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 61 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 61 of 140)
Font size
QR-code for this ebook

To further confuse the marketplace for users and competitors alike. IBM
announced a series of hardware and software products in the same time frame.
All impacted the interactive market, and were technically vague, with varying
•delivery schedules.

IBM uses the timing of product announcements and interface disclosure to
achieve two goals. First, to obtain a two-to-three-year lead in the marketplace,
therel)y making their product life at least that much longer than their competi-
tors. This time leverage in a lease business is tlie difference between significant
profit and, at best, marginal profit for any single product life cycle. Second, to
chill the market for competitors' products by putting its competitors in the
position of having to attempt to persuade customers that they can follow IBM
without the specific technical knowledge to explain how.

The tactics described in this section are within the scope of IBM's stated
policy. Under the guise of allegedly advancing technology, IB^I is able to increase
market share without a superior product. If allowed to nurture this strategy
in the communications area, the results will be disastrous for competition. The
market will be cornered.



Given the history of IBM's predatory marketing practices, it is possible to
postulate some of the effects on data processing and communications if IBM
is allowed to carry over its past practices into the new and larger world of
satellite communications, without appropriate safeguards for the public interest.
This situation would be in contrast to current common carriers Avhose systems
are transparent or neutral in relation to the devices attached to them, and whose
systems simply carry information without processing it at any point.

IBM's data processing activities would be favored, especially in a distributed
processing mode, by discriminatory services, cross-subsidation, and improper
pricing of common carrier/computer services. It might be possible under rlie
FCC's hybrid service rules (47 C.F.R. § 64.702) to offer a bundled package. IBM
could provide equipment designed to optimally interact with the communications
system because only IBM would be in a position to totally dictate the relation of
the systems. Installation and maintenance performance for IBM EDP customers
as well as more timely response to initial installations and requests for outage
corrections, could occur. IBM has been known to switch costs for one device to
another. If this practice is carried into the communications realm, an even
greater undue comi>etitive advantage will occur.

IBM has every motivation to influence the design of the satellite network in
such a way that it can be utilized for IBM's massive return to the computer serv-
ices market. A satellite communications system can be designed to favor a partic-
idar approach to nationwide computer services. IBM is presently restrainetl from
operating so-called service bureaus, but that restraint can be obviated by render-
ing ineffective the current service bureau concept.

IBM has already concentrated on distinct industry data processing applications,
such as insurance, transportation, retail point-of-sale, and tlie like, with some of
its software and product lines. These industry groupings represent .service net-
works already in place, merely waiting for IBM to "field enable"' the networks
components so that they may optionally utilize the new communications offering.
The erection of interrelated data bases within these industries is all that is neces-
sary to sound the death knell for today's service bureaus.

On the surface it would appear that the entry of IB'Sl into the communica-
tions business might )>e beneficial because to some IBIM presents the first tech-
nically competent, well-financed organization to compete with the Bell system.
However, the confrontation of these two giants will ultimately lead to a division
of the market along non-competing service lines. The al)atemeut of competition
will make losers of the customers and the public. Smaller manufacturing and
service organizations presently endeavoring to do business in this field will be
severely damaged in the early struggle, with bleak prospects once the division is

AT&T has a huge installed plant for communications. However, most of its
capacity is geared for slow speed, low bandwidth communications. The new in-
formation handling market will bo most efficiently and economically serviced by
high speed, high bandwidth communications : IB]M will make that market, by
forcing its business and data processing product lines. The growth markets of
distributed data base management, facsimile transmission, and many other serv-
ices, all demand wide transmission channels.

Bell controls the current local loop plants to an overwhelming degree, and
would continue to. However, with several hundred relatively inexpensive small
earth receiving stations, most of the integrated business system market can
be blanketed without the requirement of using established common carrier
local loops. The technology has been proven to allow "local distribution" through
the combination of small earth stations, dedicated cable or microwave, and
local internal wiring on equal or better price/performance with the current
local telephone system plant. IBM can thus neatly by-pass Bell. Incidentally,
by offering this service with its own local drop, and by offering an automated
voice/data PBX which is being actively marketed in Europe and which is in-
stalled in this country on an "experimental" basis, IBM could eliminate much
of the current local business traffic. Bell and otherwise.

AT&T is further restrained by virtue of its own domestic satellite licenses.
As a condition of allowing Bell to participate (planned now in partnership with
GT&E), the FCC proscribed data and other "specialized" traffic for three years
from onset of operation on the AT&T domestic satellite. Besides, under its
own 1950 Consent Order and under FCC rulc!. Bell is pri^hibited from offf'-iug
data processing services or equipment not available to support a tariffed com-


municatioDS service. Will the AT&T Consent Decree have to be reopened to
accommodate this latest turn?

AT&T lacks equipment to combat IBM. Its one "data processing" product en-
try — the Dataspeed 40® terminal — was almost instantaneously eclipsed by this
July 3rd announcement. Bell will have nothing to answer the variety of ap-
plications-oriented terminals and distributed processing systems IBM has al-
ready started announcing, such as the 3790.

Neither, imfortunately, will the independent terminal manufacturers, but
for a different reason. For some time now, IBM has been unilaterally develop-
ing telecommunications software and transmission protocols. The satellite sys-
tem, in whole or in part, will utilize these software products. None of the compe-
tition will be in a position to have a product line that will efficiently interact with
the communications system. The independent interactive display terminal manu-
facturers liave already seen the anticompetitive effects of such tactics in IBM's
manipulation of terminal device support. If this policy is expanded by IBM to
the cominuiiications media also, the effect on independenfs will be devastating.

IBM has maintained a bundled pricing policy whereby it quotes a single price
for hardware and software and related support.

They could thiis offer an integrated business systems package. For the past
IS months IBM's statements have increasingly emphasized "systems", aitpar-
ently in an attempt to incorporate communications as a part of their business, as
well as to justify the difficulty of identifying the interface to allow intereonnec-

The IBM/CML statement that they 'have no plans for public-switched net-
work services" is meaningless. For example, there is absolutely nothing of a
regulatory nature that precludes a domestic satellite carrier from offering
switched data services. Therefore, CML and its spokesmen can represent that
switched network services are not planned and at the same time they can be
actively developing a switched network in concert with their owners, IB:M and
COMSAT. Clearly this type of activity gives great advantage to IBM and places
its competitors at an insurmountable disadvantage.

To summarize. IBM will have the advantage of timing as it relates to prod-
uct announcements and changes, technical interface criteria, and communica-
tions costs. These advantages, when added to IBM's already overwhelming hard-
ware and software advantages will give IBM far greater market control over
the computer/communications business than it can exercise today. Those charged
with controlling commerce must understand the ramifications of adding the
dimension of communications to IBM's repertoire.


The IBM/CML acquisition should be denied because the potential for market
foreclosure and anticompetitive maneuvering is far too certain to balance any
potential positive effects.

In the event those in Government in a position to stop this merger do not see
fit to ffillow the recommendation above, consideration must be given to a drastic
resiructuring of IBM and the adoption of ironclad conditions precedent to its
entering the communications tield. These conditions will insure the interconnec-
tion of all competitive offerings — data processing or communications. Complete
divestiture is warranted to dissipate the awesome power that has been accrued.
Sul'sidiary relationships will not provide adequate protection.

V>'hen IBM's market power was localized in the traditional data processing
industry, several recommendations were made to combat its predatory practices
and reduce its overwhelming market power. Representatives of some of the com-
peting manufacturers of data processing equipment presented views to the De-
partment of .Tustice on just how this should be done. Industry commentators
have put forward their suggestions. Some of the thoughts included l)reaking the
IBM structure into many separate corporations, unbundling of products and
software, placing judicial restraint on some predatory marketing practices and
making mandatory certain interface standards for peripherals and terminals.
These suggestions must now be re-examined in the light of this latest turn of

The final plan must create an environment of transparent or neutral inter-
connection between all hardware, software and communications elements, so
that there can be true competition at each workable level. Vertical integration

I Registered TM, AT&T.


should be eliminated to reduce the kind of evils uncovered in the Western
Electric-Bell relationship. The competitive procurement requirement for domes-
tic satellite licenses must be reinstated.

This is the crucial moment for such decisions. Action must be taken prior
to the establishment of IBM's new beachhead. Only in this fashion Avill the public
interest be safeguarded. This is not a call for a Utopian solution, but is a state-
ment of the plain truth that to enter this new era with out realization of the
factors stated herein, and without means to contain the harms identified, will
forever stifle competition in the total information handling market. The de-
cisions will be difficult and implementation of plans even more so, but there
is no substitute for coming to grips with the situation now.

Senator Hart. The committee will be in order.

Our concluding witness today is the president of the Computer
Industry Association, Mr. Dan L, McGurk. Mr. McGurk?


Mv. McGurk. Thank you, Senator Hart. I won't go into any bio-
graphical information because I previously submitted that with my
testimony of last year.

I appreciate the opportunity to again appear before the subcom-
mittee. Before I go into my testimony I'd like to set the record straight
on just one point.

It is not a principal objective of my employment to make public
attacks on IBM. Unlike Mr. Katzenbach, who is paid to defend the
so-called International Business Machine Co., I'm not employed by

I serve as president of the CIA without pay. You've already heard
testimony in this set of hearings that describe the structure of the
marketplace and some of the problems for users and competitors alike
that are created by the dominance of one firm in this important

I would like today to indicate from a practical and parochial view-
point what the effects of such market dominance are to a competitor in
the industry.

