Control Data Australia Memories compiled by Brian Membrey

1973 : The IBM Settlement

"COMPUTERS: A Settlement for IBM", Time magazine, Monday, Jan. 29, 1973.

“Peace negotiators have been meeting secretly for the past year at law offices in Manhattan and Minneapolis-St. Paul to settle a long and chafing conflict between two industrial powers. On one side was International Business Machines Corp. of Armonk, N.Y., the world's largest maker of computers; on the other was Control Data Corp. of Minneapolis, which ranks fifth in the U.S. computer industry. Last week the companies announced an agreement: Control Data lawyers consented to drop the company's four-year-old antitrust suit against IBM.

"In return for dropping its suit, Control Data won a good deal. For about $16 million, it will acquire IBM's Service Bureau Corp., a subsidiary that processes customers' data and sells time on its own computers. Wall Street analysts reckon that the Service Bureau's real market value is closer to $60 million. In addition, IBM will buy services from the bureau for five years, stay out of the services business itself in the U.S. for six years and reimburse Control Data for $15 million in legal fees spent on the case. Total cost of the package to IBM: at least $80 million. William C. Norris, Control Data's one-man-gang chairman, said that the daring suit had turned out to be “one of the best management decisions in our history ...".

Still regarded as one of the most famous civil cases in U.S. legal history, the basis of  CDC Anti-Trust suit against IBM traces back to August, 1965, when just as Control Data prepared to deliver its first 6600, IBM pre-announced a new competitor, the 360/92 which was claimed to be more powerful and flexible than the 6600 and fully compatible with the remainder of it’s 360 range.

Perhaps inspired by the CDC initiative in moving Seymour Cray's design team "out-of-house" laboratories, some two hundred IBM employees were gathered together on the U.S. West Coast to work on the project (apparently dubbed during the conceptual stage as the ACS-1) away from corporate interference.

The project produced interesting computer architecture and technology, but the initial design proved incompatible with IBM's hugely successful System/360 line of computers.  Under management direction, the engineers solved the compatibility issues, but with a serious drop in performance and the ACS-1 project was dropped in 1969 without a single production machine being completed.

In reality, the early CDC 6600 models were experiencing their own difficulties, especially with memory. With penalty clauses for non-performance - suggested in one place at $90,000 per month in the huge Atomic Energy Commission contract - the technical problems along with a strong move away from outright sales to leasing and a resultant long-term cost recovery were deemed largely responsible for CDC posting a $1.8 million loss in 1966 after many years of strong profitability.

IBM continued to pledge delivery dates and accepted orders for what was now promoted as the 360/90 Series, cutting into potential Control Data business. At one point CDC was said to have been unable to book a single order for the 6600 for nearly 18 months as a result of the "ghost" computer strategy with existing IBM customers who had been considering upgrading to the 6600 waiting for the release of the mythical Model 90 series.

“FUD you, I.B.M”, says Bill Norris  

CDC founder and Chairman Bill Norris did not take this tactic lying done and firmly believed that the Model 92 was never a serious project after the initial ACS-1 proved incompatible with the 360 - one report suggested he dubbed the IBM campaign as based entirely on "FUD" - "Fear, Uncertainty and Doubt", and on 11 December, 1968 in the Federal Distict Court, St. Paul. Minnesota and in a move that caught many industry observers by surprise, CDC launched an Anti-Trust suit against IBM claiming unspecified damages and seeking trial by jury.

The complaint alleged that IBM had monopolized the market for computers in violation of section 2 of the Sherman Act [1] and that this conduct had damaged CDC’s business, entitling CDC to treble damages plus attorneys’ fees under section 4 of the Clayton Act.  The case was filed in Minnesota’s federal court.

[1] The Sherman Anti-Trust Act was a landmark federal statute passed by the U.S. Congress in 1890. In short, the purpose of the Sherman Act was not to protect competitors from harm from legitimately successful businesses, nor to prevent businesses from gaining honest profits from consumers, but rather to preserve a competitive marketplace to protect consumers from abuse and acts by a monopolist to artificially preserve that status, or disreputable  dealings to create or maintain a monopoly.

IBM immediately engaged its outside general counsel, Cravath, Swaine & Moore (CS&M), to defend the case and began organising a number of crash training courses to bring its team of predominately computer-illiterate lawyers up to speed.  Control Data were represented by their St. Paul lawyers, Messrs Oppenheimer, Wolff, Foster, Shepard and Donnelly.

And the U.S. Government

On 17 January, 1969, the U.S. Department of Justice filed suit in the U.S. District Court, New York, against IBM  (also under Section 2 of the Sherman Act) with several claims of unfair competition by bundling prices, premature product announcements when it  knew that its product was years from completion (mimicking the CDC accusation) and predatory pricing.

Over the next two or three years, several other companies took their own action - ADR (Autoflow flow-charting software); Greyhound Computer Corp, Hudson and Trans-America (all leasing companies), and Applied Data Research, Memorex, Calcomp and Telex (plug-compatible equipment) filed anti-trust suits which were transferred to the jurisdiction of the Minnesota federal court for pretrial  regardless of where they initially had been filed.

Minnesota’s U.S. District Judge Philip Neville was put in charge of managing all of these complicated cases where one of the major issues was defining the market, geography and product lines. Plaintiffs argued for definitions that produced large market shares for IBM while IBM argued for different definitions and lower market shares.

The Case - For and For

At the time, there was little publicity given to the three major lawsuits - secrecy rules in the CDC case did not allow for the dissemination of the progress of the trial or the positions taken by the principals at the trial - although they have come under considerable scrutiny since with some observers considering that IBM was, in fact, exercising common-sense marketing policies (pretty much the final assessment in the U.S. Justice Department trial).

In general, the primary complaints raised (generally with a load of statistical data in support) were that :=

(In addition, CDC specifically objected to IBM's preemptive announcements of equipment that it never intended to produce in order that prospective Control Data customers (particularly for the 6600) would hold off purchasing; Telex alleged in its suit that IBM deprived it of the right to compete by reducing prices to unreasonable levels to cut Telex out of the market; and the Justice Department suit followed Telex in complaining that IBM had introduced low-profit machines in markets where competitors have been very successful, and asked that IBM be prohibited from producing equipment on which it was not likely to make profits comparable to those it expected on other product lines).

In June, 1969 following the lodgment of the suits, IBM announced plans to drop its heavy discounting to educational institutes and to "unbundle"; previously it priced systems to include software, training and servicing, meaning (a) buyers were charged for services they often did not require or could obtain more cheaply elsewhere, and (b) potential alternate suppliers of software and services could not  assess competitive prices.  CDC itself did not “unbundle” until early 1970.

IBM and “Unbundling”

Burton Grad, ex-IBM executive in an Oral History "Association of Data Processing Service Organizations (ADAPSO) reunion workshop : IBM relations"

"The most significant suit as far as IBM was concerned was the Control Data suit in terms of its scope and the skill with which CDC pursued it. I was on IBM's unbundling task force and this is my recollection. We weren’t allowed to carry out any papers so I don't have any documentation, only what I can remember. "

"It got stirred up, apparently, by Burke Marshall and Nicholas Katzenbach, IBM attorneys, telling Tom Watson, Jr., IBM's chairman at the time, that systems engineering was a tie-in sale and that there was no way they could ever win a suit on a tie-in sale on systems engineering. That was the primary impetus for unbundling.

"But if they made the decision to unbundle systems engineering, they also had to unbundle education and field engineering where they were getting pressure from independent firms that wanted to be able to maintain any manufacturer's equipment. IBM announced the decision to unbundle in 1968 and hoped that would avoid a Justice Department suit. At least, that’s what Marshall and Katzenbach had told them: If you go ahead and do it on your own you probably won’t get sued by the government. Again, this is anecdotal. It's what I was told.

"In January of 1969 there was a big shock when the suit happened anyway. As most of you know, the government didn’t push the suit very hard for quite awhile.

(Interviewer): "for twelve years ..."

"No, for just the first three years, there was not much happening. It seemed like the government was depending upon Control Data to push ahead. But Control Data’s suit was not about software or about services; that wasn’t where their concern was. They were pushing on IBM’s so-called "fighting" machines and early announcements and those kinds of things.   They were very good".

