Keeping up with the Curve: Learning, Evolving Beliefs and the Anchoring of Expectations

(Job Market Paper) I explain the behavior of inflation, unemployment, and long-run inflation expectations in the post-war U.S. by estimating a forward-looking model in which private agents learn about structural fundamentals and the policymakers’ stabilization preferences in real-time. Learning is structural: agents understand that aggregate economic dynamics result from their optimizing behavior under imperfect knowledge. Monetary policy is conducted optimally but private agents suspect that the policymakers’ stabilization preferences are evolving, which they learn from observed policy behavior. This gives rise to a nonlinear filtering problem. The model provides a novel, information-theoretic explanation of how systematic monetary policy anchors expectations when agents are learning: an increasing emphasis on real-side stabilization has made the Fed’s policy behavior more predictable, stabilizing long-run inflation expectations while at the same time rendering short-run expectations more susceptible to supply shocks. The model offers a new explanation for the recent post-Covid inflation surge and the ``costless disinflation’’ that followed, shedding light on why it differed markedly from the crisis of the 1970s and 1980s. Download paper