Mean Ages at Parities: An Indirect Estimation

This paper is the outcome of a continuous concern that some of the basic demographic estimates provided by successive surveys in the country do not fulfil the expectation of use at their face value. In a previous paper, an indirect assessment of the birth rates from these surveys was made on the basis of internal consistency of the relevant data. It was pointed out in that paper that a rate being a ratio of the two estimates (e.g., the number of births and the population), any variation of emphasis on obtaining accuracy of the numerator or denominator makes the rate unrealistic. For example, placing more emphasis on a better coverage of births or deaths and a relatively less or no corresponding effort on the measurement of population may lead to biased estimates. Further, any inconsistency of such measurement between different surveys would make their utility for assessing even a time-trend questionable [Afzal et al. (1993)]. Whereas there is no substitute for realistically estimated birth rates (or fertility rates), an index based on relatively simple information which is less affected by the variations in adjustments of births and population is the "Mean Number of Children Ever Born per Woman" The estimates of mean number of children ever born per woman are based on a simple question about how many children have so far been born alive to a woman till the time of collection of data in a survey or a census. Based on these data, estimates of mean number of children born to women of a specific age group or collectively of all age groups within the reproductive period are easily computed. For a specific age group, these estimates reflect fertility experience since the beginning of the reproductive span, usually taken as 15 years. Thus, for a younger age group of women, the number of children ever born is less than for those in the higher age groups. Collectively for the women aged 15 to 49 years, the estimated mean number of children ever born per woman gives a composite index relating to varied exposure periods.