We develop a real business cycle model that exhibits both the persistent and forecastable movements in consumption, hours, output, and investment which are broadly consistent with U.S. data. In the model, agents solve a signal extraction problem to learn about infrequent shifts in the drift of technology growth that are obscured by transitory shocks. We estimate a Markov regime-switching model of U.S. technology growth to calibrate the shock process. Real-time inferences about the drift of technology growth exhibit similar characteristics over time to the Index of Consumer Sentiment. Learning about the drift of technology growth provides an internal propagation mechanism. The mechanism works through the effects of revisions in the expectations about future technology on the decisions of forward-looking agents.
Keywords: Technology growth, regime switching, filtering, nonlinear impulse response, propagation
Views expressed in the paper are those of the authors and do not necessarily reflect those of the Bank of Japan or Institute for Monetary and Economic Studies.