Enhancing the Micro-Foundations of Price Adjustment in Quantitative Macroeconomic Models for Policy Evaluation

点击翻译按钮获取中文摘要

There is a now a widespread literature aimed at developing macroeconomic models explicitly from underlying microeconomic theory that can confront data and be used for the evaluation of monetary and fiscal policies. While there has been considerable progress, there also has been controversy over one critical dimension: the modeling of nominal price rigidities that appear central to capturing the joint dynamics inflation and output and also to the transmission of monetary policy. While these frameworks derive aggregate relationships from individual optimization, in modeling the firms' price setting decisions they typically restrict attention to "time-dependent" rules that take the frequency of price adjustment as exogenous. Accordingly, there is an on-going controversy about how reliable these models can be, given that they take a critical feature of the model, namely the degree of price rigidity, as given.The broad goal of this project is to enrich the micro-foundations of price adjustment in macroeconomic models that can be used for empirical analysis and policy evaluation. Our starting point is to relax the strong assumption of time-dependent pricing, and instead consider the natural alternative of state-dependent pricing, where the firm is free to adjust whenever it would like, subject to a fixed adjustment cost. Of course, the virtues of state-dependent pricing have been appreciated for many years. What has stalled progress with this approach is that it leads to "Ss" pricing policies which are, in general, difficult to aggregate. This project will develop a macroeconomic framework with state-dependent pricing that is as tractable as the widely-used time-dependent models. In the first part of the project we will develop and refine an approximate solution to a simple macroeconomic framework with state-dependent pricing. It will derive a simple Phillips curve relation between inflation and real activity that evolves from state-dependent pricing. This Phillips curve will serve as a canonical baseline case, similar in spirit to the well known New Keynesian Phillips that comes from time-dependent pricing. A key step in this derivation is a "simplification" theorem that, given certain assumptions, justifies the use of an approximation for the individual firm's objective that ultimately makes the state-dependent pricing problem as tractable as corresponding time-dependent problem.The second part of the project will focus on extending the framework to capture both the microeconomic and macroeconomic evidence on price dynamics. It is well known, for example, that the baseline New Keynesian Phillips curve has difficulty capturing the high degree of persistence in inflation. The literature has developed extensions of this baseline model that significantly improve performance. Similarly, our goal is to explore extensions of the baseline state-dependent model that the lead the framework to account well for inflation dynamics. In doing so, we will consider synthesizing our state-dependent approach with alternatives that stress imperfect information and rational attention. Finally, in evaluating this model against the data, the investigators will make use of a new micro data set that contains evidence that has direct bearing on state-dependent pricing.Broader impacts: This work could lead to a number of articles in academic journals that improve our understanding of aggregate price and output dynamics. The methodology developed should be useful to applied researchers at central banks for forecasting inflation and output and also exploring the effects of alternative macroeconomic policies.