COMPUTATIONAL SOCIAL SCIENCE

Department of Computational Social Science Seminar Abstract

Friday, 23rd October: 3.00pm

Patents and Search Costs in a Multi-Agent Model of Endogenous Growth

Michael Makowsky, Department of Economics, Towson University

 

The recent literature on patent length and economic growth is characterized by a reliance on the representative agent modeling amidst monopolistic competition. This reliance obviates the importance of search costs, heterogeneous information, and the prospect of firms failing. To address this limitation we build a multi-agent model of heterogeneous firms producing a homogeneous good. A set of autonomous households serve as both consumers of the good and a fixed supply of labor. Increasing returns to scale are derivative of endogenously produced technology, but the market remains competitive due to imperfect information and costly search. Investment in research is rewarded by rents which accrue to private knowledge that is excludable due to patent rights. The benefits from multi-agent modeling is demonstrated in the non-monotone relations emergent within the model. We find that the impact of the length of patents is strongly connected to costs of search. At the lowest levels of search costs tested, an additional step of patent length has a net positive impact on growth, while at the highest level of search costs, an additional step of patent length has a net negative effect. Further, patent length has a differing impact across the distribution of economic growth. In lower growth regimes, longer patents have little, and sometimes negative, impact on growth. In the middle and upper quantiles, patents have a strong positive impact on growth, but in the highest quantile there is strong negative correlation. Search costs have a negative impact on simulated economic growth in all model simulations.