Encore episode: Angus Deaton on his Nobel Prize-winning career

Episode Summary

Executive Summary: This episode revisits Angus Deaton’s Nobel-recognized work, tracing his path from accidental economist to influential critic of over-abstract modeling. He explains his innovations in consumer demand theory, savings and aggregation, development economics, Indian poverty measurement, self-reported well-being, aid, and randomized trials. Across topics, his core theme is that micro-level behavior, measurement quality, and human agency matter more than tidy macro shortcuts or fashionable policy claims.

Main Topics: Deaton’s path into economics and methodological style (Priority: 5/5): He entered economics almost by accident after struggling in mathematics at Cambridge, but found the field rewarding because it combined writing, empirical work, and tools borrowed from other disciplines. He repeatedly describes his approach as 'tinkering' and stresses skepticism toward overly narrow academic circles. Almost Ideal Demand System and consumer choice (Priority: 5/5): Deaton explains the demand-system work he won the Nobel for: building a flexible, theory-consistent way to model how households allocate spending across goods in response to prices, income, and demographics, including cross-price effects and welfare analysis for policy and antitrust. Aggregation, savings, and the Deaton paradox (Priority: 5/5): He argues that macro-level consumption and income patterns can mislead because they average away heterogeneity across households. His 'paradox' shows why aggregate data can imply consumption should be noisier than income, while micro data reveal household-level smoothing is still plausible. Development economics, India, and poverty measurement (Priority: 5/5): Deaton describes his long engagement with India, including cohort-based 'pseudo-panels,' work on calorie consumption, and revisions to poverty estimates using better price data. He emphasizes that measurement errors can dramatically distort perceptions of poverty and growth. Critique of foreign aid and agency (Priority: 4/5): He is not anti-aid, but argues many aid programs are designed for donors rather than recipients and can weaken local accountability or entrench corrupt governments. He favors aid that builds knowledge, institutions, or bargaining power rather than directly channeling large sums through weak states. Randomized controlled trials and behavioral economics (Priority: 4/5): Deaton accepts RCTs as useful but rejects the hype around them, warning that randomization does not eliminate all confounding, blinding is often absent, and average treatment effects may not generalize. He sees behavioral economics as valuable but incomplete without welfare economics and better mechanisms. Progress, inequality, and cautious optimism (Priority: 4/5): He closes by framing growth as an 'endless dance' with inequality: innovation should be rewarded, but societies must prevent elites from pulling up the ladder behind them. He remains cautiously optimistic while insisting policy must respect evidence and institutional limits.

Key Arguments: Household-level behavior cannot safely be inferred from aggregate data; averaging across people can hide major differences in income risk, saving, and consumption. The Almost Ideal Demand System improved demand analysis by being flexible, theory-consistent, and useful for practical policy questions such as antitrust damages and welfare effects. Better measurement often changes the conclusion more than grand theory; in India, poor price-index construction and weak survey methods distorted poverty estimates. The 'poverty trap' story based on calories was overstated; in poor Indian settings, basic calories were not so expensive that people were structurally trapped by undernourishment alone. Foreign aid can harm when it bypasses local accountability, props up governments, or serves donor moral comfort more than recipient welfare; however, health interventions like PEPFAR can save lives. RCTs are informative but not definitive: they do not guarantee identical groups, are often unblinded, can suffer attrition, and often fail to answer whether results scale or generalize. Behavioral economics is useful because real people do deviate from simple rational models, but abandoning welfare analysis entirely would make public policy harder, not easier. Economic growth often benefits a small group first; the policy challenge is to reward innovation without allowing rent-seeking or plutocratic lock-in.

Data Points: Bank of England stint: 13 miserable months - Deaton’s first economics job before returning to Cambridge as a research assistant Observations in early demand work: 14 observations - He notes the limited postwar UK annual data available for early demand-system estimation Deaton and Muellbauer textbook: Widely used - Their duality-based textbook on price theory remains a standard reference Time horizon for receiving old contacts after Nobel: 40 or 50 years - He says the prize brought messages from people he had not heard from in decades Calorie cost in rural India: About 5% of a day wage - Used to argue against a binding calorie poverty trap in the poorest rural areas Indian child malnutrition: Half of all children - He cites this as a major unresolved development puzzle despite growth World War/aid reference date: 1968 - The Biafran War and televised starvation images that helped shape public aid debates Inflation example: 20 pence - He describes noticing weekly price increases during Britain’s inflation burst Inflation rate: 25–40% for a year or so - Used in his early savings/inflation observation in Britain Sample survey claim: Tiny fraction of the cost - Mahalanobis’s insight that Indian crop estimates could be done with sample surveys rather than full counts Aid example: PEPFAR - U.S. program cited as a case where aid saved lives through antiretroviral treatment RCT sample illustration: 100 people vs 100 people - He uses this to explain why randomization does not ensure identical groups Genome complexity illustration: Billions of gene pairs - Used to argue that perfect balance across all relevant factors in RCTs is impossible Growth/inequality framing: 2011 Cricket World Cup - He analogizes political resistance to revising statistics to changing cricket rules so India would not win

Pivotal Quotes: "I stumbled into economics really by accident." — Angus Deaton: Explaining how he moved from mathematics to economics at Cambridge "the endless dance between progress and inequality" — Angus Deaton: His description of the central theme of his book The Great Escape and his broader view of development "it’s not that I’m against randomized controlled trials, and I think no one in their right mind would say that" — Angus Deaton: Clarifying that his criticism is about hype, limits, and generalizability rather than outright rejection

Implications: Listeners should come away skeptical of simplistic aggregate statistics, overconfident interventions, and one-size-fits-all development policy. Deaton’s work points toward better microdata, better measurement, and more humility in economics and aid design.