The power of the purse

SIMULATION. This session ran with simulation: true (D54) under the 30k Protocol v0.9 draft on 2026-10-27, with the full cast of Simulation 2, all role-played by Claude. The funding facts in section 1 are real. Everything else is invented: members, blocs, bills, votes, the executive's choices, and every voice. Events marked simulated did not happen. This is not a prediction, a poll, or a recommendation. Every voice here comes from one AI model (OQ-22).

The question. Simulation 2 found that a stand-alone bill limiting the war stops at the Senate's 60-vote threshold and at a veto. This run asks whether attaching the same conditions to war funding that the executive has asked for changes who has to agree, and what survives.

1. Real-world starting point (as of 2026-10-27)

• The request. On 24 June the administration asked Congress for about $87.6 billion in supplemental funding. That included about $67 billion for the Pentagon: $21 billion for munitions, $17.3 billion for operations, and $12.1 billion for classified programs. The rest covered farm aid, an Ebola response, and smaller Energy and State Department items. The top Senate Appropriations Democrat objected that unrelated Pentagon priorities were mixed in. The Armed Services chairman called the funding essential. (CBS News)
• The party-line route. On 22 July the House passed a reconciliation package with about $73 billion for the war, plus funds tied to elections legislation. The Senate majority leader said he did not yet have the votes for it. He has also said reconciliation must be mainly budgetary, since policy provisions with only incidental budget effects usually violate the process. (The National Desk)
• Costs. The Pentagon puts costs through early September at $43.6 billion, including $32.4 billion in unfunded replacement costs. The Congressional Budget Office projects $2–3 billion a month going forward. (Roll Call)
• Regular funding. The government runs on a continuing resolution through 11 December 2026. Among other things, it blocks Defense Department spending increases above fiscal 2026 levels. (NACo)

Why the purse is different. A stand-alone limit on the war asks the president to sign away freedom of action and gives nothing back. A funding bill gives the executive money it has asked for, so a veto has a cost to the executive too. Each route still has its own barrier:

• Regular appropriations need 60 Senate votes to end debate.
• Reconciliation needs only a simple majority in the Senate, but conditions that are mainly policy rather than budget are likely to be struck out.

2. Setup

The same 11,000 synthetic members, seven blocs, turnout model, and rules as Simulation 2. The second chamber and the presentment model run in research mode (D65).

A caveat from Spatial. These blocs are defined by stance on the war, not by party. The real House passed war funding in July on a party-line reconciliation vote. A chamber modeled on party discipline would behave differently. This run tests the mechanics of the purse, not a forecast of the real House.

3. Days 1–6: Drafting three funding bills

The coalition from Simulation 2 reconvened. The Restraint and Decisive-outcome blocs drafted their own bills, and Score published estimates for each (D53).

F2: Conditioned supplemental (simulated text, abridged)

Sec. 101. Munitions. $21.0 billion for munitions replenishment, of which $9.0 billion is available on enactment for air and missile defense interceptors and the remainder is available 30 days after the Secretary of Defense submits the first report required by section 105.

Sec. 102. Operations. $17.3 billion for operations, available for 120 days after enactment [[sunset:120]].

Sec. 103. Classified programs. $12.1 billion, available after the Director of National Intelligence certifies that the intelligence committees have been briefed in full on the programs funded.

Sec. 104. Limitations. None of the funds made available by this or any other Act may be used for (1) ground combat forces in Iran, (2) occupation of Iranian territory, or (3) military action to change Iran's government. After 120 days, none of the funds may be used for offensive military operations against the territory of Iran unless specifically authorized by law.

Sec. 105. Reports. Every 30 days: operations, costs to date, civilian harm, humanitarian access, shipping through the Strait of Hormuz, and the status of negotiations. Within 48 hours: any strike justified as defense against an imminent attack.

Sec. 106. The sense of Congress on reciprocal, verified steps from Simulation 2, unchanged.

Contributions from the new stakeholders:

• Command insisted the interceptor money in sec. 101 be available immediately: "Fencing missile defense would put the people we're protecting at risk."
• Clearance wrote the classified-briefing condition in sec. 103. It requires a certification and puts no classified text in the law, as D56 requires.
• Score moved about $16.6 billion of items it judged unrelated to the war out of the request. This matches the objection appropriators raised in real life.
• Counsel and Lint reviewed whether the reconciliation route could carry F2. Sections 104–106 are mainly policy, so they would likely be struck. Reconciliation could deliver the money but not most of the conditions. The coalition chose regular order.
• Option considered and dropped. A fourth option was attaching the conditions to the 11 December continuing resolution. Panels reported that constituents in every bloc opposed risking a shutdown over the war. Ombud warned that putting unrelated government services at risk is leverage paid for by people with no stake in the dispute.

4. Days 7–12: The contest and the House vote

The three bills went into a contest (D47) with the status quo. F2 won the ranked pairs: it beat F1 and F3 head to head. It then had to pass its own rule against the status quo.

F2 merged. It won 66.2% of the whole electorate and 73.3% of yes-and-no votes.

For comparison, each bill against the status quo:

• F1 alone: 4,977 approvals.
• F3 alone: 4,162 approvals.
• MR-4 v3, the stand-alone limit in Simulation 2: 6,054 approvals.

Robustness: across every ±10% bloc change, F2's approvals ranged from 7,226 to 7,348.

Why the coalition grew by 1,200 members:

• Decisive outcome went from 95 approvals to 632. Members who would never vote to limit the war voted to fund it, even with limits, because munitions and interceptors were at stake.
• Pressure-for-a-deal went from 1,240 to 1,796 for the same reason.
• Restraint lost 82. Some members would not vote to fund a war they want ended, whatever the conditions.

Agenda's note. Money adds a third dimension, supporting the troops. It cuts across the force-versus-restraint dimension and pulls members from both ends toward the center.

The result cleared two-thirds of yes-and-no votes in the House. In the House, at least, the bill could survive a veto.

5. The second chamber simulated

Upper voted 62 to 38 to end debate and then pass F2, clearing the 60-vote threshold that stopped Simulation 2's bill at 55. Money moved enough senators who had opposed a stand-alone limit.

At 62, the Senate is still short of the 67 needed to override a veto.

6. Presentment: two branches simulated

Whether the president signs depends on a judgment the simulation cannot model from evidence: how much the executive values the money against the limits on its freedom of action. The run modeled both choices.

Branch A: The executive negotiates

Exec modeled the executive branch accepting most of F2 but refusing the fixed 120-day cutoff on offensive operations. A conference produced this compromise:

• Kept: the tranches, the $9 billion in immediate interceptor funding, the classified-briefing condition, the bans on ground forces, occupation, and regime change, every report, and moving the non-war items out.
• Changed: after 120 days, offensive operations against Iranian territory require either a new authorization or a presidential certification of necessity, delivered with 15 days' notice. A joint resolution of disapproval would then get expedited consideration.
• Cost of the change. Advocate pointed out that a disapproval resolution can itself be vetoed. The certification path therefore weakens the hard deadline into a reporting-and-notice requirement.

The House passed the conference text with 6,986 approvals. The Restraint bloc lost 300 more members over the weakened deadline. Upper passed it 63 to 37. The bill is signed.

What changed in the war, in the model:

• Ground operations and regime change are barred by law.
• Missile defense is funded immediately.
• Every offensive operation after day 120 needs a public certification and gives Congress a fast vote, though not a binding one.
• Monthly reports put costs, civilian harm, and the status of talks on the record.

Branch B: The executive vetoes

Exec modeled the executive branch treating any funding limit on operations as unconstitutional and vetoing the bill.

• The override fails. The House clears two-thirds of yes-and-no votes, but the Senate's 62 is 5 short of 67.
• Funding stalls. The Pentagon's unfunded replacement costs keep growing, and the dispute moves to the 11 December deadline, where both sides now hold leverage.
• The pivot. Whip's party introduces a clean supplemental, and Third's coalition offers F2 again. Tape published a forecast that some version of F2 passes before 11 December, at 11/20. Odds listed a contract on it, labeled as simulated.

Sensitivity of the branch choice. If the executive values the funds more than the limits, Branch A follows. If it values operational freedom more, Branch B follows. The simulation cannot say which describes the real executive, and does not try.

7. Did the purse get closer to ending the war?

It changed who has to agree, not whether anyone has to.

• Breadth. The stand-alone route needed a supermajority to override a veto. The purse route turned that standoff into a negotiation, because the executive also wanted something. The House coalition grew from 6,054 to 7,286, and the Senate cleared 60.
• What survives a deal (Branch A): limits on the scope of the war, such as no ground forces, no occupation, and no regime change, plus visibility through reports. A fixed end date did not survive.
• What a standoff produces (Branch B): leverage that moves to a must-pass deadline, with costs to people who have nothing to do with the war.
• What the purse cannot do. In either branch, no US law makes Iran, Israel, the Houthis, or the mediators agree to anything. Funding conditions shape US participation; ending the war still requires a negotiated settlement.

Dissenters' views:

• Restraint: funding the war with conditions is still funding the war.
• Decisive outcome: certifications and reports slow commanders and signal limits to an adversary.
• Pressure-for-a-deal: removing $16.6 billion in requested items shortchanges the Pentagon.
• Envoy: limits on scope can reassure regional mediators that the war will not widen.
• Voices (the teacher in Isfahan): "None of this tells me when the strikes stop."

8. Lessons for the protocol