Breastfeeding in the United States: Economic Analyses of Trends and Policies.

Abstract

Public health organizations expend considerable effort and resources to increase breastfeeding rates and reduce disparities on the basis of the medical literature’s conclusion that “breast is best.” Yet, due to data limitations, little is known about the evolution of breastfeeding rates and the impact of policies aimed at raising rates. My research addresses these limitations by making novel use of publicly-available survey data and applying economic methods to improve knowledge.

I carefully construct a data set of births from 1939-2009. This data set includes breastfeeding as well as demographic variables. I explore multiple weighting methods to develop a nationally-representative sample from the pooled survey data that is comparable to aggregate proprietary survey data.

The twentieth century witnessed dramatic changes and advances in infant-feeding technology, medical understanding, policy, culture, and parenting preferences. The shifts evident in the data affected demographic groups differently, leading to changes in disparities over time. Simulations of breastfeeding rates show that changes in behavior, not characteristics, primarily drove shifts in breastfeeding; however, the magnitudes of the shifts were not universal across racial groups.

Although current breastfeeding rates are at an all-time high, they are still below public health goals, particularly for working mothers. As part of a concerted initiative to raise breastfeeding rates among working women, state and federal legal protections were introduced to make it easier for mothers to pump at work. I examine the impacts of these laws on breastfeeding rates using data from multiple surveys and the exogenous variation in the existence and timing of state-level workplace lactation support laws.

Legal requirements for lactation support at work successfully raised the percent breastfed at 3 months by 2 percentage points, a substantial increase given the overall trend. In the aggregate, I rule out large effects on long-run measures of duration, but selected groups of mothers were differentially impacted. Overall, the evidence suggests that cultural shifts as well as policy changes affected breastfeeding initiation and duration, and the effects varied across subgroups. My findings and techniques will facilitate future analyses and inform policy going forward.