Raj Chetty on Teachers, Social Mobility, and How to Find Answers to Big Questions

Episode Summary

Executive Summary: Tyler Cowen interviews Raj Chetty about his personal path and research on mobility, education, and incentives. Chetty traces his interests to family history, India, Milwaukee, and Harvard, then explains his research style: use big data to answer classic questions clearly and credibly. The conversation focuses on taxes, dividends, teacher quality, segregation, adoption, and the causes of upward mobility, plus future work using social-network data.

Main Topics: Personal background and formative experiences (Priority: 5/5): Chetty links his family’s Tamil, Gandhian, and academic traditions to his interest in fairness, mobility, and data-driven social science. Childhood experiences in India and Milwaukee shaped his sensitivity to inequality and opportunity. Research strategy and competitive advantage (Priority: 5/5): Chetty explains that his edge is not just access to data, but the combination of big questions, conceptual clarity, careful execution, and team management that turns complex projects into simple, durable findings. Behavioral economics and salience (Priority: 4/5): They revisit Chetty’s work on sales taxes, showing that consumers react less to taxes when they are added at checkout. The discussion broadens to pricing, shrouding, and why firms use hidden fees and add-ons. Dividends, incentives, and corporate behavior (Priority: 3/5): Chetty discusses his paper on dividends as evidence that payout behavior is more rational than often assumed, while still puzzling. He emphasizes managerial incentives, signaling, and the role of tax changes. Teacher quality and early-childhood effects (Priority: 5/5): A major segment examines kindergarten teacher value, why effects show up later in life rather than test scores, and how experience, non-cognitive skills, and teacher selection affect long-run outcomes. Mobility, segregation, and neighborhood effects (Priority: 5/5): The conversation centers on Chetty’s mobility research: upward mobility varies by place, is linked to segregation and integration, and appears persistent over time. He argues these patterns are not mainly explained by temporary booms. Future research on social networks and technology (Priority: 4/5): Chetty describes plans to use Facebook and social-network data to identify causal effects of friendships and integration on life outcomes, and speculates that technology could either reduce or intensify segregation.

Key Arguments: Chetty’s work matters because he combines classic economic questions with newly available data and careful identification, producing findings that are both accessible and hard to overturn. The most important methodological contribution is often conceptual: turning an impossible question into one answerable with imperfect data, then executing with high clarity and rigor. Sales taxes and hidden fees show that consumers do not fully process all prices; firms exploit salience, which helps explain add-on pricing and related shrouding practices. Dividend payouts are not fully irrational; they can serve as commitment and signaling devices, especially when managers own more shares and are more aligned with shareholders. Teacher quality affects adult outcomes even when later test scores barely move, implying that schools transmit non-cognitive skills such as discipline and social behavior. Experience matters for teachers, but the more important predictor is value-added/performance; pay systems should reflect effectiveness, not only years served. Upward mobility differences across places are persistent and are not mainly driven by temporary economic booms; segregation and integration likely play a causal role. Adoption studies provide stronger evidence than twin studies for environmental effects because random assignment to families is closer to a natural experiment. Relative intergenerational mobility in the U.S. has not worsened much over 40 years, but absolute mobility has fallen sharply, which is closer to the public intuition of the fading American Dream. Future work may identify causal effects of social networks by exploiting quasi-random variation in children’s friendship exposure across grades.

Data Points: Family origin: Tamil Nadu / New Delhi - Chetty describes his family’s Tamil roots and his birth in New Delhi after his parents had moved between India and the U.S. Age at first computer exposure: 3 or 4 - His father brought home a computer from Belgium, and Chetty began simple programming as a preschooler. Age at Harvard RA work: 17 or 18 - He says Martin Feldstein gave him independence when he started as a research assistant at Harvard. Tennessee kindergarten study sample: 12,000 kids - Chetty cites this as evidence that better kindergarten classrooms improve non-cognitive outcomes. New York City school study sample: 2.5 million observations - A later study with much larger data revealed the first two years of teaching experience matter most. Teacher-experience threshold in earlier paper: 10+ years - In the original kindergarten-star analysis, experience above 10 years was used as a simple split and appeared predictive. Early childhood prediction power: about a quarter of adult variation - Chetty says early test scores and related measures explain roughly 25% of adulthood outcomes. Bay Area marriage rate for women in mid-30s: only something like a quarter or less married - He cites this to explain why household income outcomes for daughters in the Bay Area look worse than individual earnings. Income impact for boys in Baltimore: about 30% earnings loss - Chetty says growing up in concentrated-poverty areas like Baltimore is especially harmful for boys. Relative mobility change over time: roughly unchanged over 40 years - He argues the rank-order chances of moving up or down have not shifted much over recent decades. Absolute mobility change: fallen dramatically - He says children’s probability of earning more than their parents has dropped sharply. Sales-tax paper result: consumers are less discouraged by add-on taxes - Chetty explains that checkout taxes are less salient than taxes embedded in the sticker price. Teacher value-added evidence: strong predictor of later success - He notes that teachers’ test-score gains and experience are among the best predictors of long-run outcomes.

Pivotal Quotes: "I think part of what we try to do is exactly as you said, take old questions." — Raj Chetty: Chetty explains his research strategy and why big data matters less than the conceptual framing of classic economic questions. "The impact reemerges when kids become adults." — Raj Chetty: Discussing kindergarten teacher effects, he notes that test-score gains fade but adult outcomes still improve. "What I see here in Silicon Valley is that technology has completely transformed the way we live our lives through the private sector. I think it's to date had less of an impact on social issues." — Raj Chetty: He closes by arguing that technology could be applied far more aggressively to segregation, mobility, and urban policy.

Implications: Chetty’s work suggests policy should focus less on simplistic inputs and more on social networks, segregation, teacher effectiveness, and early-life environments. For researchers, the frontier is using new data to test causal mechanisms at scale.