Key Points:

• In Fiscal Year 2025, federal outlays totaled over $7 trillion across 52 general spending categories. Within each of these categories, we trace spending at the line item and subcategory level to assign a total of $4.4 trillion in spending across three age groups: retirees; working-age adults; children and young adults. We classify the remaining $2.6 trillion as “all ages” because they finance broad public goods.
• Retirees (ages 65 and older) receive $2.7 trillion, or 62 percent of the $4.4 trillion in age-assignable federal outlays, driven mainly by Social Security and Medicare.
• Working-age adults (ages 26–64) receive $1.2 trillion, or 28 percent of age-assignable outlays, spread across Medicaid, Social Security disability benefits, veterans benefits, and Marketplace subsidies.
• Children and young adults (under age 26) receive $449 billion, or 10 percent of age-assignable outlays, concentrated in Medicaid, SNAP, child nutrition, and education programs.
• The heavy expenditure share on retirees is consistent with a voting model from the field of political economy. The retiree share is predicted to increase even more with an aging population and fiscal strain.

Background

Federal spending is often described by budget function, agency, or whether it is mandatory or discretionary. Those classifications are useful for budget enforcement and program management, but they are less informative for a basic distributional question: which age groups are the ultimate recipients of federal outlays? A federal budget can expand without affecting all age groups in the same way, and large programs that appear in the same budget function can serve very different populations.

This brief reorganizes FY2025 federal outlays into four mutually exclusive age buckets: under 26, ages 26 to 64, ages 65 and older, and an all-ages residual. The resulting estimates are not a literal ledger of every federal dollar by recipient age. They are an incidence-based crosswalk that combines budget totals with program-level evidence on who benefits from major spending programs.¹

Retirees Receive the Largest Share of Age-Assignable Federal Spending

We trace spending items within 52 general spending categories totally $7.0 trillion for the 2025 Fiscal Year. Sometimes this trace is at the line item level, and often at a subcategory level, depending on availability of data for a given spending category. We are able to classify a total of $4.4 trillion across three broad age categories and assign the remaining $2.6 trillion to an all-ages residual.

As shown in Table 1, retirees (adults age 65 and older) receive $2.7 trillion, equal to 38.6 percent of total federal outlays and 61.9 percent of age-assignable spending. Working-age adults (ages 26-64) receive $1.2 trillion (27.9 percent of age-assignable), and children and young adults (under age 26) receive $449 billion (10.3 percent). More information is provided in a technical Appendix and inside a downloadable Excel file that can found in the Notes for Table 1 below. Table 1 also shows average age-assignable spending per person (per capita).

The dominance of the retiree category reflects two programs above all others: Social Security and Medicare. Social Security directs $1.3 trillion to retirees, and Medicare sends $835 billion. Together they account for 80 percent of all age-assignable spending on older adults. But the retiree total extends beyond these two entitlements. Federal employee retirement benefits ($169 billion), housing assistance for older households, Medicaid long-term-care spending, and VA medical care all contribute, making the federal budget more retiree-focused than a Social Security–only lens would suggest.

Figure 1 shows how spending breaks down across seven aggregated program categories for each age group. Social Security and Medicare together account for the vast majority of retiree spending, while Health (primarily Medicaid and Marketplace subsidies) dominates for working-age adults. Spending on younger Americans is more evenly distributed across Health, Income Security, and Education.

Working-Age Adults Receive Large Transfers Through Health, Disability, and Veterans Programs

Unlike the retiree category, no single spending program dominates for the working-age population. Instead, a broad set of health, income-support, and veteran-related programs contribute.

Medicaid is the largest single contributor to the working-age bucket. The program supports low-income adults, parents, and people with disabilities, so it does not fit neatly into a single age group. In the current crosswalk, Medicaid directs $357 billion to working-age adults — more than to any other age bucket. Social Security also sends a substantial amount to working-age adults through disability and survivor pathways, contributing $216 billion. Beyond these two programs, Medicare for disabled beneficiaries under 65 ($159 billion), VA compensation and medical care ($202 billion combined), unemployment compensation ($40 billion), housing assistance ($44 billion), and Marketplace subsidies ($90 billion) all add materially.

This pattern shows that federal spending on working-age adults is not confined to labor-market programs. Much of it flows through health insurance, disability-related support, and service-connected veteran benefits. The working-age slice of the budget is more dependent on health and disability policy than broad public discussion often assumes.

