Princeton sociologist Elizabeth Mitchell Armstrong writes in today’s New York Times about the apparent threat to infant health coming from “wealthy and powerful manufacturers” of infant formula by giving away free samples in maternity wards. Armstrong is concerned that these free samples given away to vulnerable new mothers “[do] more harm than good” because they encourage formula feeding over breast feeding.
Before proceeding, let’s stipulate: breast milk is, all other things being equal, better than formula. Now that we have gotten that out of the way, let’s continue.
What is Armstrong’s evidence that free formula samples discourage breastfeeding? She cites “a 2006 report by the Government Accountability Office [that finds maternity ward formula samples] tend to reduce breast-feeding rates among the women who receive them.”
As far as I can find, the GAO only produced one document on infant formula in 2006, and it consists mostly of slides used in a Congressional briefing. The report focuses mostly on WIC, which is responsible for about 60 percent of all U.S. formula sales. (Armstrong never mentions WIC, TANF, or any federal or state nutrition programs in her op-ed. This is a big omission; it’s like discussing the healthcare market without mentioning Medicare.)
“Finding three” in this GAO slide show states, “A majority of the studies we reviewed found lower breastfeeding rates among women receiving formula samples in [maternity] discharge packs.” That sounds damning.
But delve deeper. What the GAO found was this: “In 7 of 11 sound studies we reviewed, breastfeeding rates were lower for women receiving discharge packs.” To put it another way, about 2/3 of studies found a relationship — not a causal one, mind you, just a correlation — between discharge packs and breastfeeding rates at some period of time in the future. Four of the studies (a third of them) found no relationship between discharge packs and breastfeeding.
Here’s where it gets interesting. Appendix IV of the GAO report lists the studies used in this meta-analysis. They were published between 1986 and 1998. The median year of publication was 1990.
What is the trend in breastfeeding? The GAO report also gives us this chart:
In other words, the year in which the median study was published was the low point in breastfeeding in a generation. The trend has been upwards since then. This holds true for both WIC and non-WIC infants.
The entire thesis of Armstrong’s op-ed is that giving formula samples to discharged new moms is bad because it discourages breastfeeding. But the evidence she offers to support this is very thin. Unless discharge packs waned in popularity in the last 20 years — which would make her op-ed moot — breastfeeding seems to be rising despite the insidious machinations of formula manufacturers.
Moreover, Armstrong ignores the elephant in the room. WIC breastfeeding rates are much, much lower than non-WIC rates, and about half of all U.S. infants are WIC supported. And formula is free to WIC households. To what extent free WIC formula causes declines in breastfeeding is an empirical question, and one which I can’t find good empirics for. But intro-level microeconomics says it probably has some effect at the margin.
Ultimately, Armstrong’s op-ed hangs its hat on one piece of evidence, a slide show that doesn’t even qualify as a meta-study. What really seems to drive this op-ed is a hostility towards free enterprise and commerce. She doesn’t like the fact that formula is a “commercial product,” that there might be “commercial influence” in nutrition decisions, and that “companies provide [maternity swag] for the same reason they distribute swag to celebrities: it drives sales.”
That’s not an argument based in science or evidence. It’s one based on a revulsion to mixing capitalism with childhood. And that’s a position that one is free to hold, but let’s not confuse science with feelings.



