Blog

PWBM projects that Social Security’s combined trust fund reserves will be depleted in February 2035, when continuing program income would support approximately 85 percent of scheduled benefits. Yet the Congressional Budget Office (CBO) must construct its budget baseline assuming that scheduled benefits continue in full. A policy supplying general revenue to cover the shortfall can therefore score at $0 in additional unified-budget costs relative to that baseline, even though it would finance payments that Social Security could not otherwise make in full and on time.

Scheduled Benefits and Payable Benefits

Scheduled benefits are the amounts determined by Social Security’s eligibility rules and benefit formulas. Payable benefits are the amounts the program can finance from its legally available resources. PWBM’s Social Security Data Center projects that, on a combined Old-Age, Survivors, and Disability Insurance (OASDI) basis, the payable share falls to 85 percent at depletion and approximately 60 percent by 2100.¹ Maintaining full scheduled benefits requires additional financing or changes to the program.

Figure 1 shows the widening gap between scheduled program costs and non-interest income in PWBM’s baseline. Both are measured relative to taxable payroll, the earnings base subject to Social Security payroll taxes.

Why CBO Assumes Full Payments

The legal reason is Section 257(b)(1) of the Balanced Budget and Emergency Deficit Control Act, codified at 2 U.S.C. §907(b)(1). It requires baseline calculations to assume that funding for entitlement authority is sufficient to make the payments required by the governing benefit laws. CBO identifies this provision as the reason its Social Security baseline includes all scheduled payments even after trust fund resources become insufficient.

In 1986, CBO and the Office of Management and Budget disagreed about how to project entitlement programs whose benefits were established by law but whose funding depended on annual appropriations. Congress resolved that dispute in 1987 by requiring the projections to assume full funding.² Court decisions, including the Supreme Court’s decision in Maine Community Health Options v. United States, distinguish statutory payment obligations from appropriations, while recognizing that specific funding restrictions can limit the obligations themselves. That decision concerned health insurance payments rather than Social Security’s baseline. CBO’s obligation to assume full funding in its baseline rests on the statutory requirement.

This separates two legal questions: what benefits existing formulas specify, and what resources administrators may use to pay them. The baseline rule instructs scorekeepers to assume adequate funding for the first calculation. It does not appropriate the missing money or authorize Social Security to borrow. Applying the financing constraint first, then assuming adequate funding only for the reduced payments, would remove the very shortfall the full-funding instruction requires the baseline to accommodate. The legal requirement to assume funding in a projection does not supply the authority to spend it.

In effect, the baseline already assumes access to financing beyond Social Security’s dedicated resources. CBO’s 10-year 2026 budget outlook explicitly describes implicit transfers from the Treasury’s general fund to cover the OASI shortfall that occurs in the latter half of the 10-year budget window. This is a scorekeeping assumption, not a prediction of which legislation Congress will enact.

The Borrowing Implied by Maintaining Scheduled Benefits

PWBM’s Data Center illustrates the financing requirement if post-depletion shortfalls are borrowed at the trust fund interest rate and interest is also financed. By 2100, the accumulated amount owed reaches 202 percent of that year’s GDP (201.77 percent before rounding), equivalent to 485.37 percent of that year’s taxable payroll. The 202 percent figure can be interpreted as either accumulated borrowing, including interest, or the total accumulated cost to the government of covering the post-depletion shortfall, including interest, under the same financing assumptions. Taxable payroll denotes earnings subject to tax, rather than the tax receipts collected on those earnings.

Figure 2 separates the accumulated non-interest deficits from accumulated interest. In 2100, these components equal 78.17 percent and 123.60 percent of GDP, respectively. The result is a stock of accumulated financing needs measured against one year’s GDP, not an annual benefit cost or a present-value measure.

Why a General-Revenue Transfer Can Score at $0

Consider legislation that supplies only the general-revenue transfers needed to pay scheduled benefits, leaving benefit formulas, eligibility, and taxes unchanged. Relative to CBO’s scheduled-benefit baseline, those benefit payments and their financing are already assumed. The transfer moves resources between federal accounts; it does not itself create another payment to the public. Under those assumptions, the legislation would have $0 in additional unified-budget costs relative to the baseline.

Relative to a payable-benefit scenario, the same legislation would enable additional benefit payments and require financing. A zero score therefore does not mean that maintaining scheduled benefits requires no resources. It means those resources are already included in the benchmark used to measure the legislation. A specific bill’s score would still depend on its full provisions, including any changes to taxes, benefits, administration, or financing terms, and on whether the presentation measures the unified budget or separate on-budget and off-budget accounts.

