Dr. Yuen Ho

Human beings have implicit biases, across gender, race, and caste, which influence how they interact with others and the decisions they make. The impacts of bias across many domains are well documented, from education, to hiring, job performance, and access to credit or healthcare. Economics as a field has traditionally characterized discrimination as being based on preferences (taste-based or paternalistic) or beliefs (statistical-based). However, we propose that implicit bias would more accurately be modelled in terms of System 1 versus System 2 thinking, or automatic and intuitive versus more deliberate and logical decision making. In this framework, implicit biases are automatic and intuitive, it takes less mental effort to be biased. Likewise, a decision maker needs to expend more cognitive effort in order to overcome their implicit bias and make less discriminatory decisions. It directly follows that when a decision maker"s cognitive bandwidth is depleted, such as from fatigue or scarcity, that they should exhibit more biased behavior. Framing discrimination in this way thus leads to novel predictions not found in traditional economic models. We propose to test this hypothesis using a field experiment in Uganda, where we experimentally vary a decision maker"s cognitive stress, and examine the subsequent effects on bias in their decisions.