Topics in Macroeconomics and Forecasting

点击翻译按钮获取中文摘要

Overall, this project addresses three areas of time series econometrics and empirical macroeconomics. The first area consists of methodological and empirical issues in business cycle forecasting. The second concerns econometric inference about the long-run properties of economic time series. The third part of the project examines econometric issues about dynamic equilibrium macroeconomic models. Recent advances in computational technology and data availability have made it possible to contemplate macroeconomic forecasting with literally hundreds of economic time series. This large number of potentially important time series raises important methodological issues. This project addresses empirical and methodological issues related to parameter instability, changing cyclical patterns, and forecasting in high-dimensional systems. Economic theory suggests and studies of individual relations (such as the money-income relation) provide evidence of instability in reduced form forecasts of the business cycle. Obtaining a better empirical understanding of this instability is of interest from the perspectives of both macroeconomic theory and economic forecasting. Understanding whether, how and when relations have changed provides an empirical foundation for theoretical investigations of macroeconomic dynamics and suggest ways to develop adaptive forecasting models. This project is part of our economics of global change research agenda. It relates to the very high priority goal of developing forecasting tools for assessing the economic consequences of global climate change. The tools that are now used by government and industry to make forecasts are subject to serious methodological criticisms. The time series methods developed in this project could provide much better forecasts of the impacts of the types of structural changes in the economy anticipated from global warming and from government policies to constrain or sharply reduce emissions of greenhouse gases.