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 91 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 91 of 140)
Font size
QR-code for this ebook


some items of its peripheral equipment business, including specialized manu-
facturing facilities, but this is hardly the kind of remedy envisioned by the
proposed Act for a broad monopoly of the computer industry. Finally the respon-
sibility for architectural design of IBM's computer systems, apparently including
peripheral equipment, is centralized in the Systems Development Division, and
there may be some loss of real resource economies if peripherals are separated
from their architectural moorings.

Electronic components are in some ways a mirror image of peripherals, but
the conclusion is the same: separation from the computer business is of doubt-
ful merit. IBM once had a Components Division, but its functions have now been
absorbed into the System Products Division, which manufactures the central
processing units for IBM's computer systems. When the Components Division
was a separate entity in IBM, it had its own plants, and separation of components
manufacturing thus" appears feasible. But there may still be some economics
derived from vertical integration in components, especially now that integrated
circuits contain so much of the circuitry of a computer models on a single physi-
cal component. Since decreased vertical integration is not the medicine needed
to increase competition in the computer industry, the case for it seems dubious.
Software and services present the clearest case for letting market forces deter-
mine the nature and extent of integration in an industry or industries. They have
alwavs been thoroughly integrated with IBM's systems design and marketing
activities ; and for at least some software products, such as operating systems,
economies are attendant upon this integration. I have argued above that struc-
tural change in the computer industry may lead to decreased integration in soft-
ware and services, but the transfer of these functions can best be accomplished
in response to the market forces working against continued integration. In-



» See, for example, IBM's 1967 Annual Report, p. 12.



5635

deed, a reorganization plan that preceded tlie development of these forces would
be unsuccessful in separating software and services from the successor computer
manufacturers, because the resources — men and women — are mobile enough to
How right bacli to where the market forces are directing them.

The third layer of a reorganization plan, and the one directly indicated by the
proposed Act, is horizontal dissolution of IBM's computer systems business. It will
be a complex operation, even apart from the difficult job of unraveling a corporate
financial tangle. In the case of IBM, the important physical assets to be divided
are the manufacturing plants, the stock of leased equipment, and the laboratories.
Important intangibles are patent rights, unpatented knowledge of the technology,
software and support expertise, and customer relationships. The last of these
will generally be associated with the ownership of leased equipment, and it may
also follow such "liabilities" as exist under maintenance contracts or other ob-
ligations to owners of purchased equipment. The other intangibles are either
finely divisible, because they are carried in the persons of IBM's many em-
ployees, or they are common goods that can be assigned to all the successor
firms.

The physical assets require greater consideration. In 1968, when I last compiled
puliliely available data on this subject, IBM operated some twenty manufactur-
ing plants and thirteen laboratories for its main line of computers, plus three
components plants and a fourth planned. These counts include IBM World Ti-ade
Corporation, which conducts essentially all of IBM's operations outside the
United States.

If World Trade can be included in the reorganization plan, then the plant
structure is no obstacle to creation of half a dozen successors in the computer in-
dustry. If World Trade must be left whole, it should still be separated from all
of IBM's domestic operations, to minimize the likelihood of World Trade's enter-
ing and dominating the U.S. market as IBM reincarnated. Even without World
Trade, it should be possible to create at least four or five successors to IBM's
domestic computer business. Since IBM's policy is to encourage some specializa-
tion of manufacturing plants by product, each successor will initially have a some-
what shorter product line than IBM as a whole ; and within that product line, its
output will be heavily weighted with the products for which its plants have
had primarily responsibility. Similarly, the laboratories assigned to each suc-
cessor will tend to have a broad basic competence, but a lopsided area of expertise.

Tlie first feasibility question about this reorganization plan is whether it
destroys any economic efficiencies presently enjoyed by IBM as a whole. In the
short run, while the newly formed companies are confined to the short product
lines of their manufacturing plants, they will lose most of the advantages of
integration. These advantages, however, are of little consequence to the economy,
especially in the short run. Most of them are merchandising advantages that bene-
fit the integrated firm at the expense of its I'ivals, but that offer no gains for the
public. The others arise only in or from the design process for new equipment,
but when that occurs, the plant's product lines are changed anyway, and the
firm is no longer bound to its inherited .specialization.

Product changes thus mark the end of the short run, and in the long run the
newly formed companies have the additional option of choosing the extent of
their integration.-" Most important, the long run in the computer industry is
likely to begin very soon after dissolution, owing to the rapid pace of product
changes. The question of efficiency, at least in regard to manufacturing opera-
tions, then becomes simply the question of scale : how large must the successor
parts of IBM be in order to achieve the available economies of scale in manu-
facturing? The answer is not known, and one reason is that there is no historical
experience with computer manufacturers even one-fifth as large as IBM, except of
•course for IBM itself. However, there are no apparently substantial economies
of .scale separating the firms with five to ten percent of the market from those
having shares in the three-to-five percent range, so it seems likely that judicious
specialization and selection of products can compensate for small total size in
the computer industry.

