Do Sunspots Produce Business Cycles?

We contrast the Farmer and Guo sunspots model with the Hansen indivisible labor real business cycle model using impulse responses, growth spectra, and Watson's measure of fit for calibrated models. We find that the sunspots model is better able to reproduce the typical spectral shape of growth rates found in the data. However, the model, characterized by production externalities and aggregate increasing returns, generates excessive investment volatility and overstates high frequency behavior in employment, investment and output series. The introduction of adjustment costs, in conjunction with separate externality parameters to capital and labor inputs, can reduce these weaknesses substantially, though this may require the assumption of an implausible level of increasing returns.