John Stuart Mill called it “commonplace” for political systems to have “a party of order or stability and a party of progress or reform” (Reference Mill and Gray1991). Ralph Waldo Emerson agreed, noting that “the two parties which divide the state, the party of conservatism and that of innovation, are very old, and have disputed the possession of the world ever since it was made,” and he inferred that this “irreconcilable antagonism must have a correspondent depth of seat in the human condition” (Reference Emerson1903). The antagonism between two primal mindsets certainly pervades human history: Sparta and Athens; optimates and populares; Roundheads and Cavaliers; Inquisition and Enlightenment; Protagonus and Plato; Pope Urban VIII and Galileo; Barry Goldwater and George McGovern; Sarah Palin and Hillary Rodham Clinton. The labels “liberal” or “leftist” and “conservative” or “rightist” may be relatively recent (etymologically they are typically assumed to date to the French Revolution, but they appear to be much older; see Laponce Reference Laponce1981) but the political division they describe is ancient and universal (Bobbio Reference Bobbio1996; Jost Reference Jost2006; Jost & Amodio Reference Jost and Amodio2012; McCarty et al. Reference McCarty, Poole and Rosenthal2006). Is Emerson right in his claim that this division springs from a deep, possibly innate part of the human condition? Does political temperament vary from person to person because the physiology and psychology constituting human nature also varies from person to person? If so, how are the individuals who support parties of stability and order psychologically and physiologically different from those who support parties of progress and innovation?
Existing research offers only incomplete answers to these questions. All too often, the questions are not even asked. Folk wisdom and much scholarly research assumes political orientations are products of socialization, learned from parents and family, acquired by osmosis from sociodemographics, or conditioned exclusively by environmental situations and cultural contexts. The logic here is reasonable; authority figures encountered at impressionable early stages of life, as well as broader circumstances experienced later seem obvious sources of influence on a range of personal and social orientations including those relating to politics. Yet the effects of parental socialization on political orientations are fairly meager (bivariate correlations typically running between 0.1 and 0.3) with the exception of identification with social groups such as a political party (Jennings & Niemi Reference Jennings and Niemi1968; Niemi & Jennings Reference Niemi and Jennings1991). Adding socidemographic variables such as age, education level, and family income to models of political attitudes and behavior only modestly increases explanatory horsepower (Plutzer Reference Plutzer2002). Moreover, sociodemographic variables in and of themselves do not explain the precise factors at work in structuring preferences. In sum, political orientations do not seem to be the automatic result of parental socialization and sociodemographic circumstances.
To the surprise of many (but see Merelman Reference Merelman1971), it is increasingly clear that Emerson's intuition was right. Politics might not be in our souls, but it probably is in our DNA. More than 25 years ago Nicholas Martin and Lindon Eaves (Martin et al. Reference Martin, Eaves, Heath, Jardine, Feingold and Eysenck1986), using a standard twin design on a large sample, produced heritability estimates between 0.2 and 0.4 for attitudes on a wide variety of political issues (e.g., capital punishment, disarmament, abortion). More recent twin studies consistently confirm these findings and extend them to behaviors such as voter turnout (Alford et al. Reference Alford, Funk and Hibbing2005; Bell et al. Reference Bell, Schermer and Vernon2009; Bouchard & McGue Reference Bouchard and McGue2003; Fowler et al. Reference Fowler, Baker and Dawes2008; Hatemi et al. Reference Hatemi, Medland, Morley, Heath and Martin2007; Reference Hatemi, Alford, Hibbing, Martin and Eaves2009; Reference Hatemi, McDermott, Eaves, Kendler and Neale2013; Klemmensen et al. Reference Klemmensen, Hatemi, Hobolt, Petersen, Skytthe and Norgaard.2012; Smith & Hatemi Reference Smith and Hatemi2013). Given the many assumptions undergirding twin studies, it is important to note that alternative techniques for estimating heritability that do not rely on twins report slightly smaller but still statistically significant effects of genetics on political orientations (Benjamin et al. Reference Benjamin, Cesarini, Matthijs, Dawes, Koellinger, Magunsson, Chabris, Conley, Laibson, Johannesson and Visscher2012).
Though twin studies are valuable for assessing the general roles of heritability and various categories of environmental influence (shared and unshared), they say little about the specific sources of influence within those broad categories. Accordingly, efforts are underway to identify particular genetic regions or even particular genes that might relate to politics (Fowler & Dawes Reference Fowler and Dawes2008; Hatemi et al. Reference Hatemi, Gillespie, Eaves, Maher, Webb, Heath, Medland, Smyth, Beeby, Gordon, Montgomery, Zhu, Byrne and Martin2011; McDermott et al. Reference McDermott, Tingley, Cowden, Frazzetto and Johnson2009; Settle et al. Reference Settle, Dawes, Christakis and Fowler2010). Yet although intriguing, it is not clear genopolitics research can comprehensively illuminate the source of the “irreconcilable differences” that Mill, Emerson, and others have long suspected to be the basis of political beliefs. Any given candidate gene (or genetic region) is likely to explain only a small fraction of the variance in a complex quantitative trait like political temperament and statistically isolating meaningful relationships amongst such marginal impacts is difficult. That situation is reflected in the poor replication record of candidate gene association studies, particularly when they involve interactions with any of a large number of possible environmental influences. For example, Fowler and Dawes (Reference Fowler and Dawes2008) identified allelic variation in two genes involved in the serotonin system (the transporter 5-HTT and the degrader MAO-A) that systematically correlated with political participation. A reanalysis of the same data by Charney and English (Reference Charney and English2012) using different procedures did not reproduce that finding, and replications have fed the controversy as much as resolved it (see Deppe et al. Reference Deppe, Stoltenberg, Smith and Hibbing2013; Fowler & Dawes Reference Fowler and Dawes2013). So although twin studies suggest that political orientations may be heritable, identifying the particular genetic pathways that lead to political orientations constitutes a daunting challenge.
The same could be said, however, about identifying the particular environmental pathways that lead to political orientation. Twin studies repeatedly point to a strong influence of the unshared environment and a fairly weak role of the shared environment on political orientations. Traditional research on the correlates of political temperament backs twin study conclusions, finding a weak role of the shared environment (e.g., minimal influences of parental socialization; Jennings & Niemi Reference Jennings and Niemi1968), and efforts to identify specific environmental influences other than the “usual suspect” sociodemographics (age, education, gender, and the like) have met with at best mixed success.
The conclusion that political orientations are shaped by a combination of largely unspecified genetics and only slightly better specified features of the (mostly unshared) environment does not constitute much of an advance. Is this the best that can be done in describing the nature and derivation of political orientations that are so diverse and strongly held that they can lead to paralyzing societal divisions and sometimes violence? Here we explore the correlates of variation in political orientations at an intermediate level that is neither as proximate and overtly political as parents' political preferences nor as distal as genetic polymorphisms. This level includes the physiological and psychological processes relevant when particular classes of stimuli present themselves. The logic for our approach is straightforward. Life is about encounters: sights, sounds, smells, imaginings, objects, and people. These encounters are indisputably physiological and psychological because the systems employed to sense, process, formulate, and execute a response to stimuli are psychological and physiological. Equally indisputable is the existence of individual-level variation in these physiological and psychological mechanisms. Even if a stimulus is identical, one individual will sense, process, and respond to it differently than another.
