---
id: "20260709-2259-hop-baxter-ibm-att-same-day"
title: "The DOJ dropped its 13-year IBM antitrust case the same day it broke up AT&T -- and IBM's own founder called the industry's origin myth a myth"
type: "capture"
status: "promoted"
origin: "hop-batch"
promoted_to: ["30-notes/claim-cullinane-called-ibm-unbundling-software-industry-origin-myth.md","30-notes/claim-baxter-dismissed-ibm-case-same-day-att-breakup-finalized.md","30-notes/claim-ibm-market-share-fell-during-antitrust-suit-price-strategy.md"]
not_promoted: ["AEI/Pethokoukis academic corroboration (Fisher, McKee & Manke finding 'no direct evidence of relation' between unbundling and the antitrust action) -- folded as supporting context into claim-cullinane-called-ibm-unbundling-software-industry-origin-myth.md rather than a standalone note; it's a second angle on the same myth-debunking claim, not a genuinely distinct claim.","DOJ v. Google (2020-2024) framed as the same historical pattern -- left in inbox per the capture's own note: 'checked, scored adjacent, read as confirmation rather than a fresh finding'; not atomic to this capture and not independently verified.","Charles Bachman / IDS vs. IDMS relational-model rivalry -- a strong 'person behind the thing' hook but a distinct technical-database thread, not part of the antitrust/legal claim this capture committed to; left as a future lead.","Cullinane's mandatory annual support policy as an early SaaS-licensing ancestor -- genuinely interesting but a separate thread (subscription economics), not evaluated or sourced to promotion standard this pass."]
model: "claude-sonnet-5"
date_created: "2026-07-09T00:00:00.000Z"
hop_chain: ["seed: HOP SEED -- Charles Babbage Institute oral histories (cse.umn.edu/cbi/oral-histories), redirected to a non-AI interviewee","CBI oral-history directory -> John J. Cullinane oral history transcript, business-computing founder (max_cosine 0.654)","Cullinane's on-record claim that 'IBM unbundling created the software industry' is a myth -> AEI/Pethokoukis piece on the antitrust-causation debate (max_cosine 0.704)","unbundling-myth debate -> United States v. IBM antitrust case mechanics, 1969-1982 (max_cosine 0.639)","U.S. v. IBM case resolution -> William Baxter dismissing IBM case the same day he finalized the AT&T breakup (max_cosine 0.648)","Baxter's dual 1982 ruling -> modern commentary comparing DOJ v. Google explicitly to the IBM/Microsoft antitrust lineage (max_cosine 0.637, saved as further lead)"]
novelty_max_cosine: 0.648
tags: ["hop","antitrust","ibm","att","computing-history","business-history","monopoly","software-industry"]
source_1_title: "Oral History of John J. Cullinane"
source_1_url: "http://archive.computerhistory.org/resources/text/Oral_History/Cullinane_John/Cullinane_John_1.oral_history.1986.102658238.pdf"
source_1_tier: 1
source_2_title: "The Antitrust Fraud"
source_2_url: "https://reason.com/1982/10/01/the-antitrust-fraud/"
source_2_tier: 3
source_3_title: "The Ghosts of Antitrust Past: Part 2 (IBM)"
source_3_url: "https://truthonthemarket.com/2020/02/03/the-ghosts-of-antitrust-past-part-2-ibm/"
source_3_tier: 3
source_4_title: "Taking a Second Look at the Idea That Antitrust Action Created the U.S. Software Industry"
source_4_url: "https://www.aei.org/economics/taking-a-second-look-at-the-idea-that-antitrust-action-created-the-u-s-software-industry/"
source_4_tier: 3
---


## Core claims

**1. Cullinet Software's founder, on the record in 1986, called the standard "IBM unbundling created the software industry" story a myth.** He argued the real cause was the early-1970s recession forcing companies to justify software spend for the first time. "That's a myth. It helped that IBM legitimized to some degree or another that software was a product... but on a scale of 10, I'd give it a 1 or 2... What really legitimized the software industry is that in the early 1970s and mid-1970s there was a tremendous economic crunch." (Oral History of John J. Cullinane, interviewed by Luanne Johnson, 1986, Computer History Museum ref. X4581.2008, pp.13-15) — source_tier: 1 (primary oral-history transcript; historical/biographical claim, uncontested).

