Working Papers

The Indirect Effects of Educational Expansions: Evidence from a Large Enrollment Increase in STEM Majors
Link
Increasing access to education may have consequences that go beyond the effects on marginal students induced to enroll. It may change school quality, peer effects, and returns to skill. This paper studies the effects of an educational expansion on student learning, exploiting an Italian reform that changed the admission requirements for university STEM majors. Newly collected administrative data on 27,236 students indicate that the reform decreased learning in STEM fields due to overcrowded universities and negative peer effects. The analysis of long-run incomes suggests that the reform might have had a long-lasting negative effect on the returns to STEM degrees.
This is a shortened version of my job market paper "The General Equilibrium Effects of Educational Expansion". 
Coverage: IPR News, VoxEu, Washington Post, Science Careers. 
Does Compulsory Licensing Discourage Invention? Evidence From German Patents After WWI. (with Petra Moser and Joerg Baten)
Link
Compulsory licensing allows governments to license a patented technology without the consent of the patent owner. In recent years, many developing countries have used compulsory licensing to access patented drugs to fight public health emergencies. However, compulsory licensing is a clear violation of intellectual property and might decrease the incentives to innovate by increasing the risk of expropriation. In this paper, we exploit the 1918 US decision to violate enemy-owned patents to investigate the effects of this policy on German invention. We collected and digitized almost 80,000 German patents in chemical fields with application dates from 1900 to 1930. We find evidence that after 1918 German inventors produced more innovation in chemical fields with licensing. Controls for patent quality suggest that only a small share of this increase was due to lower quality, strategic patents. Firm-level analyses of patent data also reveal a significant increase in the number of research-active firms in fields with licensing. In the same fields, firms whose patents had been licensed began to patent more. These results indicate that compulsory licensing can promote innovation by encouraging competitors to enter fields with licensing.
Coverage: QSB Insight. 

Work in Progress

The Promotion of University STEM Education: Effects on Innovation and Entrepreneurship. (with Michela Giorcelli)
Many recent policies are designed to increase enrollment into university STEM (science, technology, engineering, and math) fields with the intended goal to foster innovation. The effects of these policies, however, are ex-ante ambiguous. For example, the students induced to enroll in STEM majors might have low ability and not produce any innovation. Moreover, the entry of low-ability students might convince the best STEM talents to move elsewhere, resulting in a net decrease of innovation. In this project, we examine a 1961 Italian reform that allowed a well-identified group of high school graduates to enroll for the first time in university STEM programs. As a consequence, freshman enrollment in these fields increased by more than 200 percent in a few years. The students allowed to enroll in STEM majors in 1961 were studying STEM-related disciplines in high school, but were previously denied access to university. Therefore, the reform replaced high school educated STEM workers with college educated STEM workers. We intend to isolate the effects of the policy on invention using a variety of techniques. At the individual level, we link the school and income data of students that were in school just before and after 1961 with information on each Italian patent that they owned or developed. At the national level, we intend to exploit differential increases of STEM skills by geographical location and by field of study.
The Intergenerational Effects of Educational Expansions. (with Massimo Anelli and Giovanni Peri)
Acquiring more education can have positive returns that extend to the next generation. If children of more educated parents are more likely to attend college, policies that aim at expanding access to higher education might increase social mobility across generations and raise children's expected income. However, identifying the effect of parents' education on their children's achievements is complicated by a wide array of confounding factors. For instance, education is positively correlated with parental income. Comparing the outcomes of children of more and less educated parents would thus not disentangle the influence of parental education and parental wealth. In this paper, we exploit a 1961 Italian reform that dramatically expanded college enrollment as a positive shock to the education level of one generation. We link school and income of more than 27,000 Italians that were in school just before and after the policy implementation with information on schooling and income of their children. We are able to isolate the role of parental education on the outcomes of the next generation because the reform increased parents' education with no effects on their income (Bianchi 2014). Preliminary results show that the children of the individuals that acquired more education through the reform were more likely to choose a high-paying field (business and engineering) in university.
Technology Transfer, Innovation, and Entrepreneurship. (with Michela Giorcelli)
In this project, we exploit a historical international policy to examine whether learning to innovate can be transferred across firms and countries. Starting in 1952, the US government promoted the transmission of technical information from US firms at the technological forefront to European firms recovering from World War II. In practice, this program organized consulting sessions of US experts in Europe and study trips of European technicians and engineers to the US. We collected data on 6,035 Italian firms located in 32 provinces spanning from 1930 to 1970. In addition, we collected information on all patents issued by the Italian Patent Office from 1940 to 2013. All the 6,035 Italian firms were initially declared eligible for the program. Due to restricted funding, however, only firms located in 5 provinces (902 firms) were eventually deemed eligible. To examine the effects of this policy on Italian innovation, we compare how patenting changed after 1952 among participating firms, relative to similar firms in ineligible provinces.