The Psychology of Voting and Election Campaigns
Theories of ethnocentrism, authoritarianism, social dominance, system justification, and moral foundations anticipate that social preferences for dominant ingroups will be stronger among political conservatives than political liberals. We found support for this hypothesis across a variety of social group attitudes and stereotypes with both implicit and explicit measures using large datasets collected via the Internet. Across domains and measurement types, both conservatives and liberals preferred dominant groups and held stereotypes aligned with normative expectations or historical precedent. However, conservatives consistently showed stronger dominant group preferences than did liberals. Such effects tended to be stronger on explicit compared to implicit measures suggesting that ideological differences in automatic evaluations are amplified in the explicit expression of attitudes and values. Results illustrate that political orientation is a marker for a broader network of beliefs and attitudes in which dominant groups are especially favored by conservatives compared to liberals, and the ideological differences are more pronounced explicitly than implicitly.
Implicit measures assess attitudes indirectly, without requiring introspection and self-report. The Affect Misattribution Procedure (AMP) is a new implicit measure of attitudes and emotions that uses misattributions rather than response time facilitation to gauge responses. It does so via an “emotional illusion,” in which emotional reactions influence impressions of ambiguous items in a way that is difficult to “undo.” Research using this approach suggests that implicit and explicit attitudes may be more closely related than is often thought. Implications will be illustrated using studies of racial attitudes, inter-group emotions, and attitudes toward political candidates.
Confidence in Attitudes: Explicit and Implicit Factors
Richard E. Petty
Ohio State University
Attitudes, our global evaluations of objects, people, and issues, are an important construct in social psychology and political science. Yet, not all attitudes are equally durable and impactful (see Petty & Krosnick, 1995, for a review). For example, attitudes held with high confidence (i.e., certainty that the attitude is correct), are more likely to direct behavior (e.g., voting, contributing to candidates) than attitudes that are equivalent in favorability but which are associated with some doubt. The current research explores two new sources of attitude certainty – explicit and implicit. In one series of studies we focus on certainty that stems from people’s explicit subjective impressions of how they arrived at their attitudes. For example, when people come to believe, whether true or not, that they arrived at their attitudes because of considerable thinking, they gain attitude certainty – increasing attitudinal impact and durability. In a second series of studies we focus on implicit factors. For example, when people have implicit evaluative associations that conflict with their explicit attitudes, people experience implicit uncertainty that renders attitudes less durable and impactful than when there is no discrepancy. Assessing both explicit and implicit certainty in relevant political attitudes in surveys can presumably aid in understanding election outcomes.
There is considerable research demonstrating that emotion is a particularly influential factor in decisionmaking, including voter decisionmaking. We know less about another influence of emotion in politics, the emotionality of the candidates themselves. Anecdotal evidence strongly suggests that extreme emotional displays are very negatively received (although perhaps more by the punditry than by actual voters). Research to date, in some cases stripping away the identities of candidates and the substance of their speech, has shown that people’s preferences for candidates are indeed affected by the quality of the candidates’ emotional displays. Experimental research has shown that emotionally expressiveness candidates are preferred if their party affiliation is not known, but the opposite is true if it is. Research on the affective tone of political messages indicates that they are most effective when they resonate with audience affect, and this may generalize to actual emotional expression by candidates. By way of promoting further research in this area to better understand and predict voters’ preferences as a function of candidate emotionality, I propose applying Ekman and Friesen’s theory of display rules, investigating the specific display rules that may be unique to political discourse and campaigns.
It is “well-established” in the theoretical literature that stereotypes are a powerful, difficult to overcome, “default” basis for person perception, and that any influence of stereotypes on person perception constitutes a social evil that needs to be eliminated. The polar opposite of this theoretical view is also well-established in the decision-making under uncertainty literature. This theoretical work concludes that people routinely ignore base-rates about groups (stereotypes) when judging individuals and that, to be as accurate as possible, people should rely on base-rates in the absence of perfectly clear, useful and relevant (“diagnostic”) individuating information. Finally, the empirical literature (over 300 studies) clearly establishes that: 1) Influences of stereotypes on person perception are, in general, weak, fragile, and fleeting; and 2) individuating information effects are, in general, powerful and pervasive.
