Understanding how deep individual differences – such as political ideology – emerge from universal, evolved cognitive biases – such as the negativity bias – is one of the key questions facing the social sciences.
From the perspective of evolutionary psychology, cognitive biases – that is, asymmetries between the subjective weights attributed to outcomes in cognitive processes and the probability of the occurrence of the outcomes – are design features that evolved to maximize expected fitness when making decisions under uncertainty (Tooby & Cosmides Reference Tooby and Cosmides1990). Uncertainty implies that errors in identifying situations accurately are possible (Haselton & Nettle Reference Haselton and Nettle2006; Johnson et al. Reference Johnson, Blumstein, Fowler and Haselton2013). By utilizing recurrent statistical relationships in the environment of evolutionary adaptedness, cognitive biases maximize fitness by weighting (1) the probability of the occurrence of errors by (2) the fitness consequences of errors if they occurred. Specifically, the negativity bias weights the probability and fitness consequences of erring by failing to identify a situation as involving potential for resource loss (a false negative) relative to the fitness consequences of erring by incorrectly identifying a potentially fitness-enhancing situation as involving potential for resource loss (a false positive). When erring by way of a false negative, the potential for fitness costs is magnified as the individual is off guard. When erring by way of a false positive, individuals shun situations that potentially involve fitness gains. Because false negatives in this case were plausibly associated with greater fitness loss over evolutionary history than false positives (Haselton & Nettle Reference Haselton and Nettle2006; McDermott et al. Reference McDermott, Fowler and Smirnov2008), this created a selection pressure for the negativity bias (even if probabilities of the errors were equal). In this perspective, cognitive biases reflect adaptive error management.
The framework of analyzing cognitive biases as adaptive error management adds to Hibbing et al.'s discussion about the relationship between individual differences in political ideology and the adaptive nature of the negativity bias in two important ways: First, the framework directly predicts the existence of the individual differences in the negativity bias and, subsequently, political ideology that Hibbing et al. struggle to reconcile with an adaptationist account. Second, the framework provides new testable predictions on the existence of contextual differences in when these differences are important and when they turn unimportant.
Only to the extent that the ratio of false negative and false positive errors in terms of expected fitness within a specific domain are constant across individuals and contexts, should biases in that domain be universally fixed (genetic noise aside). If, however, there have been systematic and evolutionarily recurrent variations across individuals and contexts, it is plausible that relevant cognitive biases evolved within a larger cognitive architecture that allows for systematic individual and contextual calibration of those biases to fit individual circumstances (see also Buss Reference Buss2009; Lukaszewski & Roney Reference Lukaszewski and Roney2011; Tooby & Cosmides Reference Tooby and Cosmides1990).
With regards to resource loss, it is highly plausible that the ratio of the fitness consequences of false negative and false positive errors, respectively, would have varied considerably and systematically from individual to individual over human evolutionary history. For example, previous research consistently shows that a range of basic individual differences influence people's abilities to guard against unexpected resource loss (e.g., Petersen Reference Petersen2013; Sell et al. Reference Sell, Tooby and Cosmides2009; Tooby & Cosmides Reference Tooby and Cosmides1996). Such differences would influence how different individuals trade off unexpected resource loss relative to the foregoing of resource gains. As consequence, such differences should be picked up by any calibrational mechanism designed to align the strength of the negativity bias with individual circumstances and, subsequently, have downstream effects on political ideology. If valid, evolutionarily recurrent correlates of loss exposure should, in part, predict modern political ideology. Some evidence suggests the existence of such a link. For example, lack of social support (Petersen Reference Petersen2013), high levels of pathogen exposure (Fincher et al. Reference Fincher, Thornhill, Murray and Schaller2008) and chronically (versus temporarily) low levels of resources (Henningham Reference Henningham1996) are all factors that ancestrally would have been associated with decreased abilities to accommodate resource loss and have all been found to be associated with social conservative ideology and/or judgments.
