A procedure for estimating a few demographic indicators of Assam state and its districts

Abstract

In India, as vital statistics reports bear flaws due to incomplete and irregular birth and death registration, researchers resort to the use of various indirect techniques of demographic estimation. In this regard, challenges faced by researchers in the area lie in the development effective indirect techniques of estimation. This paper has proposed a five parameters polynomial regression model for estimating the life expectancy at birth using child survivorship probabilities obtained from indirect techniques of estimation of infant and child mortality rates. Estimates of life expectancies at birth, infant mortality rates, under age five mortality rates and life expectancies at birth at sub-state level, i.e., district level of Assam, obtained in this paper, may provide a set of demographic indicators enormously useful to health and socio-economic planners of the state. Logit smoothing is done to child mortality estimates given by Trussell’s variant of Brass methods on children ever born (CEB) and children surviving (CS) of 2001 and 2011 census of India, and then Weibull survival function is fitted to the smoothed estimates so as to finally obtain infant and child mortality estimates with enhanced precision. A procedure is adopted here in selecting a suitable model life table family in which two established techniques are simultaneously used. MORTPAK for Windows, the UN software package for demographic measurement in developing countries, is used in abridged life tables of Assam - 23 each for male, female and the combined, from 1970 to 2016, to get unabridged life tables from the abridged wherever required.