Introduction

Populism—a Manichean worldview that pits the supposed virtue and wisdom of “ordinary” people against an allegedly corrupt and inept “elite”—has exploded as a topic of social science inquiry over the past decade (e.g., Castanho Silva et al., 2020; Hawkins & Rovira Kaltwasser, 2017; Mudde and Rovira Kaltwasser, 2013; Oliver & Rahn, 2016; Schulz et al., 2018). It remains muddled, however, as an attitudinal construct. Researchers use the “populist” label to describe a diverse and contradictory array of aggrieved perspectives that span the ideological Left and Right, and they routinely conflate populism with various other “isms” such as nationalism, nativism, authoritarianism, egalitarianism, and anti-intellectualism. The nature and dynamics of American populism are especially muddied (but see Kamens, 2019; Mudde, 2018; Oliver & Rahn, 2016; Urbinati, 2019), as evidenced by the fact that Ronald Reagan, Bill Clinton, Elizabeth Warren, and Donald Trump have all shared the moniker in recent decades. Indeed, for many political observers and commentators, populism is as populism does; if “the people” like it, it must be “populist”—conceptual coherence be damned.

In this investigation, we aim to take some steps toward clarity. We explore the dimensions, distinctions, and correlates of populism in the American mass public. Drawing upon a unique, nationally representative survey, we collected in the fall of 2020 (n = 1000), we observe two unmistakable dimensions—economic and cultural—which are equally prevalent but do not typically overlap. Indeed, despite sharing remarkable levels of (1) support for popular sovereignty, (2) contemptuousness toward political opponents, and (3) anti-Semitism, economic populists and cultural populists tend to be different types of people.

Specifically, we find that economic populists in the U.S. tend to be disproportionately young, female, secular, cosmopolitan, and “open to experience.” They are also more likely to hold humanitarian and egalitarian values, to identify as “liberal” or “progressive,” and to claim Democratic partisanship. They exhibit less confidence and emotional stability, on average, than do their less economically resentful fellow citizens, and they are usually more willing compromise with political opponents (even while loathing them, in many cases).

Cultural populists, on the other hand, appear to be more working class (in terms of income and especially education) than do those who hold cultural elites in higher esteem, and to hold more anti-intellectual, nationalistic, Christian fundamentalistic, and disciplinarian worldviews. They are also more likely to prioritize “their own,” less likely to support compromise with their opponents, and much more likely to identify as “conservative” and Republican. However, contradicting conventional wisdom, cultural populists are no more likely than more “establishment-friendly” citizens to be white, male, or authoritarian in disposition, and they seem less likely than non-populists to hold socially dominant worldviews.

Despite these distinctions between economic and cultural populists overall, about 12% of our sample respondents reveal both kinds of populism. Relative to all other respondents (those who are only economic populists, those who are only cultural populists, or those who are neither), such “ideologically-constrained populists” tend to be younger, lower in socioeconomic status, and disproportionately female. Personality-wise, they are often more conscientious and more neurotic than other Americans, but no less agreeable or open to new experiences. As for their values and beliefs, they are simultaneously more compassionate and egalitarian, on the one hand, and more disciplinarian and loyal to “their own,” on the other. Though they tend to be much more nationalistic and anti-Semitic than other Americans, they seem no more nativistic or authoritarian. Somewhat surprisingly, they appear less socially dominant and less anti-intellectual than other Americans. When it comes to ideology and partisanship, they are less likely than economic populists to identify as “liberal” and “Democratic,” but more likely than cultural populists or non-populists to do so. Perhaps most tellingly, they are much more likely than those who are just economic populists, just cultural populists, or non-populists to express social contempt toward their political opponents.

These patterns help explain the rise in proletarian anger on both the Left and the Right in America, and its inevitable policy consequences—drastic deficit spending (and thus, inflation), climate inaction, isolationism, and polarized debates around immigration, criminal justice, and education. We hope these insights serve to launch additional lines of inquiry.

Populism: what we know

As a “thin centered ideology” (Mudde, 2004) with a “chameleonic character” (Taggart, 2000), populism has a long history of being redefined, disputed, and classified in distinct ways (for a good review, see Mudde, 2017). Scholars have analyzed it as a strategy (Weyland, 2001), a mobilization device (Jansen, 2011), a mass movement (Barr, 2009), a discursive method or frame (Aslandis, 2016; Dryzek & Berejikian, 1993; Vreese et al., 2018), a political style (Moffitt & Tormey, 2014; Van Hauwaert & Van Kessel, 2018), an economically liberal/culturally conservative combination of attitudes (e.g., Swedlow & Wyckoff, 2009), and an illiberal democratic variant (Caramani, 2017; Mohrenberg et al., 2021; Mounk, 2018; Pappas, 2016; Urbinati, 2019).

Consistent across these treatments is an understanding that populism pits the perspectives and interests of ordinary “people” against those of elites. Because of this, populism is often confused or lumped together with virtually any indignant orientation that a large segment of poor and/or uneducated masses happens to possess—especially, in recent years, those associated with the far Right. Such orientations include Ethnocentric Nativism (Bonikowski & DiMaggio, 2016; Mudde, 2007; Rooduijn, 2019), Nationalism (Bonikowski, 2016; Bonikowski & DiMaggio, 2016; De Cleen & Stavrakakis, 2017; Filsinger et al., 2021; Mader & Schoen, 2019; Mader et al., 2018; Mudde, 2007), Anti-Intellectualism (Barker et al., 2021; Castanho Silva et al., 2020; Oliver & Wood, 2018), Anti-Semitism (Cremoni, 1998, but see Pollack, 1962), Authoritarianism (e.g., Dunn, 2015; Pappas, 2019; but see Bakker et al., 2021), and illiberalism more generally (Applebaum, 2020; Galston, 2018; Mounk, 2018; Pappas, 2016; Riker, 1982).

Investigations tend to be region-specific, with scholars “talking past each other” who study different parts of the world (Barr, 2009, but see Weyland, 1999; Mudde & Kaltwasser, 2013; Castanho Silva et al., 2020). Latin America has seen at least three different strains of populism in the past century, all of which manifest economic grievance. The inclusive “classical populism” of the 1930s and 1940s, the economic strife and inflation triggered “neoliberal populism” of the 1990s, and the redistributive focused “radical populism” of more recent times (De la Torre, 2017). In general, historically speaking at least, Latin American populism can be considered “inclusionary” (Mudde & Rovira Kaltwasser, 2013).

The resentments of European populists, on the other hand, tend to be more cultural—railing against immigration, regionalism, corruption, and Euroskepticism (Taggart, 2017). European populism appears to include three primary dimensions: anti-elitism, support for popular sovereignty, and an understanding of the people as homogenous and virtuous (Schulz et al., 2018; Roccatto et al., 2019). It accompanies the absence of (a) external political efficacy, (b) pluralist appreciation, and (c) political trust, but it is more than the sum of those parts (Akkerman et al., 2014; Geurkink et al., 2020; Spruyt et al., 2016); it manifests most commonly among stigmatized groups who score low on “agreeableness,” personality-wise (Bakker et al., 2021), and who struggle to find a positive social identity (Spruyt et al., 2016). Accordingly, unlike most Latin American expressions of populism, European populism tends to be “exclusionary” (Mudde and Rovira Kaltwasser, 2013).

Studies of American populism have been far less abundant, and mostly historical in nature—focusing on Jeffersonian agrarianism (e.g., Bailey, 2007), Jacksonian democracy (e.g., Goebel, 1997), and especially the late nineteenth century movement led by William Jennings Bryan (e.g., Cherny, 1994; Kazin, 2007; Rooduijn, 2014). Scholarly characterizations of the latter vary wildly—depicting it, by-turns, as xenophobic (e.g., Hofstadter, 1955), radically participatory (Goodwyn, 1976), essentially conservative (Arceneaux & Nicholson, 2012), originally progressive (Stavrakakis, 2018), protectionist (Skonieczny, 2019), and even free-market oriented (Postel, 2007).

Historians have tagged a wide range of American political figures from the twentieth to twenty-first century as “populist” as well—including Franklin D. Roosevelt, Father Coughlin, Huey Long, George Wallace, Richard Nixon, Ronald Reagan, Pat Buchanan, Ross Perot, Bill Clinton, Sarah Palin, Bernie Sanders, Elizabeth Warren and Donald Trump. About the only things these politicians all had in common is their supposed common cause with “ordinary people” and their relative (if sometimes short-lived) popularity. Indeed, populism in the United States “can be roughly divided between left-wing and right-wing variants, according to how each defines the principal foe of the people” (Lowndes, 2017)—with left-wing populists targeting economic elites and right-wing populists targeting “cultural elites” (who tend to be defined as highly educated cosmopolitans and their supposed “big government” allies). Ideologically constrained populism is sometimes defined as the marriage of such economic leftism and cultural conservatism (e.g., Swedlow, 2008; Swedlow & Wyckoff, 2009).

In recent years, though, scholarship and popular commentary on American populism have focused on the cultural side of the equation (Applebaum, 2020; Cramer Walsh, 2012; Galston, 2018; Kamens, 2019; Mounk, 2018; Mudde, 2018; Oliver & Wood, 2018)—the people who Oliver and Rahn (2016) refer to as the “Trumpenvolk.” This literature, even more so than studies of European or Latin American populism, relies primarily on impressionistic accounts, and it tends to blur the lines between populism, nativism, nationalism, anti-intellectualism, and generally illiberal attitudes.

