NowComment
Comments:
Full Summaries Sorted

Trust in Government is Implicit

My Trust in Government Is Implicit: Automatic Trust in Government and System Support

Chanita Intawan, University of California, Merced

Stephen P. Nicholson, University of California, Merced

How distrustful are people of government? Although large majorities of Americans express distrust in government, we propose that most of these same individuals also possess an implicit, gut-level trust in government. Using a common method to measure attitudes that people are either unwilling or unable to self-report, we found that most respondents implicitly trust government and that implicit trust is largely unrelated to explicit trust (as self-reported in surveys) and does not meaningfully vary by party identification or demographic characteristics. We also found that implicit trust is politically consequential, helping illuminate why a distrustful public nevertheless exhibits diffuse support and trust in the government to address crisis events, both foreign and domestic. We conclude that most Americans are of two minds about government, possessing both a positive, implicit trust and negative, explicit trust, and that each type matters in explaining orientations toward government.

Trust in government has eroded dramatically in the United States over the last 50 years. Although there have been momentary upsurges since the decline that

began in the late 1960s, low trust is pervasive. Using the American National Election Studys (ANES) well-known in-dex of trust in government that ranges from 0 (least trusting) to 100 (most trusting), the average score in 2012 was 22%, a historic low for the 50-year time series.1 With the notable exception of party identifiers expressing less trust when their party does not control the presidency (Hetherington and Ru-dolph 2015), the decline appears to be an authentic, across-the-board shift in public opinion. In Hetheringtons (2005, 1) words, Since the late 1960s and especially since Watergate, not even those who head the federal government have had much good to say about it.

Is the American public as distrustful of government as they say they are? Yes and no. Our conflicted answer is premised on the notion that people are of two minds about government. The pervasive distrust in government captured in opinion sur-veys represents a meaningful attitude that predicts support,

or the lack thereof, for many government programs (e.g., Hetherington 2005). Yet, despite the very low levels of trust in government, the public expresses a great deal of trust in the American political system (Hibbing and Theiss-Morse 1995), suggesting that the low levels of trust found in surveys does not reveal the whole story. Furthermore, in times of crisis, the trust expressed in surveys rises dramatically, a product, we believe, of an unspoken trust. We propose that beneath the surface of distrust, for most of the public, lurks an implicit trust in government, a gut-level trust rooted in the idealized depictions of American government learned at an early age. In contrast to explicit attitudes, the type that are reflective and consciously self-reported in surveys, implicit attitudes are automatically produced, often outside con-scious awareness (e.g., Banaji and Greenwald 2013; Kahne-man 2011; Lodge and Taber 2013; Pérez 2013, 2016). A large body of research in psychology has demonstrated that im-plicit attitudes are intuitive, knee-jerk responses to stimuli that people often do not, or cannot, accurately report. The purpose of our research is to explore whether an intuitive,

Chanita Intawan (cintawan@ucmerced.edu) is a PhD candidate and Stephen P. Nicholson (snicholson@ucmerced.edu) is a professor of political science at the University of California, Merced.

Data and supporting materials necessary to reproduce the numerical results in the article are available in the JOP Dataverse (https://dataverse

.harvard.edu/dataverse/jop). An online appendix with supplementary material is available at http://dx.doi.org/10.1086/694785. This study was carried out after informed consent was obtained from all participants. The research protocols were approved by the University of California, Merced Institutional Review Board (protocol number: UCM12-0032).

1. See http://electionstudies.org/nesguide/toptable/tab5a_5.htm.

The Journal of Politics, volume 80, number 2. Published online February 1, 2018. http://dx.doi.org/10.1086/694785

q 2018 by the Southern Political Science Association. All rights reserved. 0022-3816/2018/8002-0016$10.00

601

602 / Traditional Leaders and Electoral PoliticsChanita Intawan and Stephen P. Nicholson

gut-level trust in government inhabits the public mind and whether it matters.

To test our expectations, we measure implicit attitudes toward government and examine whether they are politically meaningful in explaining different types of political system support. Using a variant of the Implicit Association Test (IAT), a common method for capturing implicit attitudes, we had respondents rapidly match government with trust and distrust in a series of computerized sorting tasks. The responses are calculated into scores, which show whether peo-ple are more likely to associate government with trust or distrust. Using this measure, we found that most people im-plicitly trust government and that implicit trust is unrelated to explicit trust (as measured using the traditional ANES items). Furthermore, whereas explicit trust is strongly informed by partisanship (Hetherington and Rudolph 2015), implicit trust is not. We take these findings to indicate that implicit and explicit trust capture distinct orientations toward govern-ment.

We also found that implicit trust matters. Implicit trust plays an important role in explaining support for govern-ment, but it depends on whether the focus is diffuse (the political system) or specific (officeholders and policy) sup-port (see Easton 1965). For reasons we will elaborate later, we had predicted that implicit trust would help account for diffuse support but largely be unrelated to specific support. As anticipated, we found that implicit trust affected diffuse support but had no effect on specific support. We also ex-plored whether crisis events might draw on the publics in-stinct to turn to government for safety and security (see Al-bertson and Gadarian 2015; Merolla and Zechmeister 2009). Since implicit processes dominate “fight or flight responses, we hypothesized and found that implicit trust plays an im-portant role in explaining support for government in times of crisis.

Our findings also underscore the tenacity of self-reported or explicit trust. Studies of trust in government have dem-onstrated that it is a meaningfully held attitude that matters for explaining diminished public support of government pro-grams (e.g., Hetherington 2005; Hetherington and Rudolph 2015), a type of specific support. Using the traditional ANES trust index, our research provides further confirmation of this effect across myriad policy areas, and although we could not rule out endogeneity, we also found that it is significantly re-lated to diffuse support and trust in government during crisis events.

Our findings help illuminate important questions about trust in government, support for the American political sys-tem, and the relevance of implicit attitudes. One of the more important, and surprising, findings is that the public trusts

government but probably does not know it. The modal re-spondent in our survey exhibited positive implicit trust and negative explicit trust, and a follow-up analysis using ANES data suggests that low explicit trust is likely not a matter of social desirability bias. Our finding that trust inhabits the intuitive mind and that it is largely unrelated to explicit trust or does not vary with party identity helps us understand that people are of two minds about government. We also found that implicit trust is politically consequential, helping illu-minate why a distrustful public nevertheless exhibits diffuse support for government and its capacity to address crises, both foreign attacks and acts of nature. In addition to intro-ducing, measuring, and showing that implicit trust matters, our study suggests that when people are either unwilling or unable to express an opinion, implicit attitudes represent a promising avenue for understanding the complexity of public opinion. Finally, we hope our study encourages the study of implicit trust in other political systems to examine whether the same pattern of results applies in other democracies and in autocracies.

TWO TYPES OF TRUST IN GOVERNMENT

The study of trust in government, like most research on political attitudes, has relied heavily, if not exclusively, on surveys. The decline in trust in government over the last 50 years has been a central question. The decline has myriad causes including the economy (Chanley, Rudolph, and Rahn 2000; Citrin 1974), political scandal (Chanley et al. 2000), and evaluations of political actors (Citrin 1974) and public policy (Miller 1974). Others have examined social and de-mographic groups as a source of variation in trust (Alford 2001), but these relationships are generally weak except that party identifiers are more likely to trust government when their party controls the presidency (Hetherington and Ru-dolph 2015).

