Is there really hysteresis in the OECD unemployment rates? New evidence using a Fourier panel unit root test

We investigate the hysteresis hypothesis by proposing a heterogeneous panel unit root test that allows for gradually changing trends and cross-sectional dependence (CSD) among members using flexible Fourier form. Inconclusive results from previous studies are potentially due to very restrictive specifications with homogenous break structures and/or exogenously determined abrupt breaks. seek address these limitations employing general more capable of characterising true data generation process unemployment allowing spill-over effects bootstrapping procedure accommodate CSD must be considered in globalized world. Extensive simulations suggest failure take structural breaks into account can lead misleading conclusions about whether rate is stationarity. apply our 23 OECD countries find conclusive evidence against all countries.