FTC Staff Comment Before the FCC In the Matter of Amendment of Part 74 of the Commissions Rules Concerning FM Translator Stations

V880052
UNITED STATES OF AMERICA
FEDERAL COMMUNICATIONS COMMISSION
GUit';t.i~SmN AUTHORIZED
In the Matter of Amendment of
Part 74 of the Commission's
Rules Concerning FM Translator
Stations
)
)
)
)
)
)
MM Docket No. 88-140
RM-5416
RM-5472
~ _~_~ 	August15, 1988 	_~
,
COMMENTS OF THE STAFF OF THE BUREAU OF ECONOMICS OF
THE UNITED STATES FEDERAL TRADE COMMISSION

I. 	Introduction and Summary
In a Notice of Inquiry (NOI) released on June 2, 1988, the Federal
Communications Commission (FCC) solicits comment on its regula tions
governing the use and licensing of low-power FM commercial translators
(transla tors).l 	In these comments, the staff of the Burea u of Economics of
the Federal Trade Commission suggests that granting increased flexibility in
the use of translators may benefit consumers by permitting greater
competition for the patronage of listeners and the development of additional
listening options for consumers. 2 	We also suggest that the FCC weigh these
benefits against two potential costs of greater flexibility: the possible
inefficient use of the spectrum by translators and the current absence of
public interest obligations on translator licensees.s
Additionally, regardless of whether it decides to permit greater
fkx.ibility in translator use, the FCC also solicits comments on whether
licenses for translators should be awarded-via a lottery. ~In contrast to
lengthy comparative hearings to resolve mutually exclusive license
applications, a lottery may result in a more rapid delivery of translator
services to consumers, and we, therefore, suggest that a lottery may advance
the interests of consumers. 	To ensure that each license goes to its highest
1 	Federal Communications Commission, In the Matter of Amendment of
Part 74 of the Commission's Rules Concerning FM Translator Stations, 1\.1M
Docket No. 88-1~40, RM-5416, and RM-5472 (released June 2, 1988).
2 	These comments are the views of the staff of the Bureau of
Economics of the Federal Trade Commission. 	They are not necessarily the
views of the Federal Trade Commission or of any individual Commissioner.
Please contact staff economist Robert Rogers at (202) 326-3382 should you
have any questions regarding our comments.
3 	In these comments, we do not estimate whether these costs outweigh
the benefits of greater flexibility of translator use.

valued use under either assignment method, we also suggest that translator
licenses be transferable within a short time after receipt.
.
The remainder of this comment is organized as follows. 	Section II
describes the interest and relevant experience of the FTC staff. 	Section III
outlines the development of current FCC translator policy and the adoption
of the NOI. 	Against that background, Section IV considers the kinds of
consumer benefits that may flow from permitting flexibility in the uses of
translators. 	Section V considers some" of the costs the FCC perceives may
flow from this flexibility. 	Section VI discusses the possible consumer
advantages of a lottery as the mechanism to assign translator licenses.
Finally, Section VII summarizes our comments.
II. 	Interest and Experience of the FTC
The staff of the Federal Trade Commission, upon request by federal,
"" state, and local governmental bodies, comments on regulatory proposals that
may affect competition and increase costs without providing countervailing
benefits to consumers. 	The Federal Trade Commission staff has recently
examined the competitive and other consumer effects of various aspects of
the FCC scheme for regulating the broadcast and cable industry. 	Among the
aspects we have examined are ones that, like the question of flexibility in
the use of translators, offer the potential to increase consumer satisfaction
by increasing the array of programming available to radio listeners!
4 	See the comments of the Bureaus of Competition, Consumer
Protec'tion, and Economics of the Federal Trade Commission before the
Federal Communications Commission In the Matter of Selection of Initial
Licenses Using Random Selection or Lotteries Instead of Comparative
Hearings, Docket No, 81-768, 	December 30,1981; 	In the Matter of
Amendment of Section 73.3597 of the Commission's Rules on Applications for
Voluntarv Assignments or Transfers of Control, BC Docket No. 81-897, March
1, 1982; In the Matter of Amendment of Part 76. Subpart J, Section 76.501
2

III. 	Development oC Current Translator Policy and Notice oC Inquiry
In 1970, the FCC first authorized the use of FM translators to
"
facilitate radio service to geographic locales and populations unable to
receive adequate radio service due to distance or terrain. 6 	While recognizing
the benefits that might flow from additional radio service to these areas and
populations, the FCC also expressed concern regarding the potential adverse
economic impact of translators on existing and potential full-power FM
stations. 6 	In addition, the Commission concluded that the spectrally
efficient manner of providing FM service is via full-power stations. 7 	For
these reasons, the FCC adopted rules to ensure that FM translators remain
supplemental to, and not substitute for, full-power stations. 	Among the
rules to implement this policy are those which prohibit virtually all program
of the Commission's Rules and Regulations Relative to Elimin:ttion of the
Prohibition on Common Ownership of Cable Television Systems and Nation:tl
Television Networks, CT Docket No. 82-434, December 8, 1982; In the Ma ttcr
of Domestic Fixed Satellite Transponder Sales, CC Docket No. 82-45, April
16, 1982; In the Matter of Amendment of 47 C.F.R. Section 73.658Cj) 00 The
Svodication and Financial Interest Rule, BC Docket No. 82-345, January 31,
1983; Rcply Comments In the Matter of Amendment of 47 C.F.R. Scction
73.658 CD on The Svndication and Financial Interest Rule, BC Docket No. 82·
345, April 26, 1983; 	In the Matter of The Processing of New Domestic
Satellite Applications, Report No. DS-265, May 7, 1984; In the Mattcr of
Amendment of Part 76 of the Commission's Rules Concerning Carriage of
Television Broadcast Signals bv Cable Television Systems, MM Dockct 1'0.
85-349, February 25, 1986; 	In the Matter of Amendment of Section 73.3555
of the Commission's Rules on Broadcast Multiple OwnerShip Rules, MM
Docket No. 87-7, July 15, 1987. 	Copies of these comments are available from
the Federal Trade Commission's Office of Public Reference. 	We also note
that in December of 1923, the Report of the FTC on the Radio Industrv
contributed to the passage of the Radio Act of 1927 and the succeeding
Communications Act of 1934.
5
6
7
NOI at 2.
liL. at 15.
3

