Trustworthiness and Ideological Similarity (But Not Ideology) Promote Empathy

Abstract

The current highly polarized U.S. political culture impedes people’s ability to live and work together effectively. Here we examine one factor that may play a role: selective empathy based on shared political ideology. Across seven studies (N = 3,476), participants read about a hypothetical politician and his political ideology, trustworthiness, or both. Participants reported their empathy for the politician after learning he was fined (Studies 1–6) or injured (Study 7). When trustworthiness alone was manipulated, liberals and conservatives expressed similar levels of empathy, with greater empathy for the more trustworthy politician. However, when the politician’s ideology alone was manipulated, participants reported greater empathy for the politician who shared their ideology. When trustworthiness and ideology were manipulated, selective empathy was observed when the politician was trustworthy. Participant ideology alone had little effect on empathy. The results suggest that empathy is sensitive to both trustworthiness and ideological match, but not ideology itself.

Keywords

empathy trust political ideology political polarization

The level of political polarization in the United States has shown a sharp increase in recent years (Doherty, 2017). Most Americans now identify the degree of conflict between supporters of the Democratic and Republican parties as either strong or very strong and see these divisions as greater than those based on racial identity or socioeconomic status (Gramlich, 2017). Moreover, adherents of the two parties are likely to disagree not only on which plans and policies they prefer, but on basic facts as well (Kahan et al., 2017; Washburn & Skitka, 2018). In addition, Americans often avoid hearing political views that run counter to their own, preferring to maintain mental synchrony with like-minded individuals and avoid cognitive dissonance (Frimer et al., 2017). Because social emotions such as empathy play an important role in motivating prosocial behavior (Batson, 2010), differential empathy based on ideological similarity is likely to be an important barrier to efforts to allay ideological tensions. Some evidence supports the existence of this form of selective empathy, as individuals do indeed tend to show less empathy for people with opposing political ideologies than those with similar ideologies (Hasson et al., 2018). In the present research, we propose a new variable to help account for ideologically based disparities in empathy: trust. People may reserve empathy for trustworthy individuals because being empathic involves some degree of emotional investment, and people may not want to invest in someone who is untrustworthy and therefore likely to take advantage of them.

Ideologies involve socially shared belief systems about society that can be used as a framework for interpreting and responding to life’s problems (Jost et al., 2009). Some previous research examining the relation between ideological affiliation and empathy has suggested that liberals may be more empathic than conservatives. For example, in a large web-based study (Iyer et al., 2012; see also McCue & Gopoian, 2000) liberals reported significantly more empathic concern than both conservatives and libertarians on the widely used Interpersonal Reactivity Index (Davis, 1983). Similarly, a study that coded psychological themes from life-narrative interviews found that liberals were more likely than conservatives to focus on empathy (McAdams et al., 2008). In addition, there is evidence that people high in social-dominance orientation—a key component of conservative ideology that reflects the endorsement of social hierarchy—not only report less empathy (Bäckström & Björklund, 2007; McFarland, 2010), but also show lower brain activity in areas associated with empathy (Chiao et al., 2009). One possible explanation for this perceived “empathy gap” is that the disparities might be driven by individual differences in the types of people who develop liberal versus conservative ideologies. For example, conservative values may be associated with relatively high levels of concern with avoiding fear and anxiety, which could lead to avoidance of emotional engagement of others that may underlie empathy. In addition, conservatives are more likely than liberals to endorse just-world beliefs (Benabou & Tirole, 2006), which could lead them to experience less empathy. For example, conservatives may believe that people who experience negative outcomes are less worthy of empathy because they could have done more to prevent those outcomes, especially if they believe those people are untrustworthy.

An alternative possibility is that liberals and conservatives might not differ in their capacity for empathy, but instead may each show more empathy toward individuals who share their political ideology than those who oppose it, leading to the phenomenon of interest in the present research that we call selective empathy. In line with this possibility is evidence that liberals and conservatives show a similar level of negativity in their attributions and intolerance toward ideologically dissimilar people (Brandt et al., 2014). This worldview-conflict perspective is also in line with broader theoretical perspectives, according to which people prefer similar others and are prejudicial in response to perceived threat (Crawford & Brandt, 2020). It is also generally consistent with evidence that partisan intensity predicts cognitive rigidity for both liberals and conservatives (Zmigrod et al., 2020), and evidence that liberals and conservatives have similar physiological responses to threatening stimuli (Bakker et al., 2020).

Hasson et al. (2018) argued that one reason liberals could have shown higher levels of empathy in prior studies is that the target groups in question typically oppose conservative values. To address this possibility, Hasson et al. (2018) asked participants to report on their empathy for injured protesters who either did or did not share their political ideology. They found that liberals experienced more empathy than conservatives did, but they also found evidence of selective empathy: the level of empathy was greater when participants supported the protesters’ cause than when they did not.

If indeed people show selective empathy based on political ideology, this would be an example of a group-based emotion which depends on a view of the self as a group member and not just as a distinct individual (Goldenberg et al., 2016). This type of selective empathy may be part of a more general phenomenon in which people tend to have less empathy for outgroup members than for ingroup members (Cikara et al., 2011; Combs et al., 2009). Here, we propose that trust may play a major role in people’s willingness to empathize with others. This possibility is based on the idea that empathy involves an emotional investment, and that interpersonal trust, which by definition involves allowing oneself to be emotionally vulnerable to others (Righetti & Finkenauer, 2011), can affect people’s willingness to make such an investment. In addition, people may have concerns that the plight of an untrustworthy victim may not be as advertised, and consequently choose to keep a safe emotional distance from such an individual. A lack of empathy also precludes costly behavioral investments, as empathic concern has been shown to predict prosocial behavior (Batson, 2010; Dovidio et al., 1990; Hoffman, 2008; Toi & Batson, 1982; see Eisenberg & Miller, 1987, for a review).

If trust does indeed affect empathy, this raises the question of what the relationship between trust and political ideology might be. One possibility raised by Tetlock et al. (2013) is that ideology may serve as a cue for inferring trustworthiness when more direct information about trustworthiness is unavailable. Direct information about trustworthiness could therefore be expected to override the effect of ideology on perceived trustworthiness. Tetlock et al. (2013) tested this possibility among MBA students and executives taking continuing-education courses in the context of applying either process-based or outcome-based accountability for employees. They found that in the workplace, conservatives generally favor outcome-based accountability, whereas liberals generally favor process-based accountability. However, they found that ideology becomes immaterial when direct information about the trustworthiness of employees is available. Still, it is unclear whether information about trustworthiness might override any effects of political ideology on empathy. In addition, it is unclear whether such a pattern of effects would hold given the heightened political polarization in the United States (Doherty, 2017).

