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 86 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 86 of 140)
Font size
QR-code for this ebook


substantially lower real cost than its competitors — if there are econo-
mies of large scale — then increasing competition by eliminating IBM's
dominance would merely raise IBM's costs, not lower prices to the pub-
lic, and at least in regard to monopoly profits I do not see that it would
be aeliieving any major benefits. So one does want to find out where
IBM's monopoly profits come from, and how they relate to the factors
that make for its dominance.

One possible factor is IBM's ability to sell the same product, but at
a higlier price than its competitors. The record on that is very unclear.
Some statistical work done showed that in the mid-19G0's IBM indeed
was able to sell its computers at a higher price than competitors could
sell equivalent equipment. More recently, other statistical work has not
been able to find that relationship. It is very difficult to do it because it is
so difficult to measure computer products.

I think the most likely result of structural change in the computer
industry— and one that I think is highly desirable — will be an ira-
])rovement in product performance in the industry. There w^ill be a
decrease in product diftei'entiation, and perhaps a decrease in market-
ing activity and in the kinds of support that are provided by the
computer makers to the users. I think that there will be greater modu-
larity in computer parts, that diflerent kinds of hardware will be more
compatible with each other so it will not be necessary to buy complete
computer systems all from the same manufacturer in order to get
things that will work together.

I think the expertise in putting computer systems together will still
be needed, but it will be ti-ansferred or it will move from the computer
manufacturers, where it is part of their marketing effort, to the users
of computers who will be in a ])Osition, perhaps with expert advice
again from a reasonably competitive industry of data processing ex-
perts to put together complete computer systems from the offerings of a
variet}' of different manufacturers — some of whom produce a relatively
long product line, others of wdiom are specialists in products w^here
they have particular expertise.

I think this kind of structural change will also lead to improved
terlmological performance witli firms focusing on the development of
improved ''components'' rather than improved systems. This is very
nnicli like the debate that has gone on, particularly in connection with
weapons systems development — not computer parts, but simply weap-
ons systems in general- — whether one should be working on compo-
nents or systems. I am one of those who believes on the basis of a very
imperfect record that the components approach is better than a system's
approach.

I think also if the users are forced to put their own systems together
and learn more about them, they will do a much betfer job of using
their computers. There liave been fiascos of computers having been
oversold and not }:)erforming as they should have. There have been
])roblems of computer fraud that I think have arisen from the fact tliat
maiiagements just do not understand what is going on in the com]mter
shop. They do not understand the way in which it affects tlieir business
and, as a result, they are not able to control it. Tlie computer companies,
for other reasons, are not interested in controlling on that same level
fraud in the use of computers. I think these problems may be dealt with
by structural change in the computer industry.



5604

I mentioned a couple of other advantages of structural change in the
computer industry in my prepared statement, I think also, though I'd
like to go on now in the interests of finishing as quickly as possible, to
the question of how structural reorganization of the computer industry
might be achieved.

I think the first point, that is fairly obvious but important, is that
IBM is the only firm that needs to be reorganized. Even after we allow
for the recent changes in the market, with Honeywell and GE having
been put in the same box, and with Univac having picked up much of
RCA's share, IBM is still at least six or seven times as large as the
second largest firm in the industry. These others have market shares
under 10 percent. If we could cut IBM into pieces that were no bigger
than tluit the industr}^ would pass the test of having a four-firm con-
centration ratio under 50 percent as specified in the proposed act.

I think also that if one is reorganizing IBM, one would begin with
the obvious point of removing the noncomputer l)usinesses from IBM —
its typewriter business, its office products business in general — ^wliich,
incidentally, is also reputed to be highly profitable — its information
records business, the Federal systems division which does mostly special
purpose computer work.

