the big picture

How the Diploma Divide Is Remaking American Politics

Education is at the heart of this country’s many divisions.

Photo-Illustration: Intelligencer; Photos: Getty
Photo-Illustration: Intelligencer; Photos: Getty

Blue America is an increasingly wealthy and well-educated place.

Throughout the second half of the 20th century, Americans without college degrees were more likely than university graduates to vote Democratic. But that gap began narrowing in the late 1960s before finally flipping in 2004.

John F. Kennedy lost college-educated voters by a two-to-one margin yet won the presidency thanks to overwhelming support among white voters without a degree. Sixty years later, our second Catholic president charted a much different path to the White House, losing non-college-educated whites by a two-to-one margin while securing 60 percent of the college-educated vote. The latest New York Times/Siena poll of the 2022 midterms showed this pattern holding firm, with Democrats winning 55 percent of voters with bachelor’s degrees but only 39 percent of those without.

A more educated Democratic coalition is, naturally, a more affluent one. In every presidential election from 1948 to 2012, white voters in the top 5 percent of America’s income distribution were more Republican than those in the bottom 95 percent. Now, the opposite is true: Among America’s white majority, the rich voted to the left of the middle class and the poor in 2016 and 2020, while the poor voted to the right of the middle class and the rich.

Graph by Ohio State University political scientist Thomas Wood.

In political-science parlance, the collapse of the New Deal–era alignment — in which voters’ income levels strongly predicted their partisan preference — is often referred to as “class dealignment.” The increasing tendency for politics to divide voters along educational lines, meanwhile, is known as “education polarization.”

There are worse things for a political coalition to be than affluent or educated. Professionals vote and donate at higher rates than blue-collar workers. But college graduates also comprise a minority of the electorate — and an underrepresented minority at that. America’s electoral institutions all give disproportionate influence to parts of the country with low levels of educational attainment. And this is especially true of the Senate. Therefore, if the coalitional trends of the past half-century continue unabated — and Democrats keep gaining college-educated votes at the expense of working-class ones — the party will find itself locked out of federal power. Put differently, such a development would put an increasingly authoritarian GOP on the glide path to political dominance.

And unless education polarization is substantially reversed, progressives are likely to continue seeing their reform ambitions pared back sharply by Congress’s upper chamber, even when Democrats manage to control it.

These realities have generated a lively intra-Democratic debate over the causes and implications of class dealignment. To some pundits, consultants, and data journalists, the phenomenon’s fundamental cause is the cultural divide between educated professionals and the working class. In their telling, college graduates in general — and Democratic college graduates in particular — tend to have different social values, cultural sensibilities, and issue priorities than the median non-college-educated voter. As the New York Times’s Nate Cohn puts the point, college graduates tend to be more cosmopolitan and culturally liberal, report higher levels of social trust, and are more likely to “attribute racial inequality, crime, and poverty to complex structural and systemic problems” rather than “individualist and parochial explanations.”

What’s more, since blue America’s journalists, politicians, and activists are overwhelmingly college graduates, highly educated liberals exert disproportionate influence over their party’s actions and identity. Therefore, as the Democrats’ well-credentialed wing has swelled, the party’s image and ideological positioning have grown more reflective of the professional class’s distinct tastes — and thus less appealing to the electorate’s working-class majority.

This theory does not sit well with all Democratic journalists, politicians, and activists. Some deny the existence of a diploma divide on cultural values, while others insist on its limited political salience. Many progressives attribute class dealignment to America’s pathological racial politics and/or the Democrats’ failures of economic governance. In this account, the New Deal coalition was unmade by a combination of a backlash to Black Americans’ growing prominence in Democratic politics and the Democratic Party’s failures to prevent its former working-class base from suffering decades of stagnant living standards and declining life expectancy.

An appreciation of these developments is surely indispensable for understanding class dealignment in the United States. But they don’t tell the whole story. Education polarization is not merely an American phenomenon; it is a defining feature of contemporary politics in nearly every western democracy. It is therefore unlikely that our nation’s white-supremacist history can fully explain the development. And though center-left parties throughout the West have shared some common failings, these inadequacies cannot tell us why many working-class voters have not merely dropped out of politics but rather begun voting for parties even more indifferent to their material interests.

In my view, education polarization cannot be understood without a recognition of the values divide between educated professionals and working people in the aggregate. That divide is rooted in each class’s disparate ways of life, economic imperatives, socialization experiences, and levels of material security. By itself, the emergence of this gap might not have been sufficient to trigger class dealignment, but its adverse political implications have been greatly exacerbated by the past half-century of inequitable growth, civic decline, and media fragmentation.

