Sampling variability of own-children fertility estimates.

This paper develops methodology for estimating standard errors and confidence intervals for own-children estimates of age-specific birth rates and total fertility rates. The methodology applies to systematic samples of households, which are treated as simple random samples of women. The assumption of simple random selection, together with the treatment of sample subgroup size and reverse-survival ratios as constants instead of random variables, imply that sampling variability is slightly underestimated. The methodology is applied illustratively to fertility estimates based on the 1970 census of the Philippines.