Incorporating Diagnostic Expectations into the New Keynesian Framework

Jean-Paul L'Huillier, Federal Reserve Bank of Cleveland and Brandeis University, Sanjay R. Singh, Federal Reserve Bank of San Francisco and University of California, Davis, and Donghoon Yoo, Academia Sinica and Osaka University

Diagnostic expectations constitute a realistic behavioral model of inference. This paper shows that this approach to expectation formation can be productively integrated into the New Keynesian framework. Diagnostic expectations generate endogenous extrapolation in general equilibrium. We show that diagnostic expectations generate extra amplification in the presence of nominal frictions; a fall in aggregate supply generates a Keynesian recession; fiscal policy is more effective at stimulating the economy. We perform Bayesian estimation of a rich medium-scale model that incorporates consensus forecast data. Our estimate of the diagnosticity parameter is in line with previous studies. Moreover, we find empirical evidence in favor of the diagnostic model. Diagnostic expectations offer new propagation mechanisms to explain fluctuations.