Because political stereotypes are common, powerful, and do not evoke social desirability concerns, they seemed to be an excellent context for attempting to resolve the theoretical disarray inherent in these literatures. A preliminary experiment is presented, which examined the role of stereotypes regarding political candidates’ party membership and candidates’ policy statements (individuating information) in people’s judgments and inferences regarding those candidates’ political positions. The candidates were constructed to largely reflect the real world; therefore, people rated a leftwing Democrat, a moderate Democrat, a moderate Republican and a rightwing Republican. The results of this study were that both party and policy statements influenced those judgments and inferences, but that party usually was stronger. In an attempt to explain this pattern, a new decision-tree model of political person perception was developed and tested in a second experiment. The model is too complicated to be summarized here, but some of its main ideas are: 1. If there is clear individuating information directly related to a policy, people will decide to use that and ignore party (e.g., if one discovers that candidate X is pro-choice, one will also infer X supports stem cell research, regardless of X’s party); 2. If there is no clear individuating information about a candidate, and there is a clear ideologically-based position on the issue (e.g., liberals are usually more pro-choice conservatives), people will use party to infer a candidate’s position (i.e., they will assume a Democrat is more pro-choice than a Republican). Study Two tested these ideas by manipulating policy position orthogonally to party membership. Therefore, people made judgments about a leftwing Democrat, a rightwing Democrat, a leftwing Republican, and a rightwing Republican. Results of this experiment strongly supported the model’s predictions. Implications for political person perception in particular, and regarding rationality and irrationality in the use of stereotypes more generally, will be discussed.
Calling to Mind a Significant Other:
Implications of Research on Transference for Reactions to Political Candidates
Susan M. Andersen
New York University
Affect and motivation are known be evoked when a mental representation of a significant other is activated – for example, in the social-cognitive process transference, which occurs when a new person minimally resembles a significant other. This implicitly activates the mental representation of this significant other (Andersen, Reznik, & Manzella, 1996), and the relational self experienced in relation to that significant other (Andersen & Chen, 2002), and leads to a variety of cognitive, evaluative, affective, and motivational consequences. Significant others can be defined broadly, as a parent, a family member, a best friend, former love, current mate, or other well known person in whom the respondent feels psychologically invested. This other may be loved or detested or something in between. Consequences of the process include making biased inferences about the new person, quite immediately liking or disliking him or her, and feeling a sense of belonging and trust in relation to him or her, or a sense of disconnection and distrust. Even a positive significant other can evoke fearfulness/anxiety or disappointment or apathy under the right circumstances. In this presentation, the nature of this research and its findings will be reviewed, and the implications that can be drawn from it for people’s gut reactions to political figures (and gut reactions to how these political figures interact with each other on the public stage) will be considered.
This talk will integrates research on attitude extremity, candidate evaluation, and the perception of group variability. This proposed work will examine polarization of issue stances in the electorate over time, perceived polarization of the electorate and of the political parties on those same issues, the relationship between self-polarization and perceived polarization, and the role of these variables in determining the strength of issue-based candidate preferences.
It is a truism that political attitudes and behavior are pervasively shaped by people’s membership and identification with groups (ethnic, religious, partisan, or other). Traditional accounts of people’s reactions to their own and other groups have been heavily cognitive (whether social-cognitive models based on stereotyping processes, or social-identity models based on intergroup differentiation). Research in the last decade or so (by myself and my collaborator Diane Mackie, as well as other researchers) has begun to broaden this focus by considering the role of emotions in people’s attitudes and behaviors with respect to their own groups and others. Identifying with a group makes the group functionally part of the self, so the group acquires emotional significance. As a result, people appraise objects and events in terms of their implications for the group, not just for the individual self, and experience discrete group-based emotions such as anger, fear, guilt, satisfaction, pride, sympathy, or resentment based on group-relevant events. These emotional reactions in turn lead to specific, differentiated action tendencies or desires for action, which may include support or opposition to public policies, political candidates, or parties. This talk will present an overview of theory and research on group-based emotions and their effects, including recent evidence that group-based emotions can be clearly differentiated from the same people’s individual-level emotional responses. The overall picture is that emotions are among humans’ most fundamental tools for understanding their place in the world and the nature of current threats or opportunities, and adaptively behaving in relation to those appraisals. Given the pervasive role of group memberships and identifications in politics, group-based rather than individual emotions should be key drivers of political preferences and behaviors.
The voter’s illusion (Quattrone & Tversky, 1984) is the idea that people vote because they think that their own decision to vote (and not to abstain) is more diagnostic of the decisions of like-minded others (i.e., supporters of the same party or candidate) than of the decisions of others (i.e., supporters of competing parties of candidates). Empirically, there is a small but replicable effect such that voters’ self-reported intentions to vote are correlated with the difference between the following two judgments: “The likelihood that my party wins if I vote” and “The likelihood that my party wins if I abstain.” I argue that the voter’s illusion has a rational basis in Bayesian induction. Arguments to the contrary are internally inconsistent because they fail to distinguish between causal and diagnostic reasoning
Recent research has demonstrated that asking questions often does more than solicit a response. Across a wide range of domains, we find that asking a question about positively viewed behaviors often leads to subsequent increases in that actual behavior while asking about negatively viewed behaviors often leads to decreases in the actual behavior. For example, asking a question about the likelihood of buying a new automobile actually increases automobile purchase rates some 30+%. In addition, we have found that asking questions containing hypothetical content can similarly lead to changes in the respondent’s subsequent behavior. In the current talk I’ll review some general results on the question-behavior effect as well as specific studies conducting in a voting context, and further discuss potential implications of asking questions about voting on actual voting behavior.
