2010-23266. Distribution of the 2004 and 2005 Cable Royalty Funds

AGENCY:

Copyright Royalty Board, Library of Congress.

ACTION:

Distribution order.

SUMMARY:

The Copyright Royalty Judges are announcing the final Phase I distribution of cable royalty funds for the years 2004 and 2005.

DATES:

Effective September 17, 2010.

ADDRESSES:

The final distribution order also is posted on the Copyright Royalty Board Web site at http://www.loc.gov/​crb/​proceedings/​2007-3/​final-distribution-order.pdf.

FOR FURTHER INFORMATION CONTACT:

Richard Strasser, Senior Attorney, or Gina Giuffreda, Attorney Advisor, by telephone at [phone] or by e-mail at [e-mail].

SUPPLEMENTARY INFORMATION:

I. Background

On July 15, 2008, the Copyright Royalty Judges published in the Federal Register a notice announcing the commencement of a proceeding to determine the Phase I distribution of royalties collected from cable systems under the section 111 statutory license for the period 2004 and 2005.^[1] 73 FR 40623. The notice also requested interested parties to submit their Petitions to Participate in the proceeding no later than August 18, 2008. Petitions to Participate, all of which were joint petitions, were received from the following claimants: Public Broadcasting Service for Public TV Claimants (“PTV”); National Public Radio (“NPR”); Joint Sports Claimants (“JSC”); Canadian Claimants Group (“Canadian Claimants”); Devotional Claimants; the Motion Picture Association of America, Inc. (“MPAA”) for certain Program Supplier Claimants (“Program Suppliers”); Music Claimants;^[2] and the National Association of Broadcasters for all U.S. commercial television broadcast stations retransmitted by cable operators as distant signals during 2004 and 2005 (“CTV”). The Judges accepted these petitions. Order Announcing Negotiation Period, Docket No. 2007-3 CRB CD 2004-2005 (October 31, 2008).

After the expiration of the mandatory negotiation period, the parties were directed to submit their written direct statements on or before June 1, 2009.^[3 4] The Judges received written direct statements from Canadian Claimants; Program Suppliers; Devotional Claimants; and JSC, CTV, PTV, and Music Claimants (collectively, the “Settling Parties”). Discovery in the direct phase of the proceeding was conducted throughout June and July, and the hearings were conducted from October 6-20, 2009. The Settling Parties presented the following witnesses: James M. Trautman, Managing Director of Bortz Media & Sports Media, Inc.; Dr. Robert W. Crandall, Senior Fellow in Economic Studies at the Brookings Institution; Judith Meyka, independent consultant with clients in the cable and satellite television industry; Linda McLaughlin, Special Consultant to National Economic Research Associates, Inc.; Dr. Richard V. Ducey, Chief Strategy Officer, BIA Advisory Services; Dr. Joel Waldfogel, Ehrenkranz Family Professor of Business and Public Policy at the Wharton School of the University of Pennsylvania; Jerald N. Fritz, Senior Vice President for Legal and Strategic Affairs, Allbritton Communications Company; Seth Saltzman, Senior Vice President of Member Management in the Performing Rights Group, ASCAP; Michael O'Neill, Senior Vice President, Licensing, BMI; and William P. Zarakas, Principal, The Brattle Group.^[5]

The Canadian Claimants presented Dr. Debra J. Ringold, Dean, Atkinson Graduate School of Management, Willamette University.^[6]

The Devotional Claimants presented Dr. William Brown, Professor and Research Fellow, School of Communications and the Arts, Regent University.^[7]

The Program Suppliers presented the following witnesses: Marsha E. Kessler, Vice President of Retransmission Royalty Distribution, the MPAA; John Mansell, Jr., President/Chief Executive Officer, John Mansell Associates, Inc.; Howard B. Homonoff, Director in the Entertainment, Media and Communications advisory practice, PricewaterhouseCoopers LLP; Dr. Arthur C. Gruen, Partner/Co-Founder, Wilkofsky Gruen Associates; Paul Lindstrom, Senior Vice President, The Nielsen Company (“Nielsen”); Bruce Hoynoski, Senior Vice President and Chief Research Officer, Global Media for Nielsen; and Dr. George S. Ford, President, Applied Economics Studies, and Chief Economist, the Phoenix Center for Advanced Legal & Economic Policy Studies.^[8]

A rebuttal phase to the proceeding was requested by the parties, and written rebuttal statements were submitted by December 11, 2009. As a result of discovery on the written rebuttal statements, the Settling Parties and Program Suppliers filed a motion for adoption of a joint stipulation ^[9] regarding certain programming on Station WGN-TV (Chicago, Illinois) during the years 1998-99 and 2004-05, the adoption of which would obviate the need for the testimony of two witnesses for the Settling Parties: Dan Derian, Vice President of Research and Strategic Planning for Major League Baseball, and Marc Schader, former Senior Vice President of Programming for Tribune Broadcasting. The Judges granted the motion, and the Settling Parties withdrew the testimony of Messrs. Derian and Schader. See Order on Witnesses and Joint Stipulations, Docket No. 2007-3 CRB CD 2004-2005 (January 27, 2010); see also Tr. at 2335-36.

Rebuttal hearings were conducted February 1-5, 2010. The Settling Parties presented the rebuttal testimony of: Dr. Gregory S. Crawford, Professor of Economics, University of Warwick, United Kingdom; Jeffrey S. Berman, Senior Partner & Executive Vice President, C&R Research; Dr. Duncan; Edward S. Desser, President/Founder, Desser Sports Media, Inc.; and Mr. Trautman.^[10]

The Devotional Claimants presented the rebuttal testimony of Dr. Michael Salinger, Professor of Economics, Boston University School of Management and Managing Director of LECG.

The Canadian Claimants presented the rebuttal testimony of: Ms. Martin; Dr. Gary T. Ford, Emeritus Professor of Marketing, the Kogod School of Business, American University; Dr. John E. Calfee, Resident Scholar, American Enterprise Institute; and Dr. Brian T. Ratchford, Charles and Nancy Davidson Professor of Marketing, University of Texas at Dallas.

Program Suppliers presented the rebuttal testimony of: Ms. Kessler; Dr. John R. Woodbury, Vice President, Charles River Associates; and Mr. Mansell.^[11]

Proposed Findings of Fact and Conclusions of Law were submitted by the parties by March 17, 2010, and disputed findings were submitted by April 9, 2010. The parties also submitted Joint Agreed Findings of Fact and Conclusions of Law on April 19, 2010. Closing arguments were held on May 10, 2010, and the record to the proceeding was closed.^[12]

The Distribution Order was issued to the parties on June 29, 2010. Motions for Rehearing were filed by Program Suppliers and Canadian Claimant Group. On July 19, 2010, the Judges DENIED the Motions for Rehearing.

II. The Governing Distribution Standard

Section 803(a)(1) of the Copyright Act Provides:

The Copyright Royalty Judges shall act in accordance with this title, and to the extent not inconsistent with this title, in accordance with subchapter II of chapter 5 of title 5, in carrying out the purposes set forth in section 801. The Copyright Royalty Judges shall act in accordance with regulations issued by the Copyright Royalty Judges and the Librarian of Congress, and on the basis of a written record, prior determinations and interpretations of the Copyright Royalty Tribunal, Librarian of Congress, the Register of Copyrights, copyright arbitration royalty panels (to the extent those determinations are not inconsistent with a decision of the Librarian of Congress or the Register of Copyrights), and the Copyright Royalty Judges (to the extent those determinations are not inconsistent with a decision of the Register of Copyrights that was timely delivered to the Copyright Royalty Judges pursuant to section 802(f)(1)(A) or (B), or with a decision of the Register of Copyrights pursuant to section 802(f)(1)(D)), under this chapter, and decisions of the court of appeals under this chapter before, on, or after the effective date of the Copyright Royalty and Distribution Reform Act of 2004.

17 U.S.C. 803(a)(1).

All parties acknowledge that Congress did not set forth a statutory standard for cable royalty allocations. See, e.g., SP PCL at ¶ 6. Beginning with the Copyright Royalty Tribunal, standards were created to assist the distribution process, which changed through the years under the Tribunal and later under the Copyright Arbitration Royalty Panel (“CARP”) system administered by the Librarian of Congress.^[13] However, for purposes of this proceeding, the parties are all in agreement that the sole governing standard is the relative marketplace value of the distant broadcast signal programming retransmitted by cable systems during 2004 and 2005. See CCG PCL at ¶ 9; DPCL at ¶ 2; SP PCL at ¶ 6; PS PCL at ¶ 9.

In applying the relative marketplace value standard to this proceeding, we are cognizant of the requirements of section 803(a)(1) described above. We have considered all of the evidence and the arguments presented by the parties. To the extent that they are incorporated into our determination as to the proper distribution of the cable funds, they are accepted. To the extent they are not, they are rejected.

III. JSC, CTV, PTV and Program Suppliers Claimants' Awards

Having carefully reviewed and considered all of the evidence in the record, the Judges find that the values of the program categories at issue among these contending claimants are most reasonably delineated by a range bounded by certain results indicated primarily by the Bortz constant sum survey, to a lesser extent by the Waldfogel regression analysis and, to a slight extent, by the Gruen constant sum survey. For the reasons discussed below, the Judges find that no single methodological approach, even when ostensibly adjusted to account for acknowledged shortcomings, persuasively obviates the need for relying, at least to some small extent, on other reasonable valuation approaches that offer additional perspective from a different methodological vantage point.

The market value of the non-network programming that appears on distant signal stations that are retransmitted by cable systems is not directly measurable. That is because the price charged to the cable system for the right to retransmit such programming is not determined in a free market, but rather is determined statutorily. Therefore, the evidence adduced in this proceeding aims to show how the programming in question would be valued in a hypothetical free market that would exist but for the regulatory regime currently in place.

However, such a hypothetical free market value for non-network distant signal programming is also not directly observable, because cable operators purchase a bundle of programming when they purchase a distant signal's entire output. [“Q. And why didn't you ask them about actual expenditures by that cable system for programming? A. Well, that's not something that's really possible to do, because cable operators buy whole signals. They don't buy the individual-when they're buying distant signals, they buy entire signals that include, in—in most instances—instances, multiple types of programming or multiple categories of programming. And, therefore, they're not, in the distant signal purchase decisions, making expenditures for the—these particular categories of programming.” Tr. at 78 (Trautman).] Ergo, various alternative explanations about what induces cable system operators (the buyers) in a hypothetical distant signal market to exhibit preferences for one type of programming relative to the other types of programming that form part of the bundle on a distant signal station are the focus in this proceeding. The inducement to buy distant signals in the cable market stems from the derived demand for such distant signals as inputs in the various cable systems' channel lineups. In other words, any cable operator's demand for the programming input reflected in distant signals is only valuable to the extent that the demand for the total output of any cable system (i.e., bundles of service options) can be related to that particular input.

Analysis of the Settling Parties' Evidence

One approach to valuation, favored by the Settling Parties, explains the demand for distant signals by cable operators in terms of the strength of the cable system operators' expressed preferences for the types of programming that they identify with the distant signal. This is grounded in the notion that a cable operator's association of certain kinds of “signature programming” with a particular distant signal station tends to be the starting point for driving value. Tr. at 86 (Trautman). Thus, the Bortz survey is predicated on the notion that the cable operator respondents are focusing on “signature programming” that drives the value of the distant signal station to the cable operator. [“And I think what you're expressing there in that example is exactly what I'm talking about in terms of the dominant impression of value and the notion of signature programming. I think, on any of these distant signals, although it may—what constitutes signature programming could differ from one respondent to the next, they are, in fact, in answering this question, thinking exactly along the lines that you expressed.” Tr. at 91 (Trautman).] Following this line of analysis, the Settling Parties offer the Bortz constant sum survey of cable operators' relative preferences among certain categories of programming identifiably present on distant signal stations as determinative of the relative value of most of the categories of programming represented by the claimants in this proceeding.

