Thais Andreia Araujo de Souza

This paper aimed to study the behavior of Brazilian productivity between 2004-2014 and its impacts on growth, also verifying the influence of a few determinants. It is important due that productivity is considered a important determinant of economic growth. In the country, its behavior has been declining since 1980’s, with few recovery periods. To do this, its performance was studied by different measures, labor productivity, capital and total factor. It was also based on analyzes by sectors and regions. This study aimed to determine which regions and sectors contributed most to the low aggregate productivity. In addition, after productivity performance was verified, some growth decomposition was performed to verify the contribution of productivity to economic growth, as well as a decomposition from sectoral productivity growth to see whar most cause it. Econometric methodologies were also applied to calculate the total fator productivity in order to verify if the results obtained were similar, it was estimated pooled panel, panel data with fixed effects, dynamic panel on level and differences. Finally, considering the behavior of total factor productivity in the period, it was analyzed how a few determinants influenced it, being them human capital, innovation, infrastructure, institutional quality, trade liberalization and business environment, the analysis was performed using vector autorregressive (VAR). Through the results, it was verified that there was a greater increase of the total fator productivity in services and there was decrease in industry. There was a decrease in capital productivity in all regions and sectors, excepting services. Considering labor productivity, there was greater growth in agriculture and decline in industry. The regions that obtained best performance were North, Northeast and Mid-West. It was found that capital productivity, excepting services, contributed negatively to growth in all sectors. And considering other measures, only industry contributed negatively to growth, excepting Mid-west. It was observed that, excepting rate of participation growth, labor productivity growth rate most contributed to the growth, excepting Northeast, where productivity was the one that effectively contributed most. It was also observed that the performance of the sectoral productivity in the country was mainly due to variations in the interssectoral productivity, therefore, the low growth was due to the lack of growth within the sectors, excepting North, where the structural change represented most of variations. In relation to the comparison of the productivity estimates, it was verified that pooled panel presented the greater approximation with the methodology used in the paper, and, it was verified that in short run few changes in the shares of production inputs could increase the productivity, since, by the estimation of the dynamic panel-level it was found that the largest share of production variances was caused by its own past behavior. Finally, regarding the analysis of the determinants, it was also observed that in short run the variances in productivity were mostly explained by their own behavior, and regarding the determinants, innovation was the one that presented the greatest contribution.