International Real Business Cycles and Increasing Returns to Scale: A Formal Analysis using Likelihood Methods

One of the more common methods used to model international real business cycles is through the use of a dynamic stochastic general equilibrium (DSGE) model. Guo and Sturzenegger (1998) argue that an increasing returns to scale production technology can improve the performance of such a model. They also argue that if increasing returns are strong enough, then sunspot equilibria are possible. In this paper, we formally test the increasing returns to scale assumption and find that a model with a constant returns to scale technology has a superior out-of-sample prediction performance over a model with an increasing returns to scale production technology. Moreover, this result is robust to the degree of returns to scale and to the persistence and variance of the shocks in the model.