The college-educated population has distinct ideological tendencies and psychological sensibilities.

Educated professionals tend to be more socially liberal than the general public. In fact, the correlation between high levels of educational attainment and social liberalism is among the most robust in political science. As early as the 1950s, researchers documented the tendency of college graduates to espouse more progressive views than the general public on civil liberties and gender roles. In the decades since, as the political scientist Elizabeth Simon writes, this correlation has held up with “remarkable geographical and temporal consistency.” Across national boundaries and generations, voters with college degrees have been more likely than those without to support legal abortion, LGBTQ+ causes, the rights of racial minorities, and expansive immigration. They are also more likely to hold “post-material” policy priorities — which is to say, to prioritize issues concerning individual autonomy, cultural values, and big-picture social goals above those concerning one’s immediate material and physical security. This penchant is perhaps best illustrated by the highly educated’s distinctively strong support for environmental causes, even in cases when ecological preservation comes at a cost to economic growth.

Underlying these disparate policy preferences are distinct psychological profiles. The college educated are more likely to espouse moral values and attitudes associated with the personality trait “openness to experience.” High “openness” individuals are attracted to novelty, skeptical of traditional authority, and prize personal freedom and cultural diversity. “Closed” individuals, by contrast, have an aversion to the unfamiliar and are therefore attracted to moral principles that promote certainty, order, and security. Virtually all human beings fall somewhere between these two ideal types. But the college educated as a whole are closer to the “open” end of the continuum than the general public is.

All of these distinctions between more- and less-educated voters are probabilistic, not absolute. There are Catholic theocrats with Harvard Ph.D.’s and anarchists who dropped out of high school. A nation the size of the U.S. is surely home to many millions of working-class social liberals and well-educated reactionaries. Political attitudes do not proceed automatically from any demographic characteristic, class position, or psychological trait. At the individual level, ideology is shaped by myriad historical inheritances and social experiences.

And yet, if people can come by socially liberal, “high openness” politics from any walk of life, they are much more likely to do so if that walk cuts across a college campus. (And, of course, they are even more likely to harbor this distinct psychological and ideological profile if they graduate from college and then choose to become professionally involved in Democratic politics.)

The path to the professional class veers left.

There are a few theoretical explanations for this. One holds that spending your late adolescence on a college campus tends to socialize you into cultural liberalism: Through some combination of increased exposure to people from a variety of geographic backgrounds, or the iconoclastic ethos of a liberal-arts education, or the predominantly left-of-center university faculty, or the substantive content of curricula, people tend to leave college with a more cosmopolitan and “open” worldview than they had upon entering.

Proving this theory is difficult since doing so requires controlling for selection effects. Who goes to college is not determined by random chance. The subset of young people who have the interests, aptitudes, and opportunities necessary for pursuing higher education have distinct characteristics long before they show up on campus. Some social scientists contend that such “selection effects” entirely explain the distinct political tendencies of college graduates. After all, the “high openness” personality trait is associated with higher IQs and more interest in academics. So perhaps attending college doesn’t lead people to develop culturally liberal sensibilities so much as developing culturally liberal sensibilities leads people to go to college.

Some research has tried to account for this possibility. Political scientists in the United Kingdom have managed to control for the preadult views and backgrounds of college graduates by exploiting surveys that tracked the same respondents through adolescence and into adulthood. Two recent analyses of such data have found that the college experience does seem to directly increase a person’s likelihood of becoming more socially liberal in their 20s than they were in their teens.

A separate study from the U.S. sought to control for the effects of familial background and childhood experiences by examining the disparate “sociopolitical” attitudes of sibling pairs in which one went to college while the other did not. It found that attending college was associated with greater “support for civil liberties and egalitarian gender-role beliefs.”

Other recent research, however, suggests that even these study designs may fail to control for all of the background factors that bias college attendees toward liberal views before they arrive on campus. So we have some good evidence that attending college directly makes people more culturally liberal, but that evidence is not entirely conclusive.

Yet if one posits that higher education does not produce social liberals but merely attracts them, a big theoretical problem remains: Why has the population of social liberals increased in tandem with that of college graduates?

