Reconstruing Intolerance

Abstract

Myrdal (1944) described the “American dilemma” as the conflict between abstract national values (“liberty and justice for all”) and more concrete, everyday prejudices. We leveraged construal-level theory to empirically test Myrdal’s proposition that construal level (abstract vs. concrete) can influence prejudice. We measured individual differences in construal level (Study 1) and manipulated construal level (Studies 2 and 3); across these three studies, we found that adopting an abstract mind-set heightened conservatives’ tolerance for groups that are perceived as deviating from Judeo-Christian values (gay men, lesbians, Muslims, and atheists). Among participants who adopted a concrete mind-set, conservatives were less tolerant of these nonnormative groups than liberals were, but political orientation did not have a reliable effect on tolerance among participants who adopted an abstract mind-set. Attitudes toward racial out-groups and dominant groups (e.g., Whites, Christians) were unaffected by construal level. In Study 3, we found that the effect of abstract thinking on prejudice was mediated by an increase in concerns about fairness.

Keywords

prejudice social cognition personal values

The “American Dilemma” is . . . the ever-raging conflict between, on the one hand, the valuations preserved on the general plane which we shall call the “American Creed,” where the American thinks, talks, and acts under the influence of high national and Christian precepts, and, on the other hand, the valuations on specific planes of individual and group living, where . . . consideration of community prestige and conformity; group prejudice against particular persons or types of people; and all sorts of miscellaneous wants, impulses, and habits dominate his outlook.

—Myrdal (1944, p. xliii)

In his influential study of American race relations in the 1940s, Swedish economist Gunnar Myrdal identified a fundamental “American dilemma”—a conflict between two planes of existence in American society at that time. On the general, more abstract plane, the American Creed of fairness and equality was promoted and cherished. On the more concrete, day-to-day plane, however, many individuals in the 1940s overtly expressed biases and prejudice that conflicted with these abstract values.

Overt expressions of bias toward racial minorities are no longer tolerated as they were during the time of Myrdal’s writings (Schuman, Steeh, Bobo, & Krysan, 1997), which perhaps has resolved, or at least diminished, the conflict between the societal treatment of racial out-groups and the abstract value of fairness. However, this conflict likely remains in many Americans’ attitudes toward certain individuals, such as gay men and lesbians and members of religious out-groups (e.g., atheists and Muslims), who are perceived as being nonnormative, or deviating from Judeo-Christian values, and thus are often the targets of overt discrimination (Edgell, Gerteis, & Hartmann, 2006; Hebl, Foster, Mannix, & Dovidio, 2002; Herek, 2000). This conflict may be especially pronounced among political conservatives, who advocate for Judeo-Christian values to have public and national precedence (Republican National Committee, n.d.). We investigated how abstract and concrete mind-sets can differentially affect concerns about fairness and thereby influence prejudice toward members of nonnormative groups (specifically, gay men, lesbians, Muslims, and atheists) among political conservatives and liberals.

Members of nonnormative groups in the United States commonly face challenges—particularly from politically conservative people—to achieving equal rights and privileges. For example, gay men and lesbians are currently denied the right to marry in most states and face overt discrimination from employers, politicians, and religious leaders, especially those who are politically conservative (Herek, 2000). More than half of the respondents to a survey of a nationally representative sample (Nisbet & Shanahan, 2004) thought that the rights of Muslim Americans should be restricted, a trend that was largely driven by politically conservative and highly religious respondents. Public-opinion polls have also revealed an ideological divide with respect to the acceptance of atheists: In one recent poll, only 14% of Republicans (vs. 44% of Democrats) said they would be willing to vote for a well-qualified, party-nominated presidential candidate who was an atheist (Pew Research Center, 2007).

We propose that the discrepancy between the abstract value of fairness and a bias against certain nonnormative groups, a conflict that is more pronounced among political conservatives than among liberals, may be moderated by the mind-set that people adopt when thinking about these groups. In the present research, we used construal-level theory to examine how two different mind-sets (or “planes”)—abstract and concrete—might influence conservatives’ feelings toward nonnormative groups. A large body of research has shown that people can perceive objects, events, and individuals in either concrete (low-level) or abstract (high-level) terms (for a review, see Trope & Liberman, 2010; see also Vallacher & Wegner, 1989). Construal level has a strong influence on people’s judgments, attitudes, and behaviors, from feature perceptions and morality judgments to self-control and social perceptions (Eyal, Liberman, & Trope, 2008; Fujita, Trope, Liberman, & Levin-Sagi, 2006; Trope & Liberman, 2000).

