US Voting Patterns Are Shifting. But It’s Not Simply “Class Dealignment.”

It’s not that partisan voting patterns are becoming decoupled from class — it’s that a complicated new set of alignments, rooted in the social and occupational structures of a postindustrial economy, is emerging in the United States.

A man fills out a ballot at a voting booth on May 17, 2022 in North Carolina. (Sean Rayford / Getty Images)

By the time he published his classic book Political Man: The Social Bases of Politics in 1960, the political sociologist Seymour Martin Lipset was no longer a socialist. Even so, Lipset’s work often bore the unmistakable stamp of his youthful membership in the Young People’s Socialist League during the Great Depression. In Political Man, Lipset developed an influential theory of elections as a “democratic translation of the class struggle.” “On a world scale,” Lipset argued, “the principle generalization that can be made is that parties are primarily based on either the lower classes or the middle and upper classes.”

Such a pattern, Lipset insisted, also prevailed in the United States despite widespread pretensions to the contrary. “The Democrats from the beginning of their history have drawn more support from the lower strata of society, while the Federalist, Whig, and Republican parties have held the loyalties of the more privileged groups.”

If we believe a chorus of politicians and pundits, this historic pattern is being reversed. As the results rolled in on election night 2020, Republican senator and Potemkin populist Josh Hawley tweeted, “We are a working class party now. That’s the future.” A representative item from Axios breathlessly reported, “We’re seeing a political realignment in real time. Democrats are becoming the party of upscale voters concerned more about issues like gun control and abortion rights. Republicans are quietly building a multiracial coalition of working-class voters, with inflation as an accelerant.”

Similar pronouncements also come from the Left. For example, my friend and fellow Jacobin contributor Matt Karp argued that the 2020 election results reflected “America’s headlong march toward a party system entirely decoupled from the politics of class.” Indeed, the latest print issue of Jacobin is organized around the concept of class dealignment and its implications for left-wing politics.

It’s clear that many of the patterns of twentieth-century politics have been disrupted, not just in the United States but across the capitalist democracies. Political systems have become more fragmented, and electoral competition has become more volatile. Seemingly impregnable left-wing “walls” have been breached in places like Northern England and the American Rust Belt. Traditional political loyalties appear to have weakened, giving rise to new political identities that cut across partisan lines and put the parties’ traditional electorates up for grabs.

In this sense, the concept of class dealignment gets at something important about contemporary political life. But is it an adequate framework for understanding the political dynamics we are grappling with? I don’t think it is.

In a Jacobin interview in the “Dealignment” issue, for example, the economists Amory Gethin, Clara Martínez-Toledano, and Thomas Piketty maintain that “we do not see a dissipation of income-based left-wing voting in Western democracies . . . low-income voters continue to support left-wing parties just as much as in the past. . . .” (emphasis in original).

Their contention is supported by other scholars exploring similar questions. In a recent study of European social democracy’s electoral crisis, political scientists Tarik Abou-Chadi, Reto Mitteregger, and Cas Mudde find that “social democratic parties remain the party family that receives the highest support among both production and service workers,” followed by the center-right and the radical right in third. The traditional left parties have certainly lost voters, and in countries like France or the Netherlands, they have become an electoral also-ran.

But Abou-Chadi, Mitteregger, and Mudde find that they are losing voters in all directions, not just to the right. “Especially in the 2010s,” they observe, “social democratic parties lost the largest share of former supporters to more progressive parties” including the Greens and the radical left.

If there is still a clear link between lower-income voters and parties of the Left and center-left, then the class dealignment perspective becomes difficult to sustain. At the same time, the old patterns of electoral competition clearly have changed. The fly in the ointment is educational attainment. Republican claims that the GOP is a workers’ party now, or left-wing claims that the Democrats have become the snooty party, rest largely on the conflation of educational attainment and class location.

There was a time when this methodological tactic made a good deal of sense. As recently as the 1960s, less than 8 percent of Americans had at least a bachelor’s degree, while barely more than 40 percent graduated from high school. Higher education was largely reserved for a small sliver of the population, degree holders were largely middle to upper class, and they voted in overwhelming numbers for the Republican Party. Meanwhile, Democrats dominated among the much larger proportion of voters with no more than a high school diploma.