The proportion of millennials who endorse left-wing views on issues of race, gender, immigration, and the environment is higher than the proportion of boomers who do so. And such views are more prevalent within the baby-boom generation than they were among the Silent Generation. This cannot be explained merely as a consequence of America’s burgeoning racial diversity, since similar generational patterns have been observed in European nations with lower rates of ethnic change. But the trend is consistent with another component of demographic drift: Each successive generation has had a higher proportion of college graduates than its predecessor. Between 1950 and 2019, the percentage of U.S. adults with bachelor’s degrees increased from 4 percent to 33 percent.  

Perhaps rising college attendance did not directly cause the “high-openness,” post-material, culturally progressive proportion of the population to swell. But then, what did?

One possibility is that, even if mass college attendance does not directly promote the development of “high openness” values, the mass white-collar economy does. If socially liberal values are well suited to the demands and lifeways inherent to professional employment in a globally integrated economy, then, as such employment expands, we would expect a larger share of the population to adopt socially liberal values. And there is indeed reason to think the professional vocation lends itself to social liberalism.

Entering the professional class often requires not only a four-year degree, but also, a stint in graduate school or a protracted period of overwork and undercompensation at the lowest ranks of one’s field. This gives the class’s aspirants a greater incentive to postpone procreation until later in life than the median worker. That in turn may give them a heightened incentive to favor abortion rights and liberal sexual mores.

The demands of the professional career may influence value formation in other ways. As a team of political scientists from Harvard and the University of Bonn argued in a 2020 paper, underlying the ideological divide between social liberals and conservatives may be a divergence in degrees of “moral universalism,” i.e., “the extent to which people’s altruism and trust remain constant as social distance increases.” Conservatives tend to feel stronger obligations than liberals to their own kin and neighbors and their religious, ethnic, and racial groups. Liberals, by contrast, tend to spread their altruism and trust thinner across a wider sphere of humanity; they are less compelled by the particularist obligations of inherited group loyalties and more apt to espouse a universalist ethos in which all individuals are of equal moral concern, irrespective of their group attachments.

Given that pursuing a professional career often requires leaving one’s native community and entering meritocratic institutions that are ideologically and legally committed to the principle that group identities matter less than individual aptitudes, the professional vocation may favor the development of a morally universalistic outlook — and thus more progressive views on questions of anti-discrimination and weaker identification with inherited group identities.

Further, in a globalized era, white-collar workers will often need to work with colleagues on other continents and contemplate social and economic developments in far-flung places. This may encourage both existing and aspiring professionals to develop more cosmopolitan outlooks.

Critically, parents who are themselves professionals — or who aspire for their children to secure a place in the educated, white-collar labor force — may seek to inculcate these values in their kids from a young age. For example, my own parents sent me to a magnet elementary school where students were taught Japanese starting in kindergarten. This curriculum was designed to appeal to parents concerned with their children’s capacity to thrive in the increasingly interconnected (and, in the early 1990s American imagination, increasingly Japanese-dominated) economy of tomorrow.

In this way, the expansion of the white-collar sector may increase the prevalence of “high-openness” cosmopolitan traits and values among rising generations long before they arrive on campus.

More material security, more social liberalism.

Ronald Inglehart’s theory of “cultural evolution” provides a third, complementary explanation for both the growing prevalence of social liberalism over the past half-century and for that ideology’s disproportionate popularity among the college educated.

In Inglehart’s account, people who experience material security in youth tend to develop distinctive values and preferences from those who do not: If childhood teaches you to take your basic material needs for granted, you’re more likely to develop culturally progressive values and post-material policy priorities.

Inglehart first formulated this theory in 1971 to explain the emerging cultural gap between the baby boomers and their parents. He noted that among western generations born before World War II, very large percentages had known hunger at some point in their formative years. The Silent Generation, for its part, had come of age in an era of economic depression and world wars. Inglehart argued that such pervasive material and physical insecurity was unfavorable soil for social liberalism: Under conditions of scarcity, human beings have a strong inclination to defer to established authority and tradition, to distrust out-groups, and to prize order and material security above self-expression and individual autonomy.

But westerners born into the postwar boom encountered a very different world from the Depression-wracked, war-torn one of their parents, let alone the cruel and unforgiving one encountered by common agriculturalists since time immemorial. Their world was one of rapid and widespread income growth. And these unprecedentedly prosperous conditions engendered a shift in the postwar generation’s values: When the boomers reached maturity, an exceptionally large share of the cohort evinced post-material priorities and espoused tolerance for out-groups, support for gender equality, concern for the environment, and antipathy for social hierarchies.