In the research reported here, we investigated whether construal level can influence perceptions and attitudes toward not only objects, events, and individuals, but also groups (see also Levy, Freitas, & Salovey, 2002). We hypothesized that abstract thinking, insofar as it is related to Myrdal’s (1944) “general plane,” should lead to a reduction of prejudice, particularly prejudice toward nonnormative out-groups, because people whose construal level is more abstract should be more likely to operate under the broad societal values of fairness and justice. This hypothesis is consistent with work by Eyal et al. (2008), who found that participants who thought abstractly rather than concretely were more likely to apply their moral principles in judgments of others’ actions. Additionally, Torelli and Kaikati (2009) demonstrated that values were a stronger predictor of judgments and behaviors when people were thinking more abstractly.

In three experiments, we investigated whether thinking abstractly (vs. concretely) can increase positive feelings toward nonnormative groups (gay men, lesbians, Muslims, and atheists)—groups that experience overt prejudice that is antithetical to the value of fairness, or the “American Creed.” Moreover, because more conservative individuals show greater explicit bias toward nonnormative groups and thus presumably experience greater conflict between their concrete feelings about members of these groups and the more abstract principles of equality and fairness, we hypothesized that the predicted effect of abstract thinking on bias against nonnormative groups would be stronger for conservatives than for liberals. In our first two studies, we examined how construal level—either characteristic (Study 1) or induced (Study 2)—related to both conservatives’ and liberals’ explicit feelings toward a variety of social groups. In Study 3, we manipulated participants’ mind-sets and tested whether the societal value of fairness is indeed a mediator of the effects of construal level on bias.

Study 1

In Study 1, we investigated whether individual differences in construal level (abstract vs. concrete), assessed with Vallacher and Wegner’s (1989) Behavioral Identification Form, were related to differences in prejudice toward nonnormative social groups. We measured feelings toward different social groups using feeling thermometers, which have been shown to be reliable and precise measures of feelings toward various groups (Alwin, 1997). Our focus was on participants’ feelings toward four nonnormative groups (gay men, lesbians, Muslims, and atheists), but we also assessed feelings toward racial- ethnic minority groups (Blacks, Latinos) and dominant groups (Whites, Christians). We predicted that more politically conservative participants would display more negative feelings toward the nonnormative groups, which would be in line with results from prior research (Nosek, Banaji, & Jost, 2009), but that this effect would be moderated by individual differences in construal level. Specifically, we expected that conservatives would have less negative feelings toward nonnormative groups if they characteristically adopted a more abstract mind-set. We expected that mind-set would not be associated with feelings toward racial-ethnic minority groups (because they are legally guaranteed equal rights and because the American Creed is more commonly perceived to apply to them than to nonnormative groups) or toward dominant groups. Because liberals tend to support equal rights for nonnormative groups and thus should not experience a conflict between their abstract values and their feelings toward these groups, we did not expect mind-set to be related to their responses.

Method

Participants

Sixty-three participants (44.4% male, 55.6% female; mean age = 35.73 years, SD = 13.58) were recruited online and took part in this study in exchange for a chance to win a gift certificate.

Procedure

We assessed participants’ mind-sets using Vallacher and Wegner’s (1989) Behavioral Identification Form, which asks participants to make dichotomous choices between descriptions (concrete vs. abstract) of actions. Participants were given a list of 10 different actions (e.g., “pushing a doorbell”) and asked to indicate whether each action was best defined by a concrete description (e.g., “moving a finger”) or an abstract description (“seeing if someone is home”). For each participant, we used the proportion of actions described as abstract (vs. concrete) as our measure of mind-set (M = .55, SD = .26; Cronbach’s α = .71).

