Abstract

We estimate the union effect on wages in Chile to be between 18 and 24 per cent. We follow a two-stage procedure that allows us to correct the endogeneity of union status and to separate true from spurious dependence, by exploiting the union history of individuals using panel data. We find evidence of comparative-advantage sorting in union status, strong true state dependence and a re-distributional effect of union membership (i.e. wage gains from unionization are larger for lower-wage earners).