Since this transformation in values wasn’t rooted merely in the passage of time — but rather in the experience of abundance — it did not impact all social classes equally. Educated professionals are disproportionately likely to have had stable, middle-class childhoods. Thus, across the West, the post-material minority was disproportionately composed of college graduates in general and elite ones in particular. As Inglehart reported in 1981, “among those less than 35 years old with jobs that lead to top management and top civil-service posts, Post-Materialists outnumber Materialists decisively: their numerical preponderance here is even greater than it is among students.”  

As with most big-picture models of political development, Inglehart’s theory is reductive and vulnerable to myriad objections. But his core premise — that, all else being equal, material abundance favors social liberalism while scarcity favors the opposite — has much to recommend it. As the World Values Survey has demonstrated, a nation’s degree of social liberalism (a.k.a. “self-expression values”) tightly correlates with its per-capita income. Meanwhile, as nations become wealthier, each successive generation tends to become more socially liberal than the previous one.

Critically, the World Values Survey data does not show an ineluctable movement toward ever-greater levels of social liberalism. Rather, when nations backslide economically, their populations’ progressivism declines. In the West, recessions have tended to reduce the prevalence of post-material values and increase support for xenophobic parties. But the relationship between material security and cultural liberalism is demonstrated most starkly by the experience of ex-communist states, many of which suffered a devastating collapse in living standards following the Soviet Union’s fall. In Russia and much of Eastern Europe, popular support for culturally progressive values plummeted around 1990 and has remained depressed ever since.

Inglehart’s theory offers real insights. As an account of education polarization, however, it presents a bit of a puzzle: If material security is the key driver of social liberalism, why have culture wars bifurcated electorates along lines of education instead of income? Put differently: Despite the material security provided by a high salary, when one controls for educational attainment, having a high income remains strongly associated with voting for conservatives.

One way to resolve this tension is to stipulate that the first two theories of education polarization we examined are also right: While material security is conducive to social liberalism, the college experience and demands of professional-class vocations are perhaps even more so. Thus, high-income voters who did not go to college will tend to be less socially liberal than those who did.

Separately, earning a high income is strongly associated with holding conservative views on fiscal policy. Therefore, even if the experience of material security biases high-income voters toward left-of-center views on cultural issues, their interest in low taxes may nevertheless compel them to vote for right-wing parties.

Voters with high levels of education but low incomes, meanwhile, are very often children of the middle class who made dumb career choices like, say, going into journalism. Such voters’ class backgrounds would theoretically bias them toward a socially liberal orientation, while their meager earnings would give them little reason to value conservative fiscal policy. Perhaps for this reason, “high-education low-income voters” are among the most reliably left-wing throughout the western world.

In any case, whatever qualifications and revisions we would wish to make to Inglehart’s theory, one can’t deny its prescience. In 1971, Inglehart forecast that intergenerational value change would redraw the lines of political conflict throughout the West. In his telling, the emergence of a novel value orientation that was disproportionately popular with influential elites would naturally shift the terrain of political conflict. And it would do so in a manner that undermined materialist, class-based voting: If conventional debates over income distribution pulled at the affluent right and the working-class left, the emerging cultural disputes pulled each in the opposite direction.

This proved to be, in the words of Gabriel Almond, “one of the few examples of successful prediction in political science.”

When the culture wars moved to the center of politics, the college educated moved left.

Whether we attribute the social liberalism of college graduates to their experiences on campus, their class’s incentive structures, their relative material security, or a combination of all three, a common set of predictions about western political development follows.

First, we would expect to see the political salience of cultural conflicts start to increase in the 1960s and ’70s as educated professionals became a mass force in western politics. Second, relatedly, we would expect that the historic correlation between having a college degree and voting for the right would start gradually eroding around the same time, owing to the heightened prominence of social issues.

Finally, we would expect education polarization to be most pronounced in countries where (1) economic development is most advanced (and thus the professional sector is most expansive) and (2) left-wing and right-wing parties are most sharply divided on cultural questions.

In their paper “Changing Political Cleavages in 21 Western Democracies, 1948–2020,” Amory Gethin, Clara Martínez-Toledano, and Thomas Piketty confirm all of these expectations.

The paper analyzes nearly every manifesto (a.k.a. “platform”) put forward by left-wing and right-wing parties in the past 300 elections. As anticipated by Inglehart, the researchers found that right-wing and left-wing parties began to develop distinct positions on “sociocultural” issues in the 1970s and that these distinctions grew steadily more profound over the ensuing 50 years. Thus, the salience of cultural issues did indeed increase just as college graduates became an electorally significant demographic.

