Data-Driven Learning About Trend Productivity Growth

We investigate the causes of changing productivity growth trend perceptions using a novel state-space framework for statistically efficient estimation of growth trends in the presence of data revision. Uncertainty around contemporary US productivity growth trends has been exacerbated by data revisions that typically occur several years after the initial data release, as well as by publication lags. However, the largest source of revisions in perceived trends comes from future realizations of productivity growth. This underlines the importance of estimation uncertainty in estimates of trend productivity growth.