United States. Congress. Senate. Committee on the.

The Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) online

. (page 80 of 140)
Online LibraryUnited States. Congress. Senate. Committee on theThe Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) → online text (page 80 of 140)
Font size
QR-code for this ebook


price reduction is in excess of any value to IBM which could be attributed to
the buyback.

We understand that buybacks recently have been offered by IBM in connection
with computer procurements by AVashington State University ; the University



55G0

of California at San Diego ; the Southern Research Triangle jointly operated by
Duke University, the University of North Carolina and North Carolina State ;
Lawrence Radiation Laboratory at Livermore, California ; and Knolls Atomic
Power Laboratory operated by General Electric for the AEC. Within the last
several years, IBM has combined buybacks with large educational discounts to
provide computers practically free of charge to U.C.L.A. and Stanford University.
Only a monopolist, with the tremendous resources and market advantage avail-
able to it, can afford to engage in such indirect price cutting.

In addition to price cutting to thwart competition by means of buybacks, IBM
has abused its monopoly position by insidious price cutting in the form of extraor-
dinary programming or maintenance services offered to particular customers
for whom competition is especially serious. Such offers have been made to all
types of customers — educational institutions, governmental agencies and com-
mercial users. The price cutting takes the form both of increasing service to a
customer when it appears the customer is considering replacing IBM equipment
with competing equipment, and including extraordinary services in initial offers
where the competitive situation so demands. The follo'u-ing are examples :

(1) In recent competition for a major computer system to be installed at
NASA's Goddard Space Flight Center, IBM offered approximately 12.5 man
years of free programming suijport compared to the 30 man years which
Control Data was able to offer. At $20,000 per man year this is equal to a
$2,500,000 discount on a system where the list price was allegedlv but
about $8,000,000 initially.

(2) At Washington State University, when it became known that con-
sideration was being given to replacement of an IBM 709/1401, IBM brought
into town two customer engineers and an analyst to replace the one customer
tnmer that had been there servicing IBiM equipment. IBM told the cus-
tomer that if it purchased a 360/67 these men would be assigned to the cus-
tomer for the new equipment.

(3) In September. 1964, Sunsweet Growers contracted with IBM for a
360/30 system instead of a Control Data 3100 system, selecting IBM
partly because it promised to make available four IB^I employees to assist
in conversion of the present equipment to the new system.

3. Market Exclusion

IBM's previously discussed practices of locking in customers and price cutting-
are emjiloyed with the intent and effect of excluding and hindering competition
in the computer n^.r.rket and various segments of it. In recent competition against
IBM for certain key comjniter customers Control Data has encountered ]iarticn-
larly determined use of these practices by IB^NI to prevent Control Data from
making a sale. Examples of such instances are discussed separately below be-
cause they illustrate combined use of some or all of the practices previously
described and. in addition, involved customers to whom a sale would have par-
ticularly great positive influence upon subsequent sales to that and other
customers.

(a) During the fall of 1965. negotiations for lease of a CONTROL DATA
6400 computer system to the Missiles Division of Raytheon were in progress.
IBM was awai*e that a sale to Raytheon would open a door in New England to
Control Data, being the first installation in that area of a CONTROL DATA
6000 series computer. Such a sale would greatly assist Control Data in competi-
tion for future sales to such customers as Harvard University, MIT, Sylvania,
General Dynamics and United Aircraft. IBM. therefore, asked the customei*
what the price difference was between the IBM proposal and Control Data's lower
quotation ,and, after being given the figure, stated that they would provide
enough free rental to equate the price of the two proposals. IR;m assured
Raytheon that it could assume that the cost of IBM equipment would be as low
as the cost of Control Data equipment.

(b) Similarly, in recent competition at Washington State Uuiversitv between
an IBM 360/67 system nnd a rONTROL DATA 6^00 system, a receipt of t^he
order by Control Data would have resulted in installation of the first 6000 series
equipment in the Pacific Northwest. Aware of this, as well as of the considerable
prestige of Washington State computer personnel, IBM offered an educational
discount, gift or consignment of certain peripheral equipment, reduction of
rental on presently leased IBM equipment to 10% of normal rental for at least
a four-month period during conversion to the new equipment, furnishing of ex-



5561

tiaordin.-iry sendee personnel to influence the sale, and, probably, a buyback in
the form of a joint software research iirogram. Washington State decided to
acquire the IBM 360/67.