To sum things up, the edifice of knowledge on American populism has substantial cracks in its foundation. No previous studies have sought to paint its full panoramic portrait, almost none of the ones offering partial views bring “large N” scientific methods to the task, and the exceptions in the latter case suffer from remarkable conceptual slippage. In the analyses that follow, we try to start sealing the cracks. Specifically, we examine (a) the dimensions of American populism and their degree of overlap, (b) their relation to other psychological constructs with which they are often confused, and (c) their personal, ideational, and political correlates.

Data and analysis

We collected nationally representative survey data during the 2020 presidential election campaign period (September 20th–December 2nd; n = 1000), as part of the Cooperative Election Study (formerly the Cooperative Congressional Election Study), that is administered annually by Yougov (detailed information regarding the sampling procedure and so on is viewable here).

To measure populism, the survey asked respondents to indicate the extent to which they “agree or disagree” with a series of statements (5-point; Likert, 1932), all of which we rescaled to 0–1 for analysis. The first two statements aimed to measure economic populism:

  • (EP1) It is time for the rich to start paying their fair share again in America (mean = 0.73; SD = 0.30)

  • (EP2) It is hard to become wealthy without being a cheat or a suck up (mean = 0.52; SD = 0.31)

The second two statements aimed to measure cultural populism:

  • (CP1) I am sick of cultural elites looking down on us (mean = 0.66; SD = 0.28)

  • (CP2) The cultural elites are trying to do away with our American way of life (mean = 0.61; SD = 0.31)

A fifth statement aimed to measure democratic populism:

  • (DP) It would be better if regular people, not political elites, made decisions for the country (mean = 0.67; SD = 0.24)

These statements deliberately focus on what we consider the essence of populism, which is a dualistic understanding of “ordinary” people as good and elites (economic, cultural, and/or political) as bad.Footnote 1

How many dimensions and how prevalent?

Our first task is to assess the degree to which such American antagonism toward elites is just one thing or multiple things. How well do economic populism, cultural populism, and democratic populism hang together? We anticipate that economic populism and cultural populism are fully distinct, with little overlap between them.

To gain some purchase over this question, we conducted a factor analysis of the five survey items listed above (using maximum likelihood extraction).

As Fig. 1 displays, the factor analysis extracted exactly two factors, of similar magnitudes (eigenvalues of 1.47 and 1.12, respectively), with the cultural populism items loading strongly on the first (factor loadings = 0.77 and 0.87) and the cultural populism items loading clearly on second (factor loadings = 0.76 and 0.69). Moreover, one of the two economic populism items loads negatively on the first factor, with the same being true of one of the cultural populism items on the second factor. The democratic populism item contributes comparably to both factors, but it is much less important to either (loadings = 0.28 [first factor] and 0.22 [second factor]).Footnote 2

Fig. 1
figure 1

Factor analysis of populist dimensions. Factor loadings from maximum likelihood extraction of the two economic populism survey items (EP1 and EP2), the two cultural populism survey items (CP1 and CP2), and the democratic populism survey item (DP)

Full size image

Based on these results, we consider it safe to say that in the contemporary United States, ideological populism spanning economic and cultural dimensions is rare; it is certainly no threat to replace either liberalism or conservatism as a predominant framework for organizing political attitudes.

As such, we summed the economic resentment items into an index of Economic Populism (rescaling it to 0–1 for analysis; mean = 0.63; SD = 0.27), and we did the same with the two cultural grievance items to create Cultural Populism (mean = 0.64; SD = 0.27). To gain understanding of the relatively small but not inconsequential number of ideologically constrained populists in the sample (and the population), we created a third variable that multiplies Economic Populism by Cultural Populism (mean = 0.4; SD = 0.27). In the remainder of this investigation, we examine the correlates of these three scales. We start by considering the demographic and personality profiles of each.

Who are the populists?

To paint demographic portraits of economic and cultural populists, we estimated seemingly unrelated regression models (Zellner, 1962) using the ordinary least squares estimator, predicting each type of populism with variables for race (non-white = 28%), gender (Female = 56%), age (18–91; mean = 49 years old; SD = 18 years). Annual Gross Household Income (1–16; 1 =  < $10,000; 16 =  >  = $500,000; mean = 6.34 [~ $63,400]; SD =  ~ $34,000), and Education (1–6; 1 =  < high school graduate; 6 = post-graduate degree; mean = 3.64 [some college but no 2-year degree]; SD = 1.50).Footnote 3

As Table 1 displays, relative youth and female identity are predictive of economic populism, but race, education and household income are not (suggesting that economic populism is not class-based). For its part, cultural populism is associated with being a little older, a little less well-off financially, and quite a bit lower in educational attainment, but it is not related to gender or race (contradicting the image, somewhat, of the “angry white male”). As for the ideologically constrained populists, they tend to look a lot like the economic populists when it comes to gender and age, but the cultural populists when it comes to socioeconomic status and educational attainment.

