When African women work outside the home, their families reap more income but often with potentially "deleterious consequences on the health of their very young children," according to new Cornell University research.

Although the extra money women earn apparently raises household food consumption more than extra earnings from men do, the diminished quantity or quality of child care may have serious adverse consequences for the health of preschool age children.

This is not the case for the small proportion of women who can command good incomes in the labor market, since their higher earnings compensate for the reduction in child care time, making the overall health effects of maternal employment positive, said David Sahn, director of the Cornell Food and Nutrition Policy Program, a social science research institute at Cornell and one of the largest institutes in the world doing economic policy research on Africa.

Most women, however, fare poorly in the labor market because, relative to men, they have little education, which is a major determinant of earnings. The schooling of girls, in turn, is hindered by low household incomes and domestic responsibilities, such as child care, which impinge on their ability to attend school but does not affect boys.

"When women work, they usually must spend less time directly caring for their children, including breastfeeding, preparation of nutritious foods and getting medical care," Sahn said. "Our analysis provides evidence that these time constraints have a small but significant negative effect on the nutritional status and health of their children under 5 years of age."

These finding are a result of research undertaken by Sahn and research associate Peter Glick, both economists at Cornell, and soon to be published in the Oxford Bulletin of Economics and Statistics. It is based on their research in Conakry, Guinea, a very poor but typical African city. They sought to determine how women in the labor force fare compared with men, what factors determine how much money women earned, how women's work affects the health and nutrition of preschool age children and what determines schooling investments of older children.

The research is based on a survey of 1,725 households in Conakry conducted by the authors in 1990-91. Their work is unusual in that it differentiates among work in the informal, non-wage sector and work in the formal public and private sectors. Unlike most previous research, the statistical methodology controlled for unobserved factors that affect both maternal work and child health, which normally make it difficult to attribute causality from the former to the latter. They developed an appropriate econometric model that allows them to estimate how maternal employment, working through time and budget constraints, causally affect child health outcomes.

In a companion paper published in August in Economic Development and Cultural Change, the researchers also modeled employment outcomes and earnings of men and women. They reported that women earn far less than men in all areas of employment and that women also appear to be disadvantaged in access to formal sector employment, such as public sector jobs.

The factors behind these differences are complex, but two stand out: first, potential discrimination by employers; and second, that women have less education than men, so command lower wages or fail to qualify for many lucrative occupations.

The second finding "confirms an emerging body of evidence for developing countries showing that investments in female education will serve to reduce gender inequalities in economic opportunities and incomes," Sahn said.

At present, however, girls in Guinea are less likely to be in school, and withdraw from school earlier, than boys. Hence the focus, in another part of the research, on the reasons for these differences in the educational attainment of boys and girls. Glick and Sahn report that the higher the mother's and father's education, the higher their children's educational attainment.

"But the positive effect of mother's schooling on daughter's education was especially pronounced," Sahn said. "A plausible explanation for this is that educated women themselves have preferences for schooling daughters and also (since they are likely to have their own income or better options in the job market) can ensure that family resources are allocated to their daughter's education -- even if their spouses are not so inclined."

On the other hand, having a young sibling was a profound educational liability for girls, whereas this did not reduce the education of boys. "The burden of domestic work, such as caring for younger siblings, impinges greatly on girls' ability to get an education, yet education is a major determinant of a woman's employment and lifetime wages," said Sahn, who teaches courses in food and nutrition policy.

"Poor schooling for girls can have a devastating effect that is passed on intergerationally, since a mother's education is strongly linked to household income, daughters' education and child health."

Glick and Sahn also found that how far children would go in school is strongly linked to household income, especially for girls; this is presumably because higher income families can pay for child care rather than using a daughter, who thus has time for school.

In combination, the study on labor markets, human capital accumulation and health in Conakry indicate that labor force participation among women has benefits as well as some potential perils. As for benefits, expansion of women's employment and income earning opportunities will help women achieve economic and social parity with men -- a major policy concern throughout the developing world.

The peril is that, for poor households that cannot afford to pay for child care, there may be some nutritional risk for young children. "In other words, the net impact of maternal employment on the nutritional status of young children, accounting for both changes in income and time allocation, is generally negative for the poor in Conakry, though the effect is not very large," Sahn concluded. "This creates a dilemma: Our results, and other research, show that maternal income is much more valuable to children than father's income, because women are more likely to allocate their earnings to food and 'child goods,' but the foregone time in -- and possibly, quality of -- child care that results from greater labor market participation has negative effects on the health of young children. It may even reduce the education of older female siblings if they are obligated to substitute for their mothers in household work."

Then why do women work?

"Among the poor -- and it must be kept in mind that this is a very poor population -- the answer is simple: They have to work if the family overall is to secure basic necessities such as food and shelter. We think mothers or households may be trading off some short-term ill effects on their young children's health for the benefits their extra income has on the whole family," Sahn said.

While stressing that more work needs to be done on this issue in other settings, Sahn said that the understanding of the income and health tradeoffs of women's work has clear policy implications. Rather than discouraging involvement in the labor force of women who -- whether through necessity or aspiration -- choose to work, developing African countries need better alternatives for caring for young children. And because women's earnings have such strong benefits, policies that raise women's earnings will improve child nutrition, even taking account negative effects on child care time as women respond by working more. Thus efforts to expand women's presence in the labor market, so long as they involve improved wage or earnings incentives, will likely be quite beneficial.

Simply expanding women's employment without improving earnings, however, will not be beneficial.

"With respect to schooling, governments must work to improve information available to households on the benefits and options for education, particularly for girls, as well as defraying the short-term costs to the household of spending on their children's education. Further, efforts to target girls' schooling now will have even stronger payoffs for gender equity in future generations, since as adults these girls will be more likely to ensure that their own daughters get an education," Sahn concluded.

The research was supported by The World Bank and the U.S. Agency for International Development.