Okun’s Law: a new look in the relationship between growth and unemployment.

Abstract

This study revisits the Okun’s Law relationship both in the short and long run for a set of quarterly data of 15 countries for the post war period. Alternative unit root tests reveal that both log-GDP in constant prices and unemployment rate are non-stationary for most countries. Unit root tests (Zivot and Andrews, 1992; and Lanne et al., 2002) reveal existence of non-stationarity with a single structural break in the mentioned variables. Within a cointegrating framework we demonstrate the existence of a long run relationship between log-GDP and unemployment rate with and without the presence of structural breaks. The long run Okun coefficient estimations imply that the relationship between unemployment is negative for the majority of countries, with values varying between -1% and -4%. Nevertheless, the long run estimation values are affected by the presence of structural breaks and a linear trend. The short run analysis revisits the difference Okun’s Law equation with four alternative methodologies (Quantile regression, OLS, VECM, GARCH-ML). The short run Okun coefficient estimates vary between -1% and -2%.VECM models provide the evidence that the Okun’s Law equation is bi-lateral. Evidence of non-linearity and asymmetry in the Okun’s Law equation is found from GARCH models. The percentiles distribution of the short-run Okun coefficient implies that the unemployment rate changes response differently to different sizes of growth rate.