Frontiers: Discrimination Against Femininity in Headshots: A Field Experiment with AI-Enabled Controllable Stimuli Generation
Marketing Science · Thu
Abstract
We document an understudied form of discrimination based on femininity expressed in headshots. To do so, we use a generative adversarial network to create realistic headshots of people and then manipulate their femininity independently of other attributes such as pose, general facial expression, background, hairstyle, and clothes. Then, we devise an experimental design that allows us to identify the separate and combined causal impact of femininity and gender identity (proxied by gender pronouns) on real-life outcomes. In a field experiment within a naturalistic advertising environment, we find that prospective customers of an education company discriminated against femininity at various stages of the purchase funnel to a largely similar extent for men, women, and nonbinary people. The findings inform managers and policymakers of an important form of discrimination that would otherwise be underestimated by disregarding headshots. Methodologically, we introduce a novel framework for testing specific hypotheses pertaining to causal effects of treatments that are derived from unstructured data. The approach involves using natural text to readily identify hypothesis-relevant features from a generative AI model (in a particularly disentangled and interpretable representation space) and then creating realistic stimuli that vary controllably in those features. History: Catherine Tucker served as the senior editor. Funding: This work was supported by the Sanford C. Bernstein & Co. Center for Leadership and Ethics; Columbia Experimental Laboratory in the Social Sciences. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mksc.2024.1176 .
Keywords
gender discrimination, field experiment, interpretable machine learning, representation learning, generative AI, unstructured data, causal inference
Enable JavaScript for citation tools and additional paper interactions.