Testing for Hysteresis in Unemployment for African Countries Using Wavelet Unit Root Tests

This paper uses the wavelet unit root procedures to explore the validity of the hysteresis hypothesis in unemployment for 28 African countries covering the period 1991 through 2017. In particular, the study applied the wavelet-based unit root tests proposed by Fan and Gencay. The frequency domain procedures of DWT and MODWT involve the decomposition of the variance of the time series stochastic process into the variance in its high and low-frequency series. As a benchmark, the study implemented time domain unit root tests including DF-GLS, Phillips-Perron and KPSS. The standard unit root tests produced mixed results relative to the validity of the hysteresis hypothesis for the sample countries. However, the wavelet-based unit root tests provided more consistent results as the hysteresis hypothesis was rejected in all of the sample countries, save Rwanda. Policy implications are derived from the results.