(c) La^Yrence Radiation Laboratory, operated by the University of California
for the Atomic Energy Commission at Livermore, California is one of Control
Data's major customers. Livermore is also one of the most prestigious computer
users in the world and has for years operated the world's largest known ag-
gregation of computers. In recent competition for sale or lease of a computer
system having the power of the CONTROL DATA 6800 system, IB:M offered a
combination of inducements having the practical effect of reducing the cost to
the customer of a $4,000,000 to $6,000,000 system to approximately $1,000,000. The
inducements included a buy-back by IBM of time on the system and of pro-
gramming services to be performed by the customer ; interim installation of a
360/75 system, the rentals for which would be in part applicable against the
purchase price for the 360/95 system ultimately to be installed ; and some type
of arranj^ement whereby peripheral equipment would be provided fi'ee of charge
for a year. During negotiations IBM also indicated to the customer that it was
considering setting up an IBM research laboratory in or near Livermore,
California.

Ii. "Fighting" Machines

Preliminary investigation suggests that IBM has made use of "fighting"
macliines on various occasions. Where IBM has found gaps or weaknesses in
particular segments of its product line resulting in competitors realizing sig-
nificant sales IBM has responded by introducing equipment specifically designed
to fill the gaps and eliminate the competition, which equipment apparently has
been priced below IBM's usual margin of profit.

Data on IBM's profits on specific pieces of equipment is not available to Con-
trol Data. However, Control Data can infer on the l)asis of price/performance
information that the IBM equipment referred to below was "fighting" equipment
priced below usual profit margins. It is suggested that this area would be a
fruitful one for further investigation.

Information now available indicates at least two IBM computers which are
"fighting" machines, the 7044 and the 360/44.

(a) lOIflf. IBM's first large-scale, solid-state scientific computer was the 7090,
first installed in November, 1959. An average 7090 system sold for $3,048,000 and
leased for $67,400 per month. Suitable only for scientific applications, it had a
Gibson Scientific Mix value of 5.82 microseconds per average instruction, the mix
value being a measure of power of a computer for scientific applications, a lower
value indicating a more powerful computer.

In Jauuai-y 1960, Control Data installed its first 1004 compnter. The 1604 was
about two-thirds the IBM 7090 in performance, having a Gibson Scientific Mix
value of 8.22, and one-half of the 7090 in price, selling for $1,410,000 and leasing
for $42,000 per month. The 1604 achieved significant success, for IBM had no
equipment in this price and performance category.

IBiNI. therefore, introduced the 7044, the first such computer being installed in
June 1963. The introduction of the 7044 delivered a staggering blow to Control
Data and for all practical purposes ended marketing of the CONTROL DATA
1604, l)eing equivalent or superior to the 1604 in perfoi-mance with a Gibson
Scientific Mix value of 5.36 and renting for 25% less than the 1604 or $31,800 per
month. The 7044 sold for $1,957,000. Note that while the mix value of the 7044
and 7090 were essentially equivalent, IBM offered the 7044 at one-half the lease
price of the 7090 and two-thirds the sale price of the 7090.

(1>) 360/U- The IBM 360/44 was not a part of IBM's original April 1964
announcement of the 360 Series. It was not announced until August 1965. There
is reason to believe that it was introduced specificallv to eliminate the competition
of the CONTROL DATA 3200, a machine which was effectively taking advantage
of a gap in the IBM product line. There is also reason to believe that IBM is
realizing a much lower rate of profit on the 360/44 than on most other equipment.
This is suggested by the follo\\ing data on prices and performance of the CON-
TROL DATA 3200, the IBM 360/44. and the IBM 360/40 and 360/50 which were
a part of the original 360 Series announcement and were IBM's primary offerings
to compete with tlie CONTROL DATA 3200 before the 360/44 was announced.
Gibson mix values are in microseconds per average instruction and a lower value
indicates a more ixtwerful computer.



