When Convergence is Crisis-Driven:Evidence From Unemployment Dynamics Across Czech Districts

This article examines the evolution of persistent regional unemployment disparities in the Czech Republic and asks whether district-level unemployment rates converged or diverged between 2005 and 2024. Understanding these dynamics matters for social cohesion, economic resilience, and the design of effective employment policies. The analysis applies an absolute β-convergence framework, subsequently extended to a spatial lag specification using spatial econometric techniques. District-level unemployment data are analysed for the full period and for several sub-periods. The strongest evidence is found during the 2008–2013 downturn: both the standard and spatial models indicate crisis-driven β-convergence, with R² = 0.734. Districts that entered the crisis with initially low unemployment experienced the largest relative increases, while structurally affected districts with initially high unemployment recorded the lowest relative growth, resulting in a temporary compression of inter-district disparities. By contrast, in expansionary periods the models exhibit limited explanatory power, suggesting that unemployment dynamics are less aligned with a standard β-convergence mechanism in the absence of major macroeconomic shocks. The results underline the importance of strengthening labour-market resilience through local economic diversification, workforce reskilling, and flexibility-enhancing measures. Persistent structural disparities further point to the need for regionally differentiated policies aimed at mitigating the asymmetric regional impacts of economy-wide shocks and promoting balanced regional development.