This is an open access article distributed under the terms of the Creative Commons Attribution License (Political ideology in the United States is typically conceptualized and measured as a spectrum, with liberalism on the left and conservatism on the right. This metaphor implies that the people who place themselves along the spectrum differ in degree from each other; an individual who places themselves all the way on the left of the spectrum, is an extreme version of another individual closer to the middle. While useful, this metaphor is also limited and does not capture the full complexity of political preferences and identities (see e.g., Costello et al., 2022). The relevance of the spectrum metaphor has only increased with the rise of affective polarization (e.g., Iyengar & Krupenkin, 2018) and political sectarianism (see e.g., Finkel et al., 2020), suggesting that people who are differing in degree along the spectrum are getting further apart. And indeed, there is an increasing number of people who identify at the extreme ends of this spectrum (Twenge et al., 2016). In this paper, we sought to characterize the political identities and policy preferences associated with the left “end” of the spectrum. We find that those who, from the outside, might appear as extremely liberal, identify not as extreme Liberals, but as Leftists, who understand themselves to belong outside of the political spectrum, rather than on the end of it. This identification reflects a difference “in kind" from the broader left-wing political group.
Here, we focus on understanding self-identified Leftists in the United States, that is, people that we might broadly describe as adhering to an ideology of anti-capitalism and eliminating identity-based oppression (Bobbio, 1996). We focus on Leftists (sometimes called the “Progressive Left”; see Pew Research Center, 2021a) because they are a rapidly growing political minority group in the United States, as evidenced by increased support for progressive reform (Pew Research Center, 2021a, 2021b), the emergence of popular left-wing candidates for local and national office, and the growth of the Democratic Socialists of America as a membership group (Birch, 2020). Leftists are also underrepresented in psychological research – most psychology professors identify as liberals (e.g., Inbar & Lammers, 2016; Reinero et al., 2020; Von Hippel & Buss, 2017), and while this is usually brought up in reference to an imbalance with conservatives, it implies a lack of representation of other political identity groups among researchers as well. Leftists are also underrepresented on major research platforms like MTurk and Prolific (see Stagnaro et al., 2024 for discussion of political representativeness on these platforms), which researchers are increasingly turning to for data collection (Buhrmester et al., 2018). This broad underrepresentation highlights the importance of understanding Leftists’ political values in comparison to a typical sample in psychological research.
Finally, we also focus on understanding Leftists because there is good reason to expect that they will differ from people with other, or even similar, political identities in the United States. As we briefly review below, Leftists have a unique history in the United States that differentiates them from Liberals. This history may have led to differences in the issues each group prioritizes and moralizes (or sees as particularly relevant to their group’s values). Previous research on moralization suggests that when beliefs and attitudes become moralized (including within the realm of social issues), how people reason about them changes (e.g., Rozin, 1999; Rhee et al., 2019; Wylie et al., 2022). Thus, we predicted that, despite sharing many core priorities and values, the content and extremity of moral judgments of Leftists (and in particular, their moral opposition to the political status quo) would distinguish them from Liberals leading to meaningful differences that imply differences in kind rather than degree between them, and with consequences for support for mainstream Democratic political candidates.
Altogether, this research aims to both capture U.S. Leftists’ understanding of their own identity, and test whether Leftists’ moral judgments about the status quo distinguish them from their Liberal counterparts. In light of these goals, we begin by briefly summarizing the current two-party system in the United States, as well as providing a brief historical context for understanding the left-end of the political spectrum in the United States. Then, we present quantitative and qualitative data on the moral judgments of Leftists and Liberals, describe the consequences of moral judgments for mainstream Democratic candidate support, and document descriptive similarities and differences between the two groups.
A Brief History of Leftist Ideology in the United States
Early legislation like the New Deal (1933-1936) – which aimed to provide immediate economic recovery following the Great Depression by stabilizing banks, creating jobs, raising wages, and investing in public works – encouraged Leftists and Liberals to work together to bring about economic security (Rauchway, 2008). However, during the Cold War, fears that communism had infiltrated the United States led to a backlash against Leftist ideology and set the stage for McCarthyism (i.e., “Second Red Scare”), during which federal employees (under President Truman’s “Loyalty Order”) and private citizens (under the direction of the House Un-American Activities Committee) were screened for evidence of communist-related beliefs. The formation of anti-communist legislation in the U.S., and the simultaneous fight for radical improvement in the conditions for the working class worldwide, pushed progressives to the political margins and, in part, created the political split between Leftists and Liberals that lingers in the United States today (Schrecker, 1998, 2002). This split is evident not only in the labels that Liberals and Leftists tend to use, but also in their values (e.g., Pew Research Center, 2021a, 2021b; Proulx et al., 2023).
Like Liberals, Leftists’ values are tied to progressive change. However, unlike Liberals, Leftists are less likely to align with the Democratic party (see Pew Research Center, 2021b), and so they tend to lack mainstream political power in the predominantly two-party system in the United States (e.g., Camejo, 1970; Mikkelson, 2012; Robinson, 2017). Specifically, in the United States, two political parties dominate the government: the Democratic Party and the Republican Party. The Democratic Party typically aligns with the “liberal1” or left end of the political ideology spectrum, advocating for a strong government to regulate business, protect civil liberties, and promote individual human rights (Jost et al., 2008). The Republican Party, on the other hand, aligns most closely with the “conservative” or right end of the spectrum. Conservatives broadly advocate for traditional values, a low degree of government interference, and large support of the private sector (Jost et al., 2008). Thus, a key difference between the two is whether the government (liberals) or individuals (conservatives) should work toward solving society’s major problems (Conover & Feldman, 1984).