5562



Average system
Lease (per



Mix value



Model Installation date montti) Purchase Scientific Business

IBM:

360/40 April 1965 $12,500 $611,000 22.46 66.23-.

360/44 June 1966 9,800 422,000 4.64

360/50 Augustl965 18,000 883,000 5.71 27.14-

Control Data: 3200. __.. June 1964 11,500 541,000 3.70 22.15



It will be noted that the IBM 360/44 while having a scientific mix value essentially
equivalent to the 360/50 sells and leases for one-half its price. Further, the
CONTROL DATA 3200, priced at somewhat less than the 360/40 and substantially
less than the 360/50 although outperforming both is now faced with a machine,
the 360/44, having substantially the same scientific mix value and leasing and
selling for significantly less.

5. Refusal to Lease to Competitors.

In addition to IBM's predatory pricing policies, TB]M has recently launched a
direct attack against its competitors in their capacity as IBM customers. On
December 11, 1964, in a letter from MeWhirtle, the president of IBM's industrial
products division, to Norris, Control Data's president, IBM announced that effec-
tive that day "IBM equipment which is to he remarketed by other equipment
manufacturers (would) be available on purchase terms only." Previously, equip-
ment manufacturers requiring certain IBM equipment to complete systems to be
leased to their customers could lease the equipment from IBM for sublease to
customers.

For competing computer manufacturers, the IBM equipment that was the .sub-
ject of tliis announcement is of immense importance in assembling and in selling
their own computer systems. The CONTROL DATA 3000 and 6000 Series coin-
puter systems, for example, utilize IBM-manufactured chain printers and data
cells. In terms of percentage of system cost, the IBM equipment wlien used com-
prises ten to fifty percent of the cost of these systems.

The economic effect of the new policy on IBM's competitors is enormously
destructive. By forcing its less-financially advantaged competitors to purchase
such equipment at retail prices, IBM has placed them in an untenable position.
If Control Data makes a heavy capital investment by purchasing such equipment
outright and then attempts to achieve some return on that investment, it is not
competitive wih IBM. If, on the other hand, Control Data passes the eipupment
on at cost to its customers, thus remaining competitive, it is still crippled in its
competitive battles with IBM by having much of its capital tied up in an un-
profitable venture. Finally, if in order to free its capital. Control Data purchases
on a time-payment plan from IBM, it must lease the equipment at a loss to remain
competitive. Control Data is economically unable to comijete whichever course it
takes in response to the new IBM policy.

The effect of the new policy can be highlighted arithmetically. Assume Control
Data has $5,000,000 to utilize in building computer systems. Assume, additionally,
that 20% of each system is IBM equipment, and that each system costs $100,000.
If Control Data had to invest $80,000 in each system, the $20,000 of IBM equip-
ment being leased from IBM, Control Data could produce 62.5 computers. Because
the new IBM policy forces Control Data to invest $100,000 in each system. Control
Data is only able to produce 50 systems. The deleterious effect of the policy on
Control Data's ability to compete \vith IBM is obvious.

The economic effect cannot be avoided. Thei*e are presently no alternative
sources of supply for comparable equipment. Further, most customers are un-
willing to lease such equipment directly from IBAI for use with a Control Data
system, since that course of action would divide responsibility for system main-
tenance and would create problems of on-.site integration of the system.

IBM. in its letter of December 11, 1964, claimed that the new policy was neces-
sary to protect it "from the risk of large-scale rental discontinuances." This is
simply not the case and. in fact, is only a gloss designed to obscure IBM's
predatory purposes. First, that same risk is run with every rental customer —
only those customers who are also competitors are singled out for disadvan-
tageous treatment. Second, since IBM's competitor-customers do not manufacture



5563

the equipment they leuse from IBM, they will not iu fact discontinue the leases.
The only reason that exists for the new policy is to make it materially more
difficult for IBM's competitoi's to effectively compete with it.