By 2021, roughly 38 percent of all Americans had at least a bachelor’s degree, and over 63 percent had at least some kind of postsecondary education. If we narrow the focus to employed civilians, those numbers rise even further, reaching 70 percent with at least some postsecondary education — including the 45 percent with at least a bachelor’s degree. In 2020, President Joe Biden won well over half of voters with at least four-year degrees but struggled among voters with just some college or less. In this light, Donald Trump’s declaration of love for the “poorly educated” was less a gaffe than a characteristically frank recognition of a major part of his base.

Educational attainment is surely an important dimension of class. But it is not, in itself, determinative of one’s location in the class structure. There are many others, including a rather important one that often goes overlooked in these discussions: income.

Data from the Pew Research Center shows that Democrats continue to do well among lower-income voters. Indeed, voters with annual household incomes below $50,000 are among the party’s more reliable voting blocs. In 2020, Biden beat Trump by a nine-point margin among these voters (53-44) but lost to Trump among voters in the $50,000-$100,000 range by five points (47-52). It’s true that Biden trounced Trump by a thirteen-point margin (56-43) among voters with household incomes above $100,000. But this is the smallest income group in the electorate, and many of these voters may not be “elites” in any meaningful sense. A household headed by a registered nurse and a Teamster will cross this threshold in many parts of the country.

Gethin, Martínez-Toledano, and Piketty zero in on the crux of the dilemma. “Our finding,” they conclude, “is precisely that there has been a divergence between income and education.” While lower-income voters continue to vote for left and center-left parties in large numbers, “lower-educated voters have shifted toward conservative parties.” By conflating educational attainment with class location, advocates of the class dealignment perspective tend to misunderstand the precise nature of this shift.

Cleavages Through the Class Structure

The Piketty group’s work is very impressive and useful, but it has certain shortcomings that limit its explanatory power. Political scientists and sociologists have been working on the same questions in recent years. While their research reaches similar conclusions, it fills in some important gaps the Piketty group leaves unaddressed — namely, the precise ways occupation, education, and income interact to create political cleavages through the class structure.

The Swiss sociologist Daniel Oesch pioneered recent research into the impact of changing employment structures and occupational tasks on class voting. In an important 2006 study, Oesch sought to update frameworks of class analysis based on concepts and assumptions from the “golden age” of postwar industrial capitalism. Using Britain, Germany, Sweden, and Switzerland as case studies, he finds that “different cleavages run through the class structure as regards party support,” sending the various fractions of the working and middle classes in different partisan directions.

“Sociocultural professionals” like teachers, social workers, and nurses lean leftward, some toward the “traditional left” social democratic parties, others toward the Greens and the radical left. Business managers continue to vote center-right. Lower-skilled operatives and service workers still vote for the traditional left in large numbers, but a section of this class fraction is open to appeals from the radical right. Overall, Oesch finds evidence of continued class voting, though it “seems to involve new alliances (sociocultural professionals and the New Left) and to reverse traditional links” in some cases, as in Swiss workers shifting from the Social Democrats to the right-wing Swiss People’s Party.

The political scientists Herbert Kitschelt and Philipp Rehm have built on Oesch’s work to further refine our understanding of class-based political cleavages. In their 2014 paper “Occupations as a Site of Political Preference Formation,” they analyze how voters’ political views are shaped, to a significant extent, by the work they do in their occupations. They pay particular attention to higher-skill professionals of various kinds, as these occupational groups now constitute a large and growing proportion of employees in postindustrial societies.

As Kitschelt and Rehm note, professionals and semiprofessionals with postsecondary degrees now “equal or exceed the proportion of less skilled wage earners in the labor force” of countries like the United States. According to Census Bureau data, 43 percent of all employed civilians are now in management, professional, and related occupations, constituting the single largest occupational group by far. (The next largest group, sales and office occupations, constitutes 19 percent.) Meanwhile, the proportion employed in production occupations has dwindled to barely 5 percent.

