies, with their strong incentives to bend the truth to fit their policy agendas, are not the ones we should trust most to convey this reality. As with the similarly politicized “debate” over climate change, impartial expertsFor example, the United Nations’ Intergovernmental Panel on Climate Change (www.ipcc.ch) and the United States’ Global Change Research Program (www.globalchange.gov/nca4) publish well-publicized and influential empirical reports on climate change prepared by large teams of interdisciplinary experts.are more trustworthy voices on the empirical question of opportunity in the U.S. If citizens could develop a shared understanding of the ways in which the American economy fails to deliver meritocratic outcomes, perhaps they would demand the two parties find common ground on policies designed to combat inequality. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Forum de Gruyter

Explanations for Inequality and Partisan Polarization in the U.S., 1980–2020

Suhay, Elizabeth; Tenenbaum, Mark; Bartola, Austin
The Forum , Volume 20 (1): 32 – Apr 1, 2022
Download PDF
32 pages

Loading next page...
 
/lp/de-gruyter/explanations-for-inequality-and-partisan-polarization-in-the-u-s-1980-CMOvPWE0bh

References (28)

Publisher
de Gruyter
Copyright
© 2022 the author(s), published by De Gruyter, Berlin/Boston
ISSN
1540-8884
eISSN
1540-8884
DOI
10.1515/for-2022-2052
Publisher site
See Article on Publisher Site

Abstract

It is well-known that American partisans have grown more polarized over the past several decades. This is primarily due to growing ideological uniformity within each of the two parties among both elites and lay people, leading to greater on-average attitudinal differences (McCarty 2019). Increasing within-party uniformity stems from two main sources: people switching parties so that their partisanship better aligns with preexisting political stances (Levendusky 2009)In some cases, people may simply leave a political party that no longer reflects their policy preferences; this is especially relevant to elected officials (who may be forced to “leave” a party by citizens voting them from office).and persuasion processes within parties, led primarily by activists (Layman, Carsey, and Horowitz 2006). Partisan polarization is not only due to increasing homogeneity within the parties; some partisans have also become more extreme over time (Layman, Carsey, and Horowitz 2006). Given these changes, it is not surprising that enmity between the parties has also grown (Mason 2018).The literature on partisan polarization is voluminous, and, although some dissenters (Fiorina 2017) and debates (Krupnikov and Ryan 2022) remain, scholarly conventional wisdom has coalesced around the above conclusions. With this in mind, is there anything left to say about partisan polarization? We believe there is. Scholars who research polarization have almost exclusively focused on the relationship between Americans’ policy opinions and their partisanship. In this article, we discuss a different type of partisan polarization underappreciated by scholars: belief polarization, or disagreements over what people perceive to be true (see, e.g. Lee et al. 2021; O’Connor and Weatherall 2018).Politics is a democratic society’s main venue for normative debates—arguments over what a polity ought to do. Normative proposals that receive enough support from elected officials are enshrined into law, organizing resources and placing limits on behavior for the presumed benefit of citizens. However, these debates over “ought” inevitably involve debates over “is” as well. To win a policy debate, a person proposing a specific law must persuade others that it removes a harm or brings a benefit to people. Such arguments rely heavily on factual assertions about the nature of problems and opportunities as well as on a proposed law’s ability to address them (Suhay 2017).We argue that, as political actors’ policy proposals drift apart substantively—as is the case when polarization grows—so will their accompanying arguments. The debate over the extent of climate change is the most familiar current example of a politicized debate over facts. Democratic politicians have argued that climate change threatens humanity in specific ways and that their proposed policy solutions would remove the threat. Republicans have argued both that Democrats exaggerate the threat of climate change and that their proposed policy solutions threaten to harm the economy (Brewer 2012). Many other debates are laden with factual claims, even deeply moral ones. For example, those defending abortion rights not only champion women’s liberty but also argue that women are harmed when access to abortion is restricted (e.g. they have unsafe abortions); those challenging abortion not only champion the sanctity of life but also argue that women are harmed when they have abortions (e.g. they experience psychological damage) (Siegel 2012). Regardless of the specific argument, differing factual assertions represent an important normative problem in that partisans often cherry-pick evidence that bolsters their point of view or interpret evidence in a biased way. The tendency to transgress norms regarding accurate and unbiased reporting of evidence likely increases as policy debates become more polarized (Parkhurst 2017).In this article, we discuss a set of contested facts that rest at the center of debates over governmental efforts to regulate business, redistribute resources, and provide social welfare. These perceived facts are causal attributions for economic inequality—people’s explanations for why some flourish in the American economy while others struggle to get by (Kluegel and Smith 1986; Suhay, Klašnja, and Rivero 2021).To a significant extent, politics is about “who gets what, when, and how” (Lasswell 1936). While there is always jockeying over who ought to get what (when and how), the stakes of the competition have increased in recent decades as economic inequality has grown. Since the 1980s, incomes among the top 1% have skyrocketed; incomes among the top 10% have risen somewhat; and incomes among the bottom 90% have barely budged (Mishel, Gould, and Bivens 2015; Saez 2018). At the same time, middle- and lower-income Americans receive fewer employer-provided benefits, such