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 70 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 70 of 140)
Font size
QR-code for this ebook


develop their own capability. As with peripherals, the new components company
would also sell to the general market and other components vendors, such as
Texas Instruments, would be free to sell to the successor computer manufacturers.
One would expect then, a three-fold gain for competition. Components manufac-
ture would have a major new competitor.* Other computer manufacturers would
have access to IBM components. Other component vendors would have available
a new source of demand in the successor computer companies.

(d) Marketing, service, software, maintenance and leasing divides into two
parts : that associated directly with the products of the successor computer com-
panies and that less clearly identified with these companies. A new entity (IBM-
I ) would be created to take over all these operations which entity would promptly
turn over all outstanding leases of computer systems still in production to the
successor computer companies responsible for marketing, integration and assem-
bly of such equipment. This would assume a flow of cash to each company from
the outset." Each new computer company would thenceforth assume responsibility
for maintaining and future marketing of their respective machines by hiring
away from the new entity the staff fonnerly engaged in these operations. Simi-
larly, each new company would be responsible for acquiring personnel to market
and support IBM software which existed at the time of such divestiture. Although
pre-existing packages would be made equally available to each successor com-
pany, development of new programs would obviously thenceforth be an individual
responsibility.

The new entity (IBM-I) would then be left with office buildings, perhaps a
residual marketing staff, and responsibility for the continued leasing and main-
tenance of IBM computers (and software) no longer in production. It would have
substantial revenues from out of production IBM computers. It could continue
as a leasing, maintenance and software company, but to ensure that it did not
dominate a new service or market, the new entity would be precluded from ac-
quiring and leasing new computers. As the old machines were returned or sold,
the assets could be liquidated and the cash distributed to stockholders.

(e) Development covers the design of new computers and is, of course, import-
ant to survival in the comi)uter industry. There is a tendency, promoted by IBM
publications, to think of IBM as one massive development entity. In fact, how-



** Givpii IBM's massive size, it is posslhlo that the new components company would
dominate the semiconductor indnstrv in revenues and value of shipments. If this appears
to he the ease, further division of IBM's Components Division is possible, since several
IBM locations currently produce IBM components.

» To the extent that a successor company's major divestiture assets were in the
peripheral areas (e.p., IBM-B), any inequity resultinc from a transfer of peripheral
rentals to the company having systems responsibility could he offset with present IBM cash
reserves or other assets held by the "interim" entity.



5504

ever, such key efforts as the design of the System 360 line was done by several
teams at different locations and IBM has its present R&D activity located at at
least 24 separate laboratories. One — Manassas, Va., — serves the components
division and would be allocated to the new components company. At each of the
principal plant locations for the seven hardware companies there presently exist
major development groups, as follows :

(1) IBM-A : Rochester, Minn. ;

(2) IBM-B : San Jose, Calif. ;

(3) IBM-C : Endicott, N.Y., two labs ;

(4) IBM-4 : Poughkeepsie, X.Y., three labs ;

(5) IBM-E : Kingston, N.Y. ;

(6) IBM-F : Owego. N.Y. ; and

(7) IBM-G : Raleigh, N.C.

The development groups are concerned primarily with the products of the
plants near them, so that there would be a natural fit in assigning such groups to
adjacent companies. There would remain a few other laboratories that serve a
broader corporate function — such as the Thomas J. Watson Laboratory concerned
with basic research. These might either be allocated to those companies so as to
equalize assets and technical manpower of successor companies or, preferably,
transfer or sale of such laboratories to some sort of independent research founda-
tion could be encouraged.

2. Evaluation of the plan

(a) The divestiture plan results in economically strong companies. — The sur-
vey above, operation by operation, indicates that each successor company retains
a feasible set of activities appropriate for an independent company. Focusing on
these companies as a whole, each would still be absolutely large. Their annual
revenue is hard to predict from published IBM data, but the average would be
well in excess of 500 million dollars. Another measure of size is the "relative"
market share in the computer market. Of course, ''relative" market share depends
upon the market being defined — for example, in the computer systems market, the
shares of IBM-C, IBM-D, and IBM-E would be larger than IBM-A and IBM-B.
while the latter companies would be considerably more powerful in their respec-
tive peripheral areas. However viewed, each company's "share" would be at least
twice that of the largest non-IBM company.

Furthermore, the companies would start with substantial financial assets in
their share of a multii-billion dollar lease base. (Any sharp inequality in the di-
vision of the lease base could be equalized by the differential division of IBM's
working capital) .

The resulting companies would, of course, be much less vertically integrated
than the present IBM. But, as indicated above, the degree of vertical integration
would remain comparable to that of other computer manufacturers. There might
be particiular facilities imbalances and the management of particular successor
companies might regard some greater degree of vertical integration as desirable.
But such shifts are relatively easy. IBM now spends about $400 million annually
on plant and equipment and such a flow of investment funds would be divided
among the successor companies for new investment. The high rate of growth of
computer sales — over the ten years from 1959 to 1969 the market growth averaged
over twenty percent annually — further enhances the opportunity of the new com-
panies to alter their product lines. In addition, existing computer manufactur-
ing plants are relatively adaptable to different kinds of manufacture. The major
capital items in computer manfacture (except for components) are floor space
and easily movable test and assembly equipment. Plants can be readily adapted
to different products.

Although the new companies initially will have a limited computer line, the
fact that they would be compatible would be an initial benefit to the U.S. national
defense posture in that compatible computers would be available from multiple
sources. In addition, over time, the successor companies could diversify into wider
product lines if they so choose and. in so doing, have the financial advantages and
manufacturing flexibility de.scribed above for vertical integration.^" In sum, the
successor companies should not only be financially viable but economic successes.
Each starts with an established product, proven in the market place and with
facilities for its manufacture. Each will have access to IBM's present peripherals



^fl Of pourse. acquisitions by successor companies of each other or other companies would
he prohibited for a number of years or, at a minimum, prohibited without prior approval of
the Department of Justice.



5505

and components, together with the opportunity to buy such items on the outside
market. Each will have the financial security from a big lease base. Each will
have a development group and the financial resources to extend its product line
horizontally and vertically. Each will have a substantial fraction of IBM's exten-
sive marketing and maintenance force. .

(&) The divestiture plan represents a significant step forivard in achieving
effective competition.— The plan would still leave IBM products as virtually
dominant in separate computer (and peripheral) sub-markets. Thus, the suc-
cessor computing company would have initially the same market. While high
market share in a particular product is a significant source of IBM's market pow-
er, IBM's dominance is intensified by the compounding of market power through
its position in all the various submarkets and combining such positions with
(1) massive financial resources from a huge lease-base; (2) a marketing force
that blankets the data processing fields ; (3) a complete product line so that cus-
tomers with changing data needs will continue to be linked to IBM; and (4) a
commanding position across product lines as well as with peripherals so that IBM
can largely dictate the nature of technological change to its own advantage.

The divestiture plan strikes at these four sources of market power, even though
it fails initially to modify the fifth— large share in a specific submarket. Finan-
cial power will be reduced because the successor computer companies will no
longer be able to exploit the massive flow of funds from one lease base. The mar-
keting force will no longer be ten-to-fifteen-fold the size of competitors and blan-
ket every potential customer. And this marketing force will no longer be able to
direct each customer, as his data processing needs expand, from one IBM prod-
uct to another. Rather, in such situations the customer will evaluate on their
merits the products of now independent IBM successor companies relative to those
of competitors.

The adverse impact of IBM's monopoly power upon technological change and
innovation will be dissipated. With several successor companies, it will be no
longer possible for a single firm to paralyze the pace of change in the computer
industry by announcements of proposed sweeping product changes or impose its
standards "on the entire computer manufacturing and user community. Alterna-
tive possibilities for procurement will serve both the immediate and long term
national interest. To be sure, compatibility is important but absent IBM's mas-
sive size, one would expect the initiative for change would be more widely dis-
tributed — both among successor companies and other computer manufacturers.
Various companies would announce changes and those that met the market test
as substantial progress and were followed by competitors and widely accepted
bv users would become the industry norm. Such a freeing up of technological
change would yield a more desirable rate of technological change simply because
it would be market tested, rather than the fiat of one organization.

Finally, the divestiture plan does oifer the possibility for correction of the
dominance by successor companies in particular submarkets. To some extent,
there will be immediate competition between the successor IBM companies them-
selves, since product lines overlap — e.g.. a "large" configuration of a 360/30
system competes directly with a minimally configured 3607-^0. Also, as indicated
earlier, it is likely that successor companies will broaden their product lines and
so further compete with one another. The availability of former IBM peripherals
and components to newcomers and other computer manufacturers on equal terms
with IBM successors will encourage the rise of additional competition. The
removal of the advantages of IBM's overall size and extensive product line will
enable the present competitors of IBM to compete more effectively in each sub-
market, resulting in an increased market position of competitors.

(c) The divestiture plan does not entail any losses in social economies of
scale.— In evaluating the divestiture plan, it is important to distinguish between
"social" and "private" losses resulting from the removal of IBM's massive scale.
Social economies of scale are true gains for the economy as a whole; private
economies of scale are those accriiing to the benefit of a single firm and not to
the economy. Public policy is concerned only with social economies. And the
divestiture plan involves no loss in social economies.

As indicatefl above, the divestiture plan does not divide either individual plants
or products. Hence the only conceivable source of the loss of economies of scale
are those operations that encompass several products and plants : namely, financ-
ing, the use of common products such as peripherals and components, marketing,
providing associated customer services, and development of new products.

The ma.ssing of financial power is one example of a private economy of scale.
It is an advantage to IBM because it permits it to concentrate development ex-



5506

penditures on certain products and gain a competitive edge. It is not an advan-
tage to tlie economy because the competitive edge for one company does not
necessarily speed up teclinological progress for tlie economy. Ratlier tliere is
considerable evidence suggesting that IBM, while massive in its resources, has
not been the technological leader in computer development. IBM competitors,
with a lower level of spending, have made many important contributions to the
industry's progress.

IBM's vertical integration into peripherals and components again appears to
be primarily a private economy of scale. IBM has a private advantage in that its
extensive line of peripherals are unavailable to other computer manufacturers
except at the retail price. Yet since these terms restrict sharply the use of IBM
peripherals by other computer manufacturers, IBM's private gain is a social loss.
Since any true scale economies operable with respect to a particular peripheral
product would remain intact, the future availability of each such product to all
computer companies could thus well convert IBM's private economy to a social
economy. Moreover, given the size and growth of the present computer market,
there is evidence to suggest that any true scale economies are maximized at levels
far below IBM's current production. The same situation is applicable with re-
spect to components.

An extensive marketing service is again a private economy of scale for IBM,
giving them a competitive edge. Whether or not one massive marketing service
for a complete computer line takes less real resources in terms of manpower than
several separate sales forces for the successor companies, and so becomes a social
economy, is possible, but unlikely since the computer is not a repeat item sold
in considerable volume but, instead, each procurement must be individually
tailored and marketed. IBM's present marketing organization is generally or-
ganized according to "industry classification," which is often reflected in com-
puter system size ; such economies would thus be preserved. Moreover, such man-
power economies as may exist with respect to "repeat" customers, due to cus-
tomer loyalty or the exploitation of other competitive advantage resulting from
having a machine "on the inside," would — to the extent worthy of considera-
tion — be preserved and reflected in favor of any previously "successful" vendor
regardless of its size.

Maintenance and software are perhaps even less clear a case of minimal econo-
mies of scale. These services are now generally organized by IBM in teams
specializing in various computers or industries and so the economies of size are
limited to such relatively minor items as common office space. Much applications
software can be used by several computers but this work is now carried on by
independent software companies with some success. Also much of the existing
application software would remain available as a separate product as the result
of the "unbundling" of computer pricing.

Development is similarly unlikely to reflect significant social economies of
scale. As indicated above, IBM's development groups for specific computers and
for components are geographically separated and function in large part as
separate teams. Moreover, the fact that other computer manufacturers with
less extensive development groups have been able to keep pace and often lead
IBM's technological progress proves feasibility and suggests that the efficiency
difference between very large scale and more modest scale development is not a
decisive one for the rate of technical progress. Finally, due in part to its large
lease base, IBM has tended to announce and deliver new products in a manner
designed to "protect" its lease base and to maintain a "desired" growth rate,
thereby denying to society the benefits of competitors' advances as well as such
benefits as could even theoretically derive from its own massive development
efforts.

(d) The divestiture plan is equitable to IBM's stockholders, executives and
employees. — Although the Supreme Court has recognized the minimal considera-
tion to be given "private interests" in framing effective remedies to redress anti-
trust violations, U. S. v. duPont, 366 U.S. 316. the instant divestiture alternative
is fair to the interests of such parties. The IBM stockholders will receive stock
in all the successor companies in exchange for their present IBM holdings. In
aggregate, the same financial and physical assets will be represented by the new
stock. Stock in IBM, as opposed to that in successor companies, does provide the
financial security of diversification by reflecting wide product line and a wide
range of "data processing" activities. Yet stockholders that continue to want
.such diversification can leave their holdings distributed among successor compa-
nies. Those that wish to concentrate on particular activities can rearrange their
portfolio to do so. IBM stockholders can thus select varying mixes of activities



5507

if they so wish and. in addition, will be able to select as well as invest in
different managements of the varions successor companies.

Against this gain must be set a loss in the portion of the value of IBM stock
represeiiting monopoly profits. This is an inevitable cost of a public policy directed
at eliminating monopolization. Yet in the high growth computer industry such
losses should be modest and transitional. While it is obviously foolhardy to
predict the course of post-divestiture stock values, the risks seem less than in
other major divestitures.

As for IBM executives, divestiture will convert many IBM divisional execu-
tives now serving as second or third level management into top management
since there will be several top management groups rather than one. The execu-
tive losers from divestiture will be confined primarily to the small group of top
executives who will forfeit the power and prerogatives of administering one
of the world's largest private enterprises.

The successor companies would assume all the obligations of the pension
and benefits of the executives and employees in facilities they are assigned.
The special leasing company described above (IBM-I) will assume a cor-
responding obligation for branch and central oflice employees who are not hired
by the successor computer or components companies. Tliis company will have
both the extensive ofl3ce buildings owned by IB:M throughout the country and
the leases on older computers as assets with which to meet obligations to these
employees.

To sum up, the divestiture should preserve much of the significant equities
of IBM stockholders, executives, and employees.

B. Divestitures in related markets

Tlie preceding measures are generally directed at the domestic computer
hardware market. There remain other components of IBM that should also
l)e divested to IBM stockholders on the grounds that: (1) They can be eco-
nomically viable companies in their own right; (2) assigiiing any one to a
particular successor computer company would make successor companies unequal
in size ; and (3) the particular operations are so large in their respective markets
and so closely related to the marketing of computers that owmership by such
a computer company jeopardizes competition in the computer market.

Four IBM operations particularly fit these conditions : The Service Bureau
Corporation (SBC), the Office Product Division, Science Research Associates,
and the World Trade Corporation.

1. The divestiture of SBC

I'nder the 1956 Consent Decree, SBC is supposed to be operated as an arms-
length corporation. Yet there is considerable evidence that its computing serv-
ices promote IBM's computer business through "customer captivity" as well
as through offering IBM opportunities to institute price discrimination through
the free usage of computer time, free manpower or software support, and buy-
backs of computer time. Furthermore, it provides a sheltered market for IBM
equipment. As a result, SBC is another source of IBM market power in the
computer business.

The divestiture of SBC, through a stock distribution, would prevent this
source of market power in the computer market from being utilized by any
successor computer company.

2. The divestiture of the Office Products Division

IBM's office products operations are organized as a separate division, with
its own plants, development group, and executive organization and with revenues
in the hundreds of millions of dollars. The principal product of this division is
office typewriters in which IBM has a substantial market share. In addition,
IBM recently announced entry into the office copier market. This division should
have no difficulty operating as a separate company. AYhile at present the inter-
relationships between office products and computers are not extensive, the coupling
of magnetic-tape to typewriters and copiers could give a computer manufacturer
with a large share of outstanding office typewriters and an entry in the copier
market a powerful position in the computer market. Accordingly it is proposed
that the Office Products Division be divested into one or more separate companies.

3. The divestiture of Science Research Associates (SRA)

SRA is a recent IBM acquisition which develops and markets computer ap-
plications in the educational area. It is an asset in the marketing of computers
which should not be assigned to any one particular successor company. Since



5508

SRA has previously been successful as a separate company, it can be easily
reestablished in this status.

4. The divestiture of the World Trade Corporation

IBM has concentrated its foreign operations in a single subsidiary — IBM
World Trade Corporation. This subsidiary can be a separate company for it
already has its own overseas marketing force, extensive manufacturing and
development activities, and a distinct set of executives. As a separate company,
it would have over a billion dollars in annual revenues.

As a separate company, however, at least initially. World Trade would have to
buy some products it does not manufacture from successor companies. And it
is likely to need to license technological know-how from successor companies
to stay abreast of changing technology. But these relationships would have a
somewhat lesser impact on domestic competition than the existing ones.

The impact of a dominant position in many markets abroad would presumably
be diffused among several successor computing companies. At the same time, the
impact on American trade from divestiture would be minimal since World Trade
in effect now buys many of its products from IBM's domestic operations. Over
the longer run. World Trade could buy from other American computer manu-
facturers if their products proved superior. Thus World Trade's commanding posi-
tion in various foreign markets might become available to other American com-
puter manufacturers.^^
C. Proscriptions against certain anti-competitive practices

Since the divestiture recommendations are fairly comprehensive, the scope of
injunctive relief can be limited to a few key items. Nonetheless, since the IBM
successors will have, at least initially, very substantial market shares in particu-
lar computer and peripheral submarkets, four major provisions are required :
limitations on price discrimination, limitations against paper machines and the
like ; requirements on product disclosure, and provision for royalty-free licensing.

1. Price discrimination

Price discrimination is a traditional tactic of a dominant firm to gain and to
maintain market power — cutting prices for a group of customers or with respect
to products where competition is vigorous while maintaining high prices else-
where. IBM has made extensive use of price discrimination. Given IBM's record,



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 70 of 140)