Análise da complexidade econômica e relatedness no Brasil: dinâmicas microrregionais relacionadas ao crescimento econômico e à desigualdade de renda

Abstract

This thesis consists of three articles that examine the behavior of economic complexity and relatedness in Brazil’s micro-regions from 2002 to 2020. The economic complexity literature has been rapidly expanding, providing increasingly robust evidence on how a country’s economic structure is shaped and its implications for wealth generation, population well-being, and related topics. However, significant gaps remain regarding its effects on economic growth and formal income inequality, particularly in developing countries. At the micro-regional level, this study accounts for the dynamics of temporal and spatial heterogeneity, using reliable regional data while controlling for unobserved factors. The first article calculates the economic complexity of different activities, and analyzes the spatial distribution patterns of economic complexity and relatedness (cognitive proximity between activities) across Brazilian micro-regions through Exploratory Spatial Data Analysis (ESDA). The findings indicate that the most complex economic activities are concentrated in the manufacturing sector, followed by IT and related services. The study also identifies a positive spatial dependence of economic complexity and relatedness, revealing a high degree of spatial concentration in certain regions. This concentration is influenced by the broader economic context, as evidenced by the increased spatial clustering of economic complexity following the 2015/2016 Brazilian economic crisis. The formation of high-high clusters (high complexity and relatedness within the micro-region and its neighboring areas) is predominantly observed in the Southeast and South regions throughout the analyzed period, reflecting significant disparities in productive capabilities across the country. The second article examines the effects of economic complexity and relatedness on economic growth in Brazil’s micro-regions. To this end, dynamic spatial panel data models (SDM, SDGMM-SYS, and SDGMM-DIF) with instrumental variables were applied to a dataset covering 558 Brazilian micro-regions, incorporating employment data for 59 economic activities over the period. The results suggest that higher levels of economic complexity and relatedness are associated with an increase in micro-regional GDP per capita. Additionally, the effects of economic complexity on economic growth vary across the country’s regions, highlighting regional disparities. Finally, the third article investigates the influence of economic complexity and relatedness on regional formal income inequality in Brazil’s labor market. Employment data for each economic activity in the 558 micro-regions were analyzed using dynamic fixed-effects panel regression models and the Generalized Method of Moments (GMM) with instrumental variables. The findings reveal that both economic complexity and relatedness contribute to reducing income inequality in the formal labor market, with the impact of economic complexity being particularly pronounced in micro-regions that exhibited high initial income inequality. These results provide an original contribution to the regional economic literature in Brazil, emphasizing the positive spillover effects of economic complexity and cognitive proximity among economic activities in reducing income disparities. In summary, this thesis demonstrates that economic complexity and relatedness have positive effects on economic growth and play a significant role in mitigating formal income inequality at the micro-regional level in Brazil. These findings underscore the need for public policies that foster economic complexity, productive diversification, and regional capacity building, particularly in historically disadvantaged areas.