as health care and pensions, and face greater instability in their employment (Hacker 2019). These struggles have been exacerbated by two great economic shocks: the Great Recession of 2008 and, more recently, the COVID-19 pandemic.After some decades tacking to the political center on the economy (Hacker and Pierson 2010), the Democratic Party began to shift leftward on the economy in response to these events. Under Obama, the party passed the Affordable Care Act, raised taxes on the affluent, eased student loan burdens, tackled gender discrimination in employment, and tightened financial regulations (Rodriguez 2016). As is increasingly the case, the party would have liked to attain much more than it could accomplish given Republican opposition (Hacker and Pierson 2010). The 2020 Democratic platform under Biden was especially economically progressive (Yglesias 2020)—including proposals to greatly expand social welfare spending (e.g. education, housing, transit), raise taxes on the affluent, and expand tax credits for low- and middle-income Americans.Most political scientists and pundits begin and end their analysis of partisan politics with such proposals and policy changes. We argue that the rhetoric politicians use to persuade people to support such proposals is also worth study. The Democratic Party has long justified its left-leaning economic policies with two central claims: significant economic inequality exists between individuals and social groups, and these great inequalities are unfair because society, not individuals, are to blame for them (Brewer and Stonecash 2015; Gerring 1998). The latter proposition is especially important. It is difficult to deny that many harsh inequalities exist in the United States. Exorbitant wealth as well as homelessness are plain to see. However, such inequalities might be tolerated if they are viewed as the outcome of a meritocratic system. Democrats argue instead that “the American Dream”—success via hard work—is not a reality for many. Thus, low-income people deserve government assistance. As we discuss in the next section, these arguments have intensified in recent years as the Democratic Party has moved left on economic policy.What about the Republican Party? After a period of post-war centrism, the party moved right on the economy under Ronald Reagan and then continued its rightward march for at least three decades. Republican officeholders have loosened regulations, especially over the financial sector, and repeatedly cut taxes for the affluent (Hacker and Pierson 2010). Although their aims with respect to social welfare have often been just as conservative, Republicans have been more successful resisting new social welfare reforms proposed by Democrats than rolling back existing reforms (Grossmann and Hopkins 2016). Despite running a somewhat populist campaign, Trump largely governed in a fiscally conservative manner (Hacker and Pierson 2020).Republicans justify their economically conservative stances with a very different set of factual assertions. First, they tend to emphasize aggregate economic growth and downplay the extent of inequality. Second, Republicans argue that existing inequalities are fair—successful people have achieved success via hard work or ingenuity, and those facing difficult economic circumstances are to blame for them. Third, in response to Democrats’ instinct to use government to combat inequality, Republicans argue government efforts to intervene in business affairs, redistribute wealth, and assist those in need often do more harm than good, depressing the economic output of both firms and individuals (Brewer and Stonecash 2015; Gerring 1998). These narratives justify Republicans’ conservative economic agenda by insisting that the status quo is fine: inequality is minimal; inequalities that do exist are “just deserts”; and, even if one wished to help, government intervention in fact undermines individual and aggregate prosperity.These well-publicized factual narratives are likely influential. If they were not, parties most likely would not bother with them. Members of the public aligned with one of the political parties are enormously responsive to “party cues” (Lenz 2012; Zaller 1992). Even when exposed to information from both sides of the aisle, partisans will gravitate toward views endorsed by co-partisans, including elites and peers (Toff and Suhay 2019). Furthermore, in a fragmented and polarized media environment, those with partisan preferences are more likely than others to receive information from, and in turn be influenced by, those who share their ideological perspectives (Levendusky 2013; Stroud 2011). Most scholars interested in the public’s beliefs about inequality pay little attention to how political communication might affect them. However, Iyengar (1991) finds that Americans’ information environments shape their attributions for poverty. Jones (2020) draws on survey data from over a sixty-year period to demonstrate that partisan polarization over evaluations of the economy mirrors patterns in elite communication.These narratives are likely crucial in keeping the party base “on board” with the parties’ economic agendas and attracting swing voters to the parties. In their comprehensive study of Americans’ beliefs about inequality, Kluegel and Smith (1986) demonstrate a link between blaming society for economic inequality and support for generous social welfare and affirmative action, and vice versa for those who blame individuals (also see Hunt 2007; Hunt and Bullock 2016). Similar relationships have been found internationally (e.g. Piff et al. 2020).Experimental studies demonstrate these attributions can play a causal role in shaping attitudes. Appelbaum (2001) finds that people experimentally assigned to read vignettes about individuals who were poor through no fault of their own were more likely than other study participants to support liberal social policy views at the end of the experiment. Piff et al. (2020) find experimental subjects who played a game that taught them about the various challenges of experiencing poverty increased their situational explanations for poverty and consequently became more supportive of redistribution. In a series of articles based on experimental data, Petersen and