The effects of educational mismatch on inventor productivity. Evidence from Sweden, 2003-2010

This paper investigates the influence on inventor productivity of the imperfect occupational match, measured as the number of years of education in excess and in deficit to the required level (educational mismatch). The empirical model draws on a unique database that matches information about individual inventor characteristics, such as age, experience and gender, with patenting performance in Sweden over the period 2003-2010. The results suggest that over-educated (OE) inventors file a number of patents higher than inventors who are appropriately matched (RE), but perform poorly than well-matched inventors who hold a similar level of education. Conversely, under-educated (UE) inventors file a total number of patents lower than inventors who are well-matched (RE), but more than well-matched ones who hold the same level of education. These results conform to the hierarchical pattern of ORU model, well-documented in the literature relating the employees’ wages to educational mismatch (i.e. RE>OE>UE). We find that significant differences in returns to education across match and mismatch categories remain even after controlling for individual ability. Our findings are robust to controlling for differences between younger and older inventors, geographical areas and industry of work.