How patent rights affect university science

How do intellectual property rights influence academic science? We investigate the consequences of the introduction of software patents in the United States on the publications of university researchers in the field of computer science. Difference-in-difference estimations reveal that software scientists at US universities produced fewer publications (both in terms of quantity and quality) than their European counterparts after patent rights for software inventions were introduced. We then introduce a theoretical model that accounts for substitution and complementarity between patenting and publishing as well as for the direction of research. In line with the model's prediction, further results show that the decrease in publications is largest for scientists at the bottom of the ability distribution. Furthermore, we evidence a change in the direction of research following the reform toward more applied research.