colleagues provide evidence for an automatic “deservingness heuristic”—triggered when subjects learn about individuals who are (or are not) trying to help themselves and leading to a desire for more (or less) government assistance for similar individuals (Petersen 2012; Petersen et al. 2010). Many other scholars have converged on the notion that perceived deservingness is the key to generating a motivation to assist an individual or group suffering some type of difficulty (Brewer and Stonecash 2015; Cook and Barrett 1992; Gilens 1999; McCall 2013; Van Oorschot 2006; Weiner, Osborne, and Rudolph 2011).In the sections that follow, we describe an increasing partisan divide in beliefs about whether the American economy rewards merit or, rather, distributes rewards in a biased way. We analyze party platforms from 1980 to 2020 and survey evidence from approximately the same period. Over the past approximately two decades, Democratic party leaders and members of the public have become more pessimistic about the relationship between hard work and mobility and the existence of equal opportunity; Republican party leaders and members of the public have simultaneously become more optimistic, although these shifts are not as large as those that have occurred among Democrats. The size and timing of changes within the public suggest they are caused not only by party switching but also by within-party persuasion. Finally, we provide suggestive evidence that the public’s factual beliefs are politically consequential. Even controlling for partisanship and various demographic characteristics, perceptions of whether the U.S. economy is meritocratic are strongly associated with people’s views on taxation, social welfare policy, affirmative action, and presidential candidates. To the extent that political elites defend their economic policy positions with factual claims, those claims appear to have the intended downstream consequence of attitude change.1Explanations for Inequality in Party Platforms, 1980–2020To begin, we examine how discussions of economic inequality found in the Democratic and Republican party platforms have evolved over time. We focus on party platforms, given both their ability to predict parties’ actual policy initiatives (Gerring 1998) and their encapsulation of party messaging at a given point in time.We started by gathering the text of all party platforms from 1980 to 2020Note that the Republican platforms from 2016 and 2020 are identical. Nevertheless, we treat them as distinct in the dataset, as the Republican Party affirmed its commitment to the 2016 language at its 2020 national convention.from the website of the American Presidency Project at the University of California, Santa Barbara (www.presidency.ucsb.edu). For each party platform, all explanations for economic and related inequalities, such as education or health, were identified and categorized. The categorization scheme sought to distinguish common individual-level or group-level attributions from common societal or institutional attributions. The three individual or group categories are: (1) individual ability or character; (2) individual effort or responsibility; and (3) socialization in the family or community culture. These types of statements tended to “blame” individuals or subsets of the population for inequality. The three societal or institutional categories are: (4) access to educational or job opportunities; (5) discrimination; and, finally, (6) problematic policies or excessive government.After an initial period during which the first two authors read a sample of platforms and then created and revised the coding scheme, the second author read each platform and manually coded attributions for inequality. The first author reviewed all coded sections and suggested changes as necessary.These types of statements tended to “excuse” individuals and groups of responsibility for inequality by blaming society at large or its institutions, although note that statements in the final category ironically point the finger at the very policies intended to help the disadvantaged. See Table 1 for a summary of each coded category.The small proportion (13%) of attributions that did not fit into any of these categories are not included in our analyses.Table 1:Party platform coding scheme.Description of codeIndividual or group blameCharacter & abilityAbility and skill (or other characteristics) lead to success/positive life outcomes or lack thereofEffortEffort and personal responsibility lead to success/positive life outcomes, or lack of effort or personal responsibility leads to lack of success/negative life outcomesCulture & familyFamily or community culture leads to success/positive life outcomes (if good) or inhibits success or leads to negative life outcomes (if bad)Societal or institutional blameAccess to opportunitySociety, its structure, or its institutions are biased and thus directly cause inequality; access to opportunity and/or barriers to success are unequally distributedDiscriminationExplicit mention of inter-individual or systemic discrimination or bias against specific social groups (e.g. race, sex, disability) as a cause of inequalityExcess governmentExcess government or badly designed policies indirectly cause inequalityThe 22 platforms we reviewed contained an average of 28,387 words, ranging from 4838 to 42,092. The platforms averaged 49 attributions, from a low of 6 to a high of 97. However, the number and nature of these attributions varied dramatically by political party and over time.In Figure 1, we visualize the aggregate number of attributions for economic and related inequalities per party, per year.Note that these counts somewhat underrepresent the number of times each party mentions inequality, as platforms sometimes note inequality without providing an accompanying explanation.More often than not, Democrats provided more reasons for inequality than Republicans, reflecting greater concern over inequality. This said, the pattern is not consistent until 2008. Thereafter, Democrats discussed the reasons for inequality more frequently, year over year. Prior to 2008, the parties were more equally matched. (In two unusual years, 1988 and 1992, Democrats’ attention to inequality dropped far below Republicans’; however, this