Information-driven Business Cycles: A Primal Approach

Résumé

We develop a methodology to characterize equilibrium in DSGE models, free of parametric restrictions on information. First, we define a “primal” economy in which deviations from full information are captured by wedges in agents' expectations. Then, we provide conditions ensuring some information-structure can implement these wedges. We apply the approach to estimate a business cycle model where firms and households have dispersed information. The estimated model fits the data, attributing the majority of fluctuations to a single shock to households' expectations. The responses are consistent with an implementation in which households become optimistic about local productivities and gradually learn about others' optimism.

Mots-clés

Codes JEL

Référence

Ryan Chahrour et Robert Ulbricht, « Information-driven Business Cycles: A Primal Approach », TSE Working Paper, n° 17-784, mars 2017, révision décembre 2017.