Is polarization actually harmful? The answer depends on which kind, how much, and what it leads people to accept. This level weighs the costs and the possible benefits, and looks at the most serious danger researchers have identified: the willingness to tolerate undemocratic behavior by one's own side.
The case for some polarization
As the 1950 APSA report argued, clear party differences can be good. When parties stand for distinct things, voters have meaningful choices and can hold parties accountable. Research suggests that sharper party differences in the United States were accompanied by higher interest in politics and, in recent elections, higher turnout. Polarized eras have also been eras of moral clarity: the fight over slavery in the 1850s divided the country bitterly, and on that question compromise was not a virtue. Defenders of strong partisanship argue that calls for "unity" can be a way of asking one side to drop legitimate demands.
Gridlock
The most commonly cited cost is gridlock. The political scientist Sarah Binder has measured how often Congress fails to act on the issues on its own agenda and found that this share has risen as the parties have moved apart. In systems with separated powers, where different parties can control different branches, polarization makes divided government especially hard to run. Problems that need long-term solutions, such as public debt or infrastructure, may be left to grow. Gridlock also pushes action toward executive orders and courts, which can make policy swing back and forth with each change of power.
Trust and everyday life
Affective polarization spills beyond politics. Experiments show partisans discriminating in hiring, dating and economic exchanges, as Level 2 described. Trust in government tends to rise and fall depending on whether one's own party holds power. Hostility reaches family life too: one study of phone location data by the economists Keith Chen and Ryne Rohla found that Thanksgiving dinners in 2016 were shorter, on average, when the guests came from areas that voted for opposing parties. Some evidence suggests partisans even judge facts, such as whether the economy is improving, differently depending on who is in charge.
The democratic danger
The most serious concern is what polarization does to support for democratic rules. Steven Levitsky and Daniel Ziblatt, in How Democracies Die (2018), argued that written constitutions are not enough. Democracies also rely on two unwritten norms: mutual toleration, accepting rivals as legitimate, and forbearance, not using every legal power to the full to crush opponents. Polarization erodes both: if the other side is an enemy, any tactic to keep them from power can seem justified.
Matthew Graham and Milan Svolik tested this in experiments published in 2020. They presented American voters with pairs of candidates and varied whether a candidate supported undemocratic actions, such as ignoring court rulings or limiting the opposition. Only about 3.5 percent of voters punished a candidate from their own party for such positions by voting for the other side. The more polarized the voter, the less likely they were to defect. Voters valued democracy in the abstract, but when it conflicted with party loyalty, most chose party. Svolik and others have found similar patterns in other countries.
Violence
Some surveys have reported that a significant minority of partisans approve of political violence. These numbers deserve care. A 2022 study by Sean Westwood, Justin Grimmer and colleagues found that many such answers came from inattentive respondents or vague questions, and that when questions were specific and responses checked, support for actual violent acts was low. That is reassuring, though even small percentages of a large population can be many people, and political violence often comes from a few individuals encouraged by a climate of hostility.
Weighing it up
A balanced view holds that ideological differences are a normal part of democracy, while affective hostility that leads people to excuse rule-breaking is a real danger. The line that matters is not between people who disagree mildly and people who disagree strongly. It is between people who accept that the other side may win and govern, and people who do not.
Worked example: the election rule
Suppose a state legislature controlled by Party X proposes a law letting the legislature overturn the state's election results if it judges them "irregular." Party X's leaders say Party Y has cheated before. Party Y voters are outraged. Many Party X voters are uneasy but tell pollsters that Party Y is so dangerous that they will support the law anyway.
Graham and Svolik's findings predict this pattern: uneasy partisans tend to stay loyal. Levitsky and Ziblatt would point to forbearance: even if the law is constitutional, using it would break the norm that losers accept election results. Now reverse the labels. Would Party Y's voters be equally uneasy if their own legislature proposed the law? The test of a democratic commitment is whether you would accept the same rule with the other side in power. Treating opponents as equal citizens, with equal rights to win, is where democratic principle and basic human equality meet.