Those measurable and variable physiological and psychological signatures constitute valuable and crucial constructs in and of themselves, regardless of whether their causes are genetic, environmental, or (more likely) a combination of both. They are the tangible residue of all the genetic and experiential influences that have been retained and then incorporated for future guidance. As such, the embodied predispositions constitute inertial psychological and physiological set-points that serve as baselines for behaviors and attitudes. Individual-level variation in those predisposed response patterns goes some distance toward defining who we are as people, including the nature of our political orientations.
In this article we make the case that variations in physiological and psychological responses to a particular category of stimuli – those that are negative (or aversive) – correlate with political orientations. It is well-known that on the average people respond and pay more attention to negative than to positive stimuli (Baumeister et al. Reference Baumeister, Bratslavsky, Finkenauer and Vohs2001). Our interest, however, is in individual variation around the “average.” Certain individuals respond strongly and attend concertedly to negative stimuli; others less strongly. We reason that this variation is likely to correlate with the political positions endorsed by each individual.
Hypothesizing a connection between political orientations and psychological/physiological responses is encouraged by the intraperson longitudinal stability of each. Political scientists have documented the role of unspecified long-term forces in structuring political orientations and decisions, referring to them, alternatively, as antecedent conditions (Marcus et al. Reference Marcus, Sullivan, Theiss-Morse and Wood1995), long-term predispositions (Zaller Reference Zaller1992), or ingrained habits (Gerber et al. Reference Gerber, Green and Schachar2003; Plutzer Reference Plutzer2002). A recent study even notes that when it comes to an interest in politics, “you've either got it or you don't” (Prior Reference Prior2010; see also Alwin & Krosnick Reference Alwin and Krosnick1991; Sears & Funk Reference Sears and Funk1999). For their part, psychological and physiological response sets are also relatively stable over time (Cohen & Hamrick Reference Cohen and Hamrick2003; de Weerth & van Geert Reference De Weerth and Van Geert2002; Huizenga et al. Reference Huizenga, Koper, de Lange, Pols, Stolk, Grobbee, de Jong and Lamberts1998; Lykken Reference Lykken1999) and therefore – in theory at least – could help to explain the longitudinal stability of political orientations.
Considerable evidence suggests that liberals and conservatives are distinct on a wide variety of psychological and physiological variables. In the main sections of this article, we summarize that evidence and argue that a surprising amount of it can be integrated around the theme of differences in physiological and psychological responses to negative events and stimuli. An important preliminary step, however, is to show that political decisions in many cases are influenced by factors people do not believe are involved. Some may reject the assertion that deep physiological and psychological differences distinguish liberals and conservatives because they believe that higher level decision-making, such as that involving politics, is the product of rational, conscious responses to the objective world and therefore not influenced by forces outside of conscious awareness. This flattering view of human decision-making in the area of politics is most likely unwarranted.
2. Politics and the subconscious
Extraneous or even subthreshold factors affect a wide range of day-to-day decisions and opinions and moral, religious, and political decisions, and beliefs are not immune to such forces. People sitting in a messy, malodorous room tend to make harsher moral judgments than those who are in a neutral room (Schnall et al. Reference Schnall, Haidt, Clore and Jordan2008), and disgusting ambient odors decrease approval of gays (Inbar et al. Reference Inbar, Pizarro, Knobe and Bloom2009b; see also Inbar et al. Reference Inbar, Pizarro and Bloom2009a; Reference Inbar, Pizarro and Bloom2012a). Sitting on a hard, uncomfortable chair leads to less flexible attitudes than those offered when sitting on something soft and comfortable (Ackerman et al. Reference Ackerman, Nocera and Bargh2010). People reminded of physical cleansing – for example, by the presence of hand sanitizer – render sterner judgments than those who are not given such a reminder (Helzer & Pizarro Reference Helzer and Pizarro2011). Moral judgments can change as a result of hypnotic suggestion (Wheatley & Haidt Reference Wheatley and Haidt2005), and prompting analytical thinking lowers religiosity (Gervais & Norenzayan Reference Gervais and Norenzayan2012). Focusing exclusively on political variables, when churches are employed as polling places people's tendency to cast votes for right-of-center candidates and ballot propositions increases compared with when public schools serve as polling places (Berger et al. Reference Berger, Meredith and Wheeler2008; Rutchick Reference Rutchick2010). Mortality prompts – images of tombstones, hospitals, and the elderly – foster the adoption of conservative political positions (Jost et al. Reference Jost, Fitzsimons, Kay, Greenberg, Koole and Pyszczynski2004; Landau et al. Reference Landau, Solomon and Greenberg2004; although see Castano et al. Reference Castano, Leidner, Bonacossa, Nikkah, Perrull, Spencer and Humphrey2011). Italians who implicitly associated symbols of the United States with negative concepts were more likely to vote against the proposed expansion of a U.S. military base even though they believed themselves undecided on this issue (Galdi et al. Reference Galdi, Arcuri and Gawronski2008).
In a series of studies following the lead of Zajonc (Reference Zajonc1980), political scientist Milt Lodge and his colleagues demonstrated the importance of hot cognition or automaticity in political judgments (Lodge & Hamill Reference Lodge and Hamill1986; Lodge & Taber Reference Lodge and Taber2005). Political stimuli often produce extremely quick emotional reactions that affect more deliberate cognitive processes such as memory recall, attention, and information processing. In one study, images of a happy face flashed for too short a time to register in conscious awareness resulted in participants offering fewer reasons to oppose immigration (Lodge & Taber Reference Lodge and Taber2013), indicating that quick, preconscious responses color political judgments. These concepts receive extensive development in the work of psychologist John T. Jost and colleagues. Jost refers to preconscious biases as motivated social cognition and repeatedly demonstrates that people do not come into political situations unconstrained (Carney et al. Reference Charney2008; Jost Reference Jost2006; Jost & Amodio Reference Jost and Amodio2012; Jost et al. Reference Jost, Glaser, Kruglanski and Sulloway2003). These constraints typically operate outside of conscious awareness though people often insist that their political decisions are solely the result of conscious considerations. Even neuroscientists sometimes express surprise that political orientations are influenced by subthreshold factors (Wade Reference Wade2011).
The relevance of subthreshold factors allows for the possibility that political temperament is systematically related to a range of psychological and physiological response patterns. In the following sections we summarize research showing this possibility is in fact a reality. First, we examine liberal-conservative psychological differences as reflected in (survey) self-reports. Second, we review psychological differences that are not fully accessible to the participants themselves. Third, we describe evidence of physiological differences between liberals and conservatives. Finally, we synthesize the research by arguing that many of the correlations described are tied together by the common thread of differences in response patterns to negative stimuli.
3. Politics and self-reported psychological differences
Mass-scale political preferences systematically correlate with an astonishing variety of psychological characteristics. Perhaps the best known is authoritarianism. The Authoritarian Personality, by Adorno et al. (Reference Adorno, Frenkel-Brunswik, Levinson and Sanford1950), claimed that characteristics such as conventionalism, submission to authority and anti-intellectualism clustered into a distinct, measurable personality trait, and developed the F-scale to measure that trait. Variation on this scale correlated with a wide range of political attitudes, including self-placement on a liberal-conservative dimension. Similarly, McCloskey (Reference McCloskey1958) concluded that traits such as confidence, social behavior, mood, cognitive complexity, social behavior, and preferred leadership styles also distinguished liberals and conservatives.