Similar remarks apply to the other loci of scale economies that were discussed
previously — new product development, field maintenance, and marketing — , and
to the external economies in computer use. In each case, there is no good evidence



2« With the computer industry growing rapidly, the newly formed companies are apt to
require expansions of their nianufacturing facilities even before they introduce a new
line of products. Plant expansion facilitates the addition of different products to the line,
and it may thus cut even shorter the time the firms are restricted to their Inherited
product lines.



5636

indicating just how far tliese pconomies extend. However, the discussion above
does suggest that the important advantages of large size are not true social econ-
omies, and that the true economies can be achieved at a far smaller scale than
that of IBM.

Component production does impose a minor obstacle, because there may not
be enough component plants for each successor to have one. This is not an im-
possible situation, as there is an independent components industry, and even some
moderate-size computer manufacturers do not presently manufacture their own
components. Lack of integration may still be a disadvantage, but there is no obli-
gation to make the successors to a dissolved firm start from positions of equality,
and the unintegrated firms will of course be free to begiiu their own component
production if they find it desirable.

The final observations on reorganization concern the installed base of leased
equipment. One feasible reorganization is to distribute the leased equipment in
accord with the manufacturing capabilities. But an added competitive fillup —
and one that will probably help speed major changes in the division of labor
among hardware manufacturers, service organizations, and users — may be to give
the installed base of leased equipment to one or more successors not provided with
hardware manufacturing capabilities. The Industrial Reorganization Commission
may enjoy analyzing this particular twist early in its career.



The computer industry is changing, and many of the details of my story are
half a decade out of date. This makes them ancient history by the standards
of the computer industry. It may therefore be suggested that my analysis is no
longer relevant, having been overtaken by events. In particular, it may be sug-
gested that structural reorganization of the computer industry is not needed now,
because other structural factors have changed in a direction favorable to com-
petition. Some of the major product changes in the computer industry during the
past few years are as follows :

Minicomputers : One can now purchase, for a price in the low tens of thousands
of dollars, a computer as powerful as the largest computers in existence two
decades ago. The minicomputer is small enough to sit on a table, and its pur-
chase price is much less than the monthly rental charge for the computer sys-
tem of the mid-1950's that it matches in computing power.

The marriage of computers and communications, which enables computers to
work for a user whenever and wherever he wants.

The appearance of many new kinds of terminals, including point-of-sale termi-
nals and other on-line or real-time applicaticwis systems.

A look at the participants in the computer industry shows that they too have
changed. Two of the five leading contenders in the race for second place have left
the field, their activities having been absorbed by two that remained. In 1970,
General Electric sold its general purpose electronic computer business to Honey-
well. In 1971, RCA simply departed ; and after leaving, it sold its installed base
of computers on lease to "Univac. These consolidations have increased seller con-
centration at the top of the market, though they can also be viewed as the first
steps toward transforming the computer industry from a near monopoly inta
some kind of oligopoly. At the bottom of the market there has been a prolifera-
tion of minicomputer manufacturers and of some types of peripheral equip-
ment firms. These changes can be said by some to represent a strong increase in
competitioQi.

The last few years have also seen major developments on the private antitrust
front. Control Data settled is differences with IBINI early in 1973. Meanwhile the
Greyhound and Telex cases went to trial, with IBM winning the former and los-
ing (oin the antitrust issues) the latter ; but both cases are on appeal.

Those who use these developments as an argument against the desirability of
structural change are in my view following the path that has since 1911 led away
from effective antitrust relief. The general issues have been widely discussed.
and it is not my purpose to add to an already lengthy literature. But there is one
observation from the history of the computer industry that is germane, and it is
this : Competition in the computer industry is, as I have explained, the supposed
beneficiary of a major antitrust suit brought against IBM by the Justice Depart-
ment in 1952, and settled by consent of the parties in 1956. The consent decree in-
cluded some behavioral restraints but no significant structural reorganization,
and the result is that we are here twenty years later with IBM's dominance now
writ across a much larger and more important industry than the tabulating
equipment industry was in the 1950's.



5637

AFTERNOOX SESSION

Senator Hart. The committee will be in order. With apology again
for holding him over, we welcome our final witness for today,
Dr. Gerald Brock from the Department of Economics, University
of Arizona.

STATEMENT OF GERALD BROCK, ASSISTANT PROFESSOR OF
ECONOMICS, UNIVERSITY OF ARIZONA

Dr. Brock. Thank you, Senator. I appreciate the chance to be
here today. In the light of Mr. Katzenbach's statement on Tuesday, I
would like to clarify my role in the pending antitrust cases before I get
into my regular statement.