Footnotes

1. The combined-fund calculation is an analytical convention. The retirement and survivors (OASI) and disability (DI) funds are legally separate; PWBM projects OASI depletion earlier, in February 2033. Combining their resources would require legislation. Moreover, current law does not specify a complete procedure for distributing insufficient funds. As the Congressional Research Service explains, depletion prevents full, timely payments but leaves unresolved whether payments would be reduced or delayed. The 85 percent figure measures financing capacity; it is not an enacted instruction to reduce each check by exactly 15 percent. ↩

2. H.R. Rep. No. 100-313, at 50–53 (1987) (Conf. Rep.); Balanced Budget and Emergency Deficit Control Reaffirmation Act of 1987, Pub. L. No. 100-119, §102(a), (b)(6), 101 Stat. 754, 760, 773–74. ↩

President Trump has recommended suspending the federal excise tax on motor fuels. The estimates below assess the revenue and price effects of a suspension running from June 1 through October 1, 2026 — a 122-day window that covers the peak summer driving season.

Background

The federal excise tax on motor fuels, which finances the Highway Trust Fund, has been unchanged since 1993:

• Gasoline: 18.4 cents per gallon (18.3 cents to the Highway Trust Fund + 0.1 cents to the Leaking Underground Storage Tank trust fund)
• Highway diesel: 24.4 cents per gallon (24.3 cents + 0.1 cents LUST)

Because the tax is levied per gallon rather than as a percentage of price, its real value has eroded with inflation — but so has the revenue cost of suspending it. The estimates below use EIA supply projections for Summer 2026. Project baseline consumption patterns include a short-term demand elasticity that recognize that higher gasoline prices are likely to persist throughout the projection period relative to last year.

Gasoline Revenue Loss

Table 1 presents the projected monthly gasoline revenue loss under a base-case pump price of $4.00 per gallon. Adjusted gallons exclude the roughly 2.5 percent of gasoline consumed off-highway (e.g., agricultural and marine use) that is taxed at only 0.1 cents per gallon.

The gasoline revenue loss is sensitive to pump prices because price affects baseline consumption volume that would have existed prior to the tax suspension. At $3.50 per gallon, slightly higher demand raises the revenue loss to $8.55 billion; at $4.40 per gallon, reduced demand lowers it to $8.35 billion. In all scenarios the gasoline loss alone exceeds $8 billion.

Diesel Revenue Loss

U.S. distillate demand has averaged roughly 3.5 million barrels per day in recent months. Approximately 70 percent of distillate is on-highway (clear) diesel subject to the full 24.4-cent tax; the remainder is dyed or off-road diesel taxed at only 0.1 cents. Over the 122-day suspension window, taxable diesel consumption would total roughly 12.55 billion gallons, yielding a revenue loss of approximately $3.06 billion.

Total Revenue Impact

Combining the gasoline and diesel estimates, a four-month federal fuel tax holiday would reduce Highway Trust Fund receipts by roughly $11.5 billion. For context, total Highway Trust Fund outlays were approximately $62 billion in fiscal year 2025, meaning the suspension would forgo revenue equivalent to about 19 percent of one year’s spending.

Effect on Pump Prices

Not all of the tax cut would reach consumers. In the short run, both sides of the market are relatively inelastic: drivers need fuel regardless of small price changes, and refineries and distribution channels cannot quickly adjust output. How the savings split between the two depends on their relative elasticities. Consumers capture a share equal to $\varepsilon_s / (\varepsilon_s + |\varepsilon_d|)$, where $\varepsilon_s$ is the supply elasticity and $\varepsilon_d$ the (negative) demand elasticity, so they would receive the full cut only if supply were perfectly elastic. Because short-run gasoline supply is not perfectly elastic, suppliers can absorb part of the tax reduction as higher margins rather than passing the full savings through to pump prices.

Based on PWBM’s synthetic control estimates (see methodology in our state gas tax holiday analysis), the expected pass-through rates are 0.72 for gasoline and 0.60 for diesel. Applying these rates:

• Gasoline prices would fall by about 13.2 cents per gallon (0.72 × 18.4¢), rather than the full 18.4-cent tax.
• Diesel prices would fall by about 14.6 cents per gallon (0.60 × 24.4¢), rather than the full 24.4-cent tax.

In other words, consumers would see roughly 60 to 72 percent of the tax savings at the pump, with the remainder captured by suppliers. For a household filling a 15-gallon tank once per week over the 122-day period, the gasoline savings would amount to approximately $35.

The U.S. federal government has spent an estimated $27–28 billion on Operation Epic Fury in the first 32 days since strikes on Iran began February 28, 2026. Assuming operations continue through end of April, the total two-month direct cost lands in a range of $38 to $47 billion, with a base-case estimate of approximately $42.5 billion. These costs do not include around $5 billion in indirect costs or potential interest costs if deficit financed.

Budget Estimates (2 Months)

The estimates in the accompanying data file are built from four cost phases, each anchored to official government data or authoritative independent analysis and summarized in Table 1.