Since then, some research explicitly rolls politics into personality, whereas other research treats politics as conceptually distinct from personality. A prominent example of the former is Altemeyer's development of a scale to measure Right-Wing Authoritarianism or RWA (Altemeyer Reference Altemeyer1981; Reference Altemeyer1996). To illustrate, one RWA item asks whether respondents agree that “God's laws about abortion, pornography, and marriage must be strictly followed before it is too late, and those who break them must be strongly punished.” Altemeyer's blending of life and political tendencies to capture a personality trait of broad social relevance is part of a pattern in post-Adorno research. Wilson and Patterson (Reference Wilson and Patterson1968) measure conservatism by combining explicitly political stands on issues such as school prayer and the death penalty with preferences on broader lifestyle issues such as modern art and pajama parties (see also Wilson Reference Wilson1973). Bouchard urges combining religion, politics, and authoritarianism into a single concept (Reference Bouchard, Voland and Schiefenhövel2009). Tomkins (Reference Tomkins and White1963) and Tetlock and Mitchell (Reference Tetlock, Mitchell, Mellers and Baron1993) also conflate personality and politics. Others, however, go out of their way to tap authoritarian tendencies without explicitly invoking politics – for example, by measuring nonpolitical authoritarianism with survey items on child rearing (Feldman & Stenner Reference Feldman and Stenner1997; Hetherington & Weiler Reference Hetherington and Weiler2009; Stenner Reference Stenner2005). Those who favor more authoritarian parenting styles are significantly more likely to be political conservatives. Another longstanding concept merging politics and personality is Social Dominance Orientation (SDO) (Pratto et al. Reference Pratto, Sidanius, Stallworth and Malle1994; Sidanius & Pratto Reference Sidanius and Pratto2001) which is based on the observation that people vary in their comfort levels with group-based discrimination and dominance, with some embracing the vision of a hierarchy of groups.
Much recent research takes advantage of personality psychology's growing acceptance of a standard package of five core personality traits, known as the Big Five: conscientiousness, agreeableness, openness to new experiences, extraversion, and emotional stability (Gosling et al. Reference Gosling, Rentfrow and Swann2003; McCrae Reference McCrae1996; Mondak et al. Reference Mondak, Hibbing, Canache, Seligson and Anderson2010). Though Big Five personality batteries are not overtly political, two traits consistently discriminate political orientation across a broad range of studies: Conservatives tend to score higher on conscientiousness and liberals tend to score higher on openness to new experiences (see Caprara et al. Reference Caprara, Barbaranelli and Zimbardo1999; Gerber et al. Reference Gerber, Huber, Doherty, Dowling and Ha2010; Mondak & Halperin Reference Mondak and Halperin2008; Rentfrow et al. Reference Rentfrow, Jost, Gosling, Potter, Jost, Kay and Thorisdottir2009). Other Big Five traits do not correlate as consistently with political orientations but extraversion and emotional stability have been associated with economic (though not social) conservatism (Gerber et al. Reference Gerber, Huber, Doherty, Dowling and Ha2010; Young Reference Young2009) and elements of agreeableness have also been linked to ideology, with conservatives being more polite and liberals more empathetic (Hirsh et al. Reference Hirsh, DeYoung, Xu and Peterson2010). These personality differences encourage researchers to explore the possibility that liberals and conservatives construct and occupy different individual and social environments. For example, consistent with their tendency to report being more conscientious, conservatives' “life spaces” tend to have more cleaning supplies and organizing elements, including calendars, postage stamps, and laundry baskets, and, consistent with their penchant for new experiences, liberals tend to have more art supplies, travel materials, and greater varieties of books and music (Carney et al. Reference Charney2008).
Personality traits are far from the only psychological characteristics that discriminate political orientations. Shalom Schwartz's research focuses on the values that guide an individual's personal life, such as conformity, tradition, security, power, achievement, hedonism, stimulation, self-direction, universalism, and benevolence (Schwartz Reference Schwartz and Zanna1992; for a good overview, see Feldman Reference Feldman, Sears, Huddy and Jervis2003). Relationships among these values are stable across cultures (Piurko et al. Reference Piurko, Schwartz and Davidov2011; Schwartz Reference Schwartz2006) and are consistently related to individual-level variation in political preferences. Conservatives tend to value security and conformity and liberals tend to value self-expression and stimulation – and those values even turn out to be powerful predictors of voting behavior (Schwartz et al. Reference Schwartz, Caprara and Vecchione2010).
Jonathan Haidt and colleagues demonstrate convincingly that liberals and conservatives tend to employ different considerations when making moral judgments. Liberals rely primarily on concerns for equality and harm avoidance, whereas conservatives are more likely to take into account considerations such as purity, authority, and in-group/out-group status (Graham et al. Reference Graham, Haidt and Nosek2009; Haidt & Graham Reference Haidt and Graham2007; Haidt & Joseph Reference Haidt and Joseph2004). As was the case with personality traits and core values, these connections of moral foundations to politics apply in numerous countries (Graham et al. Reference Graham, Haidt and Nosek2009; for additional work on the political relevance of selected moral foundations, see Petersen Reference Petersen2009). The connection between purity concerns and conservatism is consistent with the previously mentioned finding that conservatives tend to have more cleaning supplies in their living spaces (Carney et al. Reference Charney2008). It is also consistent with the finding (replicated cross-nationally) that people with stronger self-reported disgust are more conservative (Inbar et al. Reference Inbar, Pizarro and Bloom2009a; Reference Inbar, Pizarro, Iyer and Haidt2012b; but see Tybur et al. Reference Tybur, Merriman, Caldwell, McDonald and Navarrete2010).
Jost et al.'s (Reference Jost, Glaser, Kruglanski and Sulloway2003) extensive meta-analysis examining the core differences between the left and the right concludes that comfort with change and attitudes toward equality are the two central variables distinguishing liberals and conservatives. The relevance of these traits argues strongly against assertions that ideology has little meaning for most people and is decreasingly relevant to modern life (Bell Reference Bell1960; Converse Reference Converse and Apter1964; Fukuyama Reference Fukuyama1992). Jost (Reference Jost2006) suggests ideology is no more likely to end than personality traits given that ideology is the political reflection of aspects of broader psychology.
Liberal-conservative differences even extend to tastes and preferences. Compared to liberals, conservatives are more likely to prefer simplicity and realism as opposed to complexity and abstractions in art (Wilson et al. Reference Wilson, Ausman and Matthews1973) and puns as opposed to unexpected incongruity in humor (Wilson Reference Wilson1990). A recently collected sample of our own shows statistically significant relationships between political conservatism and preferences for familiar as opposed to unfamiliar foods and music, for poetry that rhymes, and for novels that come to closure (Neiman Reference Neiman2012).
This last finding is consistent with a substantial body of research investigating the relationship between political beliefs and the “need for cognitive closure.” In 1993, Kruglanski et al. introduced a battery now widely used to tap preferences for closure. It includes items such as “I do not like situations that are uncertain,” “I like to have friends who are unpredictable,” and “even after I've made up my mind about something, I am always eager to consider a different opinion.” Cross-nationally this battery consistently suggests individuals who desire cognitive closure tend to self-identify as conservative (Chirumbolo et al. Reference Chirumbolo, Areni and Sensales2004; Federico et al. Reference Federico, Golec and Dial2005; Golec Reference Golec2002; Golec et al. Reference Golec, Cislak and Wesolowska2010; Kossowska & van Hiel Reference Kossowska and Van Hiel2003; Rock & Janoff-Bulman Reference Rock and Janoff-Bulman2010; van Hiel et al. Reference van Hiel, Pandelaere and Duriez2004), identify with conservative political parties (Kemmelmeier Reference Kemmelmeier1997), and adopt conservative positions on specific topics such as the death penalty and general punitiveness (Jost et al. Reference Jost, Kruglanski, Simon, Thompson, Messick and Levine1999), immigration (Chirumbolo et al. Reference Chirumbolo, Areni and Sensales2004), and a variety of other social and economic topics (Golec Reference Golec2002). A meta-analysis (Jost et al. Reference Jost, Glaser, Kruglanski and Sulloway2003) reports relationships between political conservatism and desire for cognitive closure (or related concepts such as intolerance of ambiguity and preference for order) in 20 different samples in an array of countries. As both are consistent with a desire for clear and definite answers, it is not surprising that religious fundamentalism also is related to preferences for closure (Lienesch Reference Lienesch1982).
