What Are the Solutions to Political Polarization?

Social psychology reveals what creates conflict among groups and how they can come together.

By Lee de-Wit, Cameron Brick, Sander van der Linden



What drives political polarization?

Is it simply disagreement over the great issues of the day? Not necessarily. Recent research by the More in Common Foundation found that more than three-fourths of Americans support both stricter gun laws and a pathway to citizenship for undocumented immigrants brought here as children. Roughly the same number of Americans agree “that our differences are not so great that we cannot come together.”

Are they right?

The More in Common results could be interpreted to suggest that we can build bipartisan support for specific policies by focusing more on their boring nuts and bolts. Unfortunately, however, voters don’t evaluate policies in isolation. Research has highlighted that people actively use partisan cues when evaluating different policies.

For example, a study by Carlee Beth Hawkins and Brian Nosek shows that labeling policies as “Democrat” or “Republican” can influence policy support, depending on the implicit bias of participants toward each party. A 2017 study by David Tannenbaum and colleagues finds that support for policy “nudges”—such as changing 401k retirement accounts to opt-out rather than opt-in—was heavily influenced by whether they were framed as supporting the goals of the Democratic or Republican party. This was true of regular U.S. citizens and for senior government leaders. Similarly, a 2018 study by Leaf Van Boven and colleagues finds that the majority of Republicans agree that climate change is happening—but their support for policy solutions declined when presented by Democrats.

In other words, people like policies proposed by members of their own in-group—and they don’t like ideas generated by out-groups. This dynamic is not new. Since the 1950s, social psychologists have tried to understand what pits groups against each other—and today, they’re applying these insights to figure out what is happening in the United States. This research doesn’t provide definitive answers, but it does suggest some potential solutions, from changes to the voting system to the development of common goals that might enable groups to work together.

How morality becomes partisan

The More in Common report illustrates that some of the most divisive topics often involve deep moral beliefs. For example, different political groups are very polarized on beliefs about responsibility, such as “people’s outcomes in life are determined largely by forces outside of their control,” or “people are largely responsible for their own outcomes in life.” Similarly, liberals and conservatives are very divided on questions of whether parenting should focus on cultivating a child’s curiosity versus good manners, or independence versus respect for elders.

In a new study published this year, Annemarie S. Walter and David P. Redlawsk directly pitted people’s moral concerns with their partisan identity. They presented 2,000 participants with examples of different moral violations by different actors. Based on previous research, Walter and Redlawsk had thought that the nature of the moral violation might be the most significant factor in people’s evaluations, as there are reasons to think that liberals and conservatives are concerned with some moral violations more than others. What they found, however, is that it wasn’t the nature of the moral violation that was most important. Instead, it was the political allegiance of the violator. Democrats in the study were prone to giving Democrats a pass; the same was even more true of Republicans.

This partisan influence on policy preferences and moral judgements is a cause for both hope and concern. On one hand, it reiterates a point made by Daniel Yudkin in a New York Times op-ed about the More in Common report: that the U.S. may actually be less politically polarized based on certain moral or policy issues—at least when there aren’t clear partisan associations. On the other hand, it highlights that as soon as a moral or political issue becomes associated with a particular party, it can become polarizing.

This is why it increasingly feels like U.S. politics has entered into a vicious cycle, whereby the moral and emotional language used to galvanize one side is directly antagonizing the other. The us-and-them nature of the debate has led to such a breakdown of trust that even hearing a policy proposed by the other side can be enough to trigger opposition to that policy. New policies (whatever their merit) can therefore quickly become symbols of conflict for the two sides to rally around.

What are the solutions?

This suggests that while there might be various political seeds that have helped drive the recent spike in polarization, it has gotten to a point where polarization is being exacerbated by some of the psychological processes that shape how we interpret identity and groups. This is a significant point to understand because it highlights that if we are to address polarization, we need to think not just about political solutions, but also solutions that are grounded in our understanding of social psychology.

1. Intergroup contact. The “contact hypothesis” suggests that getting to know each other can reduce prejudice between groups. However, social contact can be done well and done badly. As we discussed in a previous article, following political opponents on Twitter can make people more extreme in their political views. It turns out that many conditions have to be met for contact to reduce prejudice, including having contact be sustained, with more than one member of the group, including a genuine exchange of ideas, and between individuals of similar social rank. These conditions have been very difficult to meet in designing social policies.

One promising civic model for enabling more meaningful contact between groups in conflict involves “Citizens Assemblies,” where representative citizens are brought together to deliberate over challenging social or political issues. These assemblies can be thought of as a kind of jury duty for political deliberation, and they offer a platform for different groups to discuss issues in a way that can highlight where common ground exists and how it can be acted upon.

For example, Ireland has run several Citizens Assembles since 2016 that made policy recommendations that have been credited with advances in Ireland’s approach to climate change. Indeed, participants in a recent Citizens Assembly on Brexit, run by Alan Renwick and colleagues at University College London, came to a compromise that could resolve the current impasse surrounding the U.K.’s decision to leave the European Union.

2. Perspective taking. Perhaps one of the most important aspects of contact is that it might enable one to see things from another’s perspective. The promise of perspective taking was recently illustrated in an experiment to attempt to change support for issues faced by transgender minorities. In this intervention, a brief exchange exploring a range of issues from the perspective of a trans individual was sufficient to shift people’s attitudes on this controversial topic. Indeed, the attitude change seemed to persist even six months later, which is unusual for brief psychological interventions.