Participants then rated eight groups, using two feeling thermometers (from 0 to 100), one assessing warmth and the other assessing liking. The warmth and liking ratings for each group were highly correlated, rs > .82; we therefore averaged the two ratings to create a score for feelings toward each group (see Table 1). Exploratory factor analysis using varimax rotation confirmed that these groups fell into three categories that accounted for 84% of the variance: nonnormative groups (lesbians, atheists, gay men, and Muslims; factor loadings of .88, .87, .79, and .67, respectively), racial-ethnic minority groups (Blacks and Latinos; factor loadings of .86 and .81, respectively), and dominant groups (Whites and Christians; factor loadings of .91 and .82, respectively). Finally, participants responded to demographic questions and rated their political orientation on a scale from 1 (very liberal) to 6 (very conservative; see Jost, Federico, & Napier, 2009; see Table 1).

Table 1.

Mean Political-Orientation and Feeling-Thermometer Ratings for Studies 1, 2, and 3

Variable	Study 1 (N = 63)	Study 2 (N = 64)	Study 3 (N = 168)
Political orientation	3.25 (1.39)	2.83 (1.44)	2.60 (1.29)
Feelings toward nonnormative groups
Gay men	57.40 (25.74)	59.45 (25.67)	63.23 (25.96)
Lesbians	58.33 (24.44)	60.03 (23.80)	62.34 (25.28)
Muslims	46.79 (27.46)	50.88 (24.53)	54.78 (24.82)
Atheists	59.18 (26.57)	56.76 (24.35)	61.78 (25.59)
Feelings toward racial-ethnic minorities
Blacks	59.23 (21.56)	62.41 (18.17)	60.48 (22.24)
Latinos	61.04 (22.11)	63.34 (20.46)	61.98 (21.53)
Feelings toward dominant groups
Whites	70.60 (19.63)	68.61 (17.75)	70.57 (20.03)
Christians	63.93 (26.79)	57.54 (27.78)	61.83 (25.39)

Note: Political orientation was rated on a 6-point scale, with higher values indicating more conservative political orientations. Participants rated their warmth and liking toward each group, and these ratings were averaged to create a score for feelings toward each group; higher values indicate more positive feelings. Standard deviations are shown in parentheses.

Results and discussion

To test our main predictions, we conducted three linear regressions predicting feelings toward (a) nonnormative groups, (b) racial-ethnic minority groups, and (c) dominant groups; abstract mind-set (centered), political orientation (centered), and the interaction of political orientation and mind-set were entered as independent variables. The model predicting feelings toward nonnormative groups yielded the predicted main effect of political orientation; participants who were more liberal expressed more positive feelings toward these groups, b = −7.53, SE = 2.04, p < .001. This effect was qualified by the predicted Mind-Set × Political Orientation interaction, b = 16.88, SE = 6.89, p = .02 (see Fig. 1).

Fig. 1.

Results from Study 1: positive feelings toward nonnormative groups as a function of mind-set and political orientation. Positive feelings were rated using 100-point feeling-thermometer scales. For this analysis, “liberal” and “conservative” refer to political-orientation scores 1 standard deviation below the mean and 1 standard deviation above the mean, respectively; “concrete” and “abstract” mind-sets refer to scores 1 standard deviation below the mean and 1 standard deviation above the mean, respectively, on the Behavioral Identification Form (Vallacher & Wegner, 1989).

Simple-slopes analyses (Aiken & West, 1991) showed that liberals (1 SD below the mean political-orientation score) exhibited the same (high) level of positive feelings toward nonnormative groups regardless of construal level, b = −11.45, SE = 15.48, p = .46 (see Fig. 1). However, conservatives (1 SD above the mean political-orientation score) who had more abstract mind-sets displayed more positive feelings toward these nonnormative groups, b = 35.49, SE = 13.44, p = .01. Among concrete thinkers (1 SD below the mean Behavioral Identification Form score), there was a significant relationship between political orientation and feelings toward nonnormative groups; liberals reported more positive feelings toward these groups than conservatives did, b = −11.92, SE = 2.99, p < .001. Among abstract thinkers (1 SD above the mean Behavioral Identification Form score), however, this relationship was not significant, b = −3.14, SE = 2.63, p = .24.

Analyses of feelings toward racial-ethnic minority groups revealed only a marginally significant effect of political orientation; more conservative participants had less positive feelings toward racial-ethnic minorities, b = −3.59, SE = 2.04, p = .08. The Mind-Set × Political Orientation interaction was not significant, p = .53. Also, more conservative participants felt more positively toward dominant groups, b = 4.64, SE = 1.84, p = .02. Again, the Mind-Set × Political Orientation interaction was not significant, p = .35. Thus, as expected, construal level affected only feelings toward the nonnormative groups.