The interesting questions come when we start looking at what may
be broadly classed as the computer industry, whether we can cut such
things as the small information processing systems— like System 3 — or
the peripheral equipment or what-have-you away from the rest of
IBM's computer business. As far as small information systems are con-
cerned, I think we can cut them away. They are organized presently as a
separate division from, the main computer business of IBM, and I
cannot see any reason why they should not be separated.

Peripheral equipment and, later, components are a different prob-
lem. On the question of integration within the computer industry, I
think that questions like these should be left to the determination of
market forces. In this, I shall be disagreeing witli some of what you
have heard, I believe, this morning and perhaps at other times in the
hearing. If the computer industry can be set up with a competitive
horizontal structure so that no one firm or no few firms dominate the
production and marketing of central processing units, I think that the
market is the proper place to determine whether there are or are not
important economies in joining with the central processing business
sucli things as peripheral equipment, component software, or what-
have-you. If there are economies to be obtained by having the same
firm produce all of these things in an integrated fashion, that is what
the market solution will be.

If there are no such economies, then independent components manu-
facturers and independent manufacturers of peripheral equipment will
survive and thrive.

My main point is that if tlie computer industry is broken up so that
tliere is no longer any dominance of the central pi'ocessing business,
there is no longer any need to try to insulate that business fi'om the
otlier aspects of the computer market to prevent the spread of monop-
oly- power. There is nothing to spread. There is no need to go into the
question of Avhether there are or are not economies in these areas.

I think the same comments apply in particular to systems software.
I have already said that in regard to other software, I believe that if
the computer industry is restructured with substantial horizontal dis-



5605

solution that there wiW be less product differentiation, in particular in
tlie softAvare area and to some extent in the peripheral areas, and there
will be less integration because specialists will turn out to be better
than some of the systems companies at making certain kinds of
peripheral equipment. The systems companies, being in a competitive
market, will be forced to look for good products to complement their
central processors and will take advantage of this expertise in other
companies, and so on.

As for a dissolution plan. I would say that I believe that dissolution
of IBM horizontally is feasible. Even as long ago as 1968 there were
some 20 computer manufacturing plants and !?> laboratories in IBM
and IBM World Trade combined. Even if one looks at only the domes-
tic market, I believe the horizontal dissolution is feasible. I do not see
tliat it is necessary that all the successors start out on an equal footing.
The purpose of this operation is not to be fair to some handful of suc-
cessors of IB]M. The purpose of this structural reorganization is to
create a competitive structure for the industry ; and if in that competi-
tive structure, it turns out that for reasons of bad manag-ement or
whatever, one of the successors is not viable, I do not see that as neces-
sarily being a loss to the economy or the public.

Obviously, one does not want to put a lot of assets into an operation
that has knowingly been set up so that it cannot survive, because that
will cause economic losses. But I do not see concern with viability and
fairness as being dominant considerations in the way that perhaps
underlies some other views of what is needed in the way of horizontal
dissolution.

I think these mattei's are covered in more detail in the prepared
statement. I'd like to close with one appeal on the question of structural
reorganization.

In 1956 the Justice Department settled by consent the suit that it
had filed in 1952 against IB]\I, which Avas concerned primarily with,
the tabulating industry — the tabulating equipment industry. That set-
tlement did not involve structural reorganization. It did involve other
remedies that were presumably, in the opinion of the Department at
tliat time, satisfactory.

I think the historical record, and the mere fact that we are sitting
here talking about IBM's dominance in the computer industry, is
fairly persuasive evidence that that particular consent decree diet not
do w^hat it should have done. ^Ma^be it did the job it was intended to
do — namely to prevent domination of the tabulating industry — but it
did not do the job that in retrospect it should have done; namely, pre-
vent IB]M from dominating the computer industry.

And again, if we are concerned with what will happen in the future
we should be concerned not with a structural reorganization that will
take care of the problems that we are able to understand now as best we
can comprehend, but rather a solution that will increase competition
and not provide a base for IB]M to move into the future with whatever
it brings, and somehow transform the advantages it now possesses
into something that will enable it to dominate whatever the computer
industry will turn into over the next decades.

