The lost race against the machine : automation, education and inequality in an R&D-based growth model

We analyze the effect of automation on economic growth and inequality in an R&D-based growth model with two types of labor: highskilled labor that is complementary to machines and low-skilled labor that is a substitute for machines. The model predicts that innovationdriven growth leads to increasing automation, an increasing skill premium, an increasing population share of graduates, increasing income and wealth inequality, a declining labor share, and (in an extension of the basic model) increasing unemployment. In contrast to Piketty's famous claim that faster economic growth reduces inequality, our theory predicts that faster economic growth promotes inequality.