Historical and cultural context plays an important role in these relationships. In some postcommunist countries individuals with a strong preference for closure are more likely to support socialist economic arrangements (Golec Reference Golec2002; see also Kossowska & van Hiel Reference Kossowska and Van Hiel2003). Thorisdottir et al. (Reference Thorisdottir, Jost, Liviatan and Shrout2007) find that psychological preferences for traditionalism and rule-following lead to right-of-center preferences in both Eastern and Western Europe. On the other hand, preferences for security lead to right-of-center orientations in the West but left-of center orientations in the East (they also find that the effects of openness on politics are particular to region). Presumably the security and familiarity associated with a particular regime style (whatever the ideology) that long shaped people's lives appeals to certain personality traits. Thus, psychological tendencies may be generally related to political beliefs but the particular features and history of a polity undoubtedly modify these relationships from country to country and era to era.
In addition to a desire for cognitive closure, variations in preference for cognitively involved activity (a different concept than cognitive ability) also seem to relate to political preferences. Cacioppo et al. (Reference Cacioppo, Petty, Feinstein and Jarvis1996) developed an instrument suitable for assessing attitudes toward cognition and Sargent (Reference Sargent2004) reports that, in two separate samples, those more comfortable with cognitive effort and attributional complexity are less supportive of punitive responses to lawbreaking.
4. Politics and implicit psychological differences
The studies summarized above show liberal-conservative differences in psychological traits and tendencies but they rely almost exclusively on self-reports. It turns out that differences correlating with political orientations also extend to measures tapping implicit, subthreshold tendencies. Such measures are designed to index variation in the manner in which individuals see, pay attention to, and process stimuli (Wahlke Reference Wahlke1979; for a good summary, see Nosek et al. Reference Nosek, Graham, Hawkins, Gawronski and Payne2010). As such, they tap concepts that are much broader than politics.
A variety of measures of directed attention are available. Common protocols such as the “Emotional Stroop,” “Dot-Probe,” and “Flanker” tasks find that threatening stimuli are consistently more distracting for conservatives (Carraro et al. Reference Carraro, Castelli and Macchiella2011, McLean et al., in press). Negative stimuli such as angry faces appear to grab the attention of conservatives more than they do liberals. Eyetracking is an even more direct way to measure attention. Dodd et al. (Reference Dodd, Balzer, Jacobs, Gruszczynski, Smith and Hibbing2012) asked participants to “free view” collages of images (selected from the widely used IAPS collection) that had been pre-rated as positive (sunsets, happy children, cute animals) or negative (vomit, houses on fire, dangerous animals). They found conservatives spent significantly more time looking at negative images and were significantly quicker to “fixate” on those images, as well. In sum, across research methods, samples and countries, conservatives have been found to be quicker to focus on the negative, to spend longer looking at the negative, and to be more distracted by the negative.
Some evidence suggests conservatives have a lower bar for deeming stimuli and situations negative. When “emotionally ambiguous” faces are shown to research participants, individuals on the political right are more likely to report that the face is expressing a threatening or dominant emotion, such as anger. Those on the political left are more likely to “see” a subordinate emotion such as surprise (Vigil Reference Vigil2010). In a study of our own, a sample of 340 U.S. adults were shown a series of pre-rated IAPS images and asked to report their evaluations from favorable to unfavorable. Consistent with expectations, conservatives perceived the negative images more negatively than did liberals (p < 0.01).
Research also reports liberal-conservative differences in word usage, implicit association tests (IATs), object categorization, and exploratory behavior. Linguist George Lakoff observes that people on the left use the language of the nurturing parent and those on the right the language of the strict parent (Lakoff Reference Lakoff2002; see also, Graham et al. Reference Graham, Haidt and Nosek2009). Compared with liberals, conservatives tend to have stronger implicit attachments to tradition, stability, long-held values, conformity, and order (Jost et al. Reference Jost, Nosek and Gosling2008). Young (Reference Young2009) finds conservatives are more likely to be “hard categorizers” and liberals “soft categorizers,” suggesting that conservatives have a lower tolerance for ambiguity and are more likely to view the world in strongly defined categories (see also Rock & Janoff-Bulman Reference Rock and Janoff-Bulman2010).
Conservative-liberal differences also appear in the way individuals extract and process information from their environments. “BeanFest” is a computer game where participants must choose to accept or reject a series of differently shaped and marked beans. If the bean is accepted, the value of beans with that same shape and marking is revealed (it could be +10 or −10) and participants are rewarded for accumulating points. Strategies of play vary widely across people: Some “accept” many beans, risking points in order to acquire information, whereas others play it safe, accepting only those beans they know to have a positive value. One of the key correlates of variations in these strategies is political orientation. Liberals are significantly more exploratory than conservatives in that they choose far more unknown bean types even though doing so runs the risk of losing points (Shook & Fazio Reference Shook and Fazio2009). Differences also show up in the learning capacities of people with different political orientations. Conservatives are better than liberals at remembering which beans are “bad,” but they are also more likely to misremember the positive beans as “bad.” In short, conservatives are more likely than liberals to follow strategies that lead them to know less about positive aspects of their environment, possibly leading them to conclude that “the world is a relatively harsh place” (Shook & Fazio Reference Shook and Fazio2009).
5. Politics and physiological differences
Liberal-conservative differences in psychology appear in a variety of tasks, samples, and countries – but do these differences extend to the realm of physiology? Research on the relationship between politics and physiology is just starting to take root and often involves neuroimaging. Much of this research focuses on identifying the parts of the brain that are differentially activated by political stimuli regardless of the participant's liberal-conservative orientation (Cacioppo & Visser Reference Cacioppo and Visser2003; Knutson et al. Reference Knutson, Wood, Spampinato and Grafman2006; Lieberman et al. Reference Lieberman, Schreiber and Ochsner2003; Westen et al. Reference Westen, Blagov, Harenski, Kilts and Hamann2006). Still, some recent research reports liberal-conservative neural differences.
Amodio et al. (Reference Amodio, Jost, Master and Yee2007) analyze conflict-related anterior cingulate cortex (ACC) activity by recording two event related potentials (ERP) for 43 participants. They employ a Go/No-Go task where participants habituate to provide a “Go” response but then have to withhold that response (a situation known to be associated with enhanced ACC activity). Self-identified conservatives in this study made more mistakes in giving the habituated response, suggesting they are inclined toward greater persistence than liberals. Moreover, Amodio et al. find that conservative participants have significantly less conflict-related neural activity than liberals when response inhibition is necessary. This is consistent with research showing that conservatives are more likely to be conscientious and to favor cognitive closure and hard categorization. As Amodio et al. put it, “political orientation, in part, reflects individual differences in the functioning of a general mechanism related to cognitive control and self-regulation” (p. 1247; for parallel findings on individuals with strong religious convictions, see Inzlicht et al. Reference Inzlicht, McGregor, Hirsh and Nash2009). Schreiber et al. (Reference Schreiber, Fonzo, Simmons, Dawes, Flagan, Fowler and Paulus2013) report that during a risk-taking task, functional magnetic resonance imaging (fMRI) on 54 participants reveals those who tend to vote Republican show greater amygdala activation, whereas individuals who tend to vote Democratic show greater insula activation.
