Political Self-Confidence and Affective Polarization

Abstract

Even among those who share the same partisan commitments, some people say they despise the opposing party while others report far less animosity. Why are some people more likely to express hostility toward the opposing political party? We explore how individual-level differences in feelings of self-confidence fuel out-party animosities. Drawing on responses to a module of the 2020 Cooperative Election Study, we show that higher levels of internal political efficacy are associated with greater affective polarization. Those who feel self-assured about their political abilities are more likely to admit severing social ties with those who disagree with them and are more tolerant of discrimination against partisan opponents. In a survey experiment, we confirm that those with greater internal efficacy are also more likely to accept discrimination against a member of the opposing party. Affective polarization is greatest among those who feel the most confident of their ability to influence politics.

Those who feel confident about their ability to engage with politics are described as high in internal political efficacy. These feelings of efficacy have long been thought to be a valuable political resource. Those who feel greater internal political efficacy are more likely to discuss politics, more likely to turn out to vote, and more likely to participate in civic life (Campbell, Gurin, and Miller 1954; Campbell et al. 1960; Almond and Verba 1963; Abramson and Aldrich 1982; Verba, Schlozman, and Brady 1995; Morrell 2003). These feelings of political self-confidence are also associated with higher levels of trust in government (Almond and Verba 1963; Balch 1974). Because feelings of self-confidence within politics are positively associated with a number of behaviors tied to good citizenship, we have traditionally seen internal political efficacy as something desirable within politics (Easton and Dennis 1967). In the words of Almond and Verba (1963, p. 207), “The self-confident citizen appears to be the democratic citizen.”

However, we propose that people’s feelings of self-assurance within politics can have a darker side. We show that those with higher levels of political confidence are more likely to report hostilities toward those who hold opposing political views. When people feel self-assured about their abilities to engage in politics, they can feel empowered to fight for their side in a partisan battle. Using survey data, we show that higher levels of internal political efficacy are associated with greater affective polarization and greater negativity toward the opposing party. When people feel confident about their ability to engage with politics, they are more critical of the opposing party, more likely to avoid talking with those who do not share their views, and more tolerant of partisan discrimination. Using evidence from a survey experiment, we also show that those with greater internal efficacy are more accepting of partisan discrimination when it is to the detriment of an out-party member.

Our results help explain why some people are more likely to express negative sentiments about the opposing party than others. Even if affective polarization is on the rise, there is individual-level heterogeneity in the degree to which people express negativity toward the opposing party (Groenendyk, Sances, and Zhirkov 2020). Even among people with similar partisan allegiance, some are more likely to admit hostilities toward partisan rivals. We show that those who feel confident that they can make a difference in politics are also the most likely to express negative opinions of those who do not share their partisan views. While being confident in politics has been typically thought to be valuable to empower people to political action, we find that this self-confidence can also contribute to adversarial politics and partisan negativity. In contrast, those who are skeptical about their own ability to be effective in politics are more tolerant of those who do not share their partisan loyalties.

The effects we find are distinct to people’s particular feelings of confidence and mastery within the domain of politics. People’s general feelings of self-confidence, as captured by a measure of self-esteem, are not tied to affective polarization in the same way. Others have argued that those with feelings of self-confidence rooted outside of politics are more tolerant and less likely to discriminate against others (Sniderman 1975; Sullivan et al. 1982). We confirm the same in the case of partisan tolerance—those with high self-esteem are less likely to treat opposing partisans in hostile ways than those with low self-esteem. While people with high levels of self-confidence within politics are more likely to accept partisan discrimination, it is those with low levels of personal self-esteem who are most likely to tolerate treating people differently based on their political beliefs.

