Abstract
This paper explores a thought experiment in participatory budgeting at the federal level. Using survey data from 2019 that captured U.S. citizens’ preferences for government spending across thirteen program areas, we model a system in which individual taxpayers can direct percentages of their income‑tax payments to specific categories such as education, veterans services, infrastructure, health care, scientific research, and entitlements. We outline a technical design for a secure, identity‑verified allocation portal integrated with existing payroll and tax infrastructure, describe the legal and equity challenges inherent in such a system, and compare the resulting “citizen‑directed” budget to the actual 2019 federal budget. We also examine how regional income differences would influence the distribution of funds. The thought experiment reveals how strongly public priorities diverge from current spending patterns and highlights the practical constraints that must be addressed before implementing any form of citizen‑driven tax allocation.
Introduction
The United States federal budget reflects a complex interplay of statutory obligations, political negotiations, and macroeconomic constraints. Yet public discourse often revolves around whether elected officials spend tax dollars in ways that align with citizens’ priorities. In 2019, a nationwide survey asked Americans whether federal spending on programs ranging from veterans benefits and education to foreign aid and unemployment assistance should be increased, decreased, or kept the same. Majorities favored increases in most areas, particularly in domestic investments like infrastructure and education. This paper uses those survey responses as the basis for a thought experiment: What if individual taxpayers could allocate their own income‑tax payments across broad categories? Could such a mechanism improve the alignment between public priorities and federal expenditures?
To investigate these questions, we first propose a conceptual architecture for a tax‑allocation system that would allow citizens to designate percentages of their tax liability toward a curated list of spending categories while maintaining legal obligations for mandatory programs. We then simulate how the 2019 federal budget would look if allocations were made strictly in proportion to survey preferences, comparing the hypothetical distribution to actual outlays. Finally, we explore how variations in median household income and household counts across states affect the relative influence of different regions under a citizen‑directed system. This exploration aims not to advocate immediate policy adoption but to illuminate the tension between public desires and the realities of budgeting within a constitutional and legal framework.
Rationale for Choosing 2019
Comprehensive survey data: In March 2019 Pew Research Center surveyed 1,503 U.S. adults about whether federal spending should be increased, kept the same or decreased for 13 program areas. Majorities wanted higher spending on veterans’ benefits and education (72% each), about six‑in‑ten wanted more infrastructure spending (62%) and roughly half supported more spending on Medicare and environmental protection (55% each). This survey provides clear, recent preferences across a wide range of government functions.
Recent income and population data: The 2019 American Community Survey (ACS) is the most recent decennial census period with reliable state‑level statistics. It reports that the median U.S. household income was $65,712 with about 122.8 million households.
Available budget data: The Congressional Budget Office’s 2019 federal budget infographic summarizes outlays by major category. Social Security was the largest program ($1.0 trillion), followed by Medicare ($644 billion), defense ($676 billion), nondefense discretionary programs ($661 billion), other mandatory programs ($642 billion), Medicaid ($409 billion) and net interest ($375 billion).
Together these sources allow a grounded comparison between what the government actually spent in 2019 and what it might have spent if individual taxpayers had been allowed to allocate portions of their income‑tax payments among spending categories.
Technical Design for Citizen‑Directed Tax Allocation
Secure digital platform integrated with tax filings. Each taxpayer would log into a secure portal provided by the Internal Revenue Service (IRS) or a third‑party provider. Multi‑factor authentication and identity‑verification tools (e.g., linking Social Security numbers and biometric checks) would ensure each person can allocate only their own tax liability. The platform would incorporate accessibility standards so people with disabilities can participate.
Connection to payroll and annual returns. Employers would send withholding information (similar to existing W‑2 reporting) but would also transmit each worker’s allocation schedule. For self‑employed individuals, the allocations could be specified when quarterly estimated payments or annual tax returns are filed. The system would adjust tax withholding rates only; the underlying statutory tax rates and liability calculations would remain unchanged.
Category selection interface. Taxpayers could distribute percentages of their income‑tax payments across broad program categories (e.g., education, veterans benefits, Social Security). To avoid over‑specialization, the number of categories would be limited (similar to the 13 categories in the Pew survey) and each would be mapped to existing budget functions. Real‑time “progress bars” would display how much aggregate funding each category receives from participating taxpayers, helping users understand whether essential programs are at risk of underfunding.
Aggregation and reallocation algorithm. After the April 15 filing deadline, allocations from all taxpayers would be aggregated. Programs would receive funds proportionate to citizens’ preferences but subject to constitutional and statutory constraints (for example, legal obligations for Social Security benefits must be met). The algorithm would need to handle situations in which preferences conflict with mandatory obligations by scaling allocations proportionally across discretionary categories while still meeting legal requirements.
Transparency and feedback. The system would publish anonymized statistics showing how different regions and demographic groups allocated their taxes and what share of program budgets came from citizen allocations. Over time, Congress could use the data to adjust appropriations in line with constituents’ priorities.
