Convergence or Divergence? Bayesian Testing of Global Technology Catch-Up Models

Global technological catch-up remains constrained by absorptive capacity limitations amid persistent divergence between affluent and least-developed countries (LDCs). This study investigates the global technological convergence-divergence dynamics by comparing extended logistic and exponential catch-up models using a Bayesian non-linear hierarchical framework applied to a panel of 117 economies from 1996 to 2019 (Penn World Table 10.01). Pioneering the application of Bayesian methods to this diffusion framework, the study demonstrates that the Bayesian approach—superior to frequentist methods in handling statistical complexities—reveals convergence clubs within an otherwise divergent global pattern: emerging and advanced economies converge toward the technological frontier, while LDCs continue to diverge. Among two mechanisms of technological advancement—endogenous innovation and technology diffusion—technology adoption emerges as the dominant driver, complemented by socio-economic and institutional factors beyond human capital. The study also identifies a human capital threshold required for convergence. These findings underscore the need for countries with low human capital and structural constraints to invest in education and pursue comprehensive institutional reforms to accelerate catch-up.