Mr. Katzenbach stated on Tuesday : *

Professor Brock recently completed, under the tutelage in part of one of the
Government's principal experts, a Ph. D. thesis on the ct)mputer industry,
which is an important source of the Government's theory of the case agauist
IBM.

My Ph. D. thesis was on the computer industrv, but was done under
the direction of Harvard economists Richard Caves and Marc Rob-
erts, neither of whom have any connection with the Justice Depart-
ment suit.

I have never worked for the Justice Department or any other
participant in the pending antitrust suits in the computer industry.
I have studied the computer industry as an academic economist with
no preconceived ideas of the results of the study, or commitments to
any of the conflicting interests in the computer industry.

So far as I know, my work has had no efi'ect on the Justice Depart-
ment's case.

The computer industry is of special interest to academic economists
such as myself because of its unusual market structure and rapid
technological progress. I have been studying the industry for the
past 3 years in an attempt to bring economic analysis into the debate
surrounding potential Government action toward the computer
industry.

My statement will concentrate on how the industry's price and prod-
uct policies are related to the industry structure. By industry struc-
ture, I mean, primarily, the concentration of firms and barriers to
new firms entering the market.

IBM's market share as computed by three different sources is shown
in table 1 of my prepared statement. The three computations each
have somewhat different bases, and the figures do not exactly agree.
Plowever, they are close enough that we we may conclude with con-
fidence that IBM's market share has been in the 65- to 75-percent
range.

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

Dr. Brock. Roughly speaking, the major competition to IBM has
been the seven companies listed on table 2 who have 2 to 10 percent
of the market apiece. The seven were reduced to five with the
Honeywell-General Electric merger in 1970 and the exit of RCA
in 1971.

♦See p. 4836.



5638

Besides the companies listed, there are a large number of com-
panies with under 1 percent of the market each, largely competing
in specialty areas such as minicomputers rather than the main
general -purpose computer market.

Barriers to entry are anything which makes it difficult or impos-
sible for a new firm to enter the industry. For the computer industry,
barriers to entry consist primarily of raising capital, economies of
scale, and brand loyalty. All three problems are much greater in the
integrated systems portion of the market than in individual parts
of the computer market, such as software companies or companies
which make only input-output equipment.

Economies of scale are of minor importance in the actual manu-
facturing stage, but pose a significant barrier to new competition when
considering production of software. Producing software is like writ-
ing a book. The first copy is very expensive, but subsequent copies have
a low marginal cost.

A very small, well-managed firm would experience only slightly
higher unit costs than IBM when producing tape drives, memory
units, or central processing units, but would have an extreme cost
disadvantage if it tried to compete Avith IBM's systems software,
because almost all the cost is in producing the first copy.

Although applications software has the same theoretical economies
of scale as systems software, in practice they are not so significant
because modifications are often required for each user of application
software.

Consequently, economies of scale have not been a significant barrier
to new competition for companies producing competitive input-output
equipment, minicomputers, and service and consulting organizations.
Economies of scale have increased the difficulty of entering the in-
tegrated business systems market where complex and extensive soft-
ware support is required for success.

Tlie second barrier to entry — brand loyalty — is accounted for by
two factors. The first is the difficulty of maldng rational computer
selections. Because the major manufacturers are integrated, and parts
of one manufacturer's system generally cannot be used in another
manufacturer's system, the computer user nnist choose between com-
plete systems, including the central processor, various kinds of input-
output equipment, systems software, and various levels of consulting
help and applications programs.

It will seldom be true that one system dominates the others on all
possible points. Generally, the user will have to balance, say, a better
operating system on one machine against better input-output equip-
ment on another.

To complicate matters further, the user generally expects various
enhancements over the life of the machine, either in software or hard-
ware, which are_ not known in detail at the time of the selection.

The user also is often dependent upon the manufacturer for help in
defining his computing requirements and choosing the proper equip-
ment to meet them.

As a result of all these factors, the decision of which computei- to
purchase is seldom a perfectly rational one, but instead it is heavily
dependent upon the user's judgment about the future actions of the
various manufacturers. Equally qualified managers faced with the



5639

same information are likely to come to different decisions about which
computer best meets their needs.

As in any situation where clear choices are difficult, the established
manufacturers with good reputations have a great advantage over new
companies, even if the new companies have equal or superior products
on an objective scale.

The second factor accomiting for brand loyalty is the lack of com-
patibility among systems. In an extreme case, if all programs were
written in assembly language and the two computers had different in-
struction sets, it would probably not be profitable to switch, regardless
of the price of the new computer, because the cost of transferring ail
the programs would be equal to the value of the machine.

