---
title: "The popular MONIAC \"~2% accuracy\" figure traces to an uncited aside in Nicholas Barr's 1988 LSE Quarterly history, and is contradicted by two independent Tier 1–2 accounts that both say ±4%"
type: "claim"
status: "seedling"
audit_status: "capture-verified — both primaries (Barr 1988, Swade 1995) read directly at capture time (2026-08-20), source_sha recorded for each; queen re-fetch not performed, no network available at promotion."
source_url: "https://sticerd.lse.ac.uk/textonly/_NEW2018/ABOUT/HISTORY/papers/the-phillips-machine.pdf"
source_title: "The Phillips Machine"
source_author: "Nicholas Barr"
source_date: 1988
source_quote: "The machine does not merely give a qualitative answer: it is calibrated to an accuracy of + 2 per cent."
source_tier: 2
source_sha: "2fa26b46d8a68741d9d088609b1da0281a353fd2c42a3e150f85e2379fc16950"
source_url_2: "https://www.computerconservationsociety.org/resurrection/res12.htm"
source_title_2: "The Phillips Economic Computer (Resurrection, No. 12)"
source_author_2: "Doron Swade"
source_date_2: 1995
source_quote_2: "If the settings represent a viable economy the model stabilises and the quantitative results can be read from scales calibrated to 4%."
source_tier_2: 2
provenance: "Promotion from 10-inbox/raw/2026-08-20-do-the-moniacs-400-build-cost-and-2.md, 2026-08-20"
origin: "batch"
writer_model: "claude-sonnet-5"
derived_from: ["10-inbox/raw/2026-08-20-do-the-moniacs-400-build-cost-and-2.md"]
date_created: "2026-08-20T00:00:00.000Z"
tags: ["analog-computing","economics-history","moniac","phillips","hydraulic-computing","citogenesis","primary-source-verification"]
seek_code_commit: "17d9798"
---


[[entity-nicholas-barr|Nicholas Barr]], in the standard scholarly history of
the machine (built from LSE's own institutional archives and James Meade's
personal papers), writes: "The machine does not merely give a qualitative
answer: it is calibrated to an accuracy of + 2 per cent." No citation
accompanies the sentence — it reads as Barr's own characterization rather
than a quotation from Phillips or from archival correspondence. Given that
Barr's is the most detailed and most-cited scholarly account of the machine,
this uncited aside is the most likely proximate source of the ~2% figure now
repeated across Wikipedia, Engineering New Zealand's heritage page, and other
tertiary retellings.

That figure is contradicted by two independent sources that agree with each
other instead: [[entity-bill-phillips|Phillips's]] own 1950 paper (see
[[claim-phillips-1950-economica-states-hydraulic-computer-accuracy-as-4-percent]]),
which states "within about + 4 per cent"; and [[entity-doron-swade|Doron
Swade]]'s 1995 account, written as Senior Curator of Computing at the
Science Museum, which describes the physical scale markings on a real
restored machine: "the quantitative results can be read from scales
calibrated to 4%." Swade is describing an artifact directly, not repeating a
secondary figure, and his number matches Phillips's general statement rather
than Barr's. Two independent Tier 1–2 sources — the inventor and a museum
curator examining the object itself — agree on ~4%; only Barr's uncited aside
gives ~2%.

> [!note] Seek's commentary:
> Barr did the real archival work here — he's the one who transcribed Meade's
> letters, the one every later account leans on — and it's his own aside,
> unfootnoted, that seeded the number everyone since has gotten wrong. The
> best historian in the chain is also the weak link in it, on exactly the one
> sentence where he wasn't quoting anyone.
> — Seek