Kanai et al. (Reference Kanai, Feilden, Firth and Rees2011) provide evidence that there are liberal-conservative differences in neural structure. Using magnetic resonance imaging (MRI), they scanned the brains of 90 college students in London (and 28 more in a replication sample) and found that self-identified liberals tend to have more gray matter in the ACC, whereas self-identified conservatives tend to have increased volume in the right amygdala. Though the amygdala has been connected to intense positive, as well as negative affect processing, these results are consistent with the aforementioned self-regulating, conflict-monitoring differences between liberals and conservatives and with differences in response to threats and facial emotions (responses that have been traced to the amygdala). These similarities lead Kanai et al. to note that their results “converge with previous work to suggest a possible link between brain structure and psychological mechanisms that mediate political attitudes” (p. 677).
Physiological differences between liberals and conservatives are not limited to brain imaging. Electrodermal activity (EDA) is one of the most widely employed measures of sympathetic nervous system activation (Dawson et al. Reference Dawson, Schell, Filion, Cacioppo, Tassinary and Berntson2007) and several studies report that negatively valenced visual stimuli increase electrodermal activity in conservatives more than in liberals (Dodd et al. Reference Dodd, Balzer, Jacobs, Gruszczynski, Smith and Hibbing2012; Oxley et al. Reference Oxley, Smith, Alford, Hibbing, Miller, Scalora, Hatemi and Hibbing2008; Smith et al. Reference Smith, Oxley, Hibbing, Alford and Hibbing2011). In some of these studies EDA response to specific image categories such as disgust correlates with specific conservative issue positions such as those related to gay marriage (Smith et al. Reference Smith, Oxley, Hibbing, Alford and Hibbing2011), whereas in other studies EDA response to a wide range of aversive images correlates with broad conservatism (Dodd et al. Reference Dodd, Balzer, Jacobs, Gruszczynski, Smith and Hibbing2012). Similar research shows that physiological response to outgroup (especially ethnic) stimuli predicts attitudes and behaviors often associated with left-right conflicts on issues like affirmative action (Dambrum et al. Reference Dambrum, Despres and Guimond2003; Vanman et al. Reference Vanman, Saltz, Nathan and Warren2004). Facial electromyography (EMG) is another technique for measuring physiological response and individuals scoring high on right wing authoritarianism tend to have greater muscle activity in the corrugator region (furrowing of the brow) when viewing negative social situations (Fodor et al. Reference Fodor, Wick, Hartsen and Preve2008). Conservatives also tend to display greater blink amplitude (movement of the orbicularis occuli muscle) in response to sudden, unpleasant, and unexpected auditory prompts (Oxley et al. Reference Oxley, Smith, Alford, Hibbing, Miller, Scalora, Hatemi and Hibbing2008).
Endocrine levels are another aspect of physiology that may relate to political orientations. Although no study to date has tested and reported a connection to location on the liberal-conservative spectrum, existing research provides an indication of the possibilities. Madsen (Reference Madsen1985) finds that whole blood serotonin levels correlate with leadership and assertiveness in group situations. Testosterone levels have been shown to decrease (Stanton et al. Reference Stanton, Beehner, Saini, Kuhn and Labar2009) and cortisol levels to increase (Stanton et al. Reference Stanton, LaBar, Saini, Kuhn and Beehner2010) when favored candidates lose an election (see also Apicella & Cesarini Reference Apicella, Cesarini, Hatemi and McDermott2011; Waismel-Manor et al. Reference Waismel-Manor, Ifergaine and Cohen2011). Testosterone levels have been associated with aggressive (simulated) decision making (McDermott et al. Reference McDermott, Johnson, Cowden and Rosen2007) and oxytocin appears to increase trust toward in-group members (Kosfeld et al. Reference Kosfeld, Heinrichs, Zak, Fischbacher and Fehr2005) but may also heighten feelings of ethnocentrism (de Dreu et al. Reference De Dreu, Greer, Van Kleef, Salvi and Handgraaf2011).
6. Negativity bias and politics
As is apparent, the list of empirically demonstrated psychological and physiological differences between liberals and conservatives is long and diverse. Additional studies are needed, however, because much of the extant physiological work is based on small, geographically constrained samples and much of the psychological work relies on college undergraduates who may have yet to form stable political attitudes. Perhaps an even greater need is for theoretical integration of this burgeoning empirical literature and that is what we hope to provide in this section, though we recognize that any effort to provide a theoretical undergirding for the findings summarized will be unavoidably speculative.
Liberals and conservatives vary in their tolerance of social equality and change, their moral foundations, their values, and even their perceptions of the nature and perfectibility of the human condition (Graham et al. Reference Graham, Haidt and Nosek2009; Jost et al. Reference Jost, Glaser, Kruglanski and Sulloway2003; Pinker Reference Pinker2002, Ch. 16; Schwartz et al. Reference Schwartz, Caprara and Vecchione2010; see also Sowell Reference Sowell1987; Tomkins Reference Tomkins and White1963). As valuable as these efforts are, questions immediately arise regarding the precursors of these differences. Why do some people say they value security and some self-expression? Why do some more than others rest their moral judgments on purity and authority? Why do some have a tragic and some a utopian vision of humankind? Why do some embrace change and others avoid it? To answer these questions, it may be useful to incorporate deeper physiological and psychological differences. After all, people's answers to the survey items used to assess moral foundations, personal values, and personality traits must come from somewhere and given the important role of subthreshold forces in political orientations, variations in physiology and deep psychology are likely to play an important role.
We believe a key factor in accounting for people's political predispositions is their orientation to negatively valenced events and stimuli. Negativity bias is the principle that “negative events are more salient, potent, dominant in combinations, and generally efficacious than positive events” (Rozin & Royzman Reference Rozin and Royzman2001, p. 297; see also Baumeister et al. Reference Baumeister, Bratslavsky, Finkenauer and Vohs2001). Essentially, this principle reflects the fact that humans generally tend to respond more strongly, to be more attentive, and to give more weight to negative elements of their environment. This tendency shows up in a wide variety of socially-relevant characteristics – everything from loss aversion (Kahneman & Tversky Reference Kahneman and Tversky1984) to quick recognition of angry versus happy faces in a crowd (Hansen & Hansen Reference Hansen and Hansen1988). People generally tend to be more attuned to negative faces, words, and social information, and both the autonomic and central nervous systems tend to have measurably higher levels of activation in response to negative than positive stimuli (Rozin & Royzman Reference Rozin and Royzman2001). Good evolutionary reasons exist for negativity bias given that negative events can be much more costly in fitness terms than positive events are beneficial; to state the obvious, infection, injury, and death curtail reproductive opportunities.
For our purposes the most notable feature of negativity bias is not that it exists but that it varies so much from individual to individual (Norris et al. 2010). That some people are more attuned to potential threats, more sensitive to sources of contagion, and more desirous of in-group protections is known intuitively and amply demonstrated by a large research literature. These individual differences seem to be stable over time and generalize to a broad category of stimuli (sounds, words, and images; see Norris et al. 2010). Previous research suggests that this individual-level variation also correlates with orientations to the social world, such as risk tolerance (Baumeister et al. Reference Baumeister, Bratslavsky, Finkenauer and Vohs2001). The connection we point out now is that the empirically demonstrated individual variation in negativity bias manifests itself not just in broad social orientations, but also in political preferences.
