Essays on growth and development

Abstract

This thesis consists of three different papers under the general theme of growth and development. The first paper examines several methodological issues on the specification and estimation of the long-run or steady-state growth effects of variables. It argues that the conventional specification of the growth equation used by cross-country studies, where the dependent variable is the rate of growth of output, are inappropriate for estimating the long-run growth effects of explanatory variables. It suggests that an alternative specification based on the extended Solow (1956) growth model is suitable to estimate the long-run growth effects of variables. The growth effects of some selected variables are estimated using the conventional and alternative specification and a comparative analysis is carried out. In contrast to the estimates in the conventional specification, the empirical results show that the estimated long-run permanent growth effects in the alternative specification are small. Furthermore, the robustness of these growth effects in the two specifications is also tested with the extreme bounds analysis (EBA). The EBA results show that while the growth effects of some variables in the conventional specification are fragile, in the alternative specification the growth effects of all the variables are found to be robust by EBA. The second paper offers a new perspective on modelling the growth effects of corruption. Recent empirical studies have found the direct effect of corruption on growth is statistically insignificant. However, there is a discrepancy between these results and the intuition that corruption reduces overall productivity; as total factor productivity also depends on the quality of institutions and their efficiency. Hence, corruption should affect growth directly through productivity. This issue is addressed in this paper by developing an empirical framework based on an analytical model. The estimates show that both the direct and indirect growth effects of corruption are statistically significant, and that the direct effect is non-linear. The results are robust to different measures of corruption and when controlling for several other economic variables. Moreover, the empirical results also confirm the existence of both growth-enhancing and deteriorating effects of corruption. The third paper explores the determinants of financial development by focusing on the role of the diffusion barriers of financial technology. These barriers are measured using human genetic distance from the technology frontier. The results based on cross-sectional data for 123 countries suggest that genetic distance to the global frontier has an economically and statistically significant effect on financial development, in that countries that are genetically far from the technology leader tend to have lower levels of financial development. Genetic distance is found to have the largest effect, even after controlling for other determinants of financial development established in the literature. These findings indicate that cultural barriers to the diffusion of financial technology across borders impact financial development by influencing the follower countries’ ability to adopt and adapt innovations from the frontier.