Political scientists have typically conceptualized party polarization in terms of the question of ideological divergence. Are the platforms of the major political parties and the preferences of their followers at opposing ideological extremes? By this standard, there is clear evidence that elected officials in the U.S. leaders have become more extremist over time (McCarty, Poole, and Rosenthal 2006). Partisan voters, however, have not followed suit; the median partisan has remained centrist on most issues despite the gravitation of party elites to the ideological extremes (Fiorina, Abrams, and Pope 2008)2.

Over the past two decades, research on Ideological disagreement has given way to an alternative approach that defines mass polarization as the extent to which partisans view each other as a stigmatized out group. In the U.S., the two-party system means that voters can only affiliate with the “Democrat” group or the “Republican” group (Green, Palmquist, and Schickler 2002; Huddy, Mason, and Aaroe, 2015). Once people adopt a partisan identity, they instinctively categorize the world into a favored in group (their own party), and a disliked out group (the opposing party; see Tajfel and Turner 1979). The more central the political party to the individual’s personal identity, the stronger the degree of out-party animus (Gaertner et al. 1993).

Placed in cross-national context, partisanship is a relatively salient and powerful identify in the U.S. for several reasons. First, it develops at a young age, and rarely changes over the life cycle, notwithstanding significant shifts in personal circumstances (Sears 1975). Second, election campaigns are more frequent and last for many months (or even years). The increased duration of campaigns means that Americans are constantly reminded of their partisan identity. As I describe below, in recent decades the sense of partisan identity has elicited especially harsh evaluations of political opponents (for a review of the now extensive literature on affective polarization, see Iyengar et al. 2019; Druckman, Green, and Iyengar 2023).

The period over which mass polarization has intensified (1980-today) coincides with major changes in American society including changes in the media environment that facilitate selective exposure to news, increased social homophily, and partisan sorting or greater differentiation between Democrats and Republicans. In addition to independently inducing hostility toward opponents, each of these factors reinforces the others, further contributing to the rise of affective polarization.

First, the revolution in information technology has empowered consumers to encounter media sources on their own terms. The availability of twenty-four hour cable news channels provided partisans with their first real opportunity to obtain news from like-minded sources (Fox News first for Republicans, and MSNBC later for Democrats). In a break with the dominant paradigm of non-partisan journalism, a growing number of outlets, motivated in part by the commercial success of Fox News, began to offer biased reporting. Many of these partisan outlets depict the opposing party in harsh terms and focus disproportionately on out-party scandals (real or imagined). The creation of vast online social networks permitted extensive recirculation of news reports, even to those not particularly motivated to seek out news. Several scholars have thus singled out the technologically enhanced media environment and partisans’ ability to encounter “friendly” information providers as an especially influential agent of polarization (see, for instance, Sunstein 2017).

While there are good reasons to believe that the new media environment has contributed to the growth in partisan animus, by facilitating access to partisan news and commentary, it is possible that enhanced consumer choice also sets in motion processes that weaken polarization. As media platforms have multiplied, consumers gain access not only to more news providers, but also to entertainment providers. The availability of entertainment programming on demand enables consumers to withdraw from the political arena entirely (Prior 2007). Thus, the net impact of the increased empowerment of consumers is unclear.

In fact, despite the myriad changes in the media environment, the evidence to date demonstrating that news consumption exacerbates polarization is less than unequivocal. Early studies of online browsing behavior confirmed the tendency of partisans to self-select into distinct audiences (see, for instance, Iyengar and Hahn 2009) and that selective exposure contributed to polarization (e.g. Stroud 2010). More recent interventions that manipulate participants’ exposure to social media content, however, find only weak to negligible effects of social media usage on out-party animus (see, for instance, Nyhan et al. 2023; Alcott et al. 2020; Levy 2021).

Large-scale observational studies also record mixed findings. In the first large-scale analysis of Americans’ web browsing behavior (conducted in 2008), Gentzkow and Shapiro (2011) found that online audiences were only slightly more segregated than audiences for network or cable news. More recent studies, however -- also based on large-scale tracking of online browsing behavior -- suggest that the segregation of news audiences is increasing. A 2013 study showed that although most people relied on ideologically diverse online sources such as web aggregators (Flaxman, Goel, and Rao 2016; also see Peterson, Goel, and Iyengar 2019), audience segregation tended to increase among individuals who used search engines to locate news stories and among social media users who encountered links in their news feed. Importantly, when researchers limit the focus to online sites dedicated to news coverage, they find the expected pattern of audience segregation based on partisanship (Tyler, Grimmer, and Iyengar 2022).