Although, on average, the nonnormative groups were rated somewhat more negatively than the racial-ethnic minority groups were (a result consistent with findings from surveys of representative samples), as predicted, mind-set exclusively moderated feelings toward nonnormative groups, rather than reducing negative feelings toward out-groups in general. That is, feeling-thermometer ratings for gay men and lesbians were comparable to those for Blacks and Latinos (see Table 1), but participants’ feelings about gay men and lesbians displayed the expected Mind-Set × Political Orientation interaction (ps < .01), whereas their feelings about Blacks (p = .44) and Latinos (p = .43) did not.

The results from Study 1 are thus in line with our hypothesis that conservatives’ prejudice against nonnormative groups is reduced when they think abstractly as opposed to concretely. Indeed, there was no difference in bias at all between liberals and conservatives with abstract mind-sets. It does not seem to be the case that liberals are chronically more likely to think in abstract terms and that this accounts for their lower levels of prejudice. In fact, there was a relatively weak but reliable correlation between mind-set and political orientation, such that more conservative participants tended to endorse more abstract descriptions of actions on the Behavioral Identification Form, r(60) = .27, p = .03. In our two next studies, we sought to replicate our results from Study 1 using established experimental manipulations of construal level.

Study 2

In Study 2, we induced abstract or concrete mind-sets via a why/how paradigm in which participants must give increasingly subordinate (concrete) or superordinate (abstract) reasons for engaging in a certain behavior (Freitas, Gollwitzer, & Trope, 2004; Fujita et al., 2006; Ledgerwood, Trope, & Chaiken, 2010). Participants were asked to think about the issue of maintaining good physical health and to explain either why they would do so (abstract construal) or how they would do so (concrete construal). We assessed participants’ political orientation and their feelings about the same nonnormative, racial-ethnic, and dominant groups that were used in Study 1. We predicted that more conservative participants would display less positive feelings toward nonnormative groups, but that this effect would be less pronounced in the abstract- construal condition than in the concrete-construal condition.

Method

Participants

Sixty-four participants (46.9% male, 53.1% female; mean age = 34.72 years, SD = 13.32) were recruited online and took part in this study in exchange for a chance to win a gift certificate.

Procedure

Participants reported their political orientation (see Table 1) and were then randomly assigned to construal condition and asked to fill out a ladder questionnaire about good physical health. In the abstract condition, participants started at the bottom of the ladder and moved up, generating increasingly superordinate answers to the question of why they would maintain good physical health; in the concrete condition, they moved down the ladder, generating increasingly subordinate answers to the question of how they would maintain good physical health (Freitas et al., 2004; Fujita et al., 2006). Following this manipulation, participants used feeling thermometers to rate their feelings of warmth and liking toward the same eight groups used in Study 1 (see Table 1). The warmth and liking ratings for each group were highly correlated, rs > .79, so we again averaged them to create measures of positive feelings toward nonnormative groups, racial-ethnic minorities, and dominant groups (see Table 1).

Results and discussion

Results were consistent with our findings from Study 1. Specifically, we found the expected main effect of political orientation on feelings toward nonnormative groups, such that conservative participants showed less positive feelings toward these groups than liberal participants did, b = −11.06, SE = 2.47, p < .001. This effect was qualified by the predicted interaction of political orientation and mind-set, b = 7.18, SE = 3.41, p = .04 (see Fig. 2).

Fig. 2.

Results from Study 2: positive feelings toward nonnormative groups as a function of construal level and political orientation. Positive feelings were rated using 100-point feeling-thermometer scales. Liberals and conservatives had political-orientation scores 1 standard deviation below the mean and 1 standard deviation above the mean, respectively.

Simple-slopes analyses revealed that in the concrete condition, political orientation was significantly related to feelings toward the nonnormative groups, such that more conservative participants showed less positive feelings toward these groups, b = −11.06, SE = 2.47, p < .001. In the abstract condition, this difference was significantly reduced and no longer reliable, b = −3.88, SE = 2.36, p = .11. Examined another way, results showed that conservatives were significantly more positive toward the nonnormative groups when thinking abstractly than when thinking concretely, b = 14.74, SE = 6.97, p = .04, whereas liberals expressed the same level of positive feelings in the two conditions, b = −5.95, SE = 6.89, p = .39.