In the present research, we developed a new paradigm to address these issues. Participants read about a local politician in a hypothetical scenario. The politician is described as running for local office and as facing a significant financial penalty for filling out his campaign finance paperwork incorrectly (Studies 1–6) or as having sustained an injury while volunteering at an animal shelter (Study 7). The description of the politician was manipulated between subjects in different ways as we explored the effects of trust and ideology on empathy.

We ran seven studies using the local politician scenario. In our first two studies, we manipulated the trustworthiness of the target to test whether low trust undermines empathy. In our second two studies, we manipulated the target’s political ideology to test for selective empathy. In our fifth, sixth, and seventh studies, we manipulated both the target’s trustworthiness and the target’s political ideology to explore interactive effects (e.g., the possibility that trust overrides political affiliation when participants make empathy judgments). Studies 2, 4, and 6 were preregistered replications of Studies 1, 3, and 5, respectively; Study 7, preregistered separately, was a variation on Studies 5 and 6. Preregistrations were done via the AsPredicted website (https://aspredicted.org/kf5a4.pdf; https://aspredicted.org/88gd2.pdf).We used the results of Studies 1 and 3 to determine the sample sizes for Studies 2 and 4. Specifically, as indicated in the initial preregistration, a power analysis using G*Power revealed that a sample size of n = 400 for Studies 2 and 4 would achieve 80% power for a small effect size of f = 0.14. We doubled this sample size for Studies 6 and 7 because they involved a greater number of factorial cells.

Study 1

The goals of the first study were to find out if interpersonal trust and participant political ideology play roles in determining how much empathy people feel for others who are experiencing misfortune. First, based on the idea that people may be motivated to avoid emotional investment in those who might exploit them, we predicted that targets seen as trustworthy would elicit greater empathy from participants than targets seen as less trustworthy. To test this prediction, we developed a local politician scenario in which the target, a hypothetical local politician, suffers a misfortune. We manipulated information reflecting his trustworthiness and measured participants’ empathic responses. The second goal of the first study was to ascertain whether or not the “empathy gap”—the idea that liberals are more empathic than conservatives—would replicate previous findings. Hence, a measure of political ideology was included.

Method

Participants

After excluding 24 participants due to a failure to respond correctly to attention check questions, the final sample included 134 participants (M_age = 37.35, SD = 12.45, 84 women, 50 men). In this study and all subsequent studies, participants from the United States were recruited from Amazon’s Mechanical Turk. In addition, participants faced no significant risks, confidentiality and institutional review board’s (IRB) approval were maintained at all times, and exclusions were made prior to data analysis.

Procedure

Participants were randomly assigned to one of two conditions (low vs. high trust); in this and all subsequent studies, conditions were made to be approximately equal in size. The manipulation was adapted from the trust manipulation used by Tetlock et al. (2013) so that it was applicable to individuals rather than organizations.

The wording of the low (vs. high) trust conditions was as follows (manipulated text is shown here in bold but was presented to participants in plain text):

Joe Chambers is running for Congress in his district. He doesn’t care about the problems of people in his district and is primarily concerned with his own power and status [He cares deeply about the problems of people in his district and is primarily concerned with improving his community]. Recently, Joe filled out his campaign finance paperwork incorrectly, and has to pay 20% of the funds he raised for his campaign in fines.

All participants were then asked how much empathy they would feel for Joe on a scale from 0.0 (no empathy) to 10.0 (a lot of empathy). All scales in this and subsequent studies included specificity to one decimal place.

As a manipulation check, participants were asked how trustworthy they think Joe is on a scale from 0.0 (not trustworthy at all) to 10.0 (extremely trustworthy), with higher scores representing higher trustworthiness. Attention check questions followed.

Finally, participants indicated how socially liberal or conservative they are on a scale from 0.0 (very conservative) to 10.0 (very liberal).

Results

A one-way ANOVA revealed that participants felt significantly more trust in the high trust condition (M = 5.82, 95% CI = [5.35, 6.29]) than those in the low trust condition (M = 1.17, 95% CI = [0.71, 1.63]), F(1, 132) = 194.82, p < .001, η² = .60, demonstrating that the manipulation of trust was successful.

To test the effects of manipulated trust and participant ideology on empathy, we ran a regression analysis with the two variables entered as predictors in the first step. The main effect of the trust condition on empathy was strong and significant (b = −3.78, 95% CI = [−4.59, −2.96], p < .001). Participants empathized more with a politician in the trust condition than in the distrust condition. The main effect of participant ideology on empathy was not significant (b = 0.01, 95% CI = [−0.14, 0.16], β = .01, p = .895). In the second step, we added a trust × ideology interaction but it was not significant (b = −0.18, 95% CI = [−0.49, 0.12], p = .243), indicating that the effect of the trust manipulation on empathy was not significantly different between liberals and conservatives (see Figure 1).

Figure 1.

The effects of participant political ideology and trust on empathy in Studies 1 and 2.

Study 2

The goal of Study 2 was to replicate the findings of Study 1 in a larger sample. The prediction remained the same that targets seen as untrustworthy would elicit less empathy than targets seen as trustworthy. This study was preregistered on AsPredicted (https://aspredicted.org/kf5a4.pdf), as were Studies 4 and 6, which, similarly, are preregistered replications of Studies 3 and 5, respectively. Although, in the preregistration for Studies 2, 4 and 6, we assumed that we would divide the sample into three groups based on the ideology scores, we ultimately decided to include analyses based on categorizing this variable in the Supplemental online Material (SOM), and to instead present here analyses based on the original variable. While in our case the results were ultimately the same, relevant literature on the categorization of continuous data shows that it is not only unnecessary for statistical analysis but may in fact come with considerable costs (Royston et al., 2006) such as loss of information, reduced effect size and power, and loss of measurement reliability (Cohen, 1983; MacCallum et al., 2002).

Method

Participants

After excluding 78 participants due to a failure to respond correctly to the attention check questions, the final sample included 423 participants (M_age = 38.29, SD = 12.39, 274 women, 146 men, and 3 participants who did not indicate gender).

Procedure

The procedure was the same as that of Study 1.

Results

A one-way ANOVA revealed that participants thought Joe was significantly more trustworthy in the high trust condition (M = 5.32, 95% CI = [5.09, 5.56]) than in the low trust condition (M = 0.72, 95% CI = [0.48, 0.97]), F(1, 421) = 710.87, p < .001, η² = .63, demonstrating that the manipulation of trust was once again successful.

Replicating the effects of Study 1, the main effect of manipulated trust on empathy was significant (b = −4.19, 95% CI = [−4.58, −3.79], p < .001), but the main effect of participant ideology on empathy was not significant (b = 0.02, 95% CI = [−0.05, 0.09], β = .02, p = .574). The trust × ideology interaction was not significant either (b = −0.07, 95% CI = [−0.20, 0.07], p = .338), demonstrating no significant differences between liberals and conservatives in their empathic reactions to high versus low trust targets (see Figure 1).