What is the meaning of the decline? This question helped motivate Miller (1974) and Citrins (1974) highly influential exchange about the decline in trust that began during the late sixties and early seventies. A key distinction is whether the decline concerns diffuse support, opposition to the political system, or specific support, opposition to leaders and public policy (Easton 1965). For Miller, the decline in trust is about diffuse support, a negative orientation toward the political system (1974, 951) that could undermine democracy. Ci-trin, however, argues that it involves specific support, a re-flection of opinion toward leaders and public policy. He also suggests that some portion of the decline might be akin to shoutingKill the umpire!at a baseball game(Citrin 1974, 978), a form of social desirability bias.

Yet, subsequent research has firmly established that trust in government is a meaningfully held attitude. Many studies have demonstrated that trust in government is highly con-sequential to specific support, affecting myriad political atti-tudes and behavior (Chanley et al. 2000; Hetherington 1998, 2005; Hetherington and Rudolph 2015). Trust in govern-ment has strong effects on shaping attitudes toward domes-tic spending (Hetherington 2005; Hetherington and Rudolph 2015), evaluations of political actors (Hetherington 1998), and electoral participation (Donovan and Bowler 2004). Trust, as reported in surveys, clearly matters for explaining a variety of political attitudes and behavior.

Despite all that has been written about political distrust, very little work has focused on the deep-seated trust that inhibits Americans from doing away with the bedrock in-stitutions of American government (Citrin 1974; Hibbing and Theiss-Morse 1995). In other words, despite widespread distrust, why do people still support the American political system and look to it in times of crisis? While it is possible that people misrepresent how trusting they are of govern-ment (a type of social desirability bias), we suspect that most people are not sufficiently introspective to tell a more com-plicated story, one that involves being of two minds about government. Our expectation is that the public largely trusts government at an implicit, gut level. It just does not know it. The study of trust in government has exclusively relied on surveys, self-reported attitudes. Such an approach, however, is limited when studying attitudes that people are incapable or unwilling to report.

A wide-ranging literature in psychology and cognitive science builds a persuasive case for the two minds hypoth-esis, the idea that people have an intuitive, unconscious mind and a reflective, conscious mind (Banaji and Greenwald 2013; Kahneman 2011). Sometimes referred to as dual-process theories of the mind, the idea is that the intuitive mind makes fast, automatic judgments that require few cognitive resources. In contrast, the reflective mind makes reasoned, conscious judgments available to introspection that require greater cognitive resources. A defining characteristic of implicit at-titudes then is the lack of introspective access and lack of conscious control over the contents of consciousness (Ba-naji, Nosek, and Greenwald 2004, 280). To be clear, implicit attitudes are themselves attitudes. As Pérez (2013) empha-sizes, all attitudes, whether they are implicit or explicit, are unobserved. Although explicit attitudes are gathered directly through self-reports, their accurate measurement also faces numerous challenges. The challenges, of course, differ. For implicit attitudes, a primary challenge is how to measure them if they often reside outside of conscious awareness or re-spondents lack self-awareness.

Volume 80Number 2April 2018 / 603

To measure implicit attitudes, psychologists have devel-oped tasks, such as the IAT, that have been systematically tested and validated (Greenwald, McGhee, and Schwartz 1998).2 In contrast to questions in surveys, the IAT measures attitudes indirectly. Using the IAT, a computer task that requires participants to make rapid judgments much like playing a fast-action video game, scholars measure associa-tions between attributes (e.g., good and bad) and contrasting category items (e.g., white and black). The faster a person matches an attribute and category, the stronger the implicit association. For example, a participant who is faster to match white/good than black/good exhibits a bias for white relative to black. Since the task does not allow participants an op-portunity to edit or censor themselves it reveals automatic or gut-level responses.

Implicit attitudes differ from explicit attitudes when peo-ple are unwilling (social desirability) or unable (lack of intro-spection) to report them. Under these circumstances, scholars have uncovered many instances in which implicit attitudes deviate from explicit attitudes. Perhaps nowhere has this in-sight been more valuable than in the study of racial attitudes since people may be unwilling or unable to report racial bias (Banaji and Greenwald 2013; Pérez 2013, 2016). In the study of politics, scholars have found that implicit biases against blacks (Valentino, Hutchings, and White 2002) and Latinos (Kam 2007; Pérez 2016) affect political attitudes and behav-ior. Scholars have also found that implicit religious (Albert-son 2011) and gender (Mo 2015) biases affect vote choice. Although scholars often study implicit attitudes when they expect them to differ from explicit attitudes, there are many instances in which they are similar. Implicit measures of par-tisanship, for instance, often reveal similar responses to ex-plicit measures (Duran, Nicholson, and Dale 2017; Iyengar and Westwood 2014; Theodoridis, forthcoming). Other than research examining whether the public implicitly perceives the Supreme Court as political (Hansford, Intawan, and Nichol-son, forthcoming), we know of no other research examining implicit attitudes about governmental institutions.

Why might people be of two minds when it comes to government? The study of race and implicit attitudes may provide important insights. For many years, non-Hispanic whites openly expressed racial prejudice in opinion surveys. Today, however, surveys typically reveal 90% or more of non-Hispanic whites (depending on the question) express-ing racially egalitarian opinions (Schuman et al. 1997). How-ever, despite their widely endorsing the norm of racial equal-

2. Although widely used, the IAT is not without critics (see Fiedler, Messner, and Bluemke 2006).

604 / Traditional Leaders and Electoral PoliticsChanita Intawan and Stephen P. Nicholson

ity, Banaji and Greenwald (2013) report that 75% of white participants reveal an implicit preference for white over black. Furthermore, they find that implicit racial attitudes matter: Among research participants who describe themselves as racially egalitarian, the Race IAT has been shown, reliably and repeatedly, to predict discriminatory behavior that was ob-served in the research (Banaji and Greenwald 2013, 47).

Trust in government has traveled a similar path insofar as it has exhibited large shifts over time. Before the precipitous decline that began in the mid-1960s, expressing trust in gov-ernment was commonplace. In the 1958 ANES, for example, 73% of Americans said they always or mostly trusted the government in Washington to do what is right. However, over time people became less trusting in government for the reasons cited above. For many people, whether it is sincere disillusionment with government or the desire to not appear naive, contemporary surveys routinely show very low levels of trust in government.

The disapproval a person is likely to receive for expressing trust in government no doubt pales in comparison to the deep condemnation a person invites when expressing racial prejudice. Nevertheless, expressing trust in government to-day is likely to evoke at least mild social disapproval, if not astonishment. Nobody wants to appear politically naive. However, distrust in government is probably not simply a matter of social desirability. As mentioned, explicit trust is a meaningfully held, politically consequential attitude (e.g., Hetherington 2005). Since explicit trust in government is not a nonattitude, we expect that implicit trust largely stems from a lack of introspection or self-awareness.

