Emily Oster on Pregnancy, Causation, and Expecting Better

Episode Summary

Executive Summary: Russ Roberts and Emily Oster discuss Expecting Better, arguing that pregnancy decisions should be based on data, causation, and tradeoffs rather than folklore or blanket prohibitions. They examine caffeine, alcohol, smoking, food restrictions, cats, home birth, and delivery interventions, emphasizing that risk varies by magnitude and individual context. The conversation also reflects on econometric uncertainty and Oster’s willingness to revise conclusions when better evidence emerges.

Main Topics: Evidence-based pregnancy decisions (Priority: 5/5): Oster explains her frustration that doctors often offered rules without the underlying data, pushing pregnant women to make informed choices using evidence and judgment. Causation vs. correlation in pregnancy studies (Priority: 5/5): The discussion centers on why observational studies can mislead, especially when third factors like nausea, age, or self-selection drive apparent associations. Caffeine and alcohol moderation (Priority: 5/5): They review evidence suggesting moderate caffeine is likely safe and that small amounts of alcohol may not show harm in available studies, while excess alcohol is clearly dangerous. Food safety, listeria, and arbitrary-sounding restrictions (Priority: 4/5): Oster argues that many food warnings are not equally justified; some risks are real, but the likelihood of avoiding listeria by avoiding certain foods may be very small. Smoking, birth weight, and randomized evidence (Priority: 4/5): Unlike many pregnancy topics, smoking has stronger causal evidence because quit-intervention trials show clear benefits, especially improved birth weight. Labor, delivery, and patient advocacy (Priority: 4/5): They discuss epidurals, C-sections, doulas, and the power imbalance in childbirth, emphasizing the value of support and pre-planned preferences. How economists should think about medical evidence (Priority: 4/5): The conversation broadens to Oster’s earlier work on hepatitis B and sex ratios and to the broader challenge of drawing policy conclusions from imperfect evidence.

Key Arguments: Pregnant women need hard numbers and tradeoffs, not just reassurance or rules; decision-making under uncertainty is unavoidable. Observational studies in pregnancy are often confounded by factors like nausea, age, and self-selection, so apparent correlations may not be causal. For caffeine, evidence supports that up to about 200 mg/day is not associated with increased miscarriage risk, while very high intake may be harmful. For alcohol, the strongest evidence shows danger from heavy or binge drinking, but available studies do not show clear harm from occasional moderate drinking. Smoking is different because randomized quit-support studies show that reducing smoking improves infant birth weight and related outcomes. Many pregnancy food restrictions are only loosely connected to real risk; avoiding certain foods may change listeria risk far less than people assume. A doula or informed advocate can reduce unnecessary interventions because labor patients often cannot push back effectively in the moment. Economists can contribute usefully because the central question is causal inference, not clinical delivery care. Oster’s earlier hepatitis B hypothesis was overturned by better data, illustrating why scientists should be willing to revise conclusions when evidence improves.

Data Points: Caffeine threshold: 200 milligrams/day - Oster cites evidence that up to this level, about two 8-ounce cups of coffee, appears not to increase miscarriage risk. Coffee consumption: 3-4 cups/day - Evidence is mixed in the intermediate range; more than a little may be unclear even if very high intake looks risky. Labor length effect of epidural: About 15 minutes - Randomized trials suggest epidurals slightly lengthen labor, but by a small amount. Alcohol moderation prevalence among obstetricians: About 40% - Russ cites Oster noting that many obstetricians consider some moderate alcohol intake acceptable. Example drinking pattern: Half a glass to one glass, 3-4 times per week - Oster describes her own moderate alcohol consumption during pregnancy after reviewing the evidence. Pregnancy book limit: Stops at delivery room - Oster notes the book does not cover breastfeeding or post-delivery dietary questions. Birth weight cutoff: 2,500 grams - Discussed as the rough cutoff for low birth weight; very low birth weight is associated with more problems. Hepatitis B estimate: As much as 60% - Oster recalls her earlier speculative estimate that hepatitis B might explain a large share of male-biased sex ratios in some regions, later disproven.

Pivotal Quotes: "Are you kidding? Like you have to make these decisions. As opposed to what?" — Emily Oster: Oster rejects the idea that pregnant women can avoid decision-making by deferring entirely to authority or folklore. "Good news, ladies. If you are nauseous, you are less likely to miscarry." — Russ Roberts: Roberts summarizes Oster’s explanation that nausea is a marker of a healthy pregnancy and also a confounder in caffeine studies. "I think we have to recognize that by not all living on a rural farm in Finland away from all motor vehicle transportation and only breathing healthy Finnish air, we are all putting our babies at risk." — Russ Roberts: Roberts broadens the point to everyday risk and the impossibility of total safety in modern life.

Implications: Listeners should expect more nuance and fewer absolute rules in pregnancy guidance. The episode argues for evidence literacy, shared decision-making, and humility about medical certainty—especially where observational data can mislead and personal tradeoffs matter.