International co-movements of macroeconomic variables are a source of global risk that cannot be diversified. The first subproject will develop a general equilibrium theory of how an economy populated by imperfectly informed, rational (“rationally inattentive”) agents reacts to macroeconomic shocks. The second subproject will investigate empirically the causes of inter-national business cycles, exploring in particular how fiscal and monetary policy manage, or con-tribute to business cycles.
Many economists believe that monetary policy is capable of affecting the business cycle and that it does so primarily because prices are rigid in the short run. However, recent data suggest that firms change prices rather frequently. Traditional equilibrium models — when calibrated to the empirical frequency of price changes — imply small and transitory effects of monetary policy on the business cycle. This project will develop a new branch of equilibrium models in which substantial real effects of monetary policy are possible even when prices are perfectly flexible. Agents in the model will be assumed to be imperfectly informed about the state of the economy. The innovation of the model will be that agents decide optimally what to observe, rationally choosing a delayed and dampened reaction to certain changes in their environment. The project will study how „rationally inattentive” agents react to macroeconomic shocks, and how the optimal response of such agents affects the transmission of macroeconomic shocks and the business cycle.
The project will use multivariate time series models (such as vector autoregressions and dynamic factor models) to identify the main macroeconomic forces responsible for international business cycles. The models will attempt to decompose these forces into global and country-specific components. The models will pay special attention to the question how much and via what channels fiscal and monetary policy contribute to international business cycles. The project will develop algorithms to compute the extent of uncertainty about questions of interest (via Markov Chain Monte Carlo methods).