Estimating Total Fertility Rates for Small Areas in Brazil

Suzana M. Cavenaghi, Universidade Estadual de Campinas
Joseph E. Potter, University of Texas at Austin
Carl P. Schmertmann, Florida State University
Renato M. Assuncao, Universidade Federal de Minas Gerais

Indirect methods of fertility estimation are well known to demographers in developing countries. Recently, however, with the increased interest in estimates for small local areas, imprecision has become a major issue since all the errors and biases introduced in the calculation of indirect estimates have been enlarged by the variability due to the small size of the population exposed to risk. Additionally, most methods proposed to correct data and to estimate the TFR indirectly depend on handpicking the best estimates, a cumbersome task for a large number of areas. In this paper we propose to estimate of the level of fertility for municipalities utilizing several already developed methods to correct for non-sampling errors such as non-response and miscalculation of the reference period, as well as by taking advantage of smoothing techniques from spatial statistics to correct for sampling variation.