S6 Ep25: Duflo: Development in the 21st century

Episode Summary

Executive Summary: Esther Duflo reflects on 20 years of J-PAL and the evolution of development economics: major gains in poverty, health, and schooling came alongside a shift toward pragmatic, evidence-based policy. She explains why randomized trials changed the field, the value of long-run evaluation, and how economics now must confront climate justice—especially the responsibility of rich countries for harms borne by the global poor.

Main Topics: The rise of evidence-based development economics (Priority: 5/5): Duflo argues that development economics matured by testing interventions rigorously, learning from failures, and collaborating closely with policymakers in low-income countries. Why randomized controlled trials changed the field (Priority: 5/5): RCTs revealed how much economists and policymakers thought they knew but did not, turning failures like the textbook experiment into a catalyst for rethinking intervention design. Accumulated evidence and the limits of economic priors (Priority: 4/5): She emphasizes that single studies rarely shift beliefs; it takes many experiments to overturn entrenched ideas such as the notion that cash transfers reduce labor supply. Interdisciplinary collaboration in learning, psychology, and networks (Priority: 4/5): Duflo highlights increasingly productive work with psychologists, cognitive scientists, sociologists, statisticians, and computer scientists to understand behavior and design interventions. The importance of long-term impact evaluation (Priority: 5/5): She stresses that some programs only reveal their true value over a decade or more, especially when benefits accrue to children or emerge after graduation from a program. Climate change as development justice (Priority: 5/5): Duflo broadens her agenda to argue that emissions from rich countries are causing avoidable deaths in poor countries, and that climate action must include support and compensation. From helping poor countries to persuading rich countries (Priority: 4/5): She contrasts her earlier influence on pragmatic development policies in low-income countries with the harder task of changing behavior and political priorities in wealthy nations.

Key Arguments: Development economics has made substantial progress over the last two decades, including major reductions in poverty, infant mortality, maternal mortality, and HIV deaths. RCTs are powerful because they often reveal that intuitive or textbook assumptions are wrong, forcing economists to rethink policy design. Policy change usually requires an accumulation of evidence rather than one dramatic study, especially when prior beliefs are strong. Cash transfers do not generally make people work less; in some cases they increase work or productivity by reducing stress and improving focus. Economics benefits from sustained collaboration with psychology and other disciplines because human behavior is shaped by both cognition and incentives. Long-run follow-up is essential for interventions designed to generate lasting change, especially programs targeting ultra-poor households or education. Climate change should be understood as a justice issue: emissions by rich countries impose mortality and welfare costs primarily on poor countries. Meaningful global climate policy will require addressing both emissions reduction and support for vulnerable countries, not just promises or charity rhetoric.

Data Points: Poverty reduction: Reduced by half - Duflo cites global progress in poverty over the last two decades, especially before COVID. Infant mortality reduction: Reduced by half - She notes major health gains during the period covered by J-PAL and broader development efforts. Maternal mortality reduction: Reduced by half - Used as one example of large-scale progress in development outcomes. HIV deaths: Declining since the mid-2000s - Duflo references the sustained decline in AIDS-related mortality. J-PAL founding year: 2003 - The Abdul Latif Jameel Poverty Action Lab was established in 2003. Ultra-poor program support period: 15 to 18 months - The intervention provides intensive support before participants graduate. Ultra-poor transfer size: Equivalent to the yearly GDP in some cases - She describes the asset transfer as very large relative to local incomes. Long-term follow-up: 10 years, with 15 years soon - J-PAL followed some interventions in India over more than a decade to measure lasting effects. Scholarship study year: 2008 - Randomly selected Ghanaian secondary-school students were given scholarships in a study beginning in 2008. Climate cost estimate: $500 billion per year - Duflo’s estimate of the mortality cost of rich-country emissions borne mainly by poor countries. Carbon damage coefficient: $37 per ton of carbon - Her rough calculation multiplies emissions by a mortality-based carbon cost.

Pivotal Quotes: "there is so much we don't know, and so much that we think we know we don't know" — Esther Duflo: On why the textbook RCT on school textbooks was so important in reshaping development economics. "when you give people more, They work more because that opens the space where they are more focused and less stress" — Esther Duflo: On accumulated evidence showing cash transfers do not generally reduce labor supply. "we are literally killing people with our behavior" — Esther Duflo: On the mortality consequences of rich-country carbon emissions for poor countries.

Implications: Development policy should stay evidence-driven, long-term, and interdisciplinary. For climate, the debate shifts from charity to justice: rich countries must cut emissions and support vulnerable nations or face escalating human costs.