I'd also like to point out some of the problems inherent in our
current legal system and indicate some criteria necessary to solve this
particular problem and make a modest contribution to the task of
devising more appropriate legislation.

There are two imj^ortant constraints on the way in which a smaller
competitor must operate in the data processing industry today.

In the first place, he must adjust his total business strategy to
take account of the fact that he must compote in an environment which
has been structured by the dominance of IBM.

Tliis entails his acceptance of a set of conditions within which he
must opei'ate if he is to have a successful business.

In addition to that, however, a small competitor finds it necessary
to make realtively rapid and sometimes major changes in his business
strategy when IBM decides that the environment is due for a change.

In other words, not only is the competitive environment a difficult
one but that environment is subject to change without notice when
IBlNPs corporate objectives are not being met within that environment.

This makes the business strategy of a smaller competitor extremely
difficult, and the number of failures indicates that few are successful.


As I'm sure you have become aware during the course of these
hearings that the computer industry is an extremely complex and
diverse one.

Data is hard to acquire and interpretation is not simple. As an il-
lustration of that, I'd also like to set the record clear on Mr. Chum-
bris' questions of yesterday concerning some of the material submitted
by my colleague, Mr. Biddle.

He pointed out that in a chart that was submitted out of Judge
Larson's decree that IBM showed a world market share in lOSG of
42 ]3ercent and Sperry Rand 51 percent.

What we failed to notice was — and this is an illustration of the diffi-
cult v of understanding this business — that was a chart of total revenues
in 1956.

If you recall at that time, IBM did not sell any of their equipment.
So for the IBjNI i-evenues that was pure lease revenue — rental revenue.

For Sperry Rand it was almost pure sales revenue; for in the next
paragraph of that statement by Judge Larson it says in 1956 IBM
shipped 85 percent of all new business. At the end of 1956, IBM had
75 percent of all ADP systems outstanding.

So. you see, the facts are there but the interpretation and under-
standing of them is tricky. In fact, IBM never had as low a market
share of 42 percent except on a revenue basis for that year — which
does not represent the real market power.

The large marketplace is characterized by a very definite customer
demand for rental rather than purchase of this im})ortant capital

xVltliough there are specialized marketplaces where purchase is the
more likely means of equipment acquisition, the vast bulk of users is
accustomed by a long period of IBM market-control to leasing their
equipment on a month-to-month basis.

What this means for a small company attempting to grow was shown
by Mr. Collins earlier today. A second major environmental con-
straint to a small competitor in this industry has to do with the enor-
mous weight of software which overhangs the marketplace.

By avoiding significant software standardization IBM makes sure
that its enormous resources devoted to creating new software make it
impossible for a competitor to match it in the amount of software

I remember particularly a point in time at Scientific Data Systems,
where I used to be emplo3'ed, a small general purpose computer manu-
facturer, when we had successfully been in business for about 5 3'ears.

Animal revenues were between $50 and $100 million. We were very
proud of having accumulated a vast store of software which we esti-
mated had cost almost $30 million.

Then we found out that IB]M had invested in its 360 operating
system alone over a billion dollars. Even assuming that we were twice
as efficient in software creation — ^^-hich we Avere sure we were — that's
a very high barrier to cross when convincing a customer that vour
oft'ering is a viable alternative.

Xew firms in many industries are able to compete because of signi-
ficant technological innovation which creates a market for a newly
designed product.


Because of IBM's enormous market power in the industry such
innovation must be limited to doing just what IBM does in some better
or new way.

A new wddget to improve the data processing function cannot have
a broad market acceptance until IBM legitimizes it by coming out
with a comparable widget itself.

Until that happens, any inventor of such a widget has to have liis
marketing force spend most of its time explaining why the widget is
good; why IBM doesn't have it; and whether or not it will make a
system encompassing it noncompatible with the standard of the indus-
try— that is, IB]M.

This necessity to match IB]M on a functional basis is one reason
that there has developed an industry of plug compatible peripheral
devices and, recently, even central processing units.

Another fact of "life for a smaller competitor is that IB^Nf. hy its
size, has the ability to, and in fact does, completely blanket tlie world-
Avide computer marketplace with its sales force.

This doesn't only mean that more opportunities are available to
IB:M, but it also means that many large corporations with offices or
plants around the country and the world have a strong predisposition
to go to IBM to make "sure that if they adopt one manufacturer
throughout their company they can be assured of sales and service at
all their locations.