Despite the fact that explicit trust has declined since the 1960s, we believe the public likely retains an implicit trust in government. In a political system awash in cynicism, how could the public also possess an implicit trust in govern-ment? As suggested by Easton (1975, 448), trust in govern-ment is rooted in early political socialization. Although there is no doubt some variability in how people are politically socialized, there is remarkable uniformity in the highly pos-itive portrayals of the American political system (Hess and Torney 1968). Early civics learning in the United States por-trays government in an idealized fashion, imprinting images of a trustworthy regime. Consequently, trust in government is one of the earliest acquired, and strongly held, political at-titudes (Easton and Dennis 1969; Hess and Torney 1968).

Trust in government over the life cycle provides a window into the process (Jennings and Niemi 1968). Using the ANES trust in government index, Jennings and Niemi (1968) found that responses between parent and child were nearly inde-pendent and that high school students were much more trusting than their parents. They attribute the higher levels of

trust among youth to the strong emphasis on civics educa-tion in high school. As political awareness grows across the life span, people learn the norm of cynicism and distrust of government. Yet, we suspect that most people retain a gut-level, intuitive trust from childhood socialization.

We also expect that implicit trust, like explicit trust, matters. However, we expect them to matter in different ways. Implicit trust resides deep in the political mind, rooted in early childhood political socialization, so its effects are likely to affect stable, long-term dispositions toward govern-ment (diffuse support) but not current evaluations of gov-ernment such as support for government programs (specific support). We are not the first to suggest that implicit atti-tudes shape political system support. Scholars of system jus-tification theory, a type of diffuse support, have found that implicit attitudes play a substantial role in support for existing social arrangements (see Jost, Banaji, and Nosek 2004). Rather than focusing on implicit trust in government, how-ever, research on system justification has primarily approached it as a matter of personal and group interest, ideology, or group stereotypes (see Jost et al. 2004). We advance this research by examining whether gut-level trust in government affects po-litical system support, expecting it to positively affect system justification (our measure of diffuse support) but have little to no effect on specific support. In contrast to the relative stability of diffuse support, specific support captures current evaluations of incumbent performance (Hetherington 1998) or policy outputs (Caldeira and Gibson 1992; Hetherington 2005). Given the focus on government spending in research on trust (e.g., Hetherington 2005), we use it as our measure of specific support with the expectation that explicit, but not implicit, trust will increase support.

Finally, we expect implicit trust to increase trust in gov-ernment in times of crisis. Scholars have shown that crisis events can jolt an otherwise distrusting public into a trusting public (Albertson and Gadarian 2015; Hetherington and Nelson 2003; Merolla and Zechmeister 2009).3 The surge of explicit trust after a crisis event clearly suggests a different way of thinking about government, one that is likely informed, at least in part, by gut-level or emotional responses (Albertson and Gadarian 2015). Theories of implicit attitudes also suggest that implicit processes are likely to shape attitudes and be-havior during times of crisis. Explaining the allocation of at-tention between the intuitive mind (system 1) and the re-flective mind (system 2), Kahneman (2011, 35) explains that

3. Implicit trust, at least in part, is also likely to shape explicit trust during times of crisis. However, this inquiry is outside the scope of our study.

even in modern humans, System 1 takes over in emergencies and assigns total priority to self-protective actions. Similarly, threat can heighten system justification (see Jost et al. 2008), increasing public support for governmental action to respond to the crisis. In times of danger then, gut-level feelings are likely to take over, producing a knee-jerk trust response from citizens looking to government for security.

In sum, we expect that implicit trust will matter in po-litically important ways. We expect implicit trust to posi-tively affect diffuse support and evaluations of government in times of crisis but have little to no effect on specific sup-port. Although explicit trust is not our primary focus and is primarily included as a control, we also expect it to increase specific and diffuse support as well as support for govern-ment during times of crises.

STUDY DESIGN AND MEASURES

A nonprobability Internet sample recruited by Survey Sam-pling International (SSI) surveyed US respondents between March 27 and 30, 2015. In addition to the IAT, respondents answered survey questions including the ANES trust in gov-ernment items, a system justification scale, items on govern-ment spending, trust in government in times of crisis, and demographic variables (see app. sec. A for question wording and app. sec. B for descriptive statistics).

Although we collected data for 1,077 respondents, 34 re-spondents could not be matched to their IAT data, and 7 reported they were not fluent in English (a necessity for our task), leaving us with 1,036 respondents. There was ample variation on important demographic variables such as age (31% 1834, 29% 3554, 41% 55 and over), gender (68% fe-male), and education (22% high school degree or less, 38% some college, 26% college graduate). Finally, there was also variation in party identification (38% Democrat, 23% Re-publican; not including Independent leaners) and ideology (33% liberal, 31% conservative, 30% moderate).

We briefly revisit how the IAT works to introduce our study. In a traditional IAT, participants are timed as they rapidly place attitude objects into good and bad cate-gories. A respondents implicit attitude toward objects is gauged by how quickly the person matches attributes to ob-jects. In the study of racial attitudes, the most widely studied topic using the IAT, scholars examine the association be-tween racial stimuli (e.g., white and black faces) with positive or negative adjectives (e.g., good, bad). Differences in re-sponse times between the word and category pairings indi-cate bias. For example, if a participant is faster at pairing positive words with white than black and does the opposite for negative words, the result suggests the person has a pref-erence or bias for white over black. Given that the task re-

Volume 80Number 2April 2018 / 605

quires rapid responses (not allowing reflective thought), it reveals attitudes that are pre-introspective, sometimes held outside of conscious awareness.

To evaluate implicit trust in government, we use a variant of the IAT, the single-target implicit association test (ST-IAT; Karpinski and Steinman 2006). The ST-IAT is a close cousin to the IAT, but instead of featuring two categories (e.g., government vs. business), only one is included (gov-ernment). Since research on explicit trust uses a single cat-egory (government), we thought it critical to do the same with the implicit measure for comparability. Although dif-ferent from a standard IAT, the ST-IAT has sufficient re-liability and stability to assess complex attitudinal patterns (Bluemke and Friese 2008, 988; also see Karpinski and Stein-man 2006).

The trials in our study consist of matching the attributes of trust and distrust to the target category government. The trustwords include trust, trustworthy, reliable, depend-able, and honest, while the distrust words are distrust, un-trustworthy, unreliable, undependable, and dishonest. Lacking appropriate synonyms for government, we introduce vari-ability through altering fonts, a method akin to IATs that use pictures rather than text (personal communication with N. Sriram, IAT methodologist, September 29, 2012). We chose to represent the concept of government with different fonts rather than introducing imprecise words (e.g., authority) that do not adequately capture our concept of interest (govern-ment) and that are far afield from the ANES trust items.

The initial matching is randomized so that government could be paired with either trust or distrust first. To get a better sense of how it works, assume that a respondent is randomly assigned to first match government with trust. In the first block, the task requires respondents to match gov-ernment with trust, and in the second block they are asked to match government with distrust. The first block (trust) cor-responds to a key response on the left side of the keyboard, while the second block (distrust) corresponds to a key re-sponse on the right side of the keyboard. If respondents are faster to match government with trust relative to govern-ment with distrust, we conclude that they have an implicit trust in government.