As in Study 1, there were no effects of construal on feelings toward the racial-ethnic minority groups or dominant groups. More conservative participants reported somewhat less positive feelings toward racial-ethnic minorities, b = −3.39, SE = 2.37, p = .16, and more positive feelings toward dominant groups, b = 3.71, SE = 1.95, p = .06.

Results from our first two studies offer convergent support for the hypothesis that for politically conservative individuals, thinking with an abstract mind-set rather than a concrete mind-set can reduce expressed prejudice toward people who are viewed as somehow “deviant” from prototypical Americans. In both studies, there was no effect of construal level on feelings toward racial-ethnic minorities or feelings toward dominant groups. This finding is consistent with our conjecture that the endorsement of overt prejudice and discrimination toward nonnormative groups creates a dilemma by conflicting with the American ideal of promoting fairness and equality. In our final study, we examined the process underlying the effects that emerged in our first two studies by manipulating construal level and assessing concerns about fairness as well as feelings toward nonnormative groups. We hypothesized that abstract thinking would bring the value of fairness to the forefront of participants’ minds and thereby reduce prejudice, particularly prejudice toward groups that are perceived as deviant and that are not consistently included in the American Creed of fairness for all.

Study 3

Past research has demonstrated that emphasizing moral ideals such as fairness can serve as a means of improving intergroup relations (Does, Derks, & Ellemers, 2011). We hypothesized that the effect of mind-set on expressed prejudice that we found in our first two studies was due to a shift in the salience of central values. We reasoned that when thinking on an abstract (as opposed to concrete) level, people should be more likely to rely on broad-based moral principles such as fairness (Graham, Haidt, & Nosek, 2009). In Study 3, we tested this hypothesis by examining how construal level influenced participants’ concerns about fairness and whether shifts in the salience of values accounted for (i.e., mediated) changes in feelings toward nonnormative groups.

Method

Participants

One hundred sixty-eight participants (36.9% male, 63.1% female; mean age = 35.95 years, SD = 13.04) were recruited online and took part in this study in exchange for a chance to win a gift certificate.

Procedure

Participants completed a construal-level priming manipulation (developed by Fujita et al., 2006) in which they were randomly presented with 20 words, 5 at a time. Participants assigned to the concrete condition were asked to generate a subordinate exemplar for each word by answering the question “An example of ______ is what?” They were told to fill in the blank with each of the words presented and then answer the question for that word. For example, if one of the words presented was dog, participants could answer “poodle” (a type of dog) or even “Odie” (a specific name for a dog). Participants in the abstract condition were asked to generate a superordinate category label for each word by answering the question “_____ is an example of what?” (again, filling in the blank with each of the words presented). For example, if one of the words presented was dog, participants could answer “pet” or “animal.”

Participants next responded to four items from the Moral Foundations Questionnaire (Graham et al., 2009) that assessed concerns about fairness (e.g., “Justice is the most important requirement for a society”; α = .75). Participants used feeling thermometers to rate their feelings of warmth and liking toward the nonnormative groups, racial-ethnic minorities, and dominant groups; as in the first two studies, we averaged these ratings to calculate a composite score for positive feelings (see Table 1). Finally, participants answered a variety of demographic questions, including an item measuring political orientation (see Table 1).

Results

As in the first two studies, there was a main effect of political orientation, such that more conservative participants showed less positive feelings toward the nonnormative groups, b = −9.27, SE = 1.37, p < .001. This effect was qualified by a marginally significant Mind-Set × Political Orientation interaction, b = 3.45, SE = 1.92, p = .07. Political orientation was a stronger predictor of positive feelings in the concrete condition, b = −9.27, SE = 1.37, p < .001, than in the abstract condition, b = −5.81, SE = 1.35, p < .001. Conservatives’ feelings toward the nonnormative groups were significantly more positive in the abstract condition than in the concrete condition, b = 10.47, SE = 4.12, p = .01. There was no effect of construal level on liberals’ feelings toward these groups, b = 0.04, SE = 4.02, p = .99.

We next tested whether differences in the salience of the moral value of fairness mediated the relationship between construal level and positive feelings. We used Preacher and Hayes’s (2008) bootstrapping macro, with construal condition as the independent variable and fairness as a mediator; as covariates, we included political orientation centered at 1 standard deviation above the sample mean and the interactions of the political-orientation variable with construal condition and fairness. Thus, we evaluated whether fairness mediated the relationship between construal condition and positive feelings for conservative participants.