It has been suggested that the new IBM policy is necessary because com-
petitors leasing from it have often entered orders for quantities of equipment
in excess of their needs with the intent of cancelling these excess orders before
delivery. If IBM is unable to protect itself from such conduct, it is argued. IBM
will find itself vastly overstocked in certain e<iuipment. Of course, if such excess
ordering does exist, IBM ought to be able to protect itself. But such protection
does not require and cannot take the form of an exclusionary device. IBISI"s
interests could be equally well secured by a penalty clause for cancellation before
delivery or by a mandatory one-year lease or by some other coutractural pro-
vision. It is unnecessary for this purpose that IBM's competitors be required
to make a large capital investment in IBISI equipment. IBM's monopolistic
practice does not become any less violative because a business reason can be
advanced to support it wlien that business reason can be satisfied in other non-
monopolistic ways.

C. Miscellaneous unfair trade practices

This section of the Report deals with miseellanenus inilav.-ful practices en-
gaged in by IBM having the effect and intent of preserving its monopoly position
and preventing competition in the computer market by Control Data and other
manufacturers. Some of the practices such as disparagement of Control Data
and coercion of customers are believed to result from IBM salesmen's desperate
fear of losing any business. IBM has created an atmosphere which encourages
its salesmen to go to any length to obtain or retain computer customers. The
pressure from above stems partly from IBM's sales commission policy which
reportedly penalizes a salesman for losing an existing customer by requiring
that he repay the original commission on the applicable machine, whether or
not he was responsible for the original sale or lease. IBM salesmen are further
subjected to strict accountability and, at times, punitive action by their superiors
in any case of a lost account or lost order. In several instances, it has been noted
that an IBM salesman has Iteen taken off an account or reassigned to other
territory following loss of an account. The actions of IBIM salesmen in the market
place constitute policy in action and these, rather than any carefully drawn
written policy, are the measure of legality of IBM's conduct.

The trade practices described below are submitted in response to requests
from the Antitrust Division that Control Data report conditions encountered
in the market place. In many instances they also serve to characterize and con-
firm the anticompetitive activity and predatory intent described previously.
Viewed collectively, all of the activities complete, we submit, a pattern of attemi)t
to monopolize and abuse of monopoly power by IB]M. The practices took place
primarily in 1964 and 1965, after the .July 1963 announcement by Control Data
of its 6000 Series of computers and coincidental with various' "Stop Control
Data" statements by IBM management.

i. Disparagement of Control Data

At several accounts during 1965 competition with IBM, Control Data lias
learned that IBM sales personnel have made false representations to prospective
customei-s, including the following:

(a) Repeatedly advising a certain midwest industrial prospect that it
would be risky to deal with Control Data because the company was on
"shaky financial grounds." Representations of similar tenor were also re-
portedly made to another midwestern customer.

(b) Representing to a large west coast industrial prospect following com-
pletion of performance tests by Control Data, that the software proposed by
Control Data would not perform as proposed, either stating directlv or by
implementation that Control Data had cheated on the tests or lied" to the
customer. IBai representatives persuaded the customer to require a re-
testing in the presence of witnesses. Re-testing was in fact concluded satis-
factorily and in the presence of witnesses.

(c) Representing to a southern university and to a midwest chemical
company to the effect that Control Data's computers required excessive
maintenance, with the obvious implication that Control Data's hardware
was inferior.



5504

(d) Representing to an investor through a top level IB'Sl executive that
IBM would force Control Data into a weak position in the computer market,
the investor, being exhorted to sell any Control Data stock he might own :
further, representing to the investor that major proposals including the
University of Texas, which involved competition with Control Data, would
be personally reviewed by IBM's top management and that IBM did not in-
tend to lose* any more of the university market and was, accordingly, re-
turning to a 60% discount to educational institutions. The obvious effect of
these representations to a single investor was to open the door to widespread
repetition through investment grapevines.

2. Intimidation and Coercion of Customers

(a) Customer Procurement Personnel. IBM's practices in this area must be
considered against the following background. IBM's monopoly has created
customer resistance to changing brands of computers. Nearly all prospects for
Control Data computers are existing IBM customers and "status quo inertia" is
very widespread. Only the very highly technically qualified user personnel and
purchasing agents are able to make comparative evaluations between IBM com-
puters and other computers. These people are scarce and are found primarily
at such agencies as the Atomic Energy Commission, the Department of Defense
and large universities. Absent qualifications to critically compare competing
equipment, the tendancy is to purchase IBM because that is the best known
brand. Such a decision need not be justified to superiors, whereas purchase of a
non-IBM computer may require extensive justification. Because of this predispo-
sition to question any decision against IBM, customer personnel responsible for
deciding upon computer acquistiious are peculiarly vulnerable to intimidiation
by ovei'-aggressive salesmen.

