Revisions in Utilization-Adjusted TFP and Robust Identification of News Shocks

This paper documents large revisions in a widely-used series of utilization-adjusted total factor productivity (TFP) by Fernald (2014) and shows that these revisions can materially affect empirical conclusions about the macroeconomic effects of news shocks. The results suggest that for all its improvements over a traditional Solow residual, Fernald's measure of technology may be confounded by systematic measurement error, potentially invalidating the main identifying restriction of the literature that productivity reacts to a news shock only with a lag. Building on the large empirical literature documenting the slow diffusion of new technologies, we propose an alternative identification that does not rely on this zero impact restriction and instead accounts for most of the unpredictable variation in TFP at long horizons. We interpret the resulting shock as news because it predicts sustained future productivity growth while simultaneously generating a strong impact response of technological innovation indicators and forward-looking information variables. The identification is robust to different measurement issues, including the revisions in Fernald's TFP series, and to whether news affects (true) productivity with a lag or not. When applied to U.S. data, the shock fails to generate comovement in the main macro aggregates and therefore does not constitute a main source of business cycle fluctuations. The shock nevertheless accounts for a large part of macroeconomic fluctuations at medium and longer horizons and generates sharp impact responses of inflation and asset prices.