Graphic: The Quarterly Journal of Economics

As cultural conflict became more prominent, educated professionals became more left-wing. Controlling for other variables, in the mid-20th century, having a college diploma made one more likely to vote for parties of the right. By 2020, in virtually all of the western democracies, this relationship had inverted.

Some popular narratives attribute this realignment to discrete historical events, such as the Cold War’s end, China’s entry into the WTO, or the 2008 crash. But the data show no sudden reversal in education’s political significance. Instead, the authors write, the West saw “a very progressive, continuous reversal of educational divides, which unfolded decades before any of these events took place and has carried on uninterruptedly until today.” This finding is consistent with the notion that class dealignment is driven by gradual changes in western societies’ demographic and economic characteristics, such as the steady expansion of the professional class.

Class dealignment unfolded gradually across the West. Graphic: The Quarterly Journal of Economics
And the same general pattern played out in the U.S. Photo: World Inequaity Lab

The paper provides further support for the notion that education polarization is a by-product of economic development: The three democracies where college-educated voters have not moved sharply to the left in recent decades — Ireland, Portugal, and Spain — are all relative latecomers to industrialization.

Finally, and perhaps most important, the authors established a strong correlation between “sociocultural polarization” — the degree to which right-wing and left-wing parties emphasize sharply divergent cultural positions — and education polarization. In other words: Countries where parties are highly polarized on social issues tend to have electorates that are highly polarized along educational lines.

Graphic: The Quarterly Journal of Economics

It seems reasonable then to conclude (1) that there really is a cultural divide between educated professionals and the working class in the aggregate and (2) that this gap has been a key driver of class dealignment. Indeed, if we accept the reality of the diploma divide, then an increase of education-based voting over the past 50 years would seem almost inevitable: If you have two social groups with distinct cultural values and one group goes from being 4 percent of the electorate to 35 percent of it, debates about those values will probably become more politically prominent.

And of course, mass higher education wasn’t the only force increasing the salience of social conflict in the West over the past half-century. If economic development increased the popularity of “post-material” values, it also made it easier for marginalized groups to contest traditional hierarchies. As job opportunities for women expanded, they became less dependent on the patriarchal family for material security and thus were more liable to challenge it. As racial minorities secured a foothold in the middle class, they had more resources with which to fight discrimination.

And yet, if an increase in sociocultural polarization — and thus in education polarization — is a foregone conclusion, the magnitude of these shifts can’t be attributed to the existence of cultural divides alone.

Rather, transformations in the economic, civic, and media landscapes of western society since the 1970s have increased the salience and severity of the diploma divide.

When the postwar bargain collapsed, the center-left failed to secure workers a new deal.

To polarize an electorate around cultural conflicts rooted in education, you don’t just need to increase the salience of social issues. You also need to reduce the salience of material disputes rooted in class. Alas, the economic developments of the past 50 years managed to do both.

The class-based alignment that defined western politics in the mid-20th century emerged from a particular set of economic conditions. In the early stages of industrialization, various factors had heightened the class consciousness of wage laborers. Such workers frequently lived in densely settled, class-segregated neighborhoods in the immediate vicinity of large labor-intensive plants. This close proximity cultivated solidarity, as divisions between the laborer’s working and social worlds were few. And the vast scale of industrial enterprises abetted organizing drives, as trade unions could rapidly gain scale by winning over a single shop.

By encouraging their members to view politics through the lens of class and forcing political elites to reckon with workers’ demands, strong trade unions helped to keep questions of income distribution and workers’ rights at the center of political debate and the forefront of voters’ minds. In so doing, they also helped to win western workers in general — and white male ones in particular — unprecedented shares of national income.

But this bargain between business and labor had always been contingent on robust growth. In the postwar era of rising productivity, it was possible for profits and wages to increase in tandem. But in the 1970s, western economies came under stress. Rising energy costs and global competition thinned profit margins, rendering business owners more hostile to labor’s demands both within the shop and in politics. Stagflation — the simultaneous appearance of high unemployment and high inflation — gave an opening to right-wing critics of the postwar order, who argued that the welfare state and pro-labor macroeconomic policies had sapped productivity.