Results from the bootstrapping procedure (shown in Fig. 3) supported our predictions: Among conservatives, abstract (as opposed to concrete) thinking significantly increased endorsement of fairness, which had a significant direct effect on feelings toward the nonnormative groups. The indirect effect of construal level on positive feelings through fairness was significant, b = 4.62, SE = 2.48, p < .05 (bias-corrected confidence interval: 1.05–11.19).

Fig. 3.

Results from a multiple mediation model of fairness as a mediator of the effect of construal level on conservatives’ (1 SD above the mean score for political orientation) positive feelings toward nonnormative groups in Study 3. Standard errors are shown in parentheses. Asterisks indicate significant coefficients (*p < .01, **p < .001).

As in Study 1 and Study 2, more conservative participants tended to express more negative attitudes toward racial-ethnic minorities, b = −2.83, SE = 1.55, p = .07, and more positive attitudes toward dominant groups, b = 2.07, SE = 1.43, p = .15.

General Discussion

Results from our three studies indicate that straightforward interventions aimed at changing people’s mind-sets may be effective for improving their attitudes, at least temporarily, toward highly stigmatized social groups. Theoretically, our findings advance construal-level theory by demonstrating that manipulating construal level can influence attitudes toward groups, as well as attitudes toward objects, events, and individuals. Levy et al. (2002) found that the tendency to identify actions abstractly was correlated with empathy for and willingness to help other groups. However, we believe that our studies are the first to illustrate that varying people’s mind-sets can influence intergroup attitudes. At least for conservatives, adopting an abstract construal level reduces prejudice toward nonnormative groups.

Our work also has implications for understanding the relation between political orientation and prejudice. Whereas much prior research has focused on various factors (e.g., threat and cognitive load) that tend to make liberals more likely to endorse conservative attitudes, opinions, policies, and leaders (Jost et al., 2009; Skitka, Mullen, Griffin, Hutchinson, & Chamberlin, 2002), work investigating factors that may increase tolerance among conservatives—the focus of the present research—is rarer. Our findings suggest that at least some of conservatives’ prejudice is in conflict with their abstract values and therefore can be reduced when those values are brought to the forefront.

Ledgerwood et al. (2010, Study 3) showed that participants were more likely to vote for a policy that was in accord with their premeasured ideology when the policy was to be implemented in the distant future (and thus was construed abstractly) than when it was to be implemented in the near future (and thus was construed concretely). The researchers noted that thinking abstractly about an attitude object leads to judgments that are “shaped by what is consistently relevant . . . across different situations, including broad principles and values” (p. 34). In conjunction with the current work, this research seems to suggest that abstract thinking can reduce partisan differences insofar as everyone—conservatives and liberals alike—cares about fairness on some level. Thus, although many Americans may react to gay men and lesbians with disgust (Inbar, Pizarro, Knobe, & Bloom, 2009) or view atheists and Muslims as socially, or even physically, threatening (Edgell et al., 2006), enduring concerns about justice and fairness can perhaps mitigate discriminatory responses toward members of these groups.

Our findings also suggest promising directions for future research. First, in our studies, we used only explicit measures of prejudice and thus could not determine whether construal level had an impact on implicit attitudes. Indeed, it is conceivable that it does not: We found that construal level affected prejudice toward nonnormative groups only, and such prejudice tends to be more explicit and overt than other types of prejudices (Edgell et al., 2006).

In addition to the distinction between implicit and explicit attitudes, another aspect of intergroup bias is the endorsement of stereotypes. In fact, research has suggested that prejudice (the focus of the studies reported here) and stereotyping are only modestly related (Dovidio, Brigham, Johnson, & Gaertner, 1996). Whereas our studies suggest that construal level could be used as a tool to reduce intergroup bias, recent research by McCrea, Wieber, and Myers (2012) has shown that abstract construal increases the endorsement of stereotypes toward a variety of social groups. Future research should aim to further illuminate the differing effects of construal level on prejudice and stereotyping and the consequent impact on intergroup relations.

