United States v. American Soc. of Composers, Authors & Pub.

981 F. Supp. 199 (1997)

UNITED STATES of America, Plaintiff,
v.
AMERICAN SOCIETY OF COMPOSERS, AUTHORS AND PUBLISHERS, et al., Defendants.
In the Matter of the APPLICATIONS OF SALEM MEDIA OF CALIFORNIA, INC., et al. and New England Continental Media, Inc., et al., Applicants.
For Licenses for Their Radio Broadcasting Stations.

No. CIV.A. 13-95(WCC).

United States District Court, S.D. New York.

September 12, 1997.

White & Case, New York City (Philip H. Schaeffer, Joan M. McGivern, J. Christopher Shore, of counsel), Paul, Weiss, Rifkind, Wharton & Garrison, New York City (Allan Blumstein, of counsel), Richard H. Reimer, American Society of Composers Authors and Publishers, New York City (Richard H. Reimer, of counsel), for American Society of Composers Authors and Publishers.

Wiley, Rein & Fielding, Washington, DC (Bruce G. Joseph, Michael L. Sturm, Richard T. Pfohl, Karyn K. Ablin, of counsel), for Applicants.

OPINION AND ORDER

WILLIAM C. CONNER, Senior District Judge:

This application is before the court in the court's rate-setting capacity under Section IX of the Amended Consent Judgment ("Consent Decree") reached by the parties in United States v. American Society of Composers, Authors and Publishers, 1940-43 Trade Cas. (CCH) ¶ 56,104 (S.D.N.Y.), as amended by 1950-51 Trade Cas. (CCH) ¶ 62,595 (S.D.N.Y.1950). The Consent Decree, originally entered in 1941 and subsequently amended in 1950, settled the United States' antitrust suit against the American Society of Composers, Authors and Publishers ("ASCAP"). This court retained jurisdiction to oversee ASCAP's compliance with the terms of the Consent Decree. Section IX(A) of the Consent Decree provides that if, as is the case here, ASCAP and a music user are unable to agree on a fee for the right to perform music in ASCAP's repertory, the music user may apply to this court for determination of a reasonable fee.

BACKGROUND

The applicants in this consolidated proceeding are New England Continental Media, Inc. ("NECM") and Salem Media of California, Inc. ("Salem Media"). NECM is a group of 360 local radio stations, the majority of which broadcast religious programming with a mixed talk and music format.^[1] The remaining NECM stations broadcast primarily classical music or foreign language programming. The Salem Media Applicants are 61 local radio stations with formats substantially similar to those of the NECM stations. The NECM application covers the period from January 1, 1991 through December 31, 1995. The Salem Media application period runs from January 1, 1983 to December 31, 1990. Both the NECM and Salem Media applicants are now represented by the National Religious Broadcasters Music License Committee.

From 1986 to 1990, most local radio stations operated under a license negotiated in settlement of a rate proceeding brought by the WGN applicants. From 1991 to 1995, the majority of radio stations operated under the Group W license, which was the product of proceedings commenced by the Group W applicants. The Radio Music Licensing Committee ("RMLC"), a radio industry organization, negotiated these licenses on behalf of the respective applicants. At the time of the Group W negotiations, the RMLC represented several thousand local stations and approximately seven thousand additional stations that had consented in advance through extension agreements with ASCAP to be bound by the terms agreed upon by ASCAP and the RMLC in the Group W negotiations.

ASCAP offered the NECM stations the license terms contained in the Group W license and offered the Salem Media stations the terms set forth in the WGN license. The relevant provisions of these licenses are substantially similar except that the Group W license contains additional administrative requirements and modifies the formula by which incidental use fees are determined. The NECM applicants currently operate under an interim fee order, which in substance extends the WGN license for the NECM applicants pending the outcome of this proceeding. Those Salem applicants that have not joined the NECM proceeding continue to operate under interim licenses effective since 1982.

Applicants seek a determination by this court of a reasonable per-program license fee for their music use. Under a per-program license, ASCAP grants a music user the right to perform any of ASCAP's compositions as many times as the user wishes. In exchange, the user pays a license fee for only those programs in which it actually performs ASCAP music. See Consent Decree, § VII(B). Historically, the per-program license fee rate has consisted of a fee for feature performances of ASCAP music and a fee for incidental uses of ASCAP music. The feature performance fee is a percentage of the adjusted gross revenue that a radio station earns from programs that contain feature performances of ASCAP music. The station must report to ASCAP on a monthly basis all musical compositions that received feature performances in programs broadcast by the station, unless the station concedes that a particular program contained a feature performance of ASCAP music. The incidental use fee, which under past licenses has corresponded to a relatively small percentage of the station's adjusted gross revenue, provides ASCAP with payment for a station's broadcast of non-feature music including commercial jingles less than sixty seconds long, background music, and ambient music picked up during the coverage of public events.

A. ASCAP's Fee Proposal

In the instant proceeding, ASCAP has proposed a blanket license fee of approximately 1.6% of a local radio station's adjusted gross revenue.^[2] ASCAP has also offered a per-program license under which a radio station would pay 4.22%^[3] of its revenue subject to fee^[4] for the use of feature music. In addition, a station selecting a per-program license would incur an incidental use fee.^[5] ASCAP proposes that applicant stations operating under an interim per-program license accept an incidental use fee equal to the amount that those stations have already paid for incidental music use under the interim license. In the alternative, ASCAP asks the court to rule that applicant stations operating under either an interim per-program or blanket license, which choose a retroactive per-program license, incur a fee of 1.82% of the revenue those stations derived from weighted hours containing incidental music but no feature music. This alternative, however, would be available only to those applicant stations capable of documenting their incidental music use during the license period.

B. Applicants' Proposal

Applicants submit that a per-program license under which a station would pay a feature music fee of 1.73% of its revenue subject to fee and an incidental use fee of 0.06% of its adjusted gross revenue is appropriate. Applicants argue that the court should accept this proposal because it would make per-program licenses available to a broad range of stations and thereby provide those stations with a genuine choice between the per-program and blanket licenses, as the Consent Decree mandates. According to applicants, under the terms of their proposed license, a station broadcasting feature performances of ASCAP music at the "typical" industry level would pay roughly the same fee, exclusive of administrative costs, under the blanket or per-program license.^[6] Applicants also contend that their proposed incidental use fee more accurately reflects the value ASCAP assigns to incidental music.

Applicants urge that the court adopt a retroactive blanket license for those applicant stations that, allegedly because of the absence of a viable per-program license alternative, chose interim blanket licenses and did not maintain music-use records. Applicants argue that this blanket license should be set at 56.3% of the Group W and WGN blanket license fees to reflect applicants' limited music use. Applicants maintain that absent such a remedial blanket provision, the applicant stations lacking records of their music use would be unable to benefit from a modified per-program license if this court were to rule that ASCAP's per-program license is unreasonable.

DISCUSSION

The Consent Decree's purpose is "to limit ASCAP's ability to exert undue control of the market for music licensing rights through its control of a major portion of the music available for performance and its use of the blanket license as a means to extract non-competitive prices." United States v. ASCAP/Application of Capital Cities/ABC, Inc., 157 F.R.D. 173, 177 (S.D.N.Y.1994) (hereinafter "Buffalo Broadcasting II") (quoting Report of the Special Master) In a similar vein, in United States v. ASCAP/Application of Turner Broadcasting Sys., Inc., 782 F. Supp. 778, 790 (S.D.N.Y.1991), aff'd 956 F.2d 21 (2d Cir.1992), the court declared that "the Decree was designed to limit ASCAP's ability, by pooling copyrights for large amounts of music used in radio broadcasting, to extract unreasonable fees for performances of the music. The availability of per-program licenses, if reasonably priced as compared to the alternative blanket license, was one means of accomplishing this purpose. ..." In furtherance of this objective, the Consent Decree requires that ASCAP make a per-program license available to music users. (Consent Decree, § VII(B).) Section IX places upon ASCAP the burden of demonstrating that its proposals, if challenged by a music user, are reasonable, and Section VIII obligates ASCAP to "use its best efforts to avoid any discrimination among the respective fees fixed for the various types of licenses which would deprive the licensees or prospective licensees of a genuine choice from among such various types of licenses." Thus, the Consent Decree prohibits ASCAP from effectively denying a per-program license to eligible licensees by overpricing it in comparison to the blanket license.

In prior opinions, this court has remarked upon ASCAP's reluctance to proffer a viable per-program license. "Since the early 1940's ASCAP has viewed the per-program license as inconsistent with its business interests." United States v. ASCAP/Application of Buffalo Broadcasting Co., Civ. No. 13-95, 1993 WL 60687, at *54 (S.D.N.Y. Mar.1, 1993) (hereinafter "Buffalo Broadcasting I"). A blanket license is preferable from ASCAP's perspective because that form of license ensures ASCAP of a substantial, fixed cash payment. In contrast, a per-program license renders ASCAP's income dependent upon the level of a station's actual use of ASCAP music and may create a disincentive to the use of ASCAP music. In Turner Broadcasting, 782 F.Supp. at 810, we noted that the per-program license "serves as a counterbalance to ASCAP's market power, which is most clearly exercised by its preference for the blanket license." Thus, "the mandatory per-program option remains an integral part of the injunctive relief provided for by the decree, necessary to provide users with a viable alternative to the blanket license...." United States v. ASCAP, 586 F. Supp. 727, 729 (S.D.N.Y.1984).

A. The Group W and WGN Licenses

In support of its current license proposal, ASCAP cites the terms of the WGN and Group W license agreements. ASCAP argues that applicants and many of the radio stations that accepted those licenses are similarly situated. Thus, ASCAP contends that if applicants were to receive per-program license terms different from those agreed to by the Group W and WGN licensees, ASCAP would run afoul of the Consent Decree, which prohibits ASCAP from entering into licenses that discriminate with respect to license fees or other terms between similarly situated licensees. According to ASCAP, applicants' proposed fees would result in an unwarranted windfall for the applicants and undermine the integrity of the industry-wide bargaining process upon which ASCAP and the radio broadcasters have long relied.