Yet, it is not clear from the preferences expressed by the cable system operators who answer the Bortz survey questions where the key relative value question is limited to defining worth only “in terms of attracting and retaining subscribers,” whether the preferences so expressed would reflect actual demand in a more realistic view of a hypothetical free market. That is, the purchase of one type of channel by cable operators (such as distant signal stations) and the programming it reflects would not occur in a vacuum to the exclusion of consideration of the remaining content to bundle with that distant signal channel in the product ultimately offered to subscribers. Underlying subscriber demand for the programming that appears on a particular distant signal station is only one part of a more complex decision facing cable operators as to whether the input in question is more attractive than a cable network alternative in terms of the net revenue or profit maximization goals of the buyers. This is not a trivial concern inasmuch as the buyers in this case (cable operators) are not participants in perfectly competitive input markets or in perfectly competitive output markets for their services. In the input market for cable channel programming as well as in the output market for providing consumer subscribers with cable television services, cable system operators exercise varying degrees of market power. Therefore, it is less than realistic to assume that cable operators' programming purchases are driven only by meeting their underlying subscriber programming preferences when a myriad of other net revenue considerations may be involved in any programming decision.^[14]

One reason that more than just pure subscriber interests play a role in shaping the underlying demand for a cable operator's output is that the distant signal channels highlighted in this proceeding are not the subject of a direct choice by cable subscribers. Rather distant signal offerings are bundled together with non-distant signal broadcast channels, cable network channels and pay-per-view channels. Further, they are bundled into varying combinations of channels that are offered as different tiers of service for different prices. The bundles are packaged by the cable operator who selects the channel offerings, including any distant signal offerings. The rationale for the cable operator's decision concerning which channels to group in any tier offering and at what price, may depend not only on the impact on direct subscriber revenues, but also on such factors as advertising revenues associated with cable network channels, the relative license fee costs of various cable network channels, physical capacity constraints on the number of channels that can be transmitted over a particular cable system and even the direct ownership interests of the cable system in programming content on a given cable network.^[15] In short, the preferences expressed by the cable system operators who answer the Bortz survey, where the key relative value question is limited to defining worth only “in terms of attracting and retaining subscribers,” either may implicitly reflect more than an actual underlying subscriber demand for the programming that appears on a particular distant signal station or, alternatively, unrealistically minimize factors such as whether the input in question is more attractive than a cable network alternative in terms of the net revenue or profit maximization goals of the buyers.

This is not to say that the Bortz constant sum cable operator preference survey is substantially flawed, but rather that, given the interplay of all of the other factors described above that may color a cable operator's decision concerning the purchase of a distant signal input in a hypothetical cable market where the reality of bundling is taken into account, the Bortz survey's resulting point estimates are not a precise measure of all of those factors that may shape cable operator demand for the programming on distant signal stations. And, the Bortz study is certainly not a fully equilibrating model of supply and demand in the relevant hypothetical market, but rather a market research survey of buyer (i.e., cable operator) preferences in that market, characterized by a less than fully comprehensive explanation of what shapes those preferences. Therefore, for reasons discussed below, while the Judges find the Bortz study to be the most persuasive piece of evidence provided on relative value in this proceeding, the Bortz confidence intervals around each point estimate inspire more confidence than a strict adherence to the point estimates, particularly in relation to the larger claimants.

This is not to say that the Bortz survey should ignore the role of the subscriber growth factor in the demand for programming content or that subscriber growth is not a consideration facing cable operators in planning their programming decisions. To the contrary, as noted above, subscriber growth is one consideration facing cable operators in making programming decisions; and, underlying subscriber demand was explicitly and properly a factor which the survey respondents were asked to consider. Moreover, that there are factors other than subscriber growth considerations which may also be at work in influencing the demand for distant signal stations, does not change our finding that the Bortz survey focuses on the appropriate buyer in the hypothetical market—i.e., the cable operator.

Beyond the issue of the relevant contours of the hypothetical market, any study that purports to provide useful information on the relative value of the disparate categories of distant signal programming at issue in this proceeding must be reasonably well-founded methodologically. We find that the Bortz study is founded on a method—the constant sum survey—that has been long regarded as a recognized approach to market research. Tr. at 50 (Trautman), 1299 (Ringold), and 3007 (Gary Ford). Nevertheless, there are at least three aspects related to the execution of the Bortz survey methodology that we find additionally caution against regarding the Bortz point estimates as precise indicators of the relative value of the programming addressed in the record of this proceeding.

First, there may be bias introduced into the survey resulting from the respondents' potential misunderstanding of the exact parameters of the categories of programming they are being asked to compare in the key question (i.e., question 4) addressing valuation in the survey. [“There are—there certainly is the potential that in—in some instances, on—I would say on the—on the fringes of these categories that a respondent might be thinking that one particular thing that is of value to them is in one category, when, in fact, for purposes of these proceedings, it should fit in another.” Tr. at 83 (Trautman); and “Well, I think—first, I think that it's minor. I think that the program—there might be one or two exemptions, but the programs that are subject to miscategorization tend to be at the fringes and—and tend not to be things that drive substantial value in our service—in our survey. And, therefore, I think that the potential for spillover or for a change in result is—is limited.” Tr. at 107-08 (Trautman).] However, although such bias may well be reflected in the Bortz survey point estimates, no one in the proceeding has precisely quantified the amount or direction of such bias. Therefore, we cannot say to what degree such bias may skew the Bortz point estimates. Moreover, we find no basis for concluding that such bias takes the true relative value numbers outside of range of the confidence intervals for the valuation estimates produced by the Bortz survey. [“Q. And have you considered whether your results are reliable in light of the possibility that there might be miscategorization in the response? A. I have considered that, and—and while I indicated that there's certainly some potential for spillover or miscategorization of certain types of programming, I think I have confidence that—that within the bounds of the estimation parameters that we set forth in the survey, that our results provide an accurate indication of relative value.” Tr. at 107 (Trautman).]

Second, an acknowledged shortcoming of the Bortz survey valuations revolves around its handling of PTV and Canadian programming estimates. Because the Bortz methodology calls for surveying cable systems that contain at least one U.S. independent or network signal, cable systems which carry PTV-only or Canadian-only distant signals are excluded from the survey sample. The exclusion of such cable systems clearly biases the Bortz estimates downward for PTV and Canadian programming. The Bortz study seeks to excuse this bias on grounds that it is not possible to obtain an estimate of relative value where the cable system carries only one type of distant signal programming. But this explanation fails to adequately consider the view that: (1) A cable system that chooses only PTV or Canadian programming may be implicitly making a choice in favor of a 100% relative value score for such programming; (2) an explicit 100% relative value score for the Movies category (and concomitant 0% score for the remaining programming categories) is regarded as acceptable by the Bortz methodology in the case of a U.S. commercial station; and, (3) the latter occurrence—a 100% relative value score for the Movies category—would be recorded by Bortz even in the absence of PTV or Canadian distant signals from the responding cable operator's system. While the Bortz report acknowledges this bias (Bortz Report (SP Ex. 2) at 8-9) and the Settling Parties offer additional adjustments to purportedly remedy the problem (see infra at Section IV (Analysis of the Evidence)), the proffered remedies are not wholly satisfactory and, more importantly, obscure the basic difficulty that stems from asking cable operators to compare five different categories of programming with two types of distant signals. CCG PFF at ¶¶ 112,120. The Bortz survey may well be improved in this regard, either through the reformulation of the questions asked in the survey and/or by revisiting the underlying survey sample plan. Tr. at 2996-98 (Gary Ford); CCG PFF at ¶¶ 154-55. Yet, while this bias is troubling and proposed post-survey remedies based on the current record are discussed infra at Section IV (Conclusion and Award), it would be inappropriate to overstate the impact of this problem. No one in this proceeding maintains that it substantially affects more than a small portion of the total royalty pool (i.e., the combined PTV-Canadian portion) under any of the competing theories of royalty distributions advanced in this proceeding. Nor has it been shown that the Bortz survey's remaining non-PTV-Canadian estimates were thrown outside the parameters of their respective confidence intervals solely because of this problem. That is, the PTV-Canadian problem does not substantially affect any of the remaining categories in some disproportionate way.^[16]

Third, another acknowledged problem with the Bortz study flows from its handling of compensable as compared to non-compensable programming. [“* * * respondents to our survey are not informed that substantial portions of the movies and syndicated programming on Superstation WGN (the most widely carried distant signal) are not compensable in this proceeding because these programs are not broadcast by WGN on its over-the-air Chicago signal; thus the values that respondents to our survey attribute to these categories likely represent a `ceiling' in that respondents are considering all programming on WGN rather than just the compensable programming on WGN.” Bortz Report (SP Ex. 2) at 8.] The same issue affects the Devotional Claimants because of the presence of devotional programming on WGN that is also non-compensable. SP PFF at ¶ 686. (See also infra at Section V (Conclusion and Award)).

The Settling Parties offer some additional adjustments to the Bortz point estimates to address this problem. See SP PFF at ¶¶ 347-48. However, the Settling Parties do not incorporate their proposed adjustments explicitly into their proposed awards. Rather, the Settling Parties simply note their view that with respect to the Program Suppliers, their proposed award should only be regarded as a “ceiling” from which the Program Suppliers share should be reduced by some amount to reflect the disproportionate effect of the non-compensable programming issue. The Settling Parties clearly cannot precisely quantify an adjustment to the Bortz numbers for Program Suppliers because they recognize that

The specific amount of an appropriate reduction in the Program Suppliers' share would depend on how much of the value attributed by Bortz survey respondents to Program Suppliers programming categories was attributable to non-compensable programming on WGN, as to which there is no direct evidence, but it would be reasonable to expect that some portion of that value was attributed to non-compensable Program Suppliers programming.

SP PFF ¶ 348, n.802 (emphasis added). Further, with respect to the Devotional Claimants' share, the Settling Parties do not incorporate an explicit adjustment for this factor in their proposed award, being merely content to argue its relevance to adopting a prior lower award in place of its Bortz indicated share. See SP PFF at ¶¶ 686-87. Moreover, the method suggested by the Settling Parties for adjusting the Program Suppliers' share would produce no change in the Devotional Claimants' share—that is Dr. Waldfogel's comparison of implied royalty shares that resulted when all programming minutes on WGN were used in share calculations rather than just compensable programs showed no difference for the Devotional Claimants (a zero share in both cases). See SP PFF at ¶ 176 at Table 5. Thus, while we agree that some adjustment for this problem is reasonable, we find no reliably quantified adjustment on the record before us. However, because we focus on the confidence intervals for the Bortz estimates, rather than the Bortz point estimates themselves, we do not find that this issue alone so substantially affects the relative values of the programming so as to require us to discard those intervals as the best indicators in the record of the actual relative values of the programming of the larger claimants in this proceeding.

A number of other criticisms have been raised with respect to the Bortz survey by various claimants in this proceeding that suggest other shortcomings in terms of economic theory, statistical analysis or survey methodology. Yet, whether taken individually or viewed as a group, we do not find these other criticisms to undermine the general usefulness of the Bortz survey for the purpose offered. Certainly, none of the criticisms raised by the contending parties persuade us to “throw out the baby with the bathwater,” particularly when viewing the Bortz survey results in terms of the confidence intervals around the point estimates rather than strictly limited to the point estimates themselves. Instead, particularly in the case of the larger claimants such as JSC, CTV and Program Suppliers, we find the confidence intervals provided by the Bortz study the best starting point for evaluating an award, although we also recognize the need to give due consideration to the reasonability of adjustments to deal with acknowledged problems such as the undervaluation of PTV and Canadian programming. The Bortz intervals certainly mark the most strongly anchored range of relative programming values produced by the evidence in this proceeding. Still, other evidence produced in the record also helps to more fully delineate all of the boundaries of reasonableness with respect to the relative value of distant signal programming.

Another piece of evidence helpful to some degree in this regard is the Waldfogel regression analysis. Dr. Waldfogel's multiple regression analysis attempts to analyze the relationship between the total royalties paid by cable operators for the carriage of distant signals in 2004-05 and the quantity of programming minutes by programming category on those distant signals. In addition to considering the impact on the dependent variable (total royalties) of independent variables representing minutes of programming for eight category types, Dr. Waldfogel considered the following additional independent variables in his analysis: the number of subscribers to the cable system in the prior period, the number of activated channels (i.e., utilized capacity) for the cable system, average household income in the market in which the cable system was located, the number of channels originating locally, and dummy variables to indicate the presence of certain payment conditions (such whether a system pays any 3.75% fees or whether a system carries partially distant signals or whether a system imported only one DSE or whether a system imported less than one full DSE). See SP PFF at ¶ 156. Dr. Waldfogel's specification was similar in its choice of independent variables to a regression model utilized by Dr. Gregory Rosston to corroborate the Bortz survey results in the 1998-99 CARP proceeding. See Report of the Copyright Arbitration Royalty Panel to the Librarian of Congress, in Docket No. 2001-8 CARP CD 98-99 (“1998-99 CARP Report”) at 46 (October 21, 2003). Dr. Waldfogel offered a total minutes (i.e., compensable as well as non-compensable) version of his regression analysis as corroborative of the adjusted Bortz survey estimates. Tr. at 854 (Waldfogel).