Table 1 Demographic Covariates of Economic Populism, Cultural Populism, and Ideologically Constrained Populism
Full size table

Next, we consider the extent to which economic and cultural populists have distinct personalities from each other and from those who are not populists. Previous studies, conducted mostly in Europe, suggest that populists tend to score low on the Big-5 (OCEAN) personality trait of agreeableness (Bakker et al., 2016; Fatke, 2019; Müller, 2017), but researchers have yet to undertake comprehensive look at both economic populists and cultural populists in the United States.

We rely on the Big 5—a.k.a. OCEAN (openness, conscientiousness, extraversion, agreeableness, and neuroticism)—personality traits (see Digman, 1990 for a general review of this literature, and Gerber et al., 2011 for a review that is focused on personality as it pertains to political attitudes). We measure each with 5-point Likert-style responses (strongly disagree to strongly agree) to two statements, all of which begin with “I am someone who...” For each trait, we inverted the first statement and rescaled the response scales to statements to 0–1. We then summed responses to the two statements, and also rescaled the resulting index to 0–1. The specific statements are as follows, along with all the descriptive statistics:

Openness (mean = 0.61; SD = 0.20):

  • … has few artistic interests (inverted; mean = 0.70; SD = 0.24)

  • … has an active imagination (mean = 0.52; SD = 0.30)

Polychoric \(\rho\)= 0.13 (SE = 0.04)

Conscientiousness (mean = 0.65; SD = 0.19):

  • … can be somewhat sloppy sometimes (inverted; mean = 0.50; SD = 0.38)

  • … does a thorough job (mean = 0.19; SD = 0.19)

Polychoric \(\rho\) = 0.29 (SE = 0.04)

Extraversion (mean = 0.48; SD = 0.22):

  • … is reserved (inverted; mean = 0.62; SD = 0.26)

  • … is outgoing, sociable (mean = 0.41; SD = 0.28)

Polychoric \(\rho\) = 0.33 (SE = 0.04)

Agreeableness (mean = 0.63; SD = 0.19):

  • … tends to find fault with others (inverted; mean = 0.45; SD  = 0.25)

  • … is generally trusting (mean = 0.29; SD = 0.26)

Polychoric \(\rho\) = 0.16 (SE = 0.04)

Neuroticism (mean = 0.46; SD = 0.24):

  • … is relaxed, handles stress well (inverted; mean = 0.59; SD = 0.27)

  • … gets nervous easily (mean = 0.50; SD = 0.31)

Polychoric \(\rho\) = 0.53 (SE = 0.03)

We use these variables as the explanatory variables in another round of seemingly unrelated regression models predicting Economic Populism, Cultural Populism, and Constrained Populism. Based on demographics, we anticipate that economic populists will reveal more openness and neuroticism than non-populists or cultural populists, but we are otherwise agnostic.

As Table 2 shows, Openness is indeed negatively associated with Cultural Populism, and positively associated with Economic Populism. As we also expected, Neuroticism is positively associated with Economic Populism (as is ideological liberalism; see Mondak, 2010; Gerber et al., 2011). It is not, however, associated with Cultural Populism. In fact, the only other personality trait that predicts Cultural Populism is Conscientiousness, which is in keeping with its known relationship to conservatism (see Gerber et al., 2011). Finally, contrasting some other studies (e.g., Bakker et al., 2016, 2021), we do not find Agreeableness to be either positively nor negatively related to either form of populism. As for the ideologically constrained populists, they tend to share the neuroticism of economic populists and the conscientiousness of cultural populists.

Table 2 Personality Covariates of Economic Populism, Cultural Populism, and Ideologically Constrained Populism
Full size table

Next, we explore the worldviews of economic and cultural populists.

What do populists believe?

We begin this section by examining the degree to which economic and cultural populism are associated with a series of core values or “moral foundations” that scholars have shown to be strongly predictive of public policy preferences and political ideology in the US (e.g., Barker & Tinnick, 2006; Feldman, 1988; Feldman & Steenbergen, 2001; Haidt, 2012; Jacoby, 2006; Marietta & Barker, 2019; McCloskey & Zaller, 1984; Schwartz, 1992). Specifically, we model the predictive capacity of (1) Compassion (a.k.a. “care” [Haidt, 2012], “nurturance” [Lakoff, 1996] or “humanitarianism” (Feldman & Steenbergen, 2001), (2) Egalitarianism (Feldman, 1988), (3) Disciplinarianism (Lakoff, 1996, a.k.a. [punitive] “fairness” [Haidt, 2012], (4) Libertarianism (Iyer et al., 2012), (5) Sanctity (Haidt, 2012), and (6) In-Group Loyalty (Haidt, 2012). We measure each with a single Likert-style item (rescaled to 0–1), as follows:

