Implicit Bias Is a Public-Health Problem, and Hearts and Minds Are Part of the Solution

Abstract

Greenwald, Dasgupta, Dovidio, Kang, Moss-Racusin, and Teachman (this issue) distinguish among individual-, group-, and institutional-level interventions to reduce discrimination. We refer to the first two as “hearts and minds” approaches because they are designed to mitigate or reduce prejudice within individuals. Institutionallevel interventions, on the other hand, attempt to render individual biases irrelevant to decision processes through organizational policies and practices. On the basis of the available evidence, the authors of the target article doubt the efficacy of hearts-and-minds approaches and emphasize the prospects of institutional-level approaches from a preventative public-health perspective. As we are all steeped in the same literatures, we sympathize with the authors’ doubt. However, and despite both social and scientific attention away from individual bias and toward systemic bias, we argue that it is difficult to create systemic change by bypassing hearts and minds and that interventions attempting to change them should be preserved and refined. To that end, we offer a framework that we hope will advance research and interventions that target groups and individuals. Ultimately, we encourage an integration of individual-level and public-health perspectives to address implicit bias. Before presenting our case, it is worth noting that we have few substantive disagreements with the authors’ take on the literature as it presently sits. We are in general agreement that implicit bias is pervasive and influential, that the Implicit Association Test (IAT) is a reasonably good tool with which assess it, and that implicit bias and explicit bias are not necessarily distinct entities in people’s heads (cf. Wilson et al., 2000). We also agree, and echo work by Paluck (e.g., Paluck et al., 2021) and Lai (e.g., Lai et al., 2016), that interventions designed to weaken implicit bias (or prejudice generally) tend to disappoint once they leave the confines of a single laboratory session. Our points of departure from the authors include how implicit bias ought to be conceptualized and its modes of influence on judgments and behavior. Each of these have implications for interventions aimed at both mitigating bias’s impact and reducing bias itself. We also offer insight into two questions posed by the authors for which they believe there is insufficient data to answer: (a) “Are implicit biases introspectively (consciously) accessible?” (we think they are) and (b) “How do the association strengths measured by the IAT influence social behavior?” (we will explain below). The dual-process framework we describe may help resolve some of these disagreements, illuminate when and how individual implicit bias leads to discrimination, and explain how interventions might be improved to mitigate or reduce it. A primary point we hope to convey is that the institutional policies and practices informed by publichealth perspectives are not immune to the effects of biased individual hearts and minds. Policies and practices are proposed, adopted, enforced, and evaluated by individuals. Without buy-in by individuals who have power within organizations, policies and practices are likely to be inconsistently enforced or actively undermined (Dobbin & Kalev, 2016; Pearce & Conger, 2003). People wish to see value in their work and do not like to be told what to do. Interventions that do not seek buy-in, emphasize what not to do, or are perceived to reduce individual autonomy are likely to backfire (e.g., Does et al., 2011; Legault et al., 2011). In fact, from a historical perspective, some argue that institutional change has been driven more by individuals at the grass-roots level than by institutional leadership (e.g., Scott, 2012). For example, social-movement theory recognizes the limitations of top-down-only approaches and highlight the role of “tempered radicals,” who work within organizations to effect change from the bottom up, even without formal power (Meyerson, 2003).