The information content of capacity utilization rates for output gap estimates

From a theoretical perspective, the output gap is probably the most comprehensive and convincing concept to describe the cyclical position of an economy. Unfortunately, for practical purposes, the concept depends on the determination of potential output, which is an inherently unobservable variable. In this paper, we examine whether the real-time estimates of the output gap as published by the OECD can be improved by referring to measures of physical capital capacity utilization from business tendency surveys. These data relate directly to the stress on the current capacity to produce goods and services and are not revised. Our real-time panel data set comprises 22 countries at annual frequency with data vintages from June 1995 to June 2011 and a subset of 15 countries with quarterly data vintages from December 2003 to June 2011. We show that the real-time output gaps are informationally inefficient in the sense that quarterly survey data on capacity utilization available in real time can help produce estimates that are significantly closer to later output gap releases. This holds for all frequencies, maximum and balanced panels as well as for a vast majority of real-time individual country (i.e. time-series) forecasts.