Compassion “A decent and just community helps people who are suffering, no matter what the reason” (mean = 0.70; SD = 0.23)

Egalitarianism “A decent and just community provides equal opportunity by guaranteeing access to high quality health care, education and housing” (mean = 0.74; SD = 0.28)

DisciplinarianismFootnote 4 “A decent and just community punishes cheaters, slackers, and others who try to take advantage of the system (mean = 0.57; SD = 0.27)

Libertarianism “A decent and just community does not place any restrictions on how people speak or behave, as long as no one is getting hurt” (mean = 0.60; SD = 0.298)

Sanctity “A decent and just community shames certain types of sick or unnatural behaviors, even if no one is getting hurt (mean = 0.42; SD = 0.30)

In-Group Loyalty “A decent and just community looks out for its own people, first” (mean = 0.67; SD = 0.27).

Given what we now know from the previous models—that economic populists are more likely to be young, female, open-minded, and neurotic—we anticipate that they also tend to score higher on Compassion, Egalitarianism and perhaps Libertarianism than do those with more favorable views of the wealthy. Likewise, given our understanding that cultural populists tend to be older, of lower socioeconomic status, more closed-minded but also more conscientious than do their less culturally aggrieved fellow citizens, we anticipate that they also value compassion and equality less but discipline, sanctity, and in-group loyalty more than non-populists do.

Table 3 displays the now familiar seemingly unrelated regression equations predicting Economic Populism and Cultural Populism with these items. We see that most of our expectations were borne out by the data. Economic populists do indeed tend to exhibit higher levels of Compassion and especially Egalitarianism than do non-economic populists. They also tend to express lower levels of commitment to Discipline, Sanctity and In-Group Loyalty than do non-economic populists. They do not differ from economic non-populists with respect to Libertarianism. Cultural populists, on the other hand, are almost the mirror opposite; they tend to be less egalitarian, more punitive, more devoted to things being “pure,” and much more devoted to their own than non-cultural populists. Finally, ideologically constrained populists tend to blend the compassion and egalitarianism of economic populists with the punitiveness and “groupiness” of cultural populists. They are no more or less committed to Sanctity than non-populists (or non-constrained populists), but they do appear less committed to freedom.

Table 3 Value Covariates of Economic Populism, Cultural Populism, and Ideologically Constrained Populism
Full size table

Next, we take a look at the degree to which each type of populism overlaps independently with a set of right-wing orientations that are commonly equated with populism: Nationalism, Nativism/Ethnocentrism, Anti-Intellectualism, Authoritarianism, Christian Fundamentalism, Social Dominance Orientation, and Anti-Semitism. Based on previous research, our own intuitions, and the results we have reported so far, we anticipate that each of these is associated with cultural populism, but we are agnostic about (a) their relative degree of association to cultural populism, and (b) their relationships with economic populism or ideologically constrained populism.

We measure Nationalism—the belief that one’s country is inherently superior to others and the inclination to view international relations through a competitive rather than a cooperative lens—with a summed index of the following two Likert-style items (5-point, strongly agree to strongly disagree; rescaled to 0–1):

  • Any decent politician puts our American interests FIRST, regardless of what they think in other countries (mean = 0.68; SE = 0.28)

  • America has always been special—different and better than the rest of the world (mean = 0.56; SE = 0.32)

The two items are highly correlated (polychoric \(\rho\)  = 0.60 [SE = 0.03]; p < 0.001), producing a measure that reveals somewhat disproportionately nationalistic attitudes within the US mass public (as expected) but also reveals substantial variance (mean = 0.62; SE = 0.26).

Ethnocentric Nativism is antipathy toward non-white/non-Christian immigrants (Bennett, 1988; Fry, 2006; Knoll, 2013; Knoll & Shewmaker, 2015; Perea, 1997; Schräg, 2011; Tatalovich, 1995). We also measure it with another summed index of two Likert-style items, which we again rescaled to 0–1 for analysis:

  • If real Americans have a say, this country will never look like Mexico or Iran (mean = 0.62; SE = 0.31)

  • Immigrants should learn how to behave like real Americans (mean = 0.50; SE = 0.32)

These two items are even more strongly correlated than the Nationalism items are (polychoric \(\rho\) = 0.71 [SE = 0.03]; p < 0.001), producing a measure that reveals a highly divided but slightly nativistic American public (mean = 0.56; SE = 0.29).