Conceptually, the Waldfogel regression, with its focus on bundles of distant signals and inclusion of variables to capture both system capacity and the impact on the appetite for distant signals associated with the number of channels originating locally, may provide a richer look than the Bortz survey into factors that impact the purchasing decision of cable operators. Yet, unlike the Bortz survey, it does not purport to analyze data free from the strictures of the regulated market because the payment pools analyzed ultimately are impacted by the fee structure set in the regulated market. This raises the question of whether the Waldfogel analysis provides useful information on the key behavioral question or, alternatively, whether it merely mirrors the impact of the regulated market in its valuation. We agree with Dr. Waldfogel that the way to think about the bundle of programming that is being considered by the cable operator is to focus on its incremental value. Tr. at 890, 921, 926, and 940-41 (Waldfogel). Under that theory, Dr. Waldfogel has conceptually sought to separate the market impact of incremental signal purchasing decisions from the minimum fee issue and some other regulated fee considerations through the use of the dummy variables specified in the regression. We find, that as a result of the manner in which he has conceptualized his model, Dr. Waldfogel's regression coefficients do provide some additional useful, independent information about how cable operators may view the value of adding distant signals based on the programming mix on such signals. Although the determinants of distant signal prices in a hypothetical free market are not necessarily identified as such, some indication of what the cable operator finds valuable may be obtained by observing the way cable operators' total spending relates to the content of the bundle of distant signals purchased. That is because the cable operators are free to decide how many distant signals to purchase and, therefore, whether the addition of the content of an incremental distant signal will contribute to the net revenues of the system.

At the same time, while the Waldfogel regression analysis provides useful information, we also find that there are limits to that usefulness in corroborating the Bortz survey, largely stemming from the wide confidence intervals for the Waldfogel coefficients. Thus, the implied share of royalties calculated by Dr. Waldfogel would change substantially if the true value of the variable was at one end of the confidence interval rather than at the point estimate value used by Dr. Waldfogel in his calculations. Given the size of the standard errors around his estimates, Dr. Waldfogel concedes this imprecision. SP PFF at ¶ 184. Nevertheless, while one may question the precision of the results on this basis, it only cautions against assigning too much weight to its corroborative value.

As to the methodology employed, we find that Dr. Waldfogel employed generally reasonable methods to assure that the model's results were consistent in the face of changes in the model and that the parameter estimates did not vary in a statistically significant way across years. SP PFF at ¶¶167-68. The strident criticisms raised by Dr. Salinger and Dr. George Ford concerning the “instability” of the Waldfogel estimates over time are excessive. For example, there is no a priori reason why the two individual years examined by Dr. Salinger (by breaking the Waldfogel entire sample in two) should have exactly matching minutes coefficients. Lack of precision can result merely from the fact that all items in a population were not observed. The smaller the sample size, the fewer are the number of observations and, hence, the less precision. Then too, it is not unusual to observe the coefficients of independent variables in a model varying between two samples because all possible combinations of forces at work that result in these coefficients can seldom be fully encompassed in an efficient specification of a model. Finally, the “instability” suggested by Dr. Salinger does not extend to the signs of the coefficients—all of the minutes variables examined by Dr. Salinger continue to carry the same positive or negative sign in 2004 as they carried in 2005. Thus, any instability does not extend to the direction of the expected explanation—it is the same in both years. Dr. Salinger also raises the spectre of omitted variables with respect to the Waldfogel analysis. Tr. at 2873-74 (Salinger). But there is no evidence that the inclusion of any particular additional independent variable would improve the explanatory power of the Waldfogel regression. Nor is there any evidence in the record that the independent variables in the Waldfogel regression are correlated within an important omitted variable thereby leading to an unreliable estimate of the regression coefficients for the included variables. Without such evidence, this criticism should not be overstated because an omitted variable criticism may always be raised, since there are an almost limitless number of potential variables that may be considered for inclusion in any model of some complexity. SP PFF at ¶186.

Having carefully considered the Waldfogel analysis and various criticisms of that analysis raised by the contending parties, we find the results of this regression analysis useful in two ways—(1) to, at least in some rough way, corroborate the augmented Bortz survey results and (2) to provide an independent reasoned basis for considering movement away from the augmented Bortz point estimate for the Devotional category toward, or even beyond, either boundary of the Bortz confidence interval for that category. First, we find that, when applied to all program minutes to match the scope of the programming covered by the Bortz surveys, and when the resulting shares are compared to Bortz survey results that have been augmented to match the scope of the systems covered by the regression analysis, Dr. Waldfogel's regression analysis coefficients produce comparable share numbers for all categories except Devotional. Second, to the extent that there is imprecision in the augmented Bortz estimates, the Waldfogel regression analysis may help to identify the most imprecise point estimates and suggest a direction in which they may be adjusted further to bring them in line with what is occurring where actual decisions have been implemented. In this case, the Waldfogel analysis suggests the augmented Bortz point estimates for the Devotional category cannot be corroborated and, further, the value of the Devotional coefficient points toward a lower share for this category (consistent with our further consideration of this category, infra at Section V (Conclusion and Award)). Tr. at 922, 924 (Waldfogel).

Analysis of the Program Suppliers' Evidence

Although much less useful than the Waldfogel regression for the reasons delineated below, the Gruen survey results advocated by the Program Suppliers cannot be totally disregarded. As we have previously noted, there are factors, other than subscriber growth considerations, which may also be at work in influencing the demand for distant signal stations and that the cable operator may be best positioned to address these other considerations in a hypothetical market setting dealing with bundles of signals encompassing different programming mixes. That is why we have found that, whatever its shortcomings, the Bortz survey focuses on the appropriate buyer in the hypothetical market—i.e., the cable operator. Nevertheless, we recognize that one consideration facing cable operators, even in the subscription markets in which their cable systems may be exercising some degree of monopoly power,^[17] is the impact of programming on subscription revenues. To that extent, the preferences of subscribers as to distant signals that appear as part of the bundle of cable stations they receive may provide some relevant information, particularly if a nexus may be established between subscriber demand for such distant signals and the programming on those distant signals that drives the demand. The Gruen survey attempts to shed some light on this limited issue. Unfortunately, although not persuading us to reject the survey altogether, the various inadequacies of the Gruen approach cause us to place little weight on its findings beyond the very general notion that the highest valued categories of programming identified by the Bortz survey as a group remain the highest valued categories of programming identified by the Gruen survey and the lowest valued categories of programming identified by the Bortz survey as a group remain the lowest valued categories of programming identified by the Gruen survey.

Among the design and execution problems afflicting the Gruen survey were the lack of analysis to determine whether there was a representative sampling of demographic groups, the absence of any gender analysis, the application of valuations to the entire household rather than the survey respondent, the lack of assurance that the distant signals in question were actually viewed, and, like the Bortz survey, the failure to make an adjustment for non-compensable programming on WGN America (“WGN-A”). DPFF at ¶ 185; Tr. at 3167-68 (Ratchford); Tr. at 1915 (Gruen). Though not rendering it totally useless, the narrow focus of the study (subscriber preferences) and the difficulties largely related to the design and execution of the survey, referenced hereinabove, detract from the utility of the Gruen results, except in some very general way that confirms the broad outlines of the Bortz findings. It should be noted that many of the difficulties identified with the survey are capable of repair in the future, so that, if properly executed, it may provide some better insight into subscriber tastes to the extent such tastes play some role in cable operators' demand for distant signals as part of their offerings. For example, one issue on which the Gruen survey attempted to acquire some better information was on the definition of “live team sports”—an issue that clearly was of concern to the Judges in the context of the Bortz study. See, for example, Tr. at 81-84, 100-101 (Trautman). Still, as derived for this proceeding, we find the Gruen survey results of only slight, very general usefulness.

In addition to the Gruen survey, the Program Suppliers provided another quantitative study by Dr. George Ford on the question of relative value. Dr. Ford, in search of a market that would correspond to a hypothetical free market for the purchase and sale of the bundles of programming on distant signals, proposes a proxy for the direct observation of such a market. That proxy programming market was one that focused on local broadcast stations' purchases of exclusive broadcast rights in their own local markets.

We find that Dr. George Ford's advertising based model so far attenuated from the relevant hypothetical market as to offer no basis for reasonable estimates of the relative value of programming on distant signal stations. Moreover, questionable underlying assumptions and the methodological flaws plague the advertising based model. Finally, because we find no merit in this advertising market approach and only a slight, very general usefulness to the Gruen survey results, we reject Dr. George Ford's further suggestion of the marriage of the two approaches into a hybrid solution. See Ford WDT (PS Ex. 11) at 49-50.

Dr. George Ford's approach wholly ignores the value that may be ascribed to distant signal programming by cable operators (the buyers in the relevant hypothetical market) or even by cable subscribers (through their derived impact on demand). SP PFF at ¶¶ 423-24. Therefore, on that basis, a number of the professional economists who testified in this proceeding on the issue found the George Ford advertising based approach wanting in terms of providing any useful information. See, for example, Tr. at 229-30, 254-56 (Crandall); Tr. at 2344-46 (Crawford); Tr. at 2787-88 (Salinger); Tr. at 3060-61 (Calfee).

Furthermore, the George Ford advertising approach suffers from questionable assumptions underlying the basic tenants of his analysis or inaccurate assumptions leading to flawed adjustments of the results for particular categories of programming that do not admit of direct analysis in his approach. For example, Dr. George Ford assumes that the broadcast stations he analyzed would buy precisely the programming that was actually carried by cable systems on distant signals in 2004 and 2005. Tr. at 2199 (George Ford). But he offers no evidence to support his assertion that this is a “reasonable” assumption. Similarly, there are assumptions with respect to his determination of “prices” paid for programming on an advertising spot sales price on a “cost per thousand” or “CPM” basis that are not reasonable. As an example, he applied the CPM analysis to the Canadian programming category, even though none of the advertising data were for Canadian markets. SP PFF at ¶ 432. On the other hand, he assigns the average CPM to devotional programming even though the Devotional Claimants sell no advertising in their programs. Ford WDT (PS Ex. 11) at 35, 39 Table 6 and Johansen WDT (Devo. Ex. 2) at 7. Dr. George Ford further assumes that CTV programming did not air during prime time, resulting in no credit for Prime Time CPMs for such programming—an erroneous assumption based on the most persuasive evidence received in this proceeding. SP PFF at ¶¶ 460-61.

In short, we find that the George Ford advertising approach offers no helpful insight into the relevant hypothetical market or into the behavior of the relevant buyer in that hypothetical market—i.e., the cable operator.

In addition, even the proponent of this approach admits that, at bottom, changes in relative market values calculated between 2004 and 2005 are driven principally by the changes in viewership shares that were reported in the underlying MPAA special study. Tr. at 2286-88 (George Ford). Yet, where cable systems do not sell advertising in connection with distantly retransmitted content, a valuation dependent on ad sales tied to viewing data is untenable. Clearly, this study fails to offer a reliable means of translating viewership shares to relative value if that is its aim.

Conclusion and Award

For all of the above reasons, the Judges conclude that the Bortz intervals set the appropriate parameters for evaluating their award with respect to the JSC, CTV, and the Program Suppliers.^[18] Moreover, we do not find the Bortz estimates, either before or after various adjustments, to be so precise as to produce awards extending beyond a single decimal place. We deal with music separately as described infra at Section VI, and, therefore, divide the remainder among the JSC, CTV, Program Suppliers, Devotional Claimants, PTV and Canadian Claimants, using as our starting point the augmented Bortz survey shares as calculated by Ms. McLaughlin^[19] which includes appropriate adjustments to the PTV share at SP PFF at ¶ 317; and then, we proceed to adjust these values further to reflect the differential impact of the alternative approach we take to valuing the Canadian Claimants' and Devotional Claimants' shares. See infra at Sections IV and V. Although we provide somewhat more to the Canadian Claimants than the Bortz interval suggests for the reasons discussed infra at Section IV (Conclusion and Award), the negative effect on the remaining categories is miniscule. At the same time we provide less to the Devotional Claimants than the Bortz interval would indicate, based on the impact of the Waldfogel regression and other considerations, including the suggested direction (though difficult to quantify magnitude) of the impact of the non-compensable programming issue, as discussed supra at Section III (Analysis of the Settling Parties' Evidence) and infra at Section V (Conclusion and Award). The lower Devotional Claimants' share is divided proportionately among JSC, CTV, and PTV. However, no portion of the reduced Devotional Claimants' share is awarded to the Program Suppliers, because the latter group's Bortz share, just like that of the Devotional Claimants, includes non-compensable programming. Therefore, we decline to extend the potentially small gain from the downward adjustment of the Devotional Claimants' share to the Program Suppliers so as to recognize the differential standing of the Program Suppliers as compared to JSC, CTV and PTV with respect to non-compensable programming. The effect of this approach is to recognize and make the equivalent of a directional adjustment in the Program Suppliers' share relative to those remaining categories of programming which are largely compensable.^[20] However, the resulting positive effect on the remaining categories is small and does not place either the JSC shares or CTV shares or the share of the Program Suppliers substantially outside of its respective Bortz interval. Thus, with respect to JSC, CTV and the Program Suppliers, our award is consistent with the Bortz intervals—the strongest piece of evidence on these relative values submitted in this proceeding for our consideration—giving due consideration to the reasonability of adjustments to deal with acknowledged problems such as the undervaluation of PTV and Canadian programming.