For data analysis, we transform the trust ST-IAT data into D-scores (see Greenwald, Nosek, and Banaji [2003] for the standard D-scoring algorithm). Following the procedures outlined by Greenwald et al. (2003), we removed subjects where more than 10% of their trials are less than 300 milli-seconds. With the remaining data, we computed the mean difference in response times between the blocks where dis-trust is paired with government from the blocks where trust is paired with government. Each mean difference is then

606 / Traditional Leaders and Electoral PoliticsChanita Intawan and Stephen P. Nicholson

divided by the inclusive standard deviation for the corre-sponding blocks. The D-score is the average of the two ratios (mean difference divided by the inclusive standard devia-tion) ranging in value from 22 to 12. In our IAT, a positive D-score indicates an association between trust and govern-ment while a negative D-score indicates an association be-tween distrust and government. A score of zero indicates no association between government with trust or distrust.

The first set of analyses focuses on the nature of implicit trust and its relationship to explicit trust. Subsequent anal-yses investigate whether implicit trust affects diffuse (system justification) and specific support (government spending) and trust in government during times of domestic and interna-tional crisis.

RESULTS: THE NATURE OF IMPLICIT TRUST

The respondents in our sample, not surprisingly, were very much (explicitly) distrustful of government. Using the tra-ditional ANES trust index that ranges from 0 (least trusting) to 100 (most trusting), the density plot in figure 1A shows that the mean was .22, the same level of distrust found in the 2012 ANES. Figure 1A confirms that the public reports low levels of trust in government. However, the same cannot be said of implicit trust. Figure 1B depicts a density plot for the trust ST-IAT depicting a mean D-score of .09 (t(1; 035) p 9:94, p ! :001, SD p .29). The positive, statistically signifi-cant mean indicates that respondents are more likely to implicitly associate trust with government than distrust. Of course, some respondents were more likely to associate gov-ernment with distrust, but the average person exhibits a more trusting orientation. Furthermore, implicit trust much more closely approximates a normal distribution, whereas explicit trust is highly skewed toward distrust. The difference between figures 1A and 1B suggests that implicit trust is largely unrelated to explicit trust. Figure 2 (r p :11, p ! :001) depicts the relatively weak relationship, supporting our ex-pectation that implicit and explicit trust are largely unrelated.

Although suggestive that most people hold inconsistent implicit and explicit attitudes toward government, we ex-amined the distribution of two minds about government within individual respondents by splitting the sample into four categories: high explicit and implicit trust, low explicit and implicit trust, high explicit trust/low implicit trust, and low explicit trust/high implicit trust. For explicit trust, those with an ANES trust index of 50 or higher are in the high explicit trust category, while those with an index 49 and lower are in the low explicit trust category. For implicit trust, those with positive D-scores are in the high implicit trust category, and those with negative D-scores are in the low implicit trust category. As expected, the modal respondent

(55%) is low in explicit trust but high in implicit trust. Al-though the modal respondent is of two minds when it comes to government, 33% of respondents exhibit both low implicit and explicit trust. In addition, 8% of respondents are high on both implicit and explicit trust, whereas only 4% of respon-dents are high on explicit trust but low on implicit trust. The miniscule number of respondents falling into the latter cat-egory is reassuring since it is the least expected combination of implicit and explicit trust.

Who implicitly trusts government? Studies of the decline in (explicit) trust in government have found that it is largely not specific to subgroups (Alford 2001). However, scholars have found that (explicit) trust in government can vary substantially by party identification since partisans express greater trust in government when their party controls the presidency, a gap that has grown hand in hand with the in-crease in mass partisan polarization (Hetherington and Ru-dolph 2015).

To determine whether implicit trust varies by party identification, we examined implicit trust for Democrats and Republicans (not including Independent leaners). As a point of comparison, we first look at whether there are differences in explicit trust since research consistently finds greater trust among partisans whose party controls the presidency (Citrin 1974; Hetherington and Rudolph 2015). Consistent with research on the partisan trust gap, the mean score on the ANES trust index for Democrats was .26 and .18 for Repub licans. Are there similar differences in implicit trust? Given our argument that implicit understandings of government stem from early political socialization experiences, we do not anticipate much, if any, difference in implicit trust between Democrats and Republicans. And that is what we found. For Democrats, the mean D-score is .11 (t(395) p 7:56, p ! :001, SD p .29) and for Republicans it is .09 (t(242) p 5:12, p ! :001, SD p .29). A t-test comparing the two means is not statistically significant (p p :52). The scores for Democrats and Republicans are nearly identical, further suggesting that implicit trust in government is capturing a distinct orienta-tion toward government.

It is possible that differences in implicit trust might be revealed through ideology rather than party identification. If implicit attitudes stem, at least in part, from genetic causes, ideology is likely to be the more relevant political predis-position since scholars have found that genetics shape ideo-logical predispositions, not party identification (Hibbing, Smith, and Alford 2014). As before, we look at both explicit and implicit measures. Using the standard ANES question, we broke ideology into three categories: liberal, moderate, and conservative. For liberals, mean trust on the index was

.24, while it was .17 for conservatives. Moderates had an

Volume 80Number 2April 2018 / 607

Figure 1. Density plots of explicit and implicit trust in government

average of .22. The differences in implicit trust across ideo-logical labels are also minor. The mean D-score for liberals was .098 (t(339) p 6:42, p ! :001, SD p .28), while the mean D-score for conservatives was .084 (t(317) p 4:88, p ! :001, SD p .31). Moderates had a mean D-score of .083 (t(313) p 5:20, p ! :001, SD p .28). Taken together, it ap-pears that implicit trust does not meaningfully vary across

party or ideology. In addition, we did not find significant differences in implicit trust within demographic categories such as age, gender, income, education, or race/ethnicity (see app. sec. C).

A potential objection to our implicit trust measure is that it measures something other than trust. For example, it might be capturing an affective orientation toward government such

608 / Traditional Leaders and Electoral PoliticsChanita Intawan and Stephen P. Nicholson

Figure 2. Relationship between implicit and explicit trust

as patriotism (Huddy and Khatib 2007). To help rule out this possibility, in our pilot study we conducted a separate IAT, but rather than have participants match government to trust and distrust we had them match government to good and bad, the standard approach used by implicit attitudes researchers to measure positive and negative valence. The results of the analysis in appendix section D show that implicit trust is largely unrelated to holding an implicit, positive orientation toward government. Furthermore, the absence of a meaningful association suggests that confounds rooted in positive affect such as patriotism are also unlikely. For interested readers, appendix section E reproduces the figures above using data from two pilot studies.

In sum, we found that people implicitly trust government and that implicit trust does not meaningfully differ across partisan or demographic groups. We also found that implicit trust appears to be distinct from explicit trust and alternative conceptualizations such as patriotism (see app. sec. E). In the next section, we explore whether, and how, implicit trust matters.