We likewise measure Anti-Intellectualism, which can be defined as suspicion of intellect and especially the intellectual establishment (e.g., Hofstadter, 1963), with two Likert-style items that we summed and rescaled to 0–1:

  • Public schools and universities fill young people’s heads with all kinds of nonsense (mean = 0.47; SD = 0.36)

  • Too much education can blind you to the real truth (mean = 0.36; SD = 0.32)

These items are overwhelmingly correlated (polychoric \(\rho\) = 0.74 [SE = 0.01]; p < 0.001), producing a measure that again reveals substantial variance within the mass public but tends to be a little more pro-intellectual than anti-intellectual (mean = 0.42; SD = 0.31).

We measure Christian Fundamentalism, or the dogmatic belief in biblical inerrancy—typically but not always associated with evangelical Christian identity, with a summed (and rescaled to 0–1) index of two dichotomized items:

  • Do you consider yourself a Born Again Christian? (Yes = 27%)

  • Which of the following statements comes closest to your view of the Bible? Choice of “The Bible is the inerrant and authoritative Word of God, even in matters of history and science” (26%)

Again, these two items are again very highly correlated (polychoric \(\rho\) = 0.77 [SE = 0.0]; p < 0.001). About a quarter of the sample reveals such fundamentalist qualities, with a lot of variance (mean = 0.27; SD = 0.39).

Authoritarianism, which can be summarized as the instinct to kiss-up/kick-down and an attraction to order and rule-following, with responses to items about desirable qualities in children, which have become standard in the American politics literature (e.g., Feldman & Stenner, 1997; Hetherington & Weiler, 2009; Stenner, 2005):

“There are a number of qualities that are important for children to have, but people disagree about which ones are most important. If you had to choose, would you say it is more important for children to learn…”

  • Independence or respect for elders? (Respect = 53%)

  • Self-reliance or obedience? (Obedience = 29%)

  • Curiosity or good manners? (Good manners = 61%)

These items are reasonably intercorrelated (Cronbach’s \(\alpha\) = 0.65). Maintaining consistency with our other measures, we again summed and rescaled them, revealing an average American who is almost right in the center of the scale, but with a lot of variance around that average (mean = 0.48; SD = 0.37).

We measure Social Dominance Orientation, or support for social hierarchies (e.g., Pratto et al., 1994; Sidanius & Pratto, 1999) with a summed and rescaled index of the following Likert items:

  • It is both unnatural and unfair to try to make groups equal to each other (mean = 0.44; SD = 0.33)

  • Some groups of people are just inferior to other groups, and that’s OK (mean = 0.28; SD = 0.30)

These items are again strongly correlated, though not so much as the items in some of our other indexes (polychoric \(\rho\) = 0.63 [SE = 0.03]; p < 0.001). Our data suggest that the average American is not particularly dominant in her thinking—and much less so than s/he is authoritarian, anti-intellectual, nativistic, or nationalistic—but there is nevertheless a substantial amount of social dominance orientation within the mass public (mean = 0.36; SD = 0.27).

Finally, we measure Anti-Semitism, or the negative stereotyping of Jewish people, with a summed index of two items:

  • Jewish people are usually smart, but soft (mean = 0.36; SD = 0.24)

  • Jewish people often succeed by being shady (mean = 0.25; SD = 0.25)

The two items are again very correlated (polychoric \(\rho\) = 0.62 [SE = 0.03]; p < 0.001). The rescaled index reveals substantial overt anti-Semitism within the US public (mean = 0.31; SD = 0.22).Footnote 5

To what extent are each of these orientations independently associated with Economic Populism and Cultural Populism? Table 4 displays the results of two more seemingly unrelated regression equations, which reveal those relationships while simultaneously accounting for all of the others as well. The coefficients are decimal point increases/decreases in economic and cultural populism (on their 0–1 scales) that are associated with minimum-to-maximum differences in the orientations (with those relationships’ standard errors in parentheses).

Table 4 Orientational Covariates of Economic Populism, Cultural Populism, and Ideologically Constrained Populism
Full size table

Looking at the first results column first, we see that Economic Populism is negatively associated with Ethnocentric Nativism, Social Dominance Orientation, and especially Anti-Intellectualism. It might be negatively associated with Nationalism as well (though to a lesser and less reliable extent). It is, however, unrelated to Christian Fundamentalism or Authoritarianism (by these measures), ceteris paribus. Strikingly, and unexpectedly to us, it is strongly and positively associated with Anti-Semitism.

The second results column reveals some opposite patterns and some similar ones. As anticipated based on conventional wisdom, Cultural Populism is associated with Anti-Intellectualism, Anti-Semitism, and especially Nationalism. It might also be associated with Ethnocentric Nativism (but if so, to only a fraction of the extent to which it is associated with Nationalism), but it is not associated with Christian Fundamentalism or Authoritarianism (while holding the other orientations constant). In a surprise, it is negatively associated with Social Dominance Orientation (again, holding everything else constant).

