Son preference in Pakistan: an analysis of intentions vs. behavior
De Tray, D.
Research in Population Economics 5: 185-200
1984
ISSN/ISBN: 0163-7878 PMID: 12266412 Document Number: 407507
The study assesses the extent to which the very strong expressed preference for sons in Pakistan influences couples' actual fertility behavior. Several fertility measures and estimation techniques were used to determine whether subsequent fertility behavior is influenced by the sex composition of previous births. Khan and Sirageldin's work on sex preference and desires for additional children in Pakistan were reviewed, followed by consideration of several methods of testing the relationship between sex of children and subsequent fertility and presentation of the results of several other tests of the hypothesis that sex composition of children affects actual fertility behavior. The Khan and Sirageldin (1977) results can be translated into a stopping rule which states that the higher the proportion of boys in a family, the less likely a couple is to go on to have another child. This finding, based on intentions, conflicts with the results based on actual behavior presented here. The Khan and Sirageldin results were extended by assessing the extent to which couples translated their very strong expressed preferences for sons into actual behavior. This extension was based on a sample of women, most of whom have completed fertility (all were aged 35 or older) and all of whom had at least 1 live birth. No overall relationship was found between the sex composition of children and subsequent fertility behavior. At least 2 explanations are possible for this inconsistency between intentions and behavior with each having a different implication for population policy and for the future of population growth in general. The first explanation is that in traditional societies such as Pakistan couples are unable to control their fertility adequately. The policy implications of this view suggest family planning and contraceptive distribution programs aimed at reducing the cost of achieving whatever desired family size couples want. Another explanation is that although parents indicate that they want more boys than girls, their actions do not reflect this preference because of the dominating influence of overall desires for numbers of children regardless of the sex of those children. In other words, couples may prefer boy to girl children, but they would still rather have many girl children than few or no children at all. This explanation is consistent with theoretical findings. The results suggest that with regard to actual behavior, Pakistani preferences for having children may dominate preferences for children of a certain type.