Reddit's unemployment forums have become vital support networks during the pandemic, offering advice, emotional support, and a sense of community to those navigating job loss and financial hardship.
LA’MESSAGE is a one-way text messaging service developed by Code for America in partnership with Louisiana to broadcast reminders and guidance to residents enrolled in and eligible for SNAP, Medicaid, TANF, and WIC at key points throughout the benefits enrollment and renewal process.
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.
In response to COVID-19, the Workers Lab and Steady developed the "Income Passport" to streamline gig workers' unemployment benefit applications by pulling income data directly from gig platforms and financial accounts. This tool reduced manual verification time, helped prevent fraud, and improved workers' access to full benefits, with successful tests in Alabama and Louisiana demonstrating significant time savings and improved service delivery.
Minnesota is a good example of an organization that started small in its drive to integrate benefits programs. For instance, its recent statewide rollout of its online integrated benefit application website, MNbenefits.mn.gov, started as a pilot in 2020 with Code for America. The pilot encompassed two counties including Hennepin County, where Minneapolis is located. The pilot later expanded to four counties, then 16 and a tribal nation. The final roll out, which took 12 months to implement, included the state’s 87 counties and three tribal nations.
The team developed an AI solution to assist benefit navigators with in-the-moment program information, finding that while LLMs are useful for summarizing and interpreting text, they are not ideal for implementing strict formulas like benefit calculations, but can accelerate the eligibility process by leveraging their strengths in general tasks.
This paper introduces the problem of semi-automatically building decision models from eligibility policies for social services, and presents an initial emerging approach to shorten the route from policy documents to executable, interpretable and standardised decision models using AI, NLP and Knowledge Graphs. There is enormous potential of AI to assist government agencies and policy experts in scaling the production of both human-readable and machine executable policy rules, while improving transparency, interpretability, traceability and accountability of the decision making.
GetYourRefund.org is an online portal by Code for America that helps low-income individuals claim thousands of dollars in tax credits, even after the traditional tax deadline has passed.
The Benefit Cliffs Calculator helps case managers and public benefit recipients to prepare for benefit cliffs (i.e., declines in benefits due to an increase in earnings). It compares the net resources and benefits available to families under different employment scenarios.
The Policy Rules Database (PRD), developed by the Federal Reserve Bank of Atlanta and the National Center for Children in Poverty, consolidates complex rules for major U.S. federal and state benefit programs and tax policies into a standardized, easy-to-use format. This database allows researchers to model public assistance impacts, simulate policy changes, and analyze benefits cliffs across various household scenarios using common rules and language across different programming platforms.
Nava built flexible and reusable software and design components to make it easier for Vermonters to access their benefits. These components support Vermont’s long-term vision of integrating eligibility and enrollment processes for all of the state’s healthcare and financial benefit programs.