Finally, constrained populists seem to share the cultural populists’ nationalism (though to a lesser extent), but not their anti-intellectualism (though they are not exactly pro-intellectual, either, as the economic populists tend to be). Not surprisingly, they share both the economic and cultural populists’ relative aversion to socially dominant thinking, but they appear to be the most anti-Semitic respondents in the entire sample.

To summarize this section, economic populists and cultural populists do not share many values or worldviews in common; cultural populists fit the stereotype (as portrayed in popular media and extant scholarship) of the disciplinarian, anti-intellectual nationalist, whereas economic populists appear pretty much the opposite—compassionate, egalitarian, pro-intellectual, cosmopolitan, and multiculturalist. However, cultural populists do not appear as authoritarian or as Christian fundamentalist as is often presumed, and neither group tends to exhibit socially dominant orientations. Both brands of populism, however—in isolation and especially in concert—seem strongly related to overt anti-Semitism, which in general appears more prevalent within the US public than is perhaps often assumed in the twenty-first century.

In all our remaining analyses, we use Economic Populism, Cultural Populism, and their interaction as explanatory variables rather than outcome variables, to gauge their capacity to predict Ideological Identification, Party Identification, Resistance Political Compromise, and Social Contempt.

Ideological and partisan identities

How do populists identify politically, with respect to ideology and partisanship? We measure Ideological Identification by asking, “In general, how would you describe your own political viewpoint?” (5-point response scale: very liberal|liberal|moderate|conservative|very conservative; rescaled to 0–1; mean = 0.50; SD = 0.29).

We measure Party Identification with the standard two-part question: “Generally speaking, do you think of yourself as a Democrat, Independent, Republican, or other?” Those who answer “Democrat” or “Republican” receive the follow up question, “Do you consider yourself a strong Democrat (Republican) or a not very strong Democrat (Republican?” Those who answer “Independent” receive the follow up question “Do you lean toward either the Democratic Party or the Republican Party?” We then combine not very strong partisans and Independent “leaners” into the same categories, producing a five-point measure that we rescale to 0–1 (mean = 0.46; SD = 0.46).

Figures 2 and 3 graph the results, showing that those whose populism is exclusively economic are almost all “very liberal” and Democratic, whereas those whose populism is exclusively cultural are almost all “very conservative” and Republican. Those who are neither economically nor culturally populist also tend to be quite conservative and Republican, while those who are both economically and culturally populist tend to identify as moderately liberal and Democratic. Full tabular results are observable in the Appendix.

Fig. 2
figure 2

Populism and ideological identification

Full size image
Fig. 3
figure 3

Populism and party identification

Full size image

Intransigence and contempt

Our last set of empirical models assess the explanatory power of the populism measures as they pertain to political intransigence and contempt. We measure the former by gauging respondents’ relative Support for Political Compromise:

“Some people say that political compromise is necessary so the country can make progress on big problems. Others say that compromise is just selling out and giving in to one's enemies. What do you think? Should the politicians in Washington who share your beliefs make compromises with the other side? Or should they stand firm no matter what?” The response scale is five-points, ranging from “strongly support standing firm” to “strongly support compromise” (rescaled to 0-1; mean=.52; SD=.30).

Finally, we measure Socio-Political Contempt with a summed index of two items, with 5-point response scales (1 = “definitely willing”; 5 = “definitely not willing”:

  • How willing would you be to work on a project in your local community with someone, one-on-one, who disagrees with you politically? (mean = 2.97; SD = 1.27)

  • How willing would you be to eat dinner with someone, one-on-one, who disagrees with you politically? (mean = 2.91; SD = 1.35)

The two items are nearly perfectly correlated (polychoric \(\rho\) = 0.81; SE = 0.02). The rescaled 0–1 index reveals substantial variance (mean = 0.48; SD = 0.31).

As Fig. 4 highlights (full tabular results appear in the appendix), while Economic Populism seems to be positively related to Support for Political Compromise, Cultural Populism corresponds to markedly lower levels of Support for Compromise.

Fig. 4
figure 4

Populism and support for political compromise

Full size image

As for their correspondence to Socio-Political Contempt, Fig. 5 shows that though neither is related to it on their own, they predict dramatically higher levels of it when operating in concert.

Fig. 5
figure 5

Populism and socio-political contempt

Full size image

In general, when it comes to orientations that reflect affective political polarization, such as intransigence (as measured by antipathy toward political compromise) and contempt (as measured by the willingness to socialize or even work together with those who disagree politically), we have observed that economic populism sometimes contributes, but only when combined with cultural populism; overall, the latter is a more consistent contributor.