Interpersonal interaction is a central aspect of daily life. People interact with friends, relatives, and colleagues, each of who may have different thoughts and expectations. The experiments I will discuss show that explicit and implicit attitudes are shaped and regulated within these interpersonal interactions. When one is motivated to affiliate with others or motivated to know more, attitudes spontaneously converge toward the perceived opinions of others to create interpersonal synchrony. This synchrony, in turn, facilitates social bonding and attitude certainty. Given the goals of this conference, I will also speculate about how considering these interpersonal context effects within the larger societal milieu surrounding elections can affect understanding of the interpersonal basis of attitudes and voter behavior.
The Varieties of Psychological Experience Underlying Preference Inconsistency
Perhaps the most fundamental discrepancy between classical normative assumptions and observed behavior concerns individual preferences, which are normatively assumed to be well-ordered and consistent, but descriptively shown to be inconsistent and malleable. Not having at their disposal a clear and reliable procedure for assigning values to options, people need to construct their preferences in the context of decision. The attractiveness of options depends, among other things, on the nature of other options in the set, the procedure used to express preference, the context of evaluation, as well as the decision maker’s self conception. The varieties of psychological experience underlying preference inconsistency are reviewed, and their implications are discussed. Preference inconsistency, it is proposed, is the outcome not of distracted shortcuts or avoidable errors, but of fundamental aspects of mental life that are central to how people process information.
Identity, Belief, and Change
University of Colorado
Our research posits that social identity is a major barrier to attitude and behavioral change. Evidence and arguments–ranging from policy recommendations to constructive feedback–tend to be mistrusted when they originate from across social identity lines (e.g., party affiliations). Additionally, acceptance of counter-attitudinal information, or of information provided from across group divides, can be costly for an individuals’ sense of identity and personal integrity. A series of studies demonstrates the importance of social identity in responding to information and group conflict. Evidence is also presented showing that a theory-informed intervention strategy–wherein respondents affirm an alternative source of self-identity unrelated to the issue at hand–mitigates bias in the evaluation of evidence, resistance to persuasion, and unwillingness to compromise in negotiation across social identity lines. A final study is presented suggesting that, in an academic context, this “affirmation” intervention can produce long-term consequences on perceptions of outgroup members. Discussion centers on how the processes involved in identity and perception may be time-related, involving cycles of interaction between perceptions at one point in time and subsequent perceptions, which can become self-reinforcing and self-sustaining over the long term. Small interventions, it is suggested, can produce long-term effects if they interrupt such recursive feedback cycles.
The “end of ideology” was declared by social scientists in the aftermath of World War II. They argued that: (a) ordinary citizens’ political attitudes lack the kind of stability, consistency, and constraint that ideology requires; (b) ideological constructs such as liberalism and conservatism lack motivational potency and behavioral significance; (c) there are no major differences in content (or substance) between liberal and conservative points of view; and (d) there are few important differences in psychological processes (or styles) that underlie liberal versus conservative orientations. The end-of-ideologists were so influential that researchers ignored the topic of ideology for many years. However, current political realities, recent data from the National Election Studies, and results from an emerging psychological paradigm provide strong grounds for returning to the study of ideology. The liberalism-conservatism distinction remains a pervasive and parsimonious means of organizing thoughts, feelings, and behaviors (for both laypersons and social scientists). Studies reveal that there are indeed meaningful political and psychological differences that covary with ideological self-placement. Situational variables—including system threat and mortality salience—and dispositional variables—including openness and conscientiousness—affect the degree to which an individual is drawn to liberal vs. conservative leaders, parties, and opinions. A psychological analysis is also useful for understanding the political divide between “red states” and “blue states.”
Traditional models of goal pursuit have held that an individual consciously chooses and pursues goals. However, much recent research supports the notion that goals can be activated outside of conscious awareness by features of the environment. Once activated, the goals are pursued just as consciously chosen goals are, but without the intention or knowledge of the individual pursuing them. What are the antecedents or triggers of nonconscious goals? Various features of the environment can activate them, including significant others, brands, and situations of ego threat and temptation. Once individuals pursue the nonconscious goals, they succeed or fail at the goals to various degrees. Does success and failure at nonconscious goal pursuit lead to the same consequences as does success and failure at conscious goal pursuit? Failure at nonconscious goals leads to negative mystery mood, self-enhancement, aggressive behavior, stereotyping, and worse performance at a similar future task.
Connecting to the Wills of the People:
Voting Behavior as Goal
My talk will discuss the implications of recent self-regulatory research on social self-regulation and goal management for understanding the motivated process and dynamics of voting behavior. Specifically, this goal systems approach may help articulate how and when candidates may come to be strongly associated with the popular needs and goals of voters and when such associations may actually give rise to voting behavior. This approach may also help articulate the dynamics of voting behavior by distinguishing the motivational and cognitive factors that may influence voters’ initial choice from those that affect whether voting behavior is maintained over time.