EITM: Estimation and Analysis of Dynamic General Equilibrium Models

Northwestern University, Evanston IL

Investigators

Abstract

Formulating and evaluating economic policies are central tasks of economics. Sound economic policy is based on a fundamental understanding of the economy and the problems that policy is designed to address. This requires coherent, empirically based models that can be used to understand the economy and to evaluate both existing and proposed new policies. The author proposes five projects directed towards the estimation and analysis of dynamic general equilibrium models. Although the models ultimately will be useful for the analysis of policy more broadly, the specific focus of the research will be on monetary policy. The purpose of the first project is to better understand the role of monetary policy in determining the way technological shocks propagate through the economy. Preliminary findings suggest that these shocks affect the economy in the way predicted by standard real business cycle models, but only because of the nature of monetary policy. The analysis suggests that employment rises after a positive technology shock, and that this is due to the fact that the Federal Reserve allows the money supply to increase after such a shock. If monetary policy were instead non-accommodative, a persistent - and inefficient - fall in employment would occur in the wake of a positive technology shock. The findings highlight the potential importance of monetary policy in macroeconomic dynamics. A key part of the project, of independent interest, is the development of a set of time series tools relevant for estimating a general equilibrium model. The second project would extend the econometric estimation of general equilibrium models applied in the first project to allow for a broad range of shocks to economic fundamentals, including various types of disturbances to money demand, to technology, and to government spending. New strategies will be explored for identifying these shocks in the data. The estimated models will be used to decompose actual historical time series into components due to the various fundamental shocks. The models will be used to investigate how monetary policy has responded to shocks, and how it should respond to shocks. The third project will focus on the greatest economic convulsion experienced by the United States in the 20th century, the Great Depression of the 1930s. The author will consider Friedman and Schwartz's (1963) hypothesis that the Great Depression would not have been nearly so severe, had the US Federal Reserve followed a better monetary policy. Since the analysis concerns a counterfactual - what would have happened in the 1930s if a different monetary policy had been followed - a model is required. Accordingly, a model will be estimated which captures the basic outlines of the US economy. This will be accomplished by adding a banking sector and financial frictions to the model used in the first two projects. An estimate of the economic shocks hitting the economy in the 1930s will be incorporated into the model. The author will then use the model to investigate whether a different monetary policy response to these shocks would have prevented the economic disaster of the 1930s. In addition to shedding light on the dynamics of the Great Depression, the project is expected to produce a model that will be useful for analysis of modern economies like those of the United States or the Euro area. The fourth project will address the continuing debate over whether a small, open economy in a financial crisis should raise its interest rate and defend its exchange rate, or sharply reduce the interest rate and let the exchange rate drop. In an earlier paper, the author showed that whether one policy or the other is optimal depends on the structure of the economy. The investigator now proposes to estimate that structure using data drawn from various emerging market economies. The expectation is that the resulting structure will shed new light on the optimal strategy for the management of a financial crisis. The fifth project would explore the 'expectations trap' hypothesis that the periodic rise and fall in inflation within countries over time is a consequence of lack of commitment in monetary policy. Up to now, this hypothesis has been articulated in static models. As a consequence, it cannot address the apparent output and employment costs suffered by the US in its transition, in the early 1980s, from high inflation to low inflation. Recent computational advances will be exploited to analyze a model that incorporates the expectations trap hypothesis and appears to be capable of reproducing the transitional experience. Broader Impact: The projects will make software pertaining to the estimation and analysis of quantitative, dynamic general equilibrium models available to policy analysts in general, and to central banks in particular.