An indirect view of the fertility changes in Pakistan.

Data on children ever born from surveys conducted between 1984 and 1988 provide evidence for fertility decline in Pakistan and little evidence for a "zig zag" pattern of variations in the levels of the birth rate reported in surveys conducted between 1962 and 1988. This study is concerned with the indirect estimation of the crude birth rate (CBR). Input data is available on age specific mortality and the age distribution of the population. CBR was calculated on the basis of the life table method. The estimation of CBR from 1962-65 to 1979 revealed a declining trend, which also appeared in estimations based on survey data after 1979. Estimation of CBR based on infant and child mortality risks showed declining trends through 1979, a slight elevation in infant mortality in 1984, and then decline followed by a higher increase in 1988. The increased risk of infant mortality was interpreted as an artifact of the data. Further analysis of the correlation between infant and child mortality risks in the ordinary least squares equations indicated that 50% of the variation in infant mortality risk was explained by child mortality risk variation between 1962 and 1979. The adjusted values were then used to estimate mortality between 1984 and 1988, which indicated similar values (91/1000 births) as the 1990-91 survey data. In the ordinary least squares examination of the association between child mortality risks and the estimates of crude birth rates between 1962 and 1979, the finding was that 70% of the variation in the values of CBR were explained by variations in child mortality risks. The indirect estimation for 1984 to 1988 however showed elevated CBR, which was construed to be due to increased infant mortality risks, which may be in error. When CBR was estimated using the proportions in the population aged 0-4 years and mortality risk, the life table estimates showed a reduced rate by 3.3 for 1984 and by 1.1 for 1988, and with other substitutions, lower CBR. The conclusion from the exploration of different methods of calculating was that CBR may be even lower that 37/1000, and the rate of natural increase may be lower than 2.7%.