Essays on inflation dynamics and consumption behavior.

摘要

This dissertation studies the time-series behavior of aggregate price level and consumption. A detailed summary of the main results is provided in Chapter 1. Chapter 2 examines the long-run effects of supply shocks (such as oil shocks) on inflation in the United States. I demonstrate that the persistence of supply shocks in U.S. inflation fell considerably during the period of Volcker's disinflation (1979–1982). My empirical results suggest that the difference between the pre-Volcker and post-Volcker periods is attributable to the change in the behavior of inflation expectations—agents expected shocks to persist in the pre-Volcker period, but not in the post-Volcker period. I construct a simple model of how different monetary policies lead to different persistence equilibria.; Chapter 3 explores whether habit formation in consumer's preferences can explain two failures of the standard permanent income model: correlation of consumption growth with predicted income and lagged consumer sentiment. I establish that the sensitivity of consumption growth to both predicted income and sentiment can be to a large extent reinterpreted as a sluggish response of consumption to news. This is consistent with a model where consumers form habits about the level of their consumption. I examine the effects of the recent tax cut in the U.S. on aggregate consumption. The PIH model predicts that consumption immediately responds one-for-one to a permanent tax cut. However, if consumers form consumption habits, they are likely to respond to the tax cut slowly. Under realistic assumptions, the immediate MPC out of the tax cut is only 30%.; In Chapter 4, which is a joint work with Prof. Christopher Carroll, we reinterpret findings from the third chapter of this dissertation using a new model of consumption behavior that does not assume habit formation in consumer's preferences. We present a model where the degree of serial correlation in consumption growth is an approximate measure of the fraction of the population that does not update its macroeconomic expectations in any given period. When we estimate the model, only about 25% of households are up-to-date in their information set in any given quarter.