The Origins of Affective Polarization

As party polarization rises within Congress, so have partisan divides within the electorate. Americans are increasingly likely to report policy preferences that are consistent with their partisan and ideological orientations (Levendusky 2009; Fiorina, Abrams, and Pope 2011). While Democrats and Republicans disagree on public policy goals (Rogowski and Sutherland 2016; Webster and Abramowitz 2017), these partisan conflicts extend into social life as well. We have seen a rise in affective polarization, where partisans are increasingly likely to hold negative views of members of the opposing party (Mason 2015; Abramowitz and Webster 2016; Lelkes 2016). An increasing share of Americans say they would be unhappy if their child would marry someone who identifies with the opposing political party (Iyengar, Sood, and Lelkes 2012). Many are willing to discriminate against those who do not share their partisan views (Iyengar and Westwood 2015; Lelkes and Westwood 2017).

When thinking about why Americans are increasingly divided by partisanship, many of the explanations focus on macro-level changes. Influences like elite polarization, campaign financing, and partisan media fuel the rise of partisan animosities in the electorate (Iyengar et al. 2019). Elite polarization affects how people connect policy preference to their partisanship, and these differences in opinion about policy goals can contribute to people’s partisan animosities (Bougher 2017; Webster and Abramowitz 2017; Dias and Lelkes 2022). Affective polarization also has social and psychological origins, tied to people’s tendency to favor in-groups over out-groups (Mason 2018; Robison and Moskowitz 2019). When people spend less time with those who do not share their views, they feel more psychological freedom to dislike and discriminate against out-party members (Mason and Wronski 2018).

While much of our focus on the roots of affective polarization has focused on why partisan animosities are on the rise, it is also clear that there is important individual-level heterogeneity in how people think about the political parties. Some people are more likely to engage in partisan thinking than others (Taber and Lodge 2006; Arceneaux and Vander Wielen 2017). Even as levels of affective polarization are increasing in the public over time, this has occurred at different rates for different individuals. While some partisans are increasingly hostile toward the opposing party over time, others have not polarized in their party affect to the same degree (Groenendyk, Sances, and Zhirkov 2020).

Patterns of partisan thinking vary by levels of political sophistication, which moderates the degree to which people hold consistent partisan and ideological beliefs as well as the degree to which those beliefs are connected to their views of the political parties (Levendusky 2009; Lelkes 2018). People who lean on their moral convictions in political decision-making are prone to affective polarization (Garrett and Bankert 2020). Personality traits also play a role, where people differ in their tendencies toward criticizing the opposing side as a function of personality traits like need for closure and authoritarianism (Luttig 2017, 2018). Partisans with more authoritarian tendencies report higher levels of affect for their own party, regardless of whether they identify as Democrats or Republicans (Luttig 2018). We extend this research on individual-level heterogeneity in patterns of partisan thinking by considering how feelings of self-confidence affect levels of affective polarization.

General and Domain-Specific Self-Confidence

We propose that feelings of self-confidence affect people’s willingness to express negative sentiments about the out-party. We focus on people’s self-confidence within the domain of politics as captured by feelings of internal political efficacy. Internal efficacy reflects subjective self-evaluations of being able to understand current events and participate effectively in politics (Balch 1974; Craig, Niemi, and Silver 1990; Wolak 2018). It is distinct from related concepts like external political efficacy in its focus on self-assessment. While both internal and external political efficacy capture people’s perceptions of their political influence, external efficacy centers on perceptions of politicians and whether people believe elected officials are willing to listen and respond to their concerns. Internal efficacy reflects self-evaluations and people’s beliefs that they are personally able to understand politics and effectively act on their interests (Lane 1959; Acock, Clarke, and Stewart 1985).

We expect that people with higher levels of internal political efficacy will be more likely to be affectively polarized than those with low levels of internal efficacy. People who find politics hard to understand can be overwhelmed by the demands of engaging in political life. Rather than face these challenges, they are likely to instead withdraw or shy away from political pursuits. In contrast, people who feel politically capable are more likely to take on the partisan battles that define contemporary politics. Those with high levels of internal political efficacy are confident in what they know and believe about politics. They tend to follow politics and believe that they understand what is happening politically (Morrell 2003). People high in internal efficacy see themselves as highly capable individuals who can understand and communicate about the complexities of politics (Lane 1959).