Discussion

Providing support for our first hypothesis, Studies 1 and 2 suggest that low trust undermines empathy. In contrast to other research on political ideology and empathy (Iyer et al., 2012), liberals did not report more empathy than conservatives in either study. In contrast to the lack of influence of participant ideology on empathy, the empathic reaction of both liberals and conservatives was strongly influenced by the politician’s trustworthiness. Results suggest that differences in empathy based on political ideology, noted in previous research but absent here, might be dependent on the paradigms used to assess empathy.

Study 3

The goal of Study 3 was to discover whether selective empathy manifests utilizing a novel paradigm. Similar to Studies 1 and 2, we used a local politician scenario to elicit empathy. However, in this study, we manipulated information about the target’s political views rather than his trustworthiness. We predicted that participants would exhibit selective empathy based on ideological similarity. That is, liberals would feel more empathy for a target with liberal views than for a target with conservative views, and conservatives would feel more empathy for a target with conservative views than a target with liberal views.

Method

Participants

After excluding 33 participants due to a failure to respond correctly to attention check questions, the final sample included 159 participants (M_age = 39.06, SD = 12.68, 102 women, 56 men, 1 person did not report gender).

Procedure

Participants were randomly assigned to one of two conditions, liberal politician or conservative politician. The wording of the liberal politician (vs. conservative politician) conditions was as follows (manipulated text is shown here in bold but was presented to participants in plain text):

Joe Chambers is running for Congress in his district. His campaign speeches focus on preventing corruption and on promoting a woman’s right to have safe and legal abortions [promoting legal rights for unborn children]. Recently, Joe filled out his campaign finance paperwork incorrectly, and has to pay 20% of the funds he raised for his campaign in fines.

All participants were then asked how much empathy they would feel for Joe on a scale from 0.0 to 10.0, with higher scores representing more empathy. Attention check questions followed. Finally, participants indicated how socially liberal or conservative they are on a scale from 0.0 to 10.0, with higher scores indicating more liberal.

Results

To test the effects of the politician’s and the participant’s ideology on empathy we ran a regression analysis. Two ideological variables were entered as predictors in the first step. The main effect of politician ideology on empathy was not significant (b = 0.50, 95% CI = [−0.46, 1.45], p = .309). The main effect of participant ideology on empathy was not significant either (b = −0.002, 95% CI = [−0.17, 0.16], β = −.002, p = .979). However, the politician ideology × participant ideology interaction, which was entered in the second step, was significant (b = 0.51, 95% CI = [0.19, 0.83], p = .002, R² change = .06), indicating that liberals and conservatives differed in the empathy they expressed toward a politician of the same versus different political ideology.

A simple slope analysis showed that at high levels of social liberalism (+1 SD), the effect of politician ideology on empathy was significant and positive (b = 2.00, 95% CI = [0.68, 3.31], p = .003). In contrast, at low levels of social liberalism (−1 SD), the effect of politician ideology was not significant (b = −1.00, 95% CI = [−2.32, 0.31], p = .134). The Johnson–Neyman technique, which identifies the values of the moderator for which the regression slopes become significantly different from zero, showed that at the value of social liberalism = 1.90 (−1.40 SD), the effect became significant and negative (b = −1.60, 95% CI = [−3.20, 0], p = .05); below those levels of social liberalism, a conservative politician evoked significantly more empathy than a liberal politician (see Figure 2).

Figure 2.

The effects of politician and participant ideology on empathy in Studies 3 and 4.

Looking at it another way, in the conservative politician scenario, the effect of social liberalism was negative and significant (b = −0.24, 95% CI = [−0.46, −0.02], p = .031), but it was positive and significant in the liberal politician scenario (b = 0.27, 95% CI = [0.04, 0.51], p = .022).

Study 4

The goal of Study 4 was to replicate the previous study’s findings in a larger sample. The prediction remained the same; in alignment with previous findings on selective empathy, we hypothesized that liberals would feel more empathy for a target with liberal views than a target with conservative views, and conservatives would feel more empathy for a target with conservative views than a target with liberal views. Like Study 2, Study 4 was preregistered on AsPredicted before any data were collected.

Method

Participants

After excluding 99 participants due to a failure to respond correctly to attention check questions, the final sample included 384 participants (M_age = 38, SD = 12.35; 1 participant did not report age, 249 women, 134 men, 1 person did not report gender).

Procedure

The procedure was nearly the same as that of Study 3. The only difference was the addition of a trust measure in which participants were asked how trustworthy they think Joe is on a scale from 0.0 to 10.0, with higher scores representing more trustworthiness.

Results

To test the effects of politician and participant ideology on empathy we ran the same regression analysis as in Study 3. The main effect of politician ideology on empathy was not significant (b = 0.21, 95% CI = [−0.38, 0.79], p = .485). In contrast to Study 3, the main effect of participant ideology on empathy was weak but significant (b = 0.10, 95% CI = [0.003, 0.20], β = .10, p = .043), suggesting that liberals were overall more empathic than conservatives. More importantly, the politician ideology × participant ideology interaction entered in the second step was significant (b = 0.54, 95% CI = [0.35, 0.73], p < .001; R² change = .07), indicating that liberals and conservatives differed in their empathy toward a politician of the same versus different political ideology. A simple slope analysis showed that at high levels of social liberalism (+1 SD), the effect of manipulated ideology of the politician was significant and positive (b = 1.80, 95% CI = [1.00, 2.60], p < .001), indicating higher empathy toward a liberal politician than toward a conservative politician. However, at low levels of social liberalism (−1 SD), the effect of politician’s liberal ideology was reversed (b = −1.35, 95% CI = [−2.14, −0.56], p < .001) (see Figure 2).

Described differently, the effect of social liberalism was negative in the case of a conservative politician (b = −0.18, 95% CI = [−0.32, −0.04], p = .011), but it was positive in the case of a liberal politician (b = 0.36, 95% CI = [0.22, 0.49], p < .001).

Discussion

Studies 3 and 4 revealed consistent support for our hypothesis that liberals and conservatives both feel more empathy for others who share their ideological views than for those who do not. We also found that in Study 4, but not in Study 3, there was evidence for liberals being more empathic than conservatives. However, the effect was relatively small and should be interpreted cautiously.

