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 92 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 92 of 140)
Font size
QR-code for this ebook

on the new customers attracted, but also lower revenues on machines


plafpd earlier. Consequently, there is a tendency to concentrate more
on upgrading" one's own customer's than on taking customers away
from competitors as the rental base grows larger.

IBM's evaluation of competitors for the o70-135 concluded:

It appeared that many competitors were focusing on protecting and growing
tlieir own inventory bases and were not prepared in the near term to get around
the M135 in pursuit of IBM's lease base.

IBM has been the best example of the grow-your-customers ap-
proach to computer marketing, but not the only one. IBJM has made
a strong effort to develop close relationships with its customers
through emphasizing rental rather than sales, and before 1970 by
emphasizing a full range of services in additioii to the hardwai-e for
the basic price. IBM does iiot provide conversion aids from competi-
tive, machines and makes no attempt to directly undercut an}- given
systems manufacturer.

Because of the high cost of switching between incompatible ma-
chines a firm that wants to expand its share of the market cannot
simply cut the price of its machines. It must either provide a lower
price, together with compatibility, or provide a technological ad-
vance which increases the capabilitj^ of the machine.

Both strategies have been u.sed rcgularlj^ in the industry.
If a substantial increase in the capability of a computer can be made
through technical innovation the imiovating company can attracfc
new customers who either are not using computers or who need the
nev»' capability enough to pay the necessary conversion cost.

Two examples are the Control Data GGOO and the Burroughs master
control program. When the Control Data 6600 was delivered in 1964^
its capacity was far beyond anything else on the market. Consequently,
for vei'y large-scale computing needs, such as atomic energy research,
the 6600 could easily induce the user to switch from previous com-
puters even without compatibility.

Buri-oughs' master control progi-am foi- the B-5000 system and
related AOSP system for the military-oriented D-825 system put
l^urroughs well ahead of other companies in developing effective
multiprocessing capability which increases reliability.

Burroughs' early development of multiprocessing and virtual
memory capability allowed the company to attr.-ict customei's needing
those capabilities, in spite of compatibil'ity problems.

The innovation approach to increasing market share is only suc-
cessful if the innovation is accomplished successfully without too
great cost to the company, and if ft cannot be easily duplicated by

Transistor computers were probably the greatest single advance in
the industry so far. Philco was the fii'st company to introduce a large-
scale transistor computer, but was followed so closelv by the IBM
7090 that Philco gained little advantage in its innovation.

Time sharing has been a very impoi'tant development since the
mid-1960*s. However. General Electric ma<le earlv large investments
m time-shai-ing technology, but failed to capitalize on its innovations
bcc:vuse the costs and technical difficulties were greater than expected.
The third possible strategy is to market directlv against a par-
ticular computer manufacture!- through coj)ving des'ioii specifications
in order to achieve a high degree of compatibility. This is the most di-


rect method of expanding market share, and also the most dangerous
for the firm because it is equivalent to starting a price war. Once com-
patibility is achieved customers can more easily leave the aggressive
company as well as come to it. The price cutting company must be
concerned both with competitive response and the timing of its at-
tack in the product cycle.

• If the competitive product is introduced late in the target firm's
product cycle, it must also be competitive with the target firm's next
generation or the rental life will be too short for profitability.

The target firm may choose to cut prices as a result of the competi-
tion, but it must be concerned for the effect on its own rental base.
If only a small proportion of the rental machines are likely to move to
the competition, then the firm would be hurting itself with an across-
the-board price cut to restrain competition.

Two of the best examples of the compatibility strategy are-the
Honeywell 11-200 and the RCA Spectra 70 series. During the early
1960"s the IBIM 1401 was the most popular computer on the market.
Honeywell announced the 11-200 in December lOGo for first delivery in
July 1964, together with a program called the Liberator, which would
convert the IBM 1400 series programs into H-200 programs. The H-
200 was a great improvement over the 1400 series, for comparable
prices, and strong enough to remain competitive with IBM's re-
placement for the 1401, the 860-30.

The success of the H-200 was increased by incompatibility between
the 1401 and 360-30, making it easier for a customer to convert to
Honeywell than to upgrade to IBM. The H-200 success brought
Honeywell into an early position of stability and profitability in the
computer industry.

A less successful example of the compatibility strategy was the RCA
Spectra 70 series. Soon after the IBM 360 was announced in 1964,
RCA announced the Spectra 70, a series of four computers, each de-
signed to be compatible with, but to outperform, its 360 counterj^art.
The instructions format, and character codes were identical to
those on the 360. In spite of compatibility, the Spectra series did not
sell well against the 360. RCA was trying to sell against a largely un-
defined, rapidly changing target as various enhancements were added
to the hardware and software of the S3^stem 360, making close com-
parisons difficult for customers to make.

