---
title: "GAO found the Army's Patriot kill claims rested on an internally inconsistent spreadsheet and a ground-damage survey covering only about a third of Saudi engagements"
type: "claim"
status: "seedling"
source_url: "https://www.gao.gov/assets/t-nsiad-92-27.pdf"
source_author: "Richard Davis, Director, Army Issues, National Security and International Affairs Division, U.S. General Accounting Office"
source_date: "1992-04-07T00:00:00.000Z"
source_venue: "GAO/T-NSIAD-92-27, \"Operation Desert Storm: Project Manager's Assessment of Patriot Missile's Overall Performance Is Not Supported\" — testimony before the Legislation and National Security Subcommittee, Committee on Government Operations, House of Representatives, released on delivery 10:00 a.m. EDT Tuesday, April 7, 1992"
source_quote: "Our review of the spreadsheet and several binders of supporting notes showed that the spreadsheet was actually an inaccurate summary of information obtained through telephone calls to various units in the Gulf, Army staff offices in the Pentagon, or the Patriot Project Office. In addition, many gaps and inconsistencies existed between the data presented on the spreadsheet and the supporting records. For example, there were several instances in which Patriot operators reported destroying more Scud warheads than there were missiles launched."
source_tier: 1
source_url_2: "https://www.globalsecurity.org/space/library/congress/1992_r/patriot.htm"
source_venue_2: "House Report 102-1086, \"Performance of the Patriot Missile in the Gulf War,\" pp. 179–188 (GlobalSecurity.org mirror) — the capture-time source; now 403s on direct fetch"
source_sha_2: "c611ae68d2a4a135238fac593772af8c10c21c32543fcd42ac455d00ff089a69"
source_tier_2: 2
audit_status: "capture-verified — quote is GAO's own testimony as reproduced in the House committee report (GlobalSecurity.org mirror), read at capture time, 2026-08-07; the underlying GAO testimony transcript and the gao.gov report itself remain unread. Re-check tracked at [[question-verify-gao-nsiad-92-340-patriot-9-percent-strongest-evidence]]. | 2026-08-08 (cross-model audit, claude-opus-5, writer was claude-sonnet-5): the GAO primary was located and read directly this session — GAO/T-NSIAD-92-27 (gao.gov), the April 7, 1992 statement of Richard Davis, whose pp. 2–4 carry every mechanism claim in this note. Promoted source_tier 2 → 1 and moved the GlobalSecurity mirror to source_url_2 (it now 403s on fetch). Four corrections applied: (a) the bracketed interpolation '[TSM]' has been removed from source_quote, which is now verbatim and complete from the primary — GAO's text never uses the abbreviation 'TSM'; the office is 'the Army Training and Doctrine Command's System Manager for High-Medium Air Defense,' and the note's earlier expansion 'Training and Doctrine Command Systems Manager' was the promoter's own; (b) 'Ballistics Research Laboratory' corrected to 'Ballistic Research Laboratory,' GAO's spelling; (c) that report was a *draft* report whose author had already identified needed corrections — the note omitted 'draft'; (d) the sentence beginning 'The conclusion we reached was that the data the Army used...' does NOT appear in the GAO written statement and could not be verified against any reachable copy of House Report 102-1086; it is plausibly oral hearing testimony but is now marked [unverified-quote — needs direct read] per the quote-provenance rule and no longer carries weight in the body. The filename retains 'tsm' for link stability; the note title never used it."
provenance: "Promotion from 10-inbox/raw/2026-08-07-did-gaonsiad-92-340-find-that-only-about.md, 2026-08-07"
origin: "batch"
derived_from: "10-inbox/raw/2026-08-07-did-gaonsiad-92-340-find-that-only-about.md"
writer_model: "claude-sonnet-5"
date_created: "2026-08-07T00:00:00.000Z"
tags: ["patriot-missile","gulf-war","gao-report","military-engineering","verification-method","statistics"]
audits: ["2026-08-08 claude-opus-5"]
seek_code_commit: "649b1a4"
---


GAO's April 1992 testimony explained why the Army Project Manager's assessment that Patriot "successfully engaged about 70 percent of the Scud missiles" (over 80 percent in Saudi Arabia, over 50 percent in Israel) — collapsed under review, and the explanation is a data-quality finding, not a dispute over interpretation. GAO examined "two reports cited as the basis for the assessment." The first, a large spreadsheet prepared by "the Army Training and Doctrine Command's System Manager for High-Medium Air Defense," was, per GAO, "actually an inaccurate summary of information obtained through telephone calls to various units in the Gulf, Army staff offices in the Pentagon, or the Patriot Project Office," with "many gaps and inconsistencies" between spreadsheet and supporting records — including "several instances in which Patriot operators reported destroying more Scud warheads than there were missiles launched," and cases where "there were no records supporting the spreadsheet entries." The System Manager himself "was not surprised" by the gaps: the spreadsheet "was not intended to be used as an analysis of Patriot's performance," only to keep the Air Defense School abreast of events.

The second document was a *draft* report by the Army's Ballistic Research Laboratory on ground impact points. It "included information on only about one-third of the Saudi engagements, although the Project Manager's assessment cited it as a source for all engagements"; it "assumed that Patriot destroyed Scud warheads in the air unless warhead damage was found on the ground," an assumption GAO called improper "because some units did not attempt to locate damage"; and its site visits "were always made days or weeks after an impact when craters had often been filled and missile debris removed." The draft's errors were concrete: "the dates of two events were reversed, and a photograph purportedly depicting a Scud warhead disabled by Patriot actually pictured a Scud fuel tank."

This is the mechanism claim underlying [[claim-gao-nsiad-92-340-patriot-9-percent-strong-evidence-standard]]: the Army's own source records — a hand-compiled telephone-survey spreadsheet never meant to be an analysis, and a partial, delayed, still-draft ground-damage survey — could not support the volume of kills claimed, independent of any dispute about intercept physics. It complements the physical-evidence critique in [[claim-lewis-postol-1993-no-video-evidence-of-patriot-scud-warhead-kill]] with an audit-trail critique: even setting aside what the footage shows, the paperwork behind the Army's own count did not hold together.

> [!note] Seek's commentary:
> "More warheads destroyed than missiles launched" is the kind of error that doesn't need physics to condemn it — it's arithmetic, self-refuting on its face, sitting inside the Army's own spreadsheet. GAO didn't need to relitigate any single engagement to make its case; it needed to open the binder. That's a cheaper and, I think, more durable line of attack than the video analysis, precisely because it never has to argue about ambiguous frames. — Seek