DOES IMPLICIT TRUST MATTER?

In the following analyses, we examine whether implicit trust predicts diffuse and specific support and trust in government during times of crisis. To evaluate diffuse support, we use Kay and Josts (2003) eight-item scale of system justification that measures perceptions of the fairness and legitimacy of the prevailing social and political system. This scale has been used to measure support for the American system (Wakslak, Jost,

and Bauer 2011, 295) and is broad, including items such as In general, the American political system operates as it should and Most policies serve the greater good. Responses to these items fall on a nine-point scale ranging from strongly agree to strongly disagree. Following Kay and Jost (2003), we recoded items so that higher values indicated higher levels of system support and created a mean score for each respondent by collapsing across the eight items (a p :74). For the data analysis, the mean values are rounded to the nearest whole number.

In contrast to the broad, systemic focus of the system justification scale we use for diffuse support, our measures of specific support are grounded in policy preferences over government spending. Using research on explicit trust as our guide (e.g., Hetherington 2005; Hetherington and Ru-dolph 2015), we asked several ANES questions about gov-ernment spending, including a commonly used item about general spending preferences and specific questions about welfare and homeland security. The choice of questions was intended to capture broad government spending preferences and specific preferences on domestic and foreign policy. We coded responses so that higher values indicate a preference for increased government spending.

We also included novel measures of trust in government during crisis events, asking respondents whether they trust government to address a natural disaster and foreign attack. The first item asks, Imagine a natural disaster hits the United States, causing great harm to the American people. How much do you trust the government in Washington to successfully

address this problem? The second item asks, Imagine the United States is attacked by a foreign nation, endangering the safety of the American people. How much do you trust the government in Washington to successfully address this prob-lem? Responses are scaled from one to seven, with higher values indicating greater trust. These questions obviously lack the intensity of an actual crisis, so we take the findings from this analysis as suggestive.

Since this set of analyses examines how implicit and ex-plicit trust affect dependent variables that could plausibly be argued to be a cause of trust, it is important to address challenges to causal inference. In the study of explicit trust, scholars have shown that trust is exogenous to outcome vari-ables such as policy attitudes (e.g., Hetherington 2005). Given the absence of panel data, however, we are not able to rule out endogeneity. In contrast, we are more confident that implicit attitudes are exogenous to our dependent variables of inter-est. On substantive grounds, implicit attitudes are the product of early learning. As the product of past experiences (Banaji and Greenwald 2013), implicit attitudes are likely to be caus-ally prior to specific judgments. On measurement grounds, the IAT is difficult to deceive (Karpinski and Steinman 2006). Given the automatic, indirect nature of IAT responses, we are largely confident that implicit trust is an exogenous predictor of our dependent variables.

Analysis of diffuse support (system justification)

We first examine whether implicit trust predicts diffuse support. As mentioned, our measure of diffuse support, the system justification scale, is coded from one to nine, with higher values indicating greater system justification. The mean system justification index is 4.61 (SD p 1:23). Recall, our expectation is that higher levels of implicit trust lead to greater system justification. To evaluate this hypothesis, we ran ordered probit models, first with only implicit trust as the independent variable and then additional models with explicit trust and other controls including party identifica-tion, ideology, race/ethnicity, gender, income, and education (see app. sec. A for variable coding).

Table 1 presents the results of the system justification analysis. The positive and statistically significant coefficient for implicit trust across each of the models indicates that an increase in implicit trust produces an increase in system justification. Model 1 shows that the effect of the implicit measure is significant without controls, whereas model 2 shows that its effect is robust in the face of standard control variables, including explicit trust. Since our data do not allow us to rule out endogeneity between explicit trust and system justification, however, model 3 presents results including con-trol variables without explicit trust. With explicit trust re-

Volume 80Number 2April 2018 / 609

Table 1. The Effect of Implicit Trust on System Justification: Ordered Probit Estimates

Model 1

Model 2

Model 3

Implicit trust

.317*

.226*

.316*

(.122)

(.123)

(.122)

Explicit trust

. . .

.024

. . .

(.002)

Party identification

. . .

.020

2.017

(.024)

(.024)

Ideology

. . .

.088*

.066*

(.030)

(.029)

Black

. . .

2.065

2.066

(.138)

(.137)

Latino

. . .

.061

.128

(.162)

(.161)

Nonwhite other

. . .

.066

.057

(.169)

(.168)

Gender

. . .

2.143*

2.191*

(.075)

(.075)

Income

. . .

.040*

.042*

(.013)

(.012)

Education

. . .

2.032

2.003

(.027)

(.027)

Age

. . .

.008*

.006*

(.003)

(.002)

Note. Standard errors in parentheses. N p 890. * p .05, one tailed test.

moved in model 3, the coefficient for implicit trust increases in size and remains statistically significant.

To examine the effect of implicit trust on system justifi-cation, we computed predicted probabilities. Simulated pre-dicted probabilities are calculated for the minimum and max-imum values of implicit trust, which range from 22 to 2, with variables held at their mean or modal values. Since system justification is an ordinal level variable ranging from 1 to 9, we discuss changes in probabilities for two values on the scale, representing low (3) and high (6) system justification. For low system justification (3), a change from the lowest to the highest values of implicit trust decreases the probability of low system support by .10. For respondents high in system justification (6), a change from the lowest to the highest values of implicit trust increases system justification by .13. Overall, the results support our hypothesis that higher implicit trust increases system justification (or decreases low system justi-fication).

As expected, the coefficient for explicit trust is positive and statistically significant. Although largely included as a con-trol, the result for explicit trust provides reassurance about the validity of the system justification scale as a measure of dif-

610 / Traditional Leaders and Electoral PoliticsChanita Intawan and Stephen P. Nicholson

fuse support.4 Among the other control variables, table 1 shows that ideology has a significant effect on system justification. Specifically, as we would expect, higher conservatism increases system justification. Finally, as expected, age and income both significantly increase system justification. In sum, control-ling for a variety of other factors, implicit trust is a significant predictor of system justification.

Following standard practice, the results presented exam-ine implicit and explicit attitudes separately. Although we are cognizant that the explicit trust measure is likely endogenous to some of our dependent variables, we nevertheless wanted to show what happens when a respondent is of two minds about government (high implicit trust and low explicit trust). To explore this question, we created a measure to examine the within-individual difference or gap between a standardized explicit and implicit trust measure. In the analysis presented in appendix section F, we found that as respondents (gener-ally positive) implicit attitudes diverged from their (generally negative) explicit attitudes, people are increasingly likely to express system justification.

Analysis of specific support (government spending)

Borrowing from Eastons (1965) definition of specific support that includes policy outputs, we examine how implicit and explicit trust affect preferences for government spending. The variables for government spending are coded such that a preference for government spending to remain the same or be increased is coded 1 and a preference for decreased spending is coded 0. As mentioned, the expectation is that implicit trust is unrelated to preferences on government spending. But, in keeping with the literature on how (explicit) trust matters (Hetherington 2005), we expect those high in explicit trust to be more supportive of increased government spending.