Meanwhile, various long-term economic trends began undermining industrial unionism. Automation inevitably reduced the labor intensity of factories in the West. The advent of the shipping container eased the logistical burdens of globalizing production, while the industrialization of low-wage developing countries increased the incentives for doing so. Separately, as western consumers grew more affluent, they began spending less of their income on durable goods and more on services like health care (one needs only so many toasters, but the human desire for greater longevity and physical well-being is nigh-insatiable). These developments reduced both the economic leverage and the political weight of industrial workers. And since western service sectors had lower rates of unionization, deindustrialization weakened organized labor.

All this presented center-left parties with a difficult challenge. In the face of deindustrialization, an increasingly anti-labor corporate sector, an increasingly conservative economic discourse, an embattled union movement, and a globalizing economy, such parties needed to formulate new models for achieving shared prosperity. And they had to do so while managing rising cultural tensions within their coalitions.

They largely failed.

Countering the postindustrial economy’s tendencies toward inequality would have required radical reforms. Absent policies promoting the unionization of the service sector, deindustrialization inevitably weakened labor. Absent drastic changes in the allocation of posttax income, automation and globalization redistributed economic gains away from “low skill” workers and toward the most productive — or well-situated — professionals, executives, and entrepreneurs.

The United States had more power than any western nation to standardize such reforms and establish a relatively egalitarian postindustrial model. Yet the Democratic Party could muster neither the political will nor the imagination to do so. Instead, under Jimmy Carter, it acquiesced to various policies that reinforced the postindustrial economy’s tendencies toward inequality, while outsourcing key questions of economic management to financial markets and the Federal Reserve. The Reagan administration took this inegalitarian and depoliticized model of economic governance to new extremes. And to highly varying degrees, its inequitable and market-fundamentalist creed influenced the policies of future U.S. administrations and other western governments.

As a result, the past five decades witnessed a great divergence in the economic fortunes of workers with and without college diplomas, while the western working class (a.k.a. the “lower middle class”) became the primary “losers” of globalization.

The center-left parties’ failures to avert a decline in the economic security and status of ordinary workers discredited them with much of their traditional base. And their failure to reinvigorate organized labor undermined the primary institutions that politicize workers into a progressive worldview. These shortcomings, combined with the market’s increasingly dominant role in economic management, reduced the political salience of left-right divides on economic policy. This in turn gave socially conservative working-class voters fewer reasons to vote for center-left parties and gave affluent social liberals fewer reasons to oppose them. In western nations where organized labor remains relatively strong (such as Norway, Sweden, and Finland), education polarization has been relatively mild, while in those countries where it is exceptionally weak (such as the United States), the phenomenon has been especially pronounced.

Finally, the divergent economic fortunes of workers and professionals might have abetted education polarization in one other way: Given that experiencing abundance encourages social liberalism — while experiencing scarcity discourages it — the past half-century of inequitable growth might have deepened cultural divisions between workers with degrees and those without.

The professionalization of civil society estranged the left from its working-class base.

While the evolution of western economies increased the class distance between college graduates and other workers, the evolution of western civil societies increased the social distance between each group.

Back in the mid-20th century, the college educated still constituted a tiny minority of western populations, while mass-membership institutions — from trade unions to fraternal organizations to political parties — still dominated civic life. In that context, an educated professional who wished to exercise political influence often needed to join a local chapter of a cross-class civic association or political party and win election to a leadership position within that organization by securing the confidence of its membership.

That changed once educated professionals became a mass constituency in their own right. As the college-educated population ballooned and concentrated itself within urban centers, it became easier for interest groups to swing elections and pressure lawmakers without securing working-class support. At the same time, the proliferation of “knowledge workers” set off an arms race between interest and advocacy groups looking to influence national legislation and election outcomes. Job opportunities for civic-minded professionals in think tanks, nonprofits, and foundations proliferated. And thanks to growing pools of philanthropic money and the advent of direct-mail fundraising, these organizations could sustain themselves without recruiting an active mass membership.

Graphic: Theda Skocpol

Thus, the professional’s path to political influence dramatically changed. Instead of working one’s way up through close-knit local groups — and bending them toward one’s political goals through persuasion — professionals could join (or donate to) nationally oriented advocacy groups already aligned with their preferences, which could then advance their policy aims by providing legislators with expert guidance and influencing public opinion through media debates.

As the political scientist Theda Skocpol demonstrates in her book Diminished Democracy, college graduates began defecting from mass-membership civic organizations in the 1970s, in an exodus that helped precipitate their broader decline.

