BIM Seminar by Mohammad Rahman
Digital marketplaces remove many traditional entry barriers, especially for minorities, and facilitate micro-entrepreneurship irrespective of the personal characteristics of entrepreneurs – race, origin, ethnicity, etc. The trust mechanism in such marketplaces, however, often require identifying information of the entrepreneurs (e.g., hosts on AirBnb, Taskers on TaskRabbit, etc.), accentuating concerns about discrimination against minority entrepreneurs. In this paper, we utilize a unique quasi-random design to examine discrimination faced by minority Airbnb hosts across 10 large U.S. cities. More importantly, we evaluate the potential for a platform-controlled market design choice to alleviate such discrimination. Specifically, exploiting the discontinuity in the criteria to be a Superhost – an Airbnb certification program, we study the effect of random information shocks, enjoyed by the borderline Superhosts, on the revenue and occupancy rates of Airbnb micro-entrepreneurs (hosts). We find that an average Superhost enjoys a revenue growth of approximately 22% and increase in occupancy rate by 5% from the platform certification. Consistent with the prediction of statistical discrimination theories, borderline Black Superhosts benefit disproportionately from the information shock – an increases in revenue of approximately 55%. While the documented inequity faced by the black hosts on AirBnb is troubling, we are encouraged by the fact that platform certification programs can be effective in mitigating the discrimination faced by minority entrepreneurs on digital marketplaces.