Title: Measuring US aggregate output and output gap using large datasets
Authors: Matteo Luciani - Federal Reserve Board (United States) [presenting]
Matteo Barigozzi - Università di Bologna (Italy)
Abstract: The US aggregate output and output gap are estimated by using a non-stationary dynamic factor model estimated on a large dataset of macroeconomic indicators, together with a non-parametric trend-cycle decomposition. Two main results emerge from our analysis: first, since 2010, output growth was on average 0.4 percentage point higher than measured by GDP. Our measure's higher growth has been concentrated on the first quarter of the year, suggesting that weakness in the GDP's first-quarter growth over the past several years may be due to mis-measurement rather than problems with seasonal adjustment. Second, according to our output gap estimate, while growth in the years before the financial crisis was heavily boosted by temporary factors, hence not sustainable, growth after the financial crisis was mainly driven by permanent factors, thus suggesting that as of 2017:Q4 there is still slack in the economy.