origina tion by a commercial transla tor, any program reception by the
translator that is not off-air,8 and the use of translators merely to relay
distant programming from one market to another.9 	In addition, the FCC will
not license a translator to an existing full-power FM station if the
retransmitted signal would extend beyond the station's predicted reception
area and would lie within the reception area of an FM station licensed to
another community.lO
As urged by a number of rulemaking petitions, the FCC seeks comment
"on the value, need, and desirability of expanding FM translator
authority .... "l1 	The FCC recognizes that greater flexibility in the use of
translators can result in consumer benefits that must be weighed against the
costs of a less restrictive regulatory environment. 12 	In particular, while
expressing concern that a substantial increase in the number of translators
"could pose significant and difficult new monitoring arid enforcement
requirements for the Commission,"13 the FCC also observes that such an
increase 	"may result in programming tailored to small audiences with
8 	That is, the translator must be in the line-of-sight of the facility
transmitting the programming to be retransmitted by the translator.
9 	IQ... at 2-4.
10 	Id. at 4. 	However, any other licensee may use a translator to
retransmit the station's signal in these circumstances. 	The FCC prohibits
initiation of any "joint venture" between the translator licGnsee and such a
station until the translator is operational. 	Finally any payment made by the
retransmitted station to the translator licensee must be limited to
maintenance anCi operational expenses. 	The payments are not allowed to
compe.nsate the licensee for the costs of obtaining the license or tower
construction costs. 	Id.
11
12
13
Id. at 24.
Id. at 12-13.
IQ... at 23.
4

specialized tastes rather than to a least common denominator mass audience."
particularl y in rural areas. 14
However. the FCC notes that. according to the National Association of
Broadcasters (NAB). translator licensees may have undermined the FCC's
intent that translators remain a service secondary to full-power FM
stations. IS 	In particular. according to the NAB. the translator importation
of large. major market radio signals into the reception area of other full-
power FM stations may have resulted in "unfair competition" in that it
"disrupt[s] the balance of competition among. and drain[s] critical revenues
from. local stations. particularly in medium and small markets."16 	To remedy
these perceived abuses. the NAB contends that in general translators should
only be licensed to fill in a full-power station's reception gaps attributable
to terrain or other obstacles within the predicted coverage area of that
-_ .. - ~- ._- .. _-- .------ station-and to"provide service to areas not served by any full-power FM
station. I1 -
14 	rd. at 24.
IS 	llL. at 5. 	The NAB asserts that, among other abuses, many
translators are used solely to relay the signal of a distant FM station into a
target geographic market for the purpose of expanding the coverage area of
that distant station. 	rd. at 6.
16
11 	llL. at 6-7. 	This proposal is in contrast to the current rules which
permit a translator to import a distant full-power FM station into a market
beyond the stat,ion's predicted coverage area and into the coverage area of
another full-power station. subject to the restrictions described in note 10.
The NAB proposal would also permit AM stations to use translators in their
markets to rebroadcast their signals on the higher quality (in terms of
propagation and audio characteristics) FM band, although the NAB expresses
the concern that implementation of this proposal could lead to a diminution
of AM listenership. 	rd. at 7.
5

In reviewing the NAB's arguments, the FCC solicits comment in
particular on two issues with regard to the "unfair competition" that may be
posed by translators. 	First, the FCC expresses concern that any relaxation
of the translator rules would lead to "spectrum inefficiency," 	because the
amount of spectrum required for multiple translators to duplicate the
coverage area and the signal quality of a full-power FM station exceeds that
required for the full-power station. 	Moreover, the added competition arising
from the licensing of new translators 'may financially harm or displace
spectrally-efficient full-power stations as well as impede the licensing of new
full-power stations. I8
Second, the FCC solicits comment on whether translator licensees
should bear the same public interest responsibilities as full-power stations. 19
Full-power stations are required to maintain a studio origination capability
_0.__ '" _ ..--''- and to provide programming"rcsponsive -to issues'oflocal concern. 20 'The
FCC notes that because of the costs such requirements impose on full-
power licensees, prospective licensees may have incentives to construct
translators rather than full-power stations. 21
IV. 	Potential Benefits from Relaxation of the Translator Rules
As the FCC observes, one potentially significant consumer benefit from
relaxation of the rules governing translator use is the additional number of
listening options that translators may provide. 	Current rules prohibit
translators frolI} originating programs and full-power FM stations from
18 	IiL. at 15-18.
19 	IiL. at 16-17.
20 	IiL. at 17.
21 IiL.
6

