PART ONE
Merriam-Webster named “polarization” the word of the year in 2024.
Hopefully, by identifying a problem, the country can begin to address it. However, the term itself is deeply problematic. Polarization has too many different definitions to be meaningful without highly specific modifiers. Additionally, its most obvious interpretation of ideological differences among the general public is less pervasive than typically believed; it suggests greater political divisions than exist in reality, and none of its current main definitions directly covers the highly important finding that Americans have an overblown sense of threat from those in the other political party.
This article highlights the problems with the term “polarization”. A companion article to be published on February 6, 2025, will suggest some other language to use instead.
While it may be surprising, polarization arguably has eight different definitions. Types of polarization include ideological (distance between policy positions), affective (difference between ingroup and outgroup-like, often measured as warmth on a “ feeling thermometer ”), moral (including lack of shared moral frames between groups), and false (overestimating the extent to which the types of polarization listed above and other divisive factors are true). Each of these definitions of polarization can have two variants. Each can refer to the “mass” (the broader American public), or the “elite” (opinion leaders such as elected officials).
Many organizations try to avoid this morass of definitions by instead using the term “ toxic polarization,” essentially a catch-all for its undesirable elements. However, if most Americans hear the term, many likely do not immediately understand what these toxic aspects are, and some may hear it as liberal jargon, reminiscent of terms such as “ toxic masculinity.” If most Americans assume polarization refers to ideological differences, then many who hear toxic polarization likely assume it means “very ideologically polarized.”
This reasonable interpretation, unfortunately, does not align with academic research about the American public. Researchers recognize that party sorting has occurred among all Americans, so there are fewer conservative Democrats or liberal Republicans than in the past. However, that is different from polarization, which implies a bimodal distribution with many on the extremes and few in the middle. Both the Pew Research Center and More in Common have developed political typologies to describe the American public, and both of them only found about 15% of Americans on the ideological categories at the edges. The organization that James leads, More Like US, jointly runs Similarity Hub, which shows more than 400 examples of policy overlaps between Democrats and Republicans or supermajorities of Americans, covering every hot-button issue.
Using a term like polarization so frequently that Merriam-Webster named it the word of the year likely exacerbates false polarization. When people hear about a perceived problem like polarization again and again, they are more likely to think that this problem is real and large, even when it is actually overblown. In some original false polarization research by the organization More in Common, which calls the issue the “Perception Gap,” the more people followed the news (and likely heard about issues such as polarization), the more they incorrectly overestimated the ideological differences among those in the other political party. In a paper by leading expert Dr. Rachel Kleinfeld of the Carnegie Endowment for International Peace, she notes, “American voters are less ideologically polarized than they think they are.”
This said, some other individual definitions of polarization are fairly accurate, such as findings that members of Congress have become substantially more ideologically polarized, and the overall public has become more affectively polarized, as seen in Figure 4 here.
Maybe five years ago, a smart way to be a practitioner in this field was to say that the main goal was not to reduce mass ideological polarization and make everyone a political centrist, but rather to reduce mass affective polarization. Yet, research now shows a weak or nonexistent link between mass affective polarization and attitudes/actions more directly related to the decline of America’s system of government. These include a willingness to break democratic norms or support political violence. In 2023, leading researchers published a study with the clear title, “Interventions reducing affective polarization do not necessarily improve anti-democratic attitudes.”
The results of the Strengthening Democracy Challenge (SDC) led by Stanford, which tested 25 short interventions on about 1,000 Americans each, provide greater insight into how mass affective polarization is still somewhat of a valuable concept, but also where all eight definitions of polarization fall flat.
The SDC, which used yet another term for mass affective polarization—partisan animosity—found that the concept was fairly closely related to all sorts of variables that are problematic but often not directly catastrophic, including social distrust, social distance, and opposition to bipartisan cooperation.
Yet, mass affective polarization/partisan animosity was much less correlated with the catastrophic consequence of support for undemocratic practices, and it was not at all correlated with support for political violence.
Only three interventions achieved all the main goals of the researchers of significantly reducing partisan animosity, anti-democratic attitudes, and support for political violence. A main reason why they were so successful comes down to a factor that none of the many definitions of polarization directly covers: overblown perceptions of threat from those in the other political party.
Of the interventions that achieved all these goals, one of them focused on the threatening finding that Americans are much less supportive of breaking democratic norms than the other side believes, and another showed that members of a political party think those in the other party threateningly dehumanize them twice as much than in reality. The third intervention—a 2020 Utah gubernatorial campaign ad made jointly by both candidates—has more varied elements, but it is notable that the candidates spoke highly of accepting election results and “a peaceful transition of power.” Study participants saw this ad after January 6, and some may have seen it as a possible antidote to threats of political violence and upheaval.
Hopefully, Merriam-Webster’s choice to name “polarization” the word of the year will actually mark the peak of using this term. In the companion article to be published on February 6, 2025, we explore two better terms to use instead.
James Coan is the co-founder and executive director of More Like US. Coan can be contacted at James@morelikeus.org
Sara Weinstein is a current intern at More Like US.