ASCAP also contends that the terms of the Group W and WGN licenses provide an appropriate benchmark of reasonable fees for applicants. In ASCAP's view, there is no better measure of a reasonable fee than the price agreed to voluntarily in arms-length transactions by similarly situated stations. ASCAP asserts that because its fee proposals to applicants replicate the terms agreed to by similarly situated stations, these proposals should be deemed appropriate and reasonable for applicants. ASCAP maintains that the WGN and Group W licenses were the product of fair industry-wide bargaining in which the Radio Music Licensing Committee faithfully sought to protect the interests of stations seeking improvements in the per-program form of license. ASCAP points out that a number of RMLC-member stations selected the per-program license option and several applicant stations have realized meaningful savings by operating under the interim per-program license. ASCAP contends that the decisions of these stations to select the per-program license belie applicants' claim that the per-program license does not offer a legitimate alternative to the blanket license.

Applicants counter that ASCAP's proposed per-program license is unreasonable, anticompetitive, and thus violative of the Consent Decree. According to applicants, they are not situated similarly to the RMLC stations or to those stations that accepted licenses negotiated by the RMLC. Applicants maintain that the fact that ASCAP reached agreement with the RMLC regarding the WGN and Group W licenses does not demonstrate the reasonableness of those licenses for stations, such as applicants, that play relatively little ASCAP music. Applicants contend that, because of the low levels of music use among the applicant stations, their interest in the per-program license was far more pronounced than that of the RMLC, which was primarily concerned with negotiating a favorable blanket license for its music-intensive member stations. Moreover, in applicants' view, the execution of the Group W and WGN licenses by a sizeable number of radio stations who consented in advance with ASCAP to be bound by the license terms that the RMLC negotiated is not an accurate barometer of the reasonableness of those licenses.

1. Statistical Data

ASCAP's defense of its proposed fees is premised on the claim that applicants and the Group W and WGN licensees are similarly situated within the meaning of the Consent Decree. ASCAP does not dispute applicants' assertion that applicants, as a group, aired less music during the periods at issue than did a substantial majority of the commercial radio stations in the United States. (ASCAP Post-Trial Mem. at 23.) However, ASCAP contends that applicants' emphasis on the applicant group's average and median levels of music use is misplaced. According to ASCAP, it is far more significant that, with respect to music use, each individual applicant station is comparable to stations that accepted the WGN and Group W licenses. ASCAP also argues that even when the applicant stations are evaluated as a separate group, their formats and other characteristics are similar to those of the rest of the industry.

a. Applicants' Levels of Music Use

At trial, ASCAP relied heavily upon the testimony of its Chief Economist, Dr. Peter Boyle. Using data compiled as part of ASCAP's distribution survey, Dr. Boyle compared the music use of applicants with that of licensees under the WGN and Group W Licenses.^[7] For each station surveyed, Dr. Boyle determined the average number of ASCAP songs played per hour during the period surveyed. Dr. Boyle further determined that "for each range of feature plays per hour by the NECM and Salem Applicants, there was a similar range among the Industry and the RMLC stations." (Exh. 668, at 4.) For example, Dr. Boyle testified that for ASCAP Survey Year 1995, the survey captured twenty-five applicant stations that operated on interim blanket licenses and played fewer than three ASCAP feature performances of music per hour. (Tr. 231-32; Exh. 669A.) The survey also captured 241 Group W blanket licensees in the same range. (Tr. 231-32.) Extrapolating from these samples, Dr. Boyle estimated that those twenty-five applicants represented a total of approximately 150 to 175 applicants and that the 241 Group W licensees represented about 900 Group W licensed stations. (Tr. 232.) Dr. Boyle also concluded that there were similarities in music use between applicants and licensed stations for Survey Years 1990 and 1991. (Exh. 668, at 4; Exh. 669K-N.)

Although Dr. Boyle's interpretation of the survey data focused on the average number of feature performances per hour, he also analyzed the data presented by applicants, which considered the number of weighted hours containing at least one ASCAP feature performance. Dr. Boyle found that four NECM applicant stations captured in the 1995 survey were expected to play one ASCAP feature in 0-to-10% of their weighted hours as compared to ninety-nine Group W stations. (Exh. 688.) Dr. Boyle projected that a total of twenty-three NECM stations and 389 Group W licensees could be expected to play music in this range. (Id.) Dr. Boyle also projected that a total of forty-five applicant stations were expected to utilize ASCAP feature music within the 80%-to-100% range. (Id.)

Dr. Boyle criticized applicants for relying upon a comparison of median levels of music use to demonstrate that applicants and the RMLC stations were not similarly situated. According to Dr. Boyle, "[t]he average may be higher for the Group W licensee[s], but when you look at it line by line for the applicants, again, there are a lot of stations in the Group W license using music in a similar fashion...." (Tr. 347-48.)

Applicants do not seriously contest Dr. Boyle's conclusions that they used ASCAP feature music over a broad range of levels or that their levels of music use fell within the same ranges as those of some of the WGN and Group W licensees. However, applicants strenuously dispute the significance of Dr. Boyle's findings. Applicants challenge Dr. Boyle's methodological decision to compare the relevant groups' ranges of music use. Applicants point out that Dr. Boyle has been unable to offer any published academic, statistical, or economic literature to support the validity of his range-based analysis. According to applicants, the most appropriate method of determining whether the applicants and the RMLC stations, or those stations that executed extension agreements, were similarly situated is a comparison of the respective groups' median levels of music use.

Applicant's data-analysis expert, Ms. Barrie Kessler, testified based on her examination of the survey information that during survey years 1990 and 1991, approximately 85% of RMLC stations used ASCAP feature music in 60% or more of their weighted program hours but that fewer than one-fourth of applicants reached this level of music use. (See Exh. 711.) Ms. Kessler also found that in 1990 and 1991 combined, the median RMLC station featured at least one ASCAP composition in 97% of its weighted program hours and the median industry station featured at least one composition in 96%, while the median NECM station played at least one ASCAP feature performance in approximately 46% of its weighted hours and the median Salem station featured at least one composition in 36% of its hours. (701A, tbls. 1A and 1B). Dr. Michael Levitan, an expert in statistics, opined that the disparities between applicants' music use and that of the industry and the RMLC stations are statistically significant. (Tr. 673-75; Exh. 703, at 5.)

Last, applicants contend that ASCAP's own analysis of the survey data confirms that Group W and WGN licensees featured ASCAP compositions almost three times as frequently as did the applicant stations. In support of this assertion, applicants, relying on Dr. Boyle's analysis of feature plays per hour, calculate that the NON-RMLC Group W and WGN licensees broadcast an average of 6.46 ASCAP features per hour in contrast to the applicant stations, which averaged 2.3 ASCAP features per hour. (Apps. Post-Trial Mem. at 22; Tr. 282-84, 291-92.)

Both ASCAP and applicants have relied upon and extensively analyzed ASCAP's distribution surveys. The parties each processed the survey data differently and drew contradictory conclusions as to whether applicants' use of music suggests that applicants and the RMLC member stations, or those stations that executed extension agreements, were similarly situated. However, neither side has offered persuasive evidence that the other's substantive statistical findings are in error.

ASCAP has presented sufficient uncontroverted evidence to persuade us that a number of stations which operated under the Group W or WGN licenses used music at levels that corresponded to the applicant stations' levels of music use.^[8] Put differently, we accept ASCAP's claim that if one were to select any individual applicant, one could also identify Group W or WGN licensees with corresponding levels of music use.^[9]

Our finding that the music use levels of many individual applicant stations were comparable to those of Group W and WGN licensees does not compel us to reject the applicants' factual assertion that, as a group, applicants played significantly less feature music than did either the RMLC stations or the industry as whole. It is apparent from applicants' analysis of the 1990 and 1991 survey data that the median applicant station featured ASCAP music in approximately half as many weighted hours as did the median RMLC station or the median industry station. (701A tbls. 1A and 1B.) Applicants have also established, and ASCAP has not disproved, that for survey years 1990 and 1991 combined, less than 10% of the applicant stations played at least one ASCAP song in more than 90% of their weighted hours, while 68% of the RMLC stations played music at or above that level. (Exhs. 701A; 711.) Last, ASCAP's own analysis of the number of feature ASCAP songs played per hour reveals that in survey year 1991, for example, the Group W and WGN licensees played ASCAP compositions almost three times as frequently as did applicants. (Ex. 668, tbl. III.) Considered together, these and the other related statistics presented by applicants demonstrate that applicants, as a group, used ASCAP feature music at significantly lower levels than the RMLC stations or the Group W and WGN licensees, as group. The fact that ASCAP can construct a subgroup from the Group W and WGN licensees whose music use approximates that of the median applicant does not refute applicants' claim that, as to music use, the applicant group is dissimilar to the RMLC group and the rest of the industry.

b. The Applicant Stations' Formats and Market Sizes

In addition to contending that applicants' levels of music use were comparable to the levels of numerous stations which accepted the RMLC-negotiated licenses, ASCAP argues that applicants were similar to the Group W and WGN licensees with respect to format and appeal to large-audience markets. At trial, Robert Unmacht, the publisher of the M Street Journal, a radio-industry publication, offered his opinion that in 1992, the NECM Applicants constituted 31% of the religious-formatted stations in the industry, 50% of the fine arts-formatted stations, and 4% of the Spanish language-formatted stations. (Exhs. 665, 666.) Mr. Unmacht also observed that in 1995, 77% of the NECM Applicants were represented in large markets as compared to 58% of the RMLC stations. (Exh. 663.)

Applicants take issue with Mr. Unmacht's decision to characterize music intensive gospel and contemporary Christian music stations as having the same "religious" format as the talk-oriented Christian stations that comprise the majority of the applicant group. Applicants also argue that Mr. Unmacht's conclusions are flawed because he did not restrict his analysis to RMLC stations but rather included those stations that executed extension agreements with ASCAP. Applicants point out that 4.2% of the RMLC stations had religious, fine arts, or Spanish language formats, while 87% of the applicant stations employed one of these formats. (Exh. 781.)

We find it unnecessary to resolve the parties' dispute over whether the formats and market sizes of the applicant stations are similar to those of either the RMLC stations or the Group W and WGN licensees. The characteristic that is relevant to our determination of reasonable fees is the level of the applicants' use of ASCAP music. The amount of such music a station plays dictates whether that station will select a per-program or blanket license and, if a per-program license is selected, establishes the fee that must be paid to ASCAP. While differences between the formats and market sizes of the applicant stations and those that accepted RMLC-negotiated licenses obviously affect their revenues, these factors do not need to be separately considered if their use fees are computed as a percentage of their revenues, as we conclude should be done here. We therefore find the ASCAP survey data reflecting music use levels to be far more probative than format and market comparisons in determining whether stations are similarly situated for fee-setting purposes.

2. ASCAP's Nondiscrimination Argument

Relying on the factual premise that individual applicants and individual Group W and WGN licensees were similarly situated, ASCAP presses the argument that ASCAP may not, consistent with the Consent Decree, offer applicants a fee proposal with more favorable per-program terms than are contained in the Group W and WGN licenses. To do so, according to ASCAP, would create a situation in which individual radio stations, though similarly situated with respect to music use, would be subject to different fee structures based only upon whether they had accepted the RMLC-negotiated licenses or joined the applicant group. Such a scenario, ASCAP maintains, would be contrary to the Consent Decree's nondiscrimination provisions. Moreover, ASCAP contends that, to the extent stations which accepted the Group W and WGN licenses are disfavored in comparison to their similarly situated competitors among the applicants, the integrity of the collective, industry-wide bargaining process is imperiled.

a. Textual Grounds

In support of the theory that it is bound to offer applicants the same terms as it offered similarly situated Group W and WGN licensees, ASCAP invokes Sections VIII, IX(C), and IV(C) of the Consent Decree. (ASCAP Mem. at 24.) However, neither Section VIII, nor Section IV(C) meaningfully advances ASCAP's nondiscrimination argument. Section VIII provides:

Defendant ASCAP, in fixing its fees for the licensing of compositions in the ASCAP repertory, is hereby ordered and directed to use its best efforts to avoid any discrimination among the respective fees fixed for the various types of licenses which would deprive the licensees or prospective licensees of a genuine choice from among such various types of licenses.

Quite clearly, as the court has previously observed, "Section VIII requires ASCAP to avoid discrimination between the fees fixed for a blanket and per-program license which would have the effect of depriving licensees of a genuine choice between the two licenses." Buffalo Broadcasting II, 157 F.R.D. at 201. Section VIII offers no basis for ASCAP's claim that the Consent Decree requires "that ``applicants' (i.e., stations) which are ``similarly situated' be treated in a similar manner." (ASCAP Mem. at 15.)

Section IX(C) is similarly inapposite. Section IX(C) provides that once the court establishes a reasonable fee in a rate proceeding, ASCAP is bound to offer that fee to all users situated similarly to the applicants, except those stations that previously executed licenses without court intervention for the same period. This Section cannot be read to stand for the proposition that ASCAP must offer to applicants only those terms that ASCAP negotiated with the RMLC.^[10]

ASCAP's citation to Section IV presents a closer issue. Section IV(C) enjoins ASCAP from

[entering into ... any license for rights of public performance which discriminates in license fees or other terms and conditions between licensees similarly situated.]

The clear import of this provision is that ASCAP may not strike different licensing deals with similarly situated stations. See United States v. ASCAP/Application of Capital Cities/ABC, Inc. and CBS, Inc., 831 F. Supp. 137, 146 (S.D.N.Y.1993). Arguably then, ASCAP was indeed obligated to offer applicants the same terms as it offered the Group W and WGN licensees, since ASCAP's statistical analysis demonstrates that each applicant station was situated similarly to one or more of the licensees. However, ASCAP's suggestion that Section IV(C) precludes applicants' from obtaining a judicial determination of licensing fees or that Section IV(C) compels this court simply to adopt ASCAP's fee proposal, whether reasonable or not for applicants, is inconsistent with Section IX (A).

b. Prudential Considerations

ASCAP devotes significant effort to chronicling the "lengthy and productive history" of the industry-wide bargaining process. (ASCAP Mem. at 8.) According to ASCAP, the "dynamic efficacy" of this system will be imperiled if applicants succeed in obtaining more favorable license fee rates than ASCAP has offered to the rest of the industry. (Id. at 9.) ASCAP prophesies a scenario in which numerous radio stations, emboldened by the current applicants' success, will eschew participation in industry-wide negotiations and "await the results of litigation or future negotiations, secure in the knowledge that they could always accept the industry-negotiated licenses at any later time if they are thereafter unsuccessful in Court or at the bargaining table." (Id. at 25-6.) According to ASCAP, the result would be a balkanization of the radio industry into competing bargaining units and ultimately an end to "the rational negotiations which have been so successful for radio." (Id. at 40.)

We are not insensitive to the significant challenges ASCAP and the commercial radio industry face in reaching agreement on music licensing fees. We agree that ASCAP's "industry-wide" negotiating strategy has in the past proved to be an effective tool. However, we decline to accept ASCAP's suggestion that by reaching agreement with the RMLC, ASCAP can in effect also bind dissenting stations by subsequently invoking the sanctity of industry-wide negotiations. In our view, ASCAP's license agreements with the RMLC can fairly be characterized as the product of "industry-wide" negotiations in the loosest sense only. ASCAP does not dispute that during the Group W negotiations, the RMLC formally represented less than a quarter of the commercial radio stations in the United States. (ASCAP Mem. at 11.) Moreover, it does not appear that the members of the RMLC were selected with any precision to ensure that they reflected a truly representative cross section of the industry. As Richard Harris, chairman of the RMLC during the Group W negotiations, observed that "being a voluntary job, there is no compensation for anyone involved and you hopefully get representatives of the industry that — my goal was always to get representatives in as broad a spectrum of the industry as I could...." (Exh. 906 at 13.) Mr. Harris added:

I would say that our committee was somewhere in the eight to 12, and it was elastic too, because people would give their undying commitment and then find that they couldn't just get to the meeting and as in all typical volunteer organizations, so that there was probably a hard-core group who sat at the negotiating table of about four or five people including myself.

(Exh. 906 at 19.) When queried as to how one becomes chairman of the RMLC, Mr. Harris remarked: "Not to be facetious, but very honestly, I would like to find out myself." (Exh. 906 at 22.) Mr. Harris also candidly stated that he was unaware of exactly how many stations the RMLC did, in fact, represent. (Exh. 906 at 14.) Although we note that the RMLC seemed to be far clearer as to the identity of its contributors than those it represented (Exh. 906 at 16), we do not doubt that the RMLC sought to advance the broad interests of the industry as whole. However, it does not appear that the RMLC ever ascertained with any certainty the licensing objectives of its many member stations or the unrepresented industry stations. Given the RMLC's representation of far less than a majority of radio broadcasters, its self-appointed membership, and the apparent absence of a formal method of communicating with its members or the industry, we find that ASCAP's frequently-repeated refrain that the ASCAP-RMLC negotiations were "industry wide" is overstated.^[11] Therefore, we will not be deterred from scrutinizing ASCAP's license proposals by ASCAP's dire predictions of chaos. Either ASCAP's proposed per-program license fees comport with the Consent Decree, or they do not. The possibility that a ruling adverse to ASCAP in the instant matter may complicate future negotiations and encourage additional applicants to exercise their right to obtain a rate setting by the court, while perhaps unfortunate, cannot be allowed to dominate our analysis. Whether "industry-wide" negotiations continue to receive the court's imprimatur as an appropriate means of setting music license fees depends entirely upon whether this process yields acceptable licenses. If ASCAP feels that "[i]t is important that the process not be jeopardized," (ASCAP Mem. at 40), ASCAP need only ensure that the process produces agreements that offer reasonable fees for all music users.

3. Reasonableness of ASCAP's Proposed Per-Program Fee

Under the judicial rate-setting mechanism established by the Consent Decree, ASCAP has the burden of proof to demonstrate the reasonableness of its fee proposals. (Consent Decree, Section IX.) However, the Consent Decree does not provide any formula or criteria to aid the court in gauging the reasonableness of a royalty fee.

In Buffalo Broadcasting v. ASCAP, 744 F.2d 917, 926 (2d Cir.1984), an antitrust action in which the plaintiffs alleged that ASCAP's blanket form of license was anticompetitive because, inter alia, the per-program license did not constitute a viable alternative to the blanket license, the Second Circuit observed that the determination of whether a license is excessively priced turns on an evaluation of "whether the price of such a license, in an objective sense, is higher than the value of the rights obtained." Later, in ASCAP v. Showtime/The Movie Channel, 912 F.2d 563, 569 (2d Cir.1990), a rate proceeding under the Consent Decree, the Court provided additional guidance, indicating that the task of assessing a fee's reasonableness necessitates an appraisal of "fair market value," which the court defined as "the price a willing seller would agree to in an arms-length transaction." Recognizing that the market for music licenses does not typify the model of a competitive market and lacking the instruments with which to construct from start a model to price the rights at issue, we have in the past found that it is appropriate to evaluate the reasonableness of a fee proposal by examining prior agreements negotiated between the parties or those similarly situated. See Capital Cities, 831 F.Supp. at 144; Showtime, 912 F.2d at 577 (opinion of the trial court) ("we must look to very imperfect surrogates, particularly agreements reached either by these parties or by others for the purchase of comparable rights").

