By Howard Lavine – University of Minnesota
In their article, Kahan, Peters, Dawson and Slovic pose a fundamental question about the nature of public opinion: Why do people often hold views that are demonstrably at odds with scientific evidence? The conventional answer in political science, stemming from Philip Converse’s (1964) seminal work, is that they lack the requisite domain-relevant knowledge or reasoning ability. In this view, average citizens cannot manage to tie their political shoelaces, and therefore have little hope of holding elites accountable for their actions. This raises serious concerns about whether public opinion is “sufficiently complete and coherent to serve as a satisfactory starting point for democratic theory” (Bartels, 2003, p. 4).
The emerging literature on “motivated reasoning” that the authors engage suggests that this ability-focused explanation is incomplete. In this view, people are often more concerned with the social and psychological implications of their beliefs than with making accurate or otherwise optimal judgments. That is, getting it right—e.g., forming attitudes that align with the facts—is often subservient to the goals of self-esteem, social belonging and the reduction of uncertainty. To grasp why this is so, consider that citizens have little influence on policy outcomes through voting or expressing opinions in surveys; thus, they have little instrumental reason to become engaged in political debates. Instead, political attitudes often serve the expressive goals of reinforcing and signaling to others one’s social identities and cultural commitments. In effect, rather than thinking “what will this policy do for me (or society)?” it is “what does expressing this opinion say about me?”
Kahan et al. examine the operation of motivated reasoning within the context of judging scientific evidence. They note that few facts amenable to empirical investigation ever become symbols of group identity (or political conflict). Thus, in most instances, the effectiveness of scientific reasoning is likely to be a simple function of individual differences in cognitive ability. However, when facts do become “suffused with culturally divisive meanings,” expressive motives are likely to outpace accuracy motives; leading people to deploy their reasoning abilities in a more psychologically strategic manner.
The authors use an elegant experimental design to discriminate between two perspectives on scientific reasoning, one rooted in accuracy motivation and the other rooted in the desire to reach politically congenial conclusions. To do this, they present research participants with two scenarios that involve the task of detecting covariation between an action and its consequences. The authors’ clever hitch is that one of the scenarios is apolitical (whether using a skin cream makes one better or worse) while the other is highly partisan (whether a law banning concealed handguns increases or decreases crime). In accord with predictions, Kahan et al. find that individual differences in numeracy facilitates covariation detection in the skin cream example. However, in accord with their motivated reasoning thesis, the role of numeracy plays a more complex role in the partisan scenario. Here, the authors find that respondents use their cognitive abilities selectively, such that numeracy facilitates accurate covariation detection only when the correct answer validates the individual’s political orientation (and when a cursory glance at the evidence suggests otherwise).
What does this experimental demonstration imply about the dynamics of mass opinion? For one thing, it suggests that exposing people to the same body of evidence is unlikely to produce belief convergence among political opponents. In fact, as Kahan et al. find, the impact of numeracy in the gun control scenario actually leads to belief polarization. More generally, as partisanship becomes increasingly linked to cultural and lifestyle issues, societies are likely to become more socially polarized in ways that seem only tangentially linked to politics. Second, Kahan et al.’s findings raise theoretically interesting questions about how cognitive ability and goals interact in the production of public opinion. As prior work has shown (Lodge and Taber 2013), reasoning is not always deployed in the service of accuracy; when social and psychological goals are perceived as providing more immediate benefits, reasoning ability and accurate political judgment are likely to become uncoupled.
Third, the identity-based approach raises the question of whether citizens are at all concerned with policy substance. As Lilliana Mason (2015, p. 58) writes, “we are a nation of partisans who are prejudiced against each other, active just for the sake of winning, and increasingly angry. We might believe that we are responding to specific policy disputes, but to a very real extent we are also being driven by an automatic, basic need to defend our social group.” Partisan debates—even those over objective facts—are increasingly viewed as a struggle between in-group and out-group, animated by the desire for “positive distinctiveness”—a belief that one’s party is different and better than the other party (Tajfel and Turner 1979). This suggests that citizens may use policy conflict as an arena to litigate more basic struggles over the relative status of competing cultural groups.
In the end, Kahan et al.’s findings suggest that substantive political debate will suffer when policy debates become infused with concerns about identity conflict, creating a serious problem for democratic accountability. Although contemporary scholars are only now grappling with how the psychological distorts the political, this problem has afflicted human beings for a long time. As Talmudic scholars noted some fifteen hundred years ago, we do not see things as they are; rather, we see them as we are.
Read the full article from Kahan et al. Motivated numeracy and enlightened self-government for free in the first issue of Behavioural Public Policy here.