5102853 Before the Federal Communications Commission Washington, DC 20554 In the Matter of Applications for Consent to the Transfer of Control of Licenses General Electric Company, Transferor, To Comcast Corporation, Transferee ) ) ) ) ) ) ) ) ) ) ) ) DECLARATION OF HAL J. SINGER Introduction ..................................................................................................................................... 3 Qualifications ................................................................................................................................ 24 I. The Threat to Comcast’s Dominance in the Delivery of Video Programming ................ 26 A. The Relevant Upstream and Downstream Markets for Assessing the Merger’s Likely Effects ........................................................................................................ 27 1. The Relevant Downstream Market ........................................................... 27 2. The Relevant Upstream Markets .............................................................. 29 a. Denial of Regional Sports Programming ...................................... 29 b. Denial of Local Broadcast Programming ..................................... 31 c. Denial of National Sports Programming....................................... 32 d. Denial of Online Video Programming .......................................... 33 2. The Relevant Geographic Markets ........................................................... 34 a. Denial of Local Broadcast Programming or Regional Sports Programming................................................................................. 34 b. Denial of National Sports Programming....................................... 37 c. Denial of Online Video Programming .......................................... 37 B. Measures of Comcast’s Market Power in the Supply of MVPD Service and in the Purchase of Video Programming .......................................................................... 37 1. Market Shares ........................................................................................... 38 2. Barriers to Entry........................................................................................ 41 3. Ability to Exclude Rivals .......................................................................... 43 a. Withholding Critical Local Inputs ................................................ 43
162
Embed
Before the Federal Communications Commission Washington ...€¦ · foreclosure strategy in Philadelphia and Portland, Comcast would have then refused to supply Versus to its MVPD
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
5102853
Before the Federal Communications Commission
Washington, DC 20554
In the Matter of Applications for Consent to the Transfer of Control of Licenses General Electric Company, Transferor, To Comcast Corporation, Transferee
I. The Threat to Comcast’s Dominance in the Delivery of Video Programming ................ 26 A. The Relevant Upstream and Downstream Markets for Assessing the Merger’s
Likely Effects........................................................................................................ 27 1. The Relevant Downstream Market ........................................................... 27 2. The Relevant Upstream Markets .............................................................. 29
a. Denial of Regional Sports Programming...................................... 29 b. Denial of Local Broadcast Programming ..................................... 31 c. Denial of National Sports Programming....................................... 32 d. Denial of Online Video Programming.......................................... 33
2. The Relevant Geographic Markets ........................................................... 34 a. Denial of Local Broadcast Programming or Regional Sports
Programming................................................................................. 34 b. Denial of National Sports Programming....................................... 37 c. Denial of Online Video Programming.......................................... 37
B. Measures of Comcast’s Market Power in the Supply of MVPD Service and in the Purchase of Video Programming .......................................................................... 37 1. Market Shares ........................................................................................... 38 2. Barriers to Entry........................................................................................ 41 3. Ability to Exclude Rivals.......................................................................... 43
a. Withholding Critical Local Inputs ................................................ 43
-2-
NAVIGANT ECONOMICS
5102853
i. Regional Sports Networks ................................................ 43 ii. Local Contractors.............................................................. 48
b. Control over Local Franchise Authorities..................................... 49 B. New and Old Delivery Platforms Threaten Comcast’s Dominance ..................... 56
1. Fiber Networks Deployed by Telephone Companies ............................... 56 2. Non-Telco Cable Overbuilders ................................................................. 60 3. Direct Broadcast Satellite ......................................................................... 66 4. Internet-Only Video Distributors.............................................................. 71
II. Comcast Currently Engages in a Host of Anticompetitive Strategies to Protect Its Market Power ................................................................................................................................ 76 A. The Economic Framework for Assessing Comcast’s Exclusionary Conduct ...... 77 B. Types of Exclusionary Conduct............................................................................ 82
1. Refusing to Supply Affiliated Programming to Rival Distributors on Reasonable Terms..................................................................................... 82 a. Outright Denial ............................................................................. 82
i. Comcast SportsNet Northwest.......................................... 83 ii. Comcast SportsNet Philadelphia....................................... 86
b. Excessive Fees and Tying Carriage of RSNs to Less Popular Networks ....................................................................................... 87
c. Manipulating Zones or “Stealth Discrimination” ......................... 88 2. Denying Independent Programming Networks’ Access to Comcast’s
Subscribers................................................................................................ 89 a. Failed Attempts to Extract Equity in Exchange for Carriage ....... 93
i. MASN............................................................................... 94 ii. NFL Network .................................................................... 96 iii. Big Ten Network............................................................... 96
b. Successful Attempts to Extract Equity in Exchange for Carriage 97 i. NBA TV............................................................................ 97 ii. MLB Network................................................................... 99
3. Tying Affiliated Internet Content to the Purchase of Cable Television Service....................................................................................................... 99 a. Fancast .......................................................................................... 99 b. 2010 Olympics ............................................................................ 101
