Progress in stochastic macroeconomic modeling justifies revisiting Milton Friedman's program on the relation between macroeconomic stability and active stabilization policies. In the lecture, we use a standard new Keynesian model but depart from rational expectations by assuming that agents behave in line with adaptive learning, which increase the potential for instability in the economy.
Optimal policy under adaptive learning displays some similarity with optimal policy under commitment in the rational expectations setting. Specifically, we find that optimal policy responds in a persistent manner when expectations threaten to become unhinged. Finally, we illustrate the dynamics associated with the change from a simple regime that ignores the expectation formation, to the optimal policy that does. The results are not unlike the behavior of the U.S. economy around the Volcker transition (October 1979).