Learning New Technology: the Polarization of the Wage Distribution

This paper presents novel evidence regarding the relationship between technological progress, occupational tasks and wage inequality. By applying a counterfactual quantile regression analysis to historic U.S. data, we show that the evolution of wage inequality in the lower echelon of the wage distribution was due entirely to a reduction of within-group wage inequality, which was determined, in turn, by more homogeneous remuneration paid to workers performing routine tasks. Changes in the differential between the remuneration paid to technology-complementary and technology-substitute tasks had only a negligible impact on wage inequality among low-wage workers, which casts some doubt on the validity of basing a theory of wage inequality on routinization-biased technical change operating through a labor demand channel. To reconcile the routinization hypothesis with the data, we develop a model in which skill-heterogeneous workers face endogenous occupational choices and learning costs in connection with operating a new technology. Even in the absence of changes in wage differentials, the model argues that technical change can generate an empirically consistent non-monotone effect on wage inequality by affecting the average level of skills within different groups of workers.