Multivariate filter estimation of potential output for the United States

Estimates of potential output are an important component of a structured forecasting and policy analysis system. Using information on capacity utilization, this paper extends the multivariate filter developed by Laxton and Tetlow (1992) and modified by Benes and others (2010), Blagrave and others (2015), and Alichi and others (2015). We show that, although still fairly uncertain, the real-time estimates from this approach are more accurate than estimates constructed from naive univariate statistical filters. The paper presents illustrative estimates for the United States and discusses how the end-of-sample estimates can be improved with additional information