III. The Proposed Transaction Would Increase Comcast’s Ability and Incentive to Engage in These Anticompetitive Strategies ................................................................................... 103 A. NBC’s Marquee Network Programming ............................................................ 103
1. Local NBC Affiliates .............................................................................. 103 2. National Sports Programming................................................................. 104
B. NBC’s Marquee Online Programming ............................................................... 105 C. Ownership of These “Must-Have” Inputs Would Facilitate Comcast’s
Anticompetitive Schemes vis-à-vis Independent Networks and MVPD Rivals. 106 1. Refusing to Supply Affiliated Programming to Rival Distributors on
Reasonable Terms................................................................................... 109 a. NBC Local Affiliates .................................................................. 109 b. National Sports Programming (Versus)...................................... 111
-3-
NAVIGANT ECONOMICS
5102853
2. Denying Access to Independent Programming Networks ...................... 114 3. Tying Affiliated Internet Content to the Purchase of Cable Television
Service..................................................................................................... 115 D. Comcast’s Economists Do Not Effectively Rebut the Presumption That Comcast
Would Withhold NBCU’s Programming from MVPD Rivals ........................... 118 1. Katz’s and Israel’s Claim That Comcast Would Not Withhold NBC
Affiliates ................................................................................................. 118 a. Drs. Katz and Israel’s Critical Departure Shares Are Inaccurately
High............................................................................................. 119 b. The Katz-Israel Empirical Methodology of Actual Departure Share
Is Conducted in a Manner that Likely Understates the True Departure Shares......................................................................... 124
2. Drs. Katz’s and Israel’s Claim That Comcast Would Not Withhold Affiliated Online Content ....................................................................... 127
IV. The Commission Should Craft Remedies to Protect Competition in the Video Marketplace..................................................................................................................... 140 A. Remedies to Address Comcast’s Refusing to Supply Affiliated Programming to
Rival Distributors on Reasonable Terms ............................................................ 141 1. The Prospect of A La Carte Pricing........................................................ 142 2. Economic Proof of the Benefits of the A La Carte Restraint ................. 144
B. Remedies to Address Comcast’s Denying Access to Independent Programming Networks ............................................................................................................. 147
C. Remedies to Address Comcast’s Tying of Affiliated Internet Content to the Purchase of Cable Television Service................................................................. 151
1. I have been asked by counsel for Communications Workers of America (CWA) to
analyze the competitive effects of Comcast’s proposed joint-venture with NBC Universal
(NBCU), which will be majority-owned and managed by Comcast (the “proposed transaction”).
In particular, I have been asked to assess the proposed transaction’s likely competitive effects in
the markets for multi-video programming distribution (MVPD), including “over-the-top” (OTT)
video providers, video programming, and broadband Internet access. I have also been asked to
review the submissions by Comcast’s economists, Dr. Mark Israel and Professor Michael L.
-4-
NAVIGANT ECONOMICS
5102853
Katz, on the proposed transaction’s likely effects on the supply of affiliated online video1 and
NBCU’s local NBC affiliates.2 I understand that CWA is concerned that a potential reduction in
output, brought about by exclusionary conduct made possible by the transaction, would lead to
less investment and thus fewer jobs in the communications sector. In my opinion, CWA’s
concerns are justified.
2. The proper lens through which to view the proposed transaction’s likely effects is
“vertical foreclosure.”3 Vertical connotes the vertical relationship between Comcast, a
“downstream” MVPD distributor, and “upstream” content providers or cable networks.
Foreclosure connotes Comcast’s refusal to carry unaffiliated content on its distribution platform
or the conditioning of such carriage on equity and exclusivity; it also connotes Comcast’s refusal
to supply or interference in the supply of affiliated programming to its downstream MVPD
rivals. In the past decade, Comcast has engaged in a systematic mission of vertical foreclosure
vis-à-vis independent cable networks, with the aim of securing those rights on an exclusive basis;
once secured, Comcast withheld those critical inputs to downstream MVPD rivals such as direct
broadcast satellite (DBS) providers at reasonable terms. In this sense, foreclosure vis-à-vis
unaffiliated cable networks (step one) and foreclosure vis-à-vis rival MVPDs (step two) are part
of the same foreclosure strategy. Once Comcast secures equity in the programming, the program
access protections of the Cable Act do not provide much relief to Comcast’s MVPD rivals, as
Comcast can charge itself the same inflated rates and still satisfy the “non-discrimination”
provisions of the Act—the high license fee is an internal transfer to Comcast.
1. Mark Israel & Michael L. Katz, The Comcast/NBCU Transaction and Online Video Distribution, May 4,
2010 [hereinafter Israel & Katz Online Video]. 2. Mark Israel & Michael L. Katz, Application of the Commission Staff Model of Vertical Foreclosure to the
Proposed Comcast-NBCU Transaction, Feb. 26, 2010 [hereinafter Israel & Katz NBCU]. 3. Just as dancing is a “vertical expression of a horizontal desire,” a cable operator’s vertical foreclosure
strategies vis-à-vis an upstream programmer is ultimately intended to affect horizontal outcomes vis-à-vis rival MVPDs.
-5-
NAVIGANT ECONOMICS
5102853
3. Up until now, Comcast has executed this vertical foreclosure strategy on a local
basis only. Comcast has perfected this strategy in Philadelphia and Portland, and the results have
been harmful for MVPD rivals and consumers in those markets. Without access to must-have
local sports—the Phillies, Flyers, and 76ers in Philadelphia, and the Trailblazers in Portland—
DBS penetration has been significantly reduced (relative to what it should have been4) and cable
prices have been inflated.5 After enduring Comcast’s refusal to license CSN-Philadelphia on a
reasonable basis6 and Comcast’s interference with the local franchising process, RCN halted
overbuilding activity in the Philadelphia DMA, and associated investments that would have
produced jobs and increased competition have been quashed. The stories of how Comcast came
to acquire that programming on an exclusive basis for the purpose of denying it to its MVPD
rivals—in apparent contravention of the Cable Act—are described here. The extent of the
reduction in DBS penetration and the inflation in cable prices have been quantified by
economists and by the Federal Communications Commission (FCC).7
4. Several observers mistake the modest growth in DBS penetration in Philadelphia over the past decade as
evidence that Comcast’s exclusionary conduct has had no anticompetitive effect. But the relevant comparison is actual DBS penetration at a given point in time versus what DBS penetration should have been at the same point in time. Confusing this point is akin to believing that a parent cannot impair the growth of her child by denying it milk so long as the child continues to grow.
5. Comcast is not the only cable operator to have executed this foreclosure strategy. Time Warner refused to carry the network owned by the Charlotte Bobcats until the Bobcats turned over equity to Time Warner. Now Time Warner refuses to supply those rights to its DBS rivals in Charlotte.
6. For years, Comcast supplied CSN-Philadelphia to RCN on short-term contracts, including a contract with six-month and one-month terms, despite the fact that carriage arrangements for sports programming are typically provided under long-term contracts. See Patricia Horn, Prices Tend to Rise as Competition Lags for Cable TV, KNIGHT RIDDER TRIBUNE BUSINESS NEWS, June 3, 2001 (“SportsNet has six-month contracts with RCN and the other cable companies, Williams said.”)
7. Some have argued that total welfare has not been reduced by Comcast’s exclusionary conduct because total output, when measured by the total number of MVPD subscribers, is unchanged. See, e.g., Scott Wallsten, An Economic Analysis of the FCC’s Program Exclusivity Ban, Working Paper, June 2007. Setting aside whether consumer welfare or total welfare is the proper economic standard—there is no dispute that Comcast’s exclusionary conduct has resulted in a transfer of surplus from Comcast subscribers (its installed base plus former DBS subscribers switching to Comcast to follow the excluded content) to Comcast—the proper metric for measuring output effects is the number of MVPD subscribers with access to the excluded content. Due to high switching costs and due to the supra-competitive premium that Comcast charges for basic cable in Philadelphia attributable to the exclusionary conduct, many DBS subscribers who value the excluded content but not sufficiently to compensate for these two costs are effectively removed from the market, thereby reducing the relevant measure of output.
-6-
NAVIGANT ECONOMICS
5102853
4. Comcast recently attempted to expand its vertical foreclosure strategy on a
national level when it sought to acquire the exclusive rights to the National Football League’s
(NFL’s) Thursday Night and Saturday Night telecasts. Comcast’s objective was to place those
games on Versus, a national sports cable network owned by Comcast. Following its local
foreclosure strategy in Philadelphia and Portland, Comcast would have then refused to supply
Versus to its MVPD rivals at reasonable rates. Shortly after the NFL refused to comply with
Comcast’s demands, and instead carried those games on NFL Network, Comcast relocated NFL
Network from a digital tier to Comcast’s lowly penetrated sports tier. During the ensuing trial
between Comcast and the NFL,8 former NFL Commission Paul Tagliabue testified that not only
did Brian Roberts threaten to re-tier NFL Network if the NFL did not supply the games on an
exclusive basis to Versus, but that the larger “cable industry” would support Comcast’s refusal to
deal.9 By coordinating its carriage decision vis-à-vis national sports programmers with out-of-
region cable operators, Comcast can effectively increase its “foreclosure share” from roughly 24
percent (its current share of national MVPD subscribers) to the sum of the shares of the group’s
members.10 In addition to Mr. Roberts’s threatened group boycott, other evidence indicates that
Comcast is coordinating its carriage strategy vis-à-vis national content providers with other cable
operators.11
8. I served as the NFL’s expert in the trial. 9. See Transcript of Record, NFL Enterprises LLC v. Comcast Cable Communications LLC, File No. CSR-
7876-P, Apr. 16, 2009, 1277: 10-1279:10 (Paul Tagliabue testimony describing Comcast CEO Brian Roberts’ suggestion that the NFL’s relationship with the “cable industry” would not be “positive” on a going-forward basis.)
10. It bears noting that a foreclosure share of 20 percent, which Comcast’s national share of MVPD subscribers exceeds on its own, is presumptively anticompetitive in the antitrust literature. See PHILLIP AREEDA, IX ANTITRUST
LAW 375, 377, 387 (Aspen 1991) (indicating that 20 percent foreclosure is presumptively anticompetitive); See also HERBERT HOVENKAMP, XI ANTITRUST LAW 152, 160 (indicating that 20 percent foreclosure and an HHI of 1800 is presumptively anticompetitive).
11. Comcast, Cox, and Time Warner coordinate their strategies vis-à-vis national movie studios via iN DEMAND’s pay-per-view service. TV Everywhere’s authentication service, which facilitates the coordination of strategies vis-à-vis national content providers and over-the-top video services, was originated by Comcast and Time Warner. Finally, empirical research indicates that vertically integrated cable operators coordinated their carriage decisions with respect to independent programming. See Jun-Seok Kang, Reciprocal Carriage of Vertically
-7-
NAVIGANT ECONOMICS
5102853
5. NBCU’s national sports programming is a key driver in the proposed
transaction,12 as it would allow Comcast to expand its foreclosure strategy into the remaining
Comcast service territories that are not served by a Comcast RSN. As of the closing of the
Adelphia acquisition in 2005, the latest date on which such data are publicly available, less than
half of Comcast’s subscribers (including Adelphia subscribers) had access to a Comcast RSN.13
Accordingly, Comcast is currently failing to exploit its vertical foreclosure strategy to over half
of its footprint. No longer is this the case.
6. Comcast has not been secretive about its desire to expand its footprint in sports
programming. According to its 2008 Annual Report, Comcast seeks to expand its reach into
sports and other “live-event” programming.14 In April 2009, Comcast’s Chief Operating Officer,
Steve Burke, said “Sports is the must-have programming on cable. One way that you can hedge
yourself a bit is to get into it yourself.”15 Indeed, in its Application and Public Interest Statement
requesting the instant transfer of licenses from General Electric, Comcast argued that the
proposed transaction will allow the combined firm to expand its footprint in sports programming:
The transaction will allow for NBC’s sports programming to be distributed on Versus, Golf Channel, and Comcast’s multiple RSNs, where brand identity would be greater and opportunity cost would be lower than if the sports programming were distributed on NBCU’s current non-sports networks such as Oxygen, Bravo, or MSNBC. Similarly, by
Integrated Cable Networks: An Empirical Study, Indiana University Working Paper, August 30, 2005, at 1 (finding empirical results that “make credible an underlying premise of a 30 percent national market share limit that the Federal Communication Commission established in 1993: namely, that MSOs may tacitly collude in their carriage decisions, having the effect of restricting market access to startup cable networks in which those MSOs have no ownership interest”) (emphasis added).