Just as fitness consequences of different outcomes vary systematically across individuals, the probability of unexpectedly losing resources versus unknowingly foregoing gaining resources will vary across contexts. Contexts differ in their informational load either because of differences in available information or in assigned attention. In order to maximize expected fitness in specific situations, it is likely that natural selection geared the mind to not exclusively rely on base rate estimates but, rather, to dynamically update probability estimates depending on present information. Importantly, when probabilities that specific situation identifications constitute errors approach zero and one, respectively, other asymmetries related to making these errors will become increasingly unimportant. In other words: as certainty about outcomes increases, it will be adaptive to shift away from initial decision-making biases. In terms of cognitive architecture, this argument predicts the existence of mechanisms for not just calibrating individual differences in the strength of biases but also of mechanisms designed to temporarily deactivate a relevant bias in the face of informational certainty. Consistent with this argument, research suggests that – under specific circumstances – people's political views converge independently of strong general ideologically relevant individual differences. While individual differences in ideological outlook often fuel strong political disagreement, liberals and conservatives are surprisingly likely to agree when facing certain (rather than uncertain) information about the politically-relevant events, groups or individuals facing them (see Petersen et al. Reference Petersen, Slothuus, Stubager and Togeby2011, Reference Petersen, Sznycer, Cosmides and Tooby2012; Petersen & Aarøe Reference Petersen and Aarøe2013; Tetlock et al. Reference Tetlock, Vieider, Patil and Grant2013).
This proposed notion of ideology as an individually calibrated error management system provides an important theoretical superstructure to the argument of Hibbing et al. It embeds the analysis of political ideology within an adaptationist framework that facilitates the formulation of clear predictions on the ultimate and proximate causes of ideological differences, and on the conditions under which the relationship between elevated negativity bias and political conservatism should not apply. The notion of ideology as an error management system is also normatively important. It suggests that political polarization is not inevitable: when the problems facing society are clear, conservatives and liberals are predicted to converge in the political solutions they promote.
Understanding how deep individual differences – such as political ideology – emerge from universal, evolved cognitive biases – such as the negativity bias – is one of the key questions facing the social sciences.
From the perspective of evolutionary psychology, cognitive biases – that is, asymmetries between the subjective weights attributed to outcomes in cognitive processes and the probability of the occurrence of the outcomes – are design features that evolved to maximize expected fitness when making decisions under uncertainty (Tooby & Cosmides Reference Tooby and Cosmides1990). Uncertainty implies that errors in identifying situations accurately are possible (Haselton & Nettle Reference Haselton and Nettle2006; Johnson et al. Reference Johnson, Blumstein, Fowler and Haselton2013). By utilizing recurrent statistical relationships in the environment of evolutionary adaptedness, cognitive biases maximize fitness by weighting (1) the probability of the occurrence of errors by (2) the fitness consequences of errors if they occurred. Specifically, the negativity bias weights the probability and fitness consequences of erring by failing to identify a situation as involving potential for resource loss (a false negative) relative to the fitness consequences of erring by incorrectly identifying a potentially fitness-enhancing situation as involving potential for resource loss (a false positive). When erring by way of a false negative, the potential for fitness costs is magnified as the individual is off guard. When erring by way of a false positive, individuals shun situations that potentially involve fitness gains. Because false negatives in this case were plausibly associated with greater fitness loss over evolutionary history than false positives (Haselton & Nettle Reference Haselton and Nettle2006; McDermott et al. Reference McDermott, Fowler and Smirnov2008), this created a selection pressure for the negativity bias (even if probabilities of the errors were equal). In this perspective, cognitive biases reflect adaptive error management.
The framework of analyzing cognitive biases as adaptive error management adds to Hibbing et al.'s discussion about the relationship between individual differences in political ideology and the adaptive nature of the negativity bias in two important ways: First, the framework directly predicts the existence of the individual differences in the negativity bias and, subsequently, political ideology that Hibbing et al. struggle to reconcile with an adaptationist account. Second, the framework provides new testable predictions on the existence of contextual differences in when these differences are important and when they turn unimportant.
