Economic Growth and Convergence in the Very Long-run: The Case of Emerging Economies with a Focus on Brazil

Abstract

This doctoral thesis contains four essays on economic growth and real convergence of a relatively unexplored context, that of the Emerging Economies (EEs), focusing on the case of Brazil and embracing an overlooked very long-time span (over 100 years). The analyses encompass a diversity of methodological approaches, including bibliometric techniques, descriptive/ narrative review, and econometric time series methods. In the first essay, a bibliometric review systematically quantifies the empirical studies published on economic growth from 1991 to 2020. The main findings entail the need to encourage more scientific attention devoted to a neglected group of EEs - the Latin American and the Caribbean - and redirect the empirical growth agenda to analyze the key determinants of economic growth in the very long run (more than 100 years). The second essay provides a narrative review of Brazil's economic growth and real convergence from a very long-run perspective (1822-2019), organized by periods and sets of determinants. This historical economic analysis made it possible to identify some critical factors underlying the Brazilian backwardness: macroeconomic instability, deficient human capital formation, and poor institutional conditions. To complement the previous descriptive analysis, the third essay empirically discusses the joint impact of some critical determinants on economic growth in Brazil between 1822 to 2019. Through cointegration analysis and Granger causality tests, evidence suggests that years of schooling (human capital) have a positive and long-lasting impact on Brazil's economic growth. Moreover, institutional quality promotes growth indirectly by fostering human capital accumulation, and the increase in living standards induces pro-industry structural change. Finally, there is solid evidence that, in the period 1930-2019, industrialization, measured by the weight of industry in terms of value-added, is growth-enhancing. Whereas the third essay debates the absolute Brazilian economic performance, the fourth essay assesses the relative growth (real convergence) of Brazil towards other 6 Latin American (LA6) countries, Portugal, and the United States (US), from 1822 to 2019. The extended unit root test results favor the very long-run real convergence of Brazil towards LA6 and Portugal, but not towards the US. Moreover, the estimation results of Markov Regime-Switching models involving two regimes (real convergence and real divergence) suggest that higher human capital stocks promote Brazil's real convergence towards the other LA6 countries. In contrast, the transition to more democratic political systems (institutional quality) fosters Brazil's catching up with LA6 and the US when the country is in a real convergence regime.