Critically, while Liberals and Leftists are both better aligned with Democrats than Republicans, Leftists are against capitalism, the predominant economic ideology endorsed by both parties (to the point of invisibility; Fisher, 2009). Leftist ideology is also predicated on the rejection of Liberalism, which it claims cooperates with existing, problematic institutions and is satisfied with moderate social change (Ture, 1966). Instead, people who self-identify as Leftists tend to cohere around a shared politics of complete and fundamental transformation of social and political institutions (Kazin, 2011), typically organized around principles of anti-capitalism and anti-incrementalist (i.e., radical) attitudes toward social change (e.g., Ture, 1966).
Moral Opposition to the Status Quo
Leftists’ urgent and complete rejection of capitalism and endorsement of radical change may suggest they may morally oppose the political status quo more than Liberals do (i.e., they may consider changing the status quo as a moral mandate; Skitka, 2010). When people morally oppose an issue, this reflects a particular and consequential way of understanding an issue, especially a political one (see Bauman & Skitka, 2009 for a review). When people morally oppose (vs. dislike) an idea or action, they tend to be less tolerant of dissent, resistant to compromise, and more confident in their own beliefs (Bauman & Skitka, 2009; Bennett, 2002; Fernbach et al., 2019; Ryan, 2017). Moral opposition also leads individuals to experience moral emotions like disgust or anger (Tetlock et al., 2000) and experience their judgments as more objective (Goodwin & Darley, 2008; Heiphetz & Young, 2017), universal (Van Bavel et al., 2012), and as requiring action (Skitka & Morgan, 2014) – often in terms of institutional validation, laws, taxes, and funding (Rozin, 1999; for an overview, see Wylie et al., 2022). To the extent that Leftists morally oppose the status quo, they may be more likely to oppose incremental change, and demand drastic, radical change.
When people are morally opposed to a behavior (e.g., smoking cigarettes), they also tend to oppose harm reduction strategies (e.g., vaping) which allow people to carry on with an immoral (in their eyes) behavior at reduced risk to themselves and others (MacCoun, 2013; Wylie et al., 2022). We reasoned that for Leftists who morally oppose the status quo, incremental change (as frequently advocated by Liberals and the Democratic establishment) may resonate like a harm reduction strategy. We predicted that when Leftists morally oppose the status quo, they would be less likely to support Joe Biden in the 2020 U.S. Presidential election, and the Democratic establishment more generally, just as those who morally oppose premarital sex tend to oppose comprehensive sex education (MacCoun, 2013). As with other harm reduction strategies aimed at behaviors that some morally oppose (MacCoun, 2013; Wylie et al., 2022), we expected that a negative relationship between moral opposition to the status quo and support for Biden’s candidacy would emerge for Leftists, but not for Liberals.
Present Research
The current research aimed to explore the differences between Leftists and Liberals in the United States, focusing on a pivotal moment in time: just before the 2020 U.S. presidential election. We selected this moment because we had a specific prediction: that Leftists, but not Liberals who morally oppose the status quo, would be less likely to support Democratic candidate Joe Biden. To better understand this potentially consequential difference, and committed Leftists more generally, we also examined differences in how committed Leftists define themselves and Liberals compared to a Liberal U.S.-based convenience sample. We had specific hypotheses: First, we hypothesized that Leftists would morally oppose the status quo more than Liberals, which would predict their lower support for Joe Biden and desire to abstain from voting. Second, we hypothesized that Leftists would make more structural attributions for poverty and endorse free-market ideology less than Liberals. Third, we hypothesized that Leftists would endorse incremental change less than Liberals. Fourth, we hypothesized that Leftists would have higher support for progressive policies than Liberals.
Finally, inspired by the rich tradition of mixed-method research in psychology (Ponterotto et al., 2013) and the use of case-study analyses in political psychology (Kaarbo & Beasley, 1999), we used a mix of quantitative and qualitative methods to characterize Leftists and compare them to a convenience sample of Liberals recruited from Prolific. This approach enabled a richer understanding of the distinct identities and convictions of committed Leftists in the U.S.—especially as they differ from a more typical research sample in psychology. In this way, we also considered this an exploratory endeavor, seeking to learn about committed Leftists in their own terms.
There are benefits and drawbacks to our approach. We opted to recruit Leftists from highly involved spaces where they are traditionally found and recruited Liberals from Prolific. This provided a snapshot of how highly identified, committed Leftists were feeling leading up to a consequential presidential election compared to a typical convenience sample, but it did not let us make comparisons apples-to-apples with highly identified Liberals. Further, this moment in time in which politics was salient for many Americans, and which was critical for testing our prediction about Leftists’ opposition to Joe Biden’s candidacy and the Democratic establishment in the election, may have exaggerated differences among these two groups.
Method
Participants and Design
The week before the 2020 United States presidential election, we used systematic recruitment from Prolific and snowball sampling from non-Prolific sources. We opted to collect data from as many individuals as possible (within a short time frame), rather than set an a priori target sample size, to ensure adequate sampling for Leftists (an underrepresented group on Prolific). We recruited from Leftist Reddit communities, mutual aid groups (see Spade, 2020), community organizing collectives, and various social media platforms (e.g., Facebook). We advertised our study for left-leaning participants on Prolific; our total Prolific sample yielded 60 self-identified Leftists, and 196 self-identified Liberals. Our community sample yielded 292 self-identified Leftists, and 73 self-identified Liberals. In total we recruited 621 participants (Nleftist = 352; Nliberal = 269). Post-hoc sensitivity analyses (Faul et al., 2009) confirmed that our final sample provided 96% power to detect a significant effect with at least a Cohen’s d of 0.30.