is mainly an artifact of two unusually short Democratic platforms.)Figure 1:Number of times parties offered explanation for inequality in each platform.Next, we examine the specific causal attributions in the platforms. Figure 2 presents the proportion of each coding category per party, per year, out of the total number of coded items. Two general trends are apparent. First, societal/institutional attributions were much more common than individual/group ones. Second, as expected, Republicans mentioned individual/group attributions more often than Democrats, and Democrats mentioned societal/institutional attributions more often than Republicans.Figure 2:Six explanations for economic inequality in the party platforms, 1980–2020.These general trends overlook meaningful nuance. On the Democratic side, one sees heightened attention to “culture and family” precisely as the party tacked to the center (1988, 1992, 1996, 2000). Throughout the period under study, Democrats said relatively little about either effort or individual character; by the 2020 platform, all traces of such individual (and family) blame had disappeared. Moving to societal blame, the Democratic Party has been surprisingly consistent in its focus on disparities in access to opportunity among Americans. Noteworthy declines occurred between 1992 and 2000, as the party shifted somewhat to individual and cultural explanations, and again in 2008 and especially 2020. Interestingly, these last two dips are mostly accounted for by increased attention to discrimination. 2020 is the only year in which Democrats mentioned discrimination more than barriers to opportunity. Throughout the entire period, Democrats rarely criticized government as a cause of inequality, although there was a perceptible increase in the 1996 Clinton platform.Turning to Republicans, one sees a rise and then persistent interest in cultural explanations for inequality (with a notable spike in 2008, when John McCain ran against Barack Obama). Despite Republicans’ reputation as a party that emphasizes “hard work,” this attribution appeared at only a slightly higher rate than in Democratic platforms, although it did increase post-2000. Finally, the old-fashioned emphasis on individual character and ability has played a minimal role in Republican platforms; however, it is of note that the highest levels of attention to this theme were under Trump’s leadership. With respect to blame placed on society and its institutions, the most popular attribution in this category among Republicans is government blame. From 1980 through 1996, over half of Republicans’ attributions for inequality located the cause of inequality in government. Such interest fell sharply under George W. Bush and then increased somewhat under Romney and Trump. Despite their conservatism, Republicans have sometimes discussed uneven access to opportunity—especially under George W. Bush, reflecting his “compassionate conservatism.” Beginning in 2012, this fell sharply. Finally, Republicans did not talk much about discrimination throughout the period we study. Upon examination, the slight increase under Trump was almost entirely due to references to “reverse” discrimination.2Explanations for Inequality in the PublicWe now shift our focus to the American public. To identify trends in how Americans understand economic inequality, we exploit questions from three national survey providers: the American National Election Study (ANES), the General Social Survey (GSS), and Pew Research Center (Pew). To identify questions relevant to our inquiry, we conducted keyword searches of terms associated with causal attributions for inequality on each provider’s website. Note that we only use questions asked in at least three surveys and covering at least approximately a decade. From these searches, we identified the questions below.Beginning with the ANES, we draw on a question asked consistently over three decades. The question asks respondents whether they agree that “one of the big problems in this country is that we don’t give everyone an equal chance” (1984–2012). Respondents were assigned a 1 if they agreed with each statement and a 0 if they disagreed or selected the neutral option.Next, we rely on three questions periodically asked by Pew. The first question asks respondents if “most people who want to get ahead can make it if they’re willing to work hard” (or hard work and determination are no guarantee of success for most people) (1999–2019). The second question asks, “which is generally more often to blame if a person is poor?” Lack of effort on his or her own part (or circumstances beyond his or her control) (2010–2018). Finally, the third question asks whether respondents think “poor people today have it easy because they can get government benefits without doing anything in return” (or poor people have hard lives because government benefits don’t go far enough to help them live decently) (1997–2017).Turning to the GSS, we utilize three items from a battery of questions on Black-White economic disparities (1977–2021). The question battery begins by stating: “On the average, African-Americans [or Blacks] have worse jobs, income, and housing than white people.” This is followed by the question stem “Do you think these differences are … ” and then a series of explanations: (1) Because most African-Americans just don’t have the motivation or will power to pull themselves up out of poverty? (2) Mainly due to discrimination? (3) Because most African-Americans don’t have the chance for education that it takes to rise out of poverty? Respondents were assigned a 1 if they agreed with a given explanation and a 0 if they disagreed.The GSS carries another question on perceptions of fairness in the U.S. economy. The question reads: “Some people say that people get ahead by their own hard work; others say that lucky breaks or help from other people are more important. Which do you think is most important?” We don’t use this question because “luck”—emphasized in the alternative to “hard work”—is an unusually unpopular explanation for inequality in the U.S. and, thus, an inappropriate contrast (McCall 2013; Suhay n.d.).In each of the above cases, our outcomes are constructed as the share of respondents in each cross section who chose, or