Website CV Paper
Fields
Behavioral & Experimental, Theory, Public, Environmental & Energy
Job market paper
Discriminated but not Discouraged: Investment Responses to Algorithmic Discrimination
We study how individuals adapt their investment when an algorithm evaluates groups differently. In a pre-registered online experiment (N = 553), participants are randomly assigned to an advantaged or disadvantaged identity group and choose how much to invest to improve their probability of being hired by an algorithm that holds different prior beliefs across groups. Disadvantaged participants invest 17.7% more than advantaged participants, showing no evidence of discouragement in investment despite facing less favorable evaluation. We then test two interventions: "Role Model" treatments providing information about the behavior and outcomes of same-group predecessors, and a "No Risk" treatment that conditions payment of the investment cost on being hired. The "Role Model" treatments have limited effects, whereas "No Risk" increases investment among disadvantaged participants by 24%. This effect is significantly larger than the corresponding effect among advantaged participants. Our findings show that individuals can respond to algorithmic disadvantage by increasing rather than reducing investment, and that changing the economic consequences of investment generates a stronger behavioral response than same-group social information.