obtaining translator licenses when the translator would compete with another
full-power FM station. 	These rules may restrict the ability of translator
licensees to offer the most profitable programming and thereby artificially
reduce the number of translators and thus the number of listening options
available. 22
Moreover, some program formats that are highly valued by consumers
may not be aired under the current rules. 	As economic analysts and the
FCC itself have observed, the array of programming offered by advertiser
supported broadcast stations--measured in terms of the variety of program
formats and the number of stations offering any particular type of format--
will not necessarily be that which best satisfies consumers.23 	Advertisers
22 	Some translators may be financially viable only if the FCC permits
program origination or permits ownership of the translator by the full-power
FM station (or a joint venture between the station and the licensee). 	For
example, a translator may not be profitable unless the licensee can offer
highly specialized programming which may not be available from an existing
full-power FM station.
Ownership of the translator by a full-power FM station (or a joint
venture between the translator licensee and the FM station) may enable the
station to assume all (or part) of the financial risk in applying for a
translator license and in constructing and operating the translator. 	An
independent licensee may be willing to assume this risk only at a higher
expected return than that required by the station licensee. 	Beca use
ownership or joint ventures may provide a more assured source of revenue
for the translator licensee, each of these options may lower the cost of
applying for, constructing and operating a translator.
Further, by restricting program reception by the translator to off-air
sources, translator licensees may be unable to utilize the most cost-efficient
techniques (e.g., satellites) for program delivery. 	The FeC',s rules restricting
program reception by a translator t6 off-air sources, thus precluding the use
of satellites to relay the-signal to the translator, may be an incentive for
using translators to relay signals from one market to another.
23 	See, for example, Samuelson, The Pure Theon' of Public
Expenditure, 36 Review of Economics and Statistics at 387-389 (1954);
Steiner, Program Patterns and Preferences in Radio Broadcasting, 46
Quarterly Journal of Economics at 194-223 (1954); Spence and Owen,
Television Programming, Monopolistic Competition, and Welfare, 91 Quarterly
Journal of Economics at 103-126 (1977); Federal Communications Commission,
7

(and therefore station owners) seek to maximize audience size for any given
amount of programming and station costs rather than match formats with the
intensity of consumer preferences for specific formats.
Relaxation of the translator rules may remedy some of these
deficiencies. 	First, an increase in the number of listening options may raise
the probability that highly valued but less popular formats will be aired. 24
For example, translators operate at substantially lower power and lower cost
than full-power stations. 	Consequentt"y, in a more relaxed regulatory
environment, translators may be more likely to offer specialized programming
by locating in those areas in which listeners who prefer a distinct format
are concentrated geographically.
Second, increasing the number of stations can raise the level of
consumer satisfaction by increasing the variety of options within a format. 25
Moreover, an increase in the number of stationscan--increase the extent of
competition for the patronage of listeners and as a result, may raise the
level of expenditures on programming. 	To the extent that the existing
Report and Order In the Matter of Deregulation of Radio, BC Docket No.79-
219, February 24, 1981 and references therein.
24 	This increase may occur even in those local markets in which the
FCC's allocation of FM radio spectrum precludes the operation of additional
full-power FM stations. 	To prevent interference among full-power FM
stations, the FCC requires that stations be separated from each other on the
FM band. 	Because of their lower power, translators may be able to operate
within these separations without causing noticeable interference to the full-
power stations. 	Thus, translators may be able to utilize spectrum that might
otherwise lie faiIow. 	Further, as noted immediately below in the text,
translator licensees may have greater incentives to offer specialized
programming than do full-power stations.
25 For example, if two jazz stations permitted a greater variety of jazz
recordings than would be possible with a single jazz station, listenership
(used as a proxy for consumer satisfaction) might increase.
8

number of local radio stations is less than that required to reduce profits to
a competitive level, the introduction of additional stations may increase
station rivalry for listeners and may create incentives for stations to devote
some of their excess profits to increased programming expenditures. 26
Higher programming expenditures may indicate a higher quality of radio
service (e.g., compact discs instead of vinyl records, more entertaining or
informative on-air personalities, better audio characteristics).
In conclusion, an increase in the number of listening options may
increase consumer satisfaction by increasing the number of formats, the
within-format variety, and the quality of the programming. 	To test this
hypothesis, two economists in the FTC's Bureau of Economics conducted a
statistical analysis relating the daily average percentage of the popula tion
listening to radio to (among other factors) the number of formats in various
- ---- 	------ -radio markets. 27 -While any rise in listenershipis a highly imperfect
__ 	indicator of an increase in consumer satisfaction,28 the results suggest that
an increase in the number of stations has a statistically significant effect on
increasing the number of formats and that an increase in the number of
formats has a statistically significant effect on increasing the percentage of
26 	For a discussion of this possibility with empirical support from
television broadcasting, see Fournier, Nonprice Competition and the
Dissipation of Rents From Television Regulation, 51 Southern Economic
Journal at 754-65 (1985). 	Excess profits may result from the restrictions on
radio station entry arising from the FCC's allocation of spectrum to radio
and not necessarily from overt or tacit collusion among radio stations in a	,
particular market.
27 	The details of this analysis are provided in the attached appendix
by John R. Woodbury and Robert P. Rogers.
28 	Even if (in the extreme case) total listenership did not increase
with the rise in the number of stations or formats, it would not necessarily
mean that consumers did not place a higher dollar value on the new array of
programming.
9

the population listening to the radio. 29 	These effects exist regardless of the
number of stations and formats currently provided. 	While small in
magnitude, the listenership effect is relatively larger for markets in which
there are fewer stations and fewer formats available. 	This empirical
analysis, therefore, provides some support for the possibility that an increase
in the number of stations provides consumer benefits by enhancing the
variety of formats available.
We are not suggesting that translator licensees be limited to those
entities offering original programming or importing distant FM signals into
local markets. 	The use of translators by existing FM stations to fill in the
gaps in their predicted coverage area also offers benefits to consumers by
enabling consumers residing in the gaps to receive an additional listening
option. 	Similarly, translator use by AM stations may alter the programming
- -of the AM station in ways(e~g.,-changesirithe programming form-ili--or--------------
----- increased expenditures on programming) that may benefit consumers. 	We
only suggest that employing translators for program origination or for
importing distant FM signals into a local market (regardless of the identity
of the translator licensee) also has the potential to benefit consumers.
V. 	The Perceived Costs of Relaxing the Rules
As noted in Section III, the FCC perceives two costs that may result
from rule relaxation: a possible inefficient use of the spectrum and an
artificial incen.ive to use translators instead of full-power stations because
29 	By "statistically significant," we mean tha t these effects are
unlikely to have occurred by chance.
10

of the current absence of public interest obligations for translators. so 	While
we do not offer any definitive conclusion on these potential costs, we offer
some observations that may be useful in considering the role of translators
in the provision of radio services.
A. Inefficient spectrum use
We suggest that the appropriate standard for judging the alternative
uses of the spectrum allocated to radio is the effect different allocations
have on consumers, rather than simply on how much spectrum is allocated to
each use. 	The amount of spectrum used by a number of translators to attain
the coverage and audio quality of a full-power station is likely to be greater
than that required by a full-power station, but if the translators offer some
programming that differs from that of a full-power station, consumers may
be better off with the translators even if they require more spectrum.
Against that background, the competition afforded by a group of
translators may economically displace an existing AM or FM station. 31 	Such
an outcome may benefit consumers by providing them with listening options
they value more. 	Because consumers do not pay directly for radio services,
whether or not consumers will in fact benefit from any particular
displacement is unknown.32
30 	The FCC also expresses concern regarding a possible increase in
the agency's monitoring and enforcement costs should the translator rules be
relaxed.
I
31 	The limited range of a single translator makes displacement of a
full-power station less likely.
32 	In particular, the kind and amount of information required to know
the mix of programming that best satisfies consumers is unlikely to be
available. 	A general discussion of these information requirements C:In be
found in Federal Communications Commission, Report and Order 1:1 the
M:1tter of Deregulation of Radio, BC Docket No.79-219, February 24, 1981.
11

The FCC itself, however, has concluded that consumers in general
would experience greater benefits from a policy that relies not on regulation,
but on the interaction of advertisers and radio station owners to determine
the kind and amount of programming provided listeners. 33 	While advertiser-
supported radio may reflect only imperfectly the programming preferences of
listeners, the resulting programming may match those preferences more
closely than a regulatory determination of the programming to be offered.
Such a determination would likely be based on information less complete
than that available to licensees responding to the profit incentives created
by the demands of both listeners and advertisers.
For these reasons, a policy that permits the market to determine the
mix of both full-power stations and translators as well as the mix of
Specifically, allocation of radio station licenses between translators and full-
-power stations in a manner that best satisfies the interests of consumers
requires that the regulatory authorities be able to know the type of
programming listeners want, know how much of each type of programming
they want, and balance these demands against the cost of producing the
programming. 	Thus, regulators must possess an extraordinary amount of
information regarding current consumer valuations and the production costs
associated with the characteristics of various programming arrays that might
be offered consumers in different geographic areas. 	These characteristics
include the audio quality of the transmission, the type of on-air personalities
used, the composition of the programming in terms of music, news, weather,
and commercials, and the specific scheduling of individual musical pieces
during the day in addition to the general kind of music offered. 	Further,
regulators must not only have information on current consumer valuations
and costs of these characteristics, they must also be able to mimic the speed
with which a well-functioning real-world market would recognize and respond
to changes in these valuations and costs and to mimic the extent to which
market partici~ants would engage in risky programming and other
innovations. 	FInally, regulators must possess information on the current
value.of and changes in the value of the spectrum allocated to radio in
comparison to the value of the same spectrum in other uses, increasing or
decreasing the amount of spectrum allocated to radio appropriately.
33 	See Federal Communications Commission, Report and Order Tn the
Matter of Deregulation of Radio, BC Docket No.79-219, February 24, 1981.
12

program formats may yield greater consumer benefits than one which
protects full-power stations from economic harm. 	In addition, the latter
policy may be inconsistent with the market-oriented approach generally
articulated by the FCC. S(
Finally, the FCC has noted that the future licensing of full-power
stations could be impeded if translators are allowed to occupy a significant
part of the spectrum required by prospective full-power stations. 	Once
again, this raises the issue whether consumers would prefer the program
offerings of the translators or the full-power station. 	Assuming consumers
may prefer the programming of some of these prospective full-power
stations, there are at least two alternatives that are less restrictive than
limiting the permitted uses of translators in deference to the possible
licensing of new full-power stations. 	First, the FCC could require the
prospective full-power licensee to purchase the translator facilities. 	If the
full-power station is likely to be more profitable than the group of
translators in question, the prospective full-power licensee should be able to
buy some or all of the translators occupying the spectrum space it desires.
If the full-power station is successful, the translator licenses would be
voided by the FCC in return for the award of a full-power license. 35 	This
34 	In this regard, the FCC may wish to consider clarifying its policy
basis for distinguishing between the economic harm inflicted on a full-power
station by a translator (e.g.) importing a distant FM signal and that inflicted
on the same staIion by another local full-power station using the translator
to fill in the station's coverage gaps. 	If the programming offered by each
translator were equally attractive to listeners, the adversely affected full-
power station would experience a loss in listeners and therefore advertising
revenue in either case. 	And in either case, the adversely affected station
might become unprofitable.
35 	It is possible that in attempting to purchase a group of translators,
the full-power station may confront a translator licensee who is willing to
13

solution would be consistent with the FCC's market-oriented approach to
radio regulation and would not require the FCC to choose what kind of
station best serves consumers. 36
Second and alternatively, the FCC could condition each translator
license on the requirement that the translator license becomes void if a
license for a full-power station requires that the translator's operations
discontinue. 	Although this approach would be superior to restrictions on the
type of uses allowed for translators, the additional risk borne by translator
licensees would be likely to reduce the number of translators available to
listeners.
B. Public interest responsibilities
Currently, translator licensees are not required to bear two public
interest responsibilities borne by full-power stations: maintaining a studio
sell its facilities only at a price that captures all the additional profits that
would be earned if the full-power station were to replace the group of
translators. 	In the event of such a "hold-up," the FCC may have to act as
a court of last resort. 	"Hold-up" problems may be more likely to occur if
the full-power station can only become operational by acquiring all the
translators and if the ownership of the licenses within the group of
translators is dispersed.
But because of the lower power and the lower coverage area of any
individual translator, the full-power station may become operational even if
the entire group of translators is not acquired. 	The full-power station may
be able to use equipment to prevent the station's signal from interfering
with that of the recalcitrant translator licensee. 	To the extent that a full-
power station can become operational without having to acquire the entire
transla tor group, the incentives for a "hold-up" may be reduced.
36 	Indeeq, in a proceeding that ultimately led to the adoption of
lotteries for the assignment of low-power television licenses, the FCC noted
that "it may be exceedingly difficult for the Commission to select the
applicant that will best serve the desires of consumers." 	See Federal
Communications Commission, Notice of Proposed R ulema king In the Ma tter of
Selection of Initial Licenses Usinf! Random Selection or Lotteries Instead of
Comparative Hearings, FCC Gener::l1 Docket No. 81-768, November 5, 1981, at
3.
14

origination capability and providing issue-responsive programming. 	As the
FCC observes, these requirements impose costs on full-power stations, and
the absence of similar requirements for translators may artificially encourage
prospective station owners to operate translators rather than a full-power
station.
On the one hand, a full-power station may be more efficient than
translators, in terms of minimizing production costs and satisfying
consumers. 37 	Imposing symmetric public interest obligations on translators
would eliminate any incentive for licensees to select translators instead of
full-power stations simply for the purpose of evading the costs resulting
from these obligations.
On the other hand, imposing public interest requirements on translators
is likely to reduce the number of translators, thereby reducing the number
of listening options-and-therefore the extent of program variety available to
listeners. 	In deciding whether to impose on translators the same public
interest obligations borne by full-power stations, the FCC might wish to
weigh the loss in program variety against the gains in eliminating the
artificial incentive to select translator technology. 	The significance of these
gains will depend upon the magnitude of this incentive 38 and the value to
consumers of the public interest requirements.
VII. 	The Manner in which Licenses Are Assigned
J
37 	Operating a large enough group of translators to duplicate the
reception area of a full-power station may result in higher production costs
(e.g., maintenance or engineering) than those for a full-power station. 	But
these additional costs may be less than those resulting from compliance with
the FCC's public interest requirements for fUll-power sta tions.
38 	This incentive results from both the direct production cost savings
and the amount of increased revenue generated as public interest programs
are displaced by programs that advertisers are willing to sponsor.
15

The FCC also solicits comment on whether mutually exclusive
applications for translator licenses should be resolved by a lottery procedure
rather than by comparative hearings. 	In a different context, the FCC
previously concluded that resolving mutually exclusive license applications
through administrative hearings is unlikely to yield sufficient information to
identify those prospective licensees that will best serve consumers. 39
License assignment by lottery would reward entities that are lucky rather
than efficient.4o 	Thus, in either administrative assignment method, we
believe that resale should be permitted so that the initial licensee could
transfer the tra nsla tor license to the en ti ty that places the grea test val u e on
the use of the translator. 41
The primary advantage of lotteries over comparative hearings may be a
more rapid license assignment, thus enabling consumers to benefit from
translator services at an earlier date. 42 - Because-of the time -required to
39 	Federal Communications Commission, Notice of Proposed Rulemaking
In the Matter of Selection of Initial Licenses Using Random Selection or
Lotteries Instead of Comparative Hearings, FCC General Docket No. 81-768,
November 5, 1981, at 3; see also Letter from David T. Scheffman, Director,
Bureau of Economics, Federal Trade Commission, to Mark S. Fowler,
Chairman, Federal Communications Commission, October 29, 1986, at 3.
40 	Scheffman, supra note 39, at 3.
41 For a general discussion of this point, see the comments of the
Bureaus of Competition, Consumer Protection, and Economics of the Federal
Trade Commission before the Federal Communications Commission 	In the
Matter of Selection of Initial Licenses Using Random Selection or Lotteries
Instead of Comqarative Hearings, FCC General Docket No. 81-768, December
30, 1981, at 5-6.
42 	For example, the duration of streamlined comparative hearings for
the allocation of cellular radio licenses in the top 30 markets averaged 18
months. The use of lotteries for markets 91-120 was predicted to result in
an average delay of 12 months in the initial award of the license. 	See
K werel and Felker, Using Auctions to Select FCC Licensees, FCC Office of
Plans and Policy Working Paper No. 16, May 1985, at 12.
16

complete the proceedings, comparative hearings may considerably delay the
introduction of translator services to consumers. 	For some contemplated
uses of translators in some markets, this delay may render those uses
unprofitable.
We recognize that lotteries in conjunction with resale could result in a
deluge of lottery players and render the lottery costly and cumbersome to
administer.4.3 	This potential deluge can be reduced by, for example,
restricting the time during which applications may be filed, establishing
minimum financial or technical Qualifications for the applicants, establishing
a sufficiently large application fee, and prohibiting resale for a more
extended period of time following the award of the translator license. 	These
options can maintain the number of applications at manageable levels and
reduce the direct government costs of awarding the license. 	However, the
......_-"'-"".'- .' restriction of resale is the option most likely to harm consumer interests'
because there is still no guarantee that the "winner" from the more limited
applicant pool is the entity that could use the translator most profitably.
Therefore, the kind of lottery that is most likely to serve consumers is one
that places as few restrictions on resale as is practical and consistent with
FCC policy.
VII. 	Summary
We conclude that there may be significant consumer benefits if the FCC
were to relax the current restrictions on the permissible broadcast uses of	~
translators. 	In particular, permitting program origination or distant signal
importation by translators (regardless .of the licensee's identity) is likely to
43 	The administrative delay occurs because the FCC must log and
prescreen each application. 	Jd.
17

advance the interests of consumers by increasing the number of listening
options available. 	Although listenership is only an imperfect indicator of the
value consumers place on additional options, the empirical analysis by two of
the Bureau of Economics staff indicates that additional listening options may
increase listenership. 	In addition, by permitting the translators to receive
their programming in the most cost-effective manner, the market-determined
number of translators may be greater than would be the case if the only
allowed mode of signal reception were off-air.
With regard to the costs of relaxing the rules, we suggest that the
policy touchstone for determining whether spectrum allocated to radio is
being efficiently utilized may not be merely how much spectrum translators
will use in comparison to full-power stations. 	Whether the use of spectrum
by translators will serve consumers to a greater extent than use by a
comparable full-power station may be at least as importanC In particular, if
the translators offer some programming that differs from the comparable
full-power station (thus offering consumers greater variety), consumers may
be better off with the translators even if they utilize more spectrum. 	Thus,
a regulatory determination that full-power stations should be protected from
financial harm caused by translators may harm consumers and may be
inconsistent with the FCC's policy that the interaction of advertisers and
sta tion owners should determine the array of programming 'provided to
consumers.
If the licensing of translators limits the FCC's ability to authorize new
full-power stations, we suggest two possible resolutions of this impediment
that are less restrictive than prohibiting certain broadcast uses of
translators. 	We also suggest the kinds of concerns that the FCC may wish
18

to consider in deciding whether it should impose on translator licensees the
same public interest obligations of fUll-power stations.
Finally, as an alternative to comparative hearings, a lottery offers the
potential for a more rapid resolution of mutually exclusive license
applications for translators. 	The FCC may have to limit the number of
applications for such a lottery, but there are a number of ways to do this.
One way, proscribing resale for an extended period of time after the award
of the license, would diminish the consumer benefits from this license
assignment mechanism.
19

STATISTICAL APPENDIX!
John Woodbury and Robert Rogers
- - 	----- -- --~,-_.. --.	--,.---------- -
1 	This appendix has been prepared by the staff of the Bureau of
Economics of the Federal Trade Commission. 	It has not been reviewed by,
nor does it necessarily reflect the views of, the Commission or its
members. 	We would like to thank Alan Mathios and James Langenfeld for
helpful comments and David Eaton for valuable research assistance.

In this Appendix, we investigate in a preliminary way the relationship
between the number of radio stations and listenership. 	We do this indirectly
by first estimating the relationship between the number of formats in a
market and the number of stations. 	We then estimate the relationship
between the percentage of the market population listening to the radio and
the number of formats. 	Together, both equations permit us to estimate the
effects of increasing the number of stations on listenership.
As discussed in the text of the comment, an increase in the number of
radio stations may increase the number of programming formats offered and
the greater number of formats may increase listenership. 	We can model the
general relationship among the proportion of people listening to the radio,
the number of stations and the number of formats as two equations; one
relating the radio listenership to the number of formats and one relating the
number of formats to the number of stations. 	These relationships can be
generally depicted as follows:
FOR =
P 	= 	F(FOR, Xl' v)
F(NS, POP, Xl' U)
(I)
(2)
where FOR equals the number of formats, NS equals the number of sta tions
in a market, and P equals the proportion of population listening (called
listenership below). 	POP is the population in the radio market. 	Xl is a
vector of other market characteristics that might affect listenership and the
... _ .numberof formats respectively, and u and v are residuals. 2 -
We include POP in equation (2) because the number of formats in a
given market may be influenced by population. 	One would expect a greater
market population to increase the number of formats because such markets
may have more people with diverse tastes and may generate sufficient
advertising revenues to support differing formats. 	Consequently, stations
offering more specialized formats in such markets may be more profitable
than in markets with smaller populations.
The Xl vector consists of variables that may affect the distribution of
programming tastes across any given population and their use of radio as an
entertainment medium. 	As such, Xl may affect both P and FOR, although
we do not have any priors on whether the coefficients of the components of
Xl in either equation should be positive or negative. 	The c9mponents of Xl
are median income (Income), the population proportion of Blacks (Black) in
the market, the population proportion of Hispanics (Spanish), and the
proportion of tfie popula tion over 34 years old (Age).
2 	Because NS is determined by FCC allocation procedures, we treat NS
as exogenous, although the regula tory decision process that gives rise to
that allocation may depend on POP and Xl among other variables. 	In
addition, the number of formats may depend upon expected listenership. 	To
the extent that actual listenership is correlated with expected
listenership, endogeneity issues arise and should be considered in a more
complete analysis.

For the sample, we examined the 259 Arbitron radio markets for which
industry data are collected. 	To insure that our data points closely
correspond to real markets, we confined our sample to the 115 markets
where the local stations have a market share of over 85 per cen t. 	We used
data for the spring of 1987.s
Since it is likely that (for example) the absolute effect of the number
of stations on format is dependent on the size of the population in the
market, we estimate equations (1) and (2) in double log form:
InFOR
In [P/ (l - P)] =
Inb o + bl InNS + b2 InPop + bs InIncome
+ b. InBlack + bs InSpanish + b6 InAge + u 	(I a)
Inao + al InFOR 	+ as In Income + a. InBlack
+ as InSpanish + a6 InAge + v 	(2a)
In the listenership equation (2a) we transform P to represent the log of
the odds in order to constrain the predicted P to the zero-one interval. 	We
therefore use the minimum logit chi square method in estimating equation
(2a).· 	Since P and P/(1 - P) are proportions, they ha ve a range of from
zero to one and P has a binomial distribution with a variance of P(1 - P),
and In[P / (l - P)] has a variance of 1/ [nj (1 - P)P] where ni is the number
_________________ of people in the Arbitron sample for market "i." Thus the error term v is
heteroskedastic. 	To correct for this problem, we use a weighted least
squares technique with the following weight: __
[ni (1 - p)p]l/2.5
It would seem logical to substitute the FOR equation into the logit
equation in order to evaluate the impact of NS on P directly. 	However, this
creates the potential for a serious multicollinearity problem. 	Substitution of
(1a) into (2a) will result in In POP being an additional weighted independent
variable which will be highly correlated with other weighted variables in 2(a)
because of the correlation (.9008) between In POP and the Arbitron sample
size ni' 	Such multicollinearity can lead to large variances for the parameter
estimators thereby reducing the precision of our estimates and increasing the
probability of accepting the hypothesis that the coefficients are not
s 	Most data are from Broadcast Investment Analysts, Inc., Investing
in Radio. 1988. 'We express our appreciation to Tom Buono and Gregg Welsh
of Broadcast Investment Analysts for their permission to use the data.
Listenership data are from the market-by-market ratings publications of
Arbitron.
• 	A discussion of this method can be found in G. S. Maddala, Limited
Dependent and Qualitative Variables in Econometrics, 28-32 (1983).
5 	The basis for using these weights can be found in Maddala, id.
2

significantly different from zero when in fact the nuIl hypothesis should be
rejected. 6 	In fact, the correlation coefficients between the weighted In POP
and three of the other weighted variables (In NS, In Income, and In Age) are
0.9935, 0.9974, and -0.9511 respectively.7 	Consequently, rather than
substitute equation (la) into equation (2a) and estimate a single equation, we
estimate each equation and evaluate the impact of NS on P by making that
substitution after the two equations have been estimated.
One could argue that since POP is implicit in the FOR equation it still
has not been eliminated as a source of multicollinearity when we use FOR
only in the P equation. 	However, there are several other variables in the
format equation, and the impact of population on the number of formats,
while statisticaIly significant, is not large (the regression coefficient being
about 0.10). 	Thus its contribution to any multicollinearity problem through
its effect on FOR is likely small. 	The multicollinearity is probably less of a
problem in the FOR equation because we do not use the weighted OLS. 	For
instance while the weighted In NS and In POP variables have a correlation
coefficient of (0.9935), the correlation between the unweighted variables,
while high (0.9122), is lower than the weighted versions of the variables.
Thus estimating the FOR equation for the relationship between the
number of formats and the number of stations may enable us to obtain more
precise estimates of the effect of station numbers on listenership. 	The
method used here is one of the solutions to the multicollinearity problem
recommended by Kennedy. 	In particular, he suggests that "formaliz[ing the]
relationships among regressors" in the same manner as described above may
reduce the degree of muIticoIlinearity.8 --- 	. 	-----
Table Al reports the results for the estimation of equations (la) and
6 	See H. Theil, Principles of Econometrics 147-155 (1971).
,
7 	There are also high correlation coefficients between some of the
other variables, but they are all required in the listenership equation
for theoretical reasons. 	To exclude these relevant vari3 bles would lead
to estima tors tha t are biased.
8 	P. Kennedy, A Guide to Econometrics, 151 (I 985).
3

(2a) for two estimating methods; one is an independent estimation of each
equation where the above-described weighted method (minimum logit chi
square) is used to estimate the listenership model and an ordinary least
squares (OLS) approach is used for the format equation. 	The other method
takes into account the possibility of correlation between the residuals of the
two equations. 	It is a seemingly unrelated equations (SURE) method whereby
the two equations are estimated jointly (the variables in the listenership
ratio equation being properly weighted as in a single equation minimum logit
chi square estimate.g)
In the format equation, all the variables except Income and Age are
significantly different from zero at a 5 percent level for both estimating
techniques. 	The positive coefficient for NS is as predicted by theory: an
increase in the number of stations is associated with a greater number of
formats. 	The coefficient being considerably less than one supports the
contention of Steiner and others on the tendency of advertiser supported
radio to weight their program offerings toward "common-denominator"
programming.
The positive coefficients for the proportion of Black and Spanish people
suggests that the programming tastes of these groups contribute to format
diversity. 	The insignificant results for Income and Age are not particularly
alarming because they are not inconsistent with economic theory. 	Nothing
in the literature states that differences in age and income would necessarily
lead to more diversified tastes in radio programming, other things equal.
- In the listenership equation, the coefficient of the number of formats is
positive and significantly different from zero at the 5 percent level. 	The
proportion of Hispanics in the population is also significantly different from
zero. 	For neither method are the other demographic variables significantly
different from zero on a two tail test. 	[Since economic theory has no
prediction as to the direction of either coefficient, we rely on the two tail
test.] 	Again these insignificant results are not particularly alarming because
economic theory only predicts a possible effect for these variables.
As a whole, both equations are statistically significant; the log
likelihood ratio for the seemingly unrelated equation estimation is 200.56
which is significant at the I percent level. 	For the independent estimation
procedure, the log-likelihood ratios are 47.45 for listenership equation and
172.88 for the format model, both with significance levels of I percent.
The implications of this model for FCC policies about the use of FM
translators are as follows. 	The fact that increasing the number of formats
increases listene~rship implies that the marginal listener values additional
alternative types of radio programming. 	Thus consumers may benefit if
given "radio markets had more formats. 	To the degree that FM translators
increase the number of formats in given market areas, a flexible policy
9 For a discussion of SURE, see J. Johnston, Econometric Methods,
238-241 (1972).
4

towards FM translators will raise consumer satisfaction with the array of
programming offered.
Our model also suggests that by increasing the number of stations, a
more flexible policy toward translators will indirectly lead to increased
listenership because of the effect of the number of stations on the number
of formats. 	By multiplying the coefficient for NS in the format equation by
the coefficient for FOR in the listenership equation, we can calculate the
impact of a change in the number of stations on the proportion of the
population listening to the radio. 	For the single equation model, this
product is 0.139 times 0.196 or 0.027, and for the SURE estimation, it is
0.209 times 0.188 or 0.039. 	Asymptotically normal variances can be
calculated for these two products as: 10
For the single equation model, the square root of this variance equals 0.0158
and for the SURE estimation, it equals 0.0177. 	This gives t values of 1.71
for single equation estimation and of 2.20 for the SURE equation; thus the
product of the two coefficients is significantly greater than zero at the 5
percen t level for both models.
Thus, adding radio stations will increase the percentage of listeners in
the population by way of increasing the number of formats or at least the
expected number of formats present in a market. 	To illustrate how this
model works, consider a market with a listener percentage of 0.1653, 22
--stations, and 6 formats (all these being roughly the averages for the sample)"
Using the SURE estimates, we calculate that adding one more format would
raise the listening share of population by 0.0048 to 0.1701 (or by 2.9 per
cent of the original number of listeners). 	Adding one more station will
increase the listener share by 0.0002 raising it to 0.1655 (or by 0.1 per cent
of the original audience). 	If one were to increase the number of formats by
only increasing the number of stations, one would have to add almost twenty
stations, the derivative of FOR with respect to NS being 0.051 in this
average market.
For the smaller markets, the relative changes tend to be larger. 	For
instance for a market with 13 stations and 5 formats, the percentage
increase in listenership from the addition of one format is 3.5, and from the
addition of one station it is 0.3. 	For a market with eight stations and four
formats, the percentage increase in listenership from the aQdition of one
format is 4.5, and from the addition'of one station it is 0.4.
While these calculations do not indicate that increasing the number of
stations will greatly enlarge listenership, they do imply that the consumers
10 	G. C. Chow, Econometrics 182-184 (1983). 	In order for this
formula to apply, the estimators, at and bt, must be asymptotically
consistent. 	If NS is endogenous, contrary to our assumption in footnote
2, bt might be inconsistent in which case the application of this formula
would be inappropriate.
5

would value more listening options -- especially stations with different
formats.
6

Table Al
The Relationship between the Proportion of Population
Listening to the Radio (Listenership Ratio) and the Number
of Formats and the Relationship between the Number
of Formats and the Number of Stations 11
The Model for the Number of Formats
Variables 	Coefficients
f.Qr. Q1.S.
Constant 	0.279
Number of Stations 	0.196 (2.99)
Population 	0.098 (3.24)
Income 	-0.013 (-0.11)
Black 	0.024 (2.43)
Spanish 	0.041 (3.46)
Age 	-0.042 (-0.23)
Log Likelihood Ratio 	172.88
Coefficients
f.Qr.SURE
0.248
0.188 (2.99)
0.101 (3.52)
-0.015 (-0.13)
0.023 (2.41)
0.040 (3.55)
-0.062 (-0.36)
200.72
The Model for Listenershi p 12
Variables 	Coefficients
for OLS
Constant 	-0.771
Number of Formats 	0.139 (2.05)
Income 	-0.074 (-0.87)
Black 	0.015 (1.49)
Spanish 	0.026 (2.67)
Age 	0.246 (1.85)
Log L!kelihood Ratio 	47.45
Coefficients
f.Qr.SURE
-0.512
0.209 (3.17)
-0.120 (-1.46)
0.010 (1.05)
0.019 (2.01)
0.193 (1.48)
200.56
11 	"t-statistics" are in parentheses.
12 	For both the single equation (OLS) and the Seemingly Unrelated
Equations (SURE) approaches, this equation is estimated by a weighted
minimum logit chi square method.
7