agreed with, each italicized statement. Additional information about the coding of all three surveys can be found in the appendix. Throughout the analyses that follow, we separate respondents by partisan affiliation (Democrat or Republican). We also display trends for all participants, regardless of how they answered the party affiliation question.We begin with the ANES and Pew questions on economic inequality broadly. Figure 3 displays the percentage of Americans who agreed that “one of the big problems in this country is that we don’t give everyone an equal chance.” Agreement with this question indicates that inequalities in outcomes are the result of some sort of bias and that this problem is important to the respondent. It is worth noting, first, that Democrats and Republicans have forcefully disagreed about the existence (and importance) of a lack of equal opportunity since at least the 1980s—partisans were separated by 23 percentage points in 1984. Over time, Democrats’ endorsement is relatively steady at 60% (except for a brief rise in 1992); Republicans drifted upward initially and then downward after 1992, to 26% in 2012, creating a 35-point difference.Figure 3:Americans’ agreement we don’t give everyone an equal chance.Figure 4 shows the percentage of Americans who chose “most people who want to get ahead can make it if they’re willing to work hard” (versus hard work and determination are no guarantee of success for most people). On this more explicitly factual question, Democrats and Republicans nearly agreed with one another in 1999. At that time, 73% of Democrats agreed as did 81% of Republicans. Soon after, differences emerged. By 2019, Republicans remained at 77% agreement, but Democrats’ agreement fell to 46%, creating a 31-point difference.Figure 4:Americans’ belief most can make it if they work hard.These questions are a useful means to assess the public’s views on the fairness of the American economy in general; however, in each case, the wording leaves uncertain the extent to which people blame or excuse low-income Americans for their economic circumstances. For example, a person might agree “most people” can get ahead via hard work while simultaneously believing some people face tremendous obstacles. Figure 5 provides trends for the question “which is generally more often to blame if a person is poor … lack of effort on his or her part“ (or circumstances beyond his or her control)? In 2010, 50% of Republicans chose effort, compared to 31% of Democrats. Just two years later, Republicans climbed to 64%; after 2014, Democrats dropped to 21%. The result is an approximately 40-percentage point difference. We also point out that, on average, only 37% of Americans chose “lack of effort,” suggesting more sympathy for the poor among Americans than conventional wisdom often suggests.Figure 5:Americans’ belief the poor are poor due to lack of effort.Finally, Figure 6 presents data from a 20-year period, from 1997 to 2017, in response to the question “poor people today have it easy because they can get government benefits without doing anything in return” (or poor people have hard lives because government benefits don’t go far enough to help them live decently). This question mixes both claims about poor people’s work ethic (not “doing anything in return”) with claims about government programs (“benefits do not go far enough”). Likely for this reason, it generates some of the largest partisan differences we see. In 1997, 68% of Republicans chose “have it easy”; 43% of Democrats chose this answer. In the early 2000s, the share of respondents in all groups who agreed trended downward, with these differences remaining. But, beginning in 2004, belief polarization increased steadily through 2017. By the end of the series, 73% of Republicans said the poor “have it easy,” while only 19% of Democrats chose this answer—a difference of 54 percentage points.Figure 6:Americans’ belief poor people who receive government benefits have it easy.Thus far, we have examined questions about inequality and opportunity in the U.S. in general. However, inequality does not impact all groups equally. Black Americans have especially low levels of income and wealth relative to White Americans. While this difference is not new, the reasons offered by the parties for this difference have become more polarized in recent years—a fact evidenced by our platform analysis. Democratic elites tend to argue that discrimination against racial minorities is a major problem; Republican elites often deny this, in some cases arguing that Black Americans receive unfair advantages and implying that White Americans are the less privileged group.The GSS has been asking Americans to explain why “African-Americans [or Blacks] have worse jobs, income, and housing than White people” since the late 1970s. In Figures 7–9, we present the percentage of partisans who endorsed lack of will, lack of education, and discrimination. Across the board, partisan differences in explanations for racial inequality were less than 10 percentage points in 1977 but 3 to 6 times that difference by 2020. With respect to “lack of will,” all partisans have trended downward over time. These trends were relatively parallel through 2008; it is only after Obama’s election that polarization grew, mainly because Democrats declined by 31 points between 2008 and 2020. The resulting partisan difference is over 25 points. On education (where partisans did not differ at all in 1977), polarization gradually grew over time; however, it accelerated around 2012—Democrats’ endorsement of this reason climbed from 50 to 73% in 2020, creating an approximately 40 percentage-point partisan difference. Finally, with respect to discrimination, partisan differences were large (more than 20 points) by the 1990s; but, again, differences took off in the Obama era due to Democratic shifts—here, from 45% in 2012 to 81% in 2020. Republicans, on the other hand, have remained at 20–25% endorsement since 2002. The final partisan difference is nearly 60 points.Figure 7:Belief Black Americans have worse jobs/income/housing due to lack of will.Figure 8:Belief Black Americans have worse jobs/income/housing due to education.Figure 9:Belief Black Americans have worse jobs/income/housing due to