Do Countries Converge to Their Steady States at Different Rates?

Recent literature revisits cross-country convergence patterns over the last six decades. Whilst the debate has been about unconditional or conditional convergence, we question whether convergence rates differ across countries. Using the same dataset as in the recent studies of Kremer et al. (NBER Macroecon Annu 36:337–412, 2022) and Acemoglu and Molina (NBER Macroecon Annu 36:425–442, 2022) (GDP per capita for 108 countries over 58 years, 1960–2017), we systematically search models where the degree of heterogeneity varies from the mean group, pooled mean group, fixed effects and pooled estimators of convergence. The Bayesian Information Criterion selects the heterogeneous model whether we use the U.S., a common factor or country-specific trends as the steady state. We estimate a multi-country technological catch-up statistical model using the U.S. as the technological frontier. We show empirically that a failure to allow for heterogeneous rates of convergence creates a bias in the convergence coefficient towards zero. The long-run elasticity to the U.S.—another convergence parameter of interest—associates positively (0.79) with the country’s average growth. Countries that learn from the technological frontier also grow faster.