How does this confidence affect people’s reactions to the conflicts of politics? Consider the case when someone is confronted by beliefs that are contrary to their own. Some people will just ignore this challenge and move on with their life. Others might engage with this confrontation and try to understand why their beliefs are being directly challenged. But for someone with a high level of internal political efficacy, this kind of challenge is likely to be viewed with skepticism. Feelings of efficacy empower people toward resistance in the face of opposition (Bandura 1997). Confident in their political commitments, those high in internal efficacy will be unlikely to question their own views. Instead, they will direct their skepticism outward, questioning the credibility of their opponents and doubting the quality of their evidence (Bandura 1997).

As a result, those who are confident in their knowledge and understanding of politics are likely to not only be stubborn in their commitments, but also adversarial toward those who challenge them. Those with high levels of internal efficacy see themselves as knowledgeable experts. When people do not agree with them, it feels like a challenge to not only their beliefs, but their expertise as well. We believe that as levels of internal efficacy increase, so will levels of out-party hostility and affective polarization.

We expect that out-party animosity reflects people’s specific self-assurances about navigating politics, rather than their feelings of self-confidence generally. As past work has demonstrated, domain-specific self-confidence has distinctive origins and distinctive effects compared to general self-confidence (Lane 1959; Campbell et al. 1960; Almond and Verba 1963; Bandura 1997). While feelings of political competence have received more attention in political science than general self-esteem, both are politically consequential (Rosenberg 1962; Sniderman 1975; Carmines 1978). We expect that people’s political self-confidence drives partisan animosities, but that people’s general self-confidence does not.

Those with high self-esteem have self-confidence not attached to politics. This can be a resource to help better tolerate political challenges. Because people with high levels of self-esteem believe in their own capabilities and self-worth, self-esteem serves as a psychological resource that can lessen the threat that people feel when exposed to ideas that challenge their beliefs or group (Rosenberg 1965). While someone with high levels of domain-specific efficacy can find it hard to brush off a challenge or criticism, people with high self-esteem can do so more easily, as their self-worth is not as responsive to outside forces (Bandura 1997).

People with higher levels of self-esteem are not as threatened by ideas or beliefs they do not personally hold—and as a result, self-esteem can lead to higher levels of tolerance of out-groups and their ideas (Sniderman 1975; Sullivan et al. 1982). In contrast, people with lower levels of self-esteem are more easily set back by challenges. When things go wrong in their personal lives, people with low levels of self-esteem are more likely to assign blame to government (Baird and Wolak 2021). Those with low self-esteem are more likely to feel attacked and threatened by outside groups (Sniderman, Hagendoorn, and Prior 2004). They can respond by discriminating against the group or ideas that threaten them (Shamir and Sullivan 1983). People with low levels of self-esteem may rely on out-group derogation to boost feelings about themselves and their own group (de Zavala et al. 2020). As a result, we expect that self-esteem will not be associated with greater affective polarization. To the extent to which self-esteem predicts out-party animosities, we expect that those with high self-esteem will be less likely to express negativity toward partisan opponents than those with low self-esteem.

Data and Measures

We consider the relationship between self-confidence and affective polarization using survey responses to a module of the 2020 Cooperative Election Study. A sample of 1,000 Americans participated in the survey online. We focus on two sets of outcomes: affective polarization tied to feelings toward the political parties and willingness to discriminate against members of the other party. Our first set of models considers the relationship between feelings of self-confidence and levels of affective polarization. Using feeling thermometer items, we create measures of affect toward one’s own political party and affect toward the opposing party.¹ We also consider a measure of affective polarization based on the absolute difference between people’s rating of the Democratic Party and the Republican Party.²

We then explore how self-confidence relates to people’s negativity toward the opposing party and their tolerance for treating people differently based on their partisan views. As a measure of people’s social aversion to members of the opposing party, we ask respondents if they try to avoid spending time with those who do not share their views. We also consider a complementary measure of whether respondents admit that they have blocked someone on social media for their political opinions. We also consider people’s tolerance of partisan discrimination, asking if people agree that it is okay to treat some people differently because of the political party they favor.