Study 5

Study 5 examines how information about both the target’s trustworthiness and his political ideology would affect participants’ empathy. We again used a local politician scenario to elicit empathy along with the same manipulations of target trustworthiness and ideology from the previous studies, but now crossed factorially. We did not specify hypotheses about interactive effects but considered different possibilities. For instance, both liberals and conservatives might be negatively biased against the outgroup even when information about trust is provided. An alternative possibility is that there is no difference between liberals and conservatives in their empathy based on target ideology, and that information about trustworthiness fully overrides the role of ideological similarity; such results would support the idea of ideological information being a proxy for information about trustworthiness when information more directly indicative of trustworthiness is unavailable (Tetlock et al., 2013). The third possibility is that direct information about target trustworthiness would only partially override the effect of ideological similarity, perhaps because the effect involves some combination of outgroup bias and an uncertainty-reducing heuristic for inferring trustworthiness.

Method

Participants

After excluding 125 participants due to a failure to respond correctly to the attention check questions, the final sample included 322 participants (M_age = 38.80, SD = 12.58, 204 women, 117 men, 1 person did not indicate gender).

Procedure

We manipulated the trustworthiness and ideology of the politician. Participants were randomly assigned to one of four conditions (low trust/liberal politician, high trust/liberal politician, low trust/conservative politician, and high trust/conservative politician). The wording of the manipulation took the same form as in the previous studies, but both manipulations were included together. For example, participants in the low trust/liberal condition read the following scenario:

Joe Chambers is running for Congress in his district. He doesn’t care about the problems of people in his district and is primarily concerned with his own power and status. His campaign speeches focus on preventing corruption and on promoting a woman’s right to have safe and legal abortions. Recently, Joe filled out his campaign finance paperwork incorrectly, and has to pay 20% of the funds he raised for his campaign in fines.

All participants were then asked how much empathy they would feel for Joe on a scale from 0.0 to 10.0, with higher scores representing more empathy. As a manipulation check, participants were asked how trustworthy they think Joe is on a scale from 0.0 to 10.0, with higher scores representing more trustworthy. Attention check questions followed. Finally, participants indicated how socially liberal or conservative they are on a scale from 0.0 to 10.0, with higher scores indicating more liberal.

Results

Perceived trustworthiness

A two-way ANOVA revealed a main effect of the trustworthiness manipulation on perceived trust. Participants in the high trust condition perceived Joe as significantly more trustworthy (M = 5.54, 95% CI = [5.20, 5.88]) than those in the low trust conditions (M = 1.62, 95% CI = [1.27, 1.96]), F(1, 318) = 254.15, p < .001, η² = .44, demonstrating that the manipulation of trust was successful. The manipulation of politician ideology had no significant effect on the perception of trustworthiness (p = .163). The trust (manipulated) × politician ideology interaction was not significant either (p = .383).

Empathy

To test the effects of manipulated trust and manipulated politician ideology as well as participant ideology on empathy, we ran a regression analysis with the three variables entered as predictors in the first step. The main effect of the trust condition on empathy was significant (b = 3.29, 95% CI = [2.73, 3.86], p < .001). There was not a main effect of participant ideology on empathy (p = .455) or a main effect of politician ideology on empathy (p = .241).

In the second step, we added three two-way interactions, trust × politician ideology, trust × participant ideology, and politician ideology × participant ideology, as well as a three-way interaction, trust × politician ideology × participant ideology, which was marginally significant (b = 0.38, 95% CI = [−0.01, 0.78], p = .058).¹ Follow-up analyses revealed that the politician ideology × participant ideology interaction was significant in the high trust condition (b = 0.52, 95% CI = [0.22, 0.81], p < .001) but it was not significant in the low trust condition (p = .325). Specifically, in the high trust condition, we replicated the pattern of results found in Study 4. At high levels of social liberalism (+1 SD), the effect of manipulated ideology of the politician was significant and positive (b = 1.56, 95% CI = [0.45, 2.68], p = .006), indicating that liberals empathized more with a liberal politician than with a conservative politician. However, at low levels of social liberalism (−1 SD), the effect of politician’s liberal ideology was reversed (b = −1.39, 95% CI = [−2.61, −0.18], p = .025), indicating that conservatives empathized less with a liberal politician than with a conservative politician. Looking at the effects of social liberal ideology, there was a positive and significant effect in the liberal politician scenario (b = 0.41, 95% CI = [0.19, 0.63], p < .001), but a negative, non-significant effect in the conservative politician scenario (p = .293). In the low trust conditions, politician ideology was unrelated to empathy regardless of participant ideology, and both slopes for ideology were non-significant (see Figure 3).

Figure 3.

The effects of trust, politician ideology, and participant ideology on empathy in Studies 5 and 6.

Study 6

The goal of Study 6 was to determine whether the findings of Study 5 would replicate in a larger sample. Similar to Studies 2 and 4, Study 6 was preregistered on AsPredicted before any data were collected. Specifically, we preregistered that the effects seen when only trust or only political ideology was manipulated would also be seen when both are manipulated at once: (a) participants will have more empathy for targets who show characteristics associated with being trustworthy, as compared with not trustworthy, and (b) both participants who identify as socially liberal and participants who identify as socially conservative will show more empathy for targets whose political views are similar to their own. We did not make specific predictions about interactive effects but explored the same possibilities as outlined in Study 5.

Method

Participants

After excluding 228 participants due to a failure to respond correctly to attention check questions, the final sample included 810 participants (M_age = 36.99 SD = 11.89 [range = 18–74], 514 women, 294 men, 2 people did not indicate gender.

Procedure

The procedure was the same as that of Study 5.

Results

Perceived trustworthiness

A two-way ANOVA revealed a main effect of trust. Participants in the high trust conditions perceived Joe as significantly more trustworthy (M = 5.06, 95% CI = [4.85, 5.28]) than those in the low trust conditions (M = 1.51, 95% CI = [1.29, 1.74]), F(1, 806) = 504.71, p < .001, η² = .39, demonstrating that the manipulation of trust was successful. The manipulation of politician ideology had no significant effect on the perception of trustworthiness (p = .439). The trust (manipulated) × politician ideology interaction was not significant either (p = .547).

Empathy

To test the effects of manipulated trust and manipulated politician ideology as well as participant ideology on empathy, we ran the same regression analysis as in Study 5. The main effect of the trust condition on empathy was significant (b = 3.22, 95% CI = [2.88, 3.56], p < .001). The main effect of participant ideology on empathy was marginally significant (b = 0.06, 95% CI = [−0.002, 0.12], β = .06, p = .09).² The main effect of politician ideology was not significant (p = .755). The trust × politician ideology × participant ideology interaction was significant (b = 0.39, 95% CI = [0.15, 0.62], p = .001)³ (see Figure 3). Further analyses revealed that the politician ideology × participant ideology interaction was significant in the high trust condition (b = 0.45, 95% CI = [0.29, 0.61], p < .001) but it was not significant in the low trust condition (p = .436). In the high trust condition, at high levels of social liberalism (+1 SD), the effect of manipulated ideology of the politician was significant and positive (b = 1.27, 95% CI = [0.62, 1.92], p < .001), indicating that liberals empathized more with a liberal politician than with a conservative politician. However, in the same trust condition but at low levels of social liberalism (−1 SD), the effect of politician’s liberal ideology was reversed (b = −1.35, 95% CI = [−2.01, −0.69], p < .001), indicating that conservatives empathized less with a liberal politician than with a conservative politician. This time, the effect of social liberal ideology on empathy was significant in both scenarios (in the high trust condition). It was positive for a liberal politician (b = 0.31, 95% CI = [0.20, 0.42], p < .001) but negative for a conservative politician (b = −0.14, 95% CI = [−0.26, −0.02], p = .018). In the low trust conditions, politician ideology was unrelated to empathy regardless of participant ideology.

