Supporting a University of Southern California study on marginal tax rates

PolicyEngine is supporting Matt Unrath, an assistant professor at the University of Southern California, in a new project examining effective marginal and average tax rates facing American families. The study is funded by the U.S. Department of Health and Human Services through the Institute for Research on Poverty.

Understanding how taxes and benefits interact is notoriously difficult. A family's effective tax rate depends not just on federal and state income taxes, but on how dozens of benefit programs phase in and out as income changes. These interactions can create surprisingly high marginal tax rates that discourage work and complicate financial planning for families.

For this project, we're providing our Python microsimulation package, which researchers will run over restricted-use microdata to calculate comprehensive tax and benefit outcomes for households across the country.

For this research, the relevant components of our model include:

• Federal and state income taxes for tax years 2022-2024
• Nationwide benefit programs including SSI, WIC, and Pell Grants
• State-level program rules for SNAP, Medicaid, CHIP, ACA subsidies, and related programs across all 50 states

Our model captures critical state-level policy choices—such as broad-based categorical eligibility and expense deductions—that substantially affect household eligibility, benefit amounts, and effective tax rates. We've also worked closely with NBER staff to ensure our estimates align with TAXSIM for overlapping tax years.

As part of this project, we are extending our model back to tax year 2018 for the relevant programs and adding comprehensive state-specific TANF rules across all 50 states.

We built PolicyEngine to be fully transparent, publicly accessible, and designed so researchers can verify every calculation.

The entire codebase is available on GitHub, and anyone can use our web calculator to explore how policies affect hypothetical households. Policy researchers at Georgetown University, Vanderbilt University, University of Hawaii, and the Niskanen Center have used PolicyEngine's Python package and web interface to study tax and benefit program structures.

We're proud to support rigorous research that helps policymakers understand how the safety net actually works for American families.