Revista Brasileira de Estudos de População
versión impresa ISSN 0102-3098
BRITO, Luana Paula Gentil de; CAVENAGHI, Suzana y JANNUZZI, Paulo de Martino. Demographic estimates and projections for small domains: an evaluation of the precision for the municipalities in the State of Rio de Janeiro in 2000 and 2007. Rev. bras. estud. popul. [online]. 2010, vol.27, n.1, pp.35-57. ISSN 0102-3098. http://dx.doi.org/10.1590/S0102-30982010000100004.
Demographic projections have been used more and more frequently to generate information for planning economic, social, political and environmental development in many different countries. But the broader the levels of geographic, demographic and temporal details required, the less precise are the projections. In small domains, such as municipalities, difficulties come up that include small populations, volatility of the data on growth patterns, poor quality of information, and others. The objective of this article is to evaluate several different methodologies of demographic projections for small domains, taking the municipalities of the State of Rio de Janeiro (of which the City of Rio de Janeiro is the Capital), Brazil, as a set of examples. The techniques studied are the apportionment method (AiBi), Duchesne's relationship of cohorts and ratio correlations, used for comparisons with the results of the Federal Census of 2000 and with the Counting carried out in 2007. The comparative results with the Census of 2000 indicate that, on the average, the estimates and projections produced by applying the three methods show a precision level within acceptable margins, but the errors vary from -21.4% to 24.1%, and the projections tend to be underestimated. In contrast, the comparison of the results from the Counting of 2007 proved to be less precise. As described in this article, the method of ratio correlations, which uses symptomatic variables, was seen to be the best of the three methods used to evaluate the municipalities in the State of Rio de Janeiro.
Palabras clave : Demographic projections; Small domains; Precision estimates; Rio de Janeiro.