discrimination.3The Political Relevance of Explanations for InequalityHaving identified trends in both elite and mass level explanations for inequality, we conclude by examining the association between the American public’s beliefs about inequality and their political attitudes. To do this, we specified linear models in which we regressed various measures of respondents’ economic policy attitudes and presidential voting preferences on their causal attributions, controlling for party, education, income, age, sex, and race. Regressions were run using data from the most recent wave in which both relevant political attitude and causal attribution measures were available.The first question we examine is the ANES question “one of the big problems in this country is that we don’t give everyone an equal chance,” captured on a 5-point Likert scale. For this analysis, we created five outcomes from questions gauging respondents’ political attitudes. First, we created an index of respondents’ support for government spending by creating an additive scale of responses to three questions: support for educational spending, unemployment spending, and spending on the poor (CA = 0.62).“CA” denotes Cronbach’s Alpha, a measure of scale consistency that ranges from 0 to 1.Second, we constructed an index of respondents’ support for government action on health care and jobs by creating an additive scale of two questions asking whether it is the government’s responsibility to provide health care and jobs (CA = 0.67). Third, we measured respondents’ support for the government reducing levels of economic inequality with a single item capturing whether the government should take actions to reduce differences in income levels. Fourth, we measured whether respondents support affirmative action in the workplace for Black Americans using a single item. Fifth, and finally, we measured 2012 vote choice by creating a dichotomous indicator of the two-party vote share (1 = Obama, 0 = Romney).Next, we utilize the “most people can make it if they’re willing to work hard” question from Pew Research Center,We use only this item from Pew for two reasons. First, the question on poor people and welfare benefits (“…have it easy”) invites judgments about government programs, making it inappropriate for these analyses. Second, the most-recent iteration of the question on hard work among the poor (“…lack of effort”) was administered to only a subset of respondents, resulting in an unusually small sample size (N = approximately 200) and few political attitude items.operationalized as a dichotomous indicator of whether respondents agreed (versus disagreed) with this statement. To measure policy attitudes, we included single questions about whether it is the government’s responsibility to help the needy (dichotomous), if government should increase spending on the needy (3-point), if government should increase taxes on the rich (5-point), if government should increase taxes on large corporations (5-point), if government should provide Americans with health coverage (4-point), as well as the political party for which they intended to vote in the 2020 presidential election (two-party vote share).The final causal attribution in our series of analyses is taken from the GSS. It is agreement with societal explanations for racial differences, in which we constructed a count variable ranging from 0 to 3, representing the number of instances in which a respondent agreed with external (discrimination; education) or opposed internal (lack of will) explanations for racial disparities in socioeconomic status (CA = 0.51). We created six outcome variables using questions from the 2018 GSS.GSS analyses use data from 2018 due to problems with 2020 data availability in the cumulative file.First, we included four single items on government’s responsibility to help the poor (5-point), help the sick (3-point), and reduce income differentials (4-point) and on whether government should spend more on education (5-point). Second, we included a single item on respondents’ support for affirmative action in the workplace for Black Americans (4-point). Finally, we measured 2016 vote choice by creating a dichotomous indicator of the two-party presidential vote share.Details regarding how each of these items were measured are included in the appendix. Note that all of the variables in our analyses were recoded to range from 0 to 1 to ease interpretation. In addition, we scored all attribution items such that the societal-blame answers had the highest values and all outcome variables such that the more Democratic, liberal, or progressive answers had the highest values. Thus, we expect positive relationships between the key attributional predictors and the various political attitude outcomes.We begin with Table 2, which displays results for the “equal chance” variable. Across the different models, the coefficient ranges from a low of 0.17 (Obama vote) to a high of 0.28 (government should reduce inequality). All are statistically significant (p < 0.01). In short, those who strongly agreed that not giving everyone an equal chance was a “big problem” were, on average, around 20% higher on the various outcome variables than those who strongly disagreed with this sentiment. Note that the coefficients on partisanship range from −0.19 to −0.34 for the policy items, meaning that the “equal chance” variable is associated with these outcomes nearly as strongly as partisanship. This said, and not surprisingly, the association between partisanship and presidential vote preference is much stronger.Table 2:Importance of equal chance and political attitudes.Outcome variableGov spendingGov jobs/healthGov reduce inequalityAffirmative actionDem vote(1)(2)(3)(4)(5)Equal chance0.218c0.191c0.278c0.187c0.167c(0.009)(0.011)(0.013)(0.013)(0.016)Party (Republican)−0.199c−0.338c−0.263c−0.193c−0.965c(0.008)(0.010)(0.012)(0.011)(0.014)Education−0.036c0.006−0.100c0.051c0.071c(0.010)(0.012)(0.015)(0.014)(0.018)Income−0.078c−0.111c−0.092c−0.069c−0.025(0.010)(0.012)(0.014)(0.014)(0.018)Age−0.029c−0.068c−0.192c0.007−0.077c(0.011)(0.014)(0.016)(0.016)(0.021)Male−0.018c0.007−0.0110.001−0.002(0.005)(0.006)(0.008)(0.007)(0.009)White−0.018c−0.007−0.031c−0.133c−0.082c(0.006)(0.008)(0.009)(0.009)(0.012)Constant0.637c0.600c0.581c0.395c0.969c(0.011)(0.014)(0.016)(0.015)(0.021)Observations51514618517751823982R20.3230.3630.2980.2280.650Adjusted