Commentary on “Frontiers: Spilling the Beans on Political Consumerism: Do Social Media Boycotts and Buycotts Translate to Real Sales Impact?”

A safeguard of American democracy has long been that a diversity of interests within the population prevents the emergence of political monoliths who, as James Madison stated in 1787’s The Federalist No. 10, “are united and actuated by some common impulse of passion” (Smith et al. 2015, p. 59). We see such monoliths today: compared with the recent past, Republicans, for instance, are far more likely to be consistently conservative, and Democrats are far more likely to be consistently liberal (Levendusky 2009). The Republican Party is much more demographically homogeneous than it was 20 years ago (Mason and Wronski 2018). America was relatively well functioning, at least in the post-Civil Rights era, because people who voted for different parties still believed some of the same things and shared common experiences.

Shared experiences “provide a form of social glue. They help make it possible for diverse people to believe that they live in the same culture. Indeed they help constitute shared culture simply by creating common memories and experiences and a sense of common tasks” (Sunstein 2008, p. 105). However, as Liaukonytė et al. (2022) show, shared experiences are fading. The fact that Democrats and Republicans bought different brands of beans for a period seems relatively innocuous. Still, the phenomena point to a far more problematic sorting of lifestyle preferences and partisanship. That is, people’s partisanship is increasingly correlated with other nonpolitical preferences.

Perhaps the most pernicious effect of increased affective polarization, or interparty animosity, is that it has increasingly caused politics to bleed into ostensibly nonpolitical domains. Democrats and Republicans drive different cars, watch different television (Hetherington and Weiler 2018), and, as Liaukonytė et al. show, even eat different brands of food. Whereas scholars have long noted that American policy preferences are sorting along policy lines (Levendusky 2009), Liaukonytė et al. show that Americans’ lifestyle preferences may also be sorting nowadays. Americans seem to be inhabiting different worlds, not just based on which policies they support and which news they read but also on which products they buy. If this pattern uncovered by Liaukonytė et al. grows, the result will be that Americans will fail to see any commonalities between themselves and anyone who voted for a different party.

To be sure, the results of Liaukonytė et al. demonstrate that there are, so far, clear limits to such lifestyle polarization. In particular, the effect is circumscribed and short term. The effect lasted about three weeks and did not affect loyal consumers. This comports with work that shows that politics is, for most people, a sideshow (Krupnikov and Ryan 2022). Nonetheless, it shows that (1) polarized cultural preferences are not merely partisan cheerleading, and (2) politics has a direct impact on cultural preferences and (3) that these outcomes are driven by elite rhetoric.

The affective polarization literature is often criticized for using self-reported survey measures (Bullock and Lenz 2019). For instance, Democrats and Republicans say they would be unhappy if their son or daughter married someone who supports the other party (Iyengar et al. 2012). Most Republicans say they believe the 2020 presidential election was stolen (Badger 2020). But are these self-reports merely expressive? As discussed in one New York Times article, “Research has shown that the answers that partisans (on the left as well as on the right) give to political questions often reflect not what they know as fact, but what they wish were true. Or what they think they should say” (Badger 2020). That is, much of the apparent rising partisanship is an illusion, driven by partisans saying what they think a partisan is supposed to say or what would make their side look better.

The primary rebuttal of this claim has been to show that partisanship affects actual behavior and has real economic consequences—at least in the laboratory. For instance, McConnell et al. (2018) find that Amazon Turk workers are willing to lower their reservation wage if they think the task comes from a copartisan and are willing to take a smaller payout if it means a political rival won’t also benefit. Liaukonytė et al. extend the external validity of these claims considerably.

In addition, Liaukonytė et al. (2022) demonstrate that politics directly affects nonpolitical behavior. Although many studies have shown that Americans are geographically sorted (Bishop 2009), less willing to date people who do not share their political identity (Huber and Malhotra 2017), and are more likely to drive Priuses (if they’re Democrats) or pickup trucks (if they’re Republicans) (Hetherington and Weiler, 2018), some have questioned whether these correlations are merely incidental to politics. That is, people who live in rural areas are more likely to need a pickup truck than those who live in a city, and people who live in rural areas prefer the Republican Party (Martin and Webster 2020). This project shows that the effect of partisanship is not merely spurious. Instead, politics is driving lifestyle polarization.

This work highlights the role of political elites in driving polarization. The Goya boycott/buycott discussed in Liaukonytė et al. occurred in response to social media posts by Alexandra Ocasio Cortez and Ivanka and Donald Trump. This jibes with the conclusion of political scientists that political elites are not merely reflecting and responding to a polarized electorate but fomenting division. Interventions should aim at changing elite rhetoric, not just citizen psychology.

The careful analysis performed by Liaukonytė et al. (2022) demonstrates that although the effects are limited and not long lasting, partisanship divides us far beyond the ballot box—even to the kitchen pantry—and that political elites shoulder much of the responsibility for this discord.