The CDC Case

The CDC case was eventually set aside from the others on the base that an action involving the mainframe market and IBM’s “fighting machine” was too dissimilar to be heard conjointly with the others which involved peripherals, maintenace services or lease finance.  

The suit was eventually scheduled to commence on November 5, 1973 in a case where it was later suggested several million pages of documentary evidence would be involved, but in a move that surprised most observers (in particular the U.S. Department of Justice), IBM settled the CDC suit out-of-court on Monday, 8 January, 1973 - the three cheques above all show a date of 11 January and it appears were written almost immediately after the ink was dry on the settlement agreement.

It is probably impossible to accurately quantify, but one account suggests CDC's legal team examined over twenty million pages of IBM documentation, of which over a million were photocopied and around 150,000 documents (around half a million pages) added to its computerized database - this constituted the first time that high-speed computer technology had been used in a major legal case. Sadly, nothing remains of the equipment or techniques involved, but one suspects CDC's acquisition in 1964 of the Rabinow Corporation specialising in OCR may have paid for itself several times over!

The settlement involved the rights for CDC to acquire IBM's Service Bureau Corporation (SBC) at a nominal book price of $16 million (a valuation put the true value of SBC as between $45 and $60 million based on a rough industry standard that suggested businesses be valued at their net yearly earnings multiplied by 30). SBC earned just on $1.5 million in fiscal year 1972 based on revenue of $63 million - a drop in the ocean compared to IBM’s overall revenue of $9.5 billion).

Based on those figures, one suspects no-one at IBM lost too much sleep in allowing SBC to slip from their grasp, but perhaps they do not reflect the true value to the company of the service operation and its punched card-tabulator predecessor - a substantial part of IBM's bureau strategy was to target small data processing customers and "grow" them into tabulating equipment and later digital computers.

SBC had 40 data centres operating across the U.S., mostly operating IBM 360 models and at what might be termed the lower-end of the market offering mostly batch processing, time-sharing via Call/360 and pre-packaged software and with considerable higher levels of customer and marketing support than CDC’s two offerings - Kronos and Cybernet, which were regarded as largely for expert users capable of developing and processing software to meet their own requirements. Bob Price in his memoirs suggested that SBC was “slightly bigger” than CDC’s own data services operation.   Although now owned by CDC, SBC continued to operate under its own name and with little change to its mode of operation.

IBM also made considerable use of SBC for its own internal processing and the settlement also agreed that IBM would continue this processing with SBC at the current level of $5 million for five years, that IBM would to pay retirement and other one-off benefits of current SBC employees for ten years, estimated at $2.6 million per year [2], and that IBM would provide free rental and service of their equipment for six months, estimated at $5 million. Finally, IBM agreed that it would not compete in the service provider market for six years; it finally re-entered with a time-sharing service based in Tampa, Florida in 1982.

[2] Although never fully documented as an issue at the time, the accompanying restriction that SBC could not utilise the services of IBM employees would almost certainly meant that the existing SBC staff would have had to resign from IBM with the resultant payout of existing benefits and then, if they desired, apply for re-employment with the Control Data-managed re-invention of SBC - all-in-all, a somewhat futile and time-wasting exercise, but one that probably cost IBM a substantial sum in benefit payouts in the first year.  None of the material sighted to date mentions the likely number of IBM/SDC employees, but given the 40 data centres operating at the time, the number was in all likelihood over 1,000.

The final part of the settlement was that IBM would fund CDC Research and Development projects at $6 million a year over the years with a world-wide patent cross-licensing agreement between the two until 30 June, 1978. There was also a reimbursement of $15 million in legal fees (this later proved significant as it was ruled in another action that the payment passed ownership of the document index that CDC’s legal team had prepared back to IBM and enabled its subsequent destruction)  The SBC deal was said to more than doubled CDC's data services revenue,.  The payout overall just on $101 million, plus the $30-45 million “discount” after CDC paid for SBC.

On face value, a disastrous result for IBM - or was it?

A Disaster for IBM?

Two aspects of the case suggest not.