R20.3220.3620.2970.2270.649Residual std. error0.185 (df = 5143)0.216 (df = 4610)0.274 (df = 5169)0.259 (df = 5174)0.290 (df = 3974)Note that the Ns in each column in Table 2 and those that follow vary based on how many respondents took a clear stance on the outcome variable. Respondents who said “don't know” or otherwise did not choose one of the provided options were excluded from analysis. In addition, in some instances, outcome questions were assigned to a random subset of the survey sample. ANES 2012. ap < 0.1; bp < 0.05; cp < 0.01.Moving to Table 3, we regress five specific policy items and Democratic presidential voting onto the Pew item asking people to choose between affirming that most people can make it if they work hard and hard work is no guarantee of success. The coefficients here range from 0.11 (healthcare; rich pay more taxes) to 0.36 (help the needy). Each is again statistically significant (p < 0.01). The coefficients in the policy regressions are about half the strength or more of the partisanship variable.Table 3:Belief hard work no guarantee and political attitudes.Outcome variableGov help if debtGov help needyRich pay moreCorps pay moreGov healthDem vote(1)(2)(3)(4)(5)(6)Hard work no guarantee0.197c0.361c0.111c0.132c0.110c0.149c(0.021)(0.016)(0.008)(0.008)(0.008)(0.007)Party (Republican)−0.408c−0.714c−0.160c−0.222c−0.269c−0.819c(0.026)(0.020)(0.011)(0.010)(0.010)(0.009)Education−0.0450.140c0.028b0.031b0.055c0.101c(0.036)(0.026)(0.014)(0.013)(0.013)(0.012)Income−0.078b−0.472c−0.076c0.001−0.055c0.003(0.039)(0.028)(0.015)(0.015)(0.014)(0.013)Age0.0360.039a0.049c−0.009−0.088c−0.065c(0.029)(0.022)(0.012)(0.012)(0.011)(0.010)Male−0.012−0.048c−0.009−0.0050.011−0.028c(0.020)(0.015)(0.008)(0.008)(0.008)(0.007)White−0.025−0.093c0.076c0.009−0.022b−0.104c(0.023)(0.017)(0.009)(0.009)(0.009)(0.008)Constant0.693c1.596c0.646c0.786c0.788c0.968c(0.032)(0.023)(0.012)(0.012)(0.012)(0.011)Observations209285054178417240247455R20.1900.2890.1310.2060.2830.663Adjusted R20.1870.2880.1300.2050.2820.663Residual std. error0.438 (df = 2084)0.656 (df = 8497)0.244 (df = 4170)0.233 (df = 4164)0.227 (df = 4016)0.270 (df = 7447)Pew American Trends 53 (2019). ap < 0.1; bp < 0.05; cp < 0.01.Finally, in Table 4, we regress similar items as in the ANES analyses on the GSS racial inequality attribution measurement scale. The coefficients vary more here, from a low of 0.04 in the education model to highs of 0.25 in the affirmative action model and 0.29 in the voting model. But, in the main, the effect sizes are similar to what we saw in the Pew analysis.Table 4:Blame society for racial inequality and political attitudes.Outcome variableGov help poorGov healthSpending on educationGov reduce inequalityAffirmative actionDem vote(1)(2)(3)(4)(5)(6)Racial differences0.179c0.188c0.0400.109b0.250c0.288c(0.034)(0.037)(0.037)(0.046)(0.031)(0.043)Party (Republican)−0.244c−0.266c−0.123c−0.260c−0.199c−0.820c(0.041)(0.045)(0.044)(0.054)(0.036)(0.049)Education−0.110c−0.035−0.007−0.167c−0.091b0.042(0.042)(0.046)(0.045)(0.053)(0.037)(0.049)Income−0.052−0.006−0.0230.050−0.174c−0.179b(0.057)(0.063)(0.063)(0.082)(0.054)(0.086)Age−0.053−0.066−0.149c−0.209c−0.032−0.126b(0.051)(0.055)(0.054)(0.067)(0.044)(0.063)Male−0.031−0.065b−0.028−0.0110.013−0.048a(0.024)(0.026)(0.025)(0.031)(0.021)(0.029)White−0.037−0.0310.019−0.059−0.089c−0.235c(0.028)(0.031)(0.030)(0.040)(0.025)(0.039)Constant0.654c0.734c0.947c0.602c0.518c1.087c(0.061)(0.067)(0.067)(0.092)(0.058)(0.098)Observations499499509246967548R20.1830.1750.0480.2520.1670.555Adjusted R20.1710.1640.0340.2300.1600.550Residual std. error0.259 (df = 491)0.284 (df = 491)0.279 (df = 501)0.233 (df = 238)0.317 (df = 959)0.325 (df = 540)GSS 2018. ap < 0.1; bp < 0.05; cp < 0.01.4Discussion and ConclusionWe have argued that beliefs about the American economy—in particular, whether it is meritocratic and, therefore, fairAmericans tend to equate meritocracy with fairness; however, it is important to acknowledge that the two are not interchangeable. For example, Sandel (2020) argues that meritocratic systems can include excessive pay differentials and also reward those who do not provide value to society or who were lucky enough to inherit their talents.or instead biased and unfair—represent an important rift between the two parties. In national platforms since 1980, Democratic and Republican party elites have consistently disagreed as to the reasons why some people succeed and others struggle, and this disagreement has grown over the past two decades or so. The Democratic Party is especially notable for its recent emphasis on discrimination as an important cause of economic inequality and its omission in 2020 of any role individuals, families, and communities might play in inequality. The Republican Party has shifted over time from blaming inequality on government dysfunction to, more recently, a mix of government and person-focused blame. Despite the sympathies expressed by Trump for working class Americans, the Republican platform during his two runs for the presidency included surprisingly little language about barriers to advancement lower-income people may face, such as lack of job and educational opportunities.Partisan members of the public mirror their party’s platforms. Democratic citizens are pessimistic about whether “the American Dream” is a reality for most people, and they tend to avoid blaming the poor for falling behind. Republicans are quite optimistic and tend to blame the poor. With respect to the economic difficulties that Black Americans disproportionately face, Democrats are especially concerned that these difficulties are undeserved, whereas Republicans are skeptical of explanations that deny Black Americans are responsible. These partisan differences have grown larger over the period we examine in every case, especially so for racial inequality. This said, the timing of increasing polarization differs depending on whether survey respondents are focused on explaining economic inequality in general or between Black and White Americans. In the former case, polarization grew during the George W. Bush era or earlier; in the latter case, polarization accelerated after Obama’s