In addition, RCA could not offer the same kind of clear price ad-
vantage that Honeywell did against the 1401 because RCA was bring-
ing out its machines at approximately the same time as IBM, while
Honeywell had the advantage of several years additional technical de-
velopment. RCA's market share remained at the 2.5- to 3.5-percent
level of its pre-Spectra days.

The difficulties in following a competitive strategy with the cur-
rent structure of the computer industry make tlie profits to be earned
from competition questionable and far in the future. So long as the
company is consciously undercutting other prices or investing large
amounts_ of money in new technology, it is unlikely to be making
substantial profits. It is buying market share for future profits.

However, there is always the risk, as happen ded with RCA and
GE, that the future profits will never be realized. If no firm makes
specific compatibility attempts each can act in a semimonopolistic


manner with a good deal of market protection. Competition and gen-
eral technological progress cannot be ignored because if prices are too
hidi other firms will come into the market or customers will switch and
pay the conversion costs, but the entire market would be relatively
insensitive to exact price comparisons.

If no hrms made definite attacks on each other's customers through
compatibility and lower prices the industry as a whole would reach its
maximum profit position. This would also be the worst position for
customer welfare. It appears that the industry is moving in this direc-
tion since the exit of RCA.

In this respect, I think it might be relevant to comment on Mr.
Granfield's questions to Mr. ^Miller this morning regarding homo-
geneity of return among the various companies in the industry.

If all companies were following a strategy of simply growing their
own customers, then we would expect to see relatively homogeneous
profits. However, so long as some companies are making a specific
attempt to increase their market share their profits will show up as
lower during those years. I believe that is what has been happening
throughout most of the history of the computer industry. Honeywell,
RCA, GE, and others have been making a direct attempt to increase
their market shares and consequently have been making lower profits
than you would have expected if they were not making that attempt.

While full systems su])pliers have three strategies open to them,
partial line suppliers have only the compatibility option available.
Because they are competing for one product on another manufacturer's
system their products can be made obsolete by a change in the systems
mnkers specifications, leaving them without a stable base of customers.
Technical advances are limited by the need for compatibility with
the systems maker's equi]iment.

The partial line suppliers are made dependent on the systems sup-
plier because of the lack of independent CPU suppliers.

Consequently, at present the customer cannot purchase an entire
system without using the CPU supplied by one of the systems

An example of the benefit that can come from having independently
available CPU's can be seen in the current market for used system
360"s. Because leasing companies bought large numbers of 360's during
tlie late 1960's, many are available for remarketing and modification.
The purchased 360 CPU's have been improved with independent mem-
orj^, input-output devices, and in some cases enhancements to the
operating systems in order to make them far superior machines to
what was allowed by IBM specifications.

However, the significance of this movement is limited by the numbers
of purchased 360's available for modification, because no independent
company is manufacturing CPU's.

The problems of the independent peripheral companies in com-
peting with IBM are documented in the selection from my forthcom-
ing book, which has been presented as a written supplement to this
testimony, as well as by several of the other witnesses at these hearings.
[See exhibit 2.]

Mr. Brock. Here I will summarize the economic issues involved in
the controversy between the peripherals companies and IBj\I. The
basic fact necessary to understanding the problem is that there are


liigli barriers to entry in the systems market but low barriers to entry
in the periplierals market. In any situation where a company possesses
substantial market power in one product and less market power in a
complementary product, the most profitable policy is to tie the prod-
ucts together, refusing to sell one without the other. If the products
are totally tied together the company can clioose the prices for each
combination to optimize profit, without concern for competition. How-
ever, because of legal restraints on tying products the company mv^v
want to raise the prices on the product with monopoly power and
reduce the prices on the product subject to competition as a substitute
for tying them together. This is not as profitable as tying the product
together, but it is less likely to get the company into legal difficulties.

IBM has pursued a combination of both strategies m response to
peripherals competition. IBM has tied products together where the
tie could be technically justified, and has raised the CPU prices while
reducing peripherals prices where tying could not be defended on
technical grounds.

Exampfes of tying products together are the integration of con-
trollers with the CPU on the 370-145 and later machines, and the tying
of large quantities of minimum memory to the basic CPU price on the
370-158 and 168. Although it cannot be established for certain that
the integration was a response to competitive pressure, rather than
simply a design change to take advantage of new technology, the cir-
cumstances surrounding the introduction of integi'ation suggest that
it was in response to conipetitiA-e pressure.

Examples of raising the CPU price while lowering the peripherals
price in response to competitive pressure were tlie fixed term plan,
which reduced the price of peripherals, followed by a general CPU
price increase and the reintroduction of the 370-155 and 165 as the
370-158 and 168 with a 36 to 54 percent increase in CPU price, and a
57 percent cut in memor^^ price. Such price manipulations would not
be possible if as much competition existed in CPU production as in
peripherals, or if the CPU's were made by a separate company from
the peripherals.