We pre-registered including only participants who passed both attention checks in quantitative analyses, however, much like survey platforms that are more representative of the U.S. population (see Stagnaro et al., 2024), this led to the exclusion of an unexpectedly high number of participants (374; ~60%), likely because they are less accustomed to online surveys for research. Therefore, we decided to deviate from our pre-registered analyses and prioritize having sufficient power to detect our target effects. To mitigate the impact of this issue, we report our quantitative analyses with our full sample (excluding three participants who were under 18) everywhere. Importantly, our results are largely consistent when we compare results based on our full sample with those based on our sample of only those who passed both attention checks. We report results using our full pre-registered exclusions in the Supplementary Materials.
Our final sample for analyses consists of participants ranging in age from 18 to 75 (M = 29.97, SD = 9.98; 43% Female, 33% Male, 11% Non-binary/Other; 13% did not report gender). Of the total participants, 66% identified as White, 8% identified as Asian, 6% identified as Latino(a), 4% identified as Black or African American, < 1% identified as Native American or Native Hawaiian, and 12% did not report their race/ethnicity. We report the full demographic breakdown of our sample in the Appendix. Participants recruited via Prolific were paid approximately $12.58 per hour for their time. Participants recruited through snowball sampling were compensated by being entered into a raffle for a $100 Visa gift card and were provided an additional ideologically aligned charity incentive for community fridges in New York City.
We pre-registered that we would be collecting this data, and our pre-registration reflects a mix of specific hypotheses and exploratory aims. We take care in our results to distinguish between our exploratory and confirmatory analyses. All data exclusions, relevant hypotheses, and analyses were pre-registered on the Open Science Framework (OSF) (see Wylie et al., 2020S). To take advantage of this unique dataset, we also collected and pre-registered other variables that are beyond the scope of this paper. All research data, code for analyses, and supplemental materials (except for the potentially-identifying free responses), are available on the project’s OSF page (see Wylie et al., 2024S).
Measures
We used several scales to test our hypothesis about the relationship between moral opposition to the status quo and support for Joe Biden’s candidacy and the Democratic establishment.
Moral Opposition to the Status Quo
We used a moral opposition scale adapted from previous research on moralization (e.g., Skitka et al., 2005; Tetlock et al., 2000; see Wylie et al., 2022). This measure used the status quo as the target of moral opposition with representative items including “it is morally wrong to push an agenda in line with the political status quo” rated from 1 = Strongly disagree to 7 = Strongly agree (Cronbach’s α = .94). The same items were used to measure moral opposition of the Democratic establishment. All items were again rated from 1 = Strongly disagree to 7 = Strongly agree (Cronbach’s α = .97). Each measure had high inter-item reliability, and all items were retained in the creation of the indices, consistent with previous research (Wylie et al., 2022).
Support for Biden’s Presidency
We created items to capture support for Joe Biden’s Presidency. These items were used as an operationalization of a harm reduction policy. A total of 6 items were used (Cronbach’s α = .92). Representative items included: “I am in favor of voting for Joe Biden to reduce the harm caused by the political status quo” and “Voting for Joe Biden is a good solution to the political status quo”, which were all rated on a scale from 1 = Strongly disagree to 7 = Strongly agree (adapted from Wylie et al., 2022).
Political Identity
To better understand differences in the identity and ideology of committed Leftists and convenience-sampled Liberals, we measured political identity labels, strength of political identity, attitudes toward the economy, and attitudes toward the status quo.
Political Identity Labels
Participants indicated whether they identify as either a Leftist or a Liberal in a two-alternative forced-choice question. We present our results based on this item because we wanted to retain an adequate sample size for each group and because our research question aimed to compare Leftists to Liberals (in line with our pre-registration). Participants were also given the chance to select as many identity labels as they identified with (e.g., Democrat, Socialist; see Supplementary Materials), and they were able to select the single label that best described them (though again, we do not use these labels for analyses, in line with our pre-registration). At the end of the survey, participants filled out the traditional political orientation measure of three 7-point scales (from 1 = Extremely conservative to 7 = Extremely liberal) measuring overall political orientation, political orientation towards social issues, and political orientation towards economic issues (Cronbach’s α = .86). We used this measure because it is used as the main metaphor of the political spectrum in the United States; Liberals are believed to be on the far left of the spectrum and Conservatives are believed to be on the far right of the spectrum.
Identity Centrality and Attitude Strength
Two scales used participants’ best-selected political identity to measure centrality and strength of that identity (using piped text). We used the Identity Centrality Scale (sample item: “The fact that I am [political identity label] is an important part of my identity”, Leach et al. (2008); Cronbach’s α = .91) and the Attitude Strength Scale (sample item: “How strongly do you feel about being [political identity label]?”, Skitka et al., 2005; Cronbach’s α = .89).
Attitudes Toward the Economy
Structural Explanations for Poverty
Six items from the Attributions of Poverty Questionnaire (Bullock et al., 2003) were used to measure structural attributions of poverty. The items involved rating the importance of different factors in contributing to poverty (1 = Not at all important to 7 = Very important). One item was removed due to low inter-item correlation (α = .80, Leftists: Cronbach’s α = .76, Liberals: Cronbach’s α = .78). A representative item is “A capitalist society in which the wealth of some is contingent on the poverty of others.”
