The Cost of the Cold-Start Problem on Airbnb

Abstract

In digital markets with peer-to-peer reviews, new products encounter the so-called “cold- start” problem; Little-known products are bought too rarely and remain little known. While consumers benefit from observing reviews written by others, they do not account for the benefits of trying new products, thereby revealing the products’ quality to everyone else, in their own purchase decision. In this paper, we investigate the inefficiency arising from such insufficient social learning on Airbnb. We estimate a dynamic structural model of demand and supply for Airbnbs in Manhattan, New York, from 2016 to 2019. We then implement counterfactual tax-subsidy regimes, which boost demand for new listings relative to incumbent listings and address the cold-start problem. In our main counterfactual, we find that, on average, the rental rate of unreviewed listings should be 30% lower, leading to 25% higher demand. As a result, in the current equilibrium, the number of unreviewed listings is too large by 23%, though in total there are 2% too few listings. According to our estimates, the cold-start problem reduces social welfare by $177,241 per month, a harm that is entirely driven by a reduction in consumer surplus. In a second counterfactual, we explore tax-subsidy regimes which may also be beneficial for reasons beyond the cold-start problem.