The cohort ratio method is a demographic approach used for sex and age-specific population projections in small areas. However, due to the instability of the demographic events in small areas, this method produces unreasonable results in some age groups. In this paper, we apply a Bayesian spatial statistic technique (empirical Bayes) to reduce instability when estimating the cohort changes’ index of the cohort ratio method for small areas. When applying the method for population projections in some Brazilian small areas, from 2000 to 2010, we observe a reduction on the variability of the estimations. In addition, our method reduced the error of the projections from 2000 to 2010 when compared to the method originally proposed by Duchesne (1989), especially on the last age groups.
Esta obra está bajo una licencia internacional Creative Commons Atribución-NoComercial 4.0.
Derechos de autor 2019 Flávio Henrique Miranda de Araújo Freire, Marcos Roberto Gonzaga, Marília Miranda Forte Gomes