Childrens Program

May 1, 2009
Jonathan Gruber, Organizer

Anna Aizer, Brown University and NBER, Laura Stroud, Brown University, and Stephen Buka, Brown University
Maternal Poverty, Stress and Child Well-Being: Evidence from Siblings

Aizer, Stroud, and Buka study how maternal stress affects child outcomes. They find that in utero exposure to elevated levels of the stress hormone ”cortisol” negatively affects the cognition, health, and educational attainment of offspring. These findings are based on comparisons between siblings in order to control for unobserved differences between mothers that could otherwise bias the estimates. The results are consistent with recent experimental results in the neurobiological literature linking exogenous exposure to cortisol in utero with reduced growth in the hippocampal region of the brain and declines in offspring cognitive, behavioral, and motor development. Moreover, the researchers find that the cortisol levels of low socioeconomic status mothers are higher on average and more variable, suggesting that prenatal stress may play an important role in explaining why relatively few children born into poverty are able to escape it as adults.

Seema Jayachanran, Stanford University and NBER, and Ilyana Kuziemko, Princeton University and NBER
Why Do Mothers Breastfeed Girls Less than Boys? Evidence and Implications for Child Health in India

Because breastfeeding inhibits fertility, a mother might limit (prolong) the nursing of an infant if she wants to continue (stop) having children. Similarly, if a mother becomes pregnant while still breastfeeding, she often weans the first child sooner than she otherwise would have. Jayachandran and Kuziemko develop a dynamic programming model of breastfeeding as a function of future fertility, and provide support for its predictions using survey data from India. First, breastfeeding increases with birth order, since mothers near or beyond their desired total fertility have greater demand for the contraceptive properties of nursing. Second, if parents have a preference for having sons, then mothers with no, or few, sons are more likely to want to conceive again, so they will wean their current child sooner. Therefore, not only are girls breastfed less than boys in India, but also the gender of older siblings affects how long a child is breastfed. Furthermore, this sex-composition effect is strongest for medium birth-order children: at sufficiently low birth order, a mother will want to have more children (and thus limit breastfeeding) regardless of her children's gender composition and at sufficiently high birth order she will always want to stop having children. In unsanitary environments, breastfeeding can protect against water- and food-borne disease, and the authors indeed find that mortality has many of the same relationships with the gender and birth-order interactions as breastfeeding, and that these relationships are strongest in households without piped water. The results suggest that the gender gap in breastfeeding leads to 26,000 to 44,000 “missing girls" in India each year.

David Frisvold, Emory University, and Julie Lumeng, University of Michigan
Expanding Exposure: Can Increasing the Daily Duration of Head Start Reduce Childhood Obesity?

Coinciding with the work requirements of welfare reform in the mid-1990s, the early childhood education program, Head Start, increased the availability of full-day classes. Using unique administrative data, Frisvold and Lumeng examine the effect of full-day compared to half-day attendance on childhood obesity. They identify this effect using the elimination of a state-provided full-day expansion grant that led to an exogenous decrease in the supply of full-day classes for the program in our study. Their results suggest that full-day Head Start attendance significantly reduces the proportion of obese children at the end of the academic year.

Ann Huff Stevens, UC, Davis and NBER, and Jessamny Schaller, UC, Davis
Short Run Effects of Parental Job Loss on Children’s Academic Achievement

Stevens and Schaller study the relationship between parental job loss and children’s academic achievement using data on job loss and grade retention from the 1996, 2001, and 2004 panels of the Survey of Income and Program Participation. They find that a parental job loss increases the probability of children’s grade retention by roughly one percentage point, or 17 percent. After conditioning on child fixed effects, there is no evidence of increased grade retention prior to the job loss, suggesting a causal link between the parental employment shock and children’s academic difficulties. These effects are concentrated among children whose parents have a high school education or less. No systematic pattern of effects is found with respect to pre-displacement parental income or wealth.

Stacey Chen, Royal Holloway University of London, Yie-Chien Chen, National Taiwan University, and Jin-Tan Liu, National Taiwan University and NBER
Separate Effects of Sibling Gender and Family Size on Education: Methods and First Evidence

Parents who prefer sons tend to continue to have children until a son is born. After deciding on their set of children, those with resource constraints may divert family resources from daughters to a son. Thus, the presence of a son, relative to a daughter, has two distinct effects on his sister's outcomes: a direct effect, holding constant family size, and an indirect effect through decreasing family size. Conventional methods have taken family size as an exogenous and predetermined covariate, and assumed the indirect effect to be included in the main effect of family size. However, family size is endogenous and dependent on the sex composition of older siblings. Conventional models cannot define the indirect effect because it cannot be isolated by holding family size constant. Chen and her co-authors show that even if both child gender and family size are exogenous, use of an instrument for family size is required to isolate the direct effect from the indirect effect. Using a large and unique administrative dataset from Taiwan, they demonstrate how Instrumental-Variable methods resolve the problems of endogeneity and the causal dependence of an important covariate (family size) on treatment status (sibling sex). Furthermore, they minimize the incidence of sex-selective abortion by restricting their data to cohorts born prior to abortion legalization and prior to the prevalent practice of prenatal sex determination. Using the occurrence of twinning to instrument for family size, their results indicate that the direct causal effect of having a brother, relative to a sister, on women's juvenile mortality (1-5 years) or college attendance is nearly zero, contrary to most of the previous results based on conventional methods. The observed relationship between sibling sex composition and women's outcomes is driven mostly by the indirect effect through changing family size.

NBER Videos

National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us