election. While much scholarship has focused on increases in racial prejudice among Republicans at this juncture, our data show that increasing partisan disagreement in recent years over the causes of racial inequality has been driven mainly by Democrats, especially accelerating since Trump’s entry into presidential politics.The partisan differences across all of the questions we examine are large. In 2020, on-average differences between Democrats and Republicans range from a low of approximately 30 points to a high of nearly 60 points. To appreciate the size of these differences, consider that partisan disagreements on climate change and abortion—two of the most contentious issues in American politics today—are in the same range, depending on the specific topic and question wording (Hartig 2022; Kennedy and Johnson 2020). Note also that partisan polarization over the causes of inequality is not driven by the different racial composition of the party coalitions; in fact, if we examine the over-time trends among Whites only, between-party differences are generally larger than those in our aggregate figures.Results not shown and available from authors upon request.Why do these large differences exist? It is unwise to draw inferences about causal mechanisms from simple correlations; however, the changes in polarization that we observe over time can give us some purchase on this question. Some of the changes are likely due to party sorting—liberals moving into the Democratic Party, and conservatives moving into the Republican Party. Sorting on Racial Resentment, a measure that overlaps substantively with the GSS racial inequality items, continued through the election of Donald Trump (Abramowitz and McCoy 2018; Enders and Scott 2018). Yet, some of the changes we observe—especially between 2010 and 2020 and on beliefs about racial inequality—are too large and swift to be explained by party switching. Furthermore, it is during this period that arguments about economic and racial injustice became more common among Democratic activists and politicians, as reflected in our platform analysis. This suggests that, to some degree, partisans in the public are persuaded by political elites espousing beliefs about the nature of opportunity in the U.S. that justify their conservative or liberal policy positions. We don’t mean to suggest that most Americans read party platforms; rather, the factual claims captured in the platforms are conveyed via a variety of channels, from candidate speeches to interest group publicity to traditional and social media.As other scholars have documented, these beliefs appear to be consequential politically. Depending on the specific way in which beliefs about economic inequality are measured, their association with various economically progressive policy measures ranges from half to equal the effect of partisanship; explanations for inequality are also associated with presidential preferences, although the associations in this instance are dwarfed by that of partisanship. The policy relevance of beliefs about economic inequality in general and between Black and White Americans is similar. One might ask what would happen if we included them both on the right-hand side of the regression equations. Would explanations for generic or racial economic inequality matter more to political views? Unfortunately, we do not have the equivalent question wording in any one survey that would allow us to test this. However, in a working manuscript, we analyze comparable questions on generic and racial economic inequality (Suhay n.d.). There, we demonstrate that beliefs about the causes of economic inequality in general are more predictive of attitudes toward social welfare—contra Gilens (1995) and Nelson (1999)—whereas beliefs about the causes of racial inequality are more predictive of presidential candidate preferences.These consequential differences in factual beliefs are an underappreciated aspect of party polarization. To our knowledge, we are the first authors to provide such a broad and systematic analysis of changing attributions for inequality over time among partisans. We believe the implications are profound. Our analyses suggest there exists a “redistributive versus anti-redistributive” ideology that lies at the center of political debates over economic policy. Most likely, it originates among activists, politicians, and other political elites and is passed to broad swaths of the public. The notion that the public is “innocent of ideology” (Converse 1964; Kinder and Kalmoe 2017) simply is not the case in this domain: partisanship, policy preferences, and supporting factual narratives are all closely associated.This said, ideological thinking in this instance is nothing to celebrate. Debates over meritocracy are not merely “differences of opinion”; they reflect wildly different perceptions of ordinary Americans’ day-to-day realities and how resources are gained and lost. While our society is complex, there are truths about the extent to which the U.S. lives up to its reputation as a place where people can achieve “the American Dream” through hard work and the conditions under which the Dream is more or less likely. The political parties and their allies, with their strong incentives to bend the truth to fit their policy agendas, are not the ones we should trust most to convey this reality. As with the similarly politicized “debate” over climate change, impartial expertsFor example, the United Nations’ Intergovernmental Panel on Climate Change (www.ipcc.ch) and the United States’ Global Change Research Program (www.globalchange.gov/nca4) publish well-publicized and influential empirical reports on climate change prepared by large teams of interdisciplinary experts.are more trustworthy voices on the empirical question of opportunity in the U.S. If citizens could develop a shared understanding of the ways in which the American economy fails to deliver meritocratic outcomes, perhaps they would demand the two parties find common ground on policies designed to combat inequality.

Journal

The Forum – de Gruyter

Published: Apr 1, 2022

Keywords: party polarization; economic inequality; racial inequality; causal attributions; public opinion; political ideology

There are no references for this article.