Forecasting US economic growth in downturns using cross-country data

Abstract

The Covid-19 pandemic has created tremendous downward pressure on economic activity and revived interest in forecasting economic growth during severe downturns. However, most dynamic factor models used to forecast GDP growth include only domestic data. We construct a large data set of 77 countries representing over 90 percent of global GDP and show that including cross-country data helps produce more accurate forecasts of US GDP growth during economic downturns, but is less helpful in normal times. We provide explanations why this is the case.