25th September 2015
by Juan Antolin-Diaz, Thomas Drechsel and Ivan Petrella
We document a significant decline in long-run GDP growth in the United States, the bulk of this slowdown occurred prior to the Great Recession.
Using a Bayesian dynamic factor model that allows for changes in both the long-run growth of output and the volatility of business cycles, we document a significant decline in long-run GDP growth in the United States, the bulk of this slowdown occurred prior to the Great Recession. Failure to account for the decline in trend growth may be behind the persistent disappointments in forecasting economic activity. In a real-time forecasting evaluation exercise, we show that the proposed model is capable of detecting shifts in long-run growth in a timely and reliable way, and that it performs better at obtaining early forecasts ("nowcasts") of GDP that the standard model, which assumes long-run growth and volatility are constant.
Please click here to download the full paper.