The ability of IBM to shift price between peripherals and CPU's
reduces the beneficial effects to customers of competition m peripherals,
as well as threatening the existence of the independent peripherals

A second aspect of market structure which accounts for actions in
the peripheral market is the disruption caused by equipment installa-
tion and removal, even when compatibility is not a problem. This
means that extra revenue can be obtained when a price cut is necessary
by requiring an equij^ment exchange to take advantage of tlie price
ciit. This tactic was used effectively with both the 2314 disk drive and
the 2420-7 tape drive. In both cases IBjM needed to make price cuts
in order to remain competitive. In both cases the price-cut products
were introduced as new products with identical performance specifica-
tions, the 2314 as the 2319 and the 2420-7 as the 3420-7. The price was
reduced 31 percent on the 2319 and 34 percent on the 3420-7. The
customer could only get the price cut by physically removing the old
product and installing the new one. The freight charges and disruption
involved, as well as the lack of information or lethargy on the part of
some computer managers, allowed IBM to have a competitive low


priced product while still receiving the higher rent irom many cus-
toniei-s for some time after the price cut.

The third structural characteristic that accounts for the actions of
IB^I toward the plug compatible manufacturers is the necessary time
lag between IBM introduction of a product and competitive copying
ofit in order to insure compatibility with constantly changing IBM
systems specifications. In the early days of peripheral competition
IBM o^•erestimated this lag and erroneously felt that it would make
the plug compatible competition a minimal threat.

Later IB!M introduced the fixed term plan and more rapid minor
product changes in order to capitalize on the time lag. When a new
product is introduced customers have little incentive not to accept it
on the fixed term plan because no competitive replacement was ready.
The heavy penalties for early termination under the fixed term plan
effectively reduced the competitors* jn-ospects to those customers finish-
ing a lease. By making rapid minor product clianges, such as switch-
ing control functions between drives and control units, IBi\I was able
to further reduce the plug compatible manufacturer's marketing effec-
tiveness without reducing prices.

The foregoing analysis is not meant to suggest that there was any-
thing wrong with IBIM reducing prices. Prices of peripherals before
the competitive companies entered tlie market were much higher than
necessary to give a normal return to capital. The proper functioning
of a conipetitive market, as well as teclmological progress, should a]id
did force the prices of peripherals down.

The problem is that customers did not get all the benefit of the lower
l^rices. Some of the savings from the peripherals was merely trans-
formed into a higher price for the CPU. The beneficial effects of
competition in peripherals were thwarted by the monopoly power in
the production of CPU's.

In considering possible methods of improving the performance of
the computer industiT it is necessary to concentrate on removing the
factors which have lead to the current problems.

Because of the brand loyalty which arises from integrated systems
production, splitting IB^I into several smaller integrated sj'Stems
manufacturers would be unlikely to completely solve the problem. The
greatest cuirent competitive emphasis in the industry has been from
the independent peripherals makers, not the smaller systems suppliers.
Although the industry would be more competitive with less dominance
by one firm, it would not be likely to reach the best possible perform-
ance. A better solution is to split IBM by functions; to make separate
companies out of the production of (T^U's. peripherals, maintenance,
and marketing functions. This would prevent monopoly power in one
segment from being spread into the other segments.

There would continue to be barriers to entry to the CPU business
because of the economies of scale in the production of systems software,
but that market power could not be enhanced by control over a wide
variety of^ othei' activities.

If such a split were made each segment of IBM would have to com-
l^ete on a fair basis with other companies. If IBM were really more
efficient it would continue to dominate. However, it could not use
power in one area to manipulate standards or price ratios in order to
fight competitoi-s in another area.

40-927 — 75 52


Neither could one activity subsidize another because tliey would be
separate companies.

The proposed type of organization already exists to some extent in
the minicomputer market, and appears to be very effective. Mini-
coniputers are generally sold as distinct components : CPU, software,
peripherals, et cetera. Although the major manufacturers of mini-
computers do provide complete systems, it is also common for systems
to be made up of several manufacturer's components.

Specialized companies have been formed to choose the appropriate
minicomjniter sysiems components, write software, and deliver com-
plete packages to customers. A wide selection of peripherals is offered
by many different manufacturers with interfaces for a variety of
minicomputers, rather than for only one manufacturer's product as in
the main systems industry.

The British National Computing Centre is developing a common as-
sembly language for a substantial number of minicomputers in order
to^ provide total program compatibility. Entry into all phases of the
minicomputer business has been easy, prices have dropped rapidly, and
the entire industry has been much more competitive than the main
systems industry, in spite of relatively heavy concentration.

Consequently, it appears that the lack of competitiveness in the main
computer industij is a result of integrated systems selling rather than
concentration in itself.

Senator Hakt. Thank you very much. You are a fast reader. I know
staff' has some questions.

Mr. O'Leary. Your proposed remedy contemplates keeping IBM
facilities for the production of CPU's and the development of systems
sofeware in one company, but what about software?

Dr. Brock. I proposed in the remedy that I wrote for the subcom-
mittee that the application software go with the marketing company.
Let me explain the reasoning behind tliat. I see right now a substantial
benefit to the user in being able to buy a totally integrated package
that will include rental on a system, consulting help, and installation
support, all the kinds of things that IBM currently provides.

However, it is not necessary that that same org'anization also do
the manufacturing. By keeping the application software with the
marketing company you wouldn't be breaking up any of the natural
economies that come to the user.

It is largely a matter of avoiding user disruption, from having to
deal with too many different people.

Mr. O'Leary. In that regard, Mr. Chairman, I think the record
should reflect that Dr. Brock was a consultant for the subcommittee
and prepared a paper for the subcommittee which should be sub-
mitted for the record.

Senator Hart. It will be received.

[The document referred to appears as exhibit 3 at the end of Dr.
Brock's oral testimony.]

INIr. O'Leary. Doctor, with respect to this company which produces
CPU's and developed systems software, how long would they be barred
from integrating into, say, application software, peripheral equip-
ment, and the like ?

Dr. Brock. I think that that is a subject that requires more studv.
^ly current idea would be that they probably would not be barred at


all. Or if at all, for a sliort period of time — say 2 years — in oi-der to
make sure tliat they actually operated as a viable business and didn't
just go out and remerge.

But the idea behind my reorganization proposal was, as I think
most of the other people here have indicated, to eliminate continuing
regulation so that it is largely a one-time thing. I can see a possible
need for a short bar, just until you get things established on the new

Mr. O'Leary. "N^Hiat about the production of components?

Dr. Brock. I believe tliat if you have the production of CPU's
sejiarated from the marketing function, there is probably a necessity
to keep the production of components with the production of CPU's.
The reason for that is simply the technological advances that are
going on.

We have seen over the past few years a rapid increase in the scale
of integration of components ; that is, the number of circuits put on a
single silicon chip. And since the CPU company would only be manu-
facturing CPU's. I can imagine that it is very likely that in the future
much of the design of the CPU will actually be contained within the

We, to some extent, can see that already with companies such as
Intel, which market what they call a microcomputer, which is effec-
tively all contained in one chip.

And, if things like that should continue, and it appears that they
will, there would be very little diiference between designing a CPU and
designing a comjoonent.

Mr. O'Leary. I assume that your plan involves some sort of agree-
ment and publication of standards to insure capability. Am I right
or wrong in that regard ?

Dr. Brock. You are right that it assumes that. I believe that this
plan would force the various manufacturers to agree on standards

One of our current problems in standards is that, and I am sure you
know, the American Standards Organization works on a voluntary
basis, and almost all product standards that have been developed are
done voluntarily by consensus agreement of the various manufacturers

That works fine so long as the manufacturers see it as being within
their own interests to have standards. Our current difficulty is that
IBM really does not see it within their own interest. This is not a
criticism of IBM. It just simply is not within their interest to have
generally accepted standards.

If you have the functions separated it would be within the interest
of all the companies to have standards because the standards would
increase the market for their different products. It will be of no
benefit to the new peripheral company to have a lot of peripherals
if there is not some CPU that they can attach to.

Mr. O'Leary. Isn't there always a little bit of concern on the part
of someone who believes in antitrust. They get the idea that com-
petitors can sit down and decide what the standards will be and what
the effect of that is on new technology ?

Dr. Brock. You mean that it would be kind of a collusive agree-
ment on the standards ?

Mr. O'Leary. Do you see some problems there ?


Dr. Brock. Well, I think that there are potential problems there,
but I really think they are largely taken care of by the current stand-
ards organization procedures. Tliat is, where you have everything
done above board, and the committees get together and publish all
the results, and so forth, and different manufacturei-s have the ability
to comment on them.

If you were to say, "AVell, let these four companies, get together and
produce the standards." That would be a seiious problem.

But if it is handled tlirougli the American KStandards Organization,
even if it has a connnittee of the manufacturers working on it, I
wouldn't expect the problem, but it is a potential difficulty.

Mr. O'Leary. How did the type of structure that you described in
tlie minicomputer industry come to be? Was it simply the absence of a
firm such as IB]\I which permitted this development, or what 'I

Dr. Brock. Yes. It was the absence of a fii'm such as IBM. I think
that it was largely from the history of the minicomputer market, in
the same way that we see tlie current structure of tiie general purpose
coruputer market growing out of the tabulating machine market. ^lini-
computers started out as very specialized machines, usually for indus-
trial process contix)l. So that rather than have a general operating
system and general software you had a specific problem, and they were

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 92 of 140)