Essays in macroeconomics.

This dissertation has two major themes: (1) the effects of monetary policy and (2) productivity. The second chapter develops a theoretical, micro-founded model of pricing decisions and acquisition of information to explain the delayed response of inflation to nominal shocks as well as other facts about pricing at both the macro and micro levels. This model provides an information-based mechanism for the amplification and propagation of nominal shocks. Specifically, it is demonstrated that the combination of (a) the aggregate price level serving as a free endogenous public signal and (b) menu costs generates rigidity in price setting in flexible economies with imperfect, dispersed information. The third chapter presents a novel approach to identify structural shocks in vector autoregressions (VARs). The key idea is to impose a factor structure on observed innovations to macroeconomic variables. The chapter develops identification, estimation, inference, and selection procedures to implement this method. When applied to the U.S. macroeconomic data, this approach resolves three well-known VAR puzzles (the price level puzzle, the exchange rate puzzle, and excess policy volatility) and explains weak contemporaneous cross-variable responses (stickiness). The fourth chapter presents analysis of monetary policy rules when central bankers encounter uncertainty while measuring economic conditions. It is shown that commitment to correcting policy mistakes caused by the mismeasurement of economic conditions reduces the inflation and the volatility of economic activity. It is argued that the Federal Reserve System has been committed to correct policy mistakes and, thus, contributed to outstanding performance of the U.S. economy in the late 1990s. The fifth chapter makes three important contributions to measuring productivity and estimating returns to scale (RTS) with the micro level data. First, it is shown that because of data limitations many of the available estimates of RTS are for the revenue function, not production function. Second, the chapter proposes a full-information estimator that models the cost and the revenue functions simultaneously and deals symmetrically with productivity, factor prices and other structural shocks. Furthermore, it is proven that the share of economic profits in revenue is a robust, non-parametric diagnostic for estimates of RTS. Third, the chapter demonstrates that, in contrast to the proposed estimator, many of the popular estimators of RTS are inconsistent with simple economic arguments and tend to perform poorly in a wide range of settings.PhDEconomicsSocial SciencesUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/126697/2/3276166.pd