Optimal Central Bank Disclosure in a Business Cycle Model

I study the social value of information in a New Keynesian model of monopolistic pricesetting with dispersed information, where the central bank can choose what information to share with the public. Under flexible prices, more disclosure is welfare improving: it reduces price dispersion and moves the economy closer to the first-best allocation. Under Calvo [1983] price stickiness, however, disclosure is detrimental to welfare, because of the additional price dispersion induced by the asynchronous adjustment in prices. If the price stickiness is due to quadratic adjustment costs, as in Rotemberg [1982], disclosure is detrimental to welfare as well: prices respond more to more precise information, increasing the adjustment cost. These results challenge the conventional wisdom found in the literature. Finally, I quantify the effects of disclosure on welfare. To do so, I bring the analysis closer to the DSGE paradigm and introduce physical capital into the model. Under flexible prices, I find that moving from no disclosure at all to full disclosure is equivalent to a permanent increase in the annual steady-state consumption level of 0.35%. However, when prices are sticky, disclosure generates a welfare loss of 0.24% in annual consumption. Thus, the social value of disclosure is sensitive to the presence of price stickiness, and disclosure has negative effects on welfare in the preferred specification.