Free-Market Mentality Index
The Free-Market Mentality Index measures agreement with pro-capitalist or pro-market ideology. One item was removed, given its low inter-item correlation with other scale items (Cronbach’s α = .76). Example items include “Private ownership of business and industry should be increased”, and “Competition is good; it stimulates people to work hard and develop new ideas” rated on scales from 1 = Strongly disagree to 7 = Strongly agree (Newland, 2018).
Attitudes Toward the Political Status Quo
System Justification
We used Kay and Jost’s (2003) measure of system justification to better understand people’s willingness to justify the current system (Cronbach’s α = .82). A representative item is “Society is set up so that people usually get what they deserve,” rated on a scale from 1 = Strongly disagree to 7 = Strongly agree (Kay & Jost, 2003).
Incrementalism
We generated a scale designed to test support for incremental societal change, (α = .81). A representative item is “Gradual change can be a good thing when it means more immediate improvements in people’s lives,” rated on a scale from 1 = Strongly disagree to 7 = Strongly agree.
Free Response Items
Participants completed three free response measures, each asking them to define either a Liberal, Leftist, and Conservative. The prompt(s) read, “In a few sentences, please define what a [Liberal/Leftist/Conservative] is to you.” Here, we report only Leftists and Liberals writing about themselves and each other.
Grounded Theory Content Analysis
Codebook Development
We also used grounded theory content analysis to investigate the free response data. The codebook was developed by the primary coder using a grounded theory approach (e.g., Charmaz, 1983; Charmaz, 2006; for a review of process and sensitizing concepts, as well as final codebook see Supplementary Materials). As a result of this procedure, we were able to identify both how often a concept (like capitalism) was invoked in the responses, as well as particular ascriptions (e.g., describing Liberals as pro-capitalism).
Grounded Theory Analysis
Grounded theory analysis is particularly useful when conducting research where existing theories and frameworks are limited (or in this case, absent from the literature within social psychology as opposed to other disciplines). Because the nature of this project was exploratory, it was appropriate to use this method to discern patterns in our complex data and to develop our own novel insights. Additionally, it allowed us to determine whether our results aligned with theories from other disciplines without being restricted to them. This analysis complements our quantitative analytic strategies, providing a comprehensive account of participants’ thoughts and feelings around their responses.
Coding Procedure
A team of five researchers convened to analyze the data. The primary coder trained the other four researchers on how to use the codebook and conduct the analyses. To establish inter-rater reliability (IRR), the research team independently coded a set of 75 responses and the primary coder calculated an IRR score (see Miles & Huberman, 1994), measured against the suggested standard of 80% agreement on 95% of codes. The team resolved conflicts, until final IRR was high (> 80%), and then the team independently coded each text response. Moreover, we found that for some analyses there were empty cells, such that for example, Leftists never referred to Leftists as capitalists. To make it possible to run the analyses, we selected a cell at random to recode from a 0 to a 1 (for a similar strategy see Nam et al., 2013).
Results
Analytic Strategy
To describe our results, we will refer to committed Leftists and convenience-sample Liberals to avoid glossing over the substantial differences in recruitment techniques between the groups. We begin with our most directly impactful question: do differences in moral opposition to the status quo predict support for Joe Biden’s presidency? Once we see that these two political identities within the left differentially predict support for the Democratic candidate’s candidacy, we validate the use of these identity labels to predict differences in more typical political spectrum measures of ideology and dig deeper into the identities and ideologies of committed Leftists and convenience-sampled Liberals using both quantitative and qualitative approaches. Finally, after investigating these differences, we end with an analysis of policy support and find that there is considerable overlap in policy support among committed Leftists and convenience-sample Liberals.
All quantitative analyses were conducted using general linear models in R (R Core Team, 2020). Correlations between the dependent variables are reported in the Supplementary Materials. To analyze the free-response items, we used R statistical analysis software (R Core Team, 2020) to preprocess, compare frequencies, and implement an algorithmic approach to compare the language used when committed Leftists described Leftists, committed Leftists described Liberals, convenience-sample Liberals described Leftists, and convenience-sample Liberals described Liberals. We then used human-coded content analysis to compare themes from computer-assisted natural language processing and provide a richer characterization of themes.
Moral Opposition to the Status Quo
We hypothesized that committed Leftists specifically would evince moral opposition to the status quo, which would distinguish them from Liberals, and negatively predict support for Joe Biden’s candidacy in the 2020 U.S. election. We also predicted that moral opposition to the status quo would predict abstaining from voting in the 2020 election, but, contrary to our pre-registered predictions, the total number of abstainers was lower than anticipated, and we opted not to analyze (N = 55; 38 Leftists, 17 Liberals; 11% of the sample)2. For each of these analyses, we used ordinary least squares regression unless otherwise specified.
As predicted, we observed that committed Leftists were more morally opposed to the political status quo, b = −0.94, SE = 0.10, t(615) = −9.09, p < .001, 95% CI [-1.14, -0.73], and the Democratic establishment, b = −2.18, SE = 0.12, t(617) = −17.48, p < .001, 95% CI [-2.42, -1.93], than convenience-sample Liberals (see Figure 1). Leftists were also less likely to vote for Joe Biden in the 2020 U.S. presidential election (OR = 4.41, z = 4.93, p < .001).
