Working Paper: NBER ID: w24452
Authors: Jesse K. Anttila-Hughes; Lia Ch. Fernald; Paul J. Gertler; Patrick Krause; Eleanor Tsai; Bruce Wydick
Abstract: Infant formula use has been implicated in tens of millions of infant deaths in low and middle-income countries over the past several decades, but causal evidence of its link with mortality remains elusive. We combine birth record data from over 2.6 million infants across 38 countries in the Demographic and Health Surveys (DHS) with reconstructed historical data from annual investor reports on the timing of Nestlé entrance into infant formula country markets. Consistent with the hypothesis that formula mixed with unclean water could act as a disease vector, we find that infant mortality increased in households with unclean water sources by 19.4 per thousand births following Nestlé market entrance, but had no effect among other households. This rate is equivalent to a 27% increase in mortality in the population using unclean water and amounts to about 212,000 excess deaths per year at the peak of the Nestlé controversy in 1981.
Keywords: infant formula; infant mortality; Nestlé; public health; low and middle-income countries
JEL Codes: I14; I15; O15
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Nestlé's entry into infant formula markets (L66) | infant mortality rates among households with unclean water sources (I14) |
| Nestlé's entry into infant formula markets (L66) | excess infant deaths per year (I12) |
| infant mortality rates among households with unclean water sources (I14) | infant mortality rates among households with clean water sources (I14) |
| Nestlé's entry into infant formula markets (L66) | health risks associated with mixing formula with unsanitary water (I12) |
| education (I29) | ability to properly prepare infant formula (D18) |