PolicyEngine US is a Python-based microsimulation model of the US tax and benefit system. It models federal individual income taxes (including credits), major benefit programs, and state income taxes (currently in six states). The PolicyEngine US package can be used as a Python package, via the PolicyEngine API, or via the policyengine.org web app.
At Rules as Code Demo Day Seth Hartig from the National Center for Children in Poverty (NCCP) and Bank Street College demoed the Policy Rules Database (PRD), a collaborative effort between the Federal Reserve Bank of Atlanta and the NCCP. The primary purpose of the PRD is to simplify the interpretation of all programs by creating a common structure and a common terminology. The repository allows for research on public assistance programs and tax policies, and helps users model benefits cliffs on career pathways. The PRD is supported by a technical manual with pseudocode that helps guide integration and usage in other platforms.
We kicked off Rules as Code Demo Day with Alex Soble of 18F and Mike Gintz of 10x presenting their Eligibility APIs Initiative that explores whether APIs and rules as code might improve the efficiency and effectiveness with which federal public benefits programs communicate their policy to states. They demonstrated their original prototype, and how the open source code has now been extended into several initiatives.
Building on our February 2022 report Benefit Eligibility Rules as Code: Reducing the Gap Between Policy and Service Delivery for the Safety Net, the Beeck Center’s Digital Benefits Network (DBN) recently held a convening to share progress and potential in digitizing benefits eligibility and to begin addressing how a national approach could be started.
The DBN’s Rules as Code Community of Practice (RaC CoP) creates a shared learning and exchange space for people working on public benefits eligibility and enrollment systems — and specifically people tackling the issue of how policy becomes software code. The RaC CoP brings together cross-sector experts who share approaches, examples, and challenges. Participants are from state, local, tribal, territorial, and federal government agencies, nonprofit organizations, academia, and private sector companies. We host recurring roundtable conversations and an email group for asynchronous updates, insights, and assistance.
This report documents four experiments exploring if AI can be used to expedite the translation of SNAP and Medicaid policies into software code for implementation in public benefits eligibility and enrollment systems under a Rules as Code approach.
This article explores how AI and Rules as Code are turning law into automated systems, including how governance focused on transparency, explainability, and risk management can ensure these digital legal frameworks stay reliable and fair.
The Policy2Code Prototyping Challenge explored utilizing generative AI technology to translate U.S. government policies for public benefits into plain language and code, culminating in a Demo Day where twelve teams showcased their projects for feedback and evaluation.
The team explored using LLMs to interpret the Program Operations Manual System (POMS) into plain language logic models and flowcharts as educational resources for SSI and SSDI eligibility, benchmarking LLMs in RAG methods for reliability in answering queries and providing useful instructions to users.
