One Big Beautiful Bill Act
County-level policy effects on American households
| Federal Tax Cut | |
| Changes to Transfers | |
| State Fiscal | |
| Other Spending | |
| Per-household debt | |
| Net Effect |
Net Effect ($)
Click or tap any county to see detailed policy impact data, or use the search box to find a county by name. Use the dropdowns to switch between income groups and metrics.
Key insight: Counties with more low-income residents see net losses (red), while counties with more high-income residents see net gains (green).
Understanding the Debt Impact:
The Congressional Budget Office estimates that the OBBBA will add $3.4 trillion to the national debt over 2026-2035 — that's more than $26,000 per household in the U.S.
When "Subtract per-household debt increase" is enabled, we subtract $26,300 from all values to show what the impact would be if every household was responsible for their share of the debt.
Primary Data Source
This visualization shows the estimated county-level impacts of the One Big Beautiful Bill Act (Public Law 119-21), based on data from the CBO distributional effects analysis.
What Do the Numbers Mean?
The numbers reported for each county for each different measure are the average annual gain or loss in dollars per household, in 2025 dollars, averaged over the CBO's projection period (2025–2034). Source: CBO 61367.
Why Does the OBBBA Affect Counties Differently?
The CBO's assessment of the impact of the OBBBA breaks down the effects by income category. Since every county has a different mixture of rich and poor, counties with more rich people will benefit more from the bill's tax cuts, while counties with more poor people will be hurt more by the bill's effects on Medicaid and SNAP (Food Stamps) spending. We've taken detailed information on each county's income distribution in order to attribute gains and losses based on the composition of the population of each county. This reveals some important things about the geographic pattern of the bill's impact.
Are There Other Reasons Why the Impacts Would Differ?
Yes, but we don't address them here. The bill will have different effects on different industries and institutions. These too will cause differences.
Anything Else to Be Aware Of?
This analysis does not model:
- Economic multiplier effects — we are dubious about the use of multipliers when unemployment is low.
- Local economic ripple effects — e.g., a county losing Medicaid funding may also lose healthcare jobs, but that secondary effect isn't included here.
- Behavioral responses — while the CBO did projections of how state governments and individuals will respond to the bill and included them in their analysis, we have not done any further behavioral modeling.
If the CBO ever updates its estimates of the effect on families in different income categories we will update the site.
How the Data Were Created (technical)
- County Income Distributions: We used data from the Census Bureau's American Community Survey (ACS) on the income distribution for each county along with macro economic variables to predict the ACS type income distribution variables for the current year.
- Lognormal Mixture Model: The projected income distribution variables are used to estimate a two-component lognormal mixture model for each county using Generalized-Method-of-Moments (GMM).
- CBO Policy Effects: The Congressional Budget Office provides estimates of how the bill affects different income groups nationally.
- Weighted Aggregation: County-level effects are computed by weighting CBO's income group specific effects by the share of each county's population in each national income decile. These are computed using the county specific income distribution models estimates.
What the Metrics Mean
- Net Effect: The average annual per-household impact, combining all tax and transfer changes (averaged over 2025–2034, in 2025 dollars).
- Federal Tax Cut: This is primarily the tax cuts included in the OBBBA, but it is net of reductions in some cash transfers such as the reduced subsidies for health insurance (positive = money saved).
- Changes to Transfers: Reduction in spending on in-kind transfer payments at the state and federal level (negative = benefits cut).
- State Fiscal: Estimated state-level fiscal response including state spending reductions and some increases in transfer spending to offset federal reductions.
- Other Spending: Other spending and revenue changes (e.g. border security, changes to emissions regulations).
Note: CBO estimated effects include estimated responses of state governments and individuals to changes in laws requiring additional record keeping and more frequent assessments of eligibility.
Income Groups
- Low Income: Bottom 20% of each county's income distribution
- Middle Income: Middle 60% of each county's income distribution
- High Income: Top 20% of each county's income distribution
Data Sources
The non-partisan Congressional Budget Office is part of the legislative branch of the U.S. Government. They produced a report on the impact of the OBBBA.
The Census Bureau runs the American Community Survey, providing summary data on income distribution and demographic categories at the county level.
- We retrieved five-year averages of county level income distribution from 2014-2023
- Census Bureau API
The Bureau of Labor Statistics conducts the monthly Current Population Survey used to compute the unemployment rate.
- County level unemployment rates going back to 1990
- BLS Local Area Unemployment Statistics
- Historical and current values for the BLS's Consumer Price Index were obtained from the St. Louis Federal Reserve Bank's FRED site (CPIAUCSL)
- Historical and current values for the BLS's national Unemployment Rate were obtained from the St. Louis Federal Reserve Bank's FRED site (UNRATE)
Technical Notes
Dollar amounts shown in the map are average annual per-household estimates in 2025 dollars, averaged over 2025 to 2034 (CBO 61367). The term 'county' is used to refer to counties and county-like divisions used by the Census to denote location. These include, in some places, townships, planning areas, and other similar geographic divisions.
Visualization by Visualize Policy. Economic modeling by William T. Dickens.
Data version: March 31, 2026. Most values rounded to nearest $50 ($100 for high-income group).
Visualize Policy
William T. Dickens
Did the modeling to produce the data for the map. He is a University Distinguished Professor of Economics and Social Policy Emeritus at Northeastern University. He has been invited to serve the President's Council of Economic Advisors in both Republican and Democratic administrations and served as a Senior Economist from 1993 to 1994. He has advised central banks around the world and has been a visiting scholar with the New York and Boston Federal Reserve Banks.
Zachary Gottschalk
Helped design the website and did background research for the project. He is a student at Northeastern University and a member of Northeastern's Fiscal Challenge team.
Aadit Bhatia
Along with Zachary, Aadit was primarily responsible for the budget/debt visualization tour. He also did background research for the project. He is a student at Northeastern University.
Tamarron Austin
Did background research for the project in its early phases. He is a member of the Northeastern University Fiscal Challenge team.
Raymond Yee
Built the interactive visualization, data pipeline infrastructure, and reproducibility framework. He designed the shareable URL system, county search and selection interface, and automated testing suite. He also achieved a 160x speedup of the GMM estimation pipeline through analytical optimization and parallelization.
Anthropic's Claude Code AI and Microsoft's Copilot AI were used extensively in the development of the programs for the data pipeline and the webpage itself. All code was checked by humans for accuracy.
Visualize Policy wishes to thank many more people who helped with research, gave advice on website design, commented on the design at several points, did background research, and much more. In particular we would like to thank the wonderful people at SCIMaP whose work inspired ours, provided us with the code for their website, and helped shape the direction of our project. We are especially grateful to Alyssa Sinclair (Joan Bossert Postdoctoral Fellow, University of Penn.) for being our contact with the SCIMaP project and for very helpful advice at two crucial junctures in the development of this site.