Prior to adjusting downward for the Music Claimants' share, but after accounting for the respective shares of the Canadian Claimants and the Devotional Claimants, the shares of the Basic Fund for PTV, JSC, CTV and Program Suppliers as determined by the Judges are as follows:

	2004 (percent)	2005 (percent)
PTV	7.7	7.4
JSC	33.7	36.8
CTV	18.6	14.7
Program Suppliers	34.5	35.7

Because PTV does not participate in the 3.75% Fund, shares need only be calculated for the remaining participating claimants by adjusting the JSC, CTV, Program Suppliers, Canadian Claimants and Devotional Claimants Basic Fund shares upward to reflect PTV's non-participation. Prior to adjusting downward for the Music Claimants' share, but after accounting for the respective shares of the Canadian Claimants and the Devotional Claimants, the shares of the 3.75% Fund for PTV, JSC, CTV and Program Suppliers as determined by the Judges are as follows:

	2004 (percent)	2005 (percent)
JSC	36.7	40.0
CTV	20.3	16.0
Program Suppliers	37.6	38.9

IV. Canadian Claimants' Award

Unlike the other claimant groups, this is not the Canadian Claimants' first attempt to demonstrate to the Judges the relative marketplace value of their programming in a Phase I distribution proceeding. The Canadian Claimants litigated their distribution share vis-a-vis all the other claimants in Docket No. 2008-2 CRB CD 2000-2003, covering the royalty years 2000 through 2003. That proceeding, however, was unlike any other cable Phase I determination in the 32-plus year history of the section 111 statutory license. Instead of presenting us with competing methodologies and evidence as to the proper award for Canadian Claimants, and letting us determine relative marketplace value, the litigants restricted us, through two joint stipulations, to select one of two options: either the average of the 1998 and 1999 awards given the Canadian Claimants in the 1998-99 CARP decision, or the CARP's fee generated results—with slight modification—using 2000-03 data obtained from CDC. As described in our decision, 75 FR 26798 (May 12, 2010), we chose the latter option.

The details of the decision need not be repeated here, but there is one aspect that is worthy of reemphasis. We did not determine that the fee generation methodology used by the 1998-99 CARP, nor the modified version proposed by the Canadian Claimants, was the method to determine relative marketplace value of Canadian programming. 75 FR at 26802 (“It very well may be that there are other methods or other evidence that best represent the relative marketplace value of Canadian Claimants' programming as well as the programming of other claimant groups. Such is not the case in this proceeding, where the parties have presented us with only two choices. The Judges, therefore, do not opine as to what may be the best means of determining the relative marketplace value of Canadian Claimants' programming, or other claimant groups' programming, in future proceedings.”) (emphasis in original). No alternative methodology to determine relative marketplace value was presented. The Canadian Claimants, however, argue in this proceeding that our 2000-03 decision was an “affirmation” of the fee generation methodology to determine their award and that the decision, coupled with the 1990-92 and 1998-99 CARPs' use of fee generation for Canadian Claimants' awards, “solidifies the deference owed and the high standard that must be overcome to challenge fee generation as a viable indication of relative market value.” CCG PCL at ¶ 30. This argument is plainly wrong. We sided with the Canadian Claimants' presentation in the 2000-03 proceeding because we were given only two choices and the other claimant groups failed to demonstrate that “the fee generation approach is so arbitrary, so meritless that it is without probative value with respect to determining the Canadian Claimants' royalty share.” 75 FR at 26804. Fee generation, as used in the 2000-03, 1998-99, or 1990-92 proceedings is not given overarching weight in this proceeding. In order for it to be adopted in this proceeding, the Canadian Claimants must demonstrate that it is the best means of determining Canadian programming's relative marketplace value.

Analysis of the Evidence

As they have done in prior proceedings, the Canadian Claimants urge us to determine their award on the basis of a fee generation methodology they have developed. We discussed in detail in the 2000-03 proceeding the origin and operation of fee generation, and how it was applied by the 1998-99 CARP. See 75 FR at 26800-03. Using full-year data obtained from CDC, the Canadian Claimants demonstrated that distant Canadian broadcast signals generated 4.15% of the total Basic Fund royalty fees paid by U.S. cable systems in 2004 and 4.36% of the fees paid for 2005. For the 3.75% Fund, Canadian distant signals generated 3.50% of the 2004 royalties and 3.23% of the 2005 royalties.

In years past, the Canadian Claimants' fee generation approach would stop at this juncture. However, beginning with the 2000-03 proceeding, the Canadian Claimants performed additional computations to address two “problem” facets of the section 111 royalty payment scheme. The first difficulty occurs in analyzing royalties paid by cable operators in the Basic Fund. Under the statutory scheme, royalties are paid on a sliding scale of percentages of gross receipts obtained by cable systems for the privilege of retransmitting broadcast stations. Coupled with an additional factor that cable systems that carry no distant signals pay the same amount as if they had carried one distant signal (the so-called “minimum fee”), it is not possible to determine precisely at what royalty rate the cable system paid for the Canadian signal (or any other distant signal, for that matter). To attempt to address this, Jonda Martin, president of CDC, performed what she described as a “Min/Max” analysis, whereby she calculated royalties from cable systems as if they had paid for the Canadian distant signal at the first DSE value, and as if they had paid for it at the last DSE value. Martin WRT (CCG Ex. CDN-R-1) at 4. The purpose of this analysis was an attempt to demonstrate that the Canadian Claimants' selection of the mid-point of these royalties as actual royalties paid was a reasonable exercise. Calfee WRT (CCG Ex. CDN-R-3) Appendix B at 8.

A similar exercise was performed for the 3.75% Fund. Under the section 111 scheme, one cannot determine which signals are paid for at the 3.75% “nonpermitted” rate when more than one carried distant signal could have been identified as a Basic Fund “permitted” signal. Ms. Martin calculated cable system royalties as if cable systems paid for Canadian distant signals at the 3.75% “nonpermitted” rate, and at the basic “permitted” rate, once again in an effort to demonstrate that the selection of the mid-point for 3.75% Fund royalties paid was reasonable. Martin WRT (CCG CDN-R-1) at 5, Table 3.

Armed with Basic and 3.75% Fund fee generated royalties for 2004 and 2005, the Canadian Claimants next sought to provide the division of royalties among the program categories contained on Canadian distant signals. This was done, as it had been in the prior proceeding, by Drs. Gary Ford and Debra Ringold, who conducted a constant sum survey of large cable systems carrying distant Canadian signals in an effort to determine what value they attached to the Canadian programming (as opposed to JSC and Program Supplier programming, the only other two types of programming appearing on Canadian distant signals) contained on the Canadian distant signals. The results, presented by Dr. Ringold, showed a purported value of 59.94% for 2004 and 60.37% for 2005. Thus, of the fees generated by Canadian signals for 2004 and 2005, 59.94% and 60.37%, respectively, were attributable to Canadian programming.

The Canadian Claimants' calculations do not, however, end there. This is because the Canadian Claimants urge us to follow the distribution methodology adopted by the 1998-99 CARP for parties whose royalties were determined by means other than using their Bortz survey results. This 16-step process results in a requested award to Canadian Claimants of 2.365% of the Basic Fund and 1.586% of the 3.75% Fund for 2004,^[21] and 2.499% of the Basic Fund and 1.308% of the 3.75% Fund for 2005. CCG PFF & PCL Appendix A at 14. In the event that the Judges do not follow the 1998-99 CARP's distribution methodology, Canadian Claimants urge awards of 2.515% of the Basic Fund and 1.656% of the 3.75% Fund for 2004, and 2.665% of the Basic Fund and 1.365% of the 3.75% Fund for 2005. Id. at Appendix B, 3-4.

The Settling Parties contend that they have made significant improvements from prior proceedings to the results yielded by the Bortz survey and urge adoption of particular “augmented” point estimates for Canadian Claimants. First, they submit that the survey itself has been improved by increasing the number of large cable systems carrying a Canadian signal to 11 (18% of the total) in the 2004 Bortz survey and 13 (25.5% of the total) in the 2005 survey. SP PFF at ¶ 326. Second, to account for the exclusion from the survey of cable systems that carried only Canadian and/or PTV distant signals,^[22] they offer the testimony of economist Linda McLaughlin, who purports to mathematically compute the values the 2004 and 2005 Bortz surveys would likely have found had they not excluded these systems. These “augmented” Bortz results produce a Canadian Claimants' royalty share of 0.5% for 2004 and a range of 1.5% to 1.8% for 2005. McLaughlin WDT (SP Ex. 6) at 11, Chart 4. Third, the Settling Parties accept the observation of Dr. Gary Ford, a Canadian Claimants witness, that one large cable system which carried a distant Canadian signal, Comcast of Washington IV, was improperly excluded from the 2004 Bortz results due to a clerical error. SP PFF at ¶¶ 330-31. Finally, the Settling Parties accept the results of the Ford/Ringold constant sum surveys, whereby Dr. Ringold testified that 59.94% of 2004 Canadian signals and 60.37% for 2005 were attributable to Canadian programming.^[23]

The Settling Parties conclude that the Canadian Claimants' award should be determined by multiplying their augmented Bortz survey results for 2004 and 2005 by the Ford/Ringold constant sum survey results for Canadian programming. This yields a distribution of 1.2% for both the 2004 Basic and 3.75% Funds, and 1.0% of the Basic Fund and 1.1% of the 3.75% Fund for 2005.

The Waldfogel regression analysis, discussed supra, yielded an estimated royalty share of 2.92% for Canadian Claimants. SP PFF at ¶ 179. Not surprisingly, the Settling Parties do not advocate use of the Waldfogel number as the Canadian Claimants' award. Nevertheless, in Dr. Waldfogel's view, his regression share compares favorably to the Settling Parties' augmented Bortz shares for Canadian Claimants, more so when the Dr. Gary Ford adjustment to the augmented results is included. SP PFF at ¶¶ 180-81.

The Gruen subscriber survey yielded 0.8% for 2004 and 1.8% for 2005, respectively. Gruen WDT (PS Ex. 8) at 23, Table 3. The survey did not distinguish between the Basic Fund or the 3.75% Fund. Program Suppliers dispute use of the Ford/Ringold constant sum survey as the means for determining the division of royalties among the categories of programming contained on Canadian distant broadcast signals, but do not offer an independent basis for making such distinctions. See, PS Disputed CCG PFF & CCL at ¶¶ 82-83.

Conclusion and Award

Unburdened by the attendant limitations of the last proceeding, the Judges are free to determine distribution awards for 2004 and 2005 that best reflect the relative marketplace value of Canadian broadcast programming retransmitted by cable systems. We do not rely solely upon fee generation in general nor the specific fee generation methodology offered by the Canadian Claimants.

Our declination from use of fee generation to determine relative marketplace value stems from the Canadian Claimants' inability to demonstrate that the relationship between royalties generated by the section 111 license for Canadian signals and the overall hypothetical marketplace value of programming in this proceeding is, in the words of the Canadian Claimants' own witness, Dr. Calfee, more than “rough,” “far from perfect,” and “crude.” ^[24] The wobbly relationship between the two does not mean, as the other parties in this proceeding would have it, that we are precluded from utilizing the evidence of fee generation in shaping our award. 75 FR 26798, 26805 (May 12, 2010). What it does mean, and what we were unable to consider in the prior proceeding, is that other evidence of relative marketplace value presented by the parties should be considered. See, id. at 26820-03 (Judges' discussion of the checkered history of acceptance of fee generation in section 111 distribution proceedings).

The augmented Bortz data presented by the Settling Parties attempts to correct for prior primary criticisms; in sum, that it is does not sufficiently measure the particular circumstances of smaller claimants such as Canadian Claimants. Ms. McLaughlin's efforts to correct for cable systems excluded from the survey because they only carry a distant Canadian signal do somewhat ameliorate the under-representation of Canadian signals in the overall survey results.^[25] But, consistent with our earlier expressed concerns about the Bortz survey, there are still not enough cable systems carrying distant Canadian signals among the respondents. As a result, small adjustments to the data result in proportionately enormous increases in distribution shares. For example, when the omitted Seattle, Washington, cable system data is included in the augmented 2004 results, it produces more than a three-fold increase in the distribution share. Whether the survey sample needs to be tripled in size to be accurate, as Dr. Gary Ford suggests, is debatable, but improved response rates are necessary before the survey can be considered the best marker of relative marketplace value.