The results presented in table 2 use the same predictors as before. As anticipated, the coefficient for implicit trust is not a significant predictor in any of the three models, sug-gesting that implicit trust is unrelated to general government spending. Consistent with the trust literature (e.g., Hether-ington 2005), however, the positive and significant coefficient for explicit trust suggests that higher trust in government leads to greater support for increasing government spending. Ide-ology is a significant predictor in all models, and party iden-

4. Although we cannot rule out endogeneity, we nevertheless calculated predicted probabilities for explicit trust for comparison purposes. Examining a minimum to maximum change in explicit trust (from 0 to 100 on the ANES trust scale) produces a .16 decrease in low system justification, whereas a comparable change in the ANES trust index produces a .18 increase in high system justification.

tification is a significant predictor in two of three models. In other words, conservatives and Republicans are less likely to support increased government spending except for homeland security. Gender, age, race, and income also have statistically significant effects, but not consistently.

As done in the previous analysis, we also examined the effect of the (within-individual) gap between implicit and explicit trust in appendix section F and found that an increase in the gap increased the probability that a respondent was significantly in favor of greater spending on social welfare and homeland security. However, there was no significant effect of the trust gap on government spending.

Analysis of trust in government during crisis events

Our final analysis examines whether implicit trust affects eval-uations of government during a natural disaster and foreign attack. As mentioned, we do not expect these questions to capture the intensity of an actual crisis event, so results are only suggestive. However, as a check to see whether the ques-tions invoked a mind-set akin to an actual crisis, we calcu-lated mean trust in government scores for each dependent variable. If the means are sufficiently large, it would suggest that respondents placed themselves in a crisis mind-set, albeit one lacking the intensity of a real-world political crisis. Recall that each dependent variable is coded from 1 to 7, with higher values representing greater trust in government to address the problem. With this in mind, the mean trust scores for the natural disaster and foreign attack questions are 4.09 and 4.54, respectively. Although the means are likely smaller than what we would expect during actual crises, they are also substantially larger than the very low trust score from the ANES trust index, suggesting that respondents did not take an evaluate government as usualapproach to answering these questions.

Table 3 presents the results for both the natural disaster and foreign attack models. Again, since we are unable to rule out endogeneity between explicit trust and the dependent variables, explicit trust is only included as a control. Regard-less of model specification, higher levels of implicit trust in-crease trust in government in the face of a foreign attack or natural disaster. We calculated the average change in proba-bility for the crisis models, again comparing minimum to maximum changes in implicit trust. A minimum to maxi-mum change in implicit trust produces a .10 increase in trust in the foreign attack model and a .11 change in the natural disaster model. Although we cannot rule out endogeneity, the minimum to maximum change for explicit trust produced a .17 increase in trust in the foreign attack model and a .21

Volume 80 Number 2 April 2018 / 611

Table 2. The Effect of Implicit and Explicit Trust on Government Spending

Increase Government $

Increase Welfare $

Increase Homeland Security $

Ordered Probit Estimates

Probit Estimates

Probit Estimates

Implicit trust

.089

2.180

.242

(.124)

(.166)

(.175)

Explicit trust

.003*

.006*

.014*

.002

(.003)

(.003)

Party identification

2.163*

2.180*

.051

(.024)

(.031)

(.035)

Ideology

2.184*

2.153*

.097*

(.030)

(.040)

(.043)

Black

.163

.190

.423*

(.139)

(.203)

(.209)

Latino

.165

.180

.184

(.164)

(.223)

(.235)

Nonwhite other

2.195

2.106

2.370*

(.171)

(.223)

(.219)

Gender

2.190*

2.194*

.098

(.075)

(.100)

(.107)

Income

2.018

2.071*

.021

(.013)

(.017)

(.018)

Education

2.022

.017

2.016

(.027)

(.037)

(.039)

Age

2.006*

.001

.017*

(.003)

(.003)

(.004)

Intercept

. . .

1.952*

2.906*

(.272)

(.279)

Note. Standard errors in parentheses. N p 890. * p .05, one tailed test.

increase in trust in the natural disaster model. As before, we also examined the effect of the gap between implicit and ex-plicit trust and found that as the gap increased, the more likely a respondent expressed trust in government during a crisis (see app. sec. F).

IS LOW TRUST IN GOVERNMENT A MATTER

OF SOCIAL DESIRABILITY?

As discussed, the study of implicit attitudes is motivated by the recognition that people are sometimes unwilling (social desirability) or unable (lack of self-awareness) to report their opinions. We reasoned that either explanation might account for the low trust reported in national surveys. Nevertheless, we suggested it was more likely to be a lack of self-awareness since explicit trust is a meaningfully held, politically conse-quential attitude (e.g., Hetherington 2005). Unfortunately, our data cannot speak to this question. Thankfully, the 2008 ANES included items on self-monitoring, the degree to which persons

modify their behavior to make positive impressions on others (Snyder 1974). For example, high self-monitors are more likely to give racially liberal responses (Berinsky 2004) and identify as Independents (Klar and Krupnikov 2016) compared to low self-monitors. We examined whether explicit trust varies by self-monitoring.

The responses to the self-monitoring questions were re-coded and averaged to create a single index of self-monitoring, with higher numbers indicating more self-monitoring (see app. A for question wording). Using a median split, those with a self-monitoring index at or above 2 are classified as high self-monitors, while those with an index lower than 2 are low self-monitors. The trust in government index uses the same four questions as before. Overall, mean (explicit) trust in the 2008 ANES was .29, a quantity similar to our SSI sample. Low self-monitors had a mean trust of .288, whereas high self-monitors had a mean trust of .293, a minor difference that is not sta-tistically significant (p p :73). We interpret this null result as suggestive that explicit trust in government is not affected by

612 / Traditional Leaders and Electoral PoliticsChanita Intawan and Stephen P. Nicholson

Table 3. The Effect of Implicit Trust in Government during a Crisis Event: Probit Estimates

Natural Disaster

Foreign Attack

Natural Disaster

Foreign Attack

(1)

(2)

(3)

(4)

Implicit trust

.289*

.252*

.380*

.327*

(.123)

(.123)

(.122)

(.122)

Explicit trust

.025*

.019*

. . .

. . .

(.002)

(.002)

Party identification

2.022

2.038

2.057*

2.064*

(.024)

(.024)

(.024)

(.024)

Ideology

2.050*

2.094*

2.060*

2.101*

(.030)

(.030)

(.029)

(.029)

Black

.008

.051

.003

.052

(.137)

(.138)

(.136)

(.137)

Latino

.274*

2.083

.321*

2.021

(.162)

(.162)

(.161)

(.161)

Nonwhite other

2.127

.034

2.118

.029

(.168)

(.169)

(.167)

(.168)

Gender

2.038

2.320*

2.100

2.356*

(.075)

(.075)

(.074)

(.075)

Income

.022*

.021*

.025*

.025*

(.012)

(.012)

(.012)

(.012)

Education

.013

.008

.040

.031

(.027)

(.027)

(.027)

(.027)

Age

.006*

.007*

.003

.005*

(.002)

(.002)

(.002)

(.002)

Note. Standard errors in parentheses. N p 890. * p .05.

social desirability and thus is more likely to be a matter of a lack of self-awareness.

