Hibbing et al. provide a helpful review of the psychological underpinnings of individual differences in political ideology, especially between social liberals and social conservatives (whom for brevity we refer to as “liberals” and “conservatives,” respectively). They conclude that the principal variable underlying the difference between liberal and conservative attitudes is negativity bias.

Although we share Hibbing et al.'s view that disparities in political ideology are rooted partly in dispositional differences, we contend that they paint with too broad a brush. The data point not to global differences in negativity bias, but to differences in threat bias, most likely emanating from differences in fearfulness. In fairness, Hibbing et al. at times describe the difference between liberals and conservatives as stemming from differential sensitivity to threat, quoting Schaller and Neuberg (Reference Schaller, Neuberg, Crawford and Krebs2008): “Some people [conservatives] seem to go through life more cognizant of threats” (p. 405). Yet elsewhere, they portray this difference as originating from an overarching temperamental difference in negativity: “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” (sect. 6, para. 6).

The difference between negativity bias and threat bias is hardly semantic. The personality literature points consistently to the existence of largely orthogonal higher-order dimensions of negative emotionality (NE) and Constraint, the latter of which falls on the opposite pole of Disinhibition (Tellegen & Waller Reference Tellegen and Waller2008). NE is a pervasive dimension, similar to but broader than neuroticism, which reflects the propensity to experience unpleasant affects of many kinds, including anxiety, irritability, and mistrust. As Watson and Clark (Reference Watson and Clark1984) noted, individuals with elevated NE tend to dwell on the negative aspects of life and attend selectively to unpleasant stimuli. In contrast, Constraint is a disposition toward fearfulness and response inhibition that, according to some theorists (e.g., Fowles Reference Fowles and Sutker2002; Lykken Reference Lykken1995), reflects the activity of the Behavioral Inhibition System, a brain-based circuit that mediates sensitivity to signals of punishment and uncertainty (Gray & McNaughton Reference Gray, McNaughton and Hope1996).

Most evidence suggests that Constraint, more than NE, is the principal nexus of individual differences in threat sensitivity, especially when perceived dangers are relatively clear-cut (Depue & Spoont Reference Depue and Spoont1986). For example, individuals with elevated Constraint and its constituent traits, particularly harm-avoidance/fear, exhibit pronounced fear-potentiated startle (Kramer et al. Reference Kramer, Patrick, Krueger and Gasperi2012; Vaidyanathan et al. Reference Vaidyanathan, Patrick and Bernat2009) and habituate slowly to startle-provoking stimuli (LaRowe et al. Reference LaRowe, Patrick, Curtin and Kline2006). In contrast, NE is not consistently related to avoidance reactions to threatening stimuli, including gruesome imagery (Watson & Clark Reference Watson and Clark1984). The independence of NE and Constraint parallels the distinction between trait anxiety and trait fear, respectively (Sylvers et al. Reference Sylvers, Lilienfeld and LaPrairie2011). Trait anxiety appears to reflect a disposition to react to uncertain threats, whereas trait fear appears to reflect a disposition to react to certain threats. In factor analytic studies, trait anxiety loads primarily on NE, whereas trait fear loads primarily on Constraint (Church & Burke Reference Church and Burke1994; Tellegen & Waller Reference Tellegen and Waller2008).

These two higher-order dimensions are conflated in much of Hibbing et al.'s analysis. This confusion is problematic, because the literature suggests that liberals and conservatives differ in threat sensitivity, presumably reflecting individual differences in Constraint (see also Jost et al. Reference Jost, Glaser, Kruglanski and Sulloway2003), but not in their attunement to the negative. For example, studies in both the U.S. and Europe reveal that conservatives are either essentially identical to liberals in NE (Butler Reference Butler2000; Caprara et al. Reference Caprara, Barbaranelli and Zimbardo1999; Reference Caprara, Schwartz, Capanna, Vecchione and Barbaranelli2006; Carney et al. Reference Carney, Jost, Gosling and Potter2008; Chirumbolo & Leone Reference Chirumbolo and Leone2010; Kossowska & van Hiel Reference Kossowska and Van Hiel1999) or significantly lower than liberals in NE (Gerber et al. Reference Gerber, Huber, Doherty, Dowling and Ha2010). Vigil (Reference Vigil2010) similarly found that compared with liberals, conservatives reported modestly but significantly lower levels of emotional distress and frequencies of crying.

In contrast to the absence of clear-cut differences in NE, Carney et al. (Reference Carney, Jost, Gosling and Potter2008) found that the principal correlates of political ideology within the Five Factor Model of personality (FFM) are in the dimensions of Conscientiousness and Openness to Experience, with liberals tending to be somewhat lower in most facets of the former and somewhat higher in most facets of the latter (see also Caprara et al. Reference Caprara, Schwartz, Capanna, Vecchione and Barbaranelli2006; Gerber et al. Reference Gerber, Huber, Doherty, Dowling and Ha2010). Notably, Constraint is best accounted for by Conscientiousness and Openness within the FFM (Church Reference Church1994; cf. Digman Reference Digman1997). In sum, the principal difference between liberals and conservatives appears to lie not within with the domain of NE, but rather within Constraint and probably fearfulness in particular, manifesting itself in differential sensitivity to reasonably clear-cut threats.

This alternative conceptualization is important for at least three reasons. First, it clarifies the primary dispositional differences between liberals and conservatives, and directs efforts to understand the etiology of political ideologies away from basic affective dimensions and toward threat sensitivity. It also raises a plethora of questions, such as why the links between threat sensitivity and political affiliation are only modest, suggesting the presence of unidentified modifying variables. Second, this conceptualization may help to avert the pejorative connotations often associated with research on personality differences in political ideology (e.g., York Reference York2003). The assertion that conservatives are globally negatively biased bears few implications for adaptive functioning. In contrast, the proposition that conservatives are especially attuned to threat is not inherently disparaging, as certain hazards are genuine and necessitate attention (Barlow Reference Barlow2004). Hence, a threat bias interpretation reminds us that neither political view is inherently healthier than the other. Third, this perspective may be helpful in crafting messages designed to bridge the partisan divide (Abramowitz Reference Abramowitz2010). For example, a threat bias interpretation may imply that conservatives will be most readily persuaded by communications reassuring them that dangers arising from policy changes (e.g., immigration reform) are less dire than initially envisioned. Conversely, liberals may respond best to communications that leverage psychological attributes other than threat, such as perceived fairness (Haidt Reference Haidt2012). In this way, a more precise characterization of the wellspring of liberal-conservative differences may promote a more constructive dialogue between individuals of competing political ideologies.