Digital Performance Right in Sound Recordings and Ephemeral Recordings, 13026-13058 [2011-4995]

Agencies

• Copyright Royalty Board, Library of Congress
• Library of Congress
• Copyright Royalty Board

[Federal Register Volume 76, Number 46 (Wednesday, March 9, 2011)]
[Rules and Regulations]
[Pages 13026-13058]
From the Federal Register Online via the Government Printing Office [www.gpo.gov]
[FR Doc No: 2011-4995]



[[Page 13025]]

Vol. 76

Wednesday,

No. 46

March 9, 2011

Part II





Library of Congress





-----------------------------------------------------------------------



Copyright Royalty Board



-----------------------------------------------------------------------



37 CFR Part 380



 Digital Performance Right in Sound Recordings and Ephemeral 
Recordings; Final Rule

Federal Register / Vol. 76 , No. 46 / Wednesday, March 9, 2011 / 
Rules and Regulations

[[Page 13026]]


-----------------------------------------------------------------------

LIBRARY OF CONGRESS

Copyright Royalty Board

37 CFR Part 380

[Docket No. 2009-1 CRB Webcasting III]


Digital Performance Right in Sound Recordings and Ephemeral 
Recordings

AGENCY: Copyright Royalty Board, Library of Congress.

ACTION: Final rule and order.

-----------------------------------------------------------------------

SUMMARY: The Copyright Royalty Judges are announcing their final 
determination of the rates and terms for two statutory licenses, 
permitting certain digital performances of sound recordings and the 
making of ephemeral recordings, for the period beginning January 1, 
2011, and ending on December 31, 2015.

DATES: Effective Date: March 9, 2011.
    Applicability Dates: These rates and terms are applicable to the 
period January 1, 2011, through December 31, 2015.

FOR FURTHER INFORMATION CONTACT: Richard Strasser, Senior Attorney, or 
Gina Giuffreda, Attorney Advisor. Telephone: (202) 707-7658. E-mail: 
crb@loc.gov.

SUPPLEMENTARY INFORMATION:

I. Introduction

A. Subject of the Proceeding

    This is a rate determination proceeding convened under 17 U.S.C. 
803(b) et seq. and 37 CFR part 351 et seq., in accord with the 
Copyright Royalty Judges' Notice announcing commencement of proceeding, 
with a request for Petitions to Participate in a proceeding to 
determine the rates and terms for the digital public performance of 
sound recordings by means of an eligible nonsubscription transmission 
or a transmission made by a new subscription service under section 114 
of the Copyright Act, as amended by the Digital Millennium Copyright 
Act (``DMCA''), and for the making of ephemeral copies in furtherance 
of these digital public performances under section 112, as created by 
the DMCA, published at 74 FR 318 (January 5, 2009). The rates and terms 
set in this proceeding apply to the period of January 1, 2011 through 
December 31, 2015. 17 U.S.C. 804(b)(3)(A).

B. Statutory Background

    A lengthy review of the history of the sound recordings compulsory 
license is contained in the Final Determination for Rates and Terms in 
Docket No. 2005-1 CRB DTRA, 72 FR 24084 (May 1, 2007) (``Webcaster 
II'').\1\ This history was summarized by the United States Court of 
Appeals for the District of Columbia Circuit in Intercollegiate 
Broadcast System, Inc. v. Copyright Royalty Board, 574 F.3d 748, 753-54 
(DC Cir. 2009), as follows:
---------------------------------------------------------------------------

    \1\ The two prior webcasting proceedings often have been 
referred to informally as ``Webcaster I'' and ``Webcaster II,'' 
respectively, as opposed to the formal caption ``DTRA'' (which 
stands for ``Digital Transmissions Rate Adjustment''). In the 
current proceeding, we use the caption ``Webcasting III'' and intend 
to caption future webcasting proceedings using the term 
``Webcasting'' followed by the appropriate Roman numeral.

    [Since the nineteenth century, the Copyright Act protected the 
performance right of ``musical works'' (the notes and lyrics of a 
song), but not the ``sound recording.'' Writers were protected but 
not performers.]
    In 1995, Congress passed the Digital Performance Right in Sound 
Recordings Act. Pub. L. No. 104-39, granting the owners of sound 
recordings an exclusive right in performance ``by means of a digital 
transmission.'' 17 U.S.C. Sec.  106(6); see Beethoven.com LLC v. 
Librarian of Cong., 394 F.3d 939, 942 (D.C. Cir. 2005). The Digital 
Millennium Copyright Act of 1998, Pub. L. No. 105-304, ``created a 
statutory license in performances by webcast,'' to serve Internet 
broadcasters and to provide a means of paying copyright owners. 
Beethoven.com, 394 F.3d at 942; see 17 U.S.C. Sec.  114(d)(2), 
(f)(2). To govern the broadcast of sound recordings, Congress also 
created a licensing scheme for so-called ``ephemeral'' recordings, 
``the temporary copies necessary to facilitate the transmission of 
sound recordings during internet broadcasting.'' Beethoven.com, 394 
F.3d at 942-43; see 17 U.S.C. Sec.  112(e)(4).
    Congress has delegated authority to set rates for these rights 
and licenses under several statutory schemes. The most recent, 
passed in 2005 [sic], directed the Librarian of Congress to appoint 
three Copyright Royalty Judges who serve staggered, six-year terms. 
See 17 U.S.C. Sec.  801, et seq. These Judges conduct complex, 
adversarial proceedings, described in 17 U.S.C. Sec.  803 and 37 CFR 
Sec.  351, et seq., and ultimately set ``reasonable rates and 
terms'' for royalty payments from digital performances. 17 U.S.C. 
Sec.  114(f). * * * Rates should ``most clearly represent the rates 
and terms that would have been negotiated in the marketplace between 
a willing buyer and a willing seller.'' Id. [17 U.S.C. Sec.  
114(f)(2)(B)] ``In determining such rates and terms,'' the Judges 
must ``base [their] decision on economic, competitive and 
programming information presented by the parties.'' Id. 
Specifically, they must consider whether ``the service may 
substitute for or may promote the sales of phonorecords'' or 
otherwise affect the ``copyright owner's other streams of revenue.'' 
Id. Sec.  114(f)(2)(B)(i). The Judges must also consider ``the 
relative roles of the copyright owner and the transmitting entity'' 
with respect to ``relative creative contribution, technological 
contribution, capital investment, cost, and risk.'' Id. Sec.  114 
(f)(2)(B)(ii). Finally, ``[i]n establishing such rates and terms,'' 
the Judges ``may consider the rates and terms for comparable types 
of digital audio transmission services and comparable circumstances 
under voluntary license agreements described in subparagraph (A).'' 
Id. Sec.  114(f)(2)(B).

