Prostitutes and Brides?

In a seminal paper, Lena Edlund and Evelyn Korn (2002) introduce a puzzling stylized fact— that prostitution is “low-skilled, labor intensive, female, and well paid’’—and offer a provocative explanation: sex workers draw a compensating differential due to the foregone opportunity to “sell’’ their fertility in the marriage market. In so doing, they not only provide the first formal model of occupational choice involving prostitution, they also draw an intriguing link between the labor market and the marriage market that holds for only one occupation. When a woman chooses to become a sex worker, she relinquishes the compensation she would otherwise receive in marriage, since taboos prevent prostitutes from marrying. Thus, even in settings where prostitution is legal, it must draw an earnings premium. Beyond drawing considerable media attention, the richness of the Edlund-Korn model has made it the starting point for economists’ discussions of sex work (e.g., Marina Della Giusta, Maria Laura Di Tommaso, and Steinar Strøm forthcoming). An especially attractive feature of the paper is that it generates a number of testable predictions. To date, however, these predictions have resisted empirical testing due to paucity of data on sex workers. We utilize two large-sample datasets on sex workers, collected in Ecuador and Mexico, which we match to national labor survey data. We corroborate the existence of a sizable earnings premium for sex work, but fail to find support for the marriage-based explanation for this premium. Sex workers are actually