Expectations, Learning and Business Cycle Fluctuations

TLDR: In this article, an alternative theory based on learning dynamics is explored, where households and firms have an incomplete model of the macroeconomy, knowing only their own objectives, constraints, and beliefs.

Abstract: This paper explores an alternative theory based on learning dynamics. In the con text of an otherwise standard stochastic growth model, we consider an environment in which households and firms have an incomplete model of the macroeconomy, knowing only their own objectives, constraints, and beliefs. Agents are optimizing, have a completely specified belief system, but do not know the equilibrium map ping between primitive disturbances, the aggregate state of the economy, and mar ket clearing prices. By extrapolating from historical patterns in observed data they approximate this mapping to forecast variables that are exogenous to their decision problems, such as the rental rate of capital and the real wage. This belief structure has the property that beliefs affect the true data generating process of the economy which in turn affects belief formation. The economy is therefore self-referential: shifts in beliefs about future returns to labor and capital affect current market clear ing prices which in turn can reinforce beliefs. In this environment, current prices can become less informative about future economic conditions generating fluctuations in real activity. This kind of mechanism driving business cycle fluctuations is found in early writ ings on macroeconomic dynamics. For example, Alfred C. Pigou (1927), on page 122, writes: