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 87 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 87 of 140)
Font size
QR-code for this ebook

I am not sure that is the case, but it is my impression that probably
since 1911 there has not been an industiy like the computer industry
that one could have included in studies of that sort.

Mv. Granfield. Well, let me add that structural hypothesis does not
yield to the question this committee addresses itself to : Specifically,
the level of competition in an industry dominated by one large firm.
It is a separate question from that which the structurals are directing

themselves to. , i • j^i

]Mr. ]\Iiller. My own answer— and it may not be a good one— is the
belief that there is some sort of continuity. And if what you are saying
is that the structuralist— if you are characterizing the structuralist
hypothesis as applying only to what happens when the number of large
firms is few— hypothesis is applying when there are two or three firms
dominating an industry and saying, "Well, but if there is wily one,
the situation is difl'erent," my reply is that I, at least, am a believer m
continuity, and the whole theory of the structuralist hypothesis is that
one large firm is even better able to do, by virtue of being one, than
what two or three large firms would. So, at least, I think that the
structuralist hypothesis does purport to cover the situation of the
computer industry, or is properly extended to cover the situation of
the computer industry.

The structural hypothesis itself may not be correct. lam willing to
grant that there is dispute about the correctness of it in the first place.
But I do think it at least covers the computer industry.

Mr. Granfeeld. AVliat the industries want to show with the four, or
six, or eight firms, is, one, that they relatively earn similar rates of
return. These rates of return are above the normal rate of return, and
this is their evidence that collusion is in fact current.

If Ave look at the computer industry it is in direct contrast to the
structuralist approach, because we have one firm allegedly earning
higher than normal profits, and many other firms not earning normal

profits. .

And to include an industry such as the computer industry and more
of these kind of industries— this kind of distribution that you find—


the more you will refute the structuralist hypothesis, because it is
directly based on statistical findings of homogeneity or returns among
the leading firms of the industry.

Mr. Miller. JNly answer to that question is that there is homogeneity
of rates of return among the leading firms in the computer industry.
IBM's rate of return is the same as IBM's rcate of return. There is
only one leading firm in the computer industry that qualifies in the
antecedent part of the structural hypothesis and, obviously its i-ate
of return is homogeneous with itself.

And I would also point out that it has been remarkably stable over

]Mr. Granfield. It says nothing about the key element in structural
hypothesis data of successful tacit collusion, unless you were going
to say IBM is colluding with itself.

Mr. ]\IiLLER. I am certainly going to comnient that IB^I has at
various times been organized in division structures that encompass
different parts of the marketing in the computer industry, or the
production and marketing, and that is indeed the point. The struc-
turalist hypothesis has to do with the setting of price policies in rela-
tion to cost. And again I would say that the pricing policy of IBM
can be presumed — I certainly have no evidence — but I would presume
that the pricing policy of IBM is coordinated with itself, which again
is the nritural extension, on a continuity ])asis, of the structuralist
hj'pothesis about what happens in structurally concentrated industries.

5lr. Gr^vxfield. At least, then, when that is true, when you break
out, then, a long division which you suggest, the}^ will then simply
tacitly collude among each other and produce the same result as we
have today ; if that is true.

]Mr. ]MiLLER. That had to do with my point that three successors to
the IBM computer business is not enough ; six would be, at least as I
understand the measurements of IBM's market share. Three successors
would leave a four-firm concentration ratio of approximately 80 per-
cent, if IBM has 70 and if the next largest firm now has 10, since the
top four would be the three successors plus the next largest; while
six successors to IB^SI, if they are equal in size, would bring the four-
firm concentration ratio down to the low 40-percent level,

I make that comment primarily because I belie^•e that the Industrial
Reorganization Act sets a four-firm concentration ratio at 50 percent —
correct me if I am wrong — which is the level at which monopoly is
presumed; and breaking IB]M into six pieces will bring the industry
not very far under that line, but clearly under that line on the basis
of the figures that I have available to me as to its market shares.