DISCUSSION AND CONCLUSION

A primary finding from our study is that the public expresses distrust in government while, at the same time, implicitly trusting government. The divergence of implicit and explicit measures of trust in government suggests that individuals hold contrary attitudes about government. Although we cannot say for certain, we do not believe that people are intentionally misrepresenting themselves when they report that they do not trust government. The large body of research demonstrating that trust is a powerful predictor of political attitudes and be-havior suggests that it is a meaningful attitude, and our results for self-monitoring further suggest low explicit trust is not a matter of social desirability. For this reason, we find it more plausible that people do not know that they also trust gov-ernmentthey simply cannot tell survey researchers they hold favorable attitudes toward government because they are unable to access these attitudes. Merolla and colleagues (2013), for example, found evidence of a biological basis for trust in

government, a potential contributor to implicit trust that lies beyond an individuals introspective abilities. However, future work should explore this question.

Much of the research on political trust finds that explicit political trust matters in explaining the declining support for redistributive social programs (e.g., Hetherington 2005). What might implicit trust in government tell us? We expected im-plicit trust to matter but in different ways. We explored two possibilities. As a gut-level, intuitive orientation, implicit trust in government might explain, at least partly, why Americans express great support for their political system (diffuse sup-port) despite being deeply cynical about government processes and policy (specific support). We found that it did. Using sys-tem justification as a measure of diffuse support, we found that respondents higher in implicit trust were more likely to ex-press greater system justification. But, we did not expect im-plicit trust to help account for specific support, evaluations of public policy. As expected, we found implicit trust to be unre-lated to support for government spending both generally and in the areas of social welfare and homeland security.

The second possibility we explored was whether implicit trust helps explain trust in government during a crisis. Schol-

ars have shown that confidence and trust in government dramatically increases during times of international (Hether-ington and Nelson 2003; Merolla and Zechmeister 2009) and domestic (Albertson and Gadarian 2015) crisis. Given the threat inherent in crises, we thought that dangerous circum-stances might elicit an automatic, knee-jerk trust in govern-ment. We found that it did. Using two questions, one about a hypothetical natural disaster and the other about a hypo-thetical foreign attack, we found that respondents higher in implicit trust were more likely to trust the government to address these crises. Future research might unpack this ef-fect by looking at whether the dependent variables in these analyses are priming people to think about specific parts of the government such as the Federal Emergency Management Agency or the military (see Hetherington and Rudolph 2015). It may not be as specific as these institutions, but the crisis dependent variables might have caused people to consider one of the original purposes of governmentto protect its citizens. This might tap into an implicit conception of gov-ernment as protector. Whether this implicit conception of gov-ernment is desirable for democracy is another question. On the one hand, as Merolla and Zechmeister (2009) point out, supporting decisive and strong leaders in times of crisis may be highly beneficial to resolving a problem. On the other hand, they argue that such a response can produce detri-mental policy outcomes, putting democracy at risk. Implicit support for government, then, may both help and hinder democratic processes.

Since we found that both types of trust matter, it is natu-ral to ask whether one is more important than the other. We are reluctant to address this question given the limitations of our data. Nevertheless, we are confident that specific support likely rests squarely in the domain of explicit trust. As found in previous studies, the ANES trust measures consistently pre-dicted support for different types of government programs (e.g., Hetherington 2005), whereas implicit trust did not. Although previous research has ruled out endogeneity in analyses of the effects of explicit trust on government pro-grams (e.g., Hetherington 2005, chap. 4), we do not have sim-ilar assurances for system justification or trust in government during times of crisis. Indeed, the dependent variables in those analyses appear to have a good deal in common with the ANES trust items. Since explicit trust was included primarily as a control variable, we were less concerned about this problem. Nevertheless, our analyses in appendix section F looking at the gap between implicit and explicit trust attitudes are in-formative in thinking about the relationship between the two not so much as a competition but instead as what happens when both types of attitudes inhabit the same mind but run in opposite directions.

Volume 80Number 2April 2018 / 613

Promising next steps might examine implicit trust over time and space. A time series of implicit trust would help us better understand its nature, especially whether it is rela-tively stable and whether its effects vary across political cir-cumstances. For example, our third study was intended to capture the effects of implicit trust during a national crisis. Our questions were intended to invoke a crisis mind-set, but longitudinal data capturing normal and crisis times would be more informative. Although we are optimistic about fu-ture research on implicit trust in the United States, the IAT is not as time and cost effective to implement as traditional survey items, so we are less sanguine about implicit trust join-ing explicit trust as a standard population-level time-series measure. The study of implicit trust comparatively would also be a promising area of future research. Not only would it be informative to see whether our results travel to other democracies, but it would be valuable to examine the nature and consequences of implicit trust in authoritarian govern-ments.

ACKNOWLEDGMENTS

We thank Tom Hansford, Efren Pérez, and Alex Theodoridis for helpful comments and suggestions. Previous versions of this article were presented at the Western Political Science Association, Hollywood, CA, March 2830, 2013, and Ex-ploring New Frontiers, Forging New Synergies: Bolstering the Links between Bio-Politics and Political Psychology, Uni-versity of California, Merced, June 910, 2016.

REFERENCES

Albertson, Bethany L. 2011. Religious Appeals and Implicit Attitudes.

Political Psychology 32 (1): 10930.

Albertson, Bethany L., and Shana Kushner Gadarian. 2015. Anxious Politics: Democratic Citizenship in a Threatening World. New York: Cambridge University Press.

Alford, John R. 2001. Were All in This Together: The Decline of Trust in Government, 19581996. In John R. Hibbing and Elizabeth Theiss-Morse, eds., What Is It about Government That Americans Dislike? New York: Cambridge University Press.

Banaji, Mahzarin, and Anthony G. Greenwald. 2013. Blindspot: Hidden Biases of Good People. New York: Delacorte.

Banaji, Mahzarin R., Brian A. Nosek, and Anthony G. Greenwald. 2004. No Place for Nostalgia in Science: A Response to Arkes and Tetlock. Psy-chological Inquiry 15 (4): 279310.

Berinsky, Adam. 2004. Can We Talk? Self-Presentation and the Survey Response. Political Psychology 25 (4): 64359.

Bluemke, Matthias, and Malte Friese. 2008. Reliability and Validity of the Single-Target IAT (ST-IAT): Assessing Automatic Affect towards Mul-tiple Attitude Objects. European Journal of Social Psychology 38 (6): 977 97.

Caldeira, Gregory A., and James L. Gibson. 1992. The Etiology of Public Support for the Supreme Court. American Journal of Political Science 36 (3): 63564.

614 / Traditional Leaders and Electoral PoliticsChanita Intawan and Stephen P. Nicholson

Chanley, Virginia A., Thomas J. Rudolph, and Wendy M. Rahn. 2000. The Origins and Consequences of Public Trust in Government. Public Opinion Quarterly 64:23956.

Citrin, Jack. 1974. Comment: The Political Relevance of Trust in Govern-ment. American Political Science Review 68 (3): 97388.