Intercollegiate Broadcast System, Inc. v. Copyright Royalty Board, 574 
F.3d 748, 753-54 (DC Cir. 2009).
    Forty petitions to participate were filed in response to the 
January 5, 2009, notice of commencement of the proceeding. The great 
majority of the petitioners were webcasters. During the subsequent 
period of voluntary negotiations, settlements were reached among many 
of the parties. In addition to the negotiation phase required in this 
proceeding, 17 U.S.C. 803(b)(3), Congress enacted the Webcaster 
Settlement Acts of 2008 and 2009, which expanded the opportunities to 
resolve the issues in this proceeding, as well as the issues in 
Webcaster II. This legislation further impacted Webcasting III by 
permitting the settling parties to determine if the settlements could 
be considered as evidence before the Copyright Royalty Judges 
(``Judges'').\2\ Eight settlements were resolved under the Webcaster 
Settlement Acts. 74 FR 9293 (March 3, 2009) (three agreements); 74 FR 
34796 (July 17, 2009) (one agreement); 74 FR 40614 (August 12, 2009) 
(four agreements). The rates and terms under these settlements were the 
basis of approximately 95 percent of webcasting royalties paid to 
SoundExchange in 2008 and 2009. SX PFF at ]] 50, 51.\3\ Evidence was 
presented in this proceeding by SoundExchange, Inc. (``SX''), 
representing the owners, and three webcasters, College Broadcasters, 
Inc. (``CBI''), Live365, Inc. (``Live365''), and Intercollegiate 
Broadcasting System,

[[Page 13027]]

Inc. (``IBS'').\4\ CBI only presented evidence to support adoption of 
its settlement with SoundExchange for noncommercial educational 
webcasters. SoundExchange and Live365 presented evidence related to 
commercial webcasters. The webcasting royalties paid by Live365 to 
SoundExchange for 2008 and 2009 were less than 3 percent of total 
webcasting royalties paid to SoundExchange. SX PFF at ] 53. 
SoundExchange presented evidence related to noncommercial webcasters, 
and IBS presented evidence for small noncommercial webcasters. Written 
statements, discovery and testimony for both direct case and rebuttal 
case were filed on these issues.
---------------------------------------------------------------------------

    \2\ In the pleadings filed and during the testimony, Live365 
attempted to introduce evidence about agreements that contained 
provisions that they were not to be considered as precedential under 
the Webcaster Settlement Acts. Following the clear language of the 
statute that these agreements were not ``admissible as evidence or 
otherwise taken into account,'' 17 U.S.C. 114(f)(5)(C), these 
attempts were rejected. See, e.g., 4/19/10 Tr. at 210:9-10 
(sustaining objection to Live365's motion to enter into evidence the 
``Pure Play Agreement'').
    \3\ References to the proposed findings of fact and conclusions 
of law shall be cited as ``PFF'' or ``PCL,'' respectively, and reply 
findings and conclusions of law shall be cited as ``RFF'' or 
``RCL,'' respectively, preceded by the name of the party that 
submitted same and followed by the paragraph number. Similarly, 
references to the written direct testimony shall be cited as ``WDT'' 
preceded by the last name of the witness and followed by the page 
number. Likewise, references to the written rebuttal testimony shall 
be cited as ``WRT'' preceded by the last name of the witness 
followed by the page number. References to the transcript shall be 
cited as ``Tr.'' preceded by the date and followed by the page 
number and the name of the witness.
    \4\ After filing Written Direct Statements, RealNetworks, Inc. 
withdrew from the proceedings, and Royalty Logic, LLC, did not 
participate further.
---------------------------------------------------------------------------

    On December 14, 2010, the Judges issued their Initial Determination 
of Rates and Terms. Pursuant to 17 U.S.C. 803(c)(2)(B) and 37 CFR 
353.4, motions for rehearing were due to be filed no later than 
December 29, 2010. No motions were received.

II. Commercial Webcasters

A. Commercial Webcasters Encompassed by the National Association of 
Broadcasters-SoundExchange Agreement

    On June 1, 2009, the National Association of Broadcasters (``NAB'') 
and SoundExchange filed a settlement of all issues between them in the 
proceeding, including the proposed rates and terms. This was one of the 
Webcaster Settlement Act agreements, published by the Copyright Office 
in the Federal Register, and was filed in this proceeding, pursuant to 
17 U.S.C. 801(b)(7)(A), to be adopted as rates and terms for some 
services of commercial broadcasters for the period 2011 through 2015. 
It applies to statutory webcasting activities of commercial terrestrial 
broadcasters, including digital simulcasts of analog broadcasts and 
separate digital programming. The settlement includes per performance 
royalty rates, a minimum fee and reporting requirements that are more 
comprehensive than those in the current regulations. Section 
801(b)(7)(A) allows for the adoption of rates and terms negotiated by 
``some or all of the participants in a proceeding at any time during 
the proceeding'' provided they are submitted to the Copyright Royalty 
Judges for approval. This section provides that in such event:

    (i) The Copyright Royalty Judges shall provide to those that 
would be bound by the terms, rates, or other determination set by 
any agreement in a proceeding to determine royalty rates an 
opportunity to comment on the agreement and shall provide to 
participants in the proceeding under section 803(b)(2) that would be 
bound by the terms, rates, or other determination set by the 
agreement an opportunity to comment on the agreement and object to 
its adoption as a basis for statutory terms and rates; and
    (ii) The Copyright Royalty Judges may decline to adopt the 
agreement as a basis for statutory terms and rates for participants 
that are not parties to the agreement, if any participant described 
in clause (i) objects to the agreement and the Copyright Royalty 
Judges conclude, based on the record before them if one exists, that 
the agreement does not provide a reasonable basis for setting 
statutory terms or rates.

17 U.S.C. 801(b)(7)(A).

    The Judges published the settlement (with minor modifications) in 
the Federal Register on April 1, 2010, and provided an opportunity to 
comment and object by April 22, 2010. 75 FR 16377 (April 1, 2010). No 
comments or objections were submitted, so the provisions of 17 U.S.C. 
801(b)(7)(A)(ii) do not apply. Absent objection from a party that would 
be bound by the proposed rates and terms and that would be willing to 
participate in further proceedings, the Copyright Royalty Judges adopt 
the rates and terms in the settlement for certain digital transmissions 
of commercial broadcasters for the period of 2011-2015. 17 U.S.C. 
801(b)(7)(A). Cf. Review of the Copyright Royalty Judges Determination, 
Docket No. 2009-1, 74 FR 4537, 4540 (January 26, 2009) (review of 
settlement adoption).

B. All Other Commercial Webcasters

1. Stipulation Concerning the Section 112 Minimum Fee and Royalty Rate 
and Stipulation Concerning the Section 114 Minimum Fee
    In between the direct and rebuttal phases, SoundExchange and 
Live365 presented two settlements of issues for all remaining 
commercial webcasters not encompassed by the NAB-SoundExchange 
agreement: (1) The minimum fee and royalty rates for the section 112 
license and (2) the minimum fee for the section 114 license. These two 
settlements were included in one stipulation. The terms of the 
settlement are the same as the agreement reached and included as a 
final rule in Webcaster II, following remand. See Digital Performance 
Right in Sound Recordings and Ephemeral Recordings (Final rule), 75 FR 
6097 (February 8, 2010). The minimum fee for commercial webcasters is 
an annual, nonrefundable fee of $500 for each individual channel and 
each individual station (including any side channel), subject to an 
annual cap of $50,000. The royalty rate for the section 112 license is 
bundled with the fee for the section 114 license. There is one 
additional term in the stipulation that was not included in Webcaster 
II. The royalty rate for the section 112 license is attributed to be 5% 
of the bundled royalties. There was no objection to the stipulation. 
There was evidence presented to support the minimum fee for commercial 
webcasters and the bundled royalty rates. SX PFF at ]] 459-468, 472. No 
evidence disputed it. These provisions are supported by the parties and 
the evidence. The Judges accept and adopt these two stipulations as 
settling these issues.
2. Rate Proposals for the Section 114 License for Commercial Webcasters
    The contending parties propose vastly different rate amounts for 
the use of the section 114 license for commercial webcasters. In its 
second revised rate proposal, SoundExchange argues in favor of a 
performance rate beginning at $.0021 per performance in 2011 and 
increasing annually by .0002 to a level of $.0029 by 2015. SX PFF at ] 
118.
    Live365 also proposes a per performance fee structure. By contrast, 
under the Live365 proposal, commercial webcasters would pay $.0009 per 
performance throughout the period 2011-2015. Rate Proposal For Live365, 
Inc., Appendix A, Proposed Regulations at Sec.  380.3(a)(1).\5\
---------------------------------------------------------------------------