In the more normal case, with most programs in a language such as
COBOL or Fortran, the basic programs are usuable but some changes
are required either in the programs themselves or the job control lan-
guage. Switching computei's is conceivable in this situation, but the new
company must offer substantial price discomits in order to induce the
customer to pay the cost of conversion.

The compatibility and evaluation problems are largely the result of
selecting complete systems and the incompatibility of various manu-
facturers' specifications.

However, when both of these problems are eliminated, some residual
brand loyalty remains. Probably the most straight-forward decisions
regarding computer equipment are those made between IBM and com-
petitive, plug compatible, peripheral equipment.

Only a single piece of equipment is considered at a time, so systems
tradeoffs are not involved, and the competitors adopt IBM specifica-
tions so that compatibility is not an issue. The only uncertain issue is
the reliability of competitive claims for their product versus those of
IBM.

Even in this simplified situation an IBM study showed considerable
brand loyalty. IBM asked a sample of its disk customers what the
maximum discount was that a, competitor could offer for a replace-
ment product and the customer still remain with IBM. The results are
shown m table 3.

As can be seen from the table, only a small proportion of IBM cus-
tomers were potential targets at discounts of less than 10 percent, and
31 percent would remain with IBM even with discounts over 20 per-
cent. If the table is an accurate representation of the computer market-
place, brand loyalty is a substantial barrier to entry even hi the
plug-compatible peripheral market.

The quantity of capital required for entry into the computer mdus-
try depends heavily upon the segment entered. Service bureaus,
consulting groups, and software houses can have very low capital
requirements of a few thousand dollars for initial salaries and office
rental before payments begin.

Plug-compatible peripherals and minicomputers ha^-e capital re-
quirements beyond the ordinary range of private financing, but still
moderate in comparison with many industries. In minicomputers, the
most successful new entry in recent years. Data General, began with
$50,000 in 1968, but raised $20 million'more in the stock market between
1969 and 1971 to become an established company.

In analyzing potential entry into the memory business, IBM esti-
mated that a new company could begin with $75,000, but would need



5640

to invest $15 to $20 million more over a 4- to 6-year period before reach-
ing the break-even point.

The capital requirements for entry in the integrated systems busi-
ness are huge. When RCA left the computer market in September 1971
it took a $490 million writeoff. In addition, the company estimated it
^ould have needed a $500 to $700 million new investment over the next
5 years in order to attain profitability, suggesting total capital require-
ments of over $1 billion for a competitive systems company.

The amount of capital required is so large that it is unlikely that any
new company could raise it, and it is beyond the financing capabilities
of all but the largest corporations.

In speaking of capital costs as a barrier to entry, I would like to
respond to Mr. Granfield's remark yesterday to Mr. Collins that firms
have no difficulty raising any amount of capital if they are earning a
normal return. So long as there is perfect certainty with regard to fu-
ture earnings the statement is correct. If the amount of risk is well
defined and agreed upon by all investors then the cost of capital is still
independent of the quantity needed. However, an investment in a new
company is not only risky in the sense that there is some variance
aroimd the expected return, but also uncertain in the sense that the
mean and variance of the return is unknown.

Each investor has a different subjective probability distribution
regarding the return and risk he expects on the investment. If only
small amounts of capital are needed, it can be raised from those in-
vestors with the most favorable outlook. As lai'ger amounts are needed,
successively less favorable investors must be tapped, causing an
upward-sloping supply curve of capital to the firm rather than a
perfectly elastic one; consequently, the absolute amount of capital
retjuired does form a barrier to entiy.

Taken together, the capital cost, brand loyalt}', and economies of
scale form an almost insurmountable barrier to new entry in the
integrated systems business. In spite of ti-emendous market growth
and extensive technological changes no new companies have success-
fully entei-ed the integrated systems business since 1960.

In contrast, there has been continuous entry in peripherals, soft-
ware, and minicomputer companies, as would be expected from their
very low barriers to entry.

The extensive brand loyalty in the computer industi-y means that
a]] major manufacturers — not only IBJNI — have market power. Market
power is defined as the ability to raise prices above the cost of pro-
duction, including a normal return to capital, without being driven
out of the market.

If no manufacturei- makes a specific attempt to attack the customer
base of another manufacturer tlirough compatibility and/or con-
version aids, then all can enjoy high profits and stable market shares.
This has not been tlie pattern observed so far in the computer industry,
but appears to be becoming more important.

So long as growth in the industry was exti-emely rapid and the non-
IBM manufacturers had a relatively small customer rental base, it
was to their advantage to aggressively attempt to expand their market
sliare.

However, price competition is very expensive to companies with
large rental bases because a price cut means not only lower revenues



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 91 of 140)