In a previous rate-court proceeding, we noted that prices voluntarily agreed upon in arms-length negotiations by stations situated similarly to applicants "offer the only palpable point from which to proceed towards an estimation of fair value...." Capital Cities, 831 F.Supp. at 145. However, agreements between ASCAP and parties situated similarly to an applicant group are not necessarily determinative of a reasonable fee structure for that group. The Second Circuit has remarked that "[t]he opportunity of users of music rights to resort to the rate court whenever they apprehend that ASCAP's market power may subject them to unreasonably high fees would have little meaning if that court were obliged to set a ``reasonable' fee solely or even primarily on the basis of the fees ASCAP had successfully obtained from other users." Showtime, 912 F.2d at 570. Where, as here, agreements between ASCAP and a third party form the starting point of a rate-setting inquiry, the court must consider any "distinctive conditions" impacting those agreements in order to determine whether those agreements provide a reliable benchmark of reasonable royalties. Capital Cities, 831 F.Supp. at 145.

a. The Group W and WGN Licenses as a Benchmark of Reasonableness

ASCAP argues that the license terms negotiated by the RMLC and ASCAP constitute a benchmark for determining reasonable fees in this proceeding. According to ASCAP, the Group W licenses were the product of good-faith negotiations and were accepted by a vast majority of radio stations, including many that were situated similarly to applicants. Not surprisingly, applicants contend that the RMLC-negotiated licenses do not accurately reflect a reasonable per-program fee. According to applicants, the RMLC failed to bargain vigorously for this form of license because the RMLC's paramount objective was to secure a favorable blanket license for the music-intensive stations that comprised the overwhelming majority of the broadcasters that the RMLC represented. Moreover, applicants assert that the non-RMLC stations' acceptances of the Group W licenses are not probative of a reasonable per-program fee because those stations consented to be bound before they knew the terms of the Group W licenses.

i. Applicants' Role in the Group W Negotiations

The parties dispute whether applicants' representatives were excluded from the Group W negotiations, and if so, by whom and for what reason. Applicants argue that ASCAP and the RMLC objected to applicants' participation in the negotiations because they both wished to avoid including an independent third party that would forcefully pursue changes in the per-program license. ASCAP counters that although it was not in favor of allowing applicants to participate because applicants' presence would slow negotiations, it deferred to the RMLC, which ultimately rejected applicants' request. Moreover, according to ASCAP, had applicants genuinely desired to participate in the Group W process, their representatives could have volunteered to join the RMLC and attended the negotiations as members of that committee.

The evidence demonstrates that, after several meetings with the RMLC, the NRBMLC, which represented applicants, requested that ASCAP allow it to participate in the Group W negotiations as an independent party. (Exh. 268.) ASCAP referred the NRBMLC's request to the RMLC, stating that it would acquiesce if the RMLC favored including the NRBMLC, while indicating its preference to maintain the bilateral structure of the negotiating process. (Exh. 269.) ASCAP reiterated this position in subsequent meetings with the RMLC. (Exh. 273-43 at D02570 ("[w]e do not want [the NRBMLC] if it will make it harder to make a deal with the Committee.")) The RMLC shared this view, (Exh. 272), and thus the NRBMLC was not invited to participate in the negotiations. The applicants subsequently made no effort to affiliate with the RMLC.

The causes for applicants' nonparticipation in the Group W negotiations are of little relevance to the issue of whether the Group W per-program license provides a benchmark for determining reasonable fees in this proceeding. Although it is clear that both ASCAP and the RMLC opposed the NRBMLC's participation in the Group W negotiations, we are not prepared to find that something other than the desire to simplify and expedite the negotiation process motivated either of these parties to exclude the NRBMLC. Contrary to applicants' contention, the RMLC's actions in this regard cannot properly be construed as evidence that the RMLC had no interest in seeking a reduction in the per-program license fee. However, we do not find it surprising that applicants decided not to designate a representative to join the RMLC. Apparently, applicants perceived that the RMLC would not represent their interests and therefore chose to negotiate with ASCAP independently, as was their right under the Consent Decree.

ii. The RMLC's Commitment to the Per-Program License

At trial, ASCAP presented evidence sufficient to establish that during the Group W negotiations the RMLC sought improvements in the per-program license. (See, e.g., Exh. 237-38 at D2601-03; 237-40 at D3041-44; 273-41 at D2995-3000; 273-45 at D2943-54; Tr. 48-49.) ASCAP rebuffed these efforts by insisting that any reduction in the per-program fee would have to be offset by a corresponding increase in the blanket rate (Tr. 250-51; Exh. 906 at 50-55). In the face of ASCAP's refusal to reduce the per-program fee unconditionally, the RMLC made a calculated decision to withdraw its per-program fee demands rather than accede to an increase in the blanket fee or pursue its objectives in a rate-court proceeding. (Exh. 906, at 54-55.)

We find no reason to doubt Mr. Harris' statement that he and the RMLC were seeking "the greatest good for the greatest number." (Exh. 906, at 26.) However, from the RMLC's perspective, its goal of value maximization could not be achieved through an agreement with ASCAP that included an increased blanket license fee. According to Mr. Harris, the greatest good for the greatest number meant obtaining the best possible blanket license terms. (Exh. 906, at 54.)^[12] Given the high level of music use of the RMLC-represented stations and the industry as a whole, which impelled the majority of stations to accept a blanket license, (Tr. 107-08), because of its relative cost effectiveness, there is some reason to suspect that the RMLC may have subordinated the interests of the minority seeking reduced per-program rates to the interests of the majority favoring a blanket license. Counsel for the RMLC made the RMLC's position clear when he stated in a September 1990 letter to ASCAP, "I believe we can safely say that the Committee represents the interests of the bulk of the commercial radio stations in the United States. Nevertheless, as you are aware, the radio industry is extraordinarily diverse and there may be certain stations (such as those represented by the Religious Broadcasters' Committee) whose interests are not represented by the Committee." (Exh. 272.) In addition, the RMLC argued during the WGN negotiations that the per-program license was not a genuine choice, (Exh. 273-17, at D10017), but nonetheless consented to the terms of that license without substantial modification. (Tr. 85-86.) The RMLC subsequently accepted a similar per-program fee structure during the Group W negotiations.

Given the divergence between the bargaining objectives of the RMLC-represented stations and the applicants, which derive from the contrast in their levels of music use, we cannot find that these groups were similarly situated with respect to their interests in the per-program license.^[13] Moreover, we cannot conclude that the Group W per-program license terms reflected the RMLC's conviction that those terms were reasonable for stations with limited music use. Thus, although the terms agreed to by the RMLC and ASCAP during the Group W negotiations are at least some evidence of their reasonableness, it would be inappropriate merely to adopt those terms as a benchmark of reasonable per-program fees for the applicant group.^[14]See Buffalo Broadcasting II, 157 F.R.D. at 199, 1993 WL 60687, at *61 (finding that "since the individual stations were ... overwhelmingly interested in the blanket license, the outcome of these negotiations is not a proper indicator of a reasonable rate for the per-program license").

b. Acceptance of the Per-Program Licenses by Similarly Situated Stations

Although we have determined that the terms of the RMLC-negotiated licenses cannot be used as a benchmark of reasonable per-program fees in this proceeding, the fact remains that numerous stations situated similarly to applicants with respect to music use did accept Group W licenses. According to ASCAP, because these stations agreed to the Group W licenses, the comparable terms ASCAP has proposed for applicants should be deemed appropriate. However, acceptance of the RMLC-negotiated Group W licenses by individual radio broadcasters is of little probative value as to the reasonableness of the fees set forth in those licenses because the RMLC-represented stations and those stations that executed extension agreements with ASCAP both apparently committed themselves to accept the licenses before the terms of those licenses had been negotiated. (Exh. 12; ASCAP Reply Mem. at 19 (arguing that the RMLC members and those that executed extension agreements were not in analytically distinguishable positions)). In the absence of a mechanism that permitted these stations to signal their approval by ratifying the agreement between ASCAP and the RMLC, we do not find the stations' ex ante assent to the licenses to be particularly persuasive evidence of the reasonableness of the fee arrangements that the RMLC and ASCAP subsequently adopted.^[15]

B. The Proposed Per-Program Fees as Objectively Reasonable

ASCAP further argues that, in addition to being reasonable because of their acceptance by the vast majority of stations, the proposed fees at issue are objectively reasonable. To demonstrate the appropriateness of the proposed per-program fees, at trial ASCAP offered evidence that a hypothetical per-program station with adjusted gross revenues of $100,000 would pay $0.27 per week, or $13 per year, to play a one-hour program with feature music on Sundays. (Exh. 667-F.) A one-hour program with feature music on weekday mornings would cost $28 per year. (Tr. 191-94.) ASCAP also presented evidence to demonstrate that stations operating under per-program licenses paid relatively small fees in comparison to their revenues. For example, according to ASCAP, in 1994, fourteen selected Salem Media Group stations with total adjusted gross revenues of $27 million paid $123,673 in fees under their interim per-program licenses. (Exh. 667-H.) ASCAP observes that had these stations operated under the Group W blanket license, they would have incurred $433,376 in fees. (Id.)

Applicants point out that in computing the per-program fees that a hypothetical station would pay for playing one hour of ASCAP music, ASCAP assumed an annual revenue for that station of $100,000. However, the evidence demonstrates that the average revenue of per-program stations in 1994 exceeded $1,500,000. (Exhs. 667-B; 667-E.) Thus, ASCAP's model significantly understates the fees that an average per-program station would incur. In addition, applicants contend that ASCAP's comparison of several Salem Media stations' per-program fees to those stations' annual revenues is unilluminating absent data on the level of music used by each station. More fundamentally, applicants argue that ASCAP has presented no evidence from which this court can make a reasoned determination of the value of the rights offered by ASCAP.^[16] We agree with applicants that absent some quantification of the benefits obtained by applicants from their licensed use of ASCAP music, and of the benefits ASCAP's members enjoy from having their compositions aired, the court lacks the information necessary to assay the objective, or absolute, reasonableness of the per-program license fees.

Although we are not persuaded by ASCAP's efforts to demonstrate that its proposed fee rates are objectively reasonable, we find equally unconvincing the applicants' claim that the proposed per-program fees extract an unreasonably high fee relative to the blanket license. It is true that a station which plays ASCAP feature music in only 30% of its weighted program hours would incur a per-program fee amounting to 94% of the blanket license fee because the two fee rates are in a ratio of approximately 3:1. (Apps. Mem. at 24.) However, we discern nothing inherently unreasonable in such a ratio. A per-program license rate is not unreasonable merely because it is a substantial multiple of the blanket fee rate. Buffalo Broadcasting Co. v. ASCAP, 744 F.2d 917, 927 (2d Cir.1984). Because of the substantial difference between many stations' total revenues and their revenues subject to fee, if the per-program rate were not several times as high as the blanket rate, the blanket license, with its freedom from administrative burdens, would represent a reasonable alternative for only those stations with the most music-intensive formats.

