Análise comparativa da eficiência na produção de café entre as principais regiões brasileiras

Abstract

Agribusiness represents one of the pillars of the Brazilian economy, so that coffee is a product that contributes significantly to the good performance of the sector. Coffee production in Brazil is divided into two types, Arabica and Conilon. In the production of Arabica coffee, the regions of Minas Gerais, Espírito Santo, Paraná, São Paulo and Bahia stand out, while in Conilon the regions of Rondônia, Bahia and Espírito Santo stand out. However, the production scenario has undergone changes in recent years, following the impacts of undesirable weather events. In view of this, the objective of this dissertation was to evaluate the performance of the main producing regions of Arabica and Conilon coffees in Brazil in the 2018-2019 and 2020-2021 harvest years, through the application of Principal Component Analysis (PCA) and Envelopment Analysis techniques. of data (DEA). Therefore, the study valued the use of a hybrid approach, with the use of quantitative techniques and complementary qualitative analysis. The study also presented a descriptive character and followed the inductive logic. The completion horizon of this was 12 months for the completion of all methodological steps. Therefore, the results pointed to the presence of 4 inputs (Qty_M, Size_prop, Water_cons and Area_cafe) and 1 output (Rend_prop) for the 2018-2019 crop year, and 21 inliers in the sample set. For the 2020-2021 harvest year, 3 inputs (Area_cafe, Cred_finance and QtdInsum_prod) and 3 outputs (AdOrgan_prod, Rend_prop and Rend_tec) and 23 inliers were considered. Regarding efficiencies, 6 inefficient producers were identified for 2018-2019 and 9 for 2020-2021. The Qtd_M and Rend_prop variables had the most impact on efficiency in 2018-2019, while for 2020-2021 the Cred_financ, QtdInsum_prod, Rend_tec and AdOrgan_prod variables had the most impact. As for the existing limitations, these were related to the researcher's choice of methodological steps for the best adjustment and presentation of results, the use of a strictly financial basis to complement the data collection referring to the Campo Futuro project of CIM/UFLA and the restriction methodological principle that the number of DMUs must be three times greater than the number of original variables selected for the application of the DEA technique. In order to propose an agenda for future studies, it is suggested the replication of this study for other cultures, the expansion of the sample set in order to increase the number of original variables considered, the application and combination of new techniques such as Structural Equation Modeling (SEM), as well as a more detailed study on the efficiency of each producing region and their respective benchmarkings.