Differentiating hostile from hospitable stimuli is ubiquitous in animals. The evaluative space model (ESM; Cacioppo & Berntson Reference Cacioppo and Berntson1994; Cacioppo et al. Reference Cacioppo, Gardner and Berntson1999; Reference Cacioppo, Berntson, Norris, Gollan, Van Lange, Kruglanski and Higgins2012) is a theory of the functional structure and operating characteristics of these evaluative processes across levels of the neuraxis, ranging from spinal cord reflexes to the executive functions of the frontal lobes (e.g., impulse control). According to the ESM, physical constraints limit behavioral expressions and incline behavioral predispositions toward a bipolar organization, but this bipolar organization is posited to be the consequence of multiple operations, including motivational activation function for positivity (appetition) and the activation function for negativity (aversion). The partial segregation of positive and negative evaluative processes permits greater flexibility in the mode of these evaluative processes (e.g., reciprocal activation, coactivation/coinhibition). The result is a much more flexible and adaptable affect system of evaluative processes than would be provided were evaluative processes characterized simply as a bipolar (positive–negative) activation function.

The ESM further posits that the partial segregation of the positive and negative evaluative channels in the affect system afforded evolution the opportunity to sculpt distinctive activation functions for positivity and negativity, such that the intercept for the positive activation function (i.e., the approach motivation at zero input) is higher than the intercept for the negative activation function (producing the positivity offset), and the gain for the negative activation function is higher than for the positivity activation function (producing the negativity bias). The consequence of the positivity offset is that the motivation to approach is stronger than the motivation to withdraw at very low levels of evaluative activation (thereby motivating exploratory behavior), whereas the consequence of the negativity bias is that the motivation to withdraw is stronger than the motivation to approach at high levels of evaluative activation. We focus primarily on the negativity bias.

The theoretical rationale for the negativity bias is that it is more difficult to overcome a fatal (or a near-fatal) assault than to return to an opportunity unpursued, so it is more adaptive to err on the side of caution as threats get nearer. Human taste buds respond to sweet, salty, sour, and bitter stimuli. Most can detect sweetness in approximately one part in 200, saltiness in one part in 400, sourness in one in 130,000, and bitterness in one in 2,000,000. From the perspective of the affect system, a given amount of a negative or threat-related gustatory stimulus (e.g., most poisons taste bitter) activates a stronger affective response than the same amount of a positive (e.g., sweet) gustatory stimulus. This may be more than an epicurean curiosity; it may represent differences in the activation functions for positive and negative affective processing (see reviews by Baumeister et al. Reference Baumeister, Bratslavsky, Finkenauer and Vohs2001; Cacioppo & Gardner Reference Cacioppo and Gardner1999; Cacioppo et al. Reference Cacioppo, Gardner and Berntson1997; Larsen & McGraw Reference Larsen and McGraw2011). Moreover, the combination of spatial and affective information is essential for many approach and avoidance behaviors, and thus for survival. As predicted by the ESM, Crawford and Cacioppo (Reference Crawford and Cacioppo2002) found that the incidental learning of the likely spatial location of affective stimuli is greater for negative than positive stimuli.

Given individual variation is the engine of natural selection, the ESM predicts that there are measurable individual differences in the positivity offset and negativity bias. The underlying structure and operation of the affect system is generally outside people's awareness, and these dispositional tendencies are similarly conceived as generally lying outside awareness, but like the affect system itself these dispositional tendencies should be measurable through people's judgments of and responses to affective stimuli. Indeed, temporally stable and predictive individual differences in the positivity offset and the negativity bias have been identified (Ito & Cacioppo Reference Ito and Cacioppo2005; Norris et al. Reference Norris, Larsen, Crawford and Cacioppo2011).

Participants in Norris et al. were exposed to three different sets of stimuli (pictures, sounds, and words), and during each set they were exposed to 66 stimuli, 6 of which were neutral and low in arousal, and 30 each of which vary in their extremity of pleasant or unpleasant and arousal but which were matched on these two dimensions. Ratings of each are made using the affect matrix – a 5 (positivity: zero to maximum) by 5 (negativity: zero to maximum) matrix on which participants rate each stimulus (Larsen et al. Reference Larsen, Norris, McGraw, Hawkley and Cacioppo2009). The positivity offset was indexed by the difference between the positivity and negativity ratings of the six neutral stimuli, and the negativity bias was gauged as the difference in the rating of the six most extreme unpleasant stimuli minus the rating of the six most extreme (and initially matched on extremity and arousal) pleasant stimuli. Results revealed that individual differences in the positivity offset and negativity bias were uncorrelated, temporally stable, and generalizable across ratings of pictures, sounds, and words. Furthermore, individual differences in the positivity offset predicted the spatial learning for positive stimuli, whereas individual differences in the negativity bias predicted the spatial learning for negative stimuli (Norris et al. Reference Norris, Larsen, Crawford and Cacioppo2011). Early electrical neuroimaging research indicated that the negativity bias is associated with a larger late positive potential (Ito & Cacioppo Reference Ito and Cacioppo2000; Ito et al. Reference Ito, Larsen, Smith and Cacioppo1998; Smith et al. Reference Smith, Larsen, Chartrand, Cacioppo, Katafiasz and Moran2006), and recent work suggests that the positivity offset and negativity bias are associated differently to two serotonin receptor genes (Ashare et al. Reference Ashare, Norris, Wileyto, Cacioppo and Strasser2013). In sum, although most individuals exhibit both a positivity offset and a negativity bias, this is not true for all individuals, and stable individual differences in the positivity offset and negativity bias exist and predict what is learned about the world.

Hibbing et al. posit that individual differences in the negativity bias underlie the difference between liberals and conservatives. However, they treated any evidence that negative stimuli elicit more attention, consideration, or weight than positive stimuli as bearing on evidence for a negativity bias. This conceptualization of the negativity bias conflates the various underlying mechanisms that can produce such a result and provides little guidance for quantifying this bias. The definition and psychometrics of the negativity bias provided by the ESM may provide a means of testing the Hibbing et al. hypothesis.