Regime Shifts and Gender Disparities in Global Youth Unemployment: A Bayesian Markov Switching Analysis

Youth unemployment remains a persistent global challenge, exacerbated by economic crises such as the COVID-19 pandemic. While prior studies have examined unemployment trends, few leverage advanced econometric methods to identify latent structural shifts or assess gender disparities across regimes. This study addresses this gap by applying a Bayesian Markov Switching model to ILO modelled unemployment data (2014–2024) for 9 countries, focusing on youth (15–24 years). Our objectives are twofold: (1) to detect high and low unemployment regimes and their persistence and (2) to quantify gender disparities within these regimes. Results reveal two distinct unemployment states: a stable low unemployment regime (mean = 12.4%, duration = 3.8 years) and a volatile high unemployment regime (mean = 24.7%, duration = 1.5 years). Gender gaps widen significantly during high regimes, with female unemployment exceeding male rates by 3.6% (posterior probability > 99%). These findings underscore the cyclical vulnerability of youth labor markets and the compounding effect of crises on gender inequality. The study contributes methodologically by demonstrating the utility of Bayesian regime-switching models in labor economics and offers policy insights for targeted, gender-sensitive interventions during economic shocks.