    \5\ In addition, Live365 seeks a 20% discount applicable to this 
commercial webcasting per performance rate for certain ``qualified 
webcast aggregation services.'' This proposal is discussed infra at 
Section II.B.5.
---------------------------------------------------------------------------

    Notwithstanding the gulf between the SoundExchange and Live365 
proposed royalty amounts, there is no difference between the parties 
with respect to the basic structure of their proposed compensation 
schemes. Both SoundExchange and Live365 propose that per performance 
rates (typically stated as a fraction of a penny) be applicable in the 
case of the section 114 license. Furthermore, the per performance usage 
structure was adopted in Webcaster II. Webcaster II, 72 FR 24090 (May 
1, 2007). It remains the best structure for the reasons stated therein. 
Id. at 24089-90. Therefore, the only issues we are left to decide are 
the applicable amount of the webcaster royalty rate and whether any 
discount to that rate should be made on those occasions when certain 
types of webcasters are aggregated.
    The starting point for our determination is the applicable amount 
of the section 114 performance rate.

[[Page 13028]]

3. The Parties' Disparate Approaches To Rate Setting for the Section 
114 License for Commercial Webcasters
    Both Live365 and SoundExchange agree that the willing buyer/willing 
seller standard should be applied by the Copyright Royalty Judges in 
determining the rates for the section 114 license. Both recognize that 
those rates should reflect the rates that would prevail in a 
hypothetical marketplace that was not constrained by a compulsory 
license.
    However, in contrast to the positions of the copyright owners and 
commercial services in Webcaster II, in the instant case SoundExchange 
and Live365 do not agree that the best approach to determining rates is 
to look to comparable marketplace agreements as ``benchmarks'' 
indicative of the prices to which willing buyers and willing sellers 
would agree in the hypothetical marketplace. On the one hand, Live365 
primarily seeks to support its rate proposal by means of a modeling 
analysis that aims to determine the amount of any residue that may 
remain for compensating the sound recording input a commercial 
webcaster uses, after reducing webcaster revenues by an amount equal to 
the cost of all other inputs utilized by the webcaster in providing its 
service and also by an assumed amount of webcaster profits. By 
contrast, SoundExchange puts forward a benchmark approach in support of 
its rate proposal, similar to the primary argument it made in Webcaster 
II and an approach adopted by the Judges therein.
a. The Live365 Approach
    Live365 relies primarily on a modeling analysis provided by Dr. 
Mark Fratrik that seeks to identify the rate that commercial webcasters 
``would have been willing to pay in a negotiated settlement between a 
willing buyer and a willing seller.'' Fratrik Corrected and Amended WDT 
at 5. We find that Dr. Fratrik presumes behavioral constraints not 
found in the statutory standard and, that even if we were to ignore the 
distortions created by such added constraints, his analysis suffers 
from so many other unwarranted explicit assumptions and data defects as 
to make his analysis untenable.
i. Dr. Fratrik's Model and the Hypothetical Market
    The terms ``willing buyer'' and ``willing seller'' in the statutory 
standard simply refer to buyers and sellers who are unconstrained in 
their marketplace dealings. In other words, the buyers and sellers 
operate in a free market unconstrained by government regulation or 
interference. (See, for example, Noncommercial Educational Broadcasting 
Compulsory License (Final rule and order), 63 FR 49823, 49834 
(September 18, 1998). (``[I]t is difficult to understand how a license 
negotiated under the constraints of a compulsory license, where the 
licensor has no choice to license, could truly reflect `fair market 
value.' ''). Moreover, neither the buyers nor the sellers exercise such 
monopoly power as to establish them as price-makers and, thus, make 
negotiations between the parties superfluous. Webcaster II, 72 FR 24091 
(May 1, 2007). (``In other words, neither sellers nor buyers can be 
said to be `willing' partners to an agreement if they are coerced to 
agree to a price through the exercise of overwhelming market power.'')
    Dr. Fratrik and Live365 either misperceive the plain meaning of the 
terms of the statute or deliberately seek to expand the meaning of a 
``willing buyer'' as articulated in the willing buyer-willing seller 
standard that governs this proceeding. For them, a ``willing buyer'' is 
viewed through the lens of an additional policy consideration nowhere 
articulated in the statute--i.e., that a buyer can only be considered 
``willing'' if that buyer is able to obtain the sound recording input 
at a price that allows the buyer to earn at least a 20 percent 
operating profit margin from the use of that input. Thus, in Dr. 
Fratrik's analysis, a ``representative'' single buyer is deemed to be 
constrained in its behavior from participating in the input market for 
sound recordings unless its operating profit margin expectations in the 
output market for webcasting services are guaranteed at a level 
consistent with an industry-wide average profit margin for a 
purportedly comparable industry such as terrestrial radio. Fratrik 
Corrected and Amended WDT at 21-22.
    Nothing in the statute supports reading such a behavioral 
constraint into the hypothetical marketplace to be derived by the 
Judges in this proceeding. Indeed, a similar argument that economic 
viability based on the sufficiency of revenue streams to cover costs 
determines any individual buyer's ``willingness'' to pay for an input 
raised by Live365 in Webcaster I, was rejected in that proceeding. 
Determination of Reasonable Rates and Terms for the Digital Performance 
of Sound Recordings and Ephemeral Recordings (Final rule and order) 
(``Webcaster I''), 67 FR 45240, 45254 (July 8, 2002) (``Thus, the Panel 
had no obligation to consider the financial health of any particular 
service when it proposed the rates.'').
    Dr. Fratrik's notion of a representative entity adds an operating 
condition that distinguishes his conceptual formulation from that of a 
statistically average firm in an industry. His representative firm must 
reach one specified minimum profit margin and, therefore, can only be 
satisfied with a royalty rate sufficient to allow it to reach that 
profit margin. Any lower assumed profit margin would, ceterus paribus, 
necessarily result in a lower recommended royalty rate. Thus, Dr. 
Fratrik effectively assumes that his representative firm will never 
have a reason to operate at less than a particular operating profit 
margin (i.e., 20%).
    But there is no a priori reason to believe that a representative 
webcaster would not accept a lesser profit margin, so long as it earns 
a profit and/or finds no risk-adjusted rate of return that could be 
earned by an alternative investment. Indeed, basic microeconomic 
analysis recognizes that, in the short-run, it is in the interest of a 
firm to continue to produce even at an operating loss, so long as its 
variable costs are covered and some contribution can be made toward 
fixed costs--otherwise, the loss incurred by the firm will be even 
greater (i.e., full fixed costs if no production takes place).\6\ In 
short, Dr. Fratrik's assumption of a 20% profit margin totally ignores 
the possibility of webcasters with a whole range of potential 
acceptable operating profit margins--whether lesser or greater--that 
would be dependent on such things as varying capital investment costs 
among webcasters, changing market conditions in output markets, and the 
applicable time horizon.\7\
---------------------------------------------------------------------------

    \6\ See, for example, Varian, Hal, Intermediate Microeconomics: 
A Modern Approach, (W.W. Norton & Company, 2009) at 350, 401. 
Mansfield, Edwin and Yohe, Gary Wynn, Microeconomics: Theory and 
Applications, (W.W. Norton & Company, 2004) at 296, 407; see also 7/
28/10 Tr. at 54:2-14 (Salinger).
    \7\ In the long-run, all short-run fixed costs become variable.
---------------------------------------------------------------------------

    Still another difficulty with Dr. Fratrik's conceptual framework is 
that his single ``representative'' buyer is treated as tantamount to an 
industry. But no single firm is typically the equivalent of an industry 
on the demand side of the market, although there is the obvious 
exception where a single monopsonistic buyer constitutes the entire 
demand side of the market for a particular input. While Dr. Fratrik 
does not make the claim that his representative commercial webcaster is 
a monopsonist, his analysis effectively produces that result.

