The Impact of Economic Growth on Unemployment in Central Asia

Abstract

This study examines the relationship between economic growth and unemployment in Kazakhstan, Uzbekistan, Kyrgyzstan, Tajikistan, and Turkmenistan using a balanced annual panel for 1999–2024 (130 country-year observations). Unemployment is the dependent variable, real GDP growth is the principal explanatory variable, and consumer-price inflation is included as a macroeconomic control. The empirical strategy combines descriptive analysis, a Fisher-type panel unit root test based on Augmented Dickey–Fuller procedures, Fixed Effects (FE) and Random Effects (RE) estimation, the Hausman specification test, and a Modified Wald test for groupwise heteroskedasticity. The Hausman test (χ²=0.43, p=0.8051) favors the RE specification. In the preferred model, GDP growth has a positive and statistically significant coefficient (β=0.0645, p=0.033), while inflation is positive but statistically insignificant (β=0.0175, p=0.285). The positive growth coefficient is contrary to the conventional negative relationship predicted by Okun’s Law and suggests that output expansion in Central Asia was not uniformly employment-generating during the sample period. The result is interpreted as an association rather than a causal effect and may reflect capital-intensive growth, migration, informality, demographic pressure, and skills mismatch. Because groupwise heteroskedasticity is detected and unemployment is non-stationary in levels, statistical inference is interpreted cautiously. The study contributes region-specific evidence on the limits of translating aggregate GDP growth into broad labor-market improvements and highlights the importance of the sectoral composition and employment intensity of growth