Thank you for the opportunity to make this statement.

[Mr. Miller's prepared statement appears as exhibit 1 at the end of
his oral testimony.]



5606

Senator Hart. Thank you very much.

Mr. Nash?

iSIr. Nash. Mr. Miller, in your statement, you discuss the possibility
of restructuring IBM's facilities. You indicate that if we include
IBM's foreign facilities in world trade, we should be able to create a
half dozen or so successful companies.

If I understand your statement correcth', you have six or so fully
integrated computer systems, each making CPU system software,
peripheral equipment application software; is that right?

]Mr. Miller. Certainly in regard to the first three — CPU's peripheral
equipment, and system software — they would be fully integrated. Since
I do not contemplate any particular provision of the reorganization
plan that would prevent them from making applications software, I
assume they would be in that business.

On the other hand, I should comment — as I think you said correctly
they would be integrated — that it does not mean that the product
lines would all necessarily all be as long as IBM's present product
line. The new product lines would be as long as is typical of the other
firms in the industry. Those ai-e product lines that seem to be long
enough for the companies to survive, if not make profits. Indeed,
recently, as I see from other testimony — I believe yesterday — some of
them are even beginning to earn reasonable rates of return on their
relatively short product lines.

Mr. Nash. Given your comments about the inadequacy of the 56
percent, is it your opinion that effective relief or effective restoration
of competition through the computer industry requires eventual
divestiture ?

]Mr. Miller, Yes; but I want to emphasize that I am not talking
about effective relief, but simply about the purposes of the Industrial
Reorganization Act. I do support the proposition that, quite apart
from questions of the antitrust laws as they presently stand, structural
reorganization of the computer industry as contemplated by the act
is necessary to restore what I consider to be effective competition, in an
economic sense, to the industry.

jNIr. Nash. Now, accepting such a horizontal divestiture for the
moment, let me ask two questions.

In your judgment how would the existing manufacturei's fare under
such a program; say, for example, Honeywell, CDC, and so on?

Mr. Miller. I think they would as a group probably fare quite well.
I think, of course, that if the computer industry becomes more com-
petitive, then all of the manufacturers, including the ones that are
presently small rivals to IBjNI, will be pushed a little bit further by
market forces; and if they are pushed somewhat harder by market
forces, some of them may decide to retreat away from the heart of
the computer industry and enter more specialized areas ; some of them
may simply not be able to hack it.

As as group I do not see any problem. I think, on the othei' hand,
that if IBM were cut into a smaller number of pieces than half a
dozen — if IBM were cut up, say, into three pieces, which, of couree,
■would not be enough pieces to make the industry qualify as not being
a monopoly under the proposed act — it is possible that in the situa-
tion the other manufacturers would be worse ofi' than they are com-
peting against the single IBM.



5607

Mr. Nasii. But you believe that if six entities were formed the mar-
ket power would be sufficiently dissipated so that existing companies
could eif ectively compete ?

Mr. Miller. I said about half a dozen. If one wants to press me on
whether that really means six as opposed to seven, I think the answer
would be that my information is probably a little bit too out of date.
You would have to tell me whether IBM's present market share is
really 70, 75, or 65 percent at this time, and whether Honeywell and
Univac are really 9 and 8 percent, or whether they are not.

Mr. Nash. So. your No. 6 is an order of magnitude?

]Mr. Miller. I think my No. G is something like minimum, and I
would certainly, from what I know about the industry, say that eight
is big enough, as long as they are approximately equal in size.

In other words, eight is certainly big enough ; six, I thinlv, is prob-
al)ly good enough.

Mr. Nash. Under such a horizontal divestiture program won't bar-
riers still exist for the independent peripheral manufacturers and
software houses?