Figure 1
Differences in Moral Opposition to the Political Status Quo and Democratic Establishment Between Sampled Leftists and Liberals
Note. Error bars represent ±1 SE.
Next, we tested whether differences in moral opposition predicted differences in attitudes towards the 2020 presidential election, and whether this pattern of results differed for sampled Leftists and Liberals. For moral opposition to the political status quo, we found support for one of our main hypotheses: greater moral opposition of the status quo predicted a lower likelihood of voting for Joe Biden (OR = 2.14, z = 3.08, p = .002), and less support for Biden’s candidacy as a way of reducing the harm caused by the political status quo, b = -0.48, SE = 0.07, t(613) = -7.21, p < .001, 95% CI [-0.61, -0.35], even when controlling for age, gender, harmfulness judgments3 of the status quo, and attitude strength, b = −0.45, SE = 0.09, p < .001, 95% CI [-0.63, -0.27]. This effect was also qualified by a significant exploratory interaction, b = 0.60, SE = 0.10, t(613) = 6.03, p < .001 (see Figure 2). Simple slopes analyses revealed that this was true for committed Leftists only; the more moral opposition to the political status quo that was endorsed, the lower the support for Biden’s candidacy (b = -0.48, SE = 0.07, p < .001). The slope for convenience-sample Liberals trended in the expected direction: greater moral opposition predicted greater support for Biden (b = 0.12, SE = 0.07, p = .100).
Figure 2
Relationship Between Moral Opposition to the Political Status Quo and Support for Joe Biden’s 2020 Candidacy for Sampled Leftists and Liberals
Note. Dotted line represents neutral on the scale. Shading reflects 95% CI.
Contextualizing Differences Between Leftists and Liberals
These findings validated our prediction that differences among left-leaning participants are meaningful and potentially impactful, predicting presidential candidate support. Next, we sought to better understand the nature of the differences between them. First, to get an overall snapshot of the differences between our sample of committed Leftists and convenience-sample Liberals, we used a natural language processing approach to analyze their free response writing, comparing frequency statistics of word importance (term frequency–inverse document frequency; tf-idf) and modeling latent topics for each of the pairs. While the total number of words used by convenience-sample Liberals and committed Leftists differs slightly (WCliberal = 19.76; WCleft = 24.72), tf-idf assesses the importance of a word given the document it occurs within (rather than comparing raw frequency counts). Although these methods also require interpretation from the researcher, the initial algorithmic nature permits a robust and rich comparison of language use between the target groups (see Jackson et al., 2022). Here, we compared language used to describe Liberals and Leftists based on an individual’s self-reported ideology, removing mention of the words “Liberal”, “Left”, and “Leftist”, which occurred in the prompt. The breakdown of the top 5 words associated with each author-target pair are listed below in Figure 3. The most common term used when committed Leftists wrote about Leftists was anti-capitalism, and when they wrote about Liberals, it was incremental change. When convenience-sample Liberals wrote about Leftists they most commonly referred to them as socialist, extreme, and anti-capitalist. When convenience-sample Liberals talked about themselves, themes of choice and science came up most frequently4. We can see from this general approach that the two self-identified groups see differences between them, and do not always agree about what the other stands for. Additionally, we also used latent Dirichlet allocation (LDA) to investigate which topics were most common among the different pairings from a bottom-up approach (that assesses importance of words across the documents they appear) as a complement to our top-down content analysis. We find relatively consistent themes across the two approaches and report the findings from our content analysis in the main text in Table 1. The top 10 topics from topic modeling are reported in the Supplementary Materials.
Figure 3
List of Top 5 Most Important Words Used by Each Author-Target Pair
Note. Calculated using tf-idf, which quantifies importance of a word across a document using both frequency and a weighting term. “Left-Liberal” indicates committed Leftists talking about Liberals. “Liberal-Left” indicates convenience sample of Liberals talking about Leftists.
Table 1
Representative Quotations of Major Themes in Participants’ Writing From Grounded Theory Content Analysis
| Source and Target Group / Theme | Sample Quotation |
|---|---|
| Committed Leftists talking about Leftists | |
| Anti-capitalism | “A leftist is anyone who sees the inherent flaws in capitalism and seeks to sanction or abolish it” |
| Radical change | “A leftist is someone who recognizes the primary role of racial capitalism, white supremacy, patriarchy, and imperialism in our world. They believe that the system must be uprooted and made unrecognizably different to address human needs and wants in a way that rectifies past and current oppression” |
| Status quo | “A general term to signify an array of radical politics which seek to disrupt the status quo.” |
| Committed Leftists talking about Liberals | |
| Capitalism | “Someone who believes in the Ideology of Liberalism, believing in free markets, capitalism, and viewing society through an individualistic lens” |
| Incrementalism | “A person who wants to keep society largely the way it is, with most proposed changes being gradual, incremental, and/or minor” |
| Status quo | “Someone who generally supports social justice & progressive policies but believes in incremental reform and has faith in the Democratic status quo.” |
| Convenience-sample Liberals talking about Liberals | |
| Incrementalism | “Seeking significant changes that speak to minimizing oppression in our world, via channels that currently exist -- as opposed to starting over.” |
| Status quo | “Someone pushing for progressive evolution of laws to address current challenges.” “Someone who believes that constant growth is needed to progressively bend the arc of the universe toward justice.” “Very progressive views often striving for change now.” |
| Convenience-sample Liberals talking about Leftists | |
| Anti-capitalism | “A leftist is someone who doesn't believe in private property. Someone who wants [the] government to take over businesses.” |
| Radical change | “Someone who wants to have a revolution instead of stable steps to change, someone who may support communism” |
| Status quo | “Someone who supports more expansive overhauling of current systems with heavy government involvement/support/regulation.” |
Note. Including committed Leftists writing about Leftists and Liberals, and convenience sample Liberals writing about Leftists and Liberals. Major themes include anti-capitalism, desire for radical vs. incremental change, and opposition to the status quo.
