Is breastfeeding the key to exam success?

Episode Summary

Executive Summary: The episode examines headlines claiming longer breastfeeding improves GCSE results. Tim Harford and economist Emily Oster explain that the reported association is likely confounded by family background, education, and resources rather than causal. Better-controlled studies, including sibling comparisons and a Belarus randomized trial, find little or no effect on later cognitive outcomes, though breastfeeding does show some early-life health benefits.

Main Topics: Headline claim about breastfeeding and exam success (Priority: 5/5): The show opens by questioning newspaper reports that breastfeeding for a year makes children 38% more likely to get A grades, framing the claim as potentially misleading and overly causal. Correlation versus causation (Priority: 5/5): Emily Oster explains that longer breastfeeding is associated with higher test scores, but families who breastfeed longer differ systematically in education, resources, and cognitive ability, making causality hard to establish. Limits of observational adjustments (Priority: 4/5): The Oxford study adjusted for some factors like mother’s education and maternal cognitive test scores, but many unobserved family differences remain, limiting causal interpretation. Sibling comparisons as a stronger design (Priority: 5/5): Comparing siblings raised in the same family helps control for many background factors; Oster notes that these studies tend to show no effect of breastfeeding on later cognitive outcomes. Randomized evidence from Belarus (ProBit trial) (Priority: 5/5): The Belarus ProBit encouragement trial is presented as a gold-standard test: it increased breastfeeding rates, but found no significant differences in cognitive test performance later on. Health benefits in early life (Priority: 4/5): The discussion distinguishes long-term educational claims from better-supported early-life benefits such as fewer gastrointestinal illnesses, less eczema, possible fewer ear infections, and stronger benefits for premature babies in NICU.

Key Arguments: Longer breastfeeding is correlated with higher exam scores, but correlation alone does not prove breastfeeding causes better performance. Mothers who breastfeed longer are more likely to have higher education, more resources, and partners with higher education, all of which also predict better child outcomes. The Oxford study adjusted for some variables, but many relevant differences between families remain unmeasured. Sibling-comparison studies reduce family-level confounding and generally find no detectable cognitive benefit from breastfeeding. The Belarus ProBit randomized encouragement trial is stronger evidence and found no meaningful differences in later cognitive test scores. Breastfeeding does appear to have causal early-life health benefits, especially for digestion and some infant illnesses, and for premature infants in intensive care. The strongest evidence does not support claims that breastfeeding substantially boosts long-term exam results or cognitive outcomes.

Data Points: Exam outcome claim: 38% more likely to get A's - Headline claim about kids breastfed for at least a year Breastfeeding duration categories: Less than 2 months; 2–4 months; up to 1 year; more than 1 year - Breastfeeding exposure groups used in the Oxford study Age at test outcome: 16 years old - GCSE exam results and cognitive outcomes measured at age 16 Study location: Oxford University - Researchers analyzed breastfeeding and GCSE performance Randomized trial: ProBit trial (1990s Belarus) - Encouragement design used to increase breastfeeding rates in a randomized setting Cognitive result in randomized trial: No significant differences - ProBit trial found no meaningful later cognitive test performance differences Health outcome: Less gastrointestinal illness - Breastfed children showed lower early-life gastrointestinal illness rates Health outcome: Less eczema - Breastfeeding associated with reduced eczema in early life NICU benefit: Large impacts on digestive health for preterm babies - Breastfeeding benefits were especially strong for premature infants in intensive care

Pivotal Quotes: "correlation and causation are not the same" — Emily Oster: Explaining why the breastfeeding–exam score association should not be read as causal "kids who are breastfed are less likely to have gastrointestinal illness early in life" — Emily Oster: Describing the best-supported health benefits of breastfeeding "you just don't really see that supporting a link with breastfeeding" — Emily Oster: Summarizing the lack of strong evidence for long-term cognitive or health benefits

Implications: Listeners should be skeptical of headlines turning breastfeeding associations into parenting mandates. The best evidence supports short-term health benefits, but not major claims about improving GCSEs or long-term intelligence.