306. How to Launch a Behavior-Change Revolution

Episode Summary

Executive Summary: The episode follows Angela Duckworth and Katie Milkman launching Behavior Change for Good, a cross-disciplinary project aiming to use large-scale experiments and digital platforms to identify what actually helps people make lasting improvements in health, education, and personal finance. It highlights the promise and limits of behavioral science, the importance of real-world testing, and Danny Kahneman’s caution that meaningful gains will likely be small but still valuable.

Main Topics: Behavior Change for Good: Mission and Structure (Priority: 5/5): Duckworth and Milkman assemble a large research coalition to test behavior-change interventions across health, education, and finance using partner organizations and a custom digital experimentation platform. Behavioral Science as a Policy Revolution (Priority: 5/5): The episode situates the project within the broader behavioral economics movement stemming from Kahneman, Tversky, and Thaler, emphasizing nudges, environmental design, and decision-making biases. Real-World Experimental Platform (Priority: 5/5): The researchers plan to run A/B tests on millions of users through corporate partners, using a platform that links customer-facing programs with backend research infrastructure. Challenges of Sustaining Behavior Change (Priority: 4/5): Speakers stress that short-term interventions often fade, and that changing habits over time is much harder than changing knowledge or attitudes. Cross-Disciplinary Breakouts and Study Design (Priority: 4/5): Conference sessions generate concrete pilot ideas, including SAT prep for students and gym-exercise adherence, while exposing methodological and statistical concerns. Danny Kahneman on Limits and Expectations (Priority: 5/5): Kahneman argues that behavior change should focus on reducing restraining forces in people’s environments and warns that large social interventions usually yield only small practical effects.

Key Arguments: Human beings often undermine their own long-term well-being, so behavior change is a major lever for social improvement. A massive, real-world experimental platform can reveal which interventions work better than theory alone. Behavior change should be tested in partnership with organizations that already have large populations and data streams. Short-term shifts in knowledge or behavior are not enough; the real challenge is durable habit change. Many interventions fail because they try to increase motivation directly instead of removing environmental barriers. Most successful behavior change efforts will probably produce only modest gains, but small percentage improvements can still help millions of people. Research should be driven by the perspective of the people being influenced, not just by the designers’ assumptions. Over-promising may be necessary to attract resources, even though it risks disappointment if outcomes are limited.

Data Points: Research grant target: $100 million - MacArthur Foundation competition that inspired the project Bank of America customer reach: 47 million customers - Example of a partner organization with huge scale 24 Hour Fitness customer reach: 4 million customers - Example of a partner organization with large reach Project domains: 3 - Health, education, and personal finance are the three main focus areas Conference research launch window: Early 2018 - Planned start for pilot studies after the summit Default behavior in laptop study: About 80% - Students stayed in the no-laptop section when the default favored no laptops Student rating of laptop policy: A little over 8/10 - End-of-year survey on whether the opt-in laptop policy facilitated learning Expected impact size: A few percent - Kahneman’s estimate of practically significant improvement from large-scale interventions Behavior change leverage: Millions of people - Duckworth’s claim that even a tiny dent in behavior can materially help many people

Pivotal Quotes: "The problem is ourselves." — Angela Duckworth: Framing the central problem Behavior Change for Good is trying to solve "Most behavior change is actually not desirable, and that's one of the major things that stands in our way." — Adam Grant: Explaining why convincing people to change is difficult "You have to overpromise in order to get anything done." — Danny Kahneman: Discussing the tension between realism and mobilizing support for big social projects

Implications: The episode suggests behavior science is moving from theory to large-scale practice, but success will likely come from modest, durable improvements rather than dramatic fixes. For researchers and institutions, the key is rigorous testing, humility, and designing interventions that fit real human constraints.