The War Budget Sunset

The problem

The Authorization for Use of Military Force passed on September 14, 2001 runs sixty words. It has now outlived the enemy it named, the president who signed it, and most of the Congress that voted for it. It has been cited to justify American military operations in at least twenty-two countries — many of them unrelated to the attack it was written to answer, several of them involving organizations that did not exist in 2001. It has never been reauthorized, because it never expires. A future in which it is still operative in 2050 requires no legislation whatsoever. It requires nothing at all.

Brown University's Costs of War project puts the total price of the post-9/11 wars at roughly $8 trillion, counting Defense Department operations, State Department war costs, war-driven increases to the base budget, homeland security, care for veterans already incurred and still owed, and interest on the borrowing. That is not a partisan number; it is an accounting number, and its largest components are the ones that arrive after the shooting stops.

Almost none of it was paid for at the time. The post-9/11 wars were the first major American conflicts financed entirely by borrowing while taxes were simultaneously reduced. Every prior generation, whatever else it got wrong, understood that a war is a bill. The Revolution had its requisitions, the Civil War its income tax, the World Wars their bond drives and excess-profits levies, Korea its surtax. The generation that fought the longest war in American history is the first to have handed the entire invoice, with interest, to people who were not born when it started.

And when the wars wound down, the money did not become an asset. It became other spending. The 1990s ran the same experiment — defense fell by roughly half as a share of the economy after 1989 — and the peace dividend dissolved without leaving a monument, an endowment, or a trace.

The proposal

Three rules, applied to every future use of force: it expires, it is paid for, and what it stops costing becomes permanent.

1. Sunset. Every authorization for the use of military force expires automatically twenty-four months after enactment. Continuation requires a fresh, recorded, up-or-down vote in both chambers. No authorization may be inherited.

1. Concurrent payment. An authorization automatically triggers a dedicated war surtax sufficient to fund it, appearing as its own line on every tax return, ending when the authorization ends. Wars are funded in the regular appropriation, with a named authorization attached to every dollar — no permanent contingency accounts, no supplementals outside the process.

1. The peace dividend becomes principal. When an authorization lapses and outlays fall, the reduction is split by statute: half to debt reduction, half deposited into the American Permanent Fund (№ 001) as principal, where it cannot be re-appropriated. Peace, once, should buy something that lasts.

How it would work

The self-defense power is untouched. The President's authority to repel a sudden attack is constitutional, not statutory, and the War Powers Resolution's sixty-day window is unchanged — nothing here requires anyone to wait for Congress while missiles are inbound.

The sunset applies to authorizations, not deployments, and twenty-four months was chosen to force the vote into a different Congress than the one that started it. The 2001 and 2002 authorizations are repealed on enactment; anything currently conducted under them is re-authorized on the merits or it stops. This is not a novel demand — the Senate has voted by wide margins to repeal the Iraq authorizations, and repeal has repeatedly cleared one chamber before dying in the other. The sunset simply removes the need to ever win that fight again.

The surtax is the part that will be called unserious, so state its logic plainly: it is a price signal aimed at the decision-maker. A war that the country will not fund while fighting it is a war the country has not actually decided to fight. The friction is not a bug to be engineered away. It is the entire mechanism.

The Peace Ledger (№ 010) publishes the results each year — what the authorizations cost, what lapsed, what was deposited — on Dividend Day (№ 017), next to the check.

The numbers

Consider the counterfactual honestly. Suppose that between 2001 and 2021 a quarter of the $8 trillion had been unspent — a large claim, and we will not pretend to know which quarter — and deposited as principal under this rule's half-to-the-Fund split. That is $1 trillion of corpus. At a 5 percent real return it would today pay about $50 billion a year, forever: roughly $150 per citizen per year, arriving every October, from wars not fought.

The narrower and more defensible number is the interest. Costs of War attributes hundreds of billions of the $8 trillion to interest on war borrowing alone, and that meter is still running on debt already incurred. Concurrent payment would not have made the wars cheaper. It would have made them visible, and visible wars are shorter.

Veterans' care is the line that should end the argument about whether wars are expensive. The obligations to those who fought the post-9/11 wars peak decades after the last deployment — the peak of World War I veterans' spending arrived in the 1960s. A budget process that cannot see forty years out cannot see a war at all.

The honest objections

"Sunsets tell adversaries to wait you out." This is the serious objection and it is not fully answerable. An enemy who reads a two-year clock has a strategy available that an open-ended commitment denies them. Three mitigations, none complete: the clock renews rather than terminates; a Congress that wants to signal resolve can renew early and loudly, which is a stronger signal than silence; and the alternative on offer is not resolve but inattention — an authorization nobody has voted on in twenty-five years signals nothing to anyone, because it reflects no current decision. We would rather explain a renewal vote to an adversary than explain a forgotten war to a veteran.

"A war tax is a fantasy, and raising taxes during a conflict can be economically destructive." Sometimes, yes — wartime fiscal policy is genuinely more complicated than the moral framing suggests, and there are real cases for deficit-financing a short, existential emergency. So build the escape hatch honestly: a supermajority may suspend the surtax for a declared national emergency, on the record, with an expiration. What is not acceptable is the current default, where the deferral is silent, permanent, and unanimous.

"Congress will just rubber-stamp the renewals." Often it will, and we should concede that this proposal's realistic effect is modest rather than transformative. But recorded votes are not nothing: they end careers, they generate hearings, they force members to read the thing, and they convert a policy nobody owns into a policy with two hundred and eighteen signatures on it. Accountability is a weak force applied continuously.

"This is an anti-military proposal dressed up as budgeting." It is close to the opposite. The people who pay most for open-ended authorizations are the ones deployed under them, and the veterans' obligations that dominate the long tail of the $8 trillion are a promise we support paying in full. Nothing here cuts a benefit, a salary, or a readiness account. It changes who has to vote, and when the bill arrives.

"Splitting the peace dividend with a sovereign fund is a raid on defense." The split only ever applies to money that has already stopped being spent because a war ended. If outlays do not fall, nothing is transferred. The rule exists because the 1990s proved that an unclaimed peace dividend does not stay a dividend — it simply becomes something else, without anyone deciding.

Sources

• Brown University, Costs of War Project — post-9/11 war costs of approximately $8 trillion, including veterans' care and interest (costsofwar.watson.brown.edu)
• Authorization for Use of Military Force, Pub. L. 107-40 (2001); Congressional Research Service reporting on its citation across countries and operations (congress.gov)
• War Powers Resolution, 50 U.S.C. ch. 33
• Senate votes to repeal the 1991 and 2002 Iraq authorizations (118th Congress)
• Historical war finance: Civil War income tax, World War II bond drives and excess-profits tax, Korean War surtax
• Proposals № 001 (The American Dividend), № 010 (The Peace Ledger), № 017 (Dividend Day)