Negative situations are likely to relate to threats, whether microbial, predatory, or emotional, and people have widely varying orientations to threats. As we have seen, those individuals with politically conservative orientations display elevated physiological response to negative stimuli, devote more attention to negative stimuli, possess distinct self-reported psychological patterns when asked to imagine negative stimuli (i.e., give evidence of high disgust and high threat sensitivity), and perhaps harbor recognizable structural features consistent with elevated responsiveness to negative situations (distinctive substructures of the amygdala and perhaps even genetic differences such as a “short” allele of the dopamine receptor gene DRD4). Consistent with this line of thinking, Schaller and Neuberg observe that “some people seem to go through life more cognizant of threats” (quoted in Culotta Reference Culotta2012; see also Schaller & Neuberg Reference Schaller, Neuberg, Crawford and Krebs2008) before going on to suggest that these variations in general threat awareness likely correlate with political orientations.
Documented differences in response patterns extend beyond overtly threatening situations and into those that are more broadly negative. Environmental stimuli that are unexpected, ambiguous, uncertain, or disorderly also appear to generate more response and attention from conservatives than liberals at a variety of levels, including brain activation patterns, sympathetic nervous system response, cognitive behaviors, and self-reports. In many respects, compared with liberals, conservatives tend to be more psychologically and physiologically sensitive to environmental stimuli generally but in particular to stimuli that are negatively valenced, whether threatening or merely unexpected and unstructured. The consistency of these patterns across diverse research designs with diverse samples in different countries is difficult to miss. In fact, we know of no published study pointing in the opposite direction (i.e., that liberals respond more to negative stimuli or are more bothered by ambiguous or unexpected stimuli).
What could explain this connection? It is not surprising that those attuned to the negative in life might take steps to avoid it, perhaps by refraining from taking chances with the unknown, by following instructions, and by sticking to the tried and true. As an illustration, an adult subject in one of psychologist Jerome Kagan's longitudinal studies who was classified as “highly reactive” to novel, unfamiliar stimuli as a result of behavioral patterns detected when she was just four months old, summed up her approach to life by saying “I don't stray from the rules too much” (quoted in Henig Reference Henig2009). This is exactly the pattern we see in the personality data: Conservatives are less open to new experiences and are more conscientious. As a result, conservatives are less likely both to solicit new, potentially harmful information and to retain positive information concerning an object or perhaps a person or group (Castelli & Carraro Reference Castelli and Carraro2011; Shook & Fazio Reference Shook and Fazio2009). Consequently, not only do political positions favoring defense spending, roadblocks to immigration, and harsh treatment of criminals seem naturally to mesh with heightened response to threatening stimuli but those fostering conforming unity (school children reciting the pledge of allegiance), traditional lifestyles (opposition to gay marriage), enforced personal responsibility (opposition to welfare programs and government provided healthcare), longstanding sources of authority (Biblical inerrancy; literal, unchanging interpretations of the Constitution), and clarity and closure (abstinence-only sex education; signed pledges to never raise taxes; aversion to compromise) do, as well. Heightened response to the general category of negative stimuli fits comfortably with a great many of the typical tenets of political conservatism.
People who are highly responsive to negative sensory input may adopt a prevention focus by diminishing the possibility of negative events occurring or at least by mitigating the consequences of those events. The likelihood of negative encounters can be minimized through personal choices (e.g., not venturing into dangerous neighborhoods after dark) but, in modern democratic societies, also indirectly by political choices (e.g., advocating policies that are tough on criminals). Thus, it is reasonable to hypothesize that individuals who are physiologically and psychologically responsive to negative stimuli will tend to endorse public policies that minimize tangible threats by giving prominence to past, traditional solutions, by limiting human discretion (or endorsing institutions, such as the free market, that do not require generosity, discretion, and altruism), by being protective, by promoting in-groups relative to out-groups, and by embracing strong, unifying policies and authority figures (for an excellent discussion, focusing on promotion/prevention differences, see Janoff-Bulman Reference Janoff-Bulman2009). Such policies generally are associated with conservatism or the political right. On the other hand, individuals who appear to devote fewer psychological and physiological resources to negative encounters may not be as committed to avoiding them and thus may be more willing to condone new lifestyles, reductions in defense and police spending, assistance to out-groups, rehabilitation of criminals, and challenges to traditional authority, positions typically associated with the left (or with liberals as the phrase is used in the U.S.).
In sum, we posit that, due in all likelihood to a combination of genetic, early developmental, and later environmental factors, people's physiological and deep psychological responses to negative life situations vary widely. These variations, in turn, encourage but certainly do not mandate particular social tendencies and, more to the point of this article, particular political beliefs. Both degree of negativity bias and political dispositions obviously can change over the course of a lifetime but both change rather grudgingly and stability is more common than wild fluctuation. Although the theory that variations in negativity bias shape political beliefs has much to recommend it, many valid objections can be raised and we now address several of them.
6.1. Causal order
Do physiological and broad psychological traits shape political dispositions, or might political dispositions actually shape physiological and broad psychological traits? Our theory holds that political preferences are a natural spinoff of physiology and psychology but virtually all of the empirical studies summarized above are correlational and hence incapable of ruling out the possibility that immersion in a particular political climate might be powerful enough to lead to subsequent adjustments in those broader physiological and psychological traits. In theory at least, the role of parents and the general environment in, for example, encouraging or discouraging favorable perceptions of people in other countries and of alternative lifestyles could help to mold or to modify broader personality traits such as openness to new experiences and patterns of cognitive attention and physiological responsiveness to the novel, threatening, and unexpected. Political scientists, perhaps not surprisingly, tend to place politics at the center of social life and are not as likely as psychologists to see politics as emerging from pre-existing broader psychological tendencies. For example, Philip Converse's account of ideology is the most influential of the last half century and defines ideology narrowly, as an understanding of the particular labels that are popular at a given time and location and as a set of beliefs that is consistent with elite-defined, ephemeral, culturally idiosyncratic packages (Converse Reference Converse and Apter1964).
Teasing out the actual causal order requires either longitudinal or experimental data. Though studies containing such data are not numerous, they do exist and all of them provide evidence that politics results from rather than causes physiological and psychological traits such as negativity biases. Longitudinal data are especially difficult to come by but two studies connect early personality tendencies to later political beliefs. Both Block and Block (Reference Block and Block2006) and Fraley et al. (Reference Fraley, Griffin, Belsky and Roisman2012) correlate participant observation of play and other behavior at approximately age 4 with political orientations in early adulthood. Both works conclude that childhood temperament is clearly related to adult political beliefs. For example, the Fraley et al. (Reference Fraley, Griffin, Belsky and Roisman2012) study asked mothers of (then) 4-year-old children to report the extent to which their child was afraid of the dark or was upset by sad movies and found, exactly as our theory on negativity biases would predict, that a factor composed of these items was strongly and positively correlated with conservative political beliefs twenty years on. Children who eventually became liberals were more likely, on the other hand to score high on “activity and restlessness.”
