Short-run dynamics of Okun's Law in OECD countries: A CS-ARDL approach under cross-sectional dependence

Abstract

Purpose: This study aims to examine the short-run relationship between the unemployment rate and the annual per capita real GDP growth rate across 35 OECD countries within the framework of Okun's Law, using an annual panel dataset. Method: The study uses second-generation panel time series tools to account for cross-sectional dependence arising from common shocks and mutual interactions among countries. Within this context, the stationarity properties of the series are assessed using the CIPS test. The short-run dynamic relationship is estimated via the CS-ARDL framework, which controls for common factors through cross-sectional averages. Findings: The estimation results indicate that the growth coefficient is negative and statistically significant (β = −0.0861; p = 0.004), suggesting that increases in per capita real GDP growth are associated with a reduction in the unemployment rate in the short run. The lagged unemployment coefficient is positive and significant (λ = 0.6397; p < 0.001), pointing to pronounced persistence in unemployment. Conversely, the control variable for unemployment benefits is found to be statistically insignificant (p = 0.536). Conclusion/Contribution: The findings indicate that growth exhibits a reductive association with unemployment in the short run; however, the high degree of inertia in unemployment implies that the adjustment process is protracted. By utilizing second-generation estimators that control for cross-sectional dependence, this study addresses the common factor bias often neglected by first-generation methods. Furthermore, it transcends aggregate panel estimates by reporting country-level coefficient heterogeneity.