[[Page 13029]]

    For example, Dr. Fratrik explains that he chose to wed a 20% 
operating profit margin assumption to his cost and revenue estimates to 
``derive a resulting value for the copyrighted work.'' Fratrik 
Corrected and Amended WDT at 15, 23. In other words, Dr. Fratrik and 
Live365 effectively claim that no buyer would ever be a ``willing 
buyer'' unless the price of only the one input here analyzed (i.e., the 
royalty rate for sound recordings) is low enough to provide all buyers 
with sufficient revenue after the royalty payment to cover all other 
input costs and yield an operating profit margin of 20%. It is a claim 
that, rather than resulting from any careful analysis of the market 
demand and supply schedules, blithely ignores such analysis in favor of 
a single price point wholly determined by a single actor on the demand 
side of the market without any reference to the supply side of the 
market.\8\
---------------------------------------------------------------------------

    \8\ Dr. Fratrik implies that because the record companies 
supplying the sound recordings will incur something near zero 
incremental costs, the supply side of the market may be largely 
ignored. 4/27/10 Tr. at 1131:12-1133:19 (Fratrik). But Dr. Fratrik 
offers no empirical support for his assertion as to actual 
incremental costs. We have clearly rejected a similar contention put 
forward in Webcaster II on both empirical and theoretical grounds. 
Webcaster II, 72 FR 24094 (May 1, 2007).
---------------------------------------------------------------------------

    In other words, Dr. Fratrik's single ``representative'' buyer's 
business model is to be treated as if it is the only webcasting 
production model in the whole webcasting industry. Instead of a market 
demand curve, Dr. Fratrik puts forward the implicit assumption that the 
amount of sound recording performances demanded must be whatever his 
representative firm deems best for its particular technological and 
organizational structure. But no one firm's demand curve is equivalent 
to the market's demand curve, unless that firm is a monopsonist. 
Rather, as we have noted in Webcaster II and the CARP noted in 
Webcaster I before us, in the hypothetical marketplace we attempt to 
replicate, there would be significant variations, among both buyers and 
sellers, in terms of sophistication, economic resources, business 
exigencies, and myriad other factors. Webcaster II, 72 FR 24087 (May 1, 
2007); In the Matter of Rate Setting for the Digital Performance of 
Sound Recordings and Ephemeral Recordings, Report of the Copyright 
Arbitration Panel to the Librarian of Congress, Docket No. 2000-9 CARP 
DTRA 1&2 (``Webcaster I CARP Report'') at 24.
    Finally, even assuming the absence of the additional errors 
catalogued below, Dr. Fratrik's analysis, which focuses on past 
operating income statements to determine a royalty rate for all 
commercial webcasters in the future, fails to establish any behavioral 
information that would help to delineate the hypothetical marketplace 
we must replicate. Instead, Dr. Fratrik's analysis is largely 
mechanical and leads to an unsupported conclusion that past revenues 
and non-royalty costs, coupled with a webcaster operating profit margin 
not demonstrated to be related to past operating revenue and cost 
considerations (see infra at Section II.B.3.a.ii.), will repeatedly 
recur at the same levels in each year over the five-year period of the 
license going forward. Having tightly constrained the possibilities of 
market behavior in this manner, Dr. Fratrik's model then automatically 
produces an unchanging residue and, hence, an unchanging royalty rate 
for the whole period.\9\ This is a dubious result that flows from the 
unwarranted assumption of what amounts to a behavioral straitjacket.
---------------------------------------------------------------------------

    \9\ In addition to the flat royalty rate growth recommended by 
Dr. Fratrik over the 2011-2015 term, his recommended royalty rate of 
$0.0009 per performance would return the statutory rate to near its 
2006 statutory level.
---------------------------------------------------------------------------

    Moreover, even if Dr. Fratrik's problematic behavioral constraints 
and implicit assumptions somehow could be ignored, his analysis suffers 
from so many other unwarranted explicit assumptions and data defects as 
to make it untenable.
ii. The Specific Elements of Dr. Fratrik's Model
    Dr. Fratrik's assumptions regarding webcasting industry costs, 
revenues and profit margins are seriously flawed when viewed 
individually. Moreover, these flaws are compounded by merging revenue, 
costs and profit margin information gathered from disparate data 
sources into a single ``economic model.'' \10\
---------------------------------------------------------------------------

    \10\ Dr. Fratrik uses the term ``economic model'' to broadly 
describe his analysis. It is more closely akin to a type of pro 
forma income statement that attempts to demonstrate the expected 
effect of varying royalty rates on a firm's financial viability. In 
other words, it is an accounting model that, relying on historical 
cost and revenue data for all but royalty costs, endeavors to 
demonstrate the anticipated results of alternative royalty rates on 
projected net revenues.
---------------------------------------------------------------------------

    Dr. Fratrik begins by assuming that ``Live365's cost structure will 
serve as a good conservative proxy for the industry as it is a mature 
operator.'' Fratrik Corrected and Amended WDT at 16 (emphasis added). 
This assumption is not supported by the record of evidence in this 
proceeding which points to a wide variety of existing webcasting 
services and business models. SX PFF at ] 323. It defies credulity to 
claim, as does Live365, that all these disparate business models may be 
experiencing essentially the same unit costs. Indeed, Dr. Fratrik makes 
this assertion while recognizing that, unlike for many other 
participants in the market, at least two separate lines of business can 
be distinguished for Live365 (broadcasting services and webcasting) 
and, further, that Live365 acts as an aggregator with respect to 
webcasting. Dr. Fratrik offers no example of a comparable analogous 
participant in the industry who is structured in this manner. 
Furthermore, when he attempts to adjust Live365's costs to reflect only 
webcasting operations, he fails to adequately do so and he ignores the 
synergistic nature of Live365's various lines of business. SX PFF at ]] 
355, 357, 358. Finally, even though he argues for an additional 
aggregator discount to be applied to Live365's webcasting royalty rates 
based on monitoring and reporting savings purportedly provided to the 
collective (i.e., SoundExchange), he nowhere appears to adjust 
Live365's webcasting cost estimates to account for any resulting 
differences in costs that Live365 may incur as compared to other 
webcasters who are not aggregators. He makes no such adjustment despite 
the fact that it is the typical webcaster's unit costs he is seeking to 
model rather than the typical aggregator's unit costs. While any 
additional reporting and monitoring costs incurred by aggregators \11\ 
may be offset by fees charged to the aggregated webcasters or by the 
reduced costs of programming that Live365 would otherwise have to 
undertake in order to make comparable channel offerings as a multi-
channel broadcaster, such salient differences between the typical 
webcaster's unit costs and the typical aggregator's unit costs are not 
addressed by Dr. Fratrik's analysis. For all these reasons, the unit 
cost estimation for webcasting which Dr. Fratrik offers is seriously 
flawed.
---------------------------------------------------------------------------

    \11\ For example, Dr. Fratrik notes that, in connection with its 
aggregation services, ``Live365 has spent a considerable amount of 
time and investment establishing its software systems to accurately 
measure and document listening for each copyrighted work that is 
streamed.'' Fratrik Corrected and Amended WDT at 38 n.62.
---------------------------------------------------------------------------

    On the revenue side of his analysis, Dr. Fratrik assumes that: (1) 
Webcaster revenue comes from advertising revenue and subscription 
revenue; (2) ``publicly available industry reports from AccuStream and 
ZenithOptimedia serve as the lower and upper bounds, respectively, on 
advertising revenue measurements for the past period;'' and (3) 
Live365's subscription revenue per listening hour can be utilized as a 
proxy for gauging subscription revenues in the webcasting industry. 
Fratrik Corrected and Amended WDT at 16-17, 24-25.

[[Page 13030]]

    Live365's rate proposal in this proceeding (i.e., $.0009 per 
performance throughout the period 2011-2015), however, is apparently 
based only on Dr. Fratrik's analysis of revenues using the 
ZenithOptimedia data. Indeed, use of the Accustream revenue data 
alternative produces the anomalous result that copyright owners would 
have to pay webcasters each time the owners' sound recordings were 
performed, no matter how low a profit margin Dr. Fratrik assumed for 
webcasters in his analysis. Fratrik Corrected and Amended WDT at 26, 
Table 4; 4/27/10 Tr. at 1157:1-1158:6 (Fratrik).
    Undaunted by this anomalous result, Dr. Fratrik simply repeats his 
analysis, substituting, in part, the ZenithOptimedia advertising 
revenue data for the Accustream advertising revenue data and, in 
concert with a 20% assumed profit margin, obtains the $.0009 per 
performance royalty rate that has been proposed by Live365 to be 
applied without change throughout the period 2011-2015. Yet Dr. 
Fratrik's alternative ZenithOptimedia-based analysis does not 
completely divorce itself from the Accustream data; instead, because 
ZenithOptimedia did not provide the Aggregate Tuning Hours (``ATH'') 
numbers associated with its total advertising revenue estimate, Dr. 
Fratrik fell back on the Accustream data for a total ATH number and 
calculated advertising revenue per ATH by dividing the ZenithOptimedia 
revenue data by the Accustream ATH data. In short, Dr. Fratrik combines 
advertising revenue data based on two separate data sources without 
making a determination that the data was capable of being combined in 
this manner.
    Moreover, even Dr. Fratrik admitted that the ZenithOptimedia and 
Accustream advertising revenue estimates are ``challenging'' or 
difficult to produce because a vast number of webcasters do not report 
their revenues publicly. 4/27/10 Tr. at 1220:1-20 (Fratrik). Thus, 
these databases have clear limitations and the uncritical manner in 
which Dr. Fratrik mixes and matches data from these two separate 
advertising revenue databases and then further combines subscription 
revenue data from a third separate source (i.e., the Live365 
subscription revenue data) plainly suggests a less than rigorous 
approach to his analysis.
    Finally, with respect to revenues, Dr. Fratrik's analysis reports, 
but neither takes into account nor provides an adequate explanation 
for, the growth in the ZenithOptimedia advertising revenues forecast 
from his 2008 base through 2011 (i.e., growth from $200 million to $291 
million). Fratrik Corrected and Amended WDT, Ex. 8 at 187. It may be 
argued that growth in the level of revenues does not necessarily 
translate into growth in unit revenues. However, we find that it is 
difficult to accept Dr. Fratrik's unsupported assertion that he expects 
little improvement in such revenues on a unit basis (see Fratrik 
Corrected and Amended WDT at 5). Dr. Fratrik fails to provide any 
adequate empirical support for the implied assumption necessary to 
reach this conclusion--an assumption that the growth in performances 
will take place at precisely the pace necessary to assure that the 
anticipated growth in revenues over the relevant period will not alter 
the unit revenue ratio. Moreover, without such an implied assumption, 
it is difficult to avoid the conclusion that Dr. Fratrik's constant 
royalty rate should have been adjusted each year based on the 
implications of growing revenues for his own model. Yet, he offers no 
such adjusted royalty rate. At the very least, these changing 
advertising revenue totals call into question the reliability of the 
unchanging royalty rate derived by Dr. Fratrik from the lowest of the 
revenue totals available from the same data source (i.e., $200 million 
instead of $291 million).
    Dr. Fratrik's assumption of a 20% operating margin for webcasters 
in his analysis is not solidly supported. That operating profit margin 
is not put forward as either a historical profit margin or a forecasted 
profit margin for webcasters, but rather as a profit margin derived 
from the over-the-air broadcasting industry. SX PFF at ]] 328, 330. The 
record of evidence in this proceeding does not support the notion that 
profit margins for webcasters are likely to be similar to the more 
capital intensive terrestrial radio industry. SX PFF at ]] 332-5. 
Furthermore, we find that Dr. Fratrik failed to establish a solid basis 
for concluding that the minimum operating profit margin for his 
representative webcaster was comparable to the average firm experience 
from firms that operate on a different platform (over-the-air radio).
    Live365 argues in its proposed reply findings at ] 327 that Dr. 
Fratrik's 20% profit margin assumption is further corroborated by the 
recording industry's own expert testimony in Webcaster I (offered by 
Dr. Thomas Nagle, Chairman, Strategic Pricing Group, Inc.) which 
purportedly ``recommended that webcasters should be able to achieve 
margins between 13.2% and 21.8%.'' However, although the Nagle exhibit 
referred to by Live365 was appended to Dr. Salinger's written rebuttal 
testimony, the exhibit was only mentioned briefly in a footnote to the 
Salinger testimony and then only to make a different argument. Dr. 
Salinger, in fact, made no specific reference to any of the varying 
operating profit-margin figures utilized in that 2001 Recording 
Industry Association of America (``RIAA'') study. In other words, it 
can hardly be said that the figures in question were offered as 
``corroborative'' evidence to support Dr. Fratrik's assumptions. 
Moreover, the point of this 2001 study appears to have been to 
recommend a royalty rate based on the operating profit margins 
necessary to generate an assumed range of rates of return on investment 
for webcasters. In fact, the Nagle study utilized an operating profit 
margin in the range of 8.43% to 17.05% in order to ``arrive at the 
appropriate range for the statutory license royalty fee.'' See Salinger 
WRT, Exhibit 3 at 16 and Appendix 3 at 1. Dr. Fratrik's 20% assumption 
for webcaster operating profit margins lies substantially outside this 
range. Moreover, the CARP rejected Dr. Nagle's analysis as 
corroborating evidence in Webcaster I. [``Dr. Nagle's analysis 
necessarily relies upon a myriad of highly questionable assumptions 
that appear inconsistent with foreseeable market conditions.''] 
Webcaster I CARP Report at 73; [``We conclude that Dr. Nagle's analysis 
does not support any particular rate level.''] Id. at 74. We find it 
provides no corroborative support for Dr. Fratrik's assumed 20% 
webcaster operating profit margin in this proceeding.
    Thus, we find that Dr. Fratrik's ``model'' is based upon a series 
of assumptions and analogies that, taken individually, add such a 
degree of uncertainty or inexactitude to the resulting model as to make 
it unsatisfactory for the purpose of portraying the likely outcome of 
negotiations between willing buyers and willing sellers in the market 
for sound recording inputs that are used in webcasting services. 
Indeed, Dr. Fratrik's model does not even adequately address some of 
the modest considerations for a modeling approach laid out by Live365's 
rebuttal expert, Dr. Salinger. SX PFF at ] 307. Questionable 
assumptions, reservations about the methodological appropriateness of 
mixing disparate data sources, and concerns over the resulting 
reliability of the data used in the Fratrik model lead us to find that 
this theoretical construct suffers serious deficiencies that do not 
lend themselves to remediation.