In addition to findings that infants with stronger negativity biases are more likely to grow up to become political conservatives, a growing experimental literature suggests that manipulating the negative features of an environment can alter political orientations. Evidence indicates that mortality prompts induce greater conservatism (Bonanno & Jost Reference Bonanno and Jost2006; but see Castano et al. Reference Castano, Leidner, Bonacossa, Nikkah, Perrull, Spencer and Humphrey2011), as do disgusting situations and stimuli (Inbar et al. Reference Inbar, Pizarro, Knobe and Bloom2009b; Reference Inbar, Pizarro, Iyer and Haidt2012b). Negative outside-the-laboratory events such as the terrorist attacks of September 11, 2001, have also been found to make people more conservative on several issues (Echebarria-Echabe & Fernandez-Guede Reference Echebarria-Echabe and Fernandez-Guede2006; Huddy & Feldman Reference Huddy and Feldman2011; Huddy et al. Reference Huddy, Feldman and Weber2007; Nail & McGregor Reference Nail and McGregor2009). Whether manipulated in the lab or the real world, these adjustments in degree of negativity precede changes in political belief and are thus consistent with our theory. Finally, evidence also suggests a positive correlation between parasite load (and perhaps perceptions of parasite load) and conservative religious and social beliefs (Fincher & Thornhill Reference Fincher and Thornhill2012). It seems unlikely that beliefs could cause changes in parasite load so this is further evidence that the causal order is likely one in which beliefs are shaped by psychology and physiology rather than the other way around. On the basis of these longitudinal and experimental results, as well as common sense, we agree with Inbar et al. that “it seems unlikely political attitudes would shift a person's general emotional dispositions” (Reference Inbar, Pizarro and Bloom2009a).
6.2. Political orientations are too messy
Many scholars and in particular many political scientists assert that political issues and stances are so culturally elaborated that it is “incoherent” to expect a universal left-right or liberal-conservative dimension to appear (Charney Reference Charney2008). Yet as noted in the introduction, although names, labels, and issues may change disputes surrounding tradition and innovation, as well as progressivism and stability, in-groups and out-groups have always surfaced wherever politics are discussed openly. If the level of analysis shifts from issues-of-the-day, such as whether or not to invade Iraq and whether or not to build a wall along the border with Mexico, to bedrock principles of politics, such as the appropriate orientation of a given group with other groups, commonalities across cultures and centuries immediately become visible.
This is not to say any single explanatory factor, such as difference in negativity bias, is capable of accounting for variation in all political issues. In fact, one of the most exciting aspects of research in this area is its potential to identify those political predispositions that are closer to the core and those that are peripheral. The dimensionality of political beliefs is a matter of some debate with the evidence showing that being liberal or conservative on certain issues does not automatically translate into being liberal or conservative on others. More specifically, conservative positions on economic issues can be held without holding conservative positions on social issues, and separate dimensions of political orientation also have been observed for racial issues and even for “toughness” issues (examples of work on political dimensionality include Carmines & Stimson Reference Carmines and Stimson1990; Carsay & Layman Reference Carsay and Layman2002; Feldman Reference Feldman, Sears, Huddy and Jervis2003; Jacoby Reference Jacoby2009; Lewis-Beck et al. Reference Lewis-Beck, Nadeau and Elias2008; Weisberg Reference Weisberg1974). Sometimes even ideological subdimensions are not enough in that a person's views on a given economic issue might be inconsistent with that same person's views on other economic issues.
One claim is that deeper psychological and biological characteristics are less relevant to economic issues such as free market principles, tax codes, and the size of government than they are to social issues such as matters of reproduction, relations with out-groups, suitable punishment for in-group miscreants, and traditional/innovative lifestyles (Weaver Reference Weaver1992, p. 5; though see Gerber et al. Reference Gerber, Huber, Doherty, Dowling and Ha2010; Petersen Reference Petersen2009; Young Reference Young2009). As long as researchers assess political orientation by asking respondents about their bedrock principles and core issue positions rather than simply asking them to self-report their ideology (are you a liberal or a conservative) it is possible to push forward on these matters by correlating, for example, degree of negativity bias with first social and then economic issues. For example, Iyer et al. (Reference Iyer, Koleva, Graham, Ditto and Haidt2012) assert that libertarian beliefs, a label indicating liberal positions on social issues (limited government interference) and conservative positions on economic issues (limited government interference), exist because of an additional moral foundation based on liberty and it may not be likely that such a dimension springs directly from variations in negativity bias. Regardless, those predisposed toward both liberty and security might find it necessary to make difficult decisions on issues, such as the USA Patriot Act, that deal with tradeoffs between civil liberties and national defense.
A related set of issues surrounds the many individuals who are near the middle of the ideological spectrum (Fiorina Reference Fiorina2005). Are they also in the middle in terms of degree of negativity bias, neither as high as conservatives nor as low as liberals? Because most of the analyses reported rely on correlations of reasonably continuous variables (location on the ideological spectrum and degree of negativity bias) rather than analysis of variance techniques (ANOVAs) of discrete groups, this is likely the case but future research should pay more attention to possible nonlinearity in these relationships. In a similar vein, much more needs to be known about those individuals who tend to avoid politics. It is likely they have a physiological and psychological profile distinct from liberals, conservatives, and moderates. The larger point is that modern polities deal with an amazing array of issues and categories and it is foolhardy to expect a single trait such as negativity bias to account for all political variations.
One complicating aspect of current research arises from the fact that response to negative stimuli (like political dispositions) can be operationalized narrowly or broadly. Negative situations could be divided into disgust, threat, disorder, or the unexpected and even further parsing is possible. Disgust, for example, has not only been subdivided into core, contamination, and animal reminder (Haidt et al. Reference Haidt, McCauley and Rozin1994), but also into disgust relating specifically to microbes, to mating, or to morality (Tybur et al. Reference Tybur, Lieberman and Griskevicius2009). Thus, sometimes response to a relatively narrow stimulus type (e.g., a particular category of disgust) is tested for a correlation with broad political orientations (e.g., global liberalism or conservatism) and sometimes with positions on an individual issue (e.g., opposition to redistributive taxes); likewise, response to a broad stimulus type (e.g., all negative stimuli) is sometimes correlated with broad political orientations and sometimes with a highly specific issue stance. Is sensitivity to disgust pertinent only to attitudes regarding homosexuality, to attitudes on all sexually related issues (e.g., support for abstinence-only sex education, opposition to pornography, and opposition to abortion rights), or to conservatism more generally? Empirical evidence can be found for all of these conclusions. Different subcategories of negative stimuli appear to connect to certain political issues more than others.
Another approach to learning more about the nature of the relationship between elevated negativity bias and political conservatism is to note the instances in which it may not apply. Several examples come to mind. Conservatives are eager for protection from out-groups, criminals, and pathogens but less concerned with accidental shootings, environmental degradation, and poverty. Liberals' positions are just the opposite. If conservatives are universally more averse to negativity, it would seem that heightened response and attention to the negative should lead to equal amounts of concern over a leveled rainforest and a hostile out-group. We see this apparent incongruity as a valuable opportunity to refine understanding of the overall pattern. For example, it may be the case that conservatives are particularly attuned to threats by an identifiable, malevolent, volitional force such as a bad guy with a gun. Or, perhaps attitudes toward longer term and arguably more amorphous threats such as climate change, pollution, and income inequality are not as connected to negativity biases. This explanation would be consistent with conservatives' more concrete approach to life but is as yet empirically unverified.
6.3. Ultimate causes?
Of course, when we move the explanatory locus back a step from survey self-reports to deeper physiological and psychological forces, the issue immediately becomes the source of variations in these physiological and psychological traits. In other words, if negativity bias leads to the adoption of certain personality traits, basic values, moral foundations, and bedrock political principles, what causes variation in negativity bias in the first place? Obviously, answers to this question are even more speculative. Evolutionary psychologists actively debate the reasons for variation in personality traits (and presumably the same arguments would apply to political dispositions). Some (Figueredo et al. Reference Figueredo, Gladden, Vasques, Wolf, Jones, Corr and Matthews2009; Nettle Reference Nettle2006) say variations are adaptive in a niche or group selection sense; some (Tooby & Cosmides Reference Tooby, Cosmides, Barkow, Cosmides and Tooby1992) say that behavioral morphs that shape complex variables such as personality traits and political orientations are impossible in sexually reproducing species; and some (Cochran & Harpending Reference Cochran and Harpending2009; Thornhill et al. Reference Thornhill, Fincher and Aran2009) say that variations are the result of long-term differences in the relevant environment (Buss & Greiling Reference Buss and Greiling1999).
