The Unit Root Hypothesis for Aggregate Output May Not Hold after All: New Evidence from a Panel Stationarity Test with Multiple Breaks

Abstract

This paper investigates the hypothesis of a unit root in output for four panels of real gross domestic product (GDP) and real GDP per capita series in the Organization for Economic Cooperation and Development (OECD). For that purpose, a panel stationarity test is employed that assumes a highly flexible trend function by incorporating an unknown number of breaks in level and slope. This analysis renders clear‐cut evidence in favor of regime‐wise trend stationarity in OECD output for the three data sets of annual data, while evidence of nonstationarity is found for the data set based on quarterly data, which is composed of only seven countries over the shortest time span. Overall, the results herein stand in stark contrast to previously published results that do not control for the existence of structural change in the trend function, but accord well with those that employ panel unit root tests that allow for a single homogeneous break in the trend function.