Individual Wage Dynamics and Heterogeneity in Search Technology

Very Preliminary and Incomplete. In this paper, I present and estimate a flexible partial equilibrium on-the-job search model with the goal of better understanding wage dynamics. A very recent literature uses search models to understand the dynamics that govern the evolution of wages. This literature tries to solve the problem of lack of structural interpretation that affects univariate models of wage dynamics. This lack of structural interpretation can have important consequences in terms of policy implications. Unfortunately, these models are typically very simple and ex-ante heterogeneity plays a very limited role. If heterogeneity is not treated properly, a search model may give very misleading predictions, in particular with regards to calculations of lifetime inequality and wage growth decompositions. In this paper, I propose a model where ex ante heterogeneity affects the outcomes in several ways. Ex-ante heterogeneity affects directly the levels and the slopes of wages but it also affects, joint with realized labor market histories, the search technology used by individuals, namely arrival rates of wage offers, separation rates and reservation wages. The model is estimated using maximum likelihood techniques and data from the National Longitudianl Survey of Youth (NLSY). The results show that 1) the higher degree of flexibility introduced in this paper is necessary to explain the patterns observed in the data and 2) important calculations like wage growth decompositions, income mobility or lifetime wage inequality are indeed very sensitive to the level of heterogeneity that the model allows.