US military spending reaches 6% of GDP

Where things stand

• SIPRI's latest figure. SIPRI's April 2026 release puts US military spending at $954 billion, or 3.1% of GDP, in 2025. The database gives the precise value as 3.117%; 2024 was 3.43%. Source: the fact sheet dated 27 Apr 2026, and the database, version 1.2.
• The threshold. I took GDP implied by SIPRI for 2025 ($30.6 trillion) and grew it at about 4.5% a year. On that basis, 6% means roughly $1.9 trillion for 2026, $2.0 trillion for 2027 and $2.3 trillion for 2030. That is 2.0 to 2.4 times SIPRI's 2025 total.
• SIPRI counts money actually spent, not money approved. SIPRI says its US figure covers outlays (money actually paid out) for national defense plus foreign military aid (Yearbook 2015 chapter). Approved funding takes years to turn into outlays, which slows how fast the figure can rise.
• 2026 is already settled. US Treasury data show national defense outlays of $874 billion in fiscal 2024, $917 billion in 2025 and $965 billion in 2026 (Treasury API, Monthly Treasury Statement Table 9). So the 2026 figure SIPRI will publish in April 2027 should be about $1.0 trillion, roughly 3.1–3.2% of GDP.
• Even the full budget request falls well short. The fiscal 2027 request is $1.54 trillion in approved funding, including $350 billion through a fast-track budget bill that only needs a simple majority ("reconciliation"). If fully enacted, the White House budget office projects defense outlays at 3.6% of GDP in 2027 and 3.7% in 2028. Funding then falls 16% in real terms in 2028 (CSIS, 10 Apr 2026).
• Congress is unlikely to give even that. The Pentagon is running on a temporary funding bill until 11 December. The House's new reconciliation bill contains only $60 billion for defense, and the $67 billion war supplemental has stalled (Breaking Defense, Aug 2026).
• The Iran war is too small to matter. It has cost about $38–45 billion so far, and the Congressional Budget Office expects $2–3 billion more per month (CNN/Roll Call, Sep 2026). That is around 0.1–0.2% of GDP.

Reference class

I computed jumps in SIPRI's US series from 1949 to 2025. Only the Korean War produced a rise of 2.5 points or more within five years: from 5.1% in 1950 to 10.2% in 1951 and 13.9% in 1952. Other big buildups added less than 2 points:

• Vietnam: +1.8 points over two years.
• Reagan buildup: +1.9 points over four years.
• After 9/11: +1.8 points over eight years.

SIPRI's figure was last at or above 6% in 1988. Reaching 6% would take about 2.2–2.9 points above the current path, which only a Korea-scale or larger mobilization has ever delivered. That is one onset in 77 years, or about 1.3% a year.

Ways it could resolve YES, and timing

The main path is a sustained great-power war, most plausibly with China.

• Chance of such a war. Polymarket prices a Chinese invasion of Taiwan by the end of 2027 at about 10%; long-dated markets tend to overstate this, and there is a current thaw after the September Trump–Xi summit. I put a sustained US–China (or similar) war starting by early 2030 at about 4–5%.
• Chance it pushes SIPRI's figure to 6% in time. Given such a war, I put this at about 55–65%. Outlays lag, a short war may not be enough, and the war has to start early enough to affect a fiscal year that SIPRI publishes by April 2031.
• Smaller paths, about 0.5% combined:
  • a huge rearmament after China takes Taiwan without a US fight;
  • war with North Korea or Russia;
  • a large ground war against Iran;
  • an oddity in how SIPRI measures spending or GDP.

Timing. SIPRI publishes once a year, in late April. The chance of a YES therefore sits almost entirely in each second quarter:

SIPRI release	Fiscal year covered	Chance this release triggers YES
April 2027	2026	~0
April 2028	2027	0.25% (needs an almost immediate war)
April 2029	2028	0.8%
April 2030	2029	1.0%
April 2031	2030	1.2%

There is no release in the window after April 2031. The cumulative probability is about 3.3%.

Strongest case against my estimate

For higher: the US is already at war with Iran, the administration is seeking record budgets, and the 2027 risk window for Taiwan is real. But even a Vietnam-scale war added only about 2 points, and $100 billion is only about 0.3% of GDP. For lower: there is a thaw with China and Trump seems reluctant to defend Taiwan. Budget pressure is also high: debt is above 100% of GDP and the deficit is about $2 trillion. Spending on new weapons pays out slowly, so even a war might not lift spending fast enough. On balance I land at about 3%.

Question. By the last day of each quarter, has SIPRI reported US military expenditure at 6% of GDP or more for any year?

Definitions.

• SIPRI: the Stockholm International Peace Research Institute.
• US military expenditure as a share of GDP: the United States series for military expenditure as a percentage of GDP in the SIPRI Military Expenditure Database, using SIPRI’s military-expenditure definition and GDP denominator, exactly as SIPRI publishes it. SIPRI states that all of its figures for the United States are for the US financial year (1 October of the previous year to 30 September of the stated year) rather than the calendar year; the series counts on that basis, and no conversion to calendar years is required or applied.
• Eligible year: a year in SIPRI’s series, as SIPRI labels it, whose underlying period has ended and ended after the opening date, under the Observation period convention. SIPRI’s 2026 figure for the United States covers 1 October 2025 to 30 September 2026, so the earliest eligible year is 2026.
• 6% or more: a published value of at least 6.0 percent. Use the most precise value SIPRI makes available in the same release, including the downloadable database, without additional rounding. If SIPRI publishes only a rounded value, that value controls.

Resolves YES if SIPRI publishes an expenditure-to-GDP value meeting the threshold for at least one eligible year by the quarter-end. Estimates or provisional historical expenditure figures count; projections for an unfinished or future year do not. A revision to an eligible year can trigger YES. Once a qualifying publication has triggered YES, a subsequent downward revision does not reverse it.

Resolves NO if no qualifying publication exists by the quarter-end. Exclude budget requests, appropriations, targets, forecasts, ratios from any source other than SIPRI (including US government fiscal-year ratios), shares of government expenditure, and ratios independently calculated from spending and GDP figures. Publications repeating or revising figures for 2025 or earlier do not count. NATO, US government, World Bank, and other third-party figures cannot independently resolve the question, even if described as based on SIPRI data. No connection to AI is required.

Resolving source. SIPRI only: the SIPRI Military Expenditure Database, specifically the United States entry in its “Share of GDP” series (military expenditure as a percentage of GDP), or a SIPRI-authored publication explicitly reporting that series, such as the annual Trends in World Military Expenditure fact sheet (the 2025 edition reports it in Table 1, “Spending as a share of GDP (%)”). The database is updated annually and can revise earlier years. Apply the Public confirmation convention (A2) with this narrower source restriction, the Resolution date convention (A3), the Resolution window convention, and the Observation period convention.