← Miss-ile-Lago

Methodology

How we count. What we include. What we deliberately exclude. And who edits this.

What counts as a war cost

We count direct, trackable U.S. military expenditure tied to the conflict with Iran and aid to Israel since April 2024. That means appropriations Congress passed, weapons Congress authorized, munitions the Pentagon reports firing, equipment Iran destroyed, and operations the Defense Department ran.

If a dollar figure appears in a congressional bill, a CBO score, a CRS report, a DSCA notification, a Pentagon press briefing, or a published estimate from a named defense analyst at CSIS / JINSA / AEI / FPRI, it qualifies. If it doesn’t, it doesn’t.

What we explicitly DO NOT count

This tracker omits, on purpose:
  • Economic activity generated by defense jobs (real, but not a cost)
  • Deterrent value or foregone-conflict estimates (speculative)
  • Long-term veterans care for injuries sustained (tracked by Brown’s Costs of War separately)
  • Iranian casualties and damage (not a U.S. cost)
  • Israeli expenditure (not a U.S. cost)
  • The Pentagon’s $200B supplemental request — requested, not appropriated
  • “Opportunity cost” framings beyond the direct comparisons shown

We leave these out because the direct-cost number is already hard to argue with, and inflating it with speculative additions hands critics an easy dismissal. The goal is to be unable to be dismissed as made up.

Judgment calls we made

Weapon unit costs

We use FY2026 DoD procurement unit costs where published. These are what Congress authorizes Raytheon, Lockheed, and Boeing to charge for a new unit. When a replacement cost estimate exists (e.g. CSIS cites $3.6M per Tomahawk during the current production crunch), we note it but anchor to procurement. Procurement is defensible; replacement is situational.

Operation Epic Fury total

AEI’s April 2026 analysis puts the 40-day active-combat phase at $22.3–31B. CSIS (Cancian) cites ~$500M/day average. We use $25B— the low end of AEI’s range — for defensibility. Going higher would be within sourcing but harder to defend against gotcha questions.

Burn rate on the homepage counter

The live counter runs at $1.4B/day, which is the midpoint between CSIS’s early-phase $890M/day estimate and the Pentagon’s sustained $1.88B/day rate over Day 1–6. This reflects Epic Fury surge tempo. The pre-surge period (April 2024 – February 2026) ran at ~$26M/day — mostly Israel aid packages rather than daily ops. A phased burn-rate model is planned.

Equipment losses

We use published replacement costs (Raytheon contracts, Lockheed contracts, official USAF fact sheets). Iran’s claims are cross-checked against NYT/Bloomberg satellite imagery analysis before we book a loss as confirmed.

How often this gets updated

Every 2 weeks during active operations. Monthly otherwise. Full data-refresh workflow: verify every existing line item against current primary sources, research the gap period, flag outliers, post revised numbers, update live constants, ship. See iranwarcost.com for an upstream cross-check — though that repo is currently stale.

Corrections

If a number looks wrong, tell us what source says otherwise. We log every change in CHANGELOG.md with the date, the old value, the new value, and the source that moved it.

See the full sources index for every URL cited across every card.

Who made this

Shahzad Ahsan. Brooklyn, New York. Former Creative Strategy at Spotify, former Head of Software Marketing at Logitech. I build sites that turn complicated things into readable things. No institution funds this; no ads; no tracking beyond aggregate page views.

The rule: if a number here doesn’t have a source, it shouldn’t be here. If you find one that doesn’t, that’s a bug. Report it.