Discussion

Studies 5 and 6 revealed that a lack of trust undermines empathy, and ideological similarity becomes important when the target is not clearly untrustworthy. When participants were given reasons to distrust a politician, their ideological similarity to the politician did not appear to inform empathy—rather, empathy was uniformly low in the low trust conditions, which may be because people tend to avoid emotional investment in those who they think might take advantage of them. However, when trust was higher, ideological similarity became important for informing empathy.

It is possible—and in alignment with the idea of ideological similarity serving as an uncertainty-reducing heuristic or proxy for information about trustworthiness—that the degree to which ideological similarity influences empathy is dependent on the degree to which the politician’s trustworthiness remains uncertain. This possibility exists despite ideological similarity affecting empathy in the high trust conditions because it is unclear whether the trust conditions were equally diagnostic of trust. It is possible that the low trust manipulation was stronger than the high trust manipulation, such that participants in the low trust conditions felt relatively certain that the target was untrustworthy, but participants in the high trust conditions were not certain that the target was trustworthy. There is also the possibility that the low trust manipulation negated the manipulation of political ideology, which could explain the lack of influence ideological similarity had on empathy in the low trust conditions. Relatedly, there may be a floor effect in the low trust conditions. Moreover, from the results thus far, it is unclear whether people reduce their empathy for an ideologically dissimilar target, increase their empathy for an ideologically similar target, or some amount of both. Given these considerations as well as an interest in exploring the generalizability of the observed effects, we conducted a seventh study.

Study 7

An additional study, preregistered on AsPredicted (https://aspredicted.org/88gd2.pdf), was done to address some of the limitations of our previous studies. We used a local politician scenario again and as in Studies 5 and 6, we manipulated both trust and the politician’s political ideology and examined participants’ empathy. However, in this study, we added a neutral control political ideology condition, such that the politician is not clearly liberal or conservative, to explore if previously seen effects may be due to increased empathy for the ingroup, decreased empathy for the outgroup, or both. We also measured the perceived political ideology of the target. Moreover, we altered the wording of the trust manipulation to try to increase trust in all conditions, both to avoid a floor effect in the low trust conditions and to strengthen the high trust manipulation. Furthermore, changes to the wording of the scenario allow us to explore how specific the mechanism of the observed effect is—that is, do the results generalize to a different situation in which the negative outcome experienced is unrelated to political work? Finally, we added a multiple-item measure of empathy.

We expected similar results to those of Studies 5 and 6. In particular, we hypothesized that participants would have more empathy for trustworthy targets who share their political ideology. We also expected that ideological match would matter more when the target is trustworthy than when he is untrustworthy. We were also interested in exploring what happens when the target’s political ideology is not indicated.

Method

Participants

After excluding 487 participants due to a failure to respond correctly to the attention check questions and 2 people due to nonsensical responses to age, the final sample included 1,244 participants (M_age = 39.05, SD = 13.03, 701 women, 539 men, 4 specified other).

Procedure

We manipulated the trustworthiness and ideology of the politician. Participants were randomly assigned to one of six conditions (low trust/liberal politician, high trust/liberal politician, low trust/conservative politician, high trust/conservative politician, low trust/neutral politician, or high trust/neutral politician). The wording of the conditions is as follows, with the condition differences shown in bold, although nothing was in bold in the versions presented to participants and the second sentence was removed for the neutral conditions:

Joe Chambers, a local politician, is running for State Senate. Many of his campaign speeches focus on promoting a woman’s right to have safe and legal abortions [on promoting legal rights for unborn children]. Joe has raised a lot of money for his campaign because many people support his cause. However, in violation of the law, he used some of this money to take his family on a vacation. While Joe’s colleagues think Joe is genuinely committed to his beliefs, they do not find him to be very trustworthy. [In full compliance with the law, he only spends this money on relevant campaign matters. Joe’s colleagues think Joe is genuinely committed to his beliefs, and they find him to be extremely trustworthy in all domains.]

In his free time, Joe enjoys volunteering at his local animal shelter. Despite taking all the appropriate precautions, Joe was attacked by a recently rescued dog and had to get 6 stitches on the side of his face.

All participants were asked how much empathy they would feel for Joe on a scale from 0.0 to 10.0, with higher scores representing more empathy. They were also asked, as another measure of empathic concern, to what extent they feel tender, compassionate, and sympathetic toward Joe on scales from 0.0 to 10.0, with higher scores representing more empathy. These additional adjectives were taken from empathy research by Batson et al. (1987).

As a manipulation check, participants were asked how trustworthy they think Joe is on a scale from 0.0 to 10.0, with higher scores representing more trustworthy. Participants also indicated how socially liberal or conservative they believed Joe to be on a scale from 0.0 (very conservative) to 10.0 (very liberal). Similarly, they indicated if they thought Joe was more likely to be a Democrat or Republican on a 0.0 (Definitely a Republican) to 10.0 (Definitely a Democrat) scale, with a centered anchor labeled Equally likely to be either. Then, participants indicated how socially liberal or conservative they are on a scale from 0.0 (very conservative) to 10.0 (very liberal). They also indicated to what extent they identified as a Republican, a Democrat, and an Independent on 0.0 to 10.0 scales, higher scores being higher identification. Attention check questions followed.

Results

Perceived trustworthiness

A two-way ANOVA revealed a main effect of trust condition on perceived trustworthiness. Participants in the high trust conditions perceived Joe as significantly more trustworthy (M = 7.27, 95% CI = [7.01, 7.54]) than those in the low trust conditions (M = 3.57, 95% CI = [3.38, 3.76]), F(1, 1242) = 737.50, p < .001, η² = .37, demonstrating that the manipulation of trust was successful. Notably, as intended, the mean levels of trust for both low and high trust conditions were higher than in previous studies. The manipulation of politician ideology also had a significant effect on the perception of trustworthiness (p = .019), and there was a significant interaction between both manipulations (p = .032). We discuss this further in the Additional Exploratory Analyses section.

