Análise temporal dos setores de aglomerados subnormais dos censos 2000 e 2010: o estudo de caso da subprefeitura de São Mateus no município de São Paulo-SP

Abstract

The purpose of this research is to analyze the data on 'subnormal agglomerates' (special enumeration districts to identify the most common kind of slum in Brazil, the favelas), provided by the Brazilian Statistics Office, through the case study area, São Mateus' borough, located in Sao Paulo city, with regard to achieving temporal analysis. The population census, through the 'subnormal agglomerates' enumeration districts, are an important source of socioeconomic and demographic data on favelas, collected periodically and available to all Brazilian municipalities. However, researchers and public authorities have been subject to methodological controversies and cartographic issues that still complicate the use of this information. This research identifies and systematizes the characteristics relating to 'subnormal agglomerates'. In the sequence, this information is verified according to its occurrence and significance through the case study area, by overlaying the vector census zoning 2000-2010 and the cadastral databases from Sao Paulo city hall, in geographic information system (GIS). Improvements in the processes and results of 2010 census were observed, compared to the 2000 census. However, the 'subnormal agglomerates' data still present characteristics that complicates the accomplishment of temporal analysis. The highlighted results are: a) 96.5% of the subnormal enumeration districts in the case study, identified only in 2010 are favelas implemented until 1999; b) 42.5% of subnormal enumeration districts in 2010 were new favelas identifications; c) 55% of subnormal enumeration districts identified in 2010 are the result of subdivision or aggregation of the 2000 census enumeration districts; d) in 26% of subnormal enumeration districts in 2010 there are social housing buildings or houses not considered favelas by the municipal cadastre; e) three subnormal enumeration districts in 2010 have no correspondence with favelas registered by the municipality and four favelas registered by the municipality were not delineated as 'subnormal agglomerates' in the 2010 census. The contradictions in census data between 2000 and 2010, and the differences found in comparison to the favela municipal database become the temporal analysis less accurate. This imprecision tends to expand due the deep modifications on the enumeration districts, which limit the data comparison.