Legal and policy considerations. Significant legislative changes would be required. Article I of the U.S. Constitution grants Congress the power of the purse; a citizen‑allocation system would need to be framed as advisory or integrated into appropriations bills. Mandatory programs (e.g., Social Security and Medicare) are governed by existing entitlement laws and could not be starved of funds without changing those laws. Privacy protections would be necessary to prevent discrimination or political retaliation based on a person’s tax allocations.
Actual 2019 Federal Outlays
The table below summarizes major components of the FY 2019 federal budget. Outlays totaled about $4.4 trillion.
Major program or category | FY 2019 outlays (≈) | Share of total budget |
Social Security | $1.0 trillion | ~22.7% |
Medicare | $644 billion | ~14.6% |
Other mandatory programs (federal civilian and military retirement, unemployment compensation, veterans pensions, the Earned Income Tax Credit, SNAP and other income‑security programs) | $642 billion | ~14.6% |
Defense (discretionary) | $676 billion | ~15.4% |
Nondefense discretionary spending (transportation, education, veterans benefits, health, housing, environmental protection, etc.) | $661 billion | ~15.0% |
Medicaid | $409 billion | ~9.3% |
Net interest on the federal debt | $375 billion | ~8.5% |
These categories align with the government’s functional classification but are broader than the specific program areas asked about in the Pew survey. For example, “nondefense discretionary” spending includes many activities—education, veterans benefits, environmental protection and scientific research—that participants might have allocated differently if given a choice.
How Pew Survey Preferences Translate to a Hypothetical Budget
Survey results
The March 2019 Pew survey found that more Americans wanted increased federal spending than cuts for every program asked about. Majorities supported increases for veterans benefits and education (72%), infrastructure (62%), Medicare and environmental protection (55% each), health care (53%), scientific research (52%) and Social Security (48%). Fewer respondents wanted increases in aid to the needy abroad (35%) or assistance to the unemployed (31%), but even in these categories more people favored increases than cuts. The table below uses the percentage of respondents supporting higher spending as a simple proxy for how citizens might direct their tax dollars.
Program area (Pew survey wording) | % of respondents who would increase spending | Normalized weight* |
Education | 72% | 10.9% |
Veterans benefits | 72% | 10.9% |
Rebuilding highways, bridges and roads (infrastructure) | 62% | 9.4% |
Medicare | 55% | 8.3% |
Environmental protection | 55% | 8.3% |
Health care (not including Medicare) | 53% | 8.0% |
Scientific research | 52% | 7.8% |
Social Security | 48% | 7.2% |
Assistance to needy in the United States | 46% | 6.9% |
Anti‑terrorism defenses in the U.S. | 42% | 6.3% |
Military defense | 40% | 6.0% |
Aid to needy people around the world (foreign aid) | 35% | 5.3% |
Assistance to unemployed Americans | 31% | 4.7% |
*Normalized weight shows each category’s share of the total positive responses (the percentages above sum to 703%). It approximates how taxpayers might divide 100% of their income‑tax dollars across categories if they allocate strictly in proportion to their preference for “increasing” spending.
Hypothetical citizen‑allocated budget
Multiplying the normalized weights by $4.4 trillion of 2019 outlays yields a hypothetical “voter‑directed” budget. The resulting amounts are rounded to the nearest tenth of a billion dollars.
Program area (Pew category) | Estimated allocation if citizens direct 2019 taxes (≈) | Notes |
Education | ~$477.8 billion | Far higher than the roughly $70 billion that the federal government actually spent on education in 2019 (most education funding comes from state and local governments). |
Veterans benefits | ~$477.8 billion | Actual federal spending on veterans programs—both discretionary and mandatory—was about $194.6 billion in 2019. Citizen allocations would more than double this amount. |
Infrastructure (roads, bridges) | ~$411.4 billion | Substantially higher than 2019 federal highway and transit spending (about $90 billion), suggesting Americans want more infrastructure investment. |
Medicare | ~$365.2 billion | Much lower than the $644 billion actually spent because the survey’s support for increases (55%) is modest relative to other priorities. In reality, Medicare spending is legally mandated. |
Environmental protection | ~$365.2 billion | A huge increase compared with EPA and related spending (about $9 billion in 2019), reflecting strong public support for environmental investments. |
Health care (other than Medicare) | ~$351.6 billion | Includes programs like Medicaid and CHIP. Actual Medicaid spending was $409 billion. |
Scientific research | ~$345.0 billion | Roughly ten times the actual 2019 federal research and development budget (~$40 billion), indicating strong latent support for research. |
Social Security | ~$318.6 billion | Drastically less than the $1 trillion that actually went to Social Security benefits. This underscores the tension between preference‑based budgets and statutory entitlements. |
Assistance to needy people in the U.S. | ~$305.4 billion | Slightly lower than the $642 billion spent on income‑security and other mandatory programs. |
Anti‑terrorism defenses in the U.S. | ~$278.5 billion | Includes domestic security and homeland defense. Current funding is significantly lower (homeland security budget was about $51 billion), so citizen preferences would greatly expand it. |
Military defense | ~$265.3 billion | Less than half of the $676 billion actually spent on defense. |
Aid to needy abroad (foreign aid) | ~$232.3 billion | Much higher than the roughly $47 billion spent on international development and humanitarian aid in 2019. |
Assistance to unemployed Americans | ~$205.9 billion | Considerably larger than actual unemployment benefits and job‑training outlays (~$33 billion). |
This exercise highlights stark differences between citizen preferences and the actual budget. Respondents favored large increases for domestic programs (education, infrastructure, environmental protection, scientific research) and aid to the needy both at home and abroad, while defense, Social Security and Medicare received relatively lower weights despite constituting the bulk of current spending. Implementation would therefore require either reducing legally mandated benefits or raising taxes to fund preferred increases—difficult political choices.
How Income and Location Affect Revenue Contributions
Because taxpayers would allocate shares of their own income‑tax liability, variations in household income and the number of households across states would affect how much money flows into each program. Using median household income and household counts from the 2019 ACS, a rough estimate of each state’s total household income can be obtained by multiplying the median income by the number of households. The United States had about 122.8 million households with a median income of $65,712. Summing the product of median income and households across all states yields an estimated pool of household income; dividing each state’s estimated total by the national sum gives its share of total household income, which approximates its share of income‑tax revenue.
The table below lists the ten states with the largest estimated total household incomes in 2019. Together they accounted for roughly 55% of total U.S. household income.
State | Median household income (2019) | Households (2019) | Estimated total household income | Share of U.S. household income |
California | $80,440 | 13,157,873 | ≈$1.058 trillion | 12.86% |
Texas | $64,034 | 9,985,126 | ≈$639 billion | 7.77% |
New York | $72,108 | 7,446,812 | ≈$537 billion | 6.53% |
Florida | $59,227 | 7,905,832 | ≈$468 billion | 5.69% |
Illinois | $69,187 | 4,866,006 | ≈$337 billion | 4.09% |
Pennsylvania | $63,463 | 5,119,249 | ≈$325 billion | 3.95% |
New Jersey | $85,751 | 3,286,264 | ≈$282 billion | 3.42% |
Ohio | $58,642 | 4,730,340 | ≈$277 billion | 3.37% |
Virginia | $76,456 | 3,191,847 | ≈$244 billion | 2.97% |
Georgia | $61,980 | 3,852,714 | ≈$239 billion | 2.90% |
Implications: States with large populations and high incomes (e.g., California, Texas, New York, Florida) would contribute the largest dollar amounts to any citizen‑directed allocation system. If voters in those states disproportionately favored certain categories (for example, Californians might prioritize environmental protection and scientific research), those priorities could dominate the distribution of funds nationally. Conversely, low‑income states would have less fiscal weight, potentially exacerbating regional inequities unless the system incorporated redistributive adjustments.
Discussion of Challenges and Implications
Mandatory obligations vs. preferences. The hypothetical allocation shows that citizens might allocate only $318 billion to Social Security, far short of the $1 trillion legally owed to beneficiaries. Similar shortfalls occur for Medicare and Medicaid. Unless laws are changed to reduce benefits or raise taxes, those programs cannot simply be underfunded; the citizen‑allocation system would need to treat mandatory spending as off‑limits or require that basic obligations be funded first.
Potential for under‑funding essential services. Programs that are unpopular or less visible—such as net interest payments or regulatory enforcement—could be severely under‑funded even though they are necessary for government functioning and creditworthiness. A minimum funding floor or matching mechanism might be required to ensure basic operations.
Equity concerns. Higher‑income taxpayers would have more dollars to allocate, potentially giving them disproportionate influence over spending priorities. Policymakers would need to consider whether each person’s allocations should carry equal weight (one person, one vote) or whether dollar‑weighted allocations are acceptable.
Administrative complexity. Integrating allocations into payroll systems and ensuring compliance would require significant investment. Employers, payroll processors and the IRS would need new software and secure transmission protocols. Education campaigns would be necessary so taxpayers understand their choices and the consequences.
Political feasibility. Transferring some budget‑setting authority from Congress to individual taxpayers would represent a profound change in the U.S. constitutional framework. It could be implemented as a non‑binding survey that informs Congress rather than a binding allocation. Alternatively, a pilot program could allow taxpayers to allocate a small share (e.g., 10%) of their income tax to see how preferences align with actual program costs.
Conclusion
The 2019 Pew Research survey reveals that U.S. citizens broadly favor increased spending on many federal programs—particularly education, veterans benefits, infrastructure and environmental protection—while showing less enthusiasm for increases in defense and foreign aid. Applying these preferences to the FY 2019 federal budget produces a radically different allocation of funds: entitlement programs and defense would shrink dramatically while education, science and environmental programs would expand. Yet the exercise also illustrates the difficulty of turning preferences into policy. Many expenditures (Social Security, Medicare, veterans pensions) are mandatory under current law, and cutting them to align with voter preferences would require contentious legislative changes. A secure, citizen‑allocation system could provide valuable insights into public priorities and might be used to guide discretionary spending, but implementing it would involve complex technical, legal and equity challenges.