The vicious cycle of political polarization

Imagine two strangers meeting in a restaurant. One of them orders a vegetarian dish, mentioning their commitment to adopting a more plant-based diet to reduce their carbon footprint. Before the person can say another word, the other may already have them pegged as a Democrat.

This is possible because, in today’s America, attitudes and political identity are increasingly interwoven. At the same time, dislike and distrust along partisan lines has been on the rise. Learning a person’s position on certain issues may therefore serve as a reliable signal that indicates if someone is “with us” or “against us.” 

But how good or bad are people at reading such identity clues?  

To explore this question, we recruited a sample of U.S. citizens for a political guessing game. They first gave their own opinions on a range of political topics and then guessed another fictitious person’s political identity from a single attitude — for example, whether a participant would consider a fictitious person who tends to agree that the federal government should make it more difficult to buy a gun is more aligned with Democrats or Republicans. 

In contrast to our participants, who were provided with only a single attitude clue, we had access to comprehensive information revealing the specific attitudes held by Democrats and Republicans, respectively. We obtained this information from a statistical network model that outlined the different beliefs held by Democrats and Republicans, depicting them as a map of attitudes in which each of the two political identities is structurally encoded. This information allowed us to assess how accurately our participants could estimate the political identity of others by comparing their responses with the true values from the network map.

Strikingly, we observed a nearly perfect overlap between our model and participants’ guesses. Hence, based on our findings, we can say that American citizens clearly do seem to have a quite accurate mental representation of the attitudes separating Democrats and Republicans, and they can use this information to categorize others based on the most minimal of information.

Why should this give us reason to worry?

This feat of identity-reading is not merely an outcome of individual psychology but requires a clearly differentiated social system in which attitudes, values and even lifestyle behavior are embedded into distinguishable political narratives that are embraced by different groups. The continuous production and reproduction of these structures reflect both a form and an outcome of polarization. Thinking of it differently: In a social system with more nuanced and cross-cutting beliefs, political identities would be much harder to read, and people would need more information to do so. 

Opinion pluralism, which is vital to effective democracy, has been on the decline. Worryingly, there may be a vicious cycle turning polarization into a self-amplifying process.

In a highly structured social system in which several political and nonpolitical attitudes are neatly mapped to rivaling partisan identities, the expression of cross-cutting beliefs may quickly lead to political misrecognition. People may feel wrongly accused of sympathizing with one group or another and feel trapped in the crossfire of heated debates. Discouraging expressions of pluralism, however, may exaggerate and hence advance people’s impressions of uniformity. 

How can we fix this? 

Perhaps we should consider becoming less hasty in categorizing others as Democrats or Republicans, and recognize instead that, even within a polarized political landscape, cross-cutting political viewpoints still exist. Attitudes expressed in discussions, content liked on social media and behaviors displayed, such as ordering a vegetarian dish, may offer insights into a person’s political identity. However, using this information to create symbolic divisions based on partisan fault lines undermines the complexity that is needed for a healthy social and political discourse. If people could express and display cross-cutting positions more openly, the realm of public discourse would get progressively less identity-laden, thus helping to pave the way for depolarization.

In a similar vein, we must renew our understanding of pluralism and compromise as fundamental pillars of democratic dialogue. Even without considering moral and emotional aspects of polarization, it may be challenging, and sometimes impossible, to find common ground between people with diametrically opposed viewpoints. However, our data suggest that even on hot-button issues like environmental protection and abortion, bridges between partisans do exist. Finding and highlighting these bridges will be fundamental for rebuilding social cohesion. 

Adrian Lueders (@Adrian_Lueders) is a doctor of social psychology and leads a research group at the University of Hohenheim in Stuttgart, Germany. His current research revolves around topics regarding the interactive production of identity, ideology and culture in relation to digital advancements in communication.

Dino Carpentras (@JustaNormalDino) is a doctor of photonics and currently working as postdoctoral researcher at the ETH Zurich in Switzerland. He applies mathematical and computational modelling to research opinion dynamics and complex social systems. 

Mike Quayle (@Quayle) is an associate professor at the University of Limerick (Ireland) and honorary senior lecturer at the University of KwaZulu-Natal (South Africa). He explores the politics of identity, and how the local and immediate concerns of identity enactment can impact on broader socio-political processes. 

Philip Warncke (@pwarncke@mastodon.social) is getting his Ph.D. in political science from the University of North Carolina at Chapel Hill. His research deals with the measurement of ideology and how belief systems affect politics in the U.S. and in Europe.