Article Summary
Introduction
For people to trust in each other, strangers, and society at large, they need to believe that most people share basic values and norms. This sort of social trust is a critical requirement for democracy. It encourages cooperation and civic engagement; without it, societies struggle to work together and respond collectively to crises. Yet many in the United States now believe that the public is fundamentally divided by politics. Past work indicates that these perceptions of polarization promote partisan animosity, extreme political beliefs, and distrust in government. This paper asks if perceptions of polarization also undermine broader trust in society.
Analytical Approach
This paper explores the connection between perceptions of polarization and social trust in two studies. First, the author uses existing data from a panel survey that contacted the same representative sample of U.S. adults once a year between 2016 and 2020. In each of the five survey waves, participants were asked three questions capturing generalized social trust: whether they think most people can be trusted, try to be helpful, and try to be fair. Each wave also asked participants to indicate how much they think Americans disagree on political issues, have extreme views, and are divided and polarized.
The author then conducts a fixed-effects regression. Fixed-effects regressions examine over-time changes within individuals; that is, they compare each individual only to themselves in other survey waves. Thus, the fixed-effects regression controls for stable characteristics of individuals, and the author additionally controls for other factors that may vary over time (e.g., individuals’ levels of actual polarization). The results of this analysis thus indicate whether over-time changes in perceived polarization tend to lead to changes in social trust. However, fixed-effects regressions do not present perfectly clear evidence of causality. The author notes that any associations found in the fixed-effects analysis could be driven by reverse causality: people who experience reduced social trust may accordingly view society as more divided and polarized.
To circumvent this issue, the second study the author uses is an experiment designed to provide strong causal evidence of the link between perceived polarization and social trust. In this second study, the author directly manipulated participants’ perceptions of polarization by randomly assigning them to read a news story featuring either high polarization, low polarization, or a control story about COVID-19. After reading the story, participants answered the same three questions about generalized social trust.
To extend the analysis and provide more comprehensive information about social trust, the author also used three additional measures of social trust in this second study. The first additional measure focused on trust in the American citizenry, asking questions about how much participants trust Americans to be honest and responsible citizens. The second focused on participants’ willingness to trust strangers in everyday situations (e.g., giving a stranger a ride). Finally, the third measure focused on behavioral implications of social trust. For this measure, each participant was given a cash bonus of $2, then told they could either keep it or donate some portion to a well-known charity of their choice. Half of the participants were also randomly told that other people across the study were participating in the study, and that if at least 80% donated some of the $2 (regardless of the amount), their donations would be matched by the researcher. Conceptually, this donation matching treatment should emphasize the importance of participants’ beliefs about others; those low in social trust should be less willing to donate given their skepticism about others.
| Table of Experimental Conditions | Perceived Polarization Manipulation | |||
|---|---|---|---|---|
| High Polarization Story | Low Polarization Story | Control Story (COVID) | ||
| Donation Manipulation | Donation Matching | 1 | 2 | 3 |
| Baseline (Normal Donation) | 4 | 5 | 6 | |
Main Findings
In study 1, using 5-year nationally representative panel data, the author finds that perceptions of polarization negatively affect social trust. In study 2, the author shows that reducing polarization in particular significantly increased social trust. And it did so across different measures: people who read the story featuring low polarization felt more generalized social trust, more trust in the American citizenry, and more willingness to trust strangers in everyday situations. And, turning toward the behavioral measure, participants who read the story featuring low polarization donated more of their cash bonus to charity—though only when they were told their donations would be matched.
Thus, people who believe the American public is intensely divided over politics are much less likely to trust society at large. Moreover, the author provides some evidence as to why this effect occurred. After participants read their randomized news stories, they were asked to estimate how many Americans share their values. Consistent with the author’s theory, participants believed their values were more widely shared when they read the low polarization story.
Implications
Social trust is a relatively poorly understood variable in political science, and has rarely been connected to polarization in past work. Thus, this study not only advances our understanding of polarization and its implications for individual political behavior, but it also considers a largely new potential consequence of polarization: reduced social trust due to beliefs about Americans not sharing values.