]\Ir. Nash asked me whether I thought six was enough, and I said
I thought six was enough. Maybe I am wrong; maybe it is eight. The
structuralist hypothesis does not say that tacit collusion will occur
no matter how many firms there are in the industr}'. It says it will
occur if there is a relatively small number of relatively large firms
who together account for most of the output. I am simply offering a
judgment that in connection both with the standards in the proposed
act on which we are having the hearings, and in connection also with
such of the economic evidence that I have seen, it would appear that
six successors to IBM would break the industry into enough pieces, of
which none was particularly large, to lead to reasonably effective


Mr. Granfield. Do you know of any economic evidence that says
when we go from 8 percent concentration ratio to 40, the competition
suddenly ensues ?

They show that the price-cost margin goes to zero when that
hai)pens ?

Mr. Miller. I certainly do not know of any evidence that the price-
cost margin goes to zero. I have stated before, I believe, in my oral
testimony and certainly in my written testimony, that I am not setting
myself up as an expert on the general questions of industry structure
and performance throughout the economy; that the record on this
question, as you yourself pointed out, has been made before this com-
mittee by a variety of experts who are much better able to respond
to your question than I am.

I do not even think it is necessary or appropriate, in view of my
own qualifications, for me to try to answer that question for the record
of this committee, since it has been done better on both sides by per-
sons who preceded me.

Mr. Graxfield. The only reason I bring this up is that if you pro-
pose to restructure an industry, one should present evidence himself,
or at least have a firm grasp of the issues involved that the general
structure approaches, because that is the entire purpose of the argu-
ment ; that the structural approach is the valid way to measure com-

Mr. Miller. Let me say you are correct, that the reliance on the
hypothesis of the structural approach is certainly a major element in
my conclusion about the merits of reorganizing the industry. I clearly
do want the record to show that. And, let me simply say that I am
not prepared today to discuss or defend the merits of that proposition
beyond the few references I liave in one of my footnotes : but if you
wish, I would, given the appropriate time to prepare, be happy to
com.e back again and discuss that matter with j^ou.

Mr. Granfield. Would you agree with my comment that if econo-
mists with improved sources of data — assuming it was done in a com-
petent, judicious manner — tlint if the structural approach was tested
and found lacking this would cast a grave doubt on your whole ap-
proach to the IBM situation ?

Mr. Miller. I do not know that it should cast grave doubt. It should
certainly cast some doubt on my approach to the 1MB situation. I have
in my prepared statement at least, made comments about some of the
other problems that are not so directly related to the question of mar-
ket competition in a narrow sense; about the problems of having a
single corporation being responsible for the production and. in many
cases; the integration with the business operations of the informa-
tion handling systems for so many large companies ; and about the
impact of large corporations on the way in which government and
others react to the situation. All of these, I believe, are problems that
structural reorganization of the computer industry would go a long
wav toward resolving.

i believe, also, that my study of the computer industry, imperfect
as it is. adduces some evidence that tends to support the structural
hypothesis, at least as it applies to the computer industry, and that
is the main reason T say that if yon did find the structural hypotliesis
wanting, that would cast some doubt on my conclusions about, or my


anah'sis of the computer industry ; but it would not be such gra^■e

It might be that the structural hypothesis is not generally appli-
cable, but that the same situation or the extension of it to the single
firm dominance situation does provide a good explanation of what is
going on in the computer industry,

Mr. Granfield. What do you mean by effective competition?

Mr. Miller. It is a very difficult term to define. A simple definition
of it is that etfective competition is the situation in which the behavior
of tlie individual firms is subject to the discipline of the market and
of market forces; that firms do not have the discretion to decide what
prices they should charge or what kinds of products they jDroduce,
other than the question of going into or getting out of the industry-;
that they are not in the position to manufacture inferior products,
not necessarily deliberately, but that they simply cannot survive
by manufacturing inferior products.