Another question that arises is whether abstract thinking specifically reduces prejudice against “deviant” groups or can more generally reduce bias against any group whose societal treatment is overtly at odds with concerns about fairness. We believe that our results from Study 3 are more consistent with the latter possibility, given that concerns about fairness were found to mediate the effect of construal level on attitudes toward nonnormative groups. For instance, we would not expect abstract thinking to improve attitudes toward people such as child molesters or perpetrators of heinous crimes because negative attitudes toward these groups are generally perceived as being fair and legitimate. Additionally, it is conceivable that abstract thinking would not necessarily reduce prejudice among members of societies in which the values of fairness and equality are not a prominent part of the national discourse. In other words, construal level may not produce more favorable intergroup attitudes if there is no abstract societal value that conflicts with concrete prejudices. This suggestion is broadly consistent with our finding that construal level did not affect feelings toward racial-ethnic minority groups, despite the fact that there were no large qualitative differences in participants’ ratings of warmth and liking for racial-ethnic and nonnormative groups. It is plausible that because racial bias tends to be manifested in a more subtle way than bias against nonnormative groups is, and because racial bias is overtly condemned on a societal level, people do not experience any conflict between their values and their feelings toward racial minorities. In other words, abstract thinking is unlikely to influence feelings toward racial minorities among people who believe that racial minorities are treated fairly in society, because these people will not perceive a dilemma between their values of fairness and the treatment of racial minorities.

Thus, the effects of construal level may be critically moderated by contextual factors. Abstract construal may promote positive intergroup relations when it focuses attention on specific societal values emphasizing fairness or on inclusive social connections (e.g., general human connectedness) rather than on immediate, differing allegiances (e.g., different national identities; Gaertner & Dovidio, 2000; Wohl & Branscombe, 2005). However, abstract construal may exacerbate bias when it draws attention to beliefs or principles that form the basis of conflict (e.g., thinking of land as holy land). It could also perpetuate existing biases. Research on the linguistic intergroup bias (Maass, Salvi, Arcuri, & Semin, 1989) has revealed that people tend to describe positive behaviors of in-group members and negative behaviors of out-group members abstractly, which implies that these abstract qualities are perceived as central and stable. For example, a description of an in-group member might be “she is honest” rather than “she returned the wallet,” whereas a description of an out-group member might be “he is irresponsible” rather than “he forgot the paperwork.”

In conclusion, our results from three studies provide converging evidence that adopting an abstract mind-set (as opposed to a concrete mind-set) can reduce expressions of prejudice toward nonnormative groups, primarily among people who are politically conservative. Study 3 directly demonstrated that the influence of abstract construal on bias is mediated by an increase in the salience of concerns about fairness. Overall, this research brings construal-level theory to bear on the investigation of intergroup relations and opens several avenues for future endeavors to both understand and reduce prejudice.

Footnotes

Declaration of Conflicting Interests

The authors declared that they had no conflicts of interest with respect to their authorship or the publication of this article.

References

Aiken

L. S.

West

S. G.

(1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage.

Alwin

D. F.

(1997). Feeling thermometers versus 7-point scales: Which are better? Sociological Methods & Research, 25, 318–340.

Does

Derks

Ellemers

(2011). Thou shalt not discriminate: How emphasizing moral ideals rather than obligations increases Whites’ support for social equality. Journal of Experimental Social Psychology, 47, 562–571.

Dovidio

J. F.

Brigham

J. C.

Johnson

B. T.

Gaertner

S. L.

(1996). Stereotyping, prejudice, and discrimination: Another look. In Macrae

C. N.

Stangor

Hewstone

(Eds.), Stereotypes and stereotyping (pp. 276–319). New York, NY: Guilford Press.

Edgell

Gerteis

Hartmann

(2006). Atheists as “other”: Moral boundaries and cultural membership in American society. American Sociological Review, 71, 211–234.

Eyal

Liberman

Trope

(2008). Judging near and distant virtue and vice. Journal of Experimental Social Psychology, 44, 1204–1209.

Freitas

A. L.

Gollwitzer

P. M.

Trope

(2004). The influence of abstract and concrete mindsets on anticipating and guiding others’ self-regulatory efforts. Journal of Experimental Social Psychology, 40, 739–752.

Fujita

Trope

Liberman

Levin-Sagi

(2006). Construal levels and self-control. Journal of Personality and Social Psychology, 90, 351–367.