As our measure of political self-confidence, we rely on a scale of internal political efficacy. Higher scores indicate those who believe that they have a good understanding of political issues and believe themselves to be well qualified to participate in politics. As our measure of general self-confidence, we rely on an implicit measure of self-esteem that asks respondents how much they like their first and last name (Gebauer et al. 2008).³ Those who express less enthusiastic ratings of their name are thought to have lower levels of self-esteem. Consistent with past work that argues that general self-confidence is distinctive from domain-specific political confidence (Lane 1959; Almond and Verba 1963; Wolak 2020), we find that implicit self-esteem is uncorrelated with internal political efficacy (r = 0.03).

We include an indicator of external political efficacy as a control, based on two items that ask people whether they believe that politicians care about what people like them think and whether they believe they have a say in what government does. We also control for education, news consumption, and political knowledge. By controlling for these markers of political sophistication, we hope to better isolate the specific influence of feelings of political self-confidence apart from whether people hold the particular skills that facilitate partisan engagement (Bennett 1997). We include controls for two traditional explanations for affective polarization tied to party and policy. Because people’s policy commitments have been argued to contribute to greater affective polarization (Bougher 2017; Dias and Lelkes 2022), we include a control for the ideological consistency of people’s policy preferences. Higher scores indicate the degree to which people report uniformly liberal or conservative policy views on a slate of twenty-seven public policy preferences. We also control for the strength of people’s party affiliation (measured as a folded version of the seven-point party identification scale), given the effects of partisan identity on affective polarization (Mason 2018). We also include controls for gender, race, ethnicity, and age.

Regression results are shown in table 1.⁴ We find that those with higher levels of internal political efficacy report higher levels of affective polarization than those with low levels of internal efficacy. Moving from the lowest level of internal political efficacy to the highest level of internal efficacy is associated with a nine-point increase in affective polarization, about a third of a standard deviation difference. We next consider the relationship between internal efficacy and ratings of the two parties among the subset of partisan identifiers. We find that higher levels of internal efficacy are not associated with greater favorability to one’s own party. Instead, greater internal efficacy is associated only with greater negativity toward the opposing party. Moving from the lowest level of internal efficacy to the highest is associated with a seven-point drop in out-party affect. People who feel confident about their place in politics are more likely to dislike their partisan adversaries, but are no more likely to favor their own party. To the extent to which higher internal efficacy is tied to greater affective polarization, it seems to reflect the effects of political self-confidence on aversion toward the opposing party.

Table 1.

Internal political efficacy and affective polarization.

	Affective polarization	Own-party thermometer	Opposing party thermometer
Internal efficacy	9.397	0.107	−7.278
	(4.629)	(3.712)	(3.298)
	0.043	0.977	0.028
Self-esteem	−0.458	4.892	0.084
	(4.644)	(2.950)	(2.512)
	0.921	0.098	0.973
External efficacy	2.492	10.522	3.247
	(4.503)	(3.517)	(2.845)
	0.580	0.003	0.254
Ideological consistency	20.550	10.999	−18.991
	(3.204)	(2.928)	(2.581)
	0.000	0.000	0.000
Strength of partisanship	11.915	8.614	−1.649
	(0.929)	(0.955)	(0.713)
	0.000	0.000	0.021
News consumption	0.677	1.968	0.624
	(0.809)	(0.668)	(0.530)
	0.403	0.003	0.239
Political knowledge	1.007	−0.932	−1.011
	(0.787)	(0.610)	(0.538)
	0.201	0.127	0.061
Education	−15.127	−13.441	4.966
	(3.431)	(2.554)	(2.110)
	0.000	0.000	0.019
Female	3.616	3.041	−1.139
	(2.032)	(1.488)	(1.419)
	0.076	0.041	0.422
Black	5.048	8.532	0.107
	(3.944)	(2.352)	(2.893)
	0.201	0.000	0.970
Latino	0.281	−2.603	1.922
	(3.261)	(3.113)	(2.852)
	0.931	0.403	0.500
Asian American	8.114	2.453	−0.737
	(11.787)	(4.589)	(3.259)
	0.491	0.593	0.821
Age	0.369	0.237	−0.120
	(0.070)	(0.051)	(0.043)
	0.000	0.000	0.006
Constant	−1.493	33.086	36.589
	(5.609)	(4.384)	(3.967)
	0.790	0.000	0.000
R2	0.44	0.30	0.25
N	834	726	711

Note: 2020 CES. Regression estimates. Standard errors in parentheses. p-values in italics.

