Abstract

Economic literature has been documenting instances of discrimination on prominent online platforms. The absence of individual information prompts users to relay on ethnic or racial prejudice. In this paper, we show that an online reputation system mitigates the problem of incomplete information and in consequence, decreases ethnic discrimination. Using data from a popular ride-sharing platform, we find that minority entrants make 12% less revenue than non-minority entrants; this difference narrows down as they receive reviews. To establish a causal link between reviews and the reduction of discrimination, we exploit demand shocks caused by train strikes. Furthermore, we estimate a dynamic model of moral hazard to show how minority drivers use the reputation system to counter initial discrimination; when setting the first price, they offer a discount of 7.2%, compared to optimal static prices. We also argue that they exert additional efforts to receive higher grades. Finally, we provide two counterfactual experiments: a market without discrimination and a platform with ethnicity-blind profiles.

Reference

Emil Palikot, “The impact of online reputation on ethnic discrimination”, IAST Lunch Seminar, Toulouse: IAST, September 24, 2019, 11:45–12:45, room MS 001.