Kitschelt and Rehm predict that managers who spend their days directing subordinates and maximizing profit will not support economic redistribution, but might be amenable to moderate approaches to governance and citizenship questions. They therefore tend toward parties of the center-right. Technical professionals in engineering, design, or technology are not as strongly opposed to redistribution as managers, but are not consistently in favor of it; at the same time, they are more libertarian and inclusive on governance and citizenship. These professionals tend to be politically centrist. Finally, interpersonal or “sociocultural” professionals are more willing to support redistribution than professionals in other fields, while being the most libertarian on governance and citizenship. They tend to support parties of the center-left, and in some cases the radical left.

On these grounds, Kitschelt and Rehm make the provocative claim that it no longer makes sense to speak of a coherent “middle class” — or a “professional-managerial class” for that matter — at all.

Workers with fewer credentials, lower incomes, and less room for autonomy on the job tend to share a clear preference for redistributive economic policy, regardless of occupation. But many in this broad group are also disposed to relatively authoritarian/exclusive approaches to governance and citizenship questions, particularly if they do not have a college degree. As such, Kitschelt and Rehm posit that they tend to be politically situated on a range from the center to the Left. But a significant fraction of these workers are open to appeals to the radical right, particularly if the Left doesn’t clearly differentiate itself from the center right on economic questions.

Changing Patterns of White Partisanship

Donald Trump’s shocking 2016 victory inaugurated a cottage industry of commentary on the political behavior of the “white working class,” much of it useless. Fortunately, Kitschelt and Rehm applied the insights of their 2014 work to the question of changing patterns of partisanship among white voters in the United States, and illuminated the ways income and education interact in the process of political realignment.

In a key 2019 study on realignments in white partisan support, Kitschelt and Rehm describe the US electorate in terms of income (low vs. high) and education (low vs. high) and model their interaction. The dramatic expansion of higher education, they contend, has disrupted the traditional New Deal alignment and shaped the political demands of four main groups: low-education/low-income, low-education/high-income, high-education/low-income, and high-education/high-income.

The main conclusion of the article is that the two parties’ core constituencies during the New Deal order (low-education/low-income voters for the Democrats, high-education/high-income voters for the GOP) have become swing groups, while the former swing groups (high-education/low-income for the Democrats, low-education/high-income for the GOP) have become the core.

Lower-education voters of all income levels are shifting toward the Republicans at different speeds, while lower-income voters are split. Lower-income voters with higher levels of educational attainment are strengthening their identification with the Democrats, while those with both lower income and fewer credentials — particularly but not exclusively among white voters — can be swayed to vote for the Right depending on a candidate’s electoral appeal. Trump won a significant amount of support from low-education, low-income white voters in 2016, they contend, not just because of his reactionary views on governance and citizenship, but because he was widely perceived as not a typical Republican on economic policy issues.

In light of this research, we see that the Piketty group is correct that lower-income voters still support parties of the Left and center-left in large numbers — but also that voting behavior within this group is often contingent on educational attainment. A similar dynamic holds among high-income voters. Those with lower levels of educational attainment but high incomes (see the local gentry who often led Trump’s flotillas of “beautiful boaters”) have become ever more intensely reactionary, while those with both high education and income are torn between opposition to redistribution and support for relatively liberal approaches to governance and citizenship questions.

What’s emerging is not a dealignment of partisan loyalties, but a complicated new set of alignments rooted in the social and occupational structures of a postindustrial economy.

Strategic Implications

There is no going back to alignments that shaped the politics of industrial capitalism. Political conflict today is not organized around a relatively coherent working class on the Left against a relatively coherent bourgeois bloc on the Right. Besides, even in its halcyon days, the socialist left stood as “the champion of all the exploited and oppressed,” in the words of the old Erfurt Program — not just its industrial working class base. Noneconomic issues are simply too salient for too many people to be sidestepped or downplayed, and these are often related to questions of economic policy and redistribution in various ways.

Oesch and political scientist Line Rennwald argue in a 2018 paper that contemporary political conflict is shaped by class voting in a “tripolar” political space divided between the Left, center-right, and radical right. It continues to pit the traditional left and right against each other on economic policy, but at the same time, it pits the radical right against social democrats and conservatives on governance and citizenship questions. The radical right challenges the Left for votes from the working class, and the mainstream right for votes from small business owners and adjacent groups.

