Presentations
Slide decks from PolicyEngine talks, webinars, and meetings
PolicyEngine: data pipeline, property income and Budget planning
Follow-up to the February and March meetings: the Microcosm UK data pipeline and property income, UK work published since April, and the Autumn Budget 2026 plan.
BLS TAXSIM seminar
A 60-minute talk on the TAXSIM emulator, validation, and tax and benefit imputation for the Consumer Expenditure Surveys.
Law is an alignment problem
What if every bill had a forecast? Policy rules, models, and forecasts that resolve against reality.
What if every bill had a forecast?
Five-minute lightning talk: law as an alignment loop, one bill through the stack, a forecast the IRS will grade, and the loop run backwards.
Law is an alignment problem: backup slides
Backup slides for questions after "Law is an alignment problem" at The Overhang 2026.
The new child poverty numbers, and how state reforms would affect them
Launch webinar for the Child Poverty Impact Dashboard, four hours after the Census Bureau publishes the 2025 income and poverty estimates.
AI-driven rules engine development
Rules, data, microsimulation: AI-assisted encoding and the Axiom rules engine, the calibrated microdata underneath, and PolicyEngine graded against external scorekeepers
Six projects, one chain of checks
PrivatePrivate funder check-in — the stack, its feedback loops, and current progress
From open models to executable law
PolicyEngine and Axiom — open models, simulation-ready microdata, and encoding and verifying policy rules
L0 regularization for subnational microsimulation calibration
PolicyEngine IMA 2026 deck on using L0 regularization to prune a full Populace candidate microdataset while preserving calibration fit.
PolicyBench
PolicyEngine webinar — can AI calculate taxes and benefits?
Policy analysis in the age of AI
AI does exponentially longer work but still misses on taxes and benefits. PolicyEngine splits policy analysis in two — what the law is, and what a change does — scaled by agents: Axiom and Thesis.
The open engine for taxes and benefits
A civic-tech demo: PolicyEngine computes taxes and benefits for any household and reform — and how agents now scale that into massive digital public goods: determinism (Axiom) and prediction (Thesis).
CliffWatch
PolicyEngine deck introducing CliffWatch
Model-building and AI
Where AI is, where it is heading, and what it could mean for federal policy modeling
PolicyEngine's Open-Source Ecosystem
From policy design to benefit access: shared rules infrastructure powering measurement, navigation, and screening.
Harnessing AI for policy research
EAGxDC talk on how PolicyEngine is using AI to accelerate rigorous, transparent policy analysis.
PolicyEngine
Open microsimulation for US tax-benefit policy analysis
Reproducing a PolicyEngine result
PrivateWorking meeting with Lars Vilhuber (AEA Data Editor), Tara Watson, John Sabelhaus, and the PolicyEngine team
Making organizations legible to AI
CIO4Good panel talk — from legislative text to runnable analysis
PolicyEngine: from simple wins to strategic implications
Workshop support deck — early wins, Evo1-Evo6 encoding journey, skills exercise, strategic implications
Building data foundations for policy impact
How open-source infrastructure turns policy rules into public goods
Building state capacity through open-source policy simulation
Lightning talk on how PolicyEngine democratizes policy analysis
Rhode Island child tax credit calculator
Live demonstration of the RI CTC calculator and Governor McKee's proposal
Automating tax and benefit policy modeling with multi-agent AI
How we built a system that transforms a policy request into a complete, tested pull request in 90 minutes
Simulating tax and transfer policies' inequality impacts with PolicyEngine
Guest lecture for PPOL 6362: Inequality and public policy
Modeling state tax policy with PolicyEngine
Learn how to model and measure the impact of state tax policies using PolicyEngine
Scope of the model: policies, households, and dynamics
Overview of PolicyEngine microsimulation model structure and data
Modeling Minnesota tax policy with PolicyEngine
PrivateDemo of PolicyEngine for the Minnesota House Research Department