SATIRE (unless it isn't)
THE HOAX FILES · 09
The Study That Forgot to Weigh the Babies
The study forgot to weigh the babies. The headlines forgot to weigh the study.
Media-wide
"White doctors kill Black babies" became a talking point. Then someone controlled for birth weight.
As printed on the graphic:
Black Newborns Much More Likely to Die If Doc Is White Hoax
What they sold
A 2020 PNAS paper (Greenwood et al.) using Florida records from 1992 to 2015 reported that Black newborns had lower mortality when cared for by Black physicians. Outlets ran it as proof that racial mismatch between doctor and patient kills Black babies. The study was cited in Justice Ketanji Brown Jackson's 2023 affirmative-action dissent (REPORTED, Manhattan Institute). FNN found no specific CNN segment to pin it on. This one belongs to the mainstream press as a whole.
What the record shows
In 2024, George Borjas and Robert VerBruggen reanalyzed the same data in PNAS. They reproduced the original effect, a 0.13-percentage-point mortality gap. Then they added a control for very low birth weight, under 1,500 grams, which the original study's comorbidity controls had left out. The effect shrank to about 0.033 points and stopped being statistically significant. Their explanation: the sickest, smallest Black infants were disproportionately treated by white specialists. The specialists weren't the cause. The babies' condition was.
The damage
FNN analysis: a single unreplicated study became race-policy gospel, and a Supreme Court dissent repeated it (REPORTED). The correction got a sliver of the airtime. When "science says" turns out to mean "one regression without birth weight," people stop listening to science journalism entirely. Media trust among Democrats fell 19 points in three years (Axios on Gallup, 2025).
Closer
Same data. One missing variable. The whole headline gone. Check the controls.