We conclude that the augmented Bortz results, with the Dr. Gary Ford 2004 adjustment and the application of the Ford/Ringold survey, understate the value of Canadian programming and, therefore, represent the floor for establishing the Canadian Claimants' award. Our determination on this point is bolstered by the results of the Waldfogel regression analysis, which values Canadian programming at a higher level for both years and, to a lesser extent, the Gruen survey which yields an appreciably higher result for 2005.

Having determined the floor of the award, we turn to the weight that should be accorded the fee generation approach offered by the Canadian Claimants. We focus our attention on a “straight” fee generation approach, described in Appendix B of the Canadian Claimants' proposed findings, and not the fee generation methodology employed by the 1998-99 CARP. The CARP's approach applied to an evidentiary record, and a relationship of the parties, considerably different from this proceeding, and therefore is neither controlling nor useful here.

The Canadian Claimants' fee generation numbers for the Basic Fund are 2.515% for 2004 and 2.665% for 2005, and for the 3.75% Fund are 1.656% for 2004 and 1.365% for 2005. CCG PFF & PCL at Appendix B. We discussed above that fee generation is not persuasive as the best method for determining relative marketplace value because of the Canadian Claimants' failure to firmly link the relationship between section 111 royalties to that value. The question is whether fee generation tends to overstate or understate the value. We believe the answer is the former. The Canadian Claimants applied their fee generation methodology to royalties collected from all large cable systems in the United States, even though many, if not most, of those systems are not permitted by the section 111 license to retransmit Canadian broadcast stations. The inclusion of all royalties, rather than just those from cable operators in the “Canadian zone,” inflates Canadian Claimants' numbers. Therefore, the Judges determine that the Canadian Claimants' fee generation numbers represent the ceiling for their award.^[26]

Having determined a floor and a ceiling for the Canadian Claimants' award, the “zone of reasonableness” is framed. National Ass'n of Broadcasters v. Librarian of Congress, 146 F.3d 907, 918-19 (DC Cir.1998) (citing National Ass'n of Broadcasters v. Copyright Royalty Tribunal, 772 F.2d 922, 926 (DC Cir. 1985)). The Canadian Claimants' final awards are as follows (prior to accounting for the Music Claimants' share):

Year	Basic fund	3.75% fund (percent)	Syndex fund (percent)
2004	2.0	1.5	0
2005	2.0	1.2	0

V. Devotional Claimants' Award

The Devotional Claimants have not participated in a Phase I distribution proceeding since the 1990-92 CARP proceeding. DPCL at ¶ 102. The Devotional Claimants reached a settlement with the other Phase I parties regarding their share to the 1998-99 cable royalties and therefore did not participate in the 1998-99 CARP proceeding. See Tr. at 1368 (Opening Statement); SP PFF at p. 29 (Introduction and Summary).

Analysis of the Evidence

Devotional Claimants have consistently supported the JSC's cable operator valuations of the program categories throughout the history of their participation in these distribution proceedings. Id. Their position in this proceeding is no different: In their view, the Bortz survey continues to represent the best evidence of the relative marketplace value of the various program categories. 5/10/10 Tr. at 35 (Closing Argument). Accordingly, they argue that they are entitled to the shares afforded them by the 2004 and 2005 Bortz surveys and thus are seeking an award of 7% of the Basic Fund for each of 2004 and 2005 and 7.3% of the 3.75% Fund for each year.^[27] DPCL at ¶¶ 106-107.

Devotional Claimants argue that such an increase is warranted for several reasons. First, they note that previous awards were based primarily on the Nielsen data, not the Bortz survey. 5/10/10 Tr. at 43 (Closing Argument). If the Judges find the Bortz survey acceptable in this proceeding, then their shares should increase. Second, since the 1990-92 proceeding, their average shares under the Bortz surveys have nearly doubled from an average of 3.9 in the 1990-92 surveys to an average of 7.2 in 2004-2005. DPCL at ¶ 104. According to Devotional Claimants, such an increase constitutes “changed circumstances” thus requiring “a significant repositioning” of the Devotional Claimants' relative shares of the 2004-2005 cable royalty funds. DPFF at ¶ 17; see also DPCL at ¶ 103. Third, the Devotional Claimants assert that their 2004-2005 Bortz Survey results have been corroborated by Dr. Gruen's cable subscriber survey, which was introduced for the first time in this proceeding, and attributed a share to the Devotional Claimants of 7.3% in 2004 and 8.19% in 2005. DPFF at ¶ 190; see also Tr. at 2787 (Salinger).

Fourth, Devotional Claimants attribute the dramatic increase in their Bortz shares since the 1990-92 proceeding in part to an evolution in devotional programming over time, 5/10/10 Tr. at 44-45 (Closing Argument), and an increase in viewer avidity and loyalty. Brown WDT (Devo. Ex. 3) at 8. The evolution of programming consists of new additions in children's programming, e.g., cartoons, animated programming, and a greater emphasis on counseling, healing, and interpersonal relationships. DPFF at ¶ 146.

The increase in loyalty and avidity for devotional programming is premised on the testimony of Dr. William Brown. Brown WDT (Devo. Ex. 3) at 8-18; Tr. at 1405-1411 (Brown) (Dr. Brown identified eight factors that, in his view, demonstrated increased value to devotional programming: (1) Desire to avoid increased sex and violence on television; (2) increased desire for more moral and spiritual content on television; (3) hostility of intellectual elite toward religious faith, i.e., “culture wars”—more progressive views that man can answer all problems versus a more traditional value of looking to God for answers; (4) distrust of the news media; (5) desire for political awareness; (6) technology growth and competition; (7) threat of radical Islam and the wars in Afghanistan and Iraq; and (8) important demographic changes resulting in greater ethnic diversity).

The Settling Parties argue that Devotional Claimants are not entitled to receive their Bortz shares and should instead receive the same awards they received in the 1990-92 proceeding, namely, 1.19% of the Basic Fund and 0.91% of the 3.75% Fund for each of the 1990-92 cable royalties. SP PFF at ¶ 673. They contend that as in the 1990-92 proceeding, Devotional Claimants have not provided evidence of any price at which Devotional Claimants sold their programming nor did they provide evidence constituting a change in circumstances since the 1990-92 proceeding. Id. In other words, according to the Settling Parties, Devotional Claimants have not met their burden by failing to “provide any evidence in this proceeding about what their share of distant signal programming should be.” 5/10/10 Tr. at 109, 111 (Closing Argument).

The Settling Parties also point to the large amount of non-compensable devotional programming contained on WGN-A, which they view as inappropriately increasing the Bortz survey responses. In their view, these inflated results were confirmed by the results of the Waldfogel regression analysis, see supra at Section III (Analysis of the Settling Parties' Evidence), which produced a zero value for devotional programming, thereby further justifying Devotional Claimants' receipt of the same award as received in the 1990-92 proceeding.

The Canadian Claimants propose a method for addressing the non-compensable programming issue:

2004: 7.8% (Bortz) × 60% (WGN carried) × 10.1% (WGN compensable) + 7.8% (Bortz) × 40% (non-WGN) × 100% (non-WGN compensable) = 3.593%

2005: 6.6% (Bortz) × 60% (WGN carried) × 9.8% (WGN compensable) + 6.6% (Bortz) × 40% (non-WGN) × 100% = 3.028%.

CCG PCL at ¶ 128.

Although Canadian Claimants argue that 3.593% and 3.028% most likely should be the upper boundary of Devotional Claimants' awards, they concede that Devotional Claimants “may be entitled to more in this proceeding than as prior proceedings based on their higher results on the Bortz survey compared to 1998 and 1999.” Id. at ¶ 130.

Conclusion and Award

The Devotional Claimants seek 7% of the Basic Fund and 7.3% of the 3.75% Fund for 2004 and 2005. For the reasons stated below, we decline to give the Devotional Claimants their Bortz point estimate results and award them 3.5% of the Basic Fund and 3.8% of the 3.75% Fund for the period.

As discussed previously, we direct our consideration to the Bortz survey confidence intervals, rather than the point estimates offered by the Devotional Claimants. This results in a range of 7.1% to 8.5% for 2004 and a range of 5.8% to 7.4% for 2005. See SP PFF at ¶ 132. However, there are two factors that warrant a downward adjustment in the relative value of devotional programming: the matter of the amount and significance of non-compensable devotional programming contained on WGN-A during the period, and the results of the Waldfogel regression analysis.

WGN-A was the most widely carried distant signal by cable systems during 2004 and 2005, SP PFF at ¶ 343, and a full 90% of the devotional programming contained on the WGN-A signal was non-compensable under the section 111 license. Ducey WDT at 6; Tr. at 565 (Ducey). A decided shortcoming of the Bortz survey was its handling of compensable programming versus non-compensable programming since the survey respondents were not made aware of the issue and therefore could not confine their responses to only compensable programming. Although none of the witnesses were able to quantify the likely impact of non-compensable programming on the Bortz results, Mr. Trautman and Ms. McLaughlin each recognized that an adjustment was necessary. Tr. at 195 (Trautman); see also, Tr. at 170 (Trautman) (cable operators “don't make any such adjustment [for non-compensable programming] in the responses * * * and that some adjustment needs to be made in these proceedings to account for that fact”); Tr. at 474-76 (McLaughlin) (non-compensable programming resulted in “extra value” to Devotional Claimants that “you would want to take out”). The Judges determine that, given the widespread carriage of WGN-A among the cable systems measured by Bortz, and the predominant volume of non-compensable devotional programming contained on that signal,^[28] the Bortz results likely significantly overstate the relative value of devotional programming during the 2004-05 period.

The likelihood of overstatement is confirmed by the results of the Waldfogel regression analysis. As noted previously, Dr. Waldfogel's regression coefficients do provide some additional useful, independent information about how cable operators may view the value of adding distant signals based on the programming mix on such signals. In the case of devotional programming, his results trend in the extreme, suggesting a zero value. See supra at Section III (Analysis of the Settling Parties' Evidence). While this is certainly not the case, at a minimum, his results suggest that the Bortz results are too high and therefore require a downward adjustment.

None of the testimony offered by Devotional Claimants supports sustaining the Bortz survey point estimates, nor counsels against a downward adjustment. The testimony offered regarding growth of devotional programming and avidity and loyalty of devotional viewers was anecdotal in nature and comprised largely of unsupported opinion. See, Digital Performance Right in Sound Recordings and Ephemeral Recordings, Final rule and order, in Docket No. 2005-1 CRB DTRA (“Webcasting II”), 72 FR 24084, 24095 n.30 (May 1, 2007) (anecdotal testimony not persuasive). Devotional Claimants did not offer any survey results or data supporting these contentions, and we do not have sufficient evidence upon which to base any conclusions or adjustments.

After taking into account the adjustments just discussed, we determine that Devotional Claimants are entitled to the following awards (prior to accounting for the Music Claimants' share):

Year	Basic Fund	3.75%
2004	3.5	3.8
2005	3.5	3.8

VI. Music Claimants' Award

We now turn to Music Claimants. Music is not a stand-alone category but rather permeates all other program categories. During closing arguments the Judges posed the question whether the Music Claimants' share should be taken off of the top and the Claimants appear in general agreement that it should. 5/10/10 Tr. at 5-6, 31, 91, and 145-46 (Closing Argument).

Analysis of the Settling Parties' Evidence

To develop a benchmark for assessing the relative value of music in the distant signal marketplace for 2004 and 2005, the Settling Parties presented William P. Zarakas, an economist.^[29] Mr. Zarakas developed a music ratio conceptually similar to the ratio proffered by JSC witness Dr. George Schink in the 1998-99 CARP proceeding.^[30] Under the Schink ratio, music license fees were divided by the sum of music license fees and broadcast rights payments (i.e., total payments made by the stations and networks in the over-the-air broadcast market for the rights to broadcast the programs aired on such stations). SP PFF at ¶¶ 350 and 374. The Schink ratio was not designed specifically to measure music's value in the distant signal market, the relevant market in this proceeding, but rather was based on industry-wide television broadcast licensing fees and rights payments in the over-the-air broadcast market. Id. at ¶ 375. Indeed, the Schink ratio included music license fees and broadcast rights payments by the “Big 3” networks (ABC, CBS and NBC), even though that programming is not compensable under section 111 of the Copyright Act. Moreover, no weighting was applied to the Schink ratio in the 1998-99 CARP proceeding to account for the difference between the mix of station types retransmitted on distant signals and the stations that generally make up the entire broadcast television market. Id.

Although Mr. Zarakas determined that the Schink ratio was a reasonable method to assess the relative value of music, he concluded that the ratio inputs would need to be changed to enable the ratio to provide a more useful benchmark for assessing the relative market value of music in this proceeding. Id. at ¶¶ 375-376. In particular, Mr. Zarakas excluded from his ratio music license fees and broadcast rights payments for Big 3 network programming, which are not compensable under section 111 of the Copyright Act. Moreover, he concluded that “the market for retransmitted distant signals by cable system operators differs from the local broadcast television market in terms of the mix of programming transmitted.” SP PFF at ¶ 391. Therefore, he weighted the music ratio that he developed using distant signal subscriber instances for each different category of television stations in an effort to reflect the relative importance of the various stations actually carried by cable system operators and received by subscribers as distant signals during 2004 and 2005. Id. at ¶ 376.

To form the numerator of his ratio, Mr. Zarakas used television “blanket license” fee data that the PROs provided.^[31] These fees were agreed to by each PRO and the Television Music License Committee (“TMLC”) (an industry committee of local television broadcasters) for all local stations in the broadcast market for their local (i.e., non-Big 3 network) programming.^[32] SP PFF at ¶ 369 and 377. The Settling Parties contend that the blanket license fees are the most comprehensive, accurate data in the record and are the only data that values all music use in local broadcast markets. Id. at ¶ 377. The Settling Parties further contend that, in the absence of the compulsory license, cable systems would most likely acquire blanket licenses from the PROs for the music that they represent in the open market, as the TMLC and the Univision network do currently. Id. at ¶ 381. Mr. Zarakas included local broadcast station blanket PRO license fees of $195.5 million in 2004 and $186 million in 2005. To those totals he added the blanket license fees that Univision paid, which include license fees for local and nonlocal programming,^[33] to sum $200.8 million for 2004 and $191.7 million for 2005. These sums constituted the numerator in the music ratio and one component of the denominator. Id. at ¶ 383.

As discussed above, Mr. Zarakas used blanket license fees negotiated between the PROs and the TMLC as the numerator for his music ratio. We agree with the Settling Parties that the blanket license fees provide a useful starting point in determining the relative marketplace value of music in the over-the-air market. See also infra at Section VI (Analysis of the Program Suppliers' Evidence). As such, we find that the use of blanket license fees both in the numerator of the music ratio and as the first component of the denominator is not misplaced. The other components of the denominator, discussed below, are more problematic.^[34]

The second component of the Zarakas denominator seeks to estimate broadcast rights payments. Mr. Zarakas divides these payments into three categories: (1) Payments local television stations make for non-network programming; (2) payments made for non-Big 3 network programming; and (3) payments to local stations for programs they produce themselves. Id. at ¶ 385.

Mr. Zarakas extrapolated payments local television stations make for non-network programming from the Television Financial Report, which NAB and Broadcast Cable Financial Management Association ^[35] publish annually (known as the “NAB Survey”).^[36] The NAB Survey provides an annual average of television station expenditures for broadcast rights. Zarakas WDT (SP Ex. 27) at ¶ 36. Mr. Zarakas then calculated the total number of stations that were operating in the U.S. in 2004 (1,372) and 2005 (1,371). He then excluded “several” of these stations for 2004 and 2005 because he determined that those stations were unlikely to have been included in the NAB Survey, largely because they were too small. He then multiplied the remaining number of stations (1,187 for 2004 and 1,192 for 2005) by the average annual expenditures from the NAB Survey to estimate the total broadcast rights expense for this component for 2004 and 2005 ($2.015 billion and $2.029 respectively). Zarakas WDT (SP Ex. 27) at ¶¶ 36-37.

However, the Settling Parties provided no evidence that would bolster the accuracy of the NAB Survey numbers (e.g., what was the sample size of the respondent group and what methodology was used in the survey to ensure that it accurately represented the respondents' expenditures). Moreover, Mr. Zarakas' methodology for narrowing the number of stations to which the average expenditure number was applied appears on less firm footing. These weaknesses, which could have been easily remedied, diminish the weight we ascribe to Mr. Zarakas' ratio.

Although network programming on the Big 3 networks is not compensable under section 111 of the Copyright Act, network programming on FOX, WB, UPN and other non-Big 3 networks is compensable. The NAB Survey referenced above, however, does not estimate such programming expenditures. As a proxy, Mr. Zarakas used total programming expenses data from SNL Kagan, which the Settling Parties represent is a “recognized source of economic information for the television broadcast industry.” SP PFF at ¶ 388.^[37] SNL Kagan data did not separate broadcast rights payments from other categories of program expenses, and Mr. Zarakas did not believe he had a principled basis for determining the percentage of the programming expenses that were attributable to broadcast rights expenses. Therefore, he included the entire amount of program expenses in this component of the denominator. Zarakas WDT (SP Ex. 27) at ¶ 40. The totals were $3.254 billion for 2004 and $3.550 billion for 2005. Id. at ¶ 39, Table 4.

While Mr. Zarakas' decision to include all program expenses in this component of the denominator may have been a conservative approach on his part, this limitation diminishes the precision of the measurement. Another drawback of the SNL Kagan data: It is derived from a different source than the one that conducted the NAB Survey. Using multiple data sources in the same denominator creates a potential risk of methodological inconsistency, a weakness that was made worse by the fact that the Settling Parties did not present witnesses from either SNL Kagan or those that conducted the NAB Survey, which would have allowed an on-the-record examination of their respective methodologies so that the claimants could probe their comparability.

Mr. Zarakas was unable to use market transactions to value locally produced programming, such as local news and locally produced public affairs shows. According to Mr. Zarakas, such stations do not typically sell the broadcast rights or otherwise measure the equivalent value of such rights. Zarakas WDT (SP Ex. 27) at ¶ 41. Therefore, he estimated the number by relying on the CARP's determination of the various claimants' shares of the Basic Fund in the 1998-99 cable royalty distribution proceeding. Zarakas WDT (SP Ex. 27) at ¶ 42. In particular, he calculated the relative value that the CARP assigned to locally produced programming (using CTV's share in 1998 and 1999 as a proxy) compared to the combined local commercial television station non-network programming and non-Big 3 network programming (using the combined JSC, Program Suppliers, and Devotional Claimants' shares in 1998 and 1999 as a proxy). He then took this relative value from the 1998-99 proceeding and applied it to the relative value in this proceeding of broadcast rights in locally produced programming compared to broadcast rights payments in these other types of programming. Id. This multiplier (0.185, Zarakas WDT (SP Ex. 27) at ¶ 43) was used to derive an average factor (1.185, Zarakas WDT (SP Ex. 27) at ¶ 46 and Table 6), which Mr. Zarakas then used to develop an estimated value of broadcast rights for locally produced programming in this proceeding (approximately $975 million for 2004 and $1.03 billion for 2005). Id. at ¶ 46 and Table 7.^[38]

Use of the various claimants' shares from the 1998-99 proceeding seems to be a haphazard attempt to guesstimate a material component of the denominator of the music ratio. Such ad hoc extrapolation diminishes our confidence in the Zarakas ratio.

When all components of the denominator were combined, Mr. Zarakas determined that the estimated value of broadcast rights payments were approximately $6.2 billion in 2004 and $6.6 billion in 2005. Zarakas WDT (SP Ex. 27) at ¶ 47. He then added the blanket music license fees to each of these totals to derive a grand total denominator of $6,445.4 billion for 2004 and $6,803.6 billion for 2005. Id. at ¶ 49 and Table 8. Dividing the numerator by the denominator yields a relative market value of music of 3.1% for 2004 and 2.8% for 2005. Zarakas WDT (SP Ex. 27) at ¶ 60.

The unadjusted Zarakas percentages attempt to estimate the relative value of music in the over-the-air market. Mr. Zarakas states, however, that the unadjusted percentages are “misleading in the distant signal market because the composition of signals is different in the distant signal market compared to the over-the-air market.” Tr. at 1158 (Zarakas). Mr. Zarakas contends that “the relative value of music in the distant signal market should take into account differences in the programming mix between the local and distant signal markets.” Zarakas WDT (SP Ex. 27) at ¶ 51.^[39] As a result, Mr. Zarakas adjusts his over-the-air percentages in an effort to make them more comparable to the target distant signal market by accounting for the relative number of distant subscribers associated with three categories of television stations (i.e., Big 3 networks, non-Big 3 networks, and independent stations). Zarakas WDT (SP Ex. 27) at ¶ 54-57 and Tables 9-12.^[40] Applying this adjustment, Mr. Zarakas concludes that the relative value of music was 5.2% (from the unadjusted 3.1%) in 2004 and 4.6% (from the unadjusted 2.8%) in 2005. Zarakas WDT (SP Ex. 27) at ¶ 61. See also SP PFF at ¶ 392 and Table 12. In other words, the adjusted percentages represent increases of approximately 67.7% and 64.3% over the respective unadjusted percentages. Under either the adjusted or the unadjusted numbers, Mr. Zarakas concluded that the relative market share of music declined from 2004 to 2005 (a decline of approximately 9.7% for the unadjusted percentages compared to a decline of approximately 11.5% for the adjusted percentages).

The over-the-air market and the distant signal market may well differ in ways that could impact the relative values of music across those markets. On the record before us, however, it is not clear why those differences, if any, would translate into a variation in the market value of music of the order that Mr. Zarakas contends. In other words, given that music permeates all other programming categories, what factors make the use of music over 60% more valuable relative to other programming categories in the distant signal market than it is in the over-the-air market? The Settling Parties offer little justification for Mr. Zarakas' comparability adjustment, noting only that “the market for retransmitted distant signals by cable system operators differs from the local broadcast television market in terms of the mix of programming transmitted.” SP PFF at ¶ 391, quoting Zarakas WDT (SP Ex. 27) at 25.^[41] We do not mean to suggest that a comparability adjustment is unnecessary. Nor do we suggest that an adjustment that uses subscriber instances should be dismissed out of hand. We find, however, that the Settling Parties did not fully establish the differences in valuation that the comparability adjustment is meant to address or the efficacy of the specific adjustment that Mr. Zarakas proposes. Therefore, we cannot place full weight on Mr. Zarakas' comparability adjustment.

Analysis of the Program Suppliers' Evidence

Program Suppliers retained John R. Woodbury, PhD, a consultant, as an expert to rebut Mr. Zarakas' presentation. Dr. Woodbury questioned Mr. Zarakas' use of blanket license fees as a means for estimating the relative share of music, stating that “there is no reason to believe that the use of blanket license fees is in fact a more accurate and reliable measure of the actual music rights payments made by broadcast stations than the payments actually recorded by the PROs.” Woodbury WRT (PS Ex. 14) at ¶ 12. He noted that “to the extent that stations opt for a direct license rather than the blanket license, the payments made by the broadcast stations in the aggregate to the PROs will be less than the negotiated fee amounts used by Mr. Zarakas.” Id. at ¶ 14. Dr. Woodbury opined that “[a]t best, those blanket license fees are an upper bound on the actual payments made by broadcast stations * * *” Id. at ¶ 13. However, while the blanket fee data does not include fees that a copyright owner receives when it enters into a direct license with a broadcaster, the Settling Parties' evidence suggests that the difference between the negotiated blanket fee and the actual license fees paid, including direct license fees, is not significant. SP PFF at ¶ 382.

Dr. Woodbury also questioned the Zarakas comparability adjustment discussed above. He contended that Mr. Zarakas offered no justification for using subscriber instances to weigh station types. Tr. at 3298 (Woodbury) and Woodbury WRT (PS Ex. 14) at ¶ 25. He surmised that Mr. Zarakas did so because he assumed that the number of music performances on a distant signal is related to the number of subscribers that have access to that signal. Dr. Woodbury stated that there is no reason to believe that this is the case. Id. Dr. Woodbury noted that

it seems reasonable to think that subscriber viewership [a method that the TMLC uses to allocate blanket license fees across stations] might be related to the number of music performances of a particular show on a distant signal, but that has no relationship—no obvious relationship to the fraction of subscriber instances accounted for by a particular distant signal on a particular cable system * * *. The viewership of any distant signal on a cable system can differ for lots of reasons, even if the two systems have the same number of subscribers.