Only to the extent that the ratio of false negative and false positive errors in terms of expected fitness within a specific domain are constant across individuals and contexts, should biases in that domain be universally fixed (genetic noise aside). If, however, there have been systematic and evolutionarily recurrent variations across individuals and contexts, it is plausible that relevant cognitive biases evolved within a larger cognitive architecture that allows for systematic individual and contextual calibration of those biases to fit individual circumstances (see also Buss Reference Buss2009; Lukaszewski & Roney Reference Lukaszewski and Roney2011; Tooby & Cosmides Reference Tooby and Cosmides1990).
With regards to resource loss, it is highly plausible that the ratio of the fitness consequences of false negative and false positive errors, respectively, would have varied considerably and systematically from individual to individual over human evolutionary history. For example, previous research consistently shows that a range of basic individual differences influence people's abilities to guard against unexpected resource loss (e.g., Petersen Reference Petersen2013; Sell et al. Reference Sell, Tooby and Cosmides2009; Tooby & Cosmides Reference Tooby and Cosmides1996). Such differences would influence how different individuals trade off unexpected resource loss relative to the foregoing of resource gains. As consequence, such differences should be picked up by any calibrational mechanism designed to align the strength of the negativity bias with individual circumstances and, subsequently, have downstream effects on political ideology. If valid, evolutionarily recurrent correlates of loss exposure should, in part, predict modern political ideology. Some evidence suggests the existence of such a link. For example, lack of social support (Petersen Reference Petersen2013), high levels of pathogen exposure (Fincher et al. Reference Fincher, Thornhill, Murray and Schaller2008) and chronically (versus temporarily) low levels of resources (Henningham Reference Henningham1996) are all factors that ancestrally would have been associated with decreased abilities to accommodate resource loss and have all been found to be associated with social conservative ideology and/or judgments.
Just as fitness consequences of different outcomes vary systematically across individuals, the probability of unexpectedly losing resources versus unknowingly foregoing gaining resources will vary across contexts. Contexts differ in their informational load either because of differences in available information or in assigned attention. In order to maximize expected fitness in specific situations, it is likely that natural selection geared the mind to not exclusively rely on base rate estimates but, rather, to dynamically update probability estimates depending on present information. Importantly, when probabilities that specific situation identifications constitute errors approach zero and one, respectively, other asymmetries related to making these errors will become increasingly unimportant. In other words: as certainty about outcomes increases, it will be adaptive to shift away from initial decision-making biases. In terms of cognitive architecture, this argument predicts the existence of mechanisms for not just calibrating individual differences in the strength of biases but also of mechanisms designed to temporarily deactivate a relevant bias in the face of informational certainty. Consistent with this argument, research suggests that – under specific circumstances – people's political views converge independently of strong general ideologically relevant individual differences. While individual differences in ideological outlook often fuel strong political disagreement, liberals and conservatives are surprisingly likely to agree when facing certain (rather than uncertain) information about the politically-relevant events, groups or individuals facing them (see Petersen et al. Reference Petersen, Slothuus, Stubager and Togeby2011, Reference Petersen, Sznycer, Cosmides and Tooby2012; Petersen & Aarøe Reference Petersen and Aarøe2013; Tetlock et al. Reference Tetlock, Vieider, Patil and Grant2013).
This proposed notion of ideology as an individually calibrated error management system provides an important theoretical superstructure to the argument of Hibbing et al. It embeds the analysis of political ideology within an adaptationist framework that facilitates the formulation of clear predictions on the ultimate and proximate causes of ideological differences, and on the conditions under which the relationship between elevated negativity bias and political conservatism should not apply. The notion of ideology as an error management system is also normatively important. It suggests that political polarization is not inevitable: when the problems facing society are clear, conservatives and liberals are predicted to converge in the political solutions they promote.