Leftist vs. Liberal Political Identity
Testing Leftist vs. Liberal Identity Measure
For the previous and all following analyses, we used our Leftist vs. Liberal self-identification variable, which predicted differences on the classic 3-item political ideology scale, such that Leftists were more to the left than Liberals on political ideology (MLeft = 6.60, SDLeft = 0.78; MLib = 5.91, SDLib = 0.80; t(538) = -10.07, p < .001, d = 0.87). We prefer the self-identified Leftist vs. Liberal variable because we think it better captures how members of the Left understand themselves. And Leftists do not find this scale to be meaningfully reflective of their political identity. As one Leftist participant wrote in the overall comments section at the end of the survey: “I put liberal but LOL.”
Moreover, when participants indicated the identity labels they identify with from a larger set, (e.g., Democrat, Socialist; see Supplementary Materials), the most commonly chosen labels among our committed Leftists were “Leftist” and “Anti-capitalist”, and the most common label among our convenience-sample Liberals was “Liberal” (see Figure 4). As expected, the most frequently chosen best label among our committed Leftists was “Leftist” and the most frequently chosen among our convenience-sample Liberals was “Liberal” further validating our primary predictor. There were distinct patterns in label usage, such that some labels were popular among committed Leftists but not convenience-sample Liberals (e.g., Communist, Anarchist) and some were popular among Liberals but not Leftists (e.g., Democrat; for frequency of all label choices, see Supplementary Materials).
When we content-analyzed participants’ own writing about ‘what makes a Leftist’ and ‘what makes a Liberal’, we found that participants talked about whether Leftists and Liberals are separate groups. Using a standard generalized linear model, we found a significant participant (sampled Leftist vs. sampled Liberal writer) by target identity (writing about Leftists vs. Liberals) interaction, primarily driven by convenience-sample Liberals, z = -2.31, p = .021. Liberals (8%) were more likely to describe the two groups as similar, compared to Leftists (2%). For example, one of our sampled Liberals wrote “[A Leftist is] a liberal to the extreme.” Convenience-sample Liberals referred to this comparison more often when describing Leftists (12%) than Liberals (4%), χ2(1, N = 511) = 8.54, p = .003, and sampled Leftists were unlikely to say that the groups were the same regardless of who they were describing (Leftist = 2%; Liberal = 3%), p = .404 (see Figure 4). This, taken together with the natural language processing analysis, suggests that the boundary between the two groups is more salient for our committed Leftists than convenience-sample Liberals.
Figure 4
Frequency of Language Use When Sampled Leftists and Liberals Described Themselves as Similar or the Same
Note. Graph is scaled for clarity. Error bars represent 95% confidence intervals.
Political Identity Centrality and Attitude Strength
Committed Leftists rated their political identity as more central to their identity (M = 5.38, SD = 1.47) compared to convenience-sample Liberals (M = 4.54, SD = 1.66), t(615) = -6.68, p < .001, d = 0.54. Similarly, our sampled Leftists showed greater strength in their identification with their political labels (M = 5.49, SD = 1.30), compared to sampled Liberals (M = 4.73, SD = 1.51), t(612) = -6.67, p < .001, d = .54. However, these results must be interpreted with some skepticism since the majority of our Leftists were recruited from spaces about being a Leftist, whereas the majority of our Liberal participants were recruited from Prolific (see Supplementary Materials for differences between Leftists recruited from community samples compared to Leftists recruited on Prolific; sampled Liberals were similar across both sampling methods, but we did not reach out to Liberal special interest groups). Because of this, all subsequent analyses are reported alone and with identity centrality included as a covariate in an exploratory manner.
Primary Results: Differences Between Leftists and Liberals
Attitudes Towards the Economy
Committed Leftists made more structural attributions for poverty (M = 6.52, SD = 0.71) compared to convenience-sample Liberals (M = 5.87, SD = 0.97), t(515) = -8.83, p < .001, d = 0.79, and this effect remained when we adjusted for identity centrality, t(513) = -7.83, p < .001, ∆R2 = .01. Sampled Liberals, on the other hand, endorsed free-market mentality (M = 4.19, SD = 0.97), whereas committed Leftists did not (M = 2.77, SD = 1.05), t(521) = 15.63, p < .001, d = 1.40 (see Figure 5). This effect remained when we adjusted for identity centrality, t(519) = 14.37, p < .001, ∆R2 = .04.
Figure 5
Mean Differences Between Sampled Leftists and Liberals in Terms of Endorsement of Free Market Principles (Left), and Structural Attributions for Poverty (Right)
Note. Error bars represent ±1 SE.
Attitudes Toward Change and the Status Quo
Sampled Liberals endorsed incrementalism (M = 4.34, SD = 0.87), more than sampled Leftists did (M = 3.04, SD = 1.01), t(511) = 15.13, p < .001, d = 1.37, and were more likely to justify the current system (M = 2.61, SD = 1.05) compared to committed Leftists (M = 1.81, SD = 0.80), t(523) = 9.87, p < .001, d = 0.88, though both were below the midpoint (see Figure 6). Differences remained, after adjusting for identity centrality for both incrementalism, t(509) = 13.79, p < .001, ∆R2 = .02, and system justification, t(521) = 9.08, p < .001, ∆R2 = .00.
