Common Risks and Common Gaps

The Okun’s law puts forward a relation between the output and the unemployment gaps. I provide evidence of common non-linearities across the two quantities, highlighting common dynamics in higher order moments. I estimate a joint model for the dynamics of the marginal densities of the output gap and the unemployment gap, which can capture potential non-Gaussian features of the data through the time variation of the mean, variance and skewness. I postulate the Okun’s relation to hold for the predictive densities by assuming common cyclical components for the moments. I document a considerable reduction in the uncertainty surrounding estimates of the natural rate of unemployment, or NAIRU, in the US, as compared to estimates based on symmetric models. Simi- larly, accounting for time-varying skewness of the output delivers estimates of the output gap that are less uncertain and more stable over time with respect to CBO projections.