' This vulnerability may be intensified in the case of lower eclielon customer
procurement personnel because their superio-rs may well be former IBM em-
ployees. A substantial portion of computer customers rely upon employees or
independent consultants who are IBM alumni for advice and decisions on com-
puter acquisitions.

One of the most flagrant examples of intimidation by an IBM salesman oc-
curred in recent competition between an IBM 360/30 and a CONTROL DATA
3100 at a west coast local government agency. The customer's computer manager
advised the Control Data salesman that he was fearful of losing his job. He had
recommended the Control Data machine and believes that the IBM representa-
tive then went to his superior and stated that he was incompetent and should be
replaced. According to the computer manager, his predecessor had recommended
acquisition of a computer other than IBM and had forthwith been fired.

(b) Other Coercion. Another IBM practice has been described as a "burning
the match offer". It is exemplified by competition last September between IBM
and Control Data at a California industrial prospect. The customer notified IBM
that it was going to replace its IBM Model 7044 computer with a Control Data
conqniter as of a date certain. Due to problems in negotiating all of the contract
details with Control Data, the customer later advised IBM that it would like to
continue renting the 7044 several months past the specified change-over date.
IBM's response was that unless the customer contracted for an IBM 360 vSeries
computer within 48 hours, no extension could be granted as the Model 7044 had
lieen sold to a third party. A contract for a 360 computer would resolve the prolv
lem, however, and a Model 7044 would be made available to the customer on an
interim basis.

Similar coercion was exerted upon a major midwestern medical account by an
IBM vice president. When the customer considered procuring a Control Data
computer to supplement or replace exis-ting IBM machines, the vice president ad-
vised the customer that unless IBM equipment were ordered a team of some 27
IBINI engineers would be withdrawn. This team had been working with the
customer, free of charge, conducting advanced systems development research for
the mutual advantage of IBM and the customer.

S. Reciprocity

IBM has engaged in various forms of reciprocity with its customers and
sujipliers.

One such tactic involves the investment of large sums of money for new fac-
tories or assembly plants at or near the customer's place of business. Several such



5565

announcements have been timed to coincide with fierce competitive efforts be-
tween IBM and Control Data. The obvious economic advantage to the customer
and the entire surrounding area is a persuasive sales implement in this situation.

An example of this occurred recently in the southern part of the United States,
where IBM reportedly has leased land for its new plant from its customer and will
buy back time from the customer on a yet-to-be-installed 360/67 time sharing
computer system. One of the architects of the overall marketing arrangement is
a former employee of IBM often referred to as the "father of the 360 Series.'"
This IBM alumnus is presently an employee of the customer but reportedly is still
receiving retainer compensation from IBM.

Another example occurred during a competitive situation in a section of the
western United States. IBM announced plans to build a large manufacturing
plan there within days after announcement of a single sale by Control Data in
this geographical area, which has large potential for additional computer sales.
This announcement was made by Mr. Thomas J. Watson, Jr., coincidentally with
a dinner meeting attended by the area's most influential computer prospects.

A number of IBM gifts of buildings to accommodate university computer cen-
ters offer additional inferences of the type of reciprocity.

Incomplete investigation also has revealed a number of instances where de-
posits of funds in the million dollar range have been made by IBM in certain
banks with no legitimate business reason. Control Data does not have the facts
as to the coincidence of these deposits and competition for computer procurement
at these banks. We believe, however, that a Civil Investigative Demand would
disclose a number of instances of implied, if not contractual, reciprocity between
these banks and IBM.

Supplier reciprocity is suggested in the relationship between Fairchild Camera
and Instrument Corporation and IBM. The chairman of Fairchild's executive



Online LibraryUnited States. Congress. Senate. Committee on theThe Industrial reorganization act. Hearings, Ninety-third Congress, first session [-Ninety-fourth Congress, first session], on S. 1167 (Volume pt. 7) → online text (page 80 of 140)