Mr. Miller. Unless there is a change — which I said I foresee — in
decreased product difierentiation and so on, there will, or there might.
On the other hand I do not see it as being an important part of the
public interest to protect independent peripherals, manufacturers, or
independent software houses, or anyone like that. I think it is ini-
portant to the public interest that the forces of competition pervade-
the marketplace. If, for reasons of structure or because there are
economies, it is more efficient for one company to manufacture CPU's,
peripherals, and a good bit of software, I do not have any problem in
seeing that kind of integration.

When we talk about the automobile industry no one suggests that a
proper way to reorganize the automobile industry is to have some
companies manufacture engines while other companies manufacture
bodies.

Mr. Chumbris. I have a surprise for you. One witness testified in
Febiiiai-y that what they ought to do is have one of them make fenders,
another make bodies, another do the forging, and so forth.

Mr. Miller. I stand corrected. It is my understanding that that is
]iot what most people have in mind when they talk about reorganizing
the automobile industry.

I think the point is that if we start scratching the surface, we dis-
cover that there is integration in many industries. In the computer
industry' it is a little bit more obvious, because firms have started nib-
bling away at some of these pieces of the market, or at least trying
to nibble away, and we become concerned with them. But as a general
proposition I do not think that that kind of integration is something-
that we should be concerned about, so long as there is competition
among the manufacturers of the integrated products.

Now, I would also add my view, as I tried to make clear, that I do-
not think that integration will persist for decades, or even very many
years, if we reorganize the industry, because suddenly the small com-
puter manufacturers will discover that their money is better spent in
improving their CPU's and their systems software, and perhaps one
or a few peripheral products that they are specialists in. But they
will not all decide that they can make tape drives. They will decide



5608

that some company sitting; off on the side that is concerned primarily
with manufacturing tape drives, or disk packs, or terminals, can do a
better job than they can, and that it is in their interest to design their
CPU's so that they are complementary with, and compatible with,
these peripheral products of other manufacturers.

That is the way successor A of IBM will compete with successor B.
Successor A will say, "company X over there in the peripherals busi-
ness has a better tape drive than either I have or successor B has, and
the way in which I can do better than successor B is by designing my
CPU's to integrate with that tape drive, somebody else's disk pack,
someone else's communications equipment, and that way the user will
be able to put together a better system using my product than he will
using successor B's product, because successor B persists in tr3'ing to
put together a much more integrated system than he can really
handle.''

]Mr. Nash. In your proposal for reorganization you also discuss
four market areas which are closely related to the computer industry
itself.

The first is a small insulation processing system; and I gather you
conclude that it is feasible to construct a separate company out of
that?

Mr. Miller. I believe that is what I said, and I stand by the conclu-
sion that is feasible.

Mr. JSTash. You do indicate difficulties in separating production
from peripheral equipment, electronic components, and software.

Xow, with respect to components, you did indicate IB]M once had
a separate com]3onents division and separate plants.

You went on, then, to show technology clianges. I was wondering
whether the production facilities for components today are still sepa-
rated?

}.Ir. Miller. I do not know whether the production facilities for
components today are still separate from those of CPU's. I looked
recently at IBM's 1973 and 1972 annual reports. It was from the
annual reports of the corporation that I had gotten my previous
information about what plants did what; and I no longer find plants
being identified ns components, or systems, or what-have-you, but
simply manufacturing. So, I cannot honestly say that I know the
answer to your question.

^Ir. Nash. A number of witnesses have keyed in on the components
and pointed out, giA^en the changed technology, that the components
division is fast becomxing a key to market control.

You have indicated we might face a loss of economies if components
were divested, if I understand your statement. You didn't feel dives-
titure of components divisions was tliat important.

Could you elaborate on your views with respect to components
beino; divested ?

Mr. Miller. My point was tliat I certainly am not in a position
to say that there are substantial economies of integrating components
with processes. There may or may not be; I do not really know.
I would not want to take the position tliat nothing would be lost by
removing them.