Tr. at 3299 (Woodbury).^[42]

Dr. Woodbury contended that a better approach would have been to use the actual music rights payments that ASCAP and BMI received from broadcast stations and networks (i.e., over-the-air market participants) for 2004 and 2005 and divide those numbers by the total rights payments, which the Bureau of Census reported for 2004 ($11,710 million) and 2005 ($12,036).^[43] Dr. Woodbury stated that for 2004 the total music rights payments received by the PROs were approximately $239 million for 2004 and $234 million for 2005. Dividing these numbers by the Census data yields 2.04% for 2004 and 1.94% for 2005. Woodbury WRT (PS Ex. 14) at ¶ 22. Dr. Woodbury conceded that “[t]he approach that I have adopted * * * may to some extent understate the actual overall percentage, but my approach is tied to the underlying reality of what stations actually pay for music rights.” Id. at ¶ 23. Indeed, Dr. Woodbury conceded that he excluded direct license fees from his numerator but not from his denominator, which had the effect of understating his music rights ratios. Tr. at 3335 (Woodbury). Moreover, Dr. Woodbury conceded that the Census data he used to compile his ratios were outdated in a way that resulted in his ratios being understated compared to their value when using the revised Census data. Id. at 3327-28. He also conceded that his numerator included payments by commercial stations but that his denominator included payments by both commercial and non-commercial stations, which could have lowered his ratios. Id. at 3344-45.

Dr. Woodbury acknowledged that there are differences between the over-the-air market and the distant signal market, but he made no effort to adjust for those differences. Id. at 3347-48. Given the acknowledged flaws in Dr. Woodbury's approach, we place substantially less weight on his proposed estimates of the Music Claimants' shares compared to the weight ascribed to the Zarakas methodology. However, even the latter cannot be fully adopted by the Judges as offered.

Conclusion and Award

Despite the caveats discussed hereinabove, we find that the Zarakas ratio is useful in identifying the ceiling for a zone of reasonableness for determining the relative market value of music in the distant signal market for 2004 and 2005. This ceiling must lie below Zarakas' 5.2% adjusted ratio for 2004 and his 4.6% adjusted ratio for 2005, due to the previously noted weaknesses with respect to his ratios and his comparability adjustment. We are persuaded that the Zarakas adjusted ratios may more likely somewhat overstate rather than understate the relative value of music. On the other hand, the floor for the zone of reasonableness clearly must exceed by some substantial margin the 2.04% that Dr. Woodbury offered for 2004 and the 1.94% he calculated for 2005, in recognition of the flaws in the methodology and data on which he relied and his own admission that his ratios likely understated the relative value of music.

Within this zone of reasonableness as established by the record, we are persuaded by the greater weight we accord the Zarakas adjusted ratios as compared to the Woodbury alternative ratios, that the relative value of music lies closer to the former than the latter. That is, a value close to the upper boundary is more strongly supported than one close to the lower boundary. We find that value is 4% for 2004. We are comforted as to the reasonableness of this value in light of its congruence with the share received by the Music Claimants in their last litigated award.^[44]

We further find that the relative value of music for 2005 is 3.6%. That is because the zone of reasonableness has been shifted somewhat below the 2004 range by the evidence as discussed hereinabove. The major contending parties recognize this shift in their alternative proposals. For example, the Settling Parties' proposed shares for 2005 concede that the relative market value of music decreased from 2004 to 2005. This movement is evident both in the unadjusted and the adjusted Zarakas percentages between 2004 and 2005. Zarakas WDT (SP Ex. 27) at 31, Table 12. After rounding to the nearest single decimal place,^[45] the 2004 award is found to decline in 2005 by 0.4—a decline on the order of 10%.^[46] That is, an award of 4% in 2004 must necessarily correspond to an award of 3.6% in 2005. Both awards remain within the respective ranges which we have previously identified as setting the parameters of a zone of reasonableness for each award year.

The 4.0% award for 2004 and the 3.6% award for 2005 apply to the Basic Fund as well as the 3.75% Fund and the Syndex Fund for each of the respective award years. We take this approach because all the proposals provide a uniform award for these funds and no evidence was presented in opposition.^[47] The awards for the other claimant groups will be calculated net of the Music Claimants' awards.

The Music Claimants' final awards are as follows:

Syndex fund (percent)	Year	Basic fund (percent)	3.75% fund (percent)
2004	4.0	4.0	4.0
2005	3.6	3.6	3.6

VII. Final Awards

After adjusting downward for the Music Claimants' share (the equivalent of taking the Music Claimants' share “off the top”), the respective shares of the Basic Fund determined by the Judges are as follows:

	2004 (percent)	2005 (percent)
Music Claimants	4.0	3.6
Canadian Claimants	1.9	1.9
Devotional Claimants	3.4	3.4
PTV	7.4	7.1
JSC	32.3	35.4
CTV	17.9	14.2
Program Suppliers	33.1	34.4

Similarly, adjusting downward to account for the Music Claimants' share, the respective shares of the 3.75% Fund determined by the Judges are as follows:

	2004 (percent)	2005 (percent)
Music Claimants	4.0	3.6
Canadian Claimants	1.4	1.2
Devotional Claimants	3.7	3.7
JSC	35.3	38.6
CTV	19.5	15.4
Program Suppliers	36.1	37.5

We agree with the Settling Parties that because only Music Claimants and Program Suppliers participate in the Syndex Fund and for the reasons provided supra at Section VI (Conclusion and Award), Music Claimants should receive 4.0% of the Syndex Fund for 2004 and 3.6% of the Syndex Fund for 2005. As a result, the respective shares of the Syndex Fund determined by the Judges are as follows:

	2004 (percent)	2005 (percent)
Music Claimants	4.0	3.6
Program Suppliers	96.0	96.4

VIII. Order of the Copyright Royalty Judges

Having fully considered the record and for the reasons set forth herein, the Copyright Royalty Judges order that the 2004 and 2005 cable royalties shall be distributed according to the following percentages:

2004 Distribution

Claimant group	Basic fund (percent)	3.75% fund (percent)	Syndex fund (percent)
Music Claimants	4.0	4.0	4.0
Canadian Claimants	1.9	1.4	0
Devotional Claimants	3.4	3.7	0
PTV	7.4	0	0
JSC	32.3	35.3	0
CTV	17.9	19.5	0
Program Suppliers	33.1	36.1	96.0

2005 Distribution

Claimant group	Basic fund (percent)	3.75% fund (percent)	Syndex fund (percent)
Music Claimants	3.6	3.6	3.6
Canadian Claimants	1.9	1.2	0
Devotional Claimants	3.4	3.7	0
PTV	7.1	0	0
JSC	35.4	38.6	0
CTV	14.2	15.4	0
Program Suppliers	34.4	37.5	96.4

So ordered.

Dated: July 21, 2010.

James Scott Sledge,

Chief Copyright Royalty Judge.

William J. Roberts, Jr.,

Copyright Royalty Judge.

Stanley C. Wisniewski,

Copyright Royalty Judge.

Dated: July 21, 2010.

James Scott Sledge,

Chief, U.S. Copyright Royalty Judge.

Approved by:

James H. Billington,

Librarian of Congress.

Footnotes

1.  For a discussion of the operation of the section 111 license and the establishment of the funds for distribution, see, Distribution of 2000-2003 Cable Royalty Funds, Distribution order, in Docket No. 2008-2 CRB CD 2000-2003 (“2000-03 Distribution Order”), 75 FR 26798 (May 12, 2010).

2.  Music Claimants are comprised of the performing rights organizations (“PROs”)—the American Society of Composers, Authors and Publishers (“ASCAP”), Broadcast Music, Inc. (“BMI”), and SESAC.

3.  Prior to this deadline, the participants filed a stipulation of settlement as to NPR's claim to the 2004 and 2005 cable royalty funds and their agreement that NPR no longer needed to participate further in this Phase I proceeding. Upon notification to the Judges that all Phase II claims had been resolved, NPR moved for final distribution of their share to the 2004 and 2005 funds. The Judges granted the motion. See Order Granting Motion for Final Distribution, Docket No. 2007-3 CRB CD 2004-2005 (April 16, 2009). It is the funds remaining after this Order that are the subject of this determination.

4.  Hereinafter, references to the written direct testimony shall be cited as “WDT” preceded by the last name of the witness and followed by the exhibit number and the page or paragraph number. Similarly, references to the written rebuttal testimony shall be cited as “WRT” preceded by the last name of the witness and followed by the exhibit number and the page or paragraph number. References to the transcript shall be cited as “Tr.” followed by the page number and the name of the witness. References to the proposed findings of fact and conclusions of law shall be cited as “PFF” or “PCL,” respectively, preceded by the name of the party that submitted same (i.e., Settling Parties (“SP”), Program Suppliers (“PS”), Canadian Claimants (“CCG”) or Devotionals (“D”)) and followed by the paragraph number.

5.  The Judges also admitted the testimony of the following witnesses for the Settling Parties without live testimony pursuant to the stipulation of all parties: Dr. Gregory M. Duncan, Professor, the University of California, Berkley, and Managing Director, Huron Consulting Group, Tr. at 36-37; John F. Wilson, Senior Vice President & Chief TV Programming Executive, Public Broadcasting Service, id. at 397-98; Jonda K. Martin, President of Cable Data Corporation (“CDC”), id. at 528-29; and Alexandra Patsavas, Owner, Chop Shop Music Supervision, id. at 1009.

6.  The Judges also admitted the testimony of the following witnesses for the Canadian Claimants without live testimony pursuant to the stipulation of all parties: Janice de Freitas, Manager of the Rights Management Unit, Canadian Broadcasting Corporation/Radio-Canada, Tr. at 1270-72; Alison Smith, correspondent for the Canadian Broadcasting Corporation, id. at 1272; and Joan Fisher, Legal Counsel, Decode Entertainment, Inc., id. at 1273.

7.  The Judges also admitted the testimony of the following witnesses for the Devotional Claimants without live testimony pursuant to the stipulation of all parties: Dr. Charles F. Stanley, Senior Pastor, First Baptist Church, Atlanta, Georgia, and President, In Touch Ministries, Tr. at 1393-94; and Bruce Johansen, former President and CEO, the National Association of Television Program Executives, id. at 1394-95.

8.  The Judges also admitted the testimony of the following witnesses for the Program Suppliers without live testimony pursuant to the stipulation of all parties: Alex Paen, President, Telco Productions, Inc., Tr. at 1529; Jonda K. Martin, id. at 1529-30; Dr. Martin R. Frankel, Professor of Statistics and Computer Information Systems, Baruch College, City University of New York, id. at 1530-31; and Dr. Alan M. Rubin, Professor Emeritus and Director Emeritus, School of Communication Studies, Kent State University, id. at 1531-32.

9.  Neither the Canadian Claimants nor the Devotional Claimants objected to the adoption of the stipulation.

10.  The Judges also admitted the rebuttal testimony of two witnesses for the Settling Parties without live testimony pursuant to the stipulation of all the parties: Michael D. Topper, Vice President & Head of the Antitrust & Competition Practice, Cornerstone Research, Tr. at 2334-35; and Greg Stone, Owner/Chief Executive Officer, Greg Stone Media Consulting, id. at 2335.

11.  The Judges also admitted the rebuttal testimony of two witnesses of the Program Suppliers without live testimony pursuant to the stipulation of all the parties: Dr. Gruen, Tr. at 3238-39; and Dr. George Ford, id. at 3384-86.

12.  There remains an outstanding motion filed jointly by the parties requesting that the Judges adopt specific descriptions of the program categories at issue in this proceeding. However, at closing argument, the parties deemed the motion as no longer necessary. See, e.g., 5/10/10 Tr. at 33, 94 (Closing Argument). Consequently, the motion is denied.

13.  For a more complete discussion of how the standards for distribution have changed throughout the course of the section 111 license, see 2000-03 Distribution Order, 75 FR at 26801-02 (May 12, 2010).

14.  In markets characterized by some degree of monopoly power, consumer preferences are not honored in the same manner as in perfectly competitive markets, resulting in higher prices being charged to consumers and lesser quantities of goods/services being sold at the market price. Firms in such markets are, to varying degrees, price-makers rather than price-takers as compared to firms operating in perfectly competitive markets. So while a perfectly competitive firm is motivated to sell as much as it can produce up to the point where its marginal costs equate with the market price established by the market demand curve, a firm with some monopoly power is only motivated to sell up to the point where its marginal costs equate with the marginal revenues associated with the higher price it influences or dictates as reflected in the firm's downward sloping demand curve.

Testimony such as that offered by Judith Meyka describing the cable marketplace as competitive and declaring that the value of any particular programming to a cable operator is derived from the perceived value to the subscriber (see Meyka WDT (SP Ex. 4) at 4) is simply not credible in the face of well-documented studies showing the exercise of pricing power based on single cable operator dominance in the cable markets serving most Americans and in light of the fact that cable operators restrict their channel offerings to subscribers to bundles of channels, not just to the channels subscribers typically view. See, for example, U.S. General Accounting Office (GAO), Issues Related to Competition and Subscriber Rates in the Cable Television Industry , October, 2003 (“October 2003 GAO Report ”) at 30-31.

