Objective Methods for Regional Indices: A Multi-Criteria Analysis of Brazilian Rural Development in 2017

To prepare development indices, in general, a significant part of the academic literature used traditional weighting, that is, according to the participation of the weight of the most important variances of the latent factors obtained through factor analysis. Based on this finding, the objective of the article is to contribute by presenting a discussion on three distinct objective weighting approaches originating from multicriteria analysis: Criteria Importance Through Intercriteria Correlation (CRITIC), Gini coefficient, and Method based on the Removal Effects of Criteria (MEREC). To evaluate the methods, the Rural Development Index was used. It was constructed using data from the Brazilian Agricultural Census, originating from the Brazilian Institute of Geography and Statistics (IBGE), and representative of the 5,570 Brazilian municipalities. The three methods will be used in the latent factors and compared with traditional weighting to represent the different classifications of rural development in the regions. In summary, the article's main contribution derives from the different regional categorizations presented by the new forms of weighting, which can help researchers in the difficult task of highlighting characteristics present in the latent factors produced that are more appropriate for representing phenomena related to rural and human regional development.