We find that these effects are distinctive to people’s confidence that they can be influential in politics. Self-esteem does not predict levels of affective polarization.⁵ Consistent with past work, we find that strong partisans and those with more ideologically consistent views are more likely to dislike the out-party and report affectively polarized party ratings. But the effects of internal efficacy persist even after accounting for these policy and party forces, as well as controls for levels of political sophistication. People’s particular feelings of self-confidence within politics are uniquely associated with the expression of out-party animosities.

We next consider how feelings of political self-confidence relate to people’s social animosities and support for partisan discrimination. Given that some have argued that political consequences of affective polarization may be limited (Broockman, Kalla, and Westwood 2023), it is useful to consider whether internal efficacy is related to other forms of partisan animosity beyond affective polarization. In table 2, we first consider the relationship between self-confidence and people’s desire to avoid social ties with members of the opposing party. We find that people with higher levels of internal efficacy are more likely to say they try to avoid having political conversations with people who do not share their views. Moving from the lowest level of internal efficacy to the highest is associated with a 0.18 increase in avoidance of discussion with those who disagree (on a 0 to 1 scale), or just over a half of a standard deviation increase. We also find that greater internal efficacy is associated with a greater likelihood of blocking another person on social media due to their political views. The likelihood of having unfriended someone on social media for political reasons is 8 percent at the lowest level of internal efficacy, but 39 percent for those at the highest level of internal efficacy. People who feel more confident about their ability to engage with politics are more willing to limit their social interactions over political disagreements. People who are less assured about their political abilities are more willing to maintain social ties with those who disagree with them. This suggests that high levels of internal efficacy may serve to reinforce people’s tendencies to interact mostly within politically homogeneous political discussion networks (Mutz 2006).

Table 2.

Internal political efficacy and out-party animosity.

	Avoids talking with people with opposing views	Has blocked someone on social media due to politics	Acceptability of partisan discrimination
Internal efficacy	0.178	2.129	0.140
	(0.057)	(0.510)	(0.054)
	0.002	0.000	0.010
Self-esteem	−0.160	−1.241	−0.092
	(0.042)	(0.378)	(0.043)
	0.000	0.001	0.031
External efficacy	−0.054	0.068	0.021
	(0.047)	(0.394)	(0.048)
	0.250	0.862	0.658
Ideological consistency	0.133	0.862	0.016
	(0.034)	(0.368)	(0.032)
	0.000	0.019	0.626
Strength of partisanship	0.034	0.032	0.027
	(0.010)	(0.087)	(0.010)
	0.001	0.713	0.007
News consumption	0.002	0.023	−0.011
	(0.009)	(0.070)	(0.009)
	0.831	0.742	0.267
Political knowledge	0.005	−0.094	0.000
	(0.008)	(0.070)	(0.008)
	0.528	0.181	0.995
Education	−0.010	0.028	0.000
	(0.039)	(0.351)	(0.038)
	0.795	0.938	0.995
Female	0.014	0.667	−0.09
	(0.022)	(0.196)	(0.022)
	0.525	0.001	0.000
Black	−0.006	−0.197	0.008
	(0.035)	(0.335)	(0.041)
	0.859	0.558	0.841
Latino	0.098	0.096	0.045
	(0.039)	(0.315)	(0.037)
	0.013	0.760	0.225
Asian American	0.118	0.231	0.120
	(0.042)	(0.764)	(0.076)
	0.005	0.762	0.114
Age	0.000	−0.011	−0.001
	(0.001)	(0.006)	(0.001)
	0.838	0.052	0.088
Constant	0.328	−1.626	0.307
	(0.055)	(0.538)	(0.058)
	0.000	0.003	0.000
R2	0.12	—-	0.08
N	853	853	853

Note: 2020 CES. Regression estimates, first and third columns. Logit estimates, second column. Standard errors in parentheses. p-values in italics.

