Measuring the NAIRU with Reduced Uncertainty: A Multiple Indicator-Common Component Approach

Standard estimates of the NAIRU or natural rate of unemployment are subject to considerable uncertainty. We show in this paper that using multiple indicators to extract an estimated NAIRU cuts in half uncertainty as measured by variance. The inclusion of an Okun’s Law relation is particularly valuable. We estimate the NAIRU as an unobserved component in a state-space model and show that using multiple indicators reduces both parametric uncertainty and filtering uncertainty. Additionally, our multivariate approach overcomes the “pile-up” problem observed by other investigators.