Evidence of age exaggeration in demographic estimates for Pakistan

Abstract

‘Own-children’ fertility estimates based on Pakistan's 1973 Housing, Economic and Demographic Survey implausibly indicate a sharp rise in marital fertility at younger reproductive ages and a sharp decline at older reproductive ages between 1962–6 and l967–71. Age-specific death rates based on Pakistan's l962–5 Population Growth Estimation Experiment are too high below the age of minimum mortality and too low above this age, compared with fitted model age patterns of mortality. P/F ratio estimates of age-specific birth rates derived from Pakistan's l971 Population Growth Survey (PGS) are too low below the age of peak fertility and too high above this age, compared with parallel estimates derived from the 1975 Pakistan Fertility Survey (PFS). These changes and discrepancies are plausibly explained by a pattern of age exaggeration that increases with age in all the above-mentioned data sets except the PFS, in which the quality of age data appears to be relatively good. In all likelihood the changes and discrepancies are either not real or much smaller than estimated. A noteworthy by-product of the analysis is that the common procedure of fitting a stable population from life expectancy and the proportion of population below age 35 does not work well in Pakistan.

Support for this research was jointly provided by the Pakistan Population Census Organization, the East-West Population Institute, the United States Department of Commerce, and the United States Agency for International Development. We are grateful to Victoria Ho, Minja Choe, and Robin Loomis for programming and research assistance.

Support for this research was jointly provided by the Pakistan Population Census Organization, the East-West Population Institute, the United States Department of Commerce, and the United States Agency for International Development. We are grateful to Victoria Ho, Minja Choe, and Robin Loomis for programming and research assistance.

Notes

Support for this research was jointly provided by the Pakistan Population Census Organization, the East-West Population Institute, the United States Department of Commerce, and the United States Agency for International Development. We are grateful to Victoria Ho, Minja Choe, and Robin Loomis for programming and research assistance.