Endogenous technology cycles in dynamic R&D networks

We study the coevolutionary dynamics of knowledge creation and diffusion with formation R&D collaboration networks. The novel examination a large network over several decades reveals pronounced oscillatory (cyclical) pattern in intensity, which is not captured by existing theoretical studies. Here, we propose new model firms form collaborations others possessing complementary portfolio technologies. Innovations spillovers alter composition these portfolios time, leading to changes collaborations. show that our only able explain emergence networks, but also has important policy implications. First, demonstrate there exists critical threshold level for between collaborating must be exceeded effectively contribute economy. indicates policies promoting collaborative can successful fostering innovations if they are substantial enough so above threshold. Second, strengthening competition networks found promote fluctuations, potentially destabilizing network.