Figure 6
Mean Differences Between Sampled Leftists and Liberals in Terms of Desire for Incremental Change (Left) and System Justification (Right)
Note. Error bars represent ±1 SE.
Exploratory Results Supplementing Main Findings
Attitudes Towards the Economy
Sampled Leftists and Liberals spontaneously mentioned differences in economic ideology when describing themselves and each other. Specifically, in their free responses, we found a significant participant (committed Leftist vs. convenience-sample Liberal writer) by target identity (writing about Leftists vs. Liberals) interaction, z = -3.25, p = .001. Simple effects analyses revealed that committed Leftists made references to capitalism and anti-capitalism at roughly the same rate when describing both targets (Leftists = 33%, Liberals = 34%), z = 5.83, p < .001, while convenience-sample Liberals mentioned capitalism least when describing other Liberals (4%) compared to Leftists (13%), χ2(1, N = 511) = 41.41, p < .001. Capitalism is in the background for sampled Liberals, but the foreground for committed Leftists.
We also found that participants described Leftists (24%) but never Liberals (0%), z = 5.86, p < .001 as anti-capitalist, and described only Liberals as capitalists (20%), z = -4.81, p < .001. Only one person, a self-identified Liberal, described Leftists as capitalists. Many committed Leftists’ description of what it means to be a Leftist included mention of anti-capitalism. For example, one sampled Leftist wrote, “A leftist is anyone who in theory opposes the capitalist mode of production and the free market system.” Taken together, these results suggest that economic ideology (specifically, anti-capitalist ideology) is a guiding difference in how people understand Leftists and Liberals (see Figure 7).
Figure 7
Frequency of Language Use of Sampled Leftists and Liberals Describing Themselves and Each Other
Note. Specifically, how often they reference capitalism (including anti-capitalism; top left) and specific attributions of holding capitalist (top right) and anti-capitalist ideology (bottom). Error bars represent 95% CI.
Attitudes Toward Change and the Status Quo
We first tested for differences in overall frequency of language discussing radical (vs. incremental) change in society. We found a significant participant (sampled Leftist vs. Liberal writer) by target identity (writing about Leftists vs. Liberals) interaction effect, z = -2.22, p = .027. Simple effects analyses revealed that committed Leftists were equally likely to reference incrementalist ideology when talking about Leftists (40%) compared to Liberals (43%), p = .354. However, convenience-sample Liberals referenced radical change more frequently when talking about Leftists (32%) compared to Liberals (24%), χ2(1, N = 511) = 4.56, p = .033.
There was also a main effect of target identity on attributions of radicalism, z = 6.52, p < .001. Participants identified Leftists as radicals (25%) but not Liberals (0.4%; see Figure 8). For example, one participant who self-identified as Leftist wrote, “[A Leftist is] someone who believes we aren’t progressing fast enough and need to be more radical because non-radical methods aren’t working.” Radicalism is perceived as positive among committed Leftists, but negative among convenience-sample Liberals. We also examined differences in attributions of opposition to the status quo (see Supplementary Materials).
Figure 8
Frequency of Language Use When Sampled Leftists and Liberals Talk About Themselves and Each Other
Note. Specifically, how often ideas of incremental or radical change are referenced (top left), and how often Leftists and Liberals are described as supporting incremental change (top right) and radicalism (bottom). Error bars represent 95% CI.
Policy Support
Finally, despite these psychological differences, there was considerable overlap in policy support in our sample. We measured differences between sampled Leftists and Liberals by measuring endorsement of seven policy issues that we thought would be more strongly endorsed by Leftists than Liberals. Indeed, we found that, while Liberals and Leftists both broadly supported all policies, committed Leftists supported them more than Liberals, t(617) = 17.53, p < .001 (individual policy support differences, all p’s < .001, see Table 2). These differences remained when accounting for identity centrality, t(613) = 15.78, p < .001, ∆R2 = .05, but they were notably small in magnitude and really represent a point of relative agreement.
Table 2
Means and Standard Deviations of Self-Reported Policy Support Between Sampled Leftists and Liberals
| Policy | Political Identity | p | Cohen’s d | |||
|---|---|---|---|---|---|---|
| Leftist | Liberal | |||||
| M | SD | M | SD | |||
| Universal Healthcare | 6.85 | 0.56 | 6.41 | 1.06 | < .001 | 0.55 |
| Green New Deal | 6.68 | 0.72 | 6.13 | 1.18 | < .001 | 0.58 |
| Abolishing ICE | 6.61 | 1.03 | 5.18 | 1.72 | < .001 | 1.04 |
| Abolishing the Police | 5.84 | 1.70 | 3.48 | 1.86 | < .001 | 1.34 |
| Decriminalizing Sex Work | 6.49 | 1.15 | 5.69 | 1.48 | < .001 | 0.61 |
| De-privatizing Private Property | 6.01 | 1.33 | 4.38 | 1.61 | < .001 | 1.12 |
| Nationalizing Big Business | 5.91 | 1.39 | 4.38 | 1.63 | < .001 | 1.02 |
Note. Larger values indicate greater policy support. Alternating policies are bolded for ease of reading.