15.  See, for example, October 2003 GAO Report at 30-31. [“Most cable operators with whom we spoke provide subscribers with similar tiers of networks, typically the basic and expanded-basic tiers, which provide subscribers with little choice regarding the specific networks they purchase * * *. The manner in which cable networks are currently packaged has raised concern among policy makers and consumer advocates about the lack of consumer choice in selecting the programming they receive. Under the current approach, it is likely that many subscribers are receiving cable networks that they do not watch. In fact, a 2000 Nielsen Media Research Report indicated that households receiving more than 70 networks only watch, on average, about 17 of these networks. The current approach has sparked calls for more flexibility in the manner that subscribers receive cable service, including the option of à la carte service, in which subscribers receive only the networks that they choose and for which they are willing to pay.”] See also, U.S. Government Accountability Office, Media Programming: Factors Influencing the Availability of Independent Programming in Television and Programming Decisions in Radio, March, 2010 at 1-24. See also the testimony by Dr. Crawford for the Settling Parties and Dr. George Ford for the Program Suppliers concerning some of the economic effects of bundling as summarized in SP PFF at ¶¶ 447-49, 534.

16.  Indeed, even PTV does not object to the share accorded it under the Settling Parties' proposed shares which are based on the Bortz study as augmented by further adjustments.

17.  Just for purposes of clarity, when we say that a firm is exercising some degree of monopoly power, we mean that the firm has some influence over prices—that is, the market in which it participates is characterized by something less than perfect competition. In short, the firm may exercise market power that falls short of being a perfect monopoly, but does exercise sufficient market power to determine that it does not participate in a perfectly competitive market.

18.  Various arguments are made by some parties concerning whether or not the Judges must consider or require proof of changed circumstances, separate and apart from the estimates of relative value presented by the parties. We find, as did the 1998-99 CARP, that changed circumstances are embedded within the methodologies that provide reliable estimates of relative valuations and, therefore, have already been accounted for and are subsumed within the calculus of results. See 1998-99 CARP Report at 16, 31-2.

19.  Because Ms. McLaughlin's figures sum to slightly more than exactly 100%, we will adjust across the board to preserve the same relationships and to produce a final distribution of no more than exactly 100%.

20.  We recognize that this adjustment may not be precise. However, we agree with the Settling Parties that it would be reasonable to expect that some portion of the value assigned by Bortz survey respondents to Program Suppliers' programming was attributed to some non-compensable programming, even though there is no direct evidence in the record that delineates with specificity how much of the value attributed by Bortz survey respondents to Program Suppliers' programming categories was in fact attributable to non-compensable programming on WGN-A. See supra at 16-17 and SP PFF at ¶ 348, n.802. Furthermore, inasmuch as the Program Suppliers' programming likely involves non-compensable programming as does that of the Devotional Claimants, fairness demands that both these parties' shares should be impacted relative to the shares of the Settling Parties whose programming is largely compensable. Despite our lack of precision in our adjustment, the direction of the adjustment is correct and the magnitude of the impact on the Settling Parties' shares, though positive, is relatively small.

21.  The Canadian Claimants do not have a claim to Syndex Fund royalties.

22.  As previously noted, the Bortz survey excludes the responses of cable systems carrying only Canadian and/or PTV signals because they presumably can respond by only giving 100% value to Canadian and/or PTV programming, to the exclusion of all other program categories. SP PFF at ¶ 313.

23.  The Settling Parties accept 60% for both years. SP PFF at ¶ 336.

24.  Indeed, on the most important relative marketplace value question, the Canadian Claimants did not supply any additional testimony or support beyond the assertions of Dr. Calfee from the prior proceeding.

25.  The 2004 inclusion of the Seattle, Washington, signal discussed by Dr. Gary Ford does as well.

26.  The Settling Parties renew their argument, made in the 2000-03 proceeding, that it would be an error of law for us to adopt the Canadian Claimants' fee generation methodology as applied to the royalties collected from all large cable systems in the U.S., as opposed to only those in the Canadian zone. SP PCL at ¶ 30. We were not persuaded by the argument, particularly given the fact that fee generation had been applied to all large cable systems in the 1998-99 proceeding and had been found acceptable by the Register of Copyrights, Librarian of Congress and the United States Court of Appeals for the District of Columbia Circuit. 75 FR 26798, 26805 (May 12, 2010). In any event, we need not reconsider the argument here because we are not adopting the Canadian Claimants' fee generation approach as the method for determining their award.

27.  Devotional Claimants assert that after taking into account the Music Claimants' award, their Bortz shares fall into a reasonable range of 5.8%-8.5% and that the 7% and 7.3% they request fall within that range. DPCL at ¶¶ 106-107. The requested 3.75% Fund share is adjusted only to reflect the fact that PTV does not have any claim to the 3.75% Fund. DPFF & PCL at p. 7 (Introduction and Summary). Devotional Claimants do not seek a share of the Syndex Fund. Id. at ¶ 107.

28.  Nearly 50% of Form 3 cable systems carried WGN-A as their only distant signal and approximately 70% of Form 3 systems carried WGN-A as one of their distant signals. See SP PFF at ¶ 343.

29.  In addition to Mr. Zarakas, the Settling Parties also presented the testimony of certain other witnesses who testified about the value of music in programming generally. Based on testimony from these witnesses the Settling Parties contend that “[t]here is substantial qualitative evidence * * * that music's contribution to the overall television entertainment experience has increased over the past ten years.” SP PFF at p. 35 (Introduction and Summary). Absent quantitative corroboration, we are unable to credit significantly anecdotal and subjective opinion evidence. See Webcasting II, 72 FR at 24095 n.30 (May 1, 2007).

30.  Dr. Schink derived his data from a U.S. Census Bureau Report. 1998-99 CARP Report at 84.

31.  Mr. Zarakas identified two data sources that provide information concerning music license fees for 2004 and 2005: (1) Music blanket local television license fee data provided by the PROs; and (2) actual music license fee expenditures made by the broadcast stations. Zarakas WDT (SP Ex. 27) at ¶ 31. After 1998, individual data points for music license and broadcast rights payments were no longer available from the U.S. Census Bureau. Id. at n.17. Mr. Zarakas chose to use the blanket license fee data available from the PROs because he concluded that such negotiated fees provide strong evidence of the market value of the music licenses to the local broadcast stations and are the only available measures of total market-based prices. Id. at ¶¶ 32-33.

32.  For a negotiated annual fee, a blanket license grants the licensee unlimited use of all music in the PRO's repertoire. SP PFF at ¶ 366. The local television industry includes, among others, stations that are affiliated with the Big-3 networks with respect to non-network programming. The Big 3 networks pay separate music license fees to license music they use in their respective network programming. Zarakas WDT (SP Ex. 27) at ¶ 34 and n.19. Television stations that are affiliated with the non-Big 3 networks, with one exception, pay music license fees for stations and network programming. The Univision network pays a blanket license fee that covers all the programming for the stations that Univision owns. Id. at n.21.

33.  The fees that Univision paid totaled $5.31 million in 2004 and $5.72 million in 2005. Zarakas WDT (SP Ex. 27) at n.21. Mr. Zarakas includes the Univision blanket license fees in a category of the numerator called “other,” which totals $14.51 million in 2004 and $15.16 million in 2005. In that category he also includes blanket license fees for off-air and small stations. Id. at ¶ 34, Table 2. It is unclear what portion of the fees in the “other” category is attributable to those off-air and small stations. It is noteworthy, however, that Mr. Zarakas excludes small and “unlicensable” stations in calculating an important component of the denominator regarding broadcast rights payments. See Zarakas WDT (SP Ex. 27) at ¶ 36.

34.  Given the lack of evidence in the record to the contrary, for purposes of our analysis of Mr. Zarakas' music ratio denominator we assume that the four components he has proposed to include in the denominator represent the total of programming expenditures in the over-the-air market.

35.  The Broadcast Cable Financial Management Association Web site indicates that its name has since been changed to Media Financial Management Association (http://www.bcfm.com/​index.aspx?​PageID=​338).

36.  The NAB reports music license fees paid to PROs based on a survey of television stations. Zarakas WDT (SP Ex. 27) at n.18. By 2004, the U.S. Census Bureau no longer reported actual expenditures on music license fees by the television broadcasters as it did in the 1998 Annual Survey of Communication Services. Id.

37.  According to SNL Kagan's Web site, SNL Kagan integrates online research, data and projections in real time for the media and communications industry. http://www.snl.com/​Sectors/​Media-Communications/​.

38.  Mr. Zarakas multiplied the factor by the broadcast rights payments for local commercial television station non-network programming and non-Big 3 network programming, calculated in the previous two components of the denominator, “to form a complete estimate of broadcast rights payments applicable to the Music Ratio.” Zarakas WDT (SP Ex. 27) at ¶ 47.

39.  Mr. Zarakas reasons that although “[t]he local over-the-air market is broadcast to anyone with a television set within range of transmission * * * the market for distant signals on a cable system is dependent upon both the portfolio of signals a cable system operator elects to retransmit and upon the subscription choices made by the cable system operator's customers.” Zarakas WDT (SP Ex. 27) at ¶ 50.

40.  Mr. Zarakas' adjustment requires a multiple-step process: (1) Determine the relative numbers of distant subscribers by television station category (Zarakas WDT (SP Ex. 27) at ¶ 54 and Table 9); (2) convert those relative subscriber numbers into weights for each television station category by excluding educational, non-U.S. and low-power television stations from the distant subscriber totals (Zarakas WDT (SP Ex. 27) at ¶¶ 56-57 and Table 10); (3) determine the percentage of blanket license fees attributed to each television station category (Zarakas WDT (SP Ex. 27) at ¶ 59 and Table 11); and (4) apply the weights in step 2 to the percentages in step 3 to derive weighted percentages. Zarakas WDT (SP Ex. 27) at ¶ 60 and Table 12.

41.  See also Tr. at 1158 (Zarakas) (“[C]opyrighted content that's paid for by the local stations or the equivalent value of local programming, would be 3.1 percent * * *. But the 3.1 [percent] is somewhat misleading in the distant signal market because the composition of signals is different in the distant signal market compared to the over-the-air market.”).

42.  Dr. Woodbury also questioned Mr. Zarakas' treatment of WGN as an independent station rather than a WB affiliate for purposes of assigning a percentage music royalty due to the carriage of WGN. The Settling Parties represent that the distant signal market is dominated by WGN America, an independent station that does not retransmit any network programming and accounts for approximately half of the distant signal subscriber instances. SP PFF at ¶ 391. Dr. Woodbury contends that the “effect of this reclassification appears to have dramatically increased the weight on the percentage music rate of independent stations because WGN is apparently one of the most widely—if not the most widely—carried distant signal[s].” Woodbury WRT (PS Ex. 14) at ¶ 29 (footnote omitted).

43.  Dr. Woodbury did not include per-program license fees for SESAC because, he represents, SESAC did not offer a per-program license to local stations in 2004 and 2005. Woodbury WRT (PS Ex. 14) at ¶ 20.

44.  In the 1998-99 proceeding, the CARP awarded the Music Claimants 4.0% for the Basic Fund, the 3.75% Fund and the Syndex Fund. The Librarian adopted the CARP's determination. Distribution of 1998 and 1999 Cable Royalty Funds, Final order, in Docket No. 2001-8 CARP CD 98-99, 69 FR 3606, 3620 (January 26, 2004).

45.  We do not find the ratio evidence presented either before or after adjustments to be so precise as to warrant awards beyond a single decimal place.

46.  With respect to the Zarakas ratios, the decline from 2004 to 2005 is larger for the adjusted ratio than for the unadjusted ratio. Having found hereinabove that the upper boundary of the zone of reasonableness for the music award lies below the Zarakas adjusted ratio, a slightly less than proportionate adjustment from 4% (i.e., less than that indicated by the decline in the adjusted Zarakas ratio of 11.5%) is appropriate because the amount of variance between the adjusted and unadjusted ratios shrinks as the amount of adjustment decreases toward the limit of an unadjusted ratio. We further note, that even applying the calculated change in the Zarakas unadjusted ratio from 2004 to 2005 to the 4% 2004 award (i.e., a decline in the unadjusted Zarakas ratio of 9.7%), after rounding to the nearest single decimal, the resulting 2005 award (3.6%) would be the same as if we had applied a changed value as high as 11.2%.

47.  As the CARP noted in the 1998-99 proceeding, “[i]n past proceedings, Music has always received the same net award for each fund.” 1998-99 CARP Report at n.60. In that proceeding, no evidence was adduced in the proceeding to award a difference between the three funds.