Absolute Convergence Under Cross-sectional Dependence: Theory and Evidence

This paper proposes tests for intercept homogeneity in stationary dynamic panel framework which are robust to cross-sectional dependence. We then argue that these tests can be used as a test for absolute convergence assuming that conditional convergence holds. The proposed tests are shown to follow standard known distributions. A detailed Monte Carlo study is then carried out to evaluate the performance of these tests in terms of size and power. These proposed tests are then applied to examine absolute convergence for the EU and OECD group of countries using both per-capita income and growth rate of per-capita income. Though conditional convergence is found for both EU and OECD countries in terms of growth rates using panel unit root tests robust to cross-sectional dependence, absolute convergence is evidenced only for the EU group of countries.