What Is Political Polarization: Causes, Consequences, and Whether It Can Be Reversed

Defining Polarization

Political polarization refers to the divergence of political attitudes toward ideological extremes. But the term covers at least two distinct phenomena that researchers and commentators often conflate. Issue polarization describes the widening gap between where citizens of different parties actually stand on policy questions — whether they hold more extreme or more different views on abortion, taxes, immigration, and other issues. Affective polarization, by contrast, describes how much citizens dislike, distrust, and even loathe members of the opposing party — independent of whether their actual policy positions have moved apart. Both are real, but they are different problems with different causes and implications.

The United States provides the most extensively studied case of polarization. Survey research by Pew Research Center and academic political scientists shows that the overlap in ideological position between Democrats and Republicans has nearly disappeared over the past four decades. In 1994, the ideological distributions of the two parties substantially overlapped — there were conservative Democrats and liberal Republicans. By the 2020s, the distributions had sorted almost completely, with virtually no Republicans to the left of the median Democrat and virtually no Democrats to the right of the median Republican. Whether this represents elite polarization (political elites and activists have sorted, pulling their parties' brands with them while ordinary citizens have followed) or genuine mass polarization remains debated, but the partisan divide in Americans' expressed views on a wide range of issues has indisputably grown.

Affective polarization, some researchers argue, has grown even more dramatically than issue polarization. The percentage of Americans who say they view the opposing party "very unfavorably" has more than doubled since the 1990s. Large minorities in both parties describe the opposing party as a threat to the nation's well-being. Cross-partisan friendships and marriages have declined. Studies have shown that partisans exaggerate how extreme, how racially different, and how economically different the other party's supporters are — they harbor inaccurate stereotypes of the outgroup that are worse than reality. This "false polarization" amplifies conflict beyond what actual policy differences would warrant.

Economic Inequality and Material Grievance

Economic factors are central to many accounts of polarization. The decades since 1980 have seen substantial increases in income and wealth inequality in most advanced democracies, as gains from economic growth have accrued disproportionately to those at the top of the income distribution while wages for workers without college degrees stagnated in real terms. The communities most affected by deindustrialization — factory closures, declining union membership, hollowed-out local economies — have often become epicenters of populist anger directed at governing elites perceived as indifferent or hostile to their interests.

The geographic dimension of this economic divide reinforces polarization. Knowledge-economy cities with concentrated professional, technology, and financial employment have grown more prosperous and more educated while rural and small-town communities dependent on manufacturing and agriculture have experienced persistent economic distress. As economic geography and political identity have come to overlap — Democrats increasingly concentrated in thriving metropolitan areas, Republicans increasingly dominant in economically struggling rural areas — political conflict takes on the character of zero-sum competition between communities with different economic interests and different experiences of the social changes of the past half-century.

Thomas Piketty and other economists have connected rising inequality to rising political instability, arguing that democratic institutions become strained when economic outcomes diverge sufficiently from democratic equality norms. Relative deprivation — feeling worse off compared to others even if one's absolute situation is stable — generates political anger that is difficult to channel through normal political institutions. When people believe that the rules of the economy are rigged, that hard work no longer guarantees economic security, and that political elites are captured by economic elites, the legitimacy of normal democratic politics erodes and the appeal of outsider politicians promising radical disruption increases.

Media, Social Platforms, and Information Ecosystems

The transformation of media and information environments is frequently cited as a driver of polarization, though the evidence is more nuanced than popular accounts suggest. The decline of shared broadcast media — three major television networks reaching most of the population — and the rise of cable news, talk radio, and internet media has allowed citizens to self-select information environments that reinforce their existing beliefs and expose them to the most extreme voices in their ideological ecosystem. Fox News, MSNBC, and their online equivalents don't just report partisan perspectives; they socialize their audiences into partisan identities and feed them a constant stream of outrage about the opposing party.

Social media platforms amplify these dynamics through algorithmic content recommendation. Platforms designed to maximize engagement — through likes, shares, and comments — find that emotionally arousing content, particularly content evoking anger and fear about outgroup behavior, generates high engagement. This creates an incentive structure that rewards inflammatory content and punishes nuanced, accurate but boring information. The "filter bubble" hypothesis — that algorithms show users only content confirming their existing views — has received mixed empirical support; some studies find that exposure to outgroup content on social media actually increases polarization rather than reducing it, because algorithmically selected outgroup content tends toward the extreme. The problem may be less echo chambers than "outrage chambers."

However, the relationship between social media and polarization is complicated by timing. The increase in affective polarization in the United States began in the late 1990s — before social media existed. Polarization has increased among older Americans who use social media less intensively than younger cohorts. These patterns suggest that social media may amplify and accelerate polarization but is not its root cause. Some comparative research shows that countries with similar social media penetration have much lower levels of political polarization, suggesting that political institutional structures and party systems mediate the relationship between media environment and political division.

Partisan Sorting and Identity

Lilliana Mason's research has advanced an influential account of polarization as a consequence of identity sorting rather than changing policy positions. Her argument is that political party membership has become increasingly correlated with racial, religious, cultural, and geographic identities — all of which were previously more evenly distributed across the parties. When being a Republican or Democrat is just one identity among several, political conflict is mitigated by cross-cutting identities: you might share a party with someone but differ in religion or region. When all those identities are sorted onto the same side, partisan conflict activates multiple overlapping identity divisions simultaneously, generating much more intense hostility than any single issue difference would produce.

