Learning to Forecast the Hard Way -- Evidence from German Reunification

Abstract

How do firms learn to forecast future business conditions after major structural changes to the economy? How long does it take? We exploit German Reunification as a natural experiment, where firms in the East are treated with ignorance about the distribution of market states, to test a Bayesian learning framework. As predicted, we find that Eastern firms initially forecast future business conditions worse than Western ones, but this gap gradually closes over most of a decade following Reunification. The slow convergence stems from differences in forward expectations rather than realized market conditions. These results warn of costly and drawn out firm-level adjustments to contemporary regime changes, such as the US-China Trade War, COVID19, and Brexit.

Keywords: Organizational Learning, Mental Models, Expectation Formation, Business Cycle Forecasting, Transition Dynamics

JEL Classification: D21, D83, E32, E37