Professor Pettenuzzo’s latest work on the econometrics of policy analysis is forthcoming in the Journal of Econometrics. The title of the paper is  “Granger Causality, Exogeneity, Cointegration, and Economic Policy Analysis”, and is joint work with Halbert White (UCSD).

One of the most important roles of econometric models is to help understanding the effectiveness of alternative economic policies. Indeed, the success or failure of an economic policy in achieving its objectives usually depends on the accurate and precise estimation of a vector of parameters of interest. These parameters may link policy instruments to intermediate or final targets, or may be used to forecast the future values of variables needed to formulate the policy actions. This paper investigates the use of a particular class of econometric models, Vector Autoregressions (VAR), for policy analysis. VAR were first introduced by Christopher Sims in the early 1980s, and since then have become a very successful tool for policy analysis. Their success hinges on the fact that they provide a simple yet very flexible way of describing the intricate relationships existing among economic variables.  In this paper, we provide some new methods for reliably testing the effectiveness of policy actions when some of the variables in the model are cointegrated or, in other words, move together so closely that they appear to share the same trend. As an illustration on the applicability of our methods, we investigate the effectiveness of the U.S. Federal Reserve monetary policy and its impact on U.S. inflation, unemployment, GDP growth, and oil prices over the last 25 years.


Name (required)

Email (required)


Speak your mind

Protected by Akismet
Blog with WordPress

Welcome Guest | Login (Brandeis Members Only)