Graphic: Theda Skocpol

Combined with the descent of organized labor, the collapse of mass participation in civic groups and political parties untethered the broad left from working-class constituencies. As foundation-funded NGOs displaced trade unions in the progressive firmament, left-wing parties became less directly accountable to their less-educated supporters. This made such parties more liable to embrace the preferences and priorities of educated professionals over those of the median working-class voter.

Meanwhile, in the absence of a thriving civic culture, voters became increasingly reliant on the mass media for their political information.

Today’s media landscape is fertile terrain for right-wing populism.

The dominant media technology of the mid-20th century — broadcast television — favored oligopoly. Given the exorbitant costs of mounting a national television network in that era, the medium was dominated by a small number of networks, each with an incentive to appeal to a broad audience. This discouraged news networks from cultivating cultural controversy while empowering them to establish a broadly shared information environment.

Cable and the internet have molded a radically different media landscape. Today, news outlets compete in a hypersaturated attentional market that encourages both audience specialization and sensationalism. In a world where consumers have abundant infotainment options, voters who read at a graduate-school level and those who read at an eighth-grade level are unlikely to favor the same content. And the same is true of voters with liberal and conservative sensibilities — especially since the collapse of a common media ecosystem leads ideologues to occupy disparate factual universes. The extraordinary nature of today’s media ecology is well illustrated by this chart from Martin Gurri’s book, The Revolt of the Public:

Graphic: Martin Gurri

This information explosion abets education polarization for straightforward reasons: Since the college educated and non-college educated have distinct tastes in media, in a highly competitive attentional market, they will patronize different outlets and accept divergent facts.

Further, in the specific economic and social context we’ve been examining, the modern media environment is fertile terrain for reactionary entrepreneurs who wish to cultivate grievance against the professional elite. After all, as we’ve seen, that elite (1) subscribes to some values that most working-class people reject, (2) commandeers a wildly disproportionate share of national income and economic status, and (3) dominates the leadership of major political parties and civic groups to an unprecedented degree.

The political efficacy of such right-wing “populist” programming has been repeatedly demonstrated. Studies have found that exposure to Fox News increases Republican vote share and that the expansion of broadband internet into rural areas leads to higher levels of partisan hostility and lower levels of ticket splitting (i.e., more ideologically consistent voting) as culturally conservative voters gain access to more ideologically oriented national news reporting, commentary, and forums.

What is to be done?

The idea that education polarization arises from deep structural tendencies in western society may inspire a sense of powerlessness. And the notion that it emerges in part from a cultural divide between professionals and working people may invite ideological discomfort, at least among well-educated liberals.

But the fact that some center-left parties have managed to retain more working-class support than others suggests that the Democrats have the capacity to broaden (or narrow) their coalition. Separately, the fact that college-educated liberals have distinct social values does not require us to forfeit them.

The commentators most keen to acknowledge the class dimensions of the culture wars typically aim to discredit the left by doing so. Right-wing polemicists often suggest that progressives’ supposedly compassionate social preferences are mere alibis for advancing the professional class’s material interests. But such arguments are almost invariably weak. Progressive social views may be consonant with professional-class interests, but they typically represent attempts to universalize widely held ideals of freedom and equality. The college educated’s cosmopolitan inclinations are also adaptive for a world that is unprecedentedly interconnected and interdependent and in which population asymmetries between the rich and developing worlds create opportunities for mutual gain through migration, if only xenophobia can be overcome. And of course, in an era of climate change, the professional class’s strong concern for the environment is more than justified.

Nevertheless, professional-class progressives must recognize that our social values are not entirely unrelated to our class position. They are not an automatic by-product of affluence and erudition, nor the exclusive property of the privileged. But humans living in rich, industrialized nations are considerably more likely to harbor these values than those in poor, agrarian ones. And Americans who had the privilege of spending their late adolescence at institutions of higher learning are more likely to embrace social liberalism than those who did not.

The practical implications of this insight are debatable. It is plausible that Democrats may be able to gain working-class vote share by moderating on some social issues. But the precise electoral payoff of any single concession to popular opinion is deeply uncertain. Voters’ conceptions of each party’s ideological positioning are often informed less by policy details than by partisan stereotypes. And the substantive costs of moderation — both for the welfare of vulnerable constituencies and the long-term health of the progressive project — can be profound. At various points in the past half-century, it might have been tactically wise for Democrats to distance themselves from the demands of organized labor. But strategically, sacrificing the health of a key partisan institution to the exigencies of a single election cycle is deeply unwise. Meanwhile, in the U.S. context, the “mainstream” right has staked out some cultural positions that are profoundly unpopular with all social classes. In 2022, it is very much in the Democratic Party’s interest to increase the political salience of abortion rights.

