Income Convergence and Assimilation in a Two-Sided Matching Model

This paper develops a method suggested by the theory to estimate the structural parameters of a two-sided matching model in the context of a marriage market. A discrete marriage index is proposed to rank individuals’ measured attributes, and a matching algorithm is proposed to solve for the acceptable pool of partners numerically following the optimal reservation-match policy in the model. Estimation is based on imbedding the numerical solution of a matching model within a maximum-likelihood procedure. The structural estimation provides a unified treatment of stochastic processes that govern both marriage offers and separation and the acceptability of partners. Results from Monte Carlo experiments support the stability of the method. The usefulness of the technique is investigated through an application to the Panel Study of Income and Dynamics (1968-1993). Methods of correcting for unobserved heterogeneity are proposed.