It is the old question of the small grocery store and the supermarket
l)usiness, that Mr. Sam moves out and larger units go into the food re-
tailing business because firms are subject to the discipline of market

I think that is the best simple definition of effective competition.

Mr. Granfield. Don't you think that the exit of RCA and GE
from this industry indicates they, themselves, admit they could not
have the competitive pressures and this industry looks at that kind
of effort to be extremely competitive. That was the same as the "Mom
and Pop" store.

Mr. ^[iller. The question is not whether RCA and GE are subject
to competitive pressures, but whether the bulk of the industry, that
TO percent or so that IBM has, is subject to competitive pressures.
I hardly think one can call an industry competitive by saying that the
firms that dominate it and are not subject to competitive pressure
make the kitchen so hot thpt everybody else has to get out.

Mr. Granfield. My recollection of the dominant model in economics
indicates what happens is the domiiiant firm sets a price which maxi-
mizes his profits.

The other firms in the industry are forced to take the price set by
the pricer and they are earning a normal rate of return whereas the
dominant firm earns economic rents, but we don't see exit in that

That doesn't seem to be what happened here. It happens to be char-
acterized more by a continued, vigorous competition.

Do you ha-\'e any comment on that ?

Mr. Miller. I find the ]:)remise in your question somewhat contra-
dictory from your previous one. You correct me if I am wrong, but
I believe what you said is that we do not see exit from the computer

Maybe if you repeat your question, I will better understand how
it relates to the one before.

Mr. Graxfield. Certainly. I said I did see exit in the computer
industrv which seems to run contraiw to the — at least, the textbook
model of a firm, of an industry dominated l)y a leading firm, which he
makes his output and his pricing decisions in order to maximize his
profits, the other firms in the industry are forced to be price takers

40-927 — 75 r,0


and tliey earn a normal rate of return and under these conditions you
jionnally do not see exit.

Mr. Miller. Let me first say that I am not sure I agree with the
premise about the textbook modeL If in the textbook model tlie
dominant firm decides to lower its price a little bit below the short
run monopoly profit level, it can be presumed to expand its market
share at tlie expense of its rivals. But to go to the facts of the com-
puter industry, I don't really think that the dominant firm textbook
raodel you gave insists that there be no entry into or exit from this
industry. The model simply suggests a stable market structure, and
in a broad sense that is what I think I see in the computer industry.

IBM's market share has, according to the best figures publicly
available tliat I have seen, been approximately stable. There has been
exit but there has been also, in that sense, entry or growth of other
firms. There has been, if you wish, turnover in the non-IBM part of
the computer industry ; and I think that is a natural occurrence when
you take into account the fact that some of the other smaller fii'ms
may not be as well managed as some of the better managed among
IBM's smaller rivals, and they tend to lose market share to the better
managed ones. We see that sort of turnover, so I do think that is
consistent with the textbook model that you described.

Mr. Graxfield. In a December 1972 American Economic Review
article by Professor Belchen, in attempting to explain differential
rate of returns, the hypothesis was that there is no single segment of
a firm with superior performance.

It's a team aspect of the organization that leads to the earning of
economic rents. Is it not possible that, although we separate out and
examine intensively any one of the IBM divisions, they appear to be
not inordinately superior to competition, but it is the unique combina-
tion and information flows that IBM manages through this complex
organization that leads to their superior performance.

Is that not a possible hypothesis ?

Mr. Miller. I will certainly grant it is possible, but before I com-
ment, I'd rather hear what the next question will be.

Mr. Granfield. If this is indeed true, is it not possible that we will
lose this tremendous team efficiency if we operate IBM in the manner
you suggest?

Mr. "iSIiLLER. I am somewhat hard put to answer that question
because there may be some confusion about what we are talking of.

Are you suggesting that the loss will come about because we
separate the Office Products Division from the computer business or is
it something else ?