12. Application and Public Interest Statement, Application for Consent to the Transfer of Control of Licenses from General Electric Company, Transferor, to Comcast Corporation, Transferee, Jan. 28, 2010, at 50 [hereinafter Application].
13. As of May 2005, the total number of Comcast subscribers served by a Comcast RSN was roughly 11 million. See FCC Adelphia Docs - Exhibit AA, available at http://fjallfoss.fcc.gov/ecfs/document/view?id=6517610277.
14. Comcast SEC Form 10-K, for the fiscal year ended Dec. 31, 2008, at 29 (“We have invested and expect to continue to invest in new and live-event programming that will cause our programming expenses to increase in the future.”) (emphasis added).
15. John Ourand, Comcast’s Burke takes on critics of company’s dual strategies, SPORTS BUSINESS JOURNAL, Apr. 13, 2009.
-8-
NAVIGANT ECONOMICS
5102853
combining the NBC network with Comcast’s national sports cable networks, new opportunities will be created for the combined entity to negotiate for broader rights packages and to expand cross-promotion of broadcast and cable sports.16
Comcast may transfer some national sports content that is currently aired on NBC and other
NBC-owned networks to Versus; this is critical to understanding a key motivation of the
merger—namely, the expansion of Comcast’s sports footprint and the associated foreclosure
strategy. According to the New York Times, in its bid to rival ESPN, Versus would be “turned
over to NBC for an overhaul,” “renamed something like NBC Sports,” and “on-air and
production talent would migrate from NBC” to Versus.17 In a press release announcing the
proposed transaction, Comcast and GE acknowledged that the consolidation of sports
programming was a key merger-related synergy.18 If Comcast transfers some of that sports
content to Versus, Comcast can demand supra-competitive fees for this content throughout its
footprint (and beyond), not just in the ten DMAs served by an NBCU owned-and-operated
(O&O) affiliate.19
7. Securing additional must-have local broadcast programming is still valuable (and
provides yet another anticompetitive motivation for the proposed transaction), but that
complementary foreclosure strategy would be available only in select DMAs now served by an
NBCU O&O affiliate. NBCU owns the rights to many of today’s top national sporting events,
including the U.S. Open Championship, The Ryder Cup, Presidents Cup, Kentucky Derby,
16. Application at 50 (emphasis added). 17. Richard Sandomir, With NBC, Comcast zeroes in on ESPN, NEW YORK TIMES, Dec. 2, 2009, at 13. 18. GE Comcast Press Release, Dec. 3, 2009, attached as exhibit to Comcast SEC Form 8K, filed Dec. 4, 2009,
at 308 (“A robust sports programming lineup featuring the Olympics (through 2012), NBC Sunday Night Football, NHL/Stanley Cup, PGA Tour, US Open, Ryder Cup, Wimbledon and the Kentucky Derby, Versus, Golf Channel and . . . regional sports networks.”).
19. Comcast recently announced an agreement with NBC affiliates regarding the transfer of events to a Comcast cable network. See Bob Fernandez, Comcast reaches deal to keep sports events on free TV, PHILADELPHIA
INQUIRER, June 20, 2010. But this agreement does not fully address the potential for anticompetitive impact, as it the article acknowledges that (1) some parts of an event may migrate to a Comcast cable network and (2) the agreement is only temporary. Id. (noting that “Comcast could broadcast some parts of competitions or individual events on "one of their pay channels.”) (emphasis added).
-9-
NAVIGANT ECONOMICS
5102853
Preakness Stakes, Notre Dame football, Wimbledon, French Open, NHL Stanley Cup Finals,
Summer and Winter Olympic Games through 2012, and—perhaps most importantly—NBC
Sunday Night Football. By transferring those national sporting events from NBC’s networks to
Versus, Comcast would form a national sports network that would rival ESPN. But unlike ESPN,
which is owned by Disney, Versus will be owned by the largest MVPD, serving roughly one-
quarter of all U.S. MVPD subscribers. Consideration of Comcast’s downstream profits would
severely distort the joint venture’s pricing incentives for Versus. In particular, Comcast would be
willing to incur losses (relative to the standalone profit-maximizing price for those rights) in its
upstream content division (Versus) in exchange for ill-gotten gains in its downstream distribution
division, even if some of those downstream “benefits” were to spill into non-Comcast territories.
And MVPD consumers will be harmed in either of two possible contingencies: (a) if Comcast
merely raises the price of Versus to rival MVPDs, these higher prices will be passed through to
MVPD subscribers in the form of higher cable bills; or (b) if Comcast outright refuses to supply
this programming to rival MVPDs—for example, by moving the future marquee Versus
programming online to escape the program access rules—then non-Comcast customers will be
forced to incur switching costs to follow this programming (as they switch back to Comcast) and
higher cable bills due to the reduction in MVPD competition. Non-switching subscribers who
value the excluded content—but not by enough to compensate for these two costs—would be
harmed and would be eliminated from the market, reflecting a reduction in output.