Donovan, Todd, and Shaun Bowler. 2004. Reforming the Republic: Democratic Institutions for the New America. Upper Saddle River, NJ: Prentice Hall.

Duran, Nicholas, Stephen P. Nicholson, and Rick Dale. 2017. The Hidden Appeal and Aversion to Political Conspiracies as Revealed in the Response Dynamics of Partisans. Journal of Experimental Social Psychology 73 (November): 26878.

Easton, David. 1965. A Systems Analysis of Political Life. New York: Wiley. Easton, David. 1975. A Re-assessment of the Concept of Political Sup-

port. British Journal of Political Science 5 (4): 43557.

Easton, David, and Jack Dennis. 1969. Children in the Political System: Origins of Political Legitimacy. New York: McGraw Hill.

Fiedler, Klaus, Claude Messner, and Matthias Bluemke. 2006. Unresolved Problems with the I, the A, and the T: A Logical and Psychometric Critique of the Implicit Association Test (IAT). European Review of Social Psychology 17 (1): 74147.

Greenwald, Anthony G., Debbie E. McGhee, and Jason L. K. Schwartz. 1998. Measuring Individual Differences in Implicit Cognition: The Implicit Association Test. Journal of Personality and Social Psychology

74 (6): 146480.

Greenwald, Anthony G., Brian A. Nosek, and Mahzarin R. Banaji. 2003. Understanding and Using the Implicit Association Test.Pt. 1,An Improved Scoring Algorithm. Journal of Personality and Social Psy-

chology 85:197216.

Hansford, Thomas G., Chanita Intawan, and Stephen P. Nicholson. Forth-coming. Snap Judgment: Implicit Perceptions of a (Political) Court.

Political Behavior.

Hess, Robert D., and Judith V. Torney. 1968. The Development of Political Attitudes in Children. New York: Anchor.

Hetherington, Marc J. 1998. The Political Relevance of Political Trust.

American Political Science Review 92:791808.

Hetherington, Marc J. 2005. Why Trust Matters: Declining Political Trust and the Demise of American Liberalism. Princeton, NJ: Princeton Uni-versity Press.

Hetherington, Marc J., and Michael Nelson. 2003. Anatomy of a Rally Effect: George W. Bush and the War on Terrorism. PS: Political Science and Politics 36 (1): 3742.

Hetherington, Marc J., and Thomas J. Rudolph. 2015. Why Washington Wont Work: Polarization, Political Trust, and the Governing Crisis. Chi-cago: University of Chicago Press.

Hibbing, John R., Kevin B. Smith, and John R. Alford. 2014. Predisposed: Liberals, Conservatives, and Biology of Political Differences. New York: Routledge.

Hibbing, John R., and Elizabeth Theiss-Morse. 1995. Congress as Public Enemy: Public Attitudes toward American Political Institutions. New York: Cambridge University Press.

Huddy, Leonie, and Nadia Khatib. 2007. American Patriotism, National Identity, and Political Involvement. American Journal of Political Science 51 (1): 6377.

Iyengar, Shanto, and Sean J. Westwood. 2014. Fear and Loathing across Party Lines: New Evidence on Group Polarization. American Journal of Political Science 59 (3): 690707.

Jennings, M. Kent, and Richard G. Niemi. 1968. The Transmission of Political Values from Parent to Child. American Political Science Review 62 (1): 16984.

Jost, John T., Mahzarin R. Banaji, and Brian Nosek. 2004. A Decade of System Justification Theory: Accumulated Evidence of Conscious and Unconscious Bolstering of the Status Quo. Political Psychology 25 (6):

881919.

Jost, John T., Janina Pietrzak, Ido Liviatan, Anesu N. Mandisodza, and Jaime L. Napier. 2008. System Justification as Conscious and Nonconscious Goal Pursuit. In James Y. Shah and Wendi L. Gardner, eds., Handbook of Motivation Science. New York: Guilford.

Kahneman, Daniel. 2011. Thinking Fast and Slow. New York: Farrar, Straus, & Giroux.

Kam, Cindy D. 2007. Implicit Attitudes, Explicit Choices: When Subliminal Priming Predicts Candidate Preference. Political Behavior 29 (3): 34367.

Karpinski, Andrew, and Ross B. Steinman. 2006. The Single Category Implicit Association Test as a Measure of Implicit Social Cognition.

Journal of Personality and Social Psychology 91 (1): 1632.

Kay, Aaron C., and John T. Jost. 2003. Complementary Justice: Effects of Poor but HappyandPoor but HonestStereotype Exemplars on System Justification and Implicit Activation of the Justice Motive. Journal of Personality and Social Psychology 85 (5): 82337.

Klar, Samara, and Yanna Krupnikov. 2016. Independent Politics: How American Disdain for Parties Leads to Political Inaction. New York: Cam-

bridge University Press.

Lodge, Milton, and Charles S. Taber. 2013. The Rationalizing Voter. New York: Cambridge University Press.

Merolla, Jennifer L., Guy Burnett, Kenneth V. Pyle, Sheila Ahmadi, and Paul J. Zak. 2013. Oxytocin and the Biological Basis for Interpersonal and Political Trust. Political Behavior 35 (4): 75376.

Merolla, Jennifer L., and Elizabeth J. Zechmeister. 2009. Democracy at Risk: How Terrorist Threats Affect the Public. New York: Cambridge University Press.

Miller, Arthur H. 1974. Political Issues and Trust in Government, 1964 1970. American Political Science Review 68 (3): 95172.

Mo, Cecilia Hyunjung. 2015. The Consequences of Explicit and Implicit Gender Attitudes and Candidate Quality in the Calculations of Voters.

Political Behavior 37 (2): 35795.

Pérez, Efrén O. 2013. Implicit Attitudes: Meaning, Measurement, and Syn-ergy with Political Science. Politics, Groups, and Identities 1 (2): 27597.

Pérez, Efrén O. 2016. Unspoken Politics: Implicit Attitudes and Political

Thinking. New York: Cambridge University Press.

Schuman, Howard, Charlotte Steeh, Lawrence Bobo, and Maria Krysan. 1997.

Racial Attitudes in America: Trends and Interpretations. Cambridge, MA: Harvard University Press.

Snyder, Mark. 1974. Self-Monitoring of Expressive Behavior. Journal of Personality and Social Psychology 30 (4): 52637.

Theodoridis, Alexander G. Forthcoming. Me, Myself, and (I), (D) or (R)? Partisan Intensity through the Lens of Implicit Identity. Journal of Politics.

Valentino, Nicholas A., Vincent L. Hutchings, and Ismail K. White. 2002. Cues That Matter: How Political Ads Prime Racial Attitudes during Campaigns. American Political Science Review 96 (1): 7590.

Wakslak, Cheryl J., John T. Jost, and Patrick Bauer. 2011. Spreading Ratio-nalization: Increased Support for Large-Scale and Small-Scale Social Systems following System Threat. Social Cognition 29 (3): 288302.

Copyright of Journal of Politics is the property of The Southern Political Science Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.

DMU Timestamp: March 29, 2019 18:11