In order to combat this problem, marketing strategies of small com-
petitors must devise a way to focus their efforts by geographical or
industry application in order to be able to oifer anything approaching
such total coverage.

The net result of the structure of the environment described here
is that a small competitor has to locate either a market where IBM
has not focused attention — normally because it's too small to be of
any great interest — or to live with the environment structured ])y the
dominant company.

Most firms attempting to grow in the computer industry have started
out with the first strategy, and then in order to expand their market
have been forced to live "with a difficult environment created l)y tlie
computer industry.

But it doesn't "end there. Having developed the strategy which
miijht expect a snuill competitor to exist and grow, he has to accept
the" fact that the environment can be changed unilaterally l)y the
whim of IBM.

It matters very little if that whim comes about because the company
or some group of companies have been successful and, therefore, de-
liberate action is taken to limit their growth; or whether some other
cause creates a corporate need for IBM to change its practices in the

In either case disaster can result. If a mouse has learned to live in
a stall with an elephant, whether the elephant steps on the mouse in-
tentionally or not it hurts just as mucli.

One of the solutions to the leasing problem devised in the late
1060's came about through the creation of a number of independent
leasing companies that have been discussed earlier.

Based on the historical relationships between IBM sales prices and
rental prices this appeared to be a relatively viable business.


On the introduction of new machines, however, and after the leasing
companies ha (I begun to impact IBM's plans, IBM drastically changed
the relationship between purchase prices and rental prices, thus mak-
ing tlie leasing industry relatively uneconomic.

Shortly thereafter IBM also introduced long-term leasing plans
which had previously been offered by their competitors.

Had they been constrained by the consent decree of 1956 from of-
fering such plans until 1966, but still not offering them for 3 years,
it wasn't until 1969 that such an offering was made.

Tliis did ci-eate a new opportunity for the leasing companies —
long-term full-payout leases — but it did at the same time destroy a
small competitive advantage which the competitors of IBM had been
able to create, given the previous environment.

In fact, IBM long retained the price umbrella for all of their com-
petitors, carrying out a policy of never reducing the prices of cur-
rentl}' available products.

Although tliis policy still appears to hold in theory, starting in 1969
and 1970, IBM introduced new products with new product numbers
that turned out to be technically indistinguishable from older prod-
ucts, except for the changed model number designation and a sharply
lower price.

This is one of the pseudoinnovations which Mr. Biddle referred to
yesterday. Another such innovation, while I'm on that subject that I
would call a pseudo-one, is the changing interface between the periph-
eral units and the central processor.

The integrated file adapter and the integrated storage control were
introduced by IB]M as technological advances.

What they really were, were moving the controller into the main
frame and out of the way of the peripheral plug-compatible market.

IBM announced m 196J: a compatible set of systems which had a
cominon interface to peripheral units. And that was the basis in which
the plug-compatible market was established.

But these changing interfaces made that standard, which had been
introduced and announced by IBM, out of date, and therefore caused
great havoc among this new industry which, as somebody earlier men-
tioned, was the first new real competition to IBM in a number of years.

One can go on : Changing specifications for media ; stopping support
of previously supported software ; eliminating software interfaces in
new software releases.

However, it seems to me that this whole area can best be proven by
the following tale: The Computer Industry Association has identified
a ]X)tential problem in the future in the area of data security.

It is clear that it will be necessaiy and important to provide means
to users to prevent unauthorized access to their data files in order to
provide privacy and confidentiality.

The CIA has suggested tliat one dangerous possibility in that regard
is the creation by IBM of a hardware-software privacy lock which
prevents other manufacturers from providing access to, or compatibil-
ity witli. IBM systems.

We talked about this theory in many forms including National
Bureau of Standards seminars, National Computer Conference panels,
and so forth.


Although there has been some argument about our fear and about
our concept, the interesting thing is that there's never been any com-
ment that IBM wouldn't or couldn't do that — that it was in their best

In other words, industry leaders are so used to the dominance of
IBM in creating and changing the environment in which they operate
that they don't question that IBM could and would manipulate it in
that sort of way.

The point is that the rules of the game are difficult. But if you start
to win they change.

I'd like now to turn to the question of whether or not our current
laws and legal processes can solve the problem of IBM's monopoly
power and dominance as outlined both in my paper and other evidence
presented to your subcommittee.

A look at the history and status of the Government's antitrust action
against IBM could be instructive in that regard.

The Justice Department has sued IBM for Sherman II violations in
1932, 1952, and 1969. The latest effort, 5i^ years old, was due to go to
trial in 3 months.

I understand that this morning that was continued until next year.
Many ol)servers believed that the schedule would not be made — they
were rigJit.

Even if that schedule were to have been met and IBM were to be
held in violation of the Sherman Act it would be many years — up to 15

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 61 of 140)