As was standard in anti-trust cases under the Sherman Act, the claimant was entitled to treble the amount of damages awarded by the court if the defendant party was found in breach and that as a result, the claimant had suffered financial damage - the “treble damages” were suggested from day one of the CDC suit, although I have yet to sight anything to suggest whether CDC ever nominated a figure.

As no trial occurred, the matter of damages became irrelevant, but as a guide, another anti-trust suit brought by Telex Corporation against IBM was heard in September, 1973 and resulted in the court finding the company in breach of anti-trust laws and liable for damages payable to Telex set at $117.5 million or a cool $352.5 million when trebled.  Given the comparative size of CDC and Telex, it is probably not unreasonable to think any punitive damages awarded by the court against IBM could have reached a lazy billion dollars after they were trebled.

The “fly in the ointment” in the Telex case was they were found guilty in an IBM counter-suit heard in the same court for theft of trade secrets and industrial espionage, with IBM awarded $21.9 million as a result. The anti-trust verdict against IBM was overturned early in 1975, nullifying the damages and Telex eventually lodged a public filing with the Securities and Exchange Commission stating it was financially unable to satisfy the judgment in IBM's favour (then reduced to $18.5 million). IBM eventually settled without payment, but it was considered a major victory for them with the Justice Department case still pending.

Throughout the duration of the CDC case, the company compiled a huge computerised database of documents - there is little evidence that the U.S. Justice Department actively pursued any action during this period, depending instead of the material collated by the CDC legal team, as were others involved in actions against IBM, especially Telex.

Hidden away amongst the “big ticket” items in the settlement was an agreement that all of CDC's computerised documentation revert to IBM where it was predictably immediately destroyed. (Again, more on this to come in April, but both Telex (against CDC) and the Justice Department (against IBM) immediately launched actions to prevent further destruction of material - oops, too late! Someone pressed the DEL key!)

The removal of incriminating documents collected over four years led to IBM either winning all the alternative private suits or them being dropped because of a lack of evidence because claimants depending of the CDC data base to support their claims.

“Terrible Tommy’s” Recollections

Even at the time, some observers later suggested that IBM's settlement was designed solely to have the damaging index destroyed - their thoughts perhaps confirmed over a decade later when Thomas J. Watson Junior, the IBM Chairman at the time of the suit and apparently known as “Terrible Tommy” published an autobiography, "Father, Son and Co".., of which excerpts and some commentary appeared in the Canberra Times on 15 October, 1990 ...

"… And despite of everything, customers were still ordering 360s faster than we could build them".

At that time, however, IBM made what Watson Jr calls a "grievous mistake".

In August 1963, the tiny Control Data Corp had announced the 6600 super-computer, designed by Seymour Cray. It was far superior to anything IBM could produce. ‘Junior’ was piqued, and IBM announced that it would leapfrog it with a brand new model.

Junior notes that "the effect of this announcement was to put a chill on [CDCs] market".

"Even though our supercomputer didn't exist yet, a lot of customers decided to hold off buying from CDC until they could get a look."

CDC plunged into the red and IBM announced four different versions of this pseudo-computer before Junior conceded that "IBM was never able to beat the Cray design" and cancelled the project. But CDC sued, and this perhaps encouraged the US Justice Department to have another go at IBM. It launched yet another anti-trust suit in 1969.

IBM eventually bought off Control Data in 1972.  It sold CDC its service bureau for "a fraction of its real worth" and gave CDC $101 million worth of cash and contracts, including $15 million to cover legal fees.

"This settlement was a brilliant tactical stroke," writes Junior. CDC had indexed all IBM's documents on a computer and this "was the master link in all the other anti-trust suits against IBM". The settlement gave IBM control of the index, so we destroyed it.

"Without Control Data to help, the Justice Department's case deteriorated into a courthouse mess." This anti-trust suit dragged on for a decade, but after Ronald Reagan's election the case was suddenly dropped.

Given America's fear of Japan, and its adoption of IBM as a "national champion", it is unlikely to be reopened …".

… the Anti-Trust Division’s Vietnam

The U.S. Government Anti-Trust trial did not begin until 19 May, 1975 and lasted six years before the Justice Department determined that the case was "without merit" and withdrew it on 8 January, 1982, by which time most of the technology involved was many years out of date  - one account suggests that the court had heard testimony from 81 live witnesses, heard 928 depositions from other witnesses and 11,644 documents had been presented, the transcript of the case amounting to over 100,000 pages.