[[Page 13031]]

iii. Other Factors Put forward for Consideration
    Live365 offers several other arguments to buttress its request for 
a royalty rate that would effectively return the statutory rates to 
near their 2006 statutory level.
    First, Dr. Fratrik maintains that ``[a]s industry projections for 
more robust growth in the Internet radio advertising market have 
clearly not materialized over the past few years,'' his valuation model 
must give rise to the conclusion that a ``reduction in royalty rates 
from the prescribed rates covering 2006-2010'' is warranted. Fratrik 
Corrected and Amended WDT at 31. In so doing, he incorrectly attributes 
the annual increase in rates established in Webcaster II to projections 
of growth primarily provided by Dr. Erik Brynjolffson and Mr. James 
Griffin in that proceeding. Fratrik Corrected and Amended WDT at 12-14. 
Similarly, Live365 argues that ``[g]iven that the lofty expectations 
from the Webcasting II proceeding have not been fulfilled, it follows 
that the rates for the next five years should be set lower than the 
rates determined by the CRB [Judges] in Webcasting II.'' See Live365 
PFF at ] 38. But, quite to the contrary, the Judges' determination in 
Webcaster II did not rely on those particular predictions in setting 
rates. Indeed, the Judges expressly rejected Dr. Brynjolfsson's 
modeling attempt and specifically cited the flaws in his effort ``to 
project future growth rates'' as a basis for not relying on them. 
Webcaster II, 72 FR 24093. Moreover, the evidence in the record on 
industry growth over the 2006-2010 period which shows increased 
advertising revenues, increased performances, and increased listening 
does not support a rate reduction. It more likely would support at 
least some modest rate increase. See SX PFF at ]] 390-395, 398-401. 
While some Live365 data may show a flattening or decline for a 
particular pair of years, the overall trend of that same data does not 
show a decrease. For example, data presented by Live365 shows a year-
to-year decline in listenership from 2006 to 2007, but this is followed 
by substantial increases in 2008 and 2009 and maintenance of 2009 
levels in 2010. Overall, the trend in such listenership recorded since 
2000 has been decidedly upward, even though the growth has occurred 
unevenly from year to year. See Smallens Corrected WRT at 7, Table 1.
    Second, Live365 also contends that a downward adjustment of the 
current royalty rate is appropriate based on (1) The promotional value 
of statutory webcasting relative to its non-substitutional effect on 
other sales of music, including the promotional value to copyright 
owners stemming from the wide array of music and artists played on 
statutory webcasting services; (2) the relative creative contributions, 
technical contributions, investments, costs and risks made or borne by 
commercial webcasters compared to copyright owners; and (3) the 
relative disparate impact of certain competitive factors on webcasters 
as compared to copyright owners. After careful consideration, we find 
that the evidence submitted by Live365 on each of these claims is weak 
at best and, most certainly, too weak to establish the basis for a 
decrease in webcaster royalty rates. SX PFF at ]] 415, 419-21, 426, 
431, 446-9; SX RFF at ]] 176, 179-180. Then too, Live365 does not 
present an acceptable empirical basis for quantifying the individual 
asserted effects of these various factors and/or for deriving a method 
for translating such magnitudes into a rate adjustment. Moreover, to 
the extent that Live365 claims that the Fratrik valuation model makes 
such a quantifiable translation, we need not further address these 
issues separate from our examination of that model which we have found 
seriously flawed and an inadequate representation of the market.
b. The SoundExchange Benchmark Approach
i. The Interactive Webcasting Market Benchmark
    As in Webcaster II, SoundExchange maintains that one set of 
benchmark agreements with clear relevance for this proceeding as shown 
by an analysis prepared by its expert economist, Dr. Michael Pelcovits, 
consists of those agreements found in the market for interactive 
webcasting covering the digital performance of sound recordings. That 
is because the interactive webcasting market has characteristics 
reasonably similar to non-interactive webcasting, particularly after 
Dr. Pelcovits' final adjustment for the difference in interactivity.
    Both markets have similar buyers and sellers and a similar set of 
rights to be licensed (a blanket license in sound recordings). Both 
markets are input markets and demand for these inputs is driven by or 
derived from the ultimate consumer markets in which these inputs are 
put to use. In these ultimate consumer markets, music is delivered to 
consumers in a similar fashion, except that in the interactive case the 
choice of music that is delivered is usually influenced by the ultimate 
consumer, while in the non-interactive case the consumer usually plays 
a more passive role. This difference is accounted for in the Pelcovits 
analysis. In order to make the benchmark interactive market more 
comparable to the non-interactive market, Dr. Pelcovits adjusts the 
benchmark by the added value associated with the interactivity 
characteristic. Pelcovits Amended and Corrected WDT at 23. This results 
in a rate of $0.0036 per play for a statutory non-interactive webcaster 
as a possible outcome in the target market. Pelcovits Amended and 
Corrected WDT at 4, 33.
    The Judges find the interactive webcasting benchmark to be of the 
comparable type that the Copyright Act invites us to consider. 17 
U.S.C. 114(f)(2)(B). (``In establishing such rates and terms, the 
Copyright Royalty Judges may consider the rates and terms for 
comparable types of digital audio transmission services and comparable 
circumstances under voluntary license agreements negotiated under 
subparagraph (A).'') Nevertheless, as we indicated in Webcaster II, 
this particular Pelcovits benchmark analysis is not without warts. 
Webcaster II, 72 FR 24094 (May 1, 2007).
    In Webcaster II we recognized the potential implications of a 
benchmark analysis that focuses on only subscription services as does 
the interactive benchmark presented by Dr. Pelcovits. That is, ad-
supported non-interactive services might pay less than subscription-
based interactive services to use the same music if their advertising 
revenues failed to evolve to the point where ad-supported non-
interactive services were just as lucrative as subscription-based 
interactive services on a per-listener hour basis. In that proceeding 
the Judges indicated that to the extent that ad-supported revenues did 
not come to match subscription revenues on a per-listener hour basis 
during the 2006-2010 term and, absent clear information on the 
substitutability of the subscription and non-subscription options among 
consumers, any resulting shortfall related to ad-supported webcasting 
revenues would likely be adequately mitigated by a phase-in of the per 
performance rates to the level indicated by the benchmark analysis, 
such that the benchmark recommended rate for 2006 would not become 
effective until the last year of the term. Webcaster II, 72 FR 24094 
(May 1, 2007).
    Here, unlike the absence of data supporting this critique which we 
noted in Webcaster II, Dr. Salinger provides some empirical data to 
support the position that a benchmark which

