Emily Oster on the Pandemic

Episode Summary

Executive Summary: Emily Oster discusses her school-COVID data project, finding relatively low in-school transmission in K-12 and little evidence of schools driving major outbreaks, with risk tracking community spread more than school settings. She argues school closures impose large learning and equity costs, and then turns to how pregnancy/parenting books evolved toward a softer, more nuanced tone about uncertainty, tradeoffs, and letting children become themselves.

Main Topics: School COVID data project and methodology (Priority: 5/5): Oster explains how her team gathers district/school enrollment, mitigation, and case data every two weeks to track COVID in schools and compare across settings and age groups. What the school data shows so far (Priority: 5/5): Early results suggest low case rates among students, somewhat higher rates among staff, and no strong evidence of large K-12 super-spreader outbreaks; school cases appear to scale with community prevalence. School reopening, inequity, and policy tradeoffs (Priority: 5/5): Russ and Oster debate the costs of remote schooling, especially for lower-income families, and the uneven reality that private schools and some states reopen while many public systems remain closed. Masks, mitigation, and the politics of COVID (Priority: 4/5): They discuss indoor-outdoor transmission, mask use as a low-cost precaution, and the way public health choices became partisan and moralized. Lockdowns, Sweden, Israel, and uncertainty in pandemic policy (Priority: 4/5): Oster and Russ explore how different countries’ reopening/lockdown strategies make causal inference difficult, and why a scalpel approach plus contact tracing may be better than blanket shutdowns. Parenting books, data, and changing tone (Priority: 3/5): Oster reflects on revising Expecting Better and Cribsheet, updating recommendations with new research and softening her tone after years of writing about pregnancy and parenting tradeoffs. Parenting, identity, and transformative experience (Priority: 3/5): The conversation closes on how parenting reveals the limits of data, the challenge of raising children who are not replicas of their parents, and the importance of supporting children’s individuality.

Key Arguments: K-12 schools do not appear to be major drivers of community spread; observed school cases mostly mirror what is happening in the surrounding community. Student case rates are lower than staff rates, and high schools seem riskier than elementary schools, but still far below the transmission dynamics seen in colleges. Closing schools first, while keeping restaurants, bars, and gyms open, reflects a poor prioritization of public policy because school closures have huge educational, mental-health, and equity costs. A blanket lockdown strategy is too blunt; policymakers need a more targeted approach based on contact tracing and understanding where transmission actually occurs. Masking indoors around others is a low-cost, sensible precaution, but outdoor masking and highly symbolic mask policing are less rational and became politically distorted. Parents and families should have choice where possible because the risk calculus differs dramatically across households, especially when grandparents or multigenerational living are involved. Research and lived experience changed Oster’s own writing: as more evidence accumulated, she revised some guidance and softened confrontational language toward a more nuanced, individualized approach. Parenting is a transformative experience that data can inform only partially; children are not extensions of parents, and good parenting requires adapting to who they are.

Data Points: Schools in dataset: about 1,300 - Most recent two-week reporting period in Oster’s school-COVID project Total enrolled students tracked: about 700,000 - Students represented in the most recent poll Students attending in person: about 280,000 - Estimated in-person daily attendance In-person staff tracked: about 80,000 - Staff represented in the dataset Cases per thousand students per two-week period: about 1.5 - Most recent period; Oster notes this is relatively low and sample-selected Cases per thousand staff per two-week period: about 3 - Roughly twice the student rate School mode distribution: about 25% fully remote, 25% fully in person, 50% hybrid - Approximate distribution in the dataset Share of U.S. kids in school: 56 million - Russ cites the overall population for comparison Study cadence: bi-weekly - Schools/districts are re-contacted every other week for case reporting Number of times Oster appeared on EconTalk before: 3 - This episode is her fourth appearance Year Expecting Better first came out: 2013 - Oster reflects on revising the book years later

Pivotal Quotes: "the risks seem to be relatively low" — Emily Oster: Summary of early school-COVID findings "we aren't seeing a lot of what looks like big outbreaks in the schools" — Emily Oster: On whether K-12 schools are driving super-spreader events "we need a little bit more of a scalpel approach" — Emily Oster: On why blanket lockdowns are too blunt and contact tracing is preferable

Implications: For families and policymakers, the episode argues for reopening schools with mitigations where possible, while using targeted public-health tools instead of broad shutdowns. It also suggests future parenting advice will be more individualized, evidence-based, and comfortable with uncertainty.