Market Design Choices, Racial Discrimination, and Equitable Microentrepreneurship in Digital Marketplaces

• 成果类型: Article; Early Access
• 署名作者: Alyakoob, Mohammed; Rahman, Mohammad
• 署名单位: University of Southern California; Purdue University System; Purdue University
• 刊物名称: MANAGEMENT SCIENCE
• ISSN/ISSBN: 0025-1909
• DOI: 10.1287/mnsc.2023.01717
• 发表日期: 2025
• 关键词: discrimination; market design; quality certification; Causal Inference; Machine Learning; doubly robust methods

摘要

This study investigates the potential of market design choices, specifically quality certification, to mitigate racial discrimination in digital marketplaces. Using a regression discontinuity design (RDD) that exploits the discontinuity in Airbnb's Superhost assignment, we find that Black hosts benefit 1.5 times more from the quality certification than other hosts. Moreover, we find compelling evidence that this disproportionate increase does not come at the expense of non-Superhost Black hosts. This suggests that quality certification can effectively reduce discrimination and promote equitable outcomes. Further, our analysis utilizing a doubly robust machine learning approach reveals that the effectiveness of quality certification persists even in the prominent presence of consumer reviews, a widely used market feature and a credible user-generated quality signal. Our research underscores the judicious use of market design in fostering equity for minority entrepreneurs in digital marketplaces.