Empathy

To test the effects of manipulated trust and manipulated politician ideology as well as participant ideology on empathy, we ran the same regression analysis as in Studies 5 and 6. The only difference was that now there were three ideology conditions instead of two, as in Studies 5 and 6. We used the conservative politician condition as a reference category and we dummy-coded the variable into two variables (i.e., liberal vs. conservative and control vs. conservative).

In the first step, we tested the main effects of three predictors. The main effect of the trust condition on empathy was significant (b = 2.17, 95% CI = [1.91, 2.43], p < .001). The main effect of participant ideology on empathy was also significant (b = 0.08, 95% CI = [0.04, 0.13], p < .001). The comparison between the liberal and conservative politician was not significant (p = .383), but the comparison between the conservative politician and control condition was significant, with higher levels of empathy expressed toward the unidentified politician than toward the conservative one (b = 0.39, 95% CI = [0.07, 0.71], p = .015).

Most importantly, the trust × politician ideology (liberal vs. conservative) × participant ideology interaction, which was added in the second step, was also significant (b = 0.31, 95% CI = [0.09, 0.54], p = .006) (see Figure 4). Replicating the results from previous studies, in the high trust condition, at high levels of social liberalism (+1 SD), the effect of manipulated ideology of the politician was significant and positive (b = 1.46, 95% CI = [0.84, 2.09], p < .001), indicating that liberals empathized with a liberal politician more than with a conservative politician. Also in the high trust condition, but at low levels of social liberalism (−1 SD), the effect of politician’s liberal ideology was significant and negative (b = −1.11, 95% CI = [−1.76, −0.46], p < .001), indicating that conservatives empathized less with a liberal politician than with a conservative politician. The trust × politician ideology (control vs. conservative) × participant ideology interaction was not significant (p = .274). When we recoded the politician ideology variable to obtain a comparison between the control condition and the liberal condition, it was not significant either (p = .093). In the low trust conditions, politician ideology was unrelated to empathy regardless of participant ideology (all ps > .093).

Figure 4.

The effects of trust, politician ideology, and participant ideology on empathy in Study 7.

The results obtained for the composite measure of empathy are the same (see Table S7 in SOM).

Additional Exploratory Analyses

Perceived trustworthiness

We explored the idea that when the politician’s ideology matches participant ideology, participants report higher trust than when ideology is dissimilar. In Study 4, we explored whether there was an effect of participant ideology and politician ideology on perception of trustworthiness. Similar to the effects obtained for empathy in that study, the politician ideology × participant ideology interaction was significant (b = 0.43, 95% CI = [0.26, 0.59], p < .001; R² change = .06), indicating that liberals and conservatives differed in how trustworthy they perceived a liberal (vs. conservative) politician to be. A simple slope analysis showed that, at high levels of social liberalism (+1 SD), participants perceived the liberal politician as more trustworthy than the conservative politician (b = 1.48, 95% CI = [0.80, 2.15], p < .001). However, at low levels of social liberalism (−1 SD), this effect was reversed (b = −1.01, 95% CI = [−1.68, −0.34], p = .003).

A similar analysis was performed in Studies 5, 6, and 7. The three-way interaction of trust × politician ideology × participant ideology on perceived trustworthiness was not significant in Study 5 (b = 0.13, 95% CI = [−0.22, 0.47], p = .479). However, with trust manipulation included only as a covariate, the two-way interaction politician ideology × participant ideology was significant (b = 0.18, 95% CI = [0.01, 0.35], p = .04). At high levels of social liberalism (+1 SD), the effect of manipulated ideology of the politician was significant (b = 0.87, 95% CI = [0.17, 1.56], p = .014), indicating that liberals perceived a liberal politician as more trustworthy than a conservative politician. At low levels of social liberalism (−1 SD), the effect of politician ideology on perceived trustworthiness was not significant (p = .647). In Study 6, the three-way interaction trust × politician ideology × participant ideology was significant (b = 0.38, 95% CI = [0.17, 0.59], p < .001). In the high trust condition, conservatives (−1 SD) perceived a liberal politician as less trustworthy than a conservative politician (b = −1.27, 95% CI = [−1.87, −0.68], p < .001). At high levels of social liberalism (+1 SD), the effect of politician ideology on perceived trustworthiness was reversed (b = 1.19, 95% CI = [0.60, 1.78], p < .001). In the low trust conditions, politician ideology was unrelated to perceived trustworthiness regardless of participant ideology, which mirrors the results obtained for empathy in this study. In Study 7, the three-way interaction, trust × politician ideology (liberal vs. conservative) × participant ideology, was also significant (b = 0.29, 95% CI = [0.06, 0.51], p = .014), and the obtained pattern largely mirrored the effects obtained in Study 6. Again, in the high trust condition, liberals perceived a liberal politician as more trustworthy (b = 1.14, 95% CI = [0.51, 1.78], p < .001) but conservatives perceived a liberal politician as less trustworthy (b = −1.51, 95% CI = [−2.17, −0.84], p < .001). In the control condition, participant ideology was unrelated to perceived trustworthiness. Overall, across four studies, we saw effects of ideological similarity on perceived trustworthiness, which were parallel to the effects on empathy.

General Discussion

In the present research, we manipulated information about trust and political ideology and examined the effects on empathy. We reasoned that empathy involves a type of emotional investment in another person and that people may limit their empathy to avoid potential vulnerability with, or further investment in, people they distrust (Righetti & Finkenauer, 2011). Moreover, past research suggests that political ideology can influence empathy, with some evidence for liberals being more empathic than conservatives (e.g., Iyer et al., 2012) and some evidence for selective empathy based on congruency in political ideology (Hasson et al., 2018). To test how both trust and political ideology might ultimately interact to inform empathy, we developed a new paradigm in which participants read about a hypothetical local politician who faces a significant penalty as a consequence of filling out his campaign finance paperwork incorrectly (Studies 1–6) or suffers an injury while volunteering (Study 7). This allowed us to systematically manipulate trust and ideology alone, and in combination, between subjects, to assess the impact on empathy.

In our first two studies, we developed a trust manipulation based on one used by Tetlock et al. (2013), who examined the effects of trust within a business context. We then documented that, regardless of their political ideology, participants showed much more empathy for a trustworthy politician (i.e., one who cares deeply about the problems of people in his district and is primarily concerned with improving his community) than for an untrustworthy politician (i.e., one who does not care about the problems of people in his district and is primarily concerned with his own power and status). Thus, we found support for the hypothesized relationship between trust and empathy; this relationship may exist because people avoid investing emotional energy in, or feeling social obligations toward, people with whom they feel at risk of exploitation. In contrast, these studies showed no support for a main effect of political ideology.

