Innovation activity determines the growth rate of living standards. Unfortunately, even state of the art models of innovation are too stylized and miss important determinants of innovation activity. Hence, their qualitative and (especially) quantitative implications are inaccurate. One of the critical determinants of the market value of an innovation is its adoption path. Technologies that are adopted intensively have a higher market value than those that are not. Given the enormous cross-country variation that exists in the intensity of technology adoption, research shows that the effect of adoption on the value of an innovation is important. Of course, the dynamics of technology adoption are also endogenous. Therefore, to explore accurately the role of various science and technology policies on innovation activity there is a need for a theory that integrates the adoption and innovation decisions in a unified framework. The framework developed in this study accomplishes this in a tractable way. The framework in this study has a couple of significant virtues. First, it is easy to extend the model to accommodate realistic sectoral and cross-country heterogeneity in the costs of innovation and adoption. Second, it is quite simple to calibrate the model parameters using cross-country information on research and development expenditures and technology adoption intensity. Once the multi-country version of the model is calibrated for each of the Organization for Economic Co-operation and Development (OECD) countries, a Nash equilibrium solution is found for a game played by the 28 social planners (one per OECD country) that choose the innovation and adoption policies that maximize their country's welfare taking as given the other countries' policies. The outcome of this exercise will be a complete and accurate science and innovation policy prescription for each OECD country based on a model that is consistent with the current cross-country distribution of research and development expenses and technology adoption intensities. There are several research and policy spillovers from this study. First, the methodology presented can be used by other researchers--in economics and other disciplines--that want to model innovation and/or technology adoption and diffusion. Second, the specific predictions from the normative exercise can be used to conduct science and innovation policy in the U.S. and in the rest of OECD countries. Third, the models developed in this study will be incorporated in syllabi designed by the researchers for a second year graduate courses. The insights will be quite intuitive and can be taught at the advanced undergraduate level. Finally, various research assistants are trained, both in dynamic macroeconomic theory and in applied macroeconomics.
