Testing Asymmetric Unemployment Invariance Hypothesis in Africa: Evidence from Non-Dynamic and Dynamic Panel Threshold Methods

This study employs the data of 53 countries in Africa between 1991 and 2019 to examine the relationship between unemployment and labour force participation for the youth and working age population taking into consideration gender dichotomy. The aim is to determine the validity of Unemployment Invariance Hypothesis in the continent. For robustness, we employ both static and dynamic panel threshold regression methods to determine whether effect of unemployment on labour force participation varies across different unemployment level. First, out findings reveal that the relationship between unemployment and labour force participation is not linear with varying effects across gender and age groups. For female youths, DWH holds irrespective of unemployment levels, although the effect increases as unemployment rise above the threshold. In contrast, for male youth, below threshold, DWH holds in the static model but in the dynamic model AWEH holds. However, above the threshold, DHW dominates in both models. In the case of working-age population, for the women, AWEH holds below the threshold while DWH prevails above it. The difference between the static and dynamic models can be attributed to labour market adjustments, hysteresis effects and dynamics of household income. Given our findings, we suggest that target market interventions and skill development programmes should be prioritise by the governments in African countries, especially for the youth irrespective of their gender. Also, social safety net and if possible, introduce unemployment benefits across countries in the continent would be good policy options.