One possibility is that a strong negativity bias was extremely useful in the Pleistocene. Compared with the modern era, existence then was much more likely to be terminated prematurely at the hands of other human beings or by accidents involving wild animals or natural disasters (Pinker Reference Pinker2011). Threats were palpable and medical treatment for pathogens and injuries was ineffective. In such an environment, a heightened negativity bias would be advantageous. In modern life, on the other hand, threats are less immediate and the selection pressures for elevated negativity biases have likely been reduced, opening the door for substantial genetic variation at relevant loci.
If strong negativity biases were once selected for but now are not, it could explain why results often indicate that conservatism is in some senses better defined than liberalism. Conservatives have a negativity bias, whereas liberals do not have a positivity bias and may or may not have a negativity bias. Conservatives sometimes take umbrage at this situation, arguing that it is the result of liberal academics viewing conservatism as an aberration that needs to be explained (Will Reference Will2003). In truth, its status as a tighter, more discussed phenotype may be a result of the fact that, in contrast to proto-liberalism, proto-conservatism was once selected for.
Jencks (Reference Jencks1980) points out that relatively modest initial genetic differences across people in reading ability can easily be magnified by environmental experiences. Children proficient in reading are more likely to receive encouragement and additional opportunities to read and further hone their skills. It is likely that similar, relatively modest differences in negativity bias and associated social proclivities could be exacerbated by the environment. Individuals with slight tendencies toward caution and tradition might gravitate to those with similar tendencies, and therefore receive reinforcement for their predispositions.
A somewhat different theory that relies on group selection has been floated on occasion. It holds that societies benefit from having a mixture of those with high negativity biases and those with more modest negativity biases, of those open to out-groups and of those who are more guarded (Alford et al. Reference Alford, Funk and Hibbing2005; Nettle Reference Nettle2006). Weaver (Reference Weaver1992, p. 12) notes the dangers of a society composed entirely of what he calls “ethnocentric hawks” and “empathic doves.” Given that, except for the occasional brief (and partial) experiences such as fifth century B.C. Athens, mass-scale democracies are limited to the last couple of hundred years and even at that are still unknown in many parts of the world today (including highly populous countries such as China), the advantages of phenotypic mixtures would have to occur among the small-scale hunter-gatherer type societies that typified human existence for so long. Just as groups of spiders benefit from having a mix of social and asocial members (Pruitt & Riechert Reference Pruitt and Riechert2011) and virtually all species benefit from having individuals with different immune systems, the argument is that human groups benefit from having members who are differentially responsive and attentive to negative stimuli. If this were true, the polarization that afflicts many modern democracies may be a vestige of the mixes of the behaviorally relevant, biological predispositions that worked well in small-scale societies.
7. Conclusion: Politics and controversy
The extent to which politics evokes controversy is puzzling. Jost and Amodio ask the pertinent question: “How is it that individuals and groups can be so strongly inspired by an abstract configuration of ideas that they are willing to sacrifice even their own lives?” (Reference Jost and Amodio2012). Along with religion (another abstract configuration of ideas capable of affecting the lives of others), politics is the topic most able to produce conflict at family reunions and on the battlefield. People do not typically come to blows over whether it is better to be an introvert or an extravert, presumably because introverts do not have to worry that they will need to change their behavior as a result of the existence of extraverts. Politics, however, is unavoidably intrusive. The mere presence of liberals [conservatives] creates a very real possibility that conservatives [liberals] in the same polity will not be able to structure society in the fashion they most desire. This potential imposition of values is likely one reason politics is so emotional and explosive (Brader Reference Brader2006; Marcus et al. Reference Marcus, Newman and Mackuen2000; Redlawsk Reference Redlawsk2006; Sullivan & Masters Reference Sullivan and Masters1988; Valentino et al. Reference Valentino, Hutchings, Banks and Davis2008).
The controversial nature of politics makes research on the differences between liberals and conservatives particularly sensitive. People are quick to be defensive and to suspect that their particular ideological beliefs are being defamed. As a result, it is appropriate to note in closing that citing differences in the psychological and physiological traits of liberals and conservatives is not equivalent to declaring one ideology superior to the other. Mounting empirical evidence suggests that, compared to liberals, conservatives are more responsive and attuned to negative stimuli, patterns consistent with their tendency to advocate political solutions designed to protect against threats and disorder – real or perceived. Liberals appear not to notice, respond to, or attend to negative stimuli to the same degree, a pattern consistent with their willingness to advocate political solutions that could lead society to experience new approaches to life and governing but that could also leave society more vulnerable to threats and disorder. The relative advantages of one ideology compared to the other depend upon the circumstances. If a foreign policy threat turns out to be real, the conservative response will be extremely valuable; if it is not real, the liberal approach will be better positioned to cash in on opportunities the conservative response would miss.
Moreover, being more attuned to the dangers of the world does not make for pessimistic, fearful individuals and being less attuned to dangers does not make for carefree, hedonistic individuals. In fact, conservatives are consistently found to score higher than liberals on subjective well-being, even after controlling for socioeconomic status (Vigil Reference Vigil2010). Apparently, being responsive and attentive to negative aspects of the environment does not lead to depressive personalities. In fact, it may be that limiting the consequences of threats is a more manageable and defined goal than is pursuing novel experiences. Along these lines, it is well to remember that responding and attending to negative events is not the same thing as living in fear of them (see Aron Reference Aron1996). Turning to liberals, the desire for stimulation, self-expression and new experiences does not necessarily make for self-absorbed individuals. Liberals consistently score higher than conservatives on empathy scales (Hirsh et al. Reference Hirsh, DeYoung, Xu and Peterson2010). From an evolutionary perspective, insufficient attention and response to negative situations is clearly a problem but it is also the case that unrelenting vigilance and heightened physiological response also become problems at some point.
Finally, just as the tendency to read value judgments into the findings summarized here should be resisted, so should the tendency to conclude the results are stronger than they are. The connection of conservative political orientations and heightened orientation to negative stimuli is surprisingly consistent across designs, studies, and countries but it is also consistently modest in effect size. Many political conservatives are not particularly responsive to negative stimuli and many political liberals are. The reported effects, however, persist even when more traditional explanatory variables, such as standard sociodemographics, are included in the models. Moreover, to provide perspective, the effects of variation in negativity bias and related concepts, though modest, typically are at least as large as many of these standard variables.
A recurring feature of human history seems to be, as Atran puts it, people going “to war without understanding the transcendent drives and dreams of adversaries who see a very different world” (Reference Atran2012). Empirical evidence is increasingly documenting the psychological and physiological differences across people that can lead them to perceive the world so differently. One person focuses on threats but when facing that same situation another person focuses on opportunities. It is not surprising that these different visions of reality lead to fundamentally different sets of political preferences. By documenting that political differences are not necessarily traceable to misinformation or ignorance on the part of one side or the other, scientific understanding of the broader and deeper bases of political diversity may make it possible for Emerson's forces of tradition and innovation to live together, if not more profitably, at least less violently.
The authors acknowledge the financial support of the National Science Foundation (BCS 0826828).