In our third and fourth studies, we manipulated the ideology of the politician and found that both liberals and conservatives expressed considerably more empathy for the politician who shared their ideology. This selective empathy effect is consistent with those observed in a recent study by Hasson et al. (2018) which showed that both liberals and conservatives have similar motivations to belong, which may in turn motivate them to have more empathy toward ingroup members. This finding is also consistent with a wider body of research suggesting that people tend to experience more empathy toward ingroup members than outgroup members (Cikara et al., 2011).

In Studies 5 through 7, we manipulated both the target’s trustworthiness and his political ideology to examine their interactive effects on empathy. We found that selective empathy based on ideological similarity was observed even when participants received individuating information indicating that the politician was trustworthy. Thus, it is not always the case that ideology is only relevant when direct information about trustworthiness is unavailable (see Tetlock et al., 2013).

In Study 7, we included a control condition for political ideology in which the target was neither described as liberal nor described as conservative. The results revealed that levels of empathy between the control and conservative conditions and between the control and the liberal conditions were not significantly different, meaning the control condition levels of empathy fell between that of the two polarized conditions. This suggests that both increased empathy for the ingroup and decreased empathy for the outgroup drive differences in empathy. Hence, there may be multiple forces influencing participants’ empathic reactions, including schadenfreude and in-group favoritism.

We also conducted additional analyses for several of our studies. Our analyses regarding perceived trustworthiness revealed that the effects of ideological similarity on perceived trustworthiness mirrored the effects of ideological similarity on empathy. This suggests that ideological similarity serves as a cue for inferring trustworthiness and seems to underscore the importance of perceived trust for empathy.

Findings from Studies 5 and 6 raise the possibility that the low trust condition negated the political ideology manipulation by suggesting that the politicians were not committed to their conservative or liberal beliefs. We were able to test this possibility in Study 7, in which we asked about the perceived ideology of the politician. We found that in both trust conditions, when the ideology was clearly indicated, participants had no issues with attributing the relevant (liberal or conservative) ideology to Joe. In assessing this possibility, we also observed an interesting effect in our control condition: When the ideology of the politician was not indicated, participants seemed to project their own ideology onto Joe in the high trust condition. Specifically, in this condition, liberals perceived Joe as liberal, but conservatives perceived him as conservative. This effect did not occur in the low trust condition. The details of these analyses are presented in SOM.

Across these studies, we did not find robust support for the previously found effect of liberals expressing more empathy than conservatives (Iyer et al., 2012; McCue & Gopoian, 2000). Instead, we found that both groups showed similar levels of empathy overall (Waytz et al., 2016). Hasson et al. (2018) noted that the liberal empathy advantage may result from asking participants about targets who are preferred by liberals. They tried to rule out this possibility by asking about liberal versus conservative targets, and they still found evidence of a liberal empathy advantage. Why is it that they found such an effect and we did not? One possible explanation is that their targets were described as protesters, and that protesters may generally be viewed more negatively by conservatives than liberals (see Haidt & Graham, 2007).

Study 7 showed results that were similar to Studies 5 and 6. This is noteworthy because in Study 7, the target was presented with a very different kind of problem than in the previous studies: a dog attack rather than having to suffer the consequences of a work-related mistake. Also, in Study 7, unlike the previous studies, the politician was consistently described as very genuine and as having taken all appropriate precautions. This suggests that the results are likely to be generalizable across a wide range of empathy-eliciting situations.

There are several limitations of the present research that need to be noted. One is that our initial trust manipulation, in particular, may have influenced inferences about other personal qualities, such as likeability, pro-sociality, or social dominance, that could affect empathy, which means that we could not isolate the effects of trust. Although we did use a different trust manipulation in Study 7 and found the same results as the previous studies, we cannot rule out these potential confounds as they were not directly measured. Another limitation is that all of the data were collected in the United States with an online Mechanical Turk sample, and it is not possible to determine how our results generalize to other populations. There are also measurement limitations, and future research could address the present research questions with additional measures of empathy, for example, in relation to how people respond to others’ expressions of pain (Decety & Jackson, 2006), and with the use of additional measures of trust, such as trust games (Johnson & Mislin, 2011). A final limitation is that although our manipulation and attention check questions helped ensure only individuals who took the task seriously and read carefully were included (see Chandler et al., 2020), this restriction could have led to some sampling bias.

There are still many unanswered questions about the relations among trust, ideology, and empathy. Some of these have to do with the implications of other forms of trust, such as generalized trust, which refers to the extent to which people are generally expected to act in a trustworthy manner. If generalized trust plays a role in informing empathic responses, it would suggest that experiences that make people feel secure in their surroundings and contribute to a positive view of human nature may lead to increased levels of empathy. The role of dispositional empathic concern, which is positively associated with level of affective polarization, should also be examined (Simas et al., 2020).

Taken together, our findings provide evidence that ideological similarity is more important for empathy than ideology itself and that perceived trustworthiness is closely linked to such selective empathy. Our results are also consistent with the broader conception of empathy as a type of emotional investment that people are wary of extending to others in the absence of trust. Moreover, our results suggest that including individuating information that indicates an individual is trustworthy does not necessarily override the effects of political congruence, at least in the current political climate of the United States.

Supplemental Material

sj-docx-1-psp-10.1177_0146167220972245 – Supplemental material for Trustworthiness and Ideological Similarity (But Not Ideology) Promote Empathy

Supplemental material, sj-docx-1-psp-10.1177_0146167220972245 for Trustworthiness and Ideological Similarity (But Not Ideology) Promote Empathy by Samantha M. Stevens, Carl P. Jago, Katarzyna Jasko and Gail D. Heyman in Personality and Social Psychology Bulletin

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Katarzyna Jasko’s work was supported by a Grant from the National Science Center (2015/17/D/HS6/00362).

ORCID iDs

Samantha M. Stevens

Gail D. Heyman

Supplemental Material

Supplemental material is available online with this article.

Notes

References

Bäckström

Björklund

(2007). Structural modeling of generalized prejudice: The role of social dominance, authoritarianism, and empathy. Journal of Individual Differences, 28, 10–17.

Bakker

B. N.

Schumacher

Gothreau

Arceneaux

(2020). Conservatives and liberals have similar physiological responses to threats. Nature Human Behaviour, 4, 613–621.

Batson

C. D.

(2010). Empathy-induced altruistic motivation. In Mikulincer

Shaver

P. R.

(Eds.), Prosocial motives, emotions, and behavior: The better angels of our nature (pp. 15–134). APA Books.

Batson

C. D.

Fultz

Schoenrade

P. A.

(1987). Distress and empathy: Two qualitatively distinct vicarious emotions with different motivational consequences. Journal of Personality, 55(1), 19–39.