Although this is not the first study to do so (see e.g., Ahler, 2014; Levendusky & Malhotra, 2016), this study also emphasizes the political importance of perceptions of polarization. People do not have to be polarized themselves to act differently. Simply being exposed to others’ polarization via news or politics can be enough to meaningfully shape political behavior. Moreover, among the existing studies on perceived polarization, this paper is the first to examine explicitly non-political implications of perceived polarization. Past work in this area focuses mostly on how exposure to polarization influences ideological extremity and affective polarization; this author’s consideration of non-political consequences—in this case, social trust—represents a potentially meaningful step forward and a new way of looking at (perceived) polarization.
Questions left unanswered
Although the author provides evidence of mechanism by showing that the (non)polarized stories affected perceptions of Americans’ shared values, they note that they cannot make a definite conclusion about the mechanism by which the stories influenced social trust. Perceptions of shared norms (and perceptions of polarization more generally) are two likely candidates for the mechanism, but future work should consider other possibilities as well.
Most of the effects documented in study 2 were driven by the story featuring low polarization. Participants who read that story perceived less polarization than those who read the high polarization story or the control, but there were no differences in perceived polarization between the high polarization story and the control. That is, the author successfully reduced perceived polarization, but they could not successfully increase it. This extends to most of their findings for social trust as well—reducing perceived polarization increased social trust, but they did not find evidence that increasing perceived polarization did much at all. The author attributes the difficulty of increasing perceived polarization to a ceiling effect, wherein people already believe politics is so polarized that increasing this perception is much easier than decreasing it. This is certainly plausible (and consistent with past work by Levendusky & Malhotra, 2016), but the explanation is only speculative and, given the regularity of this finding in the literature, scholars interested in perceived polarization should examine this issue more seriously.
Finally, although the author demonstrates that perceived polarization has both attitudinal and behavioral consequences for social trust, it remains unclear whether these effects have downstream consequences for democratic behavior, endorsement of democratic norms, and so on.
Methods and Analysis
Was the study and its analyses pre-registered?: No
Did the study rely on proxy variables to measure polarization?: No
Were standard p-value thresholds used (p<.05 or 95% Confidence Intervals that don’t overlap zero)?: Yes
- Largest p-value presented as significant: 0.05
Were correlational results interpreted with causal language?: No
Limitations / Weaknesses
In the supplementary materials, the author reports the results of a series of pilot tests they used to develop the experimental treatments for study 2. They considered both a political control (focused on COVID, the topic actually used in the study) and an apolitical control (about sports). In their pilot tests, they concluded that the apolitical control did not help create variation in perceived polarization. However, the presumed mechanisms in this study are about perceived polarization and perceptions of shared norms. It seems un-intuitive that the political control condition focusing on COVID, a highly salient and divisive issue, would not activate either of these mechanisms. That might help explain why the high polarization story was so ineffective in increasing perceived polarization; conversely, it suggests that the control condition may have inflated perceived polarization and, thus, overstated the effects of the low polarization story. The author might have discussed these possibilities in more detail.
Open Data & Analyses
Does the article make the replication data publicly available?: Yes
Does the article make the replication analysis scripts publicly available?: Yes
Article Citation
Lee, A. H.-Y. (2022). Social trust in polarized times: How perceptions of political polarization affect Americans’ trust in each other. Political Behavior. https://doi.org/10.1007/s11109-022-09787-1
Bibtex
1
2
3
4
5
6
7
8
9
10
@article{lee_social_2022,
title = {Social trust in polarized times: {How} perceptions of political polarization affect {Americans}’ trust in each other},
url = {https://doi.org/10.1007/s11109-022-09787-1},
doi = {10.1007/s11109-022-09787-1},
journal = {Political Behavior},
author = {Lee, Amber Hye-Yon},
month = mar,
year = {2022},
}