Emily Oster is widely known for translating academic research into accessible advice for parents. What is less widely acknowledged is that she is an economist with no formal training in biology, medicine, epidemiology, infectious disease, or public health. This matters because she routinely publishes interpretations of biomedical evidence that fall outside her domain of expertise and those interpretations have repeatedly diverged from scientific consensus in ways that carry real‑world consequences.
A Breadth of Research That Masks Shallow Domain Knowledge
Oster’s publication record spans health economics, developmental economics, demography, gender, public‑health behavior, pregnancy, and parenting. Economists sometimes work across multiple domains, but deep expertise in biomedical risk assessment requires training she does not have. Epidemiology, teratology, infectious disease transmission, and foodborne illness all rely on methods fundamentally different from the quasi‑experimental inference tools economists use.
This mismatch is not theoretical, it shows up directly in her public claims.
Alcohol in Pregnancy: A Case Study in Misinterpreting Biomedical Evidence
Oster’s suggestion that light drinking during pregnancy shows “no consistent evidence of harm” is a textbook example of overconfidence built on underpowered data. She relies heavily on small observational studies with uncontrolled confounding, then interprets “no clear signal” as “no meaningful risk.”
Medical authorities disagree — unequivocally.
- ACOG: “There is no safe amount or type of alcohol use during pregnancy. Even moderate drinking (one drink a day) can cause lifelong problems for your baby.”
- Peter Berns, CEO of The Arc: Oster’s framing “gives the irresponsible impression that the odds of a drink here or there affecting your child aren’t high enough for women to choose to abstain.”
This is not moralism. It is risk management under uncertainty. Oster’s interpretation minimizes risk in a domain where precaution is the standard of care.
COVID‑19 School Transmission: Premature Conclusions with Policy Impact
During the pandemic, Oster published a widely circulated Atlantic piece claiming that schools “do not appear to be major spreaders of COVID‑19.” Her dataset was sparse, self‑reported, and missing critical variables. She generalized from schools with strong mitigation to schools without it, and she framed uncertainty as reassurance.
Subsequent evidence contradicted her claim:
- Bravata et al.: Increased school visits were associated with increased COVID infections and hospitalizations, especially in low‑income counties.
- Netherlands modeling: School closures reduced transmission and healthcare burden.
Epidemiologists warned at the time that her analysis was incomplete and premature. Yet her framing influenced public debate and policy decisions, particularly in communities with poor ventilation, limited PPE, and high baseline transmission.
Oster’s analysis was clearly flawed and she harmed people, particularly children, in the process, just like she has done with her position on alcohol intake during pregnancy.
This and other positions of hers should have been enough to discredit her, but of course she appeals to those who want to live life as they please with no thought of the consequences of their actions on others. This became very apparent during the COVID pandemic. When a risk is inconvenient or socially unpopular, she tends to minimize it. This makes her appealing to audiences seeking reassurance rather than accuracy. but it also means her conclusions often conflict with scientific consensus.
Oster also downplays the use of raw milk (I won’t link to her blog about it) as well as the cyclosporiasis outbreak. Regardless of intent, Oster’s public-facing analyses repeatedly downplay risks in ways that conflict with established scientific consensus and have contributed to public misunderstanding and policy missteps and harms people in the process.
Postscript: Additional links showing just how dangerous she is. I really wonder about her motivations.
https://depts.washington.edu/fasdpn/pdfs/astley-oster2013.pdf