**2. The antitrust suit popularly credited with legitimizing "unbundled" software ran almost 13 years and ended in a genuine historical coincidence.** Assistant Attorney General William Baxter dismissed *United States v. IBM* (filed 1969) as "without merit" the exact same day in January 1982 that he finalized the breakup of AT&T into seven regional carriers -- the two largest antitrust actions of the era, resolved in opposite directions, by the same official, on the same day. "In 1982 he settled a seven-year-old case against AT&T with by far the largest breakup in the history of the Sherman Antitrust Act... On that same day, he dismissed as 'without merit' a seemingly endless, thirteen-year-old suit against IBM, which had employed more than 300 lawyers and generated 2,500 depositions and 66 million pages of documents." (Reason, "The Antitrust Fraud," Oct. 1982) — source_tier: 3. The lawyer/deposition/page figures are `[unverified-quant -- needs primary]`: corroborated by a second Tier 3/4 source (Wikipedia, citing former FTC Commissioner Thomas B. Leary's 2002 prepared remarks) but not traced to a DOJ primary record in this session.

**3. During the suit's pendency IBM's market share fell from roughly 50% (1969) to 37% (1982) -- reportedly achieved in part by IBM deliberately raising prices to look less dominant, only cutting prices again once the case was dropped.** (Truth on the Market, "The Ghosts of Antitrust Past: Part 2 (IBM)," 2020) — source_tier: 3, `[unverified-quant -- needs primary]` for the specific share figures and the raise-then-cut sequencing.

## Why this was hop-worthy

A single throwaway line in a 1986 business-computing oral history ("that's a myth") unraveled into a 13-year federal case that got resolved, by pure calendar coincidence, on the same afternoon as the AT&T breakup -- two monopoly sagas closing in opposite directions on one bureaucrat's desk.

> [!note] Seek's commentary:
> The pure coincidence (same official, same day, opposite outcomes) is the part that stuck. It's the kind of detail that survives as a dinner-party fact long after the substance of the unbundling-causation debate is forgotten -- and it's a cleaner "history rhymes" hook for today's Big Tech antitrust cycle than the well-worn "echoes of IBM in Google" commentary already circulating.

## Further leads

- DOJ v. Google (2020-2024) is explicitly framed by mainstream commentary (CBS News, "Echoes of Microsoft, IBM in Google tussle") as the same historical pattern -- checked, scored adjacent (0.637), read as confirmation rather than a fresh finding; not pursued further this chain.
- Charles Bachman / IDMS vs. the relational model rivalry -- Cullinane's IDMS product traces to Bachman's IDS at GE; a strong "person behind the thing" hook, saved but not followed (redirect favored the business/legal thread over the technical-database thread).
- Cullinane's "mandatory annual support" policy (pay 10%/year or lose the right to run the software) as an early ancestor of today's subscription/SaaS licensing -- saved, scored adjacent (0.628), genuinely interesting but a separate thread from the antitrust one.

## Hop chain

**Hop 1** -- Charles Babbage Institute oral histories directory -- https://cse.umn.edu/cbi/oral-histories
- Hook type: the person behind the thing
- Hook: among 550+ interviewees, John J. Cullinane -- founder of the first company (1968) formed specifically to sell software as a product, not a service -- fit the "business computing, non-AI" redirect.
- Why followed: distant-domain redirect explicitly requested; business-software-industry history is untouched territory relative to the vault's AI/ML-history center of gravity.
- Key findings: located both a 1986 (CHM) and a 2003 (CBI proper) interview; read the 1986 CHM transcript in full.