On the other liand, if IBM is broken up horizontally into half a
dozen parts, and if four of those half-dozen parts have a components



5609

plant, then we are in a situation where the use of the components
nianufacuring as a basis for dominating the computer industry is
no h)nger possible.

Now, if it is true that there are very substantial economies of lar2:e
scale in the manufacture of compojients for computer equipment, then
Ave have a problem that there may be a structural tendency to monop-
oly in that particular business. JNIy own observation is that there is
a l■elati^'ely lari:^e number of manufacturers of semiconductois — that
is, integrated circuitry — that the computer industry is not the only
industry that uses integrated circuits ; and that there are many inde-
pendent firms, many of them even very small, that are in the business
of manufacturing circuitry. My understanding is that many of the
other computer manufacturers besides IB]M do not maintain their
own components facilities. They somehow manage to produce com-
l)uters out of components that are manufactured in conjunction with
other integrated circuit manufacturing for other purposes. Under
those circumstances, I do not see that there would be — that there is
likely to be — a major problem in the horizontal breakup of IBM's
components business.

As I said, there were four components plants — or, more precisely,
there were three in existence and a fourth being constructed — the
Inst time T carefully looked at those facts and saw them published
b}^ IBINI. With that breakup. I think it would dissipate any likely
monopoly power that presently is in the components opei'ation of IBM.

Mr. Xasit. Thank you very much.

Senator ITvrt. Mr. Chumbris ?

Mr. Ciir^iBRis. Thank you. IMr. Chairman. I have had my ques-
tions. I Avill turn it over to Dr. Granfield.

Mr. Graxfield. Thank you.

I hope to ask a set of interrelated questions which, to me as an
economist, are consistent and form a particular pattern. I hope I can
convey this to the witness and, more important, to the record.

First gf all. Mr. ]\Iiller. have you had an opportunity to read volume
1 of the hearings of the subcommittee concerning the Industrial
Eeorganization Act?

^[v. ]\riLLER. I have not read all of it. I have looked at parts of it.

^Ir. Graxfield. Then you must be aware of the fact that, some-
what contrary to your statement, there is tremendous controversv
aliout the question of whether industry structure tells us anything
about the level of competition within the industry.

Mr. Miller. I am aware that there is considerable controversv a])out
that point; yes, sir.

Mr. Graxfield. Specifically, some economists wish to direct ex-
amine the proposition and have it in a structural hvpotliesis, that
Avhen you have an industry dominated 1)y four, or six,*or eight firms,
dej^ending upon whatever concentration measure vou are using, this
indeed enhances the possibility of tacit collusion, such as firms together
simply restricting output and therefore having a higlilv competitive
price. Are you aware of this controversy ?

Mv. ]\IiLLER. Yes ; I am aware of it.

Mr. Gr.\xfield. In order to balance out the statement tliat t])ei'e
ai-c some who believe that the structui-al approach to explain economic
conduct and performance is valid, others cast doubt on this present
evidence to support their contrary position; is that not correct?



5610

Mr. Miller. What you are suggesting is that there is dispute over
the question whether dominance of an industry by a single
large firm — and let me not say "dominance," since that is what you
are'challenging — there is a dispute over the question that the possession
of a large market share by a single firm means the industry is not
competitive. You are asking me whether that is the case; whether
there is dispute over that proposition ?

Mr. Granfield. On the contrary. The structure has not dealt with
the problem of an industry dominated by a large firm,

I think if you examine your own footnotes, Professor Baine did
not include any industry dominated by a large firm in his surveys,
nor has Professor Casson or Professor Turner, or the former chief
economist of this subcommittee, Mr. Blair.

There is no such industry in the computer industry that is not con-
tained in any of the analyses.

Mv. Miller. I am certainly not going to dispute that statement.
I think it is probably right. If it is, it would not surprise me if the
reason is that no other industry has been dominated by a single large
.firm in the way that the computer industry has been.



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 86 of 140)