Economic Growth

Barro & Sala-i-Martin(2004) analyze the empirical determinants of growth. They use an cross-section empirical framework that relates the growth rate to two kinds of variables, that is, initial levels of steady-state variables(physical capital, human capital). And control variables(the ratio of the domestic investment to GDP, social infrastructure). Recent literatures suggest that panel-based model produces more efficient and consistent estimator than that of OLS or pooled regression model. This paper carries their analysis one step further and asks whether changes in the growth rate between the 1940s and the 2000s can be explained by this framework, using panel data consisted of G-5 countries. Following the implication of semi-endogeneous growth theory, we regress output growth on a constant, one year lagged output(initial income), and the determinants of steady-state income, (investment rate, population growth etc.) and the quadratic function of R&D intensity. The regression suggests more faster significant convergence. The coefficients for the determinants of steady-state income, especially for the quadratic function of R&D intensity is significant and of the expected sign.