Spending on Younger Americans Is Concentrated in Health, Nutrition, and Education

For people under age 26, Medicaid is the largest program, contributing $144 billion. SNAP follows at $44 billion, then child nutrition programs ($34 billion), Marketplace subsidies for younger enrollees ($29 billion), higher education support ($26 billion), elementary and secondary education ($24 billion), child care assistance ($21 billion), CHIP ($20 billion), TANF and family support ($18 billion), SSI for children ($15 billion), and foster care and adoption assistance ($11 billion). This is a different mix from the one that drives retiree spending: it is more fragmented and more dependent on means-tested programs.

The under-26 category combines children with young adults, which matters for interpretation. Elementary and secondary education, child nutrition, foster care, and CHIP are child-centered programs, while higher education support and a portion of Marketplace assistance reach young adults. The category therefore captures both childhood-oriented spending and spending that helps individuals transition into adulthood.

A Large Residual Cannot Be Credibly Assigned to Any Age Group

Not every federal dollar can be credibly assigned to an age group. We estimate that $2.6 trillion, or 37.5 percent of FY2025 outlays, belongs in an all-ages residual. This is not a flaw in the exercise; it reflects the actual structure of the federal budget.

The largest items in the residual are net interest ($970 billion) and national defense ($917 billion). Transportation ($146 billion), natural resources and environment ($88 billion), administration of justice ($85 billion), community and regional development ($85 billion), and several other functions also remain in the all-ages category. These programs finance public goods, broad administration, place-based investment, or mixed services for households of all ages — they are not programs where recipient age is the central organizing concept.

A crosswalk that forced those dollars into age buckets would create false precision and obscure the distinction between direct transfers to people and broader federal functions. The more useful takeaway is that most directly age-assignable spending goes to older adults, even though a large share of the federal budget operates outside a meaningful age-recipient framework.

An Age-Based View Cuts Across Standard Budget Categories

Reorganizing federal spending by age changes the way several major budget functions appear. Health spending is not just a retiree story: Medicaid and Marketplace subsidies direct large sums to younger and working-age groups, with 24 percent of age-assignable health spending going to people under 26 and 56 percent going to working-age adults. Income security is not just a working-age or anti-poverty category: it includes both retirement-related programs and child-focused supports, splitting 27 percent to younger Americans, 32 percent to working-age adults, and 41 percent to retirees. Veterans spending spans age groups as well, with 56 percent of its age-assignable spending directed to working-age adults and 41 percent to retirees.

This perspective does not replace standard budget classifications, but it adds a useful layer. A budget debate framed only around functions or mandatory versus discretionary spending can miss who ultimately benefits. An age-based view makes clear that older adults receive the largest assignable share of federal outlays, that working-age adults receive large support through health and disability-linked programs, and that spending on younger people is concentrated in a narrower set of means-tested and education-related programs.

Why Retirees Get More: The Median-Voter Model

A large part of the expenditures received by retirees is related to past earnings and payroll taxes. Almost 45 percent of Medicare remains general-revenue financed, however, including Part B and Part D. The pay-as-you-go nature of Social Security and Medicare Part A also means that many past and some current retirees received more in benefits in present value than what they paid into payroll taxes.

From a political-economy perspective, it is not obvious at first glance why transfers would go disproportionately to retirees. If transfers instead went to children and younger people — to boost their productivity and health — their future taxes during working ages could also pay for those transfers. The theory of pay-as-you-go financing from young to old has been explored in the political-economy literature, beginning with Aaron (1966).²

Consider a simple three-period model very similar to the data decomposition in this brief: young, middle age, and old. In this framework, the young include younger workers who pay payroll taxes alongside middle-age workers. Each age group votes according to self-interest on whether to start or maintain a pay-as-you-go program like Social Security or Medicare Part A.

The young vote against pay-as-you-go financing because capital markets are dynamically efficient — the average return on investment exceeds the growth rate of the economy ($r > g$).³ Pay-as-you-go financing leaves the average young person worse off because they can receive a higher risk-adjusted return in capital markets ($r$) relative to the internal rate of return from pay-as-you-go financing, which equals the growth rate of the underlying payroll tax base ($g$).

The old vote in favor because they are no longer paying into the system. They unambiguously benefit from continued transfers.

Middle-age workers hold the decisive vote. They have already paid into the system while young, making those past payroll taxes effectively sunk. Even though capital markets are dynamically efficient, conditional on being middle-aged, the present value of future pay-as-you-go benefits will typically exceed the present value of remaining payroll taxes. A clever middle-aged worker might consider voting to end pay-as-you-go financing to save their remaining payroll taxes while hoping the system re-emerges once they retire. A simple intergenerational trigger strategy — essentially a social norm — prevents this deviant behavior from being rewarded. Unless $r$ is much larger than $g$, middle-aged workers therefore support the continuation of pay-as-you-go programs. They are the median voter, and the system persists.⁴

Biology drives this equilibrium: we grow older, not younger. Supporting transfers in the opposite direction — toward younger people — is generally not competitive in the political economy, although a literature has explored bundling transfers to the young with transfers to the old through representative rather than direct voting. Representatives who deviate too much from the median voter’s preferences, however, are likely to be less competitive in general elections.

As a result, any significant reshuffling of national spending toward young people is unlikely because that outcome is not politically competitive. The share of transfers to retirees relative to the young is likely to increase even more over time due to an aging population — and hence a graying median voter — combined with the large imbalance in the U.S. federal fiscal outlook, which will place increasing strain on future spending.

Technical Appendix: Mapping FY2025 Federal Outlays to Age Groups

Purpose

This appendix documents the methods used to map FY2025 federal outlays into four age-based buckets. The main brief reports dollars and spending shares by age group; this appendix explains how those estimates were constructed, what source data were used, how mixed-age programs were treated, and where the remaining limitations lie.

Scope and Age Buckets

The controlling baseline is the Treasury FY2025 outlay total of $7,010.0 billion. The analytic file allocates each row across four mutually exclusive buckets.

Bucket	Definition	Interpretation
Younger	Under age 26	Children, teens, and young adults through age 25
Working age	Ages 26–64	Adults in the main working-age range
Retirees	Age 65 and older	Older adults, including most retirement-age beneficiaries
All ages	Residual category	Public goods, financing items, broad administration, and mixed-age programs without a credible age split

This is an age-incidence estimate, not a literal ledger of the age of every person touched by each federal dollar. The exercise preserves full reconciliation to the federal outlay baseline while using the strongest available program-level evidence on beneficiary age.

Data Sources and Source Hierarchy

The crosswalk uses a layered source strategy. Treasury provides the controlling total; OMB provides the program inventory; and agency sources provide the age-incidence evidence used to split mixed-age programs.

Source Family	Primary Use
Treasury Combined Statement, FY2025	Controlling total for federal outlays and function-level reconciliation
OMB Historical Tables	Program inventory; function and subfunction detail
SSA	Social Security and SSI beneficiary and payment-age distributions
CMS, MACPAC, and MedPAC	Medicaid, CHIP, Medicare, and Marketplace age/spending distributions
USDA/FNS	SNAP, WIC, and child nutrition age or participant distributions
NCES	Age distribution of postsecondary students for higher-education allocation
HUD	Age of householder in assisted housing
VA	Age/payment or beneficiary-type distributions for veterans benefits and medical care
OPM	Program-design evidence for federal employee retirement and disability
ACF/ACL	TANF, child care, foster/adoption, and aging-services program structure

When FY2025 program-level age data were not available, the nearest adjacent official source year was used as a proxy, with the row scaled to the FY2025 outlay amount.

Construction of the Analytic File

Multiple large language models are used to provide objective natural language interpretation of program spending by age. This approach not only supports scanning through numerous subcategories and line items; it also reduces bias from human interpretation and noise from human reading fatigue.

1. Set the baseline. Treasury FY2025 outlays define the total that the analytic file must reconcile to.
2. Build the program inventory. OMB historical tables divide broad functions into program-like rows, especially within education, health, income security, and veterans benefits.
3. Assign direct-age programs first. Programs with clear statutory or operational age targeting are assigned directly (e.g., K–12 education, child nutrition, foster care, aging services, and many retirement programs).
4. Use official age shares for mixed programs where possible. When agency tables report beneficiary-age or spending-age distributions, those shares are applied directly.
5. Interpolate only the crossing band. When source age bands do not match the requested age buckets, only the overlapping band is split.
6. Apply the intended-beneficiary rule. Spending is assigned to the person served, not necessarily the institution or household that receives payment.
7. Retain weakly identified rows in all ages. Public goods, financing costs, broad administration, and mixed-age programs without a credible age split remain in the residual bucket.
8. Record confidence. Each row is tagged with a confidence tier (A, B, or C) reflecting the strength of the age-incidence evidence.