In any case, exactly how Democrats should balance the necessity of keeping the GOP out of power with the imperative to advocate for progressive issue positions is something on which earnest liberals can disagree.

The case for progressives to be more cognizant of the diploma divide when formulating our messaging and policy priorities, however, seems clearer.

Education polarization can be self-reinforcing. As left-wing civic life has drifted away from mass-membership institutions and toward the ideologically self-selecting circles of academia, nonprofits, and the media, the left’s sensitivity to the imperatives of majoritarian politics has dulled. In some respects, the incentives for gaining status and esteem within left-wing subcultures are diametrically opposed to the requirements of coalition building. In the realm of social media, it can be advantageous to make one’s policy ideas sound more radical and/or threatening to popular values than they actually are. Thus, proposals for drastically reforming flawed yet popular institutions are marketed as plans for their “abolition,” while some advocates for reproductive rights insist that they are not merely “pro-choice” but “pro-abortion” (as though their objective were not to maximize bodily autonomy but rather the incidence of abortion itself, a cause that would seemingly require limiting access to contraception).

Meanwhile, the rhetoric necessary for cogently theorizing social problems within academia — and that fit for effectively selling policy reforms to a mass audience — is quite different. Political-science research indicates that theoretical abstractions tend to leave most voters cold. Even an abstraction as accessible as “inequality” resonates less with ordinary people than simply saying that the rich have too much money. Yet Democratic politicians have nevertheless taken to peppering their speeches with abstract academic terms such as structural racism.

Relatedly, in the world of nonprofits, policy wonks are often encouraged to foreground the racial implications of race-neutral redistributive policies that disproportionately benefit nonwhite constituencies. Although it is important for policy design to account for any latent racial biases in universal programs, there is reason to believe that, in a democracy with a 70 percent white electorate and widespread racial resentment, it is unwise for Democratic politicians to suggest that broadly beneficial programs primarily aid minority groups.

On the level of priority setting, it seems important for college-educated liberals to be conscious of the fact that “post-material” concerns resonate more with us than with the general public. This is especially relevant for climate strategy. Poll results and election outcomes both indicate that working-class voters are far more sensitive to the threat of rising energy prices than to that of climate change. Given that reality, the most politically viable approach to reducing emissions is likely to expedite the development and deployment of clean-energy technologies rather than deterring energy consumption through higher prices. In practice, this means prioritizing the build-out of green infrastructure over the obstruction of fossil-fuel extraction.

Of course, narrowing the social distance between college-educated liberals and working people would be even better than merely finessing it. The burgeoning unionization of white-collar professions and the growing prominence of downwardly mobile college graduates in working-class labor struggles are both encouraging developments on this front. Whatever Democrats can do to facilitate labor organizing and increase access to higher education will simultaneously advance social justice and improve the party’s long-term electoral prospects.

Finally, the correlation between material security and social liberalism underscores the urgency of progressive economic reform. Shared prosperity can be restored only by increasing the social wage of ordinary workers through some combination of unionization, sectoral bargaining, wage subsidies, and social-welfare expansion. To some extent, this represents a chicken-and-egg problem: Radical economic reforms may be a necessary precondition for the emergence of a broad progressive majority, yet a broad progressive majority is itself a precondition for radical reform.

Nevertheless, in wealthy, deep-blue states such as New York and California, Democrats have the majorities necessary for establishing a progressive economic model. At the moment, artificial constraints on the housing supply, clean-energy production, and other forms of development are sapping blue states’ economic potential. If such constraints could be overcome, the resulting economic gains would simultaneously increase working people’s living standards and render state-level social-welfare programs easier to finance. Perhaps the starting point for such a political revolution is for more-affluent social liberals to recognize that their affinity for exclusionary housing policies and aversion to taxation undermines their cultural values.

Our understanding of education polarization remains provisional. And all proposals for addressing it remain open to debate. The laws of political science are more conjectural than those of physics, and even perfect insight into political reality cannot settle disputes rooted in ideology.

But effective political engagement requires unblinkered vision. The Democratic Party’s declining support among working-class voters is a serious problem. If Democrats consider only ideologically convenient explanations for that problem, our intellectual comfort may come at the price of political power.

How the Diploma Divide Is Remaking American Politics