With José G. Nuño Ledesma
Abstract: We use a matched worker-workplace dataset comprising the near universe of formal private-sector workers in Mexico to estimate the contribution of average workplace-specific wage premia, worker-level characteristics, and assortative matching on the variance of Mexico’s wages between 2004 and 2018. To this end, we regress log earnings on sets of worker and workplace fixed effects and perform a decomposition of total wage variance. We find that assortative matching explains between 16% and 19% of total wage variance, while worker- and workplace-specific factors contribute between 35% to 42% and 33% to 38%, correspondingly. The importance of workplace factors in determining wage dispersion correlates negatively with regional economic development: it is lowest in the North, Mexico’s most-developed region, and largest in the South, the country’s least-prosperous region.