Benabou

Tirole

(2006). Belief in a just world and redistributive politics. The Quarterly Journal of Economics, 121(2), 699–746.

Brandt

M. J.

Reyna

Chambers

J. R.

Crawford

J. T.

Wetherell

(2014). The ideological-conflict hypothesis: Intolerance among both liberals and conservatives. Current Directions in Psychological Science, 23(1), 27–34.

Chandler

Sisso

Shapiro

(2020). Participant carelessness and fraud: Consequences for clinical research and potential solutions. Journal of Abnormal Psychology, 129(1), 49–55.

Chiao

Mathur

Harada

Lipke

(2009). Neural basis of preference for human social hierarchy versus egalitarianism. Annals of the New York Academy of Sciences, 1167, 174–181.

Cikara

Bruneau

E. G.

Saxe

R. R.

(2011). Us and them: Intergroup failures of empathy. Current Directions in Psychological Science, 20, 149–153.

10.

Cohen

(1983). The cost of dichotomization. Applied Psychological Measurement, 7(3), 249–253.

11.

Combs

D. J. Y.

Powell

C. A. J.

Schurtz

D. R.

Smith

R. H.

(2009). Politics, schadenfreude, and ingroup identification: The sometimes happy thing about a poor economy and death. Journal of Experimental Social Psychology, 45, 635–646.

12.

Crawford

J. T.

Brandt

M. J.

(2020). Ideological (a) symmetries in prejudice and intergroup bias. Current Opinion in Behavioral Sciences, 34, 40–45.

13.

Davis

M. H.

(1983). Measuring individual differences in empathy: Evidence for a multidimensional approach. Journal of Personality and Social Psychology, 44, 113–126.

14.

Decety

Jackson

P. L.

(2006). A social-neuroscience perspective on empathy. Current Directions in Psychological Science, 15, 54–58.

15.

Doherty

(2017, October 5). Key takeaways on Americans’ growing partisan divide over political values. Pew Research Center. https://www.pewresearch.org/fact-tank/2017/10/05/takeaways-on-americans-growing-partisan-divide-over-political-values/

16.

Dovidio

J. F.

Allen

J. L.

Schroeder

D. A.

(1990). Specificity of empathy-induced helping: Evidence for altruistic motivation. Journal of Personality and Social Psychology, 59, 249–260.

17.

Eisenberg

Miller

P. A.

(1987). The relation of empathy to prosocial and related behaviors. Psychological Bulletin, 101, 91–119.

18.

Frimer

J. A.

Skitka

L. J.

Motyl

(2017). Liberals and conservatives are similarly motivated to avoid exposure to one another’s opinions. Journal of Experimental Social Psychology, 72, 1–12.

19.

Goldenberg

Halperin

van Zomeren

Gross

J. J.

(2016). The process model of group-based emotion: Integrating intergroup emotion and emotion regulation perspectives. Personality and Social Psychology Review, 20(2), 118–141.

20.

Gramlich

(2017, December 19). Far more Americans say there are strong conflicts between partisans than between other groups in society. Pew Research Center. https://www.pewresearch.org/fact-tank/2017/12/19/far-more-americans-say-there-are-strong-conflicts-between-partisans-than-between-other-groups-in-society/

21.

Haidt

Graham

(2007). When morality opposes justice: Conservatives have moral intuitions that liberals may not recognize. Social Justice Research, 20, 98–116.

22.

Hasson

Tamir

Brahms

K. S.

Cohrs

J. C.

Halperin

(2018). Are liberals and conservatives equally motivated to feel empathy toward others? Personality and Social Psychology Bulletin, 44, 1449–1459.

23.

Hoffman

M. L.

(2008). Empathy and prosocial behavior. In Lewis

Haviland-Jones

J. M.

Barrett

L. F.

(Eds.), Handbook of emotions (pp. 440–455). Guilford Press.

24.

Iyer

Koleva

Graham

Ditto

Haidt

(2012). Understanding libertarian morality: The psychological dispositions of self-identified libertarians. PLOS ONE, 7(8), Article e42366.

25.

Johnson

N. D.

Mislin

A. A.

(2011). Trust games: A meta-analysis. Journal of Economic Psychology, 32, 865–889.

26.

Jost

J. T.

Federico

C. M.

Napier

J. L.

(2009). Political ideology: Its structure, functions, and elective affinities. Annual Review of Psychology, 60, 307–337.

27.

Kahan

Peters

Dawson

Slovic

(2017). Motivated numeracy and enlightened self-government. Behavioural Public Policy, 1, 54–86.

28.

MacCallum

R. C.

Zhang

Preacher

K. J.

Rucker

D. D.

(2002). On the practice of dichotomization of quantitative variables. Psychological Methods, 7(1), 19–40.

29.

McAdams

D. P.

Albaugh

Farber

Daniels

Logan

R. L.

Olson

(2008). Family metaphors and moral intuitions: How conservatives and liberals narrate their lives. Journal of Personality and Social Psychology, 95, 978–990.

30.

McCue

C. P.

Gopoian

J. D.

(2000). Dispositional empathy and the political gender gap. Women & Politics, 21, 1–20.

31.

McFarland

(2010). Authoritarianism, social dominance, and other roots of generalized prejudice. Political Psychology, 31, 453–477.

32.

Righetti

Finkenauer

(2011). If you are able to control yourself, I will trust you: The role of perceived self-control in interpersonal trust. Journal of Personality and Social Psychology, 100, 874–886.

33.

Royston

Altman

D. G.

Sauerbrei

(2006). Dichotomizing continuous predictors in multiple regression: A bad idea. Statistics in Medicine, 25(1), 127–141.

34.

Simas

Clifford

Kirkland

(2020). How empathic concern fuels political polarization. American Political Science Review, 114, 258–269.

35.

Tetlock

P. E.

Vieider

F. M.

Patil

S. V.

Grant

A. M.

(2013). Accountability and ideology: When left looks right and right looks left. Organizational Behavior and Human Decision Processes, 122, 22–35.

36.

Toi

Batson

C. D.

(1982). More evidence that empathy is a source of altruistic motivation. Journal of Personality and Social Psychology, 43, 281–292.

37.

Washburn

A. N.

Skitka

L. J.

(2018). Science denial across the political divide: Liberals and conservatives are similarly motivated to deny attitude-inconsistent science. Social Psychological and Personality Science, 9, 972–980.

38.

Waytz

Iyer

Young

Graham

(2016). Ideological differences in the expanse of empathy. In Valdesolo

Graham

(Eds.), Social psychology of political polarization (pp. 61–77). Routledge.

39.

Zmigrod

Rentfrow

P. J.

Robbins

T. W.

(2020). The partisan mind: Is extreme political partisanship related to cognitive inflexibility? Journal of Experimental Psychology: General, 149, 407–418.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.85 MB