Before the commencement, the Government estimated that the presentation of its case would last 60 days – it took three years – with one of the Justice Department's lead counsels dubbing the case ''the Anti-Trust Division's Vietnam.

It would be fascinating to have Bill Norris' private thoughts on the settlement as well as Watson's, but unfortunately nothing other than a few carefully worded comments has been traced – the last page of his scrapbook includes a press clipping on the launching of the case, but is dated some five months after the previous entry and is obviously a last-minute addition. There are three Norris Oral History with Charles Babbage Institute representatives, but the only one on-line covers just the Norris years up to the formation of CDC in 1957.

Meanwhile, here in Oz …

From memory, we didn’t hear much of the case here in Oz - or least those of us working on the peasant levels of 598 - other than an article in Between Ourselves in March ‘73 which suggested it had no effect in Australia and included a brief statement from Norris

"... we are pleased with the settlement and are confident that the results will greatly strengthen Control Data" … “the industry in general had derived great benefit from the case”.

Given that SBC had no presence in Australia, perhaps the one impact of the acquisition was the introduction of the Call/370 time-sharing service in 1978.

Its predecessor, Call/360, was the SBC time-sharing service (introduced in 1968) at the time of the settlement and continued to operate after the company was acquired by CDC, but I’m not sure what the rationale was behind introducing its successor it into Australia - it appears to have been in direct opposition to Cybernet, but perhaps was  justified on the basis of what was probably a wider range of software applications available within the commercial market..

Post-mortem (literally)

Judge Philip Neville continued with the mix of anti-trust cases through 1973 before he died on Wednesday, 13 February, 1974  of leukaemia at the University of Minnesota Hospital, aged 64. Obituary notes following his passing reveal he was a graduate of the University of Minnesota Law School admitted to the bar in 1933, and appointed United States attorney for the Minnesota district,to the Federal bench in 1967 by President Lyndon B. Johnson.

“Judge Neville's most controversial decisions on the Federal bench came in a case in 1970 in which he ruled that the University of Minnesota could not refuse to hire a person merely because he was an avowed homosexual. In that case, involving a man who was refused a job as a university librarian after it became known that he had applied for a marriage license with a male student, Judge Neville declared in his decision, “a homosexual is a human being.”   (New York Times, 16 February, 1974)

From the Charles Babbage Institute CDC image archive : three bank cheques following the 1973 settlement to the tune of a lazy $65 million dollars drawn against various IBM accounts and countersigned on the rear to the Northwestern National Bank of Minneapolis account Control Data Corporation (bottom).

The background to the cheques is however somewhat unclear - by all accounts of the settlement which was put at between $100 and $120 million, there were no cash damages involved.

The $15 million payable on the centre cheque may have been the amount specifically set aside for legal expenses; CDC’s legal firm Oppenheimer, Wolff, Foster, Shepard and Donnelly of St. Paul firm,are referenced in the countersigning below.

The other two cheques suggest that it may have been agreed that two of the other parts of the settlement that nominally extended over a period of some years were to be paid out in advance.

The agreement that IBM would pay all retirement and other benefits of Service Bureau Corporation employees for the next ten years, estimated in 1973 as $2.6 million per year and is probably the cheque cheque (top) for that sum.

The bottom cheque for $25 million in turn could be IBM’s advance payment against the $5 million per annum for five years to maintain its level of internal data processing with Service Bureau Corp.

 That leaves the minor issue of the $30 million to fund R & D projects expected to amount to $6 million a year over five years, but there are references to this effort being paid based on “time and materials” and thus not calculable until after the event.

Even without that $30 million, one can imagine the look on the face of a spotty 19-year-old teller at the Northwestern National Bank of Minneapolis when someone from CDC wandered in one morning and plonked down a deposit slip for $65 million and then perhaps for twenty bucks in change for the petty cash tin!




Telex v IBM : the winner is ... The New York Times says ...


Curiouser and Curiouser "The Anti-Trust Division's Vietnam" Service Bureau Corporation