In the United States, the parties have sorted dramatically along racial and educational lines since the 1960s. The Democratic Party, once the party of white Southern Protestants and working-class Catholics, has become the party of racial minorities, college-educated whites, and secular Americans. The Republican Party has become overwhelmingly the party of white non-college-educated Americans, evangelical Christians, and rural communities. This sorting means that when a Democrat and Republican disagree politically, they are typically disagreeing as members of communities with different racial compositions, different religious practices, different educational attainment, and different geographic contexts — not just different policy preferences. The conflict is existential rather than merely political.

Social identity theory, from social psychology, predicts that individuals strongly favor their ingroup and readily view the outgroup negatively, even when group membership is based on arbitrary criteria. Applied to partisan identity, this predicts that party membership will generate ingroup favoritism and outgroup hostility independent of actual policy differences. Experimental studies confirm this: partisans show systematic preference for ingroup members in hiring, dating, and housing decisions. This discrimination based on partisan identity is, some studies suggest, now larger in magnitude than discrimination based on race in the American context — a striking finding given the salience of racial inequality in American life and history.

Polarization and Democratic Erosion

Severe polarization threatens democracy through several mechanisms. Most directly, it makes governing impossible: when parties view each other not as legitimate opponents but as existential enemies, normal legislative compromise becomes treasonous to the base. The United States Congress has seen repeated government shutdowns, debt ceiling crises, and failure to pass routine legislation because partisan confrontation has become the dominant political logic. More insidiously, polarization erodes the informal norms that democratic institutions depend on. Constitutional scholars Steven Levitsky and Lucan Way argue that democracy requires not just formal rules but informal norms of mutual toleration (accepting the legitimacy of opponents) and institutional forbearance (not using all available legal powers aggressively). Polarization corrodes both.

The treatment of Supreme Court appointments illustrates norm erosion clearly. For most of American history, high Court nominations proceeded with bipartisan courtesy and were confirmed by large Senate majorities. The Bork hearings in 1987 began the politicization of confirmation; the Merrick Garland non-hearing in 2016 and the Amy Coney Barrett confirmation in 2020 (both executed by the Republican Senate majority using opposite reasoning about election-year appointments) represent a fundamental breakdown of institutional forbearance. Neither action was technically illegal; both were norm violations that reduced the Supreme Court's legitimacy as an institution independent of partisan politics. Other countries, notably Hungary and Poland, illustrate how a governing party with sufficient parliamentary majority can legally dismantle democratic checks and balances when norms of restraint have been abandoned.

Polarization also threatens electoral integrity. When elections are understood not as fair competitions between legitimate alternatives but as battles between good and evil, losing parties are incentivized to contest results, suppress opposing votes, and refuse to accept defeats. The January 6, 2021 attack on the US Capitol by supporters of a president who refused to accept electoral defeat is the most dramatic recent example, but attempts to restrict voting access, gerrymander districts to entrench partisan advantage, and deploy legal challenges to undermine electoral outcomes have become routine tools of polarized partisan competition. These behaviors, individually sometimes legally defensible, collectively undermine the foundational norm that elections determine who governs.

Can Polarization Be Reversed?

The evidence on depolarization is sobering but not entirely without hope. Historical comparisons show that American polarization has been severe before — the Civil War era was more polarized by almost any measure, as was the period of the 1890s–1920s featuring intense battles over immigration, labor rights, and monetary policy. The relative consensus of the mid-twentieth century (roughly 1945–1965) is historically unusual, not the norm. Polarization has also declined in some recent periods: New Zealand, after a prolonged period of polarization in the 1980s–1990s associated with its sharp economic liberalization, became substantially more moderate and consensual after adopting MMP electoral reform. Denmark and the Netherlands have shown that well-designed proportional institutions can maintain cross-party cooperation even under significant social pressures.

Institutional reforms — proportional electoral systems, ranked-choice voting, independent redistricting, campaign finance reform — can reduce polarization by changing the incentive structures politicians face. Ranked-choice voting, now used in Maine and Alaska for federal elections and in dozens of cities, rewards candidates who build broad coalitions rather than appealing only to their party's base; early evidence from Alaska suggests it produces more moderate winners in competitive races. Reforms to reduce the influence of primary elections (which select more extreme candidates) and to increase ballot access for third parties could expand the political space available for moderate candidates.

Media literacy education, cross-partisan dialogue initiatives, and "contact hypothesis" programs that bring together people from different political backgrounds have shown modest effects in reducing affective polarization in experimental settings. The contact hypothesis — that direct personal contact between members of opposing groups reduces hostility — is well-established in the prejudice literature and has been applied to partisan contact with some success. But scaling these interventions to the size of the problem is a major challenge, and they address the symptoms (hostility) without necessarily addressing the underlying causes (economic inequality, identity sorting, media incentive structures). The most honest assessment is that reducing polarization will require addressing its structural causes, not just its symptoms — a generational project rather than a quick fix.