C. ASCAP's Proposed Per-Program Licenses and the Consent Decree's Genuine-Choice Requirement

1. Applicants' Position

As noted above, applicants contend not only that ASCAP has failed to demonstrate that the proposed per-program license fee rates are reasonable within the meaning of Section IX but that the fees should be rejected because they violate the genuine-choice provision of Section VIII. Applicants assert that under ASCAP's proposal, the per-program license would be a theoretically viable alternative for stations that broadcast feature ASCAP performances in at most approximately 32.5% of their weighted program hours. (Exh. 777.) However, applicants also contend that given the administrative burdens associated with the per-program license, that form of license is realistically feasible only if a station plays ASCAP music in no more than 20%-25% of its weighted hours. Applicants observe that, according to ASCAP's 1995 survey data, only approximately 8% of the radio industry used music at or below the 32% level, (Exh. 701A.),^[17] whereas the median station in the industry played ASCAP features in approximately 93% of its weighted hours and the average station did so in 83%. (Exhs. 701A; 706; 705.)

Applicants assert that the highly limited utility of the proposed per-program license demonstrates that it is not a "genuine choice" for the typical radio station or even for a significant segment of the industry. Thus, according to applicants, the proposed licenses do not comply with the court's prior decisions in Buffalo Broadcasting I and Buffalo Broadcasting II.

In the Buffalo Broadcasting proceeding, ASCAP proposed to the All-Industry Committee of Local Television Stations, which represented approximately 960 local television stations, and to twenty local stations that were owned by the three major networks a per-program license rate that was substantially higher than the proposed blanket license rate. In Buffalo Broadcasting I, Magistrate Judge Dolinger found that ASCAP's proposed per-program license was impractical for all but a few stations that used very little ASCAP music. Buffalo Broadcasting I, 1993 WL 60687, at *57. Magistrate Judge Dolinger then held, as a matter of law, that ASCAP's proposed per-program license violated the "genuine choice" provision of the Consent Decree. Magistrate Judge Dolinger construed Section VIII of the Consent Decree to require that the per-program license fee "be set in such a way that, when compared to the blanket license fee, its use is economically feasible for a significant segment of the industry...." Id. at *66. Magistrate Judge Dolinger concluded that the per-program license for the applicants would be reasonable if it were priced so that the "typical" local television station would pay approximately the same fee under the per-program license as under the blanket license, exclusive of administrative costs. Id. at *67.

In Buffalo Broadcasting II, we found that with respect to the meaning of the genuine choice provision, Magistrate Judge Dolinger's legal conclusions withstood de novo review and his factual findings were not clearly erroneous. Moreover, we observed that "[t]o condone a pricing structure that makes the per-program license illusory for all but a few stations would serve to nullify the necessary protection from restraint of trade that the per-program license offers." Buffalo Broadcasting II, 157 F.R.D. at 202. Relying on the reasoning in Buffalo Broadcasting I, applicants argue that because the per-program license fee proposed by ASCAP in this case is not equivalent to the blanket license fee for the typical local station in the industry, ASCAP's proposal violates the Consent Decree.

2. ASCAP's Position

To refute applicants' allegation that the proposed per-program fees contravene Section VIII of the Consent Decree, ASCAP advances two principal arguments. First, ASCAP asserts that the proposed fee structure does in fact provide a viable alternative to the blanket license and thus satisfies the Consent Decree. ASCAP points out that the number of per-program licensees has risen from 55 in 1979 to 858 in 1995. (Exhs. 667-C, 667-E.) In 1994, these per-program stations generated 14.23% of the total reported gross revenues of all licensed stations. (Exh. 667-B.) ASCAP also emphasizes that the fourteen Salem Media Group stations discussed above reaped a 71% savings in 1994 by operating under interim per-program licenses rather than Group W blanket licenses. (Exh. 667-H.)

Second, according to ASCAP, we should not import to the radio-licensing context Magistrate Dolinger's conclusion that the genuine-choice provision is satisfied by a pricing structure which would enable the typical local station to operate under a per-program license at a cost, exclusive of administrative expenses, roughly equivalent to that of the blanket license. We agree with ASCAP that the television and radio industries are vastly different and that what is appropriate with respect to music licensing in the former is not necessarily appropriate for the latter. ASCAP claims, and we agree, that Magistrate Judge Dolinger's use of the "typical" station as a benchmark for determining equivalence between the per-program and blanket forms of license is inapplicable in the radio-industry context because "no station or group of stations is ``typical.'" (ASCAP Mem. at 56.) In the radio industry, over one-half of the stations broadcast ASCAP music in more than 90% of their weighted program hours, (Exhs. 701A; 707), while the music use of the others spanned the entire remainder of the spectrum, with many stations below the 30% level. (Exhs. 701A; 707; 713.) We note that applicants have offered no persuasive evidence that a similarly high degree of diversity in levels of music use characterized the television industry, which Magistrate Judge Dolinger analyzed in Buffalo Broadcasting I. With respect to the radio industry at least, we find the concept of a "typical station" to be largely unhelpful.

3. Assessing ASCAP's Proposal

With the exception of applicants' claims as to the administrative costs incurred by per-program licensees, ASCAP does not dispute applicants' factual assertions regarding the level of music use at which the per-program license becomes a viable option for a radio station or the percentage of stations currently operating under that form of license, (ASCAP Reply Mem. at 10), and we find these assertions to be adequately supported by the record evidence. However, applicants' contention that the per-program license is a cost effective option for only those stations using ASCAP feature music in less than 20% to 25% of their weighted hours is insufficiently substantiated.^[18] Applicants rely on the rather vague testimony of Russell Hauth, executive director of the NRBMLC. (Apps. Mem. at 42.) At trial, Mr. Hauth stated, "I use the 20 percent rule. If a station's music usage appears to be 20 percent or fewer of its weighted hours, of its total weighted hours, it becomes a candidate for per program license. ... [Y]ou can't be exacting about this. There's something of an art or a — feel...." (Tr. 829.)^[19] However, applicants offer no empirical data on the administrative costs associated with the per-program license, and we decline to make a factual finding based on Mr. Hauth's "feel" for those costs, which, we note, ASCAP has demonstrated can be minimal for stations that use computerized play lists. (Tr. 792-94.)

We next consider whether ASCAP's proposed fee rates satisfy Section VIII. The pricing relationship between the blanket license and the per-program license defines what proportion of stations will choose the per-program license as an economically attractive alternative. However, the generality of Section VIII offers no specific answer to the question of how many stations must be able to avail themselves of a per-program license before the "genuine choice" language of that section is satisfied. Clearly, as Magistrate Judge Dolinger observed, "[t]he reference to ``a genuine choice' ... does not mean that the per-program license must be cheaper than the blanket license for all or even most stations...." Buffalo Broadcasting I, 1993 WL 60687, at *65. It is also evident that the Consent Decree's pro-competitive objectives would be ill-served by a licensing regime that makes the per-program option a practical alternative for only a few stations. Buffalo Broadcasting II, 157 F.R.D. at 202.

Applicants urge the court to adopt Magistrate Judge Dolinger's typical-station methodology in the instant proceeding. According to applicants, a station playing ASCAP feature music at slightly below the "typical," or median, level of at least one ASCAP song in 95% of its weighted program hours, should pay the same fee under the blanket and per-program licenses. Applicants maintain that use of the Buffalo Broadcasting model in this proceeding would ensure that the per-program license is a genuine choice for a significant segment of the radio industry.

In light of applicants' heavy reliance on Buffalo Broadcasting I, we think it necessary to emphasize the limited scope of the current rate proceeding. In the Buffalo Broadcasting litigation, we accepted the principle that the fee relationship between the two forms of license should be set so as to provide a significant segment of that industry with an economically feasible per-program option when compared to the blanket license. Buffalo Broadcasting II, 157 F.R.D. at 202-03. Thus, we noted that affording a significant segment of the industry the opportunity to take the per-program license would further the Consent Decree's "goal of price stabilization in the face of ASCAP's recognized market power." Id. at 202 (quoting Magistrate's Report). However, in a decision in which we denied in part applicants' motion for summary judgment in this proceeding, we observed:

All that Buffalo Broadcasting held was that, in the context of an industry-wide application by the local television stations, the genuine choice provision is satisfied if the per-program license is an economically viable alternative for a significant number of those stations....

The equivalence formula devised by Magistrate Dolinger was developed to implement his interpretation of the genuine choice provision under the circumstances presented in the Buffalo Broadcasting application. Accordingly, it was designed to create a fee structure that was reasonable across the entire local television industry.

United States v. ASCAP/Applications of Salem Media of California, Inc. and New England Continental Media, Inc., 902 F. Supp. 411, 418 (S.D.N.Y.1995).

In contrast to the applicants in the Buffalo Broadcasting proceeding, the current applicants constitute but a small fraction of the local stations in their industry. Given that "[t]he goal of any fee-setting proceeding is to arrive at a reasonable fee for the applicants before the court," Id., the task in this rate proceeding is to determine whether ASCAP's proposed per-program fee rates provide a genuine choice for the applicant stations, not whether those fee rates make the per-program license available to a significant segment of the entire radio industry. The RMLC and ASCAP were free to agree to any terms they wished, regardless of the Consent Decree. Id. at 419. As applicants observe, "[t]his is a wholly retrospective proceeding, limited to the named Applicants." (Apps. Post-Trial Mem. at 31.)