Discussion
The current research investigated the psychology of highly committed Leftists in the United States during the weeks preceding the 2020 U.S. Presidential Election and compared them to a convenience sample of self-identified Liberals in the United States. We hypothesized and found that committed Leftists are more morally opposed to the status quo than convenience-sampled Liberals, and that for Leftists, but not Liberals, this negatively predicts support for then-candidate Joe Biden’s candidacy. To better understand this potentially consequential difference between the two groups, we used a mixed method approach to examine the differences and similarities between the identity and ideology of these two groups. Specifically, we measured attitudes about the economy, attitudes about change and the status quo, moral opposition to the status quo, and policy preferences. First, we found that economic ideology (specifically, anti-capitalist ideology) is a guiding difference in how people understand Leftists and Liberals, as evidenced by Leftists’ endorsement of structural explanations for poverty and lower endorsement of free-market ideology. We also found differences in Leftists’ and Liberals’ endorsement of change and the status quo. Liberals were more likely to endorse incrementalism and justify the current political system than Leftists, whereas Leftists sought radical change. These preferences for the status quo (or lack thereof) provide greater context for understanding why committed Leftists might oppose Joe Biden’s candidacy, in a manner akin to other instances of lack of support for harm reduction policy (MacCoun, 2013; Wylie et al., 2022). Importantly though, we found very little evidence of actual abstention from voting as a result.
Limitations and Future Directions
This research is meant to be a generative rather than an exhaustive investigation into Leftists in the United States. A significant limitation of the current research is the representativeness of our sample. We recruited Leftist participants primarily from online Leftist communities, while the majority of Liberal participants came from Prolific. This comparison is valuable because many psychological studies are conducted on Prolific, and this allows us to understand committed Leftists compared to a sample typical of psychology research. Yet, it does not allow us to compare committed Leftists and committed Liberals. To partly compensate for differences in recruitment, we report all analyses with identity centrality as a covariate to show that none of our analyses hinge on this difference. But, nonetheless, the recruitment differences limit the extent to which we can generalize our findings to other left-leaning individuals in the United States. Future research would benefit from a more representative sample of U.S. Leftists and Liberals.
The demographic composition of our sample may also limit the generalizability of our findings. Our sample of Leftists was disproportionately White (72%) and from a middle to upper class background (52%). It is possible that people of color who identify as Leftist may not prioritize the same policies and social issues as White Leftists. Similarly, there may be further differences when looking at different political subgroups. For example, there may be differences among Liberals who identify with the Democratic Party (versus those who do not) or Leftists who identify with different labels (e.g., post-materialist Leftists, Socialists, Communists). There may also be some Leftists, especially outside of the U.S., that prioritize issues related to economics, but not social issues. This might make Leftists in the U.S. feel further away, or off the spectrum, than Leftists in Europe. Future research can explore differences and distinctions between different subgroups of Leftists and Liberals–including racial, socioeconomic, and political subgroups–or look at how leftism varies across different parts of the world.
Finally, it is important to note that we collected this data at a moment of heightened political salience, in the weeks before the 2020 U.S. presidential election, and that there are many ways to characterize Leftists that we did not yet touch on. For example, there may also be differences among Leftists with regard to authoritarianism; some scholars suggest that the Left cannot be authoritarian (e.g., Nilsson & Jost, 2020) while others suggest that some on the Left can be authoritarian (Costello et al., 2022) or even that a leftist revolution cannot be non-authoritarian (e.g., Engels, 1978). Moreover, the terms “radical” and “extremist” are sometimes used interchangeably in scholarship and daily life, but these terms refer to different constructs, which Leftists in our study, but not Liberals, were sensitive to. Future research would benefit from exploring these beliefs as well as endorsement of authoritarianism, contributing to a long-standing debate within the study of those who identify with the political left.
Implications
Despite these limitations, this research makes several practical and theoretical contributions. Overall, these findings provide an illustrative example of the power of a mixed-method approach to group differences in the moral and political space. We find that traditional measures of political ideology do not fully capture the range of political identities in the United States (see Costello et al., 2023). This research draws attention to the need for measuring ideology with self-selecting political identity labels in addition to the more common political ideology Likert scales. If researchers are using a typical political ideology spectrum from liberal to conservative in university subject pools, they may not be capturing the full range of political identities among their participants; there may be Leftists in their sample, people who identify as progressives, that may be selecting “extremely liberal” when it does not well-represent their identity as they would describe it.
Consistent with extant theorizing (e.g., Jost, 2021), we also find that there is disagreement in how left-leaning political groups see each other: Liberals perceive Leftists as different only in degree, whereas Leftists perceive Liberals as a different political group altogether. We find that one of the primary ways that Leftists distinguish themselves from Liberals is in their moral opposition to the status quo; Leftists may be unwilling to vote for Democratic candidates as a way to minimize the harm associated with the political status quo. This research can help stimulate discussions around how to correctly characterize these groups (and others) within left-leaning ideology. Prior research has shown that when different subgroups are properly acknowledged, cooperation and collective outcomes are improved (Dovidio et al., 2007; Roberge & Van Dick, 2010). We hope that in characterizing Leftists and Liberals, we can help foster cooperation between the two groups.
Conclusion
We sought to characterize self-identified Leftists in terms of how they see themselves and what policies they support. We compared them to a Liberal convenience sample recruited from Prolific. On quantitative scales and in their own words, highly committed Leftists see themselves as a distinct identity subgroup within Left-wing political ideology in the United States and outside the traditional ideology spectrum–defined by their rejection of capitalism and their preference for radical (vs. incremental) change.