Mr. GrxVnfield. That is correct.

]\Ir. Miller. If we are thinking of IBM as a conglomerate, a com-
pany operating in a number of industries, and if what you are sug-
gesting is that the efficiency of the operations has to do with the fact
that the management is very good, I am ceitainly not familiar with
any substantial body of economic evidence to suggest the fact that
either the efficiencies of a large conglomerate corporation are depend-
ent on its being a conglomerate, nor am I familiar with any evidence
to suggest that the quality of management is destroyed by putting it
into smaller pieces.

Tliere is another way of responding to tliat }:)rol)lem. I certainly
am not disputing that there is something about IB]\I that allows it to


earn substantial excess profits ; and I also accept and indeed vigorously
espouse the proposition that one of tlie benelits that will result from
structural change in the computer industry will be the creation of
the situation in which those monopoly profits will no longer be
earned by IBM. Thus, if what you are asking me is whether I think
the reorganization of IBM will cause a change in the situation that
creates this problem, the answer is, yes, of course, that is one of the
[)urposes of a reorganization: to do something to deprive that collec-
tion that we call IBM, whether it is management information flows
or anything else, of the ability to so control its destiny that, among
otlier things, it is able to charge prices so far above the cost of its
doing business that it earns substantial monopoly profits.

Mr. Granfield. Well, this is exactly the point I wish to bring out.
It's whether these are monopoly profits or are they economic rents?
Economic rents is superior management, and if we break it up IBM
loses the economic rents, and the market loses that efficiency to the
detriment of the consumer of those profits.

To quote from your testimony, you indicate that one of the ef-
ficiencies or one of the talents of IBM when tlicy first encountered
ISperry Eand was they learned a great deal about managing customer
services in the computer business; and although it didn't look like
selling typewriters would lead to a better computer division, in fact,
that seems to be a transfer of ability.

I don't Ivuow how it would be inhibited. This kind of efficiency ulti-
mately led to better customer service, that it would be lost as it would
be in many firms if you attempt to divest it along the lines you

^Ir. I\IiLLER. I would comment primarily about the facts. "WHiat IBM
transferred was not expertise in selling typewriters. I do not know
even how important IBM was in the typewriter business in the early
1950's. What IBM transferred was its expertise in selling tabulating
equipment, punched card equipment, into the successor of the punched
■ card equipment business ; namely, the computer industry.

The point I would now make is not that expertise in selling tabulat-
ing equipment is important to the sale of computers and therefore,
siiould be preserved. I think the important point to make is that IBM
dominated the tabulating equipment business to a much greater ex-
tent, even, than — at least, according to what little evidence there is
on that point — it dominates the computer industry. One way in which
I might characterize your argument — it is not entirely fair but it may
be a beginning point for exchanging views — is that you seem to be
suggesting that J3ecause IBM dominated the tabulating industry and
was able to dominate the tabulating industry, whetherthat was good
or bad for the economy at the time, it is important that we allow IBM
to transfer that monopoly to the computer industry, because there is
some efficiency in letting computers grow out of tabulating equipment;
and because IBINI dominates the computer industry we ought to let that
dominance go on forever, including all the other industries that may
eventually grow out of the computer industry, because we are going
to wrencli the system somehow if we ever try to do something about it.

I will not buy that.

Mr. Granfield. If you force IB^M to divest itself along the lines
that you suggest, however, all of the benefits would be lost. This was
in pre^dous testimony.


]Mr. Miller. The response I would make is that in my analysis of
the industry I have tried to identify specific areas where I think
IBM's advantages lie. I have tried to distinguish between those that
involve what I call true savings of resources, or true social cost sav-
ings, and those that are really private pecuniary advantages that have
nothing to do with benefits to the public but simply give IBM ad-
vantage over its rivals in the computer industry. I have found, by my
analysis, that most of the explanation of IB^NI's dominance lies in that
latter group of advantages.