8. In addition to seeking control of NBCU’s national sports programming, Comcast
has recently acquired equity interests in several national sports networks, including MLB
Network, NHL Network, and NBA TV. (Comcast also owns the Golf Channel, a national sports
network.) Consistent with the discrimination hypothesis outlined here, Comcast re-tiered those
-10-
NAVIGANT ECONOMICS
5102853
three national sports networks from its lowly penetrated sports tier to its highly penetrated
“Digital Classic Tier” shortly after it acquired equity in those networks. Indeed, this is precisely
how Comcast extracts equity from rights holders of sports programming; refusal to grant equity
lands an independent programmer on the sports tier, which severely limits advertising
opportunities. And the best way out of this predicament is to offer equity to Comcast. These
acquisitions of national sports networks alongside NBCU’s national sports-programming
portfolio will vastly increase Comcast’s market power vis-à-vis its MVPD rivals throughout
Comcast’s service area.
9. In the Internet Age, sports programming is critical to an MVPD because viewers
demand to see it in real-time, often on their big-screen televisions. This “must-have” nature of
sports programming is recognized by advertisers, which explains why such programming
commands high advertising rates. By acquiring NBCU’s national sports content, Comcast seeks
to exploit this unique opportunity to protect its cable television profits. In particular, Comcast
could withhold affiliated national sports programming from downstream rivals to impair MVPD
competition. Non-sports or non-event programming that can be watched on delay without any
significant diminution in viewer utility is relatively less valuable, as MVPD subscribers
increasingly have alternative means for accessing that content after it originally airs, including
via the Internet. But Comcast has a solution for that problem as well (originally called “TV
Everywhere” and rebranded as “Fancast”).
10. In addition to securing national sports programming, Comcast is acquiring local
must-have programming via NBCU’s ten owned-and-operated (O&O) NBC affiliates. These
properties will also increase Comcast’s market power vis-à-vis MVPD rivals, but only in the
DMAs in which NBCU owns a local NBC affiliate. In the seven local markets in which NBCU
-11-
NAVIGANT ECONOMICS
5102853
owns a broadcast affiliate and Comcast owns an RSN (the “overlapping markets”)—Chicago,
Philadelphia, Dallas-Ft. Worth, San Francisco-Oakland-San Jose, Washington, DC
(Hagerstown), Miami-Ft. Lauderdale, and Hartford-New Haven—Comcast will realize a
substantial increase in market power vis-à-vis its MVPD rivals. In particular, Comcast can now
target customers of rival MVPDs who were unwilling to switch on account of access to local
sports but might switch on account of access to NBC. As it turns out, Comcast serves
approximately 70 percent of MVPD subscribers in the Philadelphia DMA, and it serves
approximately 60 percent of MVPD subscribers in the Chicago, Miami, and San Francisco
DMAs. With such significant downstream market shares, Comcast can count on recouping any
upstream losses associated with the withholding of the NBC affiliate with ill-gotten downstream
gains. And given the fact that Comcast is often the largest provider of local advertising in these
local markets, the acquisition of an NBC affiliate would increase Comcast’s ability to raise local
advertising rates.20
11. Relative to a standalone NBC broadcast affiliate, Comcast’s ownership reduces
the cost to the affiliate of withholding programming from Comcast’s MVPD rivals. When faced
with the prospect of not watching NBC programming long-term,21 some MVPD customers in the
20. According to an analysis by Booz Allen based on 2005 advertising data, Comcast’s share of local
advertising sales would increase from 25 to 39 percent in Philadelphia, and from 24 to 36 percent in San Francisco. See Comments of NBC, filed in NBC Media Ownership Comments, FCC 06-121 (Oct. 2006). The increase in the Herfindahl-Hirschman index (HHI) caused by the proposed transaction would be approximately 700 and 500, and the post-merger HHI would be approximately 2,400 and 1,900 in Philadelphia and San Francisco, respectively. According to the Horizontal Merger Guidelines, such levels would create a presumption of an exercise in market power. See Department of Justice and Federal Trade Commission Horizontal Merger Guidelines, § 1.51 (“Mergers producing an increase in the HHI of more than 50 points in highly concentrated markets post-merger potentially raise significant competitive concerns, depending on the factors set forth in Sections 2-5 of the Guidelines. Where the post-merger HHI exceeds 1800, it will be presumed that mergers producing an increase in the HHI of more than 100 points are likely to create or enhance market power or facilitate its exercise.”).
21. Comcast has offered an economic report by Drs. Michael Katz and Mark Israel that purports to show that MVPD subscribers would not switch to Comcast when faced with a short-term lack of a broadcast affiliate based on a handful of Dish-related anecdotes that have no import here. See Israel & Katz NBCU, supra. In particular, Drs. Katz and Israel examine a few cases in which a single DBS operator (Dish) lost access to a broadcast network. In those cases, Dish’s customers likely understood that the carriage dispute would be resolved shortly, as the local
-12-
NAVIGANT ECONOMICS
5102853
relevant DMAs will switch to Comcast; as a result, Comcast will earn incremental (downstream)
profits. These profits will offset the cost to Comcast of withholding the NBC affiliate.
Recognizing these tradeoffs, Comcast will be able to bargain for higher programming prices for
programming previously owned by NBC than NBC would otherwise be able to negotiate on its
own. This is true even if the loss in upstream programming profit from withholding the NBC
affiliate exceeds the gain in downstream MVPD profit due to customer switching. And MVPD
consumers will be harmed because these higher programming prices will be passed through to
them in the form of higher cable bills.
12. Figure 1 shows the areas of Comcast’s service territory that are most vulnerable to
an exercise of market power made possible by the proposed transaction.
