LETTER TO JOSEPH LEVINE, FROM DOUGLAS H. GINSBURG

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP05T02051R000200270001-9
• Release Decision: RIFPUB
• Original Classification: K
• Document Page Count: 9
• Document Creation Date: December 22, 2016
• Document Release Date: November 28, 2011
• Sequence Number: 1
• Publication Date: February 8, 1984
• Content Type: LETTER

Declassified and Approved For Release 2011/11/28: CIA-RDP05TO2051 R000200270001-9
C'initcb Q)tatc Dtpartmrnt of htlstitt
WASHINGTON, D.C. 20530
February 8, 1984
Joseph Levine, Esquire
office of the General Counsel
Department of Commerce
Washington, D.C. 20230
We have reviewed the draft legislation, provided to us in.
December 1983 and dated October 26, 1983, "to establish a
system to promote the use of land remote-sensing satellite
data, and for other purposes." We have the following general
comments about the competitive implications of the bill.
I. Summary of the Biil
The bill provides for the transfer to the private sector of
the responsibility for operating land remote-sensing
satellites, a function now performed by the "Landsat" system
operated by the National Oceanic and Atmospheric
Administration. Title I of the bill states that it is the
policy of the United States to commercialize those
remote-sensing functions that lend themselves to private sector
development. Section 103(b). The Government's right to
acquire and to disseminate land remote-sensing data on a
non-discriminatory basis must, however, be preserved.
Section 103(a).
In general terms, Title II of the bill provides for the
Secretary of Commerce to contract with a private party to
operate the Government's current Landsat system. Title III
contains certain interim provisions designed to assure
continued availability of the data through private sector
operation of new privately-owned land remote-sensing systems
after the "practical demise" of the Landsat system. Title IV
sets the conditions under which private operators of future
-.d re:iote-sensing systems will be licensed. Title V provides
i Declassified and Approved For Release 2011/11/28: CIA-RDP05TO2051 R000200270001-9
 Declassified and Approved For Release 2011/11/28: CIA-RDP05T02051 R000200270001-9
for continued federal research and development. Title VI
requires that in all circumstances users are to have
non-discriminatory access to the data. Although by its express
terms the bill applies only to the land remote-sensing system,
and not to the meteorological system, Title VII specifically
prohibits sale of the Government's meteorological system.
Specifically, the contract provided for in Title II will be
awarded on a competitive basis, and will be for "such limited
duration as may be necessary" to fulfill the conditions of the
sale of the operational rights. Section 201(c). Any sale of
the Landsat data by the contractor must be through an
organizationally separate unit that acquires the data on a
non-discriminatory basis at non-discriminatory terms.
Section 202(b). The Secretary of Commerce shall award the
contract on the basis of financial return to the Government,
technical competence, ability to satisfy all conditions of the
sale, marketing ability, absence of conflicts of interest
affecting access to the data, and ability of the contractor to
effect a smooth transition from Government to private
operation. Section 203(b). Operation thus shall be conducted
in a way that will prevent any "competitive advantage to the
operator." Section 202(b).
Under Title III, a contract for the purchase of data
required by the Government is to be awarded to a private party
or parties by an "open competitive process". Section 302(a).
The Government is to prepay to the operat(-,.c a portion of the
capital cost of the system, Section 302(b), and to enter a
six-year requirements contract, Section 302(c). Surplus
Landsat equipment can be used or purchased by the operator,
Section 302(d), who may also use other civilian satellites upon
immediate reimbursement of related costs. Section 302(e). The
United States Government, but not other buyers, will recover a
five percent rebate on purchases of data. Section 303(b).
Again, there is to be no "competitive advantage" accruing to
the operator. Section 303(c).
The licensing provisions in Title IV require the Secretary
of Commerce to license qualified private parties, consortia of
private parties, and consortia of private and governmental
parties wishing to operate their own land remote-sensing
systems. Section 401. Licensees must operate any system under
the following conditions: (1) operation must preserve and
promote United States national security; (2) data must be
available to all potential users on a non-discriminatory basis;
(3) the system shall be administered by a central entity (in
the case of a consortium); and (4) the license shall not
protect the holder from "fair competition" from other
~crnaees. Section 402(b).
Declassified and Approved For Release 2011/11/28: CIA-RDP05T02051 R000200270001-9
 Declassified and Approved For Release 2011/11/28: CIA-RDP05T02051 R000200270001-9
II. Discussion
As a preliminary matter, we question the threshold
c.termination to commercialize the land remote sensing system.
A private firm would choose to operate the system only if it
believed it could derive sufficient revenues from buyers of
land remote-sensing data to more than offset its costs. Thus,
if the initial fixed costs of investing in the system were very
high, high prices would have to be paid by data users to make
commercialization feasible. Since users pay very little at
present for these data, it is not at all clear whether the
market would be able to support a private Landsat system absent
large continuing government subsidies. 1/ Even assuming,
however, that potential users valued Landsat data highly, the
system operator could sustain profitable prices only if it
could encrypt the data and thereby prevent satellite signal
piracy.
The questionable outlook for commercialization is further
clouded by Sections 103, 402(b), and 601(a) of the proposed
bill, which require that all potential users be granted access
on a "non-discriminatory" basis. It may be that only a system
of discriminatory pricing -- under which different users are
charged different prices according to the value they place on
the data, and thus the price they are willing to pay for it --
would yield sufficient revenues to cover costs and make
commercialization attractive. 2/ Under such a system revenues
derived from customers willing to pay higher prices might
substitute for the subsidy now provided by the Government.
This suggests that, in order to maximize the likelihood of
successful commercialization without a government subsidy, the
bill should be modified to omit any requirement that the prices
charged for data necessarily be "non-discriminatory."
1/ It is not clear that total demand for remote-sensing data
is sufficient for full commercialization of the land
remote-sensing system.' See Report of the National Oceanic and
Atmospheric Administration in Response to P.L. 97-324 (January,
1983); The National Academy of Public Administration, Space
Remote Sensinc and the Private Sector: An Essay (March, 1983).
2/ For a good description of price discrimination in
"decreasing cost" industries, see generally A. Kahn, 1 The
Economics of Regulation 123-158 (1970). In order for
discriminatory pricing to be effective, the system operator
~,ould have to be able to prevent arbitrage -- the resale of
data by customers charged a low price to customers charged a
`._ch price. Arbitrage might be prevented by enforceable
contractual clauses prohibiting such resales.
Declassified and Approved For Release 2011/11/28: CIA-RDP05T02051 R000200270001-9
 Declassified and Approved For Release 2011/11/28: CIA-RDP05T02051 R000200270001-9
If foreign policy considerations require that foreign
governments' concerns about the availability of data be
accommodated, a requirement that foreign governments be given
access to these data on equal terms could be added. 3/
We also wish to point out that, assuming the aggregate
social benefits of operating Landsat exceed the costs, and that
commercialization is feasible, the transfer of the Landsat
system from public to private hands may inefficiently restrict
the usage of remote-sensing data.' Economic welfare is
maximized when a good is sold to all buyers willing to pay at
least marginal cost -- the cost of supplying one more unit of
the good. If the marginal cost of supplying remote-sensing
data is extremely low, the socially optimal use of these data
is assured by charging a correspondingly low price for them --
as the Government may be doing at present. If, under private
ownership, higher prices were charged (to ensure that total
revenues covered total operating costs), some current users who
value the data at an amount equal to or greater than marginal
cost -- but at less than the privately charged price -- would
be inefficiently deprived of the data. 4/
In addition, while the proposed bill does not expressly
suggest that private Landsat operators should be regulated, it
should be noted that its repeated references to "non-
discrimination" (see text at p. 3, supra) evoke a statutory
3/ Under such a "most-favored nation" clause, the treatment of
foreign government users would not necessarily be linked to the
terms afforded domestic purchasers of data. Of course, due to
commercialization, foreign governments would be charged higher
prices after the demise of Landsat than they presently pay for
data -- unless the United States Government explicitly
subsidized private sales to foreign states.
If the treatment of foreign private users raised no foreign
policy concerns, they could be sold data under the same terms
as domestic users. It is, of course, possible that foreign
governments might request preferential treatment for
politically influential private users. United States
Government agreement to such special treatment would, once
again,, require the payment of federal subsidies to the Landsat
operator or operators.
4/ This inefficiency probably would exist to some extent
whether the land remote-sensing industry became competitive or
monopolistic. If several competitors could profitably serve
the market -- as the bill apparently assumes.-- competition
somewhat constrain prices (and, incidentally, limit the
use of price discrimination). Nevertheless, price probably
woucld not reach the very low level of marginal cost, and some
:o old-be users would be denied access to the data,
notwithstanding their willingness to pay an amount equal to or
m_rg nal cost.
Declassified and Approved For Release 2011/11/28: CIA-RDP05T02051 R000200270001-9
 Declassified and Approved For Release 2011/11/28: CIA-RDP05T02051 R000200270001-9
term of art that historically has been applied to justify rate
of return regulation in areas such as telecommunications.
.ccordingly, there is some the danger that "non-discrimination"
might be cited to support the imposition of rate of return
regulation on commercialized satellite systems. Unless
provision of satellite data involves a natural monopoly (see
n. 4, supra), however, or there is some other "market failure,"
there is no economic rationale for regulation of land remote
sensing at all.
Moreover, even the existence of natural monopoly would not
necessarily justify regulation. Rate of return and other forms
of natural monopoly regulation are economically justifiable
only if regulatory costs outweigh regulatory benefits. 5/ In a
4/ continued
If, instead, the industry became a natural monopoly
(because one firm could serve the entire market at lower cost
than multiple firms), the extent of the inefficient reduction
in output would hinge on the pricing scheme adopted. If the
satellite operator were restricted to charging a single
monopoly price, many users would be inefficiently deprived. of
data. If, on the other hand, the monopolist were allowed to
price discriminate, and charge different users different
prices, inefficient output restriction would be less
pronounced. The efficient output level would be achieved,
however, only if the monopolist could "perfectly discriminate"
-- charge each potential user the exact amount it would be
willing to pay for data. Such an outcome would be extremely
unlikely.
5/ Regulation is justifiable generally where market failures
result in inefficient levels of production of goods and
services or no production at all. Some forms of regulation,
however, have counter-productive tendencies. Rate base and
rate of return regulation reduces incentives for efficiency,
and emphasizes service competition at the expense of price
competition. Since a "normal" rate of return is guaranteed on
service expenditures that are included in a regulated firm's
rate base, the regulated company has an incentive to
"overinvest" in services, with little regard to cost. In
addition, rate regulation consumes significant resources of
both the regulated firm in supporting rate requests and the
regulating agency in sorting out the large volume of data
submitted with these rate requests. A rate regulatory process
can also provide a licensee with incentives for inefficiency in
i=s investments and operations, and for rigidity in its rate
st_uctures any service offerings. -See Joskow & Noll,
Peculation in Theory and Practice: An Overview, in G. Fromm
Studies in Public Regulation 1 (1981); S. Breyer,
=eculation and Its Reform 36-59 (1982).
I Declassified and Approved For Release 2011/11/28: CIA-RDP05T02051 R000200270001-9
 Declassified and Approved For Release 2011/11/28: CIA-RDP05TO2051 R000200270001-9
technologically dynamic area such as land remote-sensing, the
static nature of regulation may actually cause more damage by
retarding innovation than would be caused by allowing a
temporary monopolist to set prices without constraints. 6/
We are also troubled about the possibility that the
contractor that assumes responsibility for supplying the
Government with data during the six-year interim period will be
provided, as a result of the contract, with a significant
subsidy that will lessen the possibility of competitive entry
into the market (assuming no natural monopoly). Section 302
provides that a party or parties contracting for the provision
of data to the Government must be capable of providing, at a
minimum, the amount of data used by the Government during
fiscal year 1983. To encourage such production. Section 302
also provides for the Government to prepay a portion of the
capital cost of providing this capability. This prepayment
subsidy would be at least partially repaid to the Government
through the five percent rebate on the Government's data
purchases, but it might not be fully repaid during the contract
period. The award of a six-year Government requirements
contract likewise gives the contractor a competitive advantage
over any potential entrants into the market.
While we recognize the possible need for a subsidy to
assure that land sensing data will be available, we are
concerned that such a contract may make it impossible, as a
practical matter, for any competing systei..3 to be formed.
Under such a contract, the subsidized system may be little
different, from the point of view of competition, from a
Government-owned and operated Landsat system. The contractor's
incentives to innovate and to provide better and cheaper
services may be significantly dampened due to the protected
position it holds by reason of both the subsidy and the
Government requirements contract it receives.
In order to minimize the competitive advantage of the
initial contractor the term of the contract should be subject
to downward negotiation between the Secretary of Commerce and
the prospective contractor. We recommend that the statutory
6/ There are currently prospects for competition from foreign
(e.g., French and Japanese) launched and operated systems.
These systems may not be-profit-making enterprises-, and may
involve significant government subsidies. They may,
nevertheless, represent a competitive check on the exercise of
monopoly power by a United States firm. _
Declassified and Approved For Release 2011/11/28: CIA-RDP05TO2051 R000200270001-9
 Declassified and Approved For Release 2011/11/28: CIA-RDP05TO2051 R000200270001-9
language be amended to provide that the six-year period is an
outside limit for the length of the contract. Likewise, five
.percent should be made the minimum discount to be given to the
United States Government. A higher percentage discount would
make it more likely that the Government prepayment subsidy
could be repaid or substantially reduced by the end of the
contract, and less likely that a competitive advantage will
continue after the period of operation as an interim system.
Moreover, we believe that the bill should not require the
contractor to sell value-added derivative data through a
separate subsidiary. 7/ See Section 202(b). A separate
subsidiary is justifiable only as'a means to prevent a
regulated monopolist from using profits in regulated markets to
cross-subsidize low cost sales in unregulated markets and
thereby disadvantage competitors. The justification for
utilizing a separate subsidiary is entirely lacking in the case
of the land remote-sensing system, if. as the proposed
legislation appears to contemplate, this field is not a natural
monopoly and will not be price-regulated.
The foregoing discussion highlights the broad competitive-
issues raised by the bill. There are also more specific
definitional questions raised by the bill as currently
drafted. Section 203(b)(5) imposes as a criterion for
selection as a provider of land remote-sensing data the
"absence of any conflicts-of-interest which could inhibit
non-discriminatory access to such data." 'That language could
conceivably be construed to bar the system operator from
providing value-added services regardless of whether it is a
regulated monopolist. The general requirements for separation
of activities in Section 201(b) and non-discriminatory access
in Sections 103(a), 402(b), and 601(a) may not be sufficient to
preclude such an interpretation of the statute. Second, it is
unclear what is meant in Sections 202(b) and 303(c) by the
requirement that licensed operators have no "competitive
advantage." That language might be read to suggest that
value-added services could be barred or limited and to sanction
specific licensing conditions not enumerated in the statute.
The-bill could be clarified by affirmatively stating that any
operator should be permitted to provide value-added services.
Additionally, the criterion in Section 402(b)(4) that no
license shall protect the holder from "fair competition" should
be clarified, so as to fulfill two distinct and important
7/ See attached letter from Wayne D. Collins, Deputy Assistant
:attorney General, to Joseph Levine, Department of Commerce,
rust 25, 1983.
Declassified and Approved For Release 2011/11/28: CIA-RDP05TO2051 R000200270001-9
 Declassified and Approved For Release 2011/11/28: CIA-RDP05TO2051 R000200270001-9
functions. First, specific language should be added to insure
against any implicit repeal of the antitrust laws. 8/ Second,
Section 402(b)(4) should preclude an unduly restrictive
licensing process that might exclude potential entrants. The
section should specifically provide that the Secretary is not
to grant exclusive licenses, or to consider the economic
effects of the entry of additional firms into the land
remote-sensing business. Thus, replacing the "fair
competition" language with a provision clarifying the
Secretary's powers, and inserting an antitrust savings clause
would more effectively serve the purpose of preserving
competition.
Finally, the participation of federal agencies in
consortia, even with private parties, raises both substantive
and definitional problems. Section 403(b)(3) limits such
activities to those that "will not compete with U.S. private
sector activities." However, both operational and value-added
services of such mixed public and private sector consortia
would seem inherently likely to compete with purely private
sector activities in the land remote-sensing industry.
Moreover, the possibility of mixed consortia may in itself
distort the incentives of the private sector to develop land
remote-sensing services. Potential private entrants would be
deterred if they feared that the participation of federal
agencies might entail public subsidies for the consortium with
which they would have to compete. This suggests that
individual Government agency needs should De met contractually
through the private sector alone, not through Government
partnership arrangements, so that all potential providers may
have the opportunity to compete for any necessary Government
subsidies. On balance, permitting Government participation in
consortia appears unnecessary and possibly damaging to
competition.
8/ We would be happy to assist in the drafting of such a
provision and the accompanying explanation, which should make
clear that the antitrust laws remain fully applicable except
insofar as the legislative scheme or essential fairness
considerations require a limited exemption. Although we have
not settled on final wording, such a provision might provide
that "Nothing in this Act shall be deemed to create an
exemption or defense to any action under the federal antitrust
laws, as defined in Section 1 of the Clayton Act (15 U.S.C.
12), or the Federal Trade Commission Act (15 U.S.C.
41-58); provided, that a contract awarded in accordance with
Section 302 of this Act shall not be deemed to violate such
statutes.
Declassified and Approved For Release 2011/11/28: CIA-RDP05TO2051 R000200270001-9
 Declassified and Approved For Release 2011/11/28: CIA-RDP05TO2051 R000200270001-9
As reflected above, we believe, there are substantial
competitive and definitional questions raised by the bill as
drafted. We would be pleased to consult further with you on
possible modifications of the bill, should you so desire.
Douglas H. Ginsburg
Deputy Assistant Attorney General
Antitrust Division
I Declassified and Approved For Release 2011/11/28: CIA-RDP05TO2051 R000200270001-9