You are now proposing sort of an agnostic view of things, at least as
I see it; namely, that one possibility is that somehow by some not care-
fully documentable process, some kind of synergistic process of com-
bining all of these things in one group, we somehow achieve an entity
that in unspecified ways is more efficient than smaller entities, than
IBM's smaller rivals. And you are suggesting that we will lose what-
ever these benefits are, these things that come about in unspecified
ways, if we break up IBM.

Without a more precise description from the proponents of that
view as applied particularly to the comjniter industry, of exactly what
economics result from this combination, I am totally unable to respond.

Even if I had a more detailed description I would add that when
it comes down to detailed facts about this or that economy in the
computer industry, or exactly what the situation is, my experience,
while it stretches over a period of years, is limited essentially to the
interaction with personnel of two companies plus wluit I can glesin
from having read the trade literature in a period that ended some 5
years ago, because I have not done work on the industry since then
except in connection with the Justice Department suit in which I was
involved, and which I cannot bring to bear at all on Avhat I say today.
Therefore, I can only repeat that my analytical finding that these
advantages are resident somewhere else.

Mr. Graxfield. Well, my only point — and I certainly don't wish to
belabor or bore you with it — is that there are two ways economists ex-
plain an industry dominated by a single firm: One is as alleged in
numerous antitrust suits against IBM, tluit it is doing something illegal
to preserve its rate of return. The other view is that IB^I sim])ly is
a superior performer. The source of that superior performance is not
necessarily easily separable out into what the economists either look
at as production economies of scale, or marketing economies of scale.
or distribution economies.

They may all also be present. But the most significant source of
efficiency for that superior performance simply may be the organiza-
tional structure and the djaiaminism of that organization.

For example, I would argue that if you examine the relevant coi'-
porate structures of Sears and Montgomery Ward, their buying prac-
tices and so forth, you would find remarkable similarities; that year
after year Sears outperforms Montgomery Ward.

Some allege that simply because of superior management capalnlity
of the Sears executives, and Sears spends a great deal more money and
time training their executives, this is where the rate of return accrues
to Sears, this team group of managers; and that IBM tradition has
been to spend more in the trade and promotion of the human capital.

At least that is an alternative hypothesis to explain superior per-
formance not only of IBM but of any firm in a particular industry that


far outperforms its competitors. And the resolution of that uhimately,
in terms of economics, is that if it is not found that the firm has con-
structed artificial barriers that, indeed its performance is explained by
superior organizational efficiency; and that is a subject that as a
problem has not yet been resolved.

That is tlie only thing I wanted to bring out. There are two alterna-
tive hypotheses here. We do not have any resemblance of sufficient
evidence to make a decision as to which hypothesis is correct.

I would submit that both are equally valid hypotheses to explain
superior performance.

^Ir. Miller. That has been a long question. There are a couple of
points that I would like to respond to : One, primarily to keep the rec-
oi-d straight, I do not think I have said that IB^I's dominance is due to
its having done something. I certainly have not said that it is due to
IBjjiTs having done something in ^•iolation of the antitrust laws. I have
tried not to make any comments about IBM's position in regard to the
present antitrust laws.

As a general proposition, though, I subscribe to what I think in
general terms you said was your hypothesis ; namely, that IBM is a
(lominant firm in tlie industry because of something that is, not some-
thing that it necessarily does, if you want to distinguish between some-
thing that is and something that IBM does.

The question before this committee in its concern with the Industrial
Eeorganization Act, as I understand it, is not whether the economy
would be better off if we prevented firms from doing things — to use
an active verb — that monopolize or control an industry. The question
is whether the U.S. economy and the public interest would be better
served by correcting, once and for all. situations that involve a non-
competitive structure in the market, and in particular in an important

That is why I go l^ack to IBM's dominance in the tabulating indus-
try and say my view is that IBM got its large market share in the
computer industry primarily by virtue of having dominated the tabu-

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 87 of 140)