Matching Platforms and HIV Incidence: An Empirical Investigation of Race, Gender, and Socioeconomic Status.

Although recent work has examined the adverse implications of Internet-enabled matching platforms, limited attention has been paid to whom the negative externalities accrue. We examine how the entry of platforms for the solicitation of casual sex influences the incidence rate of human immunodeficiency virus (HIV) infection by race, gender, and socioeconomic status. Using a census of 12 million patients subjected to a natural experiment in Florida, we find a significant increase in HIV incidence after platform implementation, with the largest effect accruing to historically at-risk populations (i.e., African Americans) despite documented lower rates of Internet utilization. Strikingly, our analysis reveals that HIV incidence increases in historically low-risk populations as well (e.g., individuals of higher socioeconomic status) and that men and women experience similar penalties. Identifying granular effects across subpopulations allows us to offer additional insight into the mechanisms by which matching platforms increase HIV incidence. We estimate the cumulative effect of platform entry over the five-year period of the study as 1,149 additional Floridians contracting HIV at a cost of $710 million.