Internal efficacy is also associated with a greater willingness to discriminate against others due to their political views. Moving from the lowest level of internal efficacy to the highest is associated with a 0.14 increase in a person’s willingness to treat someone differently because of their political beliefs, or about half a standard deviation shift on this 0 to 1 scale. Most people say that it is not okay to treat people differently because of their political beliefs. But when people feel confident about their ability to engage effectively in politics, they are more willing to say that it is okay to discriminate against rival partisans. Higher levels of internal efficacy are associated with higher levels of partisan intolerance, while those who are less confident about their capacity to influence politics believe that people should not be treated differently because of the political beliefs they hold.

The effects of political self-confidence are distinctive from those of self-esteem. In fact, higher levels of self-esteem tend to be associated with lower levels of partisan animosity. People with low self-esteem are more likely to say they avoid conversations with those who do not share their views, while those with high self-esteem are more willing to spend time with those who disagree with them politically. Those with high self-esteem are less likely to unfriend someone on social media for their political beliefs. At the lowest level of self-esteem, the predicted probability of blocking someone on social media is 46 percent, which drops to 22 percent at the highest level of self-esteem. Furthermore, people with high levels of self-esteem are less likely to believe that it is okay to treat someone differently because of their political beliefs. The predicted tolerance of partisan discrimination is 0.27 at the highest level of self-esteem, and climbs to 0.36 at the lowest level of self-esteem, just under a third of a standard deviation difference. These findings confirm past studies that show that high self-esteem is associated with greater political tolerance (Sniderman 1975; Sullivan et al. 1982). They also highlight the distinctiveness of domain-specific self-confidence in driving people’s out-party animosities. It is the particular self-assurance people feel within the political domain that drives their animus toward political opponents.

Experimental Evidence

As additional evidence of the relationship between internal political efficacy and people’s hostility toward the opposing party, we also draw on evidence from a survey experiment. We consider heterogeneity in reactions to partisan discrimination as a function of levels of internal efficacy. In the postelection wave of the 2020 Cooperative Election Study module, survey respondents were given a short vignette about a person vetting applicants for an internship.⁶ In the treatment conditions, she is described as rejecting the application after seeing that the prospective intern was the president of either a College Democrats or College Republicans group. In the control condition, the applicant is described as president of a nonpartisan college group.⁷ Using respondents’ partisanship, we determine whether the vignette describes discrimination against a copartisan or a member of the opposing party.⁸

We consider two outcomes. In the first, people were asked to assess whether they personally felt it was acceptable for the hiring employee to reject the application. In a second question, people were asked how likely it is that they would act in the same way if they found themselves in this situation. We find that people are more tolerant of discrimination against opposing partisans, but rate the acceptability of discrimination against co-partisans similarly to the control condition. In the nonpartisan control condition, most people rated the behavior as unacceptable, with a mean acceptability rating of 0.22 on a 0 to 1 scale.⁹ In the shared partisanship condition, the mean was 0.18, statistically indistinguishable from the control condition (t = 1.42, p = 0.16). However, when people read about a member of the opposing party being rejected for the internship, the mean rating of acceptability was 0.29, which is significantly higher than in the control condition (t = −2.79, p = 0.005). On the whole, people do not believe it is okay to treat someone differently based on their partisan loyalties. But their willingness to tolerate partisan discrimination is greater when it involves penalties against an out-party member.

