Comments are provided on Professor Sabot's paper on human capital accumulation in Pakistan. The findings on gender gaps in school attendance and the acquisition of cognitive skills, and the impact of the quality of rural schooling on cognitive skills and labor productivity, reflected a shortage of schools (supply problem); variation in quality determined variations in cognitive skills. There have been general assumptions that gender differences in education in developing countries are related to the prevalence of Islam, but this decomposition model says otherwise. Inadequate supply clearly in this study is the reason for educational differences by gender. The reasons that the gender gap persists in spite of new school construction in rural villages is not adequately explained. A question is raised about whether supply is strictly separate from demand, and whether parents who want schooling for their daughters press government into building in their village. When the gender gap in preschool ability is eliminated, girls outperform boys, but not enough attention, for those not educators, was devoted to explaining the measurement tool (Raven's Colored Progressive Matrices) and methodology. Quality of school issues, which were found to be related to cognitive skill acquisition, were not defined in such a way as to specify important factors. Cognitive skill of teachers, access to electricity, availability of desks, chairs, blackboards, and chalk, and provision of books and other educational material are plausible causative factors determining quality. A constraint analysis is required to determine the most severe constraints. A good guess would be that teachers are a major link to quality; quality might be improved with careful recruitment and adequate training. The methods used to calculate the cost benefit of improving quality of primary schools over increasing the number of middle schools is questioned as to whether the equation covers all the differences in costs. The question is whether improving the quality of teachers is captured in the calculation of costs. The study goes a long way in correcting wrong perceptions and in directing educational reform in appropriate cost effective ways.

The distillation of Zeba Sathar's article on the determinants of fertility decline and child mortality decline is that marriage age and contraceptive surgery could be important factors in bringing about changes in both fertility and child mortality in Pakistan. The concern is that 80 out of 115 million Pakistanis live in rural areas where marriage age is very low and program efforts are limited or nonexistent. The question is raised about how to effectuate changes in attitudes in rural areas to increase marriage age. Another point is made about the simplicity of explanations for fertility and mortality change, when the reality is a complex host of interactive socioeconomic, cultural, social, and program factors that are responsible for fertility at present levels. The suggestion is for development of a more appropriate model of fertility at the micro level which illuminates the interaction of these factors in determining fertility. Sathar is reported to have concluded that the impact of infant and child mortality on fertility was inconclusive. The changing patterns of fertility are likely to bring about a change in the demand for children and a lesser preference for gender; this status change for women will further reduce child mortality and fertility. Large family sizes are postulated to be associated with close spacing and greater concentrations of children under 5 years of age competing for physical resources and having a high risk of infection with inadequate parental attention and care. These conditions occur in families with low income and little parental education. Institutional and community services also affect child mortality. There are also examples of educational opportunity and income equality as factors bringing about demographic change in Sri Lanka and Kerala, India. The author speculates that an outcome of development is increased educational attainment and more equitable distribution of income. Low levels of maternal education are associated with child mortality. Sex preferences and greater allocation of resources to male children undermine the worth of females. The author questions whether parents really do care more for sons at the expense of daughters.

Sathar and Irfan's paper on Pakistan's 1979-80 Population, Labor Force, and Migration Survey provides a useful starting point for additional research, although the objective of interpreting the results of this survey is somewhat premature since the findings have not been fully compiled. The authors' estimates of the determinants of fertility, however, is flawed by the absence of precise, sharply focused hypotheses. It is unclear what is being measured by the dependent variable--children ever born to women aged 15-50 years. The rationale for splitting the survey population into 4 variables (urban metropolitan areas, urban nonmetropolitan areas, rural farms, and rural nonfarm areas) is also unclear. The entry into a presentation of regression results is often precipitous, without clarification as to what factors such as tractor ownership measure. The coefficient of the results on breastfeeding is inconsistent with the results on the effects of income and well-being on fertility. The paper also would have benefited from more detailed study of birth intervals. There has been considerable research on Pakistani fertility, which could have been brought into greater focus.