Emily Oster on Infant Mortality

Episode Summary

Executive Summary: Emily Oster discusses her infant-mortality paper with Russ Roberts, showing that the U.S. gap with Europe is partly measurement-driven, partly due to higher prematurity, and largely concentrated after the first month of life. The U.S. performs well in neonatal care once birth weight is controlled, but lags later due to social and household factors, especially among lower-SES groups.

Main Topics: Defining and comparing infant mortality across countries (Priority: 5/5): The conversation begins by clarifying the standard definition of infant mortality and why international comparisons can be misleading due to differences in live-birth reporting, especially for very early gestations. Data sources and micro-level linked birth-death records (Priority: 5/5): Oster explains the natality and mortality files used in the U.S. and comparable data from Finland and Austria, emphasizing the value of linked microdata for separating causes by gestational age, birth weight, and timing of death. Measurement issues and reporting adjustments (Priority: 5/5): A significant portion of the U.S.-Europe gap shrinks once countries are made comparable by excluding births under 22 weeks, under 500 grams, and plural births; reporting differences explain about a third of the raw gap. Prematurity and birth-weight distribution (Priority: 5/5): The U.S. has more preterm and low-birth-weight infants than comparison countries, and adjusting for birth weight explains a large share of the gap with Finland, though less of the gap with Austria. Neonatal versus post-neonatal mortality (Priority: 5/5): The paper’s most surprising result is that, conditional on birth weight, the U.S. does very well in the first week/month after birth but does much worse from one month to one year, suggesting the main problem is not hospital care. Socioeconomic status, inequality, and policy relevance (Priority: 4/5): Mortality differences are concentrated among lower-SES groups in the U.S.; high-SES U.S. mothers look similar to comparable women in Europe. Oster argues that poverty and social support matter more than medical technology for the remaining gap. Policy implications: home visiting and broader social context (Priority: 4/5): The discussion turns to interventions like home-visiting programs and the limits of importing Scandinavian-style parenting or health models into the U.S. without addressing broader inequality and labor-market disadvantage.

Key Arguments: Raw infant mortality comparisons overstate the U.S. disadvantage because countries differ in how they report very early live births; after standardizing definitions, part of the gap disappears. Higher prematurity and a heavier share of low-birth-weight births explain a substantial portion of the remaining U.S. disadvantage, especially relative to Finland. Conditional on birth weight, the U.S. is not lagging in neonatal care; the country does quite well in the first month of life, which weakens the case that NICU quality is the main culprit. The excess U.S. mortality is concentrated in the post-neonatal period, where deaths are more often associated with SIDS, accidents, and some assaults rather than congenital abnormalities. Socioeconomic differences inside the U.S. are much larger than in Finland or Austria; high-SES U.S. mothers have infant outcomes similar to European peers, implying that the problem is uneven distribution of risk rather than a universal national failure. Policy solutions should focus more on social supports for mothers and infants at home, such as home visiting, and on reducing poverty and labor-market disadvantage, rather than only on hospital-based interventions.

Data Points: U.S. infant mortality rate: about 6 per 1,000 live births - Baseline national rate discussed for the United States Finland/Sweden infant mortality rate: about 3 per 1,000 live births - Frontier comparison countries used in the discussion Raw U.S.-Europe gap: about 3 deaths per 1,000 - Difference before adjustments for reporting and comparability Adjusted U.S.-Europe gap: about 2 deaths per 1,000 - Gap after limiting to comparable reporting categories Excess infant deaths in the U.S.: about 12,000 per year - Implied by the raw rate difference Infant mortality in 1900: about 1 in 10 - Historical U.S. level used to show long-run improvement Current historical improvement: roughly 20-fold decline since 1900 - Comparison of infant mortality over the century India infant mortality rate: about 40 per 1,000 - Illustrates conditions in a large developing country Some poor-country infant mortality rates: over 100 per 1,000 - Examples include Afghanistan and some African countries Home births in the U.S.: about 0.5% - Used to explain that missing birth reporting is limited Share of the U.S.-Europe gap explained by reporting: about one-third - Result from excluding non-comparable very early births and very small infants Birth-weight adjustment effect vs Finland: closes about 75% of the gap - Counterfactual based on matching Finland’s birth-weight distribution Birth-weight adjustment effect vs Austria: closes less than one-third of the gap - Counterfactual showing Austria is more similar to the U.S. on birth weight Average Finnish birth-weight advantage: more than 200 grams heavier - Compared with U.S. babies on average U.S. preterm birth rate: about 10% - Used to illustrate the scale of prematurity U.S. excess post-neonatal mortality gap: about 1 per 1,000 - Difference from about one month to one year of life Annual deaths implied by post-neonatal gap: about 4,000 deaths per year - Illustrative magnitude of the later-life mortality difference Cook County, south side Chicago infant mortality: about 16 per 1,000 - Example of severe within-U.S. disparities among low-SES Black women

Pivotal Quotes: "the U.S. is doing very well early on. And then when we move to the period from a month to a year of life... that is the period in which the U.S. is really lagging behind other countries." — Emily Oster: Summarizing the paper’s key finding on timing of infant deaths "the U.S. is not falling down... in the medical technology part of this. Instead, the place that the U.S. seems to be doing worse is in the more mundane issues of what is happening before birth... and then what's happening... after people are home with their infants." — Emily Oster: Her broad interpretation of the results "the issues that the U.S. is having are not issues about the kind of medical technology part of this." — Emily Oster: Final takeaway on what drives the mortality gap

Implications: The U.S. should not assume higher infant mortality mainly reflects bad hospitals. The bigger opportunity is reducing prematurity, supporting low-SES mothers, and improving home-based and community interventions that address conditions after discharge.