Are those high in internal efficacy more likely to accept discrimination against an opposing partisan? We test this by looking for heterogeneous treatment effects as a function of internal efficacy as well as self-esteem. As shown in Supplementary Material table A2, the effects of the opposing partisan treatment vary as a function of internal efficacy, but not self-esteem. In figure 1, we plot treatment effects at low and high levels of internal efficacy. People evaluate the acceptability of discriminating against a member of the opposing party differently as a function of their feelings of political self-confidence. Those at the lowest level of internal efficacy frown at the prospects of this kind of partisan discrimination, where the predicted level of perceived acceptability of discrimination is 0.08. Those who feel the highest level of internal efficacy are far more tolerant of this behavior, with a predicted level of perceived acceptability of 0.38. As levels of internal efficacy climb, people are increasingly willing to see partisan discrimination as acceptable. In the previous survey analysis, we found that internal efficacy better explained out-party animosities rather than in-party affect. The results of this experimental analysis mirror this—where levels of political self-confidence moderate people’s assessments of out-party discrimination more so than in-party bias.

Figure 1.

Heterogeneous effects of internal efficacy on tolerance of out-party discrimination.

We also asked people whether they felt that they would do the same thing if they found themselves in this situation. On a scale where 0 means very unlikely to do the same thing and 1 means very likely to do the same thing, most said they would be unlikely to engage in partisan discrimination. In the control condition, the mean score is 0.18. We fail to find a significant effect associated with the copartisan treatment (mean = 0.13, t = −1.85, p = 0.065). However, we again find that people are more willing to discriminate against members of the opposing party. Compared to the control condition, people were significantly more likely to say they would personally reject the application from someone in the opposing party (mean = 0.33, t = 6.93, p = 0.0001).

Our main interest is in whether people at different levels of internal efficacy respond to these treatments differently. When considering heterogeneous treatment effects, we again find that those with high levels of internal efficacy were the most likely to say they would discriminate against an applicant for a position based on his political beliefs. At the lowest level of internal efficacy, the predicted likelihood of rejecting the out-party application is 0.12, which climbs to 0.44 for those at the highest level of internal political efficacy. To the extent to which people are willing to tolerate discrimination against a member of the opposing party, this effect is largest from people with high levels of internal efficacy. Those who feel most confident in their ability to influence politics are the most likely to choose to discriminate against a member of the rival party. These experimental analyses are limited in that this exploration of heterogeneous treatment effects cannot provide causal evidence. Nonetheless, they are useful in demonstrating the effects of internal efficacy in another decision setting beyond affective polarization (Broockman, Kalla, and Westwood 2023). These results indicate that those high in internal efficacy are more likely to accept discrimination against members of the opposing party.

Conclusions

Self-confidence has been thought to be a valuable psychological resource within politics. When people feel like they can influence politics, they are more likely to feel empowered to pursue their political interests. The efficacious are thought to be good citizens: attentive and engaged (Almond and Verba 1963). Yet we show that this kind of confidence has its negatives. People with higher levels of internal efficacy also report higher levels of affective polarization. Their confidence within politics is associated with greater resistance to talking to those who do not share the same views as well as greater willingness to tolerate discrimination against partisan opponents. In contrast, those who are more tentative in politics and less confident of their ability to make a difference are more open to spending time with those who hold contrary views. They are less likely to say that it is okay to treat someone differently because of their political beliefs.

Many people find politics to be a difficult domain and admit that they do not understand political issues as well as they would like to. These reservations have been sometimes thought to be a negative in politics, where lack of confidence keeps people from being able to effectively act on their interests. Yet our findings demonstrate an upside for lower levels of political self-assurance. People who admit that they are less confident about their political competence are less likely to derogate partisan rivals. Those with high levels of internal efficacy are not only more likely to be party activists (Almond and Verba 1963), but they are also more likely to be partisan loyalists.