Interpolation Conventions

Interpolation is used sparingly, only when an official age band crosses one of the requested cutoffs.

Source Band	Rule
21–64 → <26 vs 26–64	5/44 to younger, 39/44 to working age
18–64 → <26 vs 26–64	8/47 to younger, 39/47 to working age
18–34 → <26 vs 26–64	8/17 to younger, 9/17 to working age
Adult <45 veteran band	8/27 to younger, 19/27 to working age

Confidence Coding

Code	Meaning	Examples
A	Direct age-target program design or official age table closely aligned to the requested buckets	K–12 education, SSI, child nutrition, aging services
B	Official mixed-age source with interpolation or a strong proxy	Medicare, SNAP, housing assistance, VA medical care
C	Conservative residual or inferential split without a direct payment-by-age table	Federal employee retirement, Medicaid disability residual, broad mixed functions left in all ages

Selected High-Impact Assumptions

Social Security. Estimated from SSA benefit-type and beneficiary-age tables rather than treated as a single retiree-only program. This produces a retiree-dominant allocation while preserving non-zero younger and working-age components tied to disability, survivor, and child benefits.

Medicare. Allocated using the official spending split between beneficiaries under age 65 and beneficiaries age 65 and older. The under-65 component is then divided between younger and working-age adults using enrollment-by-age evidence.

Medicaid. Uses spending-age evidence rather than enrollment-age evidence. The main allocation is based on CMS spending shares for children, non-disabled adults, and older adults. The disability component is split inferentially across younger, working age, and retirees and is treated as lower confidence.

Federal employee retirement and disability. Treated as overwhelmingly retiree-oriented (90 percent), with modest working-age disability and survivor components (9 percent) and a very small younger survivor component (1 percent), based on OPM program design and annuitant evidence.

Housing assistance. Allocated using the age of the household head in HUD-assisted housing. This is conservative with respect to younger beneficiaries because many assisted households include children even when the householder is an older adult.

Quality Checks

• The analytic file contains 52 rows.
• Row amounts reconcile to the Treasury FY2025 baseline of $7,010.0 billion.
• Bucket totals also reconcile to $7,010.0 billion.
• Each row sums to 100 percent across the four buckets.
• 29 rows are retained fully in all ages, reflecting a deliberate decision not to force weak age assignments.

Limitations

1. The crosswalk is an estimate of age incidence, not a direct administrative tabulation of spending by person-level age.
2. Some rows use adjacent-year official evidence scaled to FY2025 spending levels because FY2025 age tables were not available.
3. Housing, Medicaid disability, VA medical care, and federal employee retirement/disability rely on strong but imperfect proxies rather than direct payment-by-age tables.
4. The all-ages bucket remains large by construction because public goods, financing items, and several mixed-age functions are large in the federal budget.
5. Some tax-administered outlays remain embedded in the outlay baseline and could not be cleanly age-mapped.
6. Negative or offsetting budget lines are preserved as they appear in the fiscal accounts so that the crosswalk continues to reconcile to federal budget totals.

This analysis was produced by Kent Smetters.

Footnotes

1. The Technical Appendix at the end of this brief documents the methods, data sources, and confidence coding used to construct the age-based crosswalk. ↩

2. Henry Aaron, “The Social Insurance Paradox,” Canadian Journal of Economics and Political Science 32, no. 3 (1966): 371–374. For a good overview of the subsequent literature, see Panu Poutvaara, “On the Political Economy of Social Security and Public Education,” Journal of Population Economics 19, no. 2 (2006): 345–365. For a more rigorous treatment with Condorcet voting, see Sita Nataraj Slavov, “Age Bias in Fiscal Policy: Why Does the Political Process Favor the Elderly?,” The BE Journal of Theoretical Economics 6, no. 1 (2006). ↩

3. For a discussion of the role of dynamic efficiency in the context of Social Security, see Penn Wharton Budget Model, “Six Options to Restore Social Security’s Financial Balance,” March 2026. ↩

4. This is a standard intergenerational “trigger strategy” — essentially a social norm that punishes deviant voting behavior across generations. ↩