A second societal-level change associated with the growth of out-party animus in the U.S. is the phenomenon of sorting. Over time, the percentage of “sorted” partisans, i.e., partisans who identify with the party most closely reflecting their ideology, has steadily increased (Levendusky 2009). When most Democrats [Republicans] are also liberals [conservatives], they are less likely to encounter conflicting political ideas and identities (Roccas and Brewer 2002) and are more likely to see non-identifiers as socially distant. As partisan and ideological identities have come into alignment, other salient social identities, including race and religion, have also converged with partisanship. Democrats are increasingly the party of women, non-whites, professionals, and residents of urban areas, while Republican voters are disproportionately older white men, evangelical Christians and residents of rural areas. This decline of crosscutting identities is at the root of affective polarization according to Mason (2015, 2018). She has shown that those with consistent partisan and ideological identities became more hostile towards the out party without necessarily changing their ideological positions, and those that have aligned religious, racial, and partisan identities react more emotionally to information that threatens their partisan identities or issue stances. In essence, sorting has made it much easier for partisans to make generalized inferences about the opposing side, even if those inferences are inaccurate.

A third potential cause of strengthened polarization is social homophily. I have already referred to strengthened processes of socialization by which families come to agree on their partisan loyalties. Family agreement creates an inter-personal echo chamber that facilitates polarization. When family members identify with the same party, they also express more extreme positions on the issues and harbor hostile views toward their opponents; partisan agreement within the family strengthens polarization (see Iyengar, Konitzer, and Tedin 2018).

Given the importance of family socialization to the development of partisan attitudes, the rate at which any given society undergoes polarization will be conditional on the extent to which partisans grow up in homogeneous environments. Recent simulations by Klofstad, McDermott, and Hatemi (2013) suggest that spousal agreement rapidly induces ideological polarization, reaching a stable equilibrium by the 11th generation, but with most of the increased polarization occurring as early as the fifth generation (Klofstad, McDermott, and Hatemi 2013, pp. 530–531). We would similarly expect generations to move increasingly apart on their feelings toward the opposing party to the extent family members share these sentiments.

The literature on affective polarization has grown exponentially over the past decade. The elevated level of out-party animus has set off alarm bells within the academy and at civic organizations. The pressing concern appears to be that partisan animus motivates anti-social behaviors such as denying the results of legitimate elections, the use (or at least approval) of violence as a means of advancing partisan interests, and the strategic spreading of misinformation. The fear is that polarization and such related changes in behavioral norms constitute democratic “backsliding” which now presents an imminent threat to democratic institutions and norms.

Against this backdrop of a “crisis for democracy” (Voelkel et al. 2023), it is not surprising that major foundations with pro-democracy agendas have encouraged and funded (in the millions of dollars) the development of “treatments” for polarization. In the first large-scale assessment of potential treatments, Voelkel et al. (2023) tested the effects of twenty-five interventions designed to reduce out-party animus and anti-democratic predispositions. These interventions all represented brief psychological nudges designed to alter the beliefs and attitudes of partisans such as providing partisans with corrective information concerning the composition of the opposing party, the reasons underlying opposing partisans’ preferences, or highlighting the anti-democratic consequences of polarization. In fact, many of these interventions did “work” as intended -- they significantly reduced out-party animus and willingness to express anti-democratic attitudes -- although the long-term persistence of these effects is unknown.

To this writer, the efforts to develop treatments appear shortsighted and premature. For one thing, there is no compelling evidence that out-party animus contributes to the anti-democratic outcomes of concern (see, for instance, Broockman, Kalla, and Westwood 2023; for a review of the evidence, see Druckman, Green, and Iyengar 2023)). For another, researchers have yet to arrive at a clear understanding of the principal causes of affective polarization. This gives rise to scholars tailoring their treatments to the symptoms rather than the underlying causes of polarization. In short, there is an insufficient theoretical or evidentiary basis for designing appropriate treatments. There are also concerns about the real-world applicability of some of the proposed interventions. Many represent “counterfactuals” in the sense that the stimuli and experiences encountered in the treatment are rarely found in everyday life. For instance, engaging in “deliberation” with fellow citizens, accompanied by presentations from experts on both sides of a policy debate, is not an everyday opportunity for most Americans. I would also call attention to the significant costs that would be required to scale up some of the more “promising” treatments; in even the most optimistic scenarios, the costs are prohibitive.

All told, the intellectual development of the field of affective polarization remains at the descriptive rather than explanatory stages. In the absence of rigorous evidence of causal effects, the prescription of treatments is unlikely to yield long-term benefits and may even yield unanticipated counterproductive effects as occurred with five of the tested depolarizing interventions in the case of support for undemocratic practices (Voelkel et al., 2023). Organizations seeking to promote the development of harmonious inter-party relations will be better served by investing in basic research into the causal mechanisms at play.