Matching and Sorting in Online Dating

Abstract:     
This paper studies the economics of match formation using a novel data set obtained from a major online dating service. Using detailed information on the users' attributes and interactions, we estimate a model of mate preferences. Based on the estimated preferences, we use the Gale-Shapley algorithm to predict the stable matches among the users of the dating site. Comparing the predicted and observed matching patterns, we find that the Gale-Shapley model explains the matches achieved by the online dating market well, and that the matches are approximately efficient (within the class of stable matches). We then explore whether the estimated mate preferences, in conjunction with the Gale-Shapley algorithm, can explain the matching patterns in offline, real-world marriages. We find that we can predict assortative mating patterns that are similar to those observed in marriages. As the Gale-Shapley algorithm predicts match outcomes in the absence of search costs, we conclude that mate preferences, not search frictions, explain much of the strong degree of sorting along various attributes in marriages.

Keywords: matching, market design, Gale-Shapley, marriage, sorting

JEL Classifications: C78, D02, D78, J12

Working Paper Series