We agree with applicants that the legal principles articulated in Buffalo Broadcasting I apply with equal force to the radio industry. Were this an industry-wide rate proceeding, Section VIII's nondiscrimination provision would be satisfied by a per-program fee that offered a genuine choice for a significant segment of the industry.^[20] Here, however, the applicant group constitutes a subset of the industry, and a per-program fee rate that serves as a reasonable alternative for a significant segment of the stations in that applicant group is appropriate.

The data presented by applicants demonstrate that a significant segment of the applicant stations did, in fact, have a genuine choice between the two forms of license. During the three survey years for which applicants presented data, an average of approximately twenty-nine percent of the NECM stations captured in the ASCAP survey played ASCAP features in thirty percent or less of their weighted program hours. (Exhs. 701A; 713.) The available data regarding the Salem applicants indicate that an average of approximately thirty-three percent of those stations played ASCAP features in thirty percent or less of their weighted hours. (Id.) Given that approximately thirty-two percent is the "cross-over" point below which the per-program license becomes more cost effective than the blanket license, it is apparent that the proposed perprogram licenses were an option for slightly less than one-third of the applicant stations, which is clearly a substantial segment of the group. Even were we to accept applicants' unsupported claim that inherent administrative expenses associated with the per-program license reduced the "cross-over" point to between twenty and twenty-five percent, the proposed per-program licenses would still be available to more than twenty-one percent of the surveyed applicant stations.

Although we have found that ASCAP may not rely upon the Group W and WGN per-program licenses as conclusive evidence that the rates it has proposed in this proceeding are reasonable, we reject applicants' contentions that ASCAP's proposed fee rates contravene the Consent Decree's genuine-choice provision and that we should apply Buffalo Broadcasting's equivalence methodology here. Because ASCAP's proposals as to the feature-music fee component of the per-program licenses provide a genuine choice for a significant segment of the applicant stations, and because applicants have suggested no valid alternative, we find the feature-music fee proposed by ASCAP to be acceptable.

D. The Incidental Use Fee

In addition to their objections to the feature music fee rate, applicants challenge the incidental use fee component of ASCAP's proposed per-program licenses. ASCAP takes the position that the Consent Decree does not empower the rate court to set an incidental use fee that is based on a percentage of the licensee's adjusted gross revenue ("AGR"). As ASCAP observes, in Buffalo Broadcasting II, we vacated the incidental use fee portion of the per-program license set by Magistrate Judge Dolinger. Buffalo Broadcasting II, 157 F.R.D. at 204-05. In that proceeding, we accepted ASCAP's contention that the fixed incidental use fee was, in effect, a hybrid of the per-program and blanket forms of license, which enabled a station to obtain "a mini blanket license" for the unlimited use of incidental music. Id. at 204. Because the incidental use fee was neither a true per-program nor a true blanket license, and the Consent Decree only requires ASCAP to offer those two types of license, we concluded that the court was without authority to impose a license that included a flat fee for incidental music use. Id. at 204-05.

Given this constraint on the court's authority, ASCAP now suggests a consensual arrangement between the parties under which ASCAP would accept in satisfaction of applicants' obligation to pay for incidental music use the fees that the applicants have already remitted pursuant to the interim per-program licenses. In the alternative, ASCAP proposes that the court adopt a true per-program, incidental use fee of 1.82% of the revenue attributable to programs which contained incidental but no feature ASCAP music. (Exh. 276.) According to ASCAP, this proposal would comport with the Consent Decree because stations would pay incidental use fees only for those programs that actually contained incidental music. ASCAP, however, refuses to consent to a determination by this court of a reasonable flat-rate incidental use fee. (ASCAP Reply Mem. at 16.)

Applicants reject ASCAP's first suggestion that they agree to an incidental use payment that is based on the rates incorporated in their interim per-program licenses. Applicants cite Buffalo Broadcasting I, in which Magistrate Judge Dolinger set an incidental use fee of 7.5% of the maximum possible per-program license fee. Magistrate Judge Dolinger arrived at this rate because it corresponded to the percentage of ASCAP's total revenues that ASCAP allocated to its members as compensation for the incidental use of ASCAP music. Buffalo Broadcasting I, 1993 WL 60687, at *77. Applicants contend that under ASCAP's "stand-in-place" proposal, which is predicated on the interim per-program license fee of 0.24% AGR for incidental performances, they would pay 15% of the blanket license fee and thus considerably more than Magistrate Judge Dolinger found reasonable in Buffalo Broadcasting I.

Applicants also assert that ASCAP's performance-based fee proposal of 1.82% of revenue from programs making only incidental use of ASCAP music is untenable because it would require applicants to compile retrospective records of their incidental music uses. Applicants argue that no ASCAP license for at least twenty years has contained this type of incidental use fee, and therefore applicants had no reason to maintain records of this information for the relevant periods. Moreover applicants contend that even had they been given notice of this requirement, the systematic tracking of incidental music use is wholly impracticable and renders ASCAP's proposal unreasonable. Indeed, at trial, ASCAP acknowledged that the record-keeping burden associated with its proposal would not be feasible: "Our proposal ... contemplated an incidental use fee that complied with the decree that was a percentage of the revenue from the programs that contained incidental use ASCAP music. We now know from this trial that that's impossible." (Tr. 764-65.)

Applicants maintain that if the court were to adopt ASCAP's use-based fee proposal, the per-program form of license would cease to be a genuine choice for any radio station because of the impossibility of complying with the record-keeping requirements. Thus, applicants urge the court to reconsider its prior determination that it is beyond the court's authority to order a fixed-rate incidental use fee.

We recognize the problem created by our previous indication that the court is empowered to order ASCAP to offer only a per-program license that includes an incidental use fee component which is based on a station's actual use of incidental music. Buffalo Broadcasting II, 157 F.R.D. at 205. However, such a fee structure would impose an unacceptable reporting burden on per-program licensees and render the per-program form of license an unrealistic alternative. Therefore, ASCAP's proposal must be rejected because an unusable per-program license would be inconsistent with the Consent Decree's procompetitive objectives and the preferential protection the Consent Decree accords the per-program form of license.

After careful consideration, we find it necessary to revisit our determination in Buffalo Broadcasting II and Salem Media. In reaching our conclusions in Buffalo Broadcasting II and Salem Media, we relied on the assumption that, although a fixed fee is an "eminently sensible" method by which to compensate ASCAP for the use of incidental music, it would also be reasonable for a station to monitor and report its incidental music use as required by a performance-based fee. See Salem Media, 902 F.Supp. at 419. Our assumption, at least for the radio industry, now appears to have been unjustified. Given the impracticality of ASCAP's current proposal, the court deems it appropriate to set an incidental use fee that is based on a fixed percentage of the applicant stations' adjusted gross revenues. Such a fee does not directly conflict with any express provision of the Consent Decree and is, in fact, consistent with the Consent Decree's objectives.

Applicants submit that an incidental use fee of 0.06% AGR is appropriate.^[21] Applicants assert that their proposed fee is reasonable because it more than compensates ASCAP for the value that ASCAP, in its distribution of royalties, has assigned to incidental performances. Specifically, according to applicants' expert, Barry M. Massarsky, a former senior economist at ASCAP, ASCAP's distribution of royalty credits for incidental radio performances equaled approximately 1% to 2% of ASCAP's distributions for feature performances between 1992 and 1995.^[22] (Exh. 704, at 7-8.) In terms of the Group W blanket license fee of approximately 1.6% AGR, this range of percentages equates to roughly 0.03% of adjusted gross revenue, about one-half of applicants' proposed incidental use fee of 0.06%.

ASCAP attacks applicants' proposal on two principal grounds. First, ASCAP argues that Mr. Massarsky's analysis is based on the fundamental misconception that the distribution credits that ASCAP uses to disburse copyright royalty income among ASCAP's members accurately reflect the "value" that ASCAP and its licensees place on incidental music uses. (Exh. 678, at 2.) According to ASCAP, "the setting of weights for distribution credits is the product of a process, subject to government and court review, full of choices and compromises among the membership." (Id.) ASCAP maintains that because stations on per-program licenses typically use less feature music than do stations on blanket licenses, incidental music has a relatively higher value for per-program stations. Therefore, ASCAP reasons that incidental music use should generate a proportionately greater share of the fees incurred by per-program licensees.^[23]

Second, ASCAP contends that Mr. Massarsky failed in his analysis to account for the fact that ASCAP applied a "feature multiplier" to the distribution credits ASCAP attributed to feature performances by radio blanket licensees. The purpose of this multiplier was to account for nonbroadcast feature uses of ASCAP music by, for example, bars and restaurants, whose music use ASCAP did not survey directly. According to ASCAP, Mr. Massarsky's error led him to overstate the credits for performances of feature works by blanket-licensed stations in the ASCAP survey and to understate the share of credits attributable to incidental uses.

Clearly, ASCAP's system of royalty distributions was, and continues to be, the product of negotiation among different classes of ASCAP members. Thus, we accept ASCAP's assertion that ASCAP's distributions "were based on factors other than merely yield to ASCAP's members." (ASCAP Mem. at 74.) In addition, there is no dispute that ASCAP's weighting process is subject to government and judicial review. However, we are unpersuaded that ASCAP's distributions do not bear a significant, albeit less than direct, relation to the value ASCAP's members assign to musical performances. Indeed, it would be surprising, and probably inequitable, if there were no correlation between royalty distributions and the relative worth ASCAP's members ascribe to different uses of music.

Absent from ASCAP's critique of applicant's proposal is any alternative, reasoned suggestion as to how incidental music is, or should be, valued for licensing purposes.^[24] Thus, we conclude that an incidental use fee that is predicated on the share of distribution credits allocated by ASCAP to incidental music is appropriate. See Buffalo Broadcasting I, 1993 WL 60687, at *77. Mr. Massarsky's decision not to reduce his calculation of feature credits by the number of credits attributable to nonradio performances of music does not undermine the validity of the valuation principle upon which applicants' proposal rests.^[25] Based on the foregoing, we accept as reasonable applicants' proposal of an incidental use fee of 0.06% of adjusted gross revenues.

E. ASCAP's Proposed Administrative Requirements

Pursuant to the terms of the Group W per-program license, a station must report the following information to ASCAP with respect to any broadcast of recorded nonincidental music: (1) title; (2) name of composer, author and publisher; (3) name of performing artist; (4) name of record company; and (5) all other information as to composer, author and publisher in full as shown on the label. (Exh. 7, Group W Local Station Per Program Radio License ¶ 5; see Exh. 6, WGN Local Station Per Program Radio License ¶ 5.) However, applicants contend that these reporting requirements are unduly burdensome because they obligate stations to report more information than is necessary for ASCAP to determine whether the music is part of ASCAP's repertory. In support of this claim, applicants cite the testimony of ASCAP's former director of radio licensing, David Hochman, who acknowledged that ASCAP has not fully enforced the reporting requirements at issue. According to Mr. Hochman: "Obviously, the more information the station gives us, the easier it is to identify the work. We have never rejected a music log because they didn't have one or two of the items." (Tr. 128.) In addition, ASCAP's monitors, who scrutinize ASCAP survey tapes of radio broadcasts and compile data so that music can be identified, generally rely upon the name of the artist and the title of the performance. (Tr. 271-72.)

Applicant stations that have been operating under an interim ASCAP blanket license will have the right to choose retroactively to adopt and report under the per-program license terms ordered by this court. Thus, contrary to ASCAP's assertion, applicants' challenges to ASCAP's proposed reporting requirements are not moot. In our earlier Opinion and Order addressing applicants' motion for summary judgment, we stated: "ASCAP may only impose those reporting requirements ... that are reasonably necessary to ensure that ASCAP receives adequate and accurate music use information from per-program licensees." Salem Media, 902 F.Supp. at 423. Based upon the evidence presented at trial, we agree with applicants that ASCAP's reporting requirements are excessive. ASCAP has offered no compelling reason why stations should be obligated to report more information than ASCAP relies upon internally to identify its music. Thus, we conclude that ASCAP's proposed reporting requirement should be modified to require that stations must document a composition's title and the identity of either the performer, composer, or recording artist.^[26]

F. Applicants' Remedial Blanket License Proposal

Applicants contend that many of their number had no recourse but to operate under interim blanket licenses during the pendency of this litigation because the interim per-program licenses available to them did not constitute a viable alternative. Although the record is unclear as to the exact number of applicant stations that accepted an interim blanket license, its appears that approximately two-thirds of the applicant stations fall into this category. (Tr. 11.) Applicants do not dispute that the interim blanket licensees did not keep records of their music use and that these stations, with the exception of the ones that may be able to reconstruct their music use based on per-program license records retained by BMI, would be unable to select a final per-program license in this proceeding.

Applicants argue that these stations used significantly less music on average than did the other members of the industry that selected the blanket license.^[27] Therefore, applicants urge that the interim blanket licensees should be given the opportunity to pay proportionately less than those industry stations. Applicants propose a remedial blanket license fee equal to 56.3% of the Group W and WGN blanket license fees.

We reject applicants' request for a remedial blanket license. Having concluded above that ASCAP's proposed per-program license constitutes a "genuine choice" within the meaning of the Consent Decree because it was a reasonable alternative for a significant segment of the applicant group, we see no reason why those applicant stations that selected an interim blanket license are entitled to the remedy applicants propose. Given applicants' acknowledgment that the RMLC obtained fair and reasonable blanket fees from ASCAP, the fact that the interim blanket licensees, as a group, used less music than the rest of the blanket licensees in the industry does not justify selectively discounting the negotiated rate of the blanket license.

We also find that the interim blanket licensees' failure to keep music-use records deprives them of the option to convert retroactively to per-program licenses. These stations' apparently had so little confidence that they would prevail in their challenge to ASCAP's proposed per-program licenses that they did not bother to maintain the records necessary for a per-program license.

CONCLUSION

The court finds the terms of ASCAP's proposed licenses to be reasonable and nondiscriminatory with the following modifications: (1) the incidental music fee under the per-program form of license shall be set at 0.06% of the licensee's adjusted gross revenue; (2) at minimum, music reports shall list the title of the composition and one other identifying characteristic (composer, author, publisher, or performer); and (3) consistent with the court's ruling on summary judgment, Salem Media, 902 F.Supp. at 421, works licensed independently of ASCAP and performances that would not give rise to copyright liability shall be excluded from the fee calculation.

Applicants shall have 120 days from the entry of this Order to notify ASCAP in writing whether they opt for a retroactive blanket or per-program license with fees as now set by the court. Applicants that operated under an interim blanket license and desire to select the per-program license retroactively shall be required to document their feature music use.

SO ORDERED.

NOTES

[1] Throughout the preliminary stages of this rate proceeding, the parties have disputed the exact number of applicant stations. However, without waiving its right to object to the standing of individual applicants, ASCAP has agreed that applicants' letter of August 14, 1996, (Exh. 18), listing the applicant stations, is accurate. We note that several of the stations appearing on that list were applicants for less than the entire application period and that, as a consequence, the precise number of applicants has varied over time.

[2] The blanket license has traditionally been the principal licensing mechanism for frequent users of music. Under a blanket license, the licensee pays a specified fee each year, and in exchange receives the right to make unlimited use, during the licensing period, of all of the music in ASCAP's repertory. Under the Group W license, the blanket license fee rate rises annually from 1.575% of adjusted gross revenue in 1991 to 1.615% of adjusted gross revenue in 1995. (Exh. 7, Local Station Blanket Radio License ¶ 5A.) Under the WGN license, the blanket license fee rate is 1.725% of adjusted gross revenue. (Exh. 6, Local Station Blanket Radio License ¶ 5.)

[3] The Group W license sets the feature music fee at 4.22%. (Exh. 7, Local Station Per Program Radio License ¶ 4.)

[4] "Revenue subject to fee" is the percentage of a station's adjusted gross revenue attributable to programs that contain feature performances of ASCAP music. Because of the difficulty of attributing revenue to specific programs, the per-program license contains a formula that approximates attributable revenue by using an average of the number of program hours that contain feature performances of ASCAP music, weighted to give greater effect to performances during times when the listening audience is larger.

[5] Under the WGN per-program license, the incidental use fee is 48 times the licensee's highest one-minute card rate. (Exh. 6 ¶ 4.) The Group W incidental use fee is 0.24% of adjusted gross revenue. (Exh. 7 ¶ 4.)

[6] Rather than creating fee equivalence (uncorrected for administrative costs) between the per-program and blanket licenses for a station that plays ASCAP feature music in approximately 33% of its weighted hours, as applicants contend is the case under the Group W license, applicants' proposal would establish 90% as the crossover point. Assuming that a "typical" local radio station, one whose music use reflects the industry median, reports music use in approximately 90% of its weighted program hours, revenue subject to fee would be 0.90 × AGR. At applicant's proposed per-program license fee rate, the total per-program fee would be (0.0173) (0.90)(AGR) + (0.0006) (AGR) = 0.01617(AGR), or 1.617% of adjusted gross revenue. This is comparable to ASCAP's blanket license fee rate of 1.615% of adjusted gross revenue.

[7] Both ASCAP and applicants relied on the 1990, 1991, and 1995 ASCAP survey data as the basis for their analyses of applicants' music use characteristics. The parties do not dispute that approximately one-quarter of all radio stations are sampled for their music usage in each annual survey, which ASCAP conducts for the purpose of distributing ASCAP royalties to its members. (Tr. 205-213.) Moreover, the parties agree that the stations surveyed provide a representative sample of the industry's and of the relevant groups' music use. (Tr. 267.)

[8] We note applicants' contention that ASCAP's efforts to compare the applicant group with the industry group based on their respective ranges of music use finds no support in the literature of statistical analysis. However, applicants mischaracterize ASCAP's methodology, which is not premised on a comparison of the relevant groups' characteristics but rather on the theory that individual applicant stations and individual Group W or WGN licensees are similarly situated with respect to music use.

[9] Both Dr. Boyle and ASCAP contend that the applicant stations' music use was comparable not only to that of the Group W and WGN licensees but to the music use of stations that the RMLC represented. (Exh. 668 ¶ 5 ("for each range of feature plays per hour by the ... Applicants, there was a similar range among the Industry and the RMLC stations"); ASCAP Post-Trial Mem. at 18 ("radio stations represented by the NRBMLC ... are comparable in amount of music used ... to those stations which were RMLC members....").) However, ASCAP's Proposed Findings of Fact omit this assertion and the tables summarizing Dr. Boyle's analysis do not address music use by the RMLC stations, although Dr. Boyle's full report does contain two tables that appear to support the claim that some RMLC stations used ASCAP music at levels similar to that of the applicants. (Exh. 668, tbls. III(C), VI(C).)

[10] Sections VIII and IX(C) of the Consent Decree are relevant only to the extent that they expressly refer to individual licensees and users rather than groups of stations and therefore offer some support for ASCAP's claim that its individual-station-based analysis of the survey data provides the proper methodology for determining whether applicants are situated similarly to other stations in the industry.

[11] We find it noteworthy that prior to the Group W negotiations, the RMLC had referred to itself as the All Industry Music License Committee. Although certainly not dispositive, the fact that the committee with whom ASCAP sought to negotiate music licenses felt compelled to drop "All Industry" from its title certainly does little to bolster ASCAP's claim that the RMLC "represent[ed] the collective radio broadcasting industry." (ASCAP Mem. at 26.) Contrary to ASCAP's assertion, the fact that approximately seven thousand broadcasters, which the RMLC expressly did not represent, consented to be bound by the RMLC-brokered agreement cannot be said to have transformed the RMLC into an industry-wide body.

Additionally, in its October 1, 1990 correspondence to the many radio stations not formally represented by the RMLC, ASCAP advised: "You can make your own application to ASCAP and, later, to the Court. You can do this alone or jointly with other stations." (Exh. 12, Questions & Answers.) We are somewhat troubled that ASCAP now argues, in effect, that the instant application to the court should be summarily rejected because it threatens to unsettle the RMLC negotiations that ASCAP previously advised unrepresented stations they could forgo, if they wished to proceed independently.

