PolicyEngine ecosystem map

NSF POSE · Winter 2026
Arrow types ideas / feedback funding technical / code data / analysis influence Gov / Validation Funding Users / Research Integrators Contributors / Team Risks Publish research papers (citations, tenure) Score reforms quickly (faster than competitors) Embed eligibility calcs (no rules engine needed) Validate against CBO/JCT (transparency) Foundations: policy research Clients: custom analysis NSF: infrastructure (different value exchanges) Fact-check policy claims (interactives) PolicyEngine policyengine-us policyengine-uk policyengine-core us-data uk-data microdata policyengine.org (web app) custom interactives REST API / policyengine.py Models · Data · Apps · API All open source (AGPL) 10 Government Economists Budget Analyst, Policy Economist (CBO, JCT, Treasury, SSA) Validate, inform scoring 5 Policymakers / Staffers Legislative Dir, Committee Staff (Hill offices, state leg) Quick what-if, constituent Qs 10 Policy Foundations (Arnold, Pritzker) Big grants → research back 5 Research Clients (CRFB, Niskanen, think tanks) Paid analysis / custom tools 5 Government Grants (NSF POSE) Infrastructure funding 5 Data Journalists Reporter, Data Editor (NYT, WaPo) Fact-check, build interactives 10 Academic Researchers Professor, PhD, Postdoc (Econ, Public Policy depts) Publications, grants, teaching 15 Think Tank Analysts Policy Analyst, Fellow, Director (CRFB, Niskanen, Urban, AEI) Fast, credible methodology 5 Benefit Platform Builders Product Manager, Engineer (Benefits screeners, fintech) API integration, auto-updates 10 Policy Advocates Exec Director, Comms Dir (CBPP, advocacy orgs) Credible backing, media reach 5 OSS Contributors Software Eng, Data Scientist (External volunteers) Portfolio, impact, learning 10 PolicyEngine Team Analyst, Engineer (Core team) Build models, demo usage 5 Data/Tool Partners CTO, Product Lead (Census, IRS) Data sharing, co-develop 10 Non-Users (Potential) Using alternatives (TRIM3, etc) Why not? Cost? Trust? 3 Competitors CEO/CTO of rival tools Collaborate? 0 Misusers Bad-faith analysts Cherry-pick results $ grants research NSF $ validate methodology what-if analysis ideas docs reports feedback API analysis amplify PRs, fixes data embeds barriers? How can we scale? • Champions / Evangelists • Citations as validation • Training workshops Assumptions to test: • Think tanks want OSS? • Platforms will use API? • Gov economists engage? • Contributors exist?