Social Segregation Profiles Based on FANNY and GOM Methods: An Application for Belo Horizonte, Brazil, 1996

Ricardo A. Garcia, Centro de Desenvolvimento e Planejamento Regional (CEDEPLAR)

The purpose of this paper is to compare results of the application of two different methodologies, FANNY – Fuzzy Analysis – and GoM – Grade of Membership, on the construction of a social segregation index. The main difference between the methodologies is that, while FANNY is applicable to continuous data, GoM is applicable to discrete variables. An index can be understood as a synthetic measure that reduces the dimensionality of a group of variables, maintaining the variability of the original information. In this paper, we target a measure that could best reproduce the social segregation definition, anchored in two factors: inequality and poverty. Two profiles were created, trying to represent the two socioeconomic opposite conditions what, in a general way, represent good and bad socioeconomic conditions. The exploitation of these tools can be fundamental for urban planning and public policies, especially in developing countries.