In his book The Better Angels of Our Nature, Pinker argues that the printing press may have had an important role in increasing levels of empathy following the Enlightenment by making it easier to read stories framed in the perspective of others. Indeed, Pinker speculates that some of the literature written from the perspective of black slaves may have been instrumental to the abolition of slavery.

Considering the revolution in communication technology in our lifetimes, social media may have done more to promote taking sides than seeing the world through the eyes of another. Social media companies, and the governments that regulate them, clearly need to reflect on the extent to which these platforms encourage “side taking” instead of “perspective taking.”

3. Superordinate goals. One of the clearest solutions from the psychological literature is that identity-based conflicts require common goals or a “superordinate” sense of identity to bring people back together. In other words, we need a large sense of ourselves that is able to bridge smaller differences. This need to create a superordinate identity has clearly been intuitive to rulers for centuries, who would use various traditions and ceremonies to help build alliances between different countries and cultures.

Of course, superordinate goals also come with a potential risk. Whenever we form an in-group, we also create out-groups. As Richard Dawkins recently tweeted:

National pride has evil consequences. Prefer pride in humanity. German pride gave us Hitler, American pride gave us Trump, British pride gave us Brexit. If you must have pride, be proud that Homo sapiens could produce a Darwin, Shakespeare, Mandela, Einstein, Beethoven.

Unfortunately, drawing a parallel between Trump and Hitler is perhaps itself an illustration of the polarized nature of modern discourse. Dawkins does have a point, however: The use of a superordinate identity such as American or European has potential risks. So, should we just think of ourselves as humans—or is the idea of “humanity” too abstract? Former British prime minister Theresa May famously criticized such a universalist perspective, stating: “If you believe you are a citizen of the world, you are a citizen of nowhere.”

Is that true? Research suggests that a universalist perspective might well have underappreciated benefits. Sam McFarland and colleagues recently reviewed this topic and found that those who identify highly as citizens of the world are indeed more empathic. Of course, those who are more empathic might simply identify more as international citizens. This idea warrants further testing, particularly as McFarland and colleagues identify several factors that might serve to further develop this sense of international citizenship.

4. Proportional voting. While searching for psychological solutions to polarization, it’s important not to ignore the context in which political decisions are made, and to think about the way in which different political systems will engage with, and exacerbate, aspects of our psychology.

The U.S. is one of the few countries to be dominated by just two political parties. This fact is almost certainly a reflection of the “winner take all”/“first past the post” voting system. Many countries employ a proportional (or mixed) system, which means that if a party gets 5 percent of the popular vote, they will receive 5 percent of the seats in a given representative body. In the U.S., this party would almost certainly get no representation—which could worsen the us-and-them dynamic of the U.S. political system.

For now, however, there isn’t systematic evidence comparing the extent of identity-based politics with the political system used in the U.S. Unfortunately, that kind of large-scale, cross-country research is often the most difficult to obtain, but could be exactly what we need to understand how different voting systems might influence polarization.

That said, there is evidence that more proportional systems have higher levels of voter turnout (at least for supporters of smaller parties). In turn, that increase in turnout is correlated with citizens being more likely to report feeling that their vote makes a difference. This doesn’t necessarily stop politics becoming less polarized, but it might make it harder for the extremes to come to dominate.

5. Voting for policies, not for parties. Another potential solution to identity-based policy preferences is to hold direct referendums on specific issues. Among large territories, California and Switzerland both regularly use referendums to address complex policy topics. Referendums are used less frequently in other countries to try to resolve controversial topics, as was the case with gay marriage in Australia or voting reform in New Zealand. When designed well, referendums might cut across existing partisan divides, and if a clear majority is reached, they can signal a new social norm that can help a country move forward.

For those who are familiar with the fallout from the recent Brexit referendum in the U.K., however, this suggestion would probably seem a little laughable. Contemptuous, even. Indeed, there is evidence that the referendum in the U.K. has itself spilled over into a new form of emotional polarization, as recent data from YouGov highlights that (especially younger) “Remain” supporters would not want to see a close relative marry a “Leaver” (a member of the opposing political camp). There’s another problem as well: While Australians did indeed vote to legalize same-sex marriage, it could have gone the other way; allowing the majority to vote on the civil and human rights of a minority is very risky.

Like many complex political systems, however, referendums can be designed well and designed badly. In countries with more established systems of direct democracy, the U.K. referendum wouldn’t have even been legal. For example, in Switzerland, referendums have to be about precisely defined changes to the law, not vaguely defined outcomes. In New Zealand’s referendum on the voting system, an independent educational body was created to inform both sides of the debate without taking a position (as the British government controversially did during the Brexit referendum).

The psychological impact of more direct voting systems is worthy of further enquiry. When poorly implemented, referendums risk causing new fault lines along which polarization can manifest. When well implemented, referendums might cut across existing lines of polarization and help establish a new social norm that can move a country forward.

From reframing issues to tap into a superordinate sense of identity, to promoting forms of contact that encourage perspective taking, social psychology does offer some useful ideas for thinking about how to tackle polarization. Furthermore, social psychology provides insights into the potential implications of different kinds of voting systems and the way in which they might exacerbate or diminish identity-based politics. As we have been careful to try and illustrate, however, experiments in social psychology do not yield off-the-shelf solutions that would be effective in all political contexts. Nevertheless, the farther that modern politics sinks into a self-fulfilling cycle of identity-based polarization, the more we’ll need new insights from social science.