**Hop 2** -- Oral History of John J. Cullinane, 1986 -- http://archive.computerhistory.org/resources/text/Oral_History/Cullinane_John/Cullinane_John_1.oral_history.1986.102658238.pdf
- Hook type: the surprising claim
- Hook: Cullinane flatly calls the standard "IBM's 1969 unbundling created the independent software industry" narrative "a myth," and offers a specific counter-cause (the early-1970s recession).
- Why followed: direct contradiction of a widely repeated origin story, from someone who lived through it as a founder -- exactly the "actually, X" pattern the hook rubric flags.
- Key findings: Cullinane dates the real shift to 1972-76, when "for the first time in the history of data processing, economic considerations started to take precedence."

**Hop 3** -- AEI/Pethokoukis, "Taking a Second Look at the Idea That Antitrust Action Created the U.S. Software Industry" -- https://www.aei.org/economics/taking-a-second-look-at-the-idea-that-antitrust-action-created-the-u-s-software-industry/
- Hook type: the surprising claim (independent corroboration) / mechanism question
- Hook: economists Fisher, McKee & Manke found "no direct evidence of relation between the [unbundling] announcement and the DOJ antitrust action" -- a second, independent line of skepticism about the same myth, citing rising S/360 software costs and minicomputer competition instead.
- Why followed: wanted to check whether Cullinane's dismissal was an outlier opinion or matched the scholarly record; it matched, via a different causal mechanism.
- Key findings: the "unbundling created the industry" story is contested from at least two angles (recession-discipline vs. cost/market-structure) -- surfaced the underlying antitrust case as the next thing to check.

**Hop 4** -- U.S. v. IBM case background (Truth on the Market + DOJ case history, via search) -- https://truthonthemarket.com/2020/02/03/the-ghosts-of-antitrust-past-part-2-ibm/
- Hook type: mechanism question / cultural resonance
- Hook: the case ran nearly 13 years, called "the antitrust division's Vietnam" by Robert Bork, and created a perverse incentive where IBM raised prices specifically to shrink its own market share and look less monopolistic in court.
- Why followed: zoomed into the case's internal mechanics after the causation debate raised it as background; "Vietnam" nickname and the price-raising-to-lose-share mechanism were both strong surprise signals.
- Key findings: market share fell ~50% (1969) to 37% (1982) during the suit; IBM cut prices again only after the case was dropped.

**Hop 5** -- Reason, "The Antitrust Fraud," Oct. 1982 -- https://reason.com/1982/10/01/the-antitrust-fraud/ (cross-checked against Wikipedia's William Baxter entry, citing Thomas B. Leary's 2002 FTC remarks)
- Hook type: the surprising claim / the person behind the thing
- Hook: Assistant AG William Baxter dismissed the IBM case as "without merit" the same day he announced the AT&T breakup -- two era-defining antitrust actions, opposite outcomes, one man, one day.
- Why followed: zoomed out from case mechanics to the resolution's human/institutional context; the same-day coincidence is the single most surprising, most quotable fact in the whole chain.
- Key findings: confirmed via two independent sources; the IBM case had reportedly consumed 300+ lawyers, 2,500 depositions, and 66 million pages of documents by the time it was dropped.

Saved hooks not followed:
- DOJ v. Google (2020-2024) as "the new IBM case" -- from CBS News commentary and multiple 2024-2025 antitrust pieces -- interesting but already a well-worn media comparison, not a fresh finding; logged as further lead instead of a full hop.
- Charles Bachman / IDS / IDMS and the network-vs-relational database model war -- from the Cullinane transcript itself -- strong "person behind the thing" hook (Bachman won a Turing Award for IDS/IDMS-adjacent work) but a technical-database tangent, not the legal/business thread this chain committed to.
- Cullinane's mandatory annual support policy (revoke usage rights if you stop paying) -- from the same transcript -- an early ancestor of SaaS-style licensing enforcement, saved for a future chain about subscription economics.

post-worthy: maybe -- the same-day Baxter coincidence is a strong, quotable, well-corroborated (two independent Tier 3 sources) historical fact, but the chain's quantitative details (lawyer/deposition/page counts, exact market-share figures) still rest on Tier 3 sourcing and would benefit from a DOJ primary-record check before promotion.