Discussion and conclusion

This analysis contributes to the rather scant body of empirical scholarship focusing on attitudinal populism in the United States, by (1) identifying distinct strains of American anti-elite sentiment, (2) analyzing their demographic, personality, and value correlates; (3) distinguishing these populist typologies from other oft-misconstrued orientations (and also observing their connections), and (4) investigating the extent to which these populist variants are predictive of party identification, ideology, support for compromise, and socio-political contempt. We find that economic populism, cultural populism, and ideologically constrained populism (both economically populist and culturally populist attitudes working in concert) are all operating within the American public, though the latter is uncommon. Economic populists are disproportionately younger and female, while cultural populists are inordinately older, less educated, and less well-off financially. The ideologically constrained populists look like economic populists when it comes to age and gender, but more like the cultural populists when it comes to educational attainment and income.

These populist strains are also connected with psychological personality traits. Economic populism is associated with higher levels of openness and neuroticism, while cultural populism is connected to lower levels of openness and higher levels of conscientiousness. The ideologically constrained populists display the neurotic traits of economic populists and the conscientiousness of cultural populists, but do not have a connection with openness like either of the others. We did not detect any support for populism’s connection with the psychological personality trait of agreeableness—a prior finding in the literature.

Our variants of populism are also related to a subset of core values or “moral foundations.” Economic populists tend to score higher on the values of compassion and egalitarianism while displaying less disciplinarianism and in-group loyalty. We do not find evidence that cultural populism is related to compassion, but we do observe that it opposes economic populism on the other values, as cultural populists are more likely to be disciplinarians, in-group loyalists, and non-egalitarians.

We also note that our variants of populism are conceptually distinct from, but also correlated with certain right-wing orientations that are often confused with populism. Economic populists are less likely to display nationalism, nativism, anti-intellectualism, and social dominance orientation, but are more likely to hold anti-Semitic attitudes. Cultural populism is strongly associated with nationalism and anti-intellectualism, but like economic populism, also tends to be espoused by those who score lower in social dominance orientation and higher in anti-Semitism. Ideologically constrained populists show the same indirect relationship with social dominance orientation and direct relationship with anti-Semitism as the others do, but they tend to mimic the nationalist tendencies of cultural populists.

In our sample, economic populism is strongly predictive of both liberal ideology and Democratic party identification, while cultural populism is strongly predictive of conservatism and Republican party identification. The remaining non-populists respondents trend to the Right politically on both measures while the ideologically constrained populists trend slightly to the Left, but neither to the same extent as our economic or cultural populists. We also find that cultural populists are less likely to support political compromise and that ideologically constrained populists exhibit more socio-political contempt on average.

Our exploratory analysis of these different veins of populism operating in the United States enhances our picture of who the populists are, how they are wired, and how they behave. This study advances the investigation of populism as an attitudinal construct, and identifies a number of covariates that are related to distinct populist strains currently present in the United States. The identification of many of these covariates may be useful for future empirical researchers who want to understand public policymaking in an age where politicians on both sides of the aisle must answer to the “peasants with pitchforks.” Future studies could also delve into more specificity regarding the potential origins and causal pathways of populism and these covariates.

Another important takeaway for researchers of populism (both in the American and Comparative communities) is that separate types of populism can come from different places—as illustrated by our observation of economic populism coming from a younger and more female segment of the American public, while cultural populists tended to be of older, and less well educated origins.

Even outside of the U.S. setting, scholars of populism can incorporate these valuable findings into their analyses. The importance of taking conceptually distinct but clearly related orientations into account when studying populism is paramount. In addition, as there has been some evidence that moral foundations are stable across different cultural contexts (Doğruyola et al., 2019), researchers of populism around the world should be sure to include and examine these value correlates. Comparative comparisons to this study using populism abroad to predict left and right leaning ideology and party membership could also identify the extent to which American populism is integrated in the socially sorted American parties, relative to populism elsewhere in the world. Our guess is that these strains of populism in the United States are perhaps more predictive of ideology than elsewhere in the world, but not necessarily partisanship- as many of the multi-party systems in Europe have parties that are explicitly populist in nature.

This analysis may be of specific interest to scholars of populism in Latin America and in Europe. The economic populists in the U.S. seem to share the economic grievance focused attitudes and inclusionary style of Latin American populism, while the cultural populists in the U.S. seem to share the nationalist and anti-intellectual elements of European populism. Perhaps, the U.S. is an ideal setting for future research from scholars of either of these camps, especially to better understand understudied questions such as: what happens when these types of populism are both separately present in a national population and what are the causes and effects of ideologically constrained populism (when one holds both economically populist and culturally populist anti-elite attitudes simultaneously)?

Indeed, studies investigating populism’s predictiveness of willingness to compromise should be undertaken in Europe and Latin America. Our projection is that, like in the United States, the economic populists of Latin America may display more of a willingness to compromise politically, while the European populists may be more likely to prefer standing firm. It is our hope that this analysis both encourages and assists future studies of populism around the world, adding important nuance and context to what has historically been a rather nebulous concept.