Phase 1 (Days 1–6, Feb 28 – Mar 5) averaged roughly $2.1 billion per day — dominated by the use of expensive standoff munitions. The Pentagon briefed the Senate on March 10–11 that just the first six days cost at least $11.3 billion in unbudgeted outlays, primarily munitions replacement. The Center for Strategic and International Studies (CSIS) revised that figure upward to approximately $12.7 billion once attrition, equipment losses, and base damage were included. The high burn rate reflects the cost structure of the opening salvo: Tomahawk cruise missiles (~$3.5 million each), SM-3 and SM-6 interceptors (up to $4–5 million each), and AGM-154 glide bombs (~$600,000 each) were expended by the hundreds in the first days.

Phase 2 (Days 7–24, Mar 6 – Mar 23) saw costs fall sharply to approximately $601 million per day, as the military shifted to cheaper precision-guided munitions — primarily JDAM kits costing under $100,000 each — and Iranian ballistic missile launches dropped by roughly 90% from their peak, dramatically cutting U.S. air defense interceptor expenditures. CSIS estimated cumulative costs at $16.5 billion by Day 12, and the Center for American Progress (CAP), applying the same $601 million/day rate forward, projected costs crossing $25 billion around Day 26. Cumulative costs through the end of Phase 2 (March 23) stand at approximately $23.5 billion.

Phase 3 (Days 25–32, Mar 24 – Mar 31) applies a further step-down to approximately $500 million per day, consistent with CSIS’s forward projection as operations settled into a sustained but lower-intensity tempo and ceasefire negotiations began — though no deal was reached. This adds roughly $4 billion, bringing the total through March 31 to approximately $27.5 billion. The Low and High columns in the data file are blank through the end of Phase 3, as these days are treated as known point estimates rather than scenarios.

Phase 4 (Days 33–62, Apr 1 – Apr 30) is fully projected. With no ceasefire in effect as of April 1, the base case continues at $500 million per day for all 30 days of April, adding $15 billion. A low scenario of $350 million/day applies if operations wind down significantly or a ceasefire takes hold early in the month; a high scenario of $650 million/day reflects possible re-escalation.

End-of-April cumulative totals by scenario are shown in Table 2:

Not Included

The figures above count only direct, unbudgeted federal military spending — and even on that narrow definition, they likely understate the true cost. The Pentagon’s $11.3 billion figure explicitly excluded pre-strike repositioning costs (estimated at ~$630 million by AEI’s Elaine McCusker), routine O&M costs for ships, aircraft, and personnel in theater (~$196 million in the first six days alone, per CSIS), and allied resupply transfers. Extending those omitted routine O&M costs across all 62 days adds a further $2–3 billion not captured in any public estimate. Beyond accounting gaps, long-term veterans’ care and disability benefits for the almost 300 U.S. service members already injured thus far will accumulate over several decades.

References

• CENTCOM Press Release — “U.S. Forces Launch Operation Epic Fury,” U.S. Central Command, February 28, 2026.
• DoD/Pentagon Fact Sheet — “Operation Epic Fury Fact Sheet: The First 29 Days,” U.S. Department of Defense, March 28–29, 2026.
• Mark F. Cancian, “Iran War Cost Estimate Update: $11.3 Billion at Day 6, $16.5 Billion at Day 12,” Center for Strategic and International Studies, March 12, 2026.
• Mark F. Cancian, “$3.7 Billion: Estimated Cost of Epic Fury’s First 100 Hours,” Center for Strategic and International Studies, March 5 (published March 16), 2026.
• Center for American Progress, “By the End of the Week, the Trump Administration’s War in Iran Will Likely Have Cost $25 Billion,” March 23, 2026.
• Elaine McCusker (AEI), “The Dollars and Cents of Military Action Against Iran,” Forbes, February 26, 2026.

Several states are considering suspending state-level gasoline and diesel taxes (excluding the federal tax). Table 1 presents the projected impact on the per-gallon price of gasoline and diesel, as well as the estimated revenue loss to each state from suspending these taxes for two months. Revenue loss estimates are adjusted for leakage due to existing partial exemptions and refunds (e.g., diesel used in farming and food processing, and gasoline used in off-highway agricultural activities).

The One Big Beautiful Bill Act (OBBBA) instituted a cap on the marginal tax benefit of itemized deductions of 35 percent beginning in 2026. We present conventional 10-year revenue forecasts and distributional tables for two options: Lower the cap on the marginal tax benefit of itemized deductions from 35 percent to 32 percent (Option 1) or to 24 percent (Option 2), beginning in 2026.

Table 1 shows that the 32 percent cap would raise $110 billion over ten years. The 24 percent cap would raise $475 billion over ten years.

Table 2 presents conventional distributional analysis. Option 1 mostly impacts those in the top 10 percent of the income distribution while Option 2 impacts those in the top 20 percent.