Gaertner

S. L.

Dovidio

J. F.

(2000). Reducing intergroup bias: The common ingroup identity model. Philadelphia, PA: Psychology Press.

10.

Graham

Haidt

Nosek

B. A.

(2009). Liberals and conservatives rely on different sets of moral foundations. Journal of Personality and Social Psychology, 96, 1029–1046.

11.

Hebl

M. R.

Foster

J. B.

Mannix

L. M.

Dovidio

J. F.

(2002). Formal and interpersonal discrimination: A field study of bias toward homosexual applicants. Personality and Social Psychology Bulletin, 28, 815–825.

12.

Herek

G. M.

(2000). The psychology of sexual prejudice. Current Directions in Psychological Science, 9, 19–22.

13.

Inbar

Pizarro

D. A.

Knobe

Bloom

(2009). Disgust sensitivity predicts intuitive disapproval of gays. Emotion, 9, 435–439.

14.

Jost

J. T.

Federico

C. M.

Napier

J. L.

(2009). Political ideology: Its structure, functions, and elective affinities. Annual Review of Psychology, 60, 307–338.

15.

Ledgerwood

Trope

Chaiken

(2010). Flexibility now, consistency later: Psychological distance and construal shape evaluative responding. Journal of Personality and Social Psychology, 99, 32–51.

16.

Levy

S. R.

Freitas

A. L.

Salovey

(2002). Construing action abstractly and blurring social distinctions: Implications for perceiving homogeneity among, but also empathizing with and helping, others. Journal of Personality and Social Psychology, 83, 1224–1238.

17.

Maass

Salvi

Arcuri

Semin

G. R.

(1989). Language use in intergroup contexts: The linguistic intergroup bias. Journal of Personality and Social Psychology, 57, 981–993.

18.

McCrea

S. M.

Wieber

Myers

A. L.

(2012). Construal level mind-sets moderate self- and social stereotyping. Journal of Personality and Social Psychology, 102, 51–68.

19.

Myrdal

(1944). An American dilemma: The Negro problem and modern democracy. New York, NY: Harper.

20.

Nisbet

E. C.

Shanahan

(2004). Restrictions on civil liberties, views of Islam, & Muslim Americans [Special Report of the Media & Society Research Group, Cornell University]. Retrieved from www.yuricareport.com/Civil%20Rights/CornellMuslimReportCivilRights.pdf

21.

Nosek

B. A.

Banaji

M. R.

Jost

J. T.

(2009). The politics of intergroup attitudes. In Jost

J. T.

Kay

A. C.

Thorisdottir

(Eds.), The social and psychological bases of ideology and system justification (pp. 480–506). Oxford, England: Oxford University Press.

22.

Pew Research Center. (2007). Republicans lag in engagement and enthusiasm for candidates. Retrieved from http://people-press.org/files/legacy-pdf/307.pdf

23.

Preacher

K. J.

Hayes

A. F.

(2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879–891.

24.

Republican National Committee. (n.d.). 2008 Republican platform. Retrieved from http://www.gop.com/2008Platform/2008platform.pdf

25.

Schuman

Steeh

Bobo

Krysan

(1997). Racial attitudes in America: Trends and interpretations. Cambridge, MA: Harvard University Press.

26.

Skitka

L. J.

Mullen

Griffin

Hutchinson

Chamberlin

(2002). Dispositions, ideological scripts, or motivated correction? Understanding ideological differences in attributions for social problems. Journal of Personality and Social Psychology, 83, 470–487.

27.

Torelli

Kaikati

(2009). Values as predictors of judgments and behaviors: The role of abstract and concrete mindsets. Journal of Personality and Social Psychology, 96, 231–247.

28.

Trope

Liberman

(2000). Temporal construal and time-dependent changes in preference. Journal of Personality and Social Psychology, 79, 876–889.

29.

Trope

Liberman

(2010). Construal-level theory of psychological distance. Psychological Review, 117, 440–463.

30.

Vallacher

R. R.

Wegner

D. M.

(1989). Levels of personal agency: Individual variation in action identification. Journal of Personality and Social Psychology, 57, 660–671.

31.

Wohl

M. J. A.

Branscombe

N. R.

(2005). Forgiveness and collective guilt assignment to historical perpetrator groups depend on level of social category inclusiveness. Journal of Personality and Social Psychology, 88, 288–303.