FIGURE 1: COMCAST’S SERVICE TERRITORY MOST VULNERABLE TO AN EXERCISE OF MARKET POWER
affiliate was incurring significant losses with no offsetting downstream benefits. Moreover, Drs. Katz’s and Israel’s measure of the treatment effect—Comcast’s increase in market share—fails to capture intra-DBS switching (that is, from Dish to DirecTV that may have occurred in the handful of events studied), which could not occur if Comcast decided to withhold an NBC affiliate from both DBS providers. Accordingly, Drs. Katz’s and Israel’s proposed benchmarks likely understate the extent of switching that would occur here. Because DBS customers would quickly recognize that Comcast’s withholding of an NBC affiliate was likely a permanent strategy, the switching to Comcast would likely be as significant as the DBS-to-cable switching that has occurred in Philadelphia due to Comcast’s denial of access to CSN SportsNet Philadelphia to DBS rivals. Stated differently, by injecting an offsetting benefit to the withholding strategy into the calculus, Comcast’s withholding of a (now-Comcast-owned) NBC affiliate becomes more credible to MVPD rivals and their subscribers than NBCU’s withholding of an NBC affiliate.
Served by a Comcast RSN
Served by a NBCU O&O Affiliate
Overlap
Non-Overlap
Served by a Comcast RSN Only
Served by a NBCU O&O Affiliate Only
Increase in the price/outright denial of Versus
Tying of weaker NBC national cable networks to Comcast RSN
Tying of weaker Comcast national cable networks to NBC affiliate
-13-
NAVIGANT ECONOMICS
5102853
In all markets within Comcast’s service territory, including the non-overlapping areas and the
areas served by a Comcast RSN only in Figure 1, Comcast will enjoy newfound market power
vis-à-vis its MVPD rivals owing to its control of NBCU’s national sports programming. This
exercise of market power would take the form of outright denial of access or higher prices for
Versus—that is, higher prices than what would be charged by an independent network that
owned the same programming rights. In the markets within Comcast’s territories served by an
NBCU O&O affiliate, Comcast would either deny access or raise the price of that affiliate
network, particularly in the markets already served by a Comcast RSN, as those happen to be
markets in which Comcast generally enjoys local market shares in excess of 50 percent.
13. Although NBC’s remaining video programming, which includes Syfy, Bravo,
mun2, and Oxygen, may not constitute must-have content,22 Comcast would still be able to
extract inflated rates for this content under the proposed transaction. Specifically, Comcast could
bundle this content with a compelling network—that is, “tying” product—in the form of a
Comcast RSN (or a combined RSN and a local NBC affiliate). Comcast could then extract supra-
competitive prices for the “tied” products by threatening to charge a “penalty price” for its RSN
(or the combined RSN and local NBC affiliate) were it purchased on a standalone basis. Indeed,
22. It bears noting that the must-have feature of the programming is only relevant with respect to whether
Comcast refuses to supply affiliated programming to rival MVPDs. Affiliation of any programming, regardless of its must-have nature, creates the incentive for a vertically affiliated cable operator such as Comcast to discriminate vis-à-vis independent, similarly situated networks. For example, with ownership of NBCU’s national news networks, MSNBC and CNBC, Comcast would have a fresh incentive to discriminate in the carriage, tiering, and pricing of independent news networks that compete with MSNBC and CNBC.
Increase in the price/outright denial of NBC
-14-
NAVIGANT ECONOMICS
5102853
Comcast has allegedly executed this very strategy in Chicago, with CSN-Chicago serving as the
tying product and the new Comcast-affiliated network, called Sprout, serving as the tied product.
According to a complaint brought by Dish Network,23 Comcast threatened to charge a penalty
price for CSN-Chicago unless Dish purchased Sprout as part of a bundle, presumably at a supra-
competitive price. The proposed transaction significantly increases the potential for such
anticompetitive bundling, as the new tying product is significantly stronger for Comcast than a
Comcast RSN on its own in the overlapping markets. And in markets served by a Comcast RSN
but not served by an NBC O&O affiliate, including Sacramento, Baltimore, and Detroit,
Comcast can now tie NBC’s other network programming (for example, Syfy or Oxygen) to a
Comcast RSN—a tying strategy that was not as easily available to NBCU.
14. Last but not least, the proposed transaction would allow Comcast to impair the
development on the Internet as an alternative platform for video delivery. As part of the proposed
transaction, Comcast would acquire NBCU’s (32 percent) interest in Hulu.com, a joint venture of
NBCU, Fox (News Corp), and ABC Networks (Disney). By February 2010, the number of
videos watched on Hulu.com had increased to 912 million, making Hulu.com the second most
popular online video content portal.24 Among the most popular shows watched on Hulu.com in
the first quarter of 2010 were ABC’s Modern Family, Fox’s Glee, and Fox’s Family Guy.25
Comcast will also acquire an interest in NBCU’s other online content, including NBC.com,
CNBC.com, and iVillage.com. And Comcast will acquire access to Universal’s library of
23. Arbitration Demand, Dish Network v. Comcast Corporation LLC, AAA Case No. ________ (filed Jan. 27,
2010), ¶¶ 3, 13. 24. comScore Releases February 2010 U.S. Online Video Rankings, available at
25. Brian Stelter, Viacom and Hulu part ways, NEW YORK TIMES, Mar. 2, 2010, available at http://www.nytimes.com/2010/03/03/business/media/03hulu.html
-15-
NAVIGANT ECONOMICS
5102853
movies, which could be released for Comcast’s cable customers before they are released to
Comcast’s MVPD rivals.