Last, we consider ASCAP's concerns for the future of collective negotiations to be somewhat overstated. Given the significant expense and uncertainty of success in rate-court litigation, there is scant likelihood that a vast number of stations, which have heretofore agreed to accept the RMLC-negotiated licenses, would elect to negotiate with ASCAP independently in the future. As ASCAP itself observes: "No single station or small group of stations could easily afford such negotiations, or if negotiations fail, to seek judicial relief." (ASCAP Mem. at 8.)

[12] In support of the claim that the RMLC was committed to the per-program license, ASCAP argues that the RMLC negotiators who attended the Berkshire, WGN, and Group W meetings included representatives of stations that held or eventually switched to that form of license. For example, Mr. Harris, whose involvement with the RMLC began in approximately 1986 (Exh. 906, at 12), also served as an executive of Westinghouse Broadcasting, the owner of several radio stations (Id. at 6-7). ASCAP points out that during the Group W license period, the number of Westinghouse stations that operated under per-program licenses increased from five to eight and the number of blanket licenses decreased from eleven to six. In addition, after the RMLC and ASCAP reached agreement on the WGN licenses, the lead applicant in that proceeding, WGN, accepted a per-program license.

However, even though some of the RMLC negotiators apparently had a direct economic interest in the per-program license, the fact remains that the vast majority of RMLC-represented stations and of the industry as a whole, which ASCAP contends the RMLC faithfully represented, were blanket license holders at the time of the Group W negotiations. Moreover, the fact that WGN and several other stations have adopted a clever strategy of bunching their broadcasts of ASCAP music in selected program hours to obtain maximum benefit from their per-program licenses, (Tr. 42-43), suggests that a per-program fee rate that is favorable for these stations might not be attractive to other stations with less programming flexibility. Thus there remains at least some doubt whether the RMLC adequately represented the interests of limited music users and obtained per-program license terms that were reasonable for those stations.

[13] We note ASCAP's claim that since 1977, the RMLC has actively sought changes in the per-program license. (ASCAP Mem. at 30.) However, ASCAP does not identify any meaningful modifications that the RMLC actually obtained following the Berkshire agreement, with the exception of ASCAP's consent to eliminate provisions allowing either the RMLC or ASCAP to terminate the per-program license. (Tr. 38-44.) As ASCAP notes, the fee relationship between the blanket and per-program licenses has remained fundamentally unchanged at approximately 3:1. (ASCAP's Proposed Finding of Fact No. 44.) Moreover, ASCAP has offered no persuasive evidence that Magistrate Judge Dolinger erred when he found that in the early radio license negotiations, radio station negotiators accepted higher per-program fees because they were not interested in that form of license. Buffalo Broadcasting I, 1993 WL 60687, at *61.

[14] Applicants have presented extensive evidence in the form of music-use data and statements by RMLC Chairman Harris to support their claim that the RMLC-negotiated per-program licenses are not adequate benchmarks of reasonable fees. However, applicants' proffers pertain almost exclusively to the Group W negotiation period. Applicants have offered very little specific evidence that the RMLC was predominantly concerned with the blanket license or willing to accept less favorable per-program terms in order to secure a better blanket license agreement during the WGN negotiations. Moreover, applicants have made no showing as to the levels of music use by applicants, the RMLC-represented stations, or the industry at the time of the WGN negotiations. Nevertheless, having determined that the Group W per-program license negotiated by the RMLC is not an appropriate benchmark of reasonable rates, we decline to rely on its predecessor, the WGN per-program license, which contains a substantially similar fee structure and was also negotiated by the RMLC.

[15] It is of course true that this court, pursuant to its supervisory authority under the Consent Judgment, issued an Order approving the Group W license fees as reasonable and nondiscriminatory. In that Order, however, we cabined our approval of the per-program license by stating that the terms of that license complied with the Consent Decree as applied only to the Group W applicants. (Final Order dated October 8, 1991 ¶ 2.) Thus, the court's acceptance of the Group W licenses cannot be invoked to support the claim that the terms of the Group W per-program license provide a benchmark of reasonableness in this proceeding. Similarly, although the Final Order approving the WGN licenses did not contain the same explicit limiting language as did the Group W Final Order, we do not consider that Order to be probative of the reasonableness of the terms of the WGN license for the applicant group.

[16] In addition, applicants protest that the proposed per-program fee's objective, or absolute, reasonableness is irrelevant to this proceeding. However, if ASCAP could somehow demonstrate that the price of the challenged per-program license is commensurate with the "value of the rights" offered by ASCAP, then, pursuant to the Second Circuit's position in Buffalo Broadcasting Co. v. ASCAP, 744 F.2d 917, 927 (2d Cir.1984), the per-program license would constitute a "realistic alternative" to the blanket license.

[17] Applicants note that of those stations that comprised the 8% that operated under a per-program license, more than half reported no feature music use at all. Thus, less than 4% of stations actually making any feature performances of music during the survey year found the license usable. (Tr. 138, 188.)

[18] We reject applicants' contention that, because ASCAP bears the ultimate burden of proving the reasonableness of its fee proposals, ASCAP must offer evidence to refute every unsubstantiated factual claim made by applicants.

[19] At trial, applicants also presented the equally general testimony of Edward Atsinger, president of Salem Communication Corporation, who stated that "if the station originated music in as little as 20 percent of its [weighted] hours, subject to fee, it becomes a very problematic problem as to whether there is any economic benefit. In terms of pure economics, the cross-over point is 32 percent." (Tr. 468.)

[20] In addition to our dissatisfaction with the concept of a "typical" station in the radio industry, addressed above, we are skeptical that Buffalo Broadcasting's equivalence model would provide an appropriate basis upon which to set a per-program fee rate for a music-intensive group of applicants such as the radio industry as a whole. Magistrate Judge Dolinger estimated that the "typical" television station used ASCAP music in approximately 75% of its non-network programs, Buffalo Broadcasting I, 1993 WL 60687, at *67, whereas applicants have offered evidence that the median industry radio station played music in 95% of its weighted programs hours. Although the music use data relied upon by Magistrate Judge Dolinger differed from that presented in this proceeding, it appears that the radio industry during the relevant period was, on average, significantly more music intensive than was the television industry in the period reviewed by Magistrate Judge Dolinger. Thus although, as applicants observe, the Consent Decree applies to the licensing of music in both industries, it does not necessarily follow that the equivalence model developed for the television industry is appropriate for the radio industry. In our view, even if the median station were deemed "typical," it would be unreasonable to impose a fee structure under which a radio station using music in 90% or more of its hours would have an economically neutral choice between a per-program license and a blanket license.

[21] Because ASCAP takes the position that the court may not set an incidental use fee that is a flat percentage of a station's adjusted gross revenue, ASCAP has not proposed any such fee in this proceeding.

[22] Applicants assert that since 1992, ASCAP has heavily subsidized its distribution of member royalties related to incidental performances with collections from per-program licensees. Thus, applicants contend that ASCAP's post-1991 feature distributions, which constitute 96.5% of ASCAP's total radio distributions, are the best indicator of ASCAP's actual valuation of incidental performances relative to feature performances for that period. (Exh. 704, at 6-7.) According to applicants, ASCAP's distributions prior to 1992 reflect a valuation for radio incidental performances of approximately 2%. (Exh. 704, at 6-8.)

[23] ASCAP also observes that for Survey year 1990, 166 blanket license stations captured in the survey averaged fewer than two feature performances of ASCAP music per hour. For these stations, incidental performances received 14.9% of the total credits. (Exh. 678, at 5.) In ASCAP's view, "[i]t is interesting to note that the 14.9% share of credits attributable to incidental uses on stations with fewer than two ASCAP features per hour happens to correspond to the ratio of the base fee under the per program license to blanket license fee that Mr. Massarsky cites [in] his report." (Id.)

[24] In light of our determination that the RMLC-negotiated licenses cannot be adopted as a benchmark of reasonableness in this proceeding, the fact that Mr. Harris accepted a 0.24% AGR fee during the Group W negotiations does not establish that rate as reasonable.

[25] Mr. Massarsky acknowledged at trial that the feature multiplier added approximately 30% to the credits that ASCAP assigned to feature uses of music. (Tr. 634.) According to ASCAP, as a result, "the share for incidental performances that Mr. Massarsky calculated for the 1992-95 Survey Years is only 20-31% of the correct figure." (Exh. 678, at 9.) It appears to us, however, that had Mr. Massarsky excluded the credits attributable to the feature multiplier, this would have increased the incidental credit share of total credits by approximately 30%. Even assuming a 30% increase from 2% to 2.6% in the percentage of credits ASCAP assigned to incidental performances in ASCAP's blanket distributions, applicants' proposal of 0.06% AGR is still reasonable because 2.6% of the blanket fee of approximately 1.6% AGR, is 0.046%.

[26] Applicants also contend that it is unreasonable for ASCAP to be permitted to charge programs as "subject to fee" without identifying to the licensee the composition that triggered the charge. (Apps. Mem. at 64 n. 30.) Applicants further maintain that the license ordered by this court should not require a final count of weighted hours subject to fee from the stations. (Id.) "In practice," according to applicants, only ASCAP is capable of computing that figure. (Id.) Thus, in applicants' view, ASCAP should be required to provide stations with data on their monthly weighted hours subject to fee as a condition precedent to the licensees' obligation to make a final report.

Applicants raise these claims in a brief footnote that is unsupported by any citations to the record. Moreover, ASCAP has offered no response whatsoever to applicants' claims. If applicants wish to pursue these matters, applicants are directed to submit, within fourteen days, a thorough and adequately supported statement of their position, and ASCAP is directed to respond within fourteen days thereafter.

[27] ASCAP does not dispute that non-applicant blanket stations featured ASCAP music in an average of 87.7% of their weighted hours during the years for which data is available. (Exh. 701A, tbl. 6.) Nor does ASCAP appear to take issue with applicants' assertion that the interim NECM blanket licensees featured ASCAP music in an average of 48.6% of their weighted hours, which is 55.4% of the usage the industry blanket licensees. Id.