Recommendation in equilibrium for the NYC High School Match
Talkat Bridging Prediction and Intervention Problems in Social Systems (Simons Institute, 2026) · 2026-01-12..01-18·
summary
Collaboration with NYC Public Schools designs and deploys an email-based informational intervention that recommends nearby high-performing high schools with individualized admissions likelihoods; a match simulator built from lottery numbers and admission rules achieves 95% accuracy on counterfactual admits; recommendations are congestion-aware so capacity effects are guaranteed within design; a conservative RCT yields a rise from 10% to 16% application rate to nudged schools (about a 60% relative increase) without breaking admission guarantees.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:33Two NYC High School Match projects
- 5:39Need for recommender system
- 9:44Disparities in applications
- 13:14Measuring undermatching
- 16:56Geographic school selection
- 19:16Undermatching in portfolios
- 21:36First-choice application patterns
- 24:24Descriptive study findings
- 25:31Randomized controlled trial design
- 27:50Capacity-constrained nudge challenges
- 29:51RCT intervention results
- 30:58Perfect simulations for interventions
- 31:31Future for government recommenders
related talks
30mLLM-Based Recommendation Systems
35mDating Algorithms: What Are They Really Optimizing?
27mDreadful Frailties in Propensity Score Matching and How to Fix Them
27mThe ethical algorithm
41mRecommender Systems: Crafting Effective Personalised Experiences
40mNext Best Action: Maximizing Customer Value
23mWhat cyber security can teach us about COVID-19 testing
16mMaking recommendations explainable