Innovations, patent races and endogenous growth

This paper presents a model of innovations and endogenous economic growth with two main assumptions: first, the cost of searching for innovations differs across innovations, and second, innovations take time to find. The paper shows that given these two assumptions together, competition leads to patent races and to duplication of innovative activity. The paper then shows that duplication significantly reduces the effect of scale on growth. It also shows that competitive R&D creates too much research on easy innovations, and too little research on the difficult ones. Finally, the paper shows that risk sharing might increase duplication and reduce growth.