Our results provide additional evidence of the upsides of weaker political commitments. Within the electorate, some people are deeply concerned about politics, while others go days without giving a thought to anything political. Those who are not as involved in politics are less likely to be affectively polarized, and report less extreme views and weaker partisan commitments (Krupnikov and Ryan 2022). The least politically sophisticated are less likely to evaluate new information in biased ways (Taber and Lodge 2006). The least politically engaged may be the most likely to receive deliberative rewards from conversations with those who do not share their views (Mutz 2006). Being interested, engaged, and confident about politics may promote partisan engagement and political participation, but may also contribute to partisan stubbornness, leading people to being hardheaded and adversarial in the face of political challenges. Those who feel most empowered within politics are among the most likely to express partisan hostilities.

Our results speak to the factors that contribute to affective polarization, and why some are more likely to admit to biases against partisan rivals. People’s animosities toward the opposing party reflect not just their partisan and ideological commitments—they also reflect how people think about their place in politics. People’s feelings of mastery within politics help explain some of the individual-level heterogeneity we see in levels of reported affective polarization. Those who believe in their ability to make a difference in politics seem to be more likely to have internalized the fault lines of partisan politics. Those who feel greater internal efficacy are ready to engage in political battles and are prepared to derogate the other side. Those who doubt their ability to really effect change in politics not only spend less time on political pursuits, but also show less psychological commitment to inter-party battles.

Our results highlight how people’s self-confidence within politics has different consequences than feelings of self-esteem. Like internal efficacy, self-esteem has been thought to be part of a “democratic personality” (Lasswell 1951), associated with outcomes like political interest, discussing politics, and participation (Rosenberg 1962; Carmines 1978; Rosenberg and Owens 2001). But people who feel a sense of self-confidence detached from politics are not any more prone to affective polarization. Instead, those with high self-esteem are less likely to block people on social media who disagree with them and are more willing to talk politics with members of the opposing political party. They are less willing to tolerate discrimination on partisan grounds. This is consistent with past work that shows that high self-esteem is good for political tolerance (Sniderman 1975; Sullivan et al. 1982; de Zavala et al. 2020). People with high self-esteem may be more likely to affiliate with a political party (Wolak and Stapleton 2020), but they are no more likely to express hostility toward the opposing party. It is people’s distinctive sense of self-confidence within the domain of politics that contributes to affective polarization, rather than their perceptions of their self-worth. People are the most tolerant of partisan discrimination when they are confident in their own political capabilities.

Footnotes

1

We use items that ask about ratings of the political parties to mirror the approach used in the American National Election Study. The affective polarization item includes both partisans and independents. For the in-party and out-party ratings measures, we classify congruent party ratings based on self-reported partisanship. Leaners are coded as partisans for these items, while those who do not affiliate with a party or do not lean toward one of the parties are excluded from these ratings.

2

Question wordings for all items are reported on Supplementary Material pp. 2–4. On Supplementary Material p. 2, we provide additional details about the 2020 CES, including the dates of the survey, the cooperation rate, and how respondents enter the sample.

3

The authors demonstrate that this name-liking measure is positively associated with other measures of implicit self-esteem, including those based on implicit association tests and ratings of the letters in one’s name.

4

We apply survey weights in the analyses reported in tables 1 and 2.

5

On Supplementary Material pp. 8–11, we present replication models using the 2013 ANES Recontact Study. With this second dataset, we confirm that internal efficacy is associated with greater affective polarization and greater negativity toward the opposing party. We find that self-esteem, as measured with an alternate measure of explicit self-esteem, is associated with lower affective polarization and warmer views of the opposing party.

6

The full wording is included on Supplementary Material p. 4.

7

Two nonpartisan conditions were considered, describing the applicant as either president of the Government and Politics Club or the Video Game Club. We pool these as a single control condition because people do not evaluate the vignette differently across the political and apolitical conditions.

8

We assign independents based on which party they lean toward in the branched partisanship wording. For pure independents, we determine party favorability based on which party was rated more warmly in the pre-election party feeling thermometer items.

9

We do not use survey weights in the examination of heterogeneous treatment effects in this experiment.

Supplementary Material

Supplementary Material may be found in the online version of this article: https://doi.org/10.1093/poq/nfad064.

Data Availability

Replication data and documentation are available at https://doi.org/10.7910/DVN/IRUNWG.