[[Page 13032]]

reflects a weighted average of revenues obtained from subscribers and 
non-subscribers may result in a lower estimated royalty rate than Dr. 
Pelcovits' benchmark which focuses on only subscription rates. Salinger 
WRT at 10-11. Therefore, we are not persuaded that Dr. Pelcovits' 
benchmark estimates are sufficiently reflective of the hypothetical 
target market as to support the immediate implementation of a royalty 
rate equivalent to the $0.0036 outcome estimated by Dr. Pelcovits. Some 
further downward adjustment to his recommendation to adequately address 
the subscription/non-subscription revenue level differences may well be 
in order, although the magnitude of such an adjustment is not clear.
    While Dr. Salinger shows that there is likely some ``upward bias'' 
introduced into the Pelcovits analysis through its focus on only 
subscription-based services in the benchmark market, the amount of such 
upward bias is not persuasively determined. Non-interactive webcasters 
in the market like Live365 often provide both subscription and non-
subscription offerings. 7/28/10 Tr. at 40:10-15 (Salinger). Therefore, 
subscription-based revenues clearly must be considered. Moreover, the 
data used by Dr. Salinger to support his criticism, as Dr. Salinger 
admits, is not without its shortcomings. 7/28/10 Tr. at 98:2-104:6 
(Salinger). Similarly, Dr. Fratrik admitted that the ZenithOptimedia 
and Accustream advertising revenue estimates are ``challenging'' or 
difficult to produce because a vast number of webcasters do not report 
their revenues publicly. 4/27/10 Tr. at 1220:1-20 (Fratrik). There is 
also the difficulty of segmenting intermingled revenues from webcasting 
business models that may often directly and/or indirectly depend on 
both subscription and nonsubscription lines of business, as well as 
potentially on other sources of revenue. 7/28/10 Tr. at 40:10-15, 92:1-
19 (Salinger); Ordover WRT at 10-11. Nevertheless, Dr. Salinger's 
critique is sufficiently supported to raise legitimate concerns about 
the potential for upward bias in the Pelcovits estimates. It is only 
the magnitude of the potential upward bias that is not clearly 
quantified. What is clear from the record of evidence in this 
proceeding is that $0.0036 can be no more than the upper bounds of the 
range of possible rates reasonably applicable to the target market and 
that the most likely prevailing rate in that market is currently lower 
than $0.0036.
    Dr. Salinger also criticizes the Pelcovits interactive webcasting 
benchmark analysis for: (1) Relying only on contracts with the four 
major record companies to the exclusion of the independent record 
labels; (2) ignoring the downward trend in the effective play rates 
paid by interactive services by utilizing the average rate in his 
calculations; and (3) inappropriately constructing the hedonic 
regression model that is used as one alternative measure of 
interactivity in the analysis. Salinger WRT at 15-21.
    The first of these criticisms fails for lack of persuasive evidence 
in the record that the use of independent record contracts would have 
made a material difference. SX RFF at ]] 101-103.
    Although the second and third criticisms have some merit, the 
Judges find that these criticisms indicate that the Pelcovits 
interactive webcasting benchmark may overstate the likely prevailing 
market rate in the target market without necessarily rendering the 
Pelcovits analysis fatally flawed. With respect to the second 
criticism, Dr. Salinger acknowledged that this concern could be 
addressed by multiplying the recommended rate by 0.8737.\12\ SX PFF at 
] 209. Such an adjustment, of course, would reduce the recommended 
rate. SoundExchange offers no evidence that such an adjustment is 
unwarranted and even appears to endorse such an approach by performing 
this exact calculation with respect to the $0.0036 rate and reducing it 
to $0.0031. See SX PFF at ] 210. But SoundExchange's calculation was 
applied to the highest possible outcome Dr. Pelcovits lists for his 
benchmark analysis (i.e., $0.0036), when in fact, Dr. Pelcovits 
indicates that his rate after substitution adjustment would result in a 
``range of recommended rates'' with a ``simple average of $0.0033.'' 
Thus, it appears that this $0.0033 average also requires adjustment to 
meet Dr. Salinger's criticism (e.g., to approximately $0.0029). This is 
not a trivial consideration in light of the fact that in Webcaster II, 
it was Dr. Pelcovits' recommended rates after the substitution 
adjustment that formed the basis for SoundExchange's rate proposal and 
that formed the basis for the determination by the Judges of a royalty 
rate to be achieved by the end of the term in 2010 (i.e., a per play 
rate of $0.19). See Webcaster II, 72 FR 24096 (May 1, 2007). In any 
event, the validity of this criticism of the Pelcovits approach 
regarding the effective per play rate clearly erodes the weight to be 
accorded to the $0.0036 figure.
---------------------------------------------------------------------------

    \12\ The 0.8737 multiplier represents the value of a ratio where 
the numerator consists of the effective per play rate for 2009 
(i.e., 0.01917) and the denominator consists of the average 
effective play rate over the three years in question (i.e., 
0.02194).
---------------------------------------------------------------------------

    Dr. Salinger also criticizes the Pelcovits hedonic regression 
analysis that formed the basis for one of the alternative measures of 
interactivity in the interactive webcasting benchmark approach. Dr. 
Salinger expressed concerns about the use of certain fixed effects 
variables (alternatively described as dummy variables) in the 
specification of the regression model and about the broad confidence 
interval surrounding the estimated interactivity coefficient in the 
hedonic regression. Salinger WRT at 20; 21 n.31 and Exhibit 6; 7/28/10 
Tr. at 66:4-69:22 (Salinger). These criticisms have some merit, 
especially in light of Dr. Pelcovits' admitted lack of familiarity with 
some of the relevant economic literature, including recent literature 
cautioning against the indiscriminant use of dummy variables in certain 
hedonic estimations. 4/20/10 Tr. at 373:18-376:15 (Pelcovits). 
SoundExchange, in response to this criticism, claims that any problem 
associated with the hedonic regression is negated by Dr. Pelcovits' use 
of other methods that result in rates almost identical to the $0.0036 
average. See, for example, SX RFF at ] 107. However, this does not 
wholly obviate the impact of any resulting overstatement. The rate 
associated with the hedonic regression is the highest of the three 
values that are used to calculate the $0.0036 average. Removing the 
rate associated with the hedonic regression from the average would, in 
this case, reduce the average. Thus, this criticism of the Pelcovits 
approach additionally erodes the weight that the Judges accord to the 
$0.0036 figure.
    In short, the potential for upward bias or actual demonstrated 
upward bias in the Pelcovits estimates persuade us that $0.0036 can be 
no more than the upper bounds of the range of possible rates reasonably 
applicable to the target market and that the most likely prevailing 
rate at the present time in that market is significantly lower than 
$0.0036.
ii. The National Association of Broadcasters and SiriusXM Agreements
    In addition to the interactive webcasting benchmark, Dr. Pelcovits 
offers a second benchmark based on the average of rates established for 
the 2011-2015 term in precedential Webcaster Settlement Act Agreements 
(``WSA agreements'') between SoundExchange and the National Association 
of Broadcasters and between SoundExchange and SiriusXM (``SiriusXM 
agreement'' or ``Commercial

[[Page 13033]]