Their framework is crucial for understanding electoral competition today, because it details how the relationship between social classes and their traditional parties has been weakened and modified without disappearing.

As Kitschelt and Rehm put it in their most recent paper, “the core constituencies of left and right parties have changed, and the old core groups have become cross-pressured groups” open to switching their votes from election to election. Members of what they call the “new working class” — typically workers in sociocultural services, disproportionately women with professional training and high educational attainment but modest salaries — “provide the backbone of a progressive push in favor of both economic redistribution as well as libertarian social governance and cultural tolerance for difference.”

The fact that voters fitting this profile have become the core leftist constituency across the capitalist democracies suggests that left-wing parties are responding to structural processes as much as they’re choosing to appeal to specific sections of the electorate. In their study of shifts in white partisanship, Kitschelt and Rehm draw on American National Election Study data to estimate that the low-education/low-income group’s size dropped from 50-60 percent in the 1950s to under 40 percent in 2016, while the high-education/low-income group grew from “almost nothing to one-sixth of white respondents.”

In such a context, the Left has to make some kind of appeal to the “professional managerial class” if it wants to win elections, particularly in a winner-take-all, two party system. Karp is right to insist that the Clinton-Blair-Schröder turn to neoliberal politics was a choice, and a terrible one at that. But it was, at least in part, an attempt to deal with a real electoral dilemma the Left hasn’t yet figured out.

Even though members of the “new” and “old” working classes tend to share a preference for progressive economic policy, they may still prefer specific policy approaches that could be in tension with each other. Along these lines, the political scientists Silja Häusermann and Hanspeter Kriesi argue that the “economic conflict dimension can no longer be analyzed in terms of ‘more versus less welfare spending.’” Members of the “new” working class might want to prioritize social investments in higher education or labor market training that may not have an immediate material payoff; members of the “old” working class may want to prioritize public pensions, unemployment benefits, and other social protections aimed at near-term consumption.

Anyone with experience formulating a bargaining strategy for a large and diverse union local will immediately recognize this kind of dilemma. These are not necessarily zero-sum trade-offs, but different segments of the Left’s social base will often have different priorities, and balancing them all will require constant coalition-building and maintenance.

These dynamics pose a challenge to left-wing politics in all capitalist democracies, but the precise nature of the challenge will vary according to local conditions including electoral systems. In proportional representation systems, the “Brahmin left” and the “traditional left” (or the center-right and the radical right, for that matter) can support their own parties with their own political profiles, and work to form coalitions when possible or necessary. Here in the United States, all social and political groups are shoehorned into one of two big coalition parties, so this fragmentation is expressed as intraparty conflict.

This situation may not be ideal, but there’s no way around it for the foreseeable future. We cannot bank on the intervention of dei ex machina like electoral reform or a labor upsurge that somehow shatters and supplants the two-party system.

The political unity of the working class and the old left is often overstated in retrospect. The once-mighty Swedish Social Democrats, for example, won more than 50 percent of the vote just twice in the twentieth century. They formulated concepts like folkhemmet (“people’s home”) and the “wage-earner alliance” to broaden their appeal beyond their core base in the industrial working class. There is a long history of working-class conservatism, and left-wing movements have everywhere won support from professionals, intellectuals, and class traitors.

There have always been divisions in the working class and on the Left along numerous lines, and the degree of social fragmentation and differentiation has only increased in the last few decades. This makes the politics of alliance and coalition even more important than it was in the past. Considering the ongoing degradation of professional work, it may even hold better majoritarian possibilities for socialist politics in the long term — provided, of course, that the Left does a better job of organizing itself and winning back some layers of lower-education voters.

Toward the end of his last book, Socialism: Past and Future, Michael Harrington surveyed the transformations that restructured the working class, undermined organized labor, and gave rise to new social movements based on race, ethnicity, gender, age, and ecology. Writing over thirty years ago, he recognized that “one of the most fundamental challenges to a new socialism, therefore, is to find a way to politically unify social forces that in recent years have spent a fair amount of time frustrating and fighting one another.” The Left will not advance unless it recognizes this terrain and deals with it as it is.