WIC Participation and the Initiation and Duration of Breastfeeding

Year: 2001

Research Center: Institute for Research on Poverty, University of Wisconsin-Madison

Investigator: Chatterji, Pinka, Karen Bonuck, Nandini Deb, and Simi Dhawan

Institution: Montefiore Medical Center, Einstein College of Medicine

Project Contact:
Pinka Chatterji, Assistant Professor
Einstein College of Medicine
Department of Epidemiology and Social Medicine
111 East 210th Street
Bronx, NY 20467
Phone: 718-519-0772


This study measured the effect of participation in the Supplemental Nutrition Program for Women, Infants, and Children (WIC) after the birth of a child on one important health behavior: the initiation and persistence of breastfeeding. In the United States, many low-income mothers do not initiate breastfeeding at all and very few breastfeed for at least 4 months, although most pediatricians recommend breastfeeding exclusively for at least 4 to 6 months. Recognizing this problem, during the 1990s, the U.S. Department of Agriculture implemented many new policies intended to promote and support breastfeeding among WIC participants. While the new policies promote breastfeeding, the program provides infant formula in food packages, a major incentive for mothers to participate in WIC. The provision of infant formula in WIC food packages may work against the program’s emphasis on breastfeeding.

To build on previous research on breastfeeding among WIC participants, this study used data from the children of the National Longitudinal Survey of Youth—a large, longitudinal data set that includes children born between 1990 and 1995. Average annual expenditures on breastfeeding education and promotion increased from $9 million in 1990 to $36 million in 1995. The authors employed econometric methods to account for self-selection into the WIC program and linked mothers’ breastfeeding practices to State WIC and Medicaid policies.

Very little is known about the effect of WIC participation on breastfeeding practices among participants during the 1990s. Previous research suggests that a woman’s participation in WIC during pregnancy leads to important health benefits for children, including increases in birth weight, reductions in iron deficiency anemia, and reductions in infant mortality. Much less is known, however, about the benefits of WIC participation after pregnancy, during infancy, and during early childhood. Although research on the effectiveness of the WIC program focuses almost exclusively on a woman’s participation during pregnancy, most WIC participants are not pregnant women, but rather are infants, young children, and post-partum women. A large proportion of WIC funds are targeted at these groups, but little is known about whether WIC improves health behaviors and outcomes for them. This information is necessary to evaluate the overall effectiveness and cost-effectiveness of the program, as well as to better target existing funds.

The main empirical challenge in this study was to account for the possibility that women select into WIC based on unobservable characteristics that also determine breastfeeding practices. For example, mothers disinclined toward breastfeeding might be more likely to participate in WIC. In this case, standard estimation methods will lead to a biased and inconsistent estimate of the effect of WIC on breastfeeding. The authors used three methods to account for this problem:

  1. They estimated standard probit models with a rich set of covariates intended to capture many factors associated with breastfeeding decisions.
  2. They estimated the models using an instrumental variables approach with State-level WIC and Medicaid policies as identifying instruments.
  3. They took advantage of data available on siblings to estimate heteroscedastic, family-level fixed effects models that control for differences across families in unobservable characteristics.

The results of the study suggest that WIC participation has a negative effect on initiating breastfeeding, but it is not clear whether the program affects breastfeeding persistence. Baseline results, which are preferable to instrumental variables results, indicate that WIC participation has a strong, negative effect on breastfeeding initiation, but no effect on breastfeeding for at least 16 weeks. A heteroscedastic fixed-effects model suggests that within families, WIC participation is associated with fewer weeks of breastfeeding. It is important to note that these results are based on a sample of mothers who are older and probably more advantaged than the national WIC population. It is also important to note that expenditures on WIC breastfeeding education and promotion have increased significantly since 1995, which would be expected to influence the effect of the program on breastfeeding.

To generate implications for policy, more qualitative and quantitative research is needed that confirms the results from this small body of work. However, if these results can be replicated using other data, they imply that, while WIC has the potential to affect infant feeding practices, the program faces the challenge of increasing breastfeeding rates among participants, while continuing to offer infant formula to mothers who decide not to breastfeed.