Webcasters agreement''). Pelcovits Amended and Corrected WDT at 22.
    While these precedential WSA agreements certainly pertain to rates 
to be paid by non-interactive webcasters in the commercial webcasting 
market at issue in this proceeding, the buyers' and sellers' 
circumstances are not comparable to those that would prevail in the 
absence of the Webcaster Settlement Act. Rather than a single seller, 
the sellers in the hypothetical market we are to consider consist of 
multiple record companies. Webcaster II, 72 FR 24087, 24091 (May 1, 
2007); Webcaster I, 67 FR 45244 (July 8, 2002). Thus, in Webcaster II 
we found that the fact that there were multiple buyers and multiple 
sellers in the benchmark market as well as in the target market 
supported a benchmark analysis. Webcaster II, 72 FR 24093 (May 1, 
2007). While the applicable law does not require a perfectly 
competitive benchmark market, the market must be at least 
``competitive'' in the sense that buyers and sellers have comparable 
resources and market power. Webcaster II, 72 FR 24093 (May 1, 2007); 
Webcaster I, 67 FR 45245 (July 8, 2002). This would be generally 
consistent with free market principles. Yet, the buyers' and sellers' 
circumstances underlying the WSA agreements were not comparable to 
market conditions that would prevail in the absence of the WSA. That 
legislation permitted a single seller representative to enter into 
negotiations with buyers in the market with respect to rates that would 
be permitted to supplant the statutory rates previously established in 
the 2006-2010 period, as well as with respect to rates applicable to 
the 2011-2015 period. Even Dr. Pelcovits admits that ``[e]ach of these 
contracts, of course, was negotiated in the shadow of the regulatory 
scheme and against the background of statutory rates previously set by 
this Court. To that extent, they may or may not represent the same 
outcome that would result in a pure market negotiation with no 
regulatory overtones.'' Pelcovits Amended and Corrected WDT at 15. 
Therefore, we find that these precedential WSA agreements, which may be 
fairly characterized as single-seller agreements reached under atypical 
marketplace conditions, cannot satisfy the comparability requirements 
for an appropriate benchmark.
    However, we further find that, because the NAB-SoundExchange and 
SiriusXM-SoundExchange agreements clearly govern the rates for a 
substantial number of commercial webcasters over the relevant 2011-2015 
period (Pelcovits Amended and Corrected WDT at 15) and the commercial 
webcasters covered by these agreements are competitors with the other 
commercial webcasters who comprise the remainder of the non-interactive 
webcasting services (Salinger WRT at 24; Smallens Corrected WRT at 21), 
these agreements are a useful gauge of the weight to be assigned to the 
rates suggested by the interactive webcasting benchmark discussed supra 
at Section II.B.3.b.i. Moreover, nothing in the Webcaster Settlement 
Act constrains us from using these agreements for that purpose. See 17 
U.S.C. 114(f)(5)(C).
    The NAB-SoundExchange and SiriusXM agreements provide for royalty 
rates on a per performance basis. For the five-year period beginning 
2011, the NAB-SoundExchange agreement sets the following rates: $0.0017 
for 2011, $0.0020 for 2012, $0.0022 for 2013, $0.0023 for 2014 and 
$0.0025 for 2015. For the same period, the SiriusXM agreement sets the 
following rates: $0.0018 for 2011, $0.0020 for 2012, $0.0021 for 2013, 
$0.0022 for 2014 and $0.0024 for 2015. Pelcovits Amended and Corrected 
WDT at 15. Two characteristics of these rates are noteworthy. First, 
the 2011 rate is slightly less than the current 2010 statutory rate of 
$0.0019 and the rates in the precedential WSA agreements covering the 
years 2009 and 2010 were somewhat lower than the corresponding 
statutory rate for those years. Pelcovits Amended and Corrected WDT at 
15. Second, the rates in the NAB-SoundExchange and SiriusXM agreements 
over their entire term are substantially lower than the range of annual 
rate possibilities suggested for implementation pursuant to the 
proposed interactive benchmark ($0.0036) or the interactive benchmark 
after Dr. Pelcovits' substitution adjustment ($0.0033) or the 
interactive benchmark adjusted to give a more likely reading of the 
impact of downward trend in the effective play rates paid by 
interactive services ($0.0031).
    Thus, we find that these negotiated rates indicate that the 
interactive benchmark may likely overstate the prevailing market rate 
in the target market even when subjected to Dr. Pelcovits' substitution 
adjustment or Dr. Salinger's adjustment to mitigate the impact of 
downward trend in the effective play rates paid by interactive 
services. As a consequence, we further find that the interactive 
benchmark, even when subjected to these alternative adjustments, 
provides for rates near the upper bounds of the range of possible rates 
reasonably applicable to the target market, when the most likely 
prevailing rate in that market appears to be lower than the interactive 
benchmark rates. In other words, the NAB-SoundExchange and SiriusXM 
agreements lend weight to the need for a further downward adjustment in 
the benchmark rate to reflect a prevailing rate in the target market 
closer to the current statutory rate.
    Dr. Fratrik contends that the royalty rates in the NAB-
SoundExchange agreement must overvalue the input in question, because 
the NAB received a particularly valuable concession with respect to the 
waiver of performance complement rules as part of the rate agreement. 
See Fratrik Corrected and Amended WDT at 43-44. [``Consequently, these 
terrestrial broadcasters, already with the programming established to 
webcast, should be willing to pay more than other webcasters in order 
to relieve themselves of these provisions.'' (emphasis added)]. This 
claim of a one-sided benefit to broadcasters is not adequately 
supported in the record. The testimony of Dr. Pelcovits, Dr. Ordover 
and Mr. McCrady indicates that the waivers had value to both the NAB 
and to the record companies. Pelcovits Amended and Corrected WDT at 20 
n.21; Ordover WRT at 5, 18; McCrady WDT at 5-6. There is no clear 
evidence in the record to support either the notion that the limited 
performance complement waiver in the NAB-SoundExchange agreement was a 
largely one-sided benefit accruing only to the broadcasters or that 
broadcasters did, in fact, pay more than other webcasters to obtain 
these provisions.
    Dr. Fratrik also contends that terrestrial broadcasters were 
willing to pay more because they have fewer other costs to cover than 
pure webcasters. But Dr. Fratrik offers less than persuasive evidence 
of major cost differences between pure webcasters and broadcasters who 
engage in webcasting generally or between pure webcasters and the more 
limiting case of those broadcasters who exclusively simulcast. Dr. 
Fratrik appears to center his analysis on the latter case. Of course, 
focusing on this latter comparison simplifies from the reality of the 
market by assuming that all the webcasting performed by broadcasters 
consists of simulcasting when, in fact, the NAB-SoundExchange agreement 
provides for other types of webcasting (e.g., through side channels). 
See SX Ex. 102-DP at Article 1.1(d), 4.2. In addition to that 
analytical shortcoming, Dr. Fratrik's analysis suffers from other 
unsupported conclusions. Dr. Fratrik's cost-based contention appears to 
largely rest on the notion that simulcasters, unlike other

[[Page 13034]]

commercial webcasters, have no additional programming costs as those 
costs have already been paid in connection with their over-the-air 
operations. See Fratrik Corrected and Amended WDT at 41. But no 
specific empirical data in the record unambiguously supports this 
asserted relative difference. For example, Dr. Fratrik's conclusion 
ignores the wide range of business models utilized by commercial 
webcasters, including that of Live365, a webcaster that is apparently 
paid to put on programming designed by its clients as opposed to 
incurring a cost for originating such programming itself. Floater 
Corrected WDT at 4-8; 4/27/10 Tr. at 1274:5-16; 1301:1-4