This paper estimates the effect of PhD programs on local innovation by leveraging the introduction of doctoral degrees in Italy in 1983 and a novel dataset at the individual level matching PhD applicants to inventors. I exploit the staggered opening of PhD programs across Italian provinces in a staggered Difference-in-Difference (DiD) design. Thanks to the centralised allocation of PhD programs across universities, I was able to identify a suitable control group. I find a permanent increase in local patent production of 46% in provinces that opened a PhD program. I use the PhD entry exam scores to estimate the individual returns from PhD degrees in terms of innovation in a RDD. The results are in line with the aggregate estimates and suggest PhD programs increase innovation thanks to PhD graduates becoming inventors.

Leveraging novel data from a recent policy in the Italian labor market, we construct a new measure of the intensity of firm-provided general training characterizing individual workers. We use this measure to test the relationship between labor market competition and firms’ willingness to invest in the training of general versus industry-specific skills. The results suggest that, on average, workers located in more competitive labor markets receive less general training.