15. Video programming delivered via the Internet is a significant threat to all MVPD
distributors, as acknowledged by Comcast itself in the Commission’s most recent video
competition proceeding.26 The recent increase in the quantity of video programming available
online along with a burgeoning ecosystem of software and hardware that enable customers to
view video delivered over the Internet on their televisions has transformed the Internet into a
closer substitute to cable television. The transition of Internet video from a complement to a
substitute for cable television can be seen in recent viewing data. According to comScore, the
average American web user spent about ten minutes a day in early 2009 (slightly under six hours
per month) viewing online video, compared with roughly 300 minutes spent watching live
television (slightly over 150 hours per month),27 suggesting that the two platforms were mild
substitutes not long ago. Because the audience for online video is young and growing, and
because the broadcast networks are replicating their content online, however, the Internet has
quickly emerged as a serious threat to cable television. Over the course of 2009, the average
amount of time that web users spent watching videos online more than doubled to nearly thirteen
hours per month (from slightly under six hours per month in the beginning of the year).28
Independent analyses confirm this trend. Parks Associates reported that the number of U.S.
26 In the FCC’s thirteenth annual report on video competition, Comcast noted that “Many networks have
jumped head-first into Internet video, providing consumers with an interactive alternative to traditional TV-set viewing.” Comcast Comments at 29-30 (emphasis added). Comcast provided further evidence that Internet video is a substitute to cable television: “All of these modalities of communications are important to younger consumers, all are part of the paradigm shift to a ‘what-you-want-when-you-want-it’ world, and all of them compete with traditional and not-so-traditional video distribution technologies for time, attention, and dollars.” Id. at 59 (emphasis added).
27. Time Warner and the internet: After the divorce, THE ECONOMIST, May 7, 2009. 28. comScore data shows 2009 was a blistering year for online video, VIDEO NUZE, available at
broadband households watching premium online content doubled in 2009; as of April 2010, over
25 million U.S. broadband households regularly watched full-length television shows online, and
over 20 million watched movies online.29 The Convergence Consulting Group estimated that,
from 2008 to 2010, 800,000 U.S. households disconnected their cable television service and
watched their television online; that number was also expected to double by 2011.30
16. Some of the most compelling online video is now available from Hulu.com,
programmers’ websites (such as Comedy Central), and the growing libraries of services like
Apple’s iTunes, Amazon’s Unbox, Amazon’s Video on Demand service, Netflix’s “Watch
Instantly” streaming service, and sites offering free content such as Joost. New hardware from
firms such as Apple (Apple TV) and TiVo (the recently released TiVo Premier), and new
software running on gaming consoles, DVD players, and increasingly built into televisions
themselves are making it easier for subscribers to access the Internet from their televisions.
Indeed, a new class of “over-the-top” (OTT) video providers, such as Boxee and Roku, aims to
reach Internet users with subscription-based or advertising-supported streaming-video services.
The success of the OTT business models depends critically on access to online content. And
investment by broadband access providers in faster and wider networks depends on the
development of this ecosystem, as the demand for the pipe is derived from the demand of the
content that rides over the pipe.
17. Cable companies have recognized the threat of their video distribution business
being cannibalized by their Internet business. As Glenn Brit, CEO of Time Warner Cable,
29. Parks Associates finds over 25 million U.S. broadband households regularly watch full-length TV shows
online, Apr. 20, 2010, available at http://www.fiercetelecom.com/press_releases/parks-associates-finds-over-25-million-u-s-broadband-households-regularly-watch-full.
30. Ryan Fleming, New Report Shows More People Dropping Cable TV for Web Broadcasts, Apr. 16, 2010, available at http://www.digitaltrends.com/computing/new-report-shows-that-more-and-more-people-are-dropping-cable-tv-in-favor-of-web-broadcasts.
-17-
NAVIGANT ECONOMICS
5102853
acknowledged in May 2009: “The reality is, we’re starting to see the beginnings of cord cutting
where people, particularly young people, are saying all I need is broadband.”31 In April 2009,
Comcast’s president and chief operating officer, Steve Burke, likened television viewers’
movement to online video to “wildfire.”32 According to Melinda Witmer, Time Warner Cable’s
programming chief, OTT video providers are actual (and not just potential) competitors: “We
wake up every day and there is some new competitor out there—a Roku or a Boxee. People like
to think of cable operators as monopolists, but we face a lot of competition just to keep the
business we have.”33 Despite this threat, incumbent cable operators still maintain an advantage
over OTT video providers by virtue of their control over video content, as explained by Time
Warner executive Jeffrey Bewkes: “We’re fortunately in a position where this [Internet video]
doesn’t cost us much money. We have an advantage and we’re going to use that advantage.”34
18. Beholden to cable operators for license fees and access to advertisers, some
broadcast networks appear to have fallen in line. For example, Hulu.com has blocked Boxee’s
ability to access Hulu.com video feeds. According to Jeff Zucker, CEO of NBC Universal, “We
want to find an economic model that makes sense.”35 Although broadcasters may weakly prefer
viewers to access their content via cable television rather than via the Internet (depending on
which platform generates more advertising revenue), a combined Comcast-NBCU would have
31. Christopher Lawton, More Households Cut the Cord on Cable, WALL STREET JOURNAL, May 28, 2009,
available at http://online.wsj.com/article/SB124347195274260829.html. 32. Tom Lowry, Cable TV: Pushing to Become More Web-like, BUSINESSWEEK, Apr. 16, 2009, available at
http://www.businessweek.com/magazine/content/09_15/b4126050298367.htm [hereinafter Cable TV]. 33. Ronald Grover, Tom Lowry & Cliff Edwards, Revenge of the Cable Guys, BUSINESSWEEK, Mar. 11, 2010,
available at http://www.businessweek.com/magazine/content/10_12/b4171038593210_page_2.htm [hereinafter Revenge of the Cable Guys].
34. Knowledge at Wharton, Cable TV Follows Its Subscribers to the Internet, July 22, 2009, available at http://knowledge.wharton.upenn.edu/article.cfm?articleid=2295. In the same article, the former chief economist of the FCC, Gerald Faulhaber, described Internet video as “very disruptive,” a technology that “attacks the model of cable television.” Id.
35. Dan Frommer, NBC Boss Explains Why Boxee Users Can’t Have Hulu, Mar. 18, 2009, available at http://www.businessinsider.com/nbc-boss-explains-why-boxee-users-cant-have-hulu-2009-3.
-18-
NAVIGANT ECONOMICS
5102853
even greater economic incentive to block such services, as the combined firm would fully
internalize the losses associated with cancelled cable television subscriptions. Indeed, the Hulu-
Boxee disagreement appears to have started in February 2009,36 one month before the Comcast-
GE negotiations over NBCU began by most press accounts.37 Accordingly, it is certainly
plausible that Comcast exerted some pressure on Hulu.com to not allow Hulu.com content to
appear on Boxee.
19. Comcast has addressed this increasing threat of Internet-based television
primarily in three ways. First, Comcast has made some of the video content it distributes
available online, but only to paying customers of both its cable television and cable modem
services through its TV Everywhere service, which is branded in Comcast markets as
“Fancast.”38 The proper lens to view this conduct is a tie-in, with Comcast’s cable television and
cable modem services serving as the tying product and the online content serving as the tied
product. Second, Comcast requires that unaffiliated cable networks not make their content
available online as a condition of being carried on Comcast’s cable television systems.39 The
proper lens to view this conduct is an exclusive deal; Comcast insists that a cable network not
license its content to OTT video providers or post content to its own website as a condition of
carriage on Comcast’s cable system. Third, Comcast maintains complete control of the set-top
box used to access its cable television service by requiring its cable television customers to
36. See, e.g., Chadwick Matlin, What Happened Between Boxee and Hulu, and When, THE BIG MONEY, Feb. 8,
2010. 37. See, e.g., Tim Arango, G.E. Makes It Official: NBC Will Go to Comcast, NEW YORK TIMES, Dec. 3, 2009
(“The deal’s genesis lies in frequent flirtations over the last several years between Comcast and General Electric, although serious talks began in March.”).
38. Fancast provides access to network shows and movies integrated with television-related news and a viewing guide for their video service. Information about Comcast’s Fancast is available at http://www.fancast.com.
39. Cable TV, supra, at *1 (“As the cable companies scramble to get these technologies out of the lab, they are trying to prevent cable networks from putting their shows online. In recent months, Big Cable has reminded USA, Bravo, TNT (TNT), and hundreds of other channels that the cable companies provide about half their revenues, in the form of fees to carry their shows. The implicit warning: put your content online and forfeit nearly $30 billion. ‘If I don’t have a customer,’ says Time Warner Cable’s Britt, ‘the programmers aren’t going to have a customer.’”).
-19-
NAVIGANT ECONOMICS
5102853
purchase the set-top box from Comcast to enjoy the full suite of interactive services. The proper
lens to view this conduct is another tie-in; in this case, the tied product is the set-top box. It bears
noting that Comcast’s customers can access Fancast only via their computers, another reflection
of Comcast’s intent to sever the Internet from the television.
20. These exclusionary strategies will slow the growth of Internet-only television and
thereby protect Comcast’s market power in the supply of cable television and cable modem
service. Because the marginal (or imputed) price of accessing Comcast’s online portfolio is zero
as a result of the tie-in (and Comcast’s pricing structure), providers of OTT video seeking to
compete against Comcast will be forced to raise revenues from advertisers only, which could
undermine their business model. By preventing unaffiliated cable networks from posting their
content online as a condition of carriage, and by denying access to its affiliated online content,
Comcast can deny a critical input to OTT video providers, further impairing Internet-only
television. When Comcast’s affiliated online properties are combined with those of Time
Warner, which also participates in TV Everywhere, the resulting collection of withheld content
could significantly impair OTT providers. A harmful byproduct of this foreclosure of OTT video
providers is that it would attenuate the value of the Internet ecosystem for rival broadband access
providers, which could reduce their incentives to invest in broadband infrastructure. By denying
its online portfolio to in-region rival Internet service providers (“ISPs”)—that is, by tying
Fancast to Comcast’s cable modem service—Comcast and other members of TV Everywhere can
induce substitution away from fiber/DSL connections offered by telephone companies, further
weakening broadband access rivals. Finally, by controlling the set-top box, Comcast can ensure
that its video subscribers cannot access the Internet from the comfort of their televisions. An
independent set-top-box maker would be more inclined to add features, including the ability to
-20-
NAVIGANT ECONOMICS
5102853
connect to and download video from the Internet via a Wi-Fi connection, that could threaten a
cable provider’s cable television revenues.
21. Comcast’s acquisition of NBCU’s online content portfolio would facilitate these
exclusionary strategies, both on a unilateral basis and on a coordinated basis with other cable
operators. Denying access to NBCU’s online content portfolio, especially Hulu.com, would
significantly impair the ability of OTT providers to compete effectively for online video
customers. And to the extent that TV Everywhere does not provide access to rival ISPs on
reasonable terms, denying access to NBCU’s online content portfolio would also impair the
ability of in-region ISP rivals to compete effectively for broadband subscribers. Although it may
be possible to achieve these outcomes through (exclusive) contracting, having these must-have
assets under ownership greatly facilitates the exclusionary objective, as the possibility of
defection by the affiliated, online content provider (Hulu.com and others) decreases.40 When
done in conjunction with Time Warner and other cable operators that join TV Everywhere, the
foreclosure share associated with the exclusionary strategy increases, which in turn increases
Comcast’s market power in the supply of online video. Independent cable networks and movie
studios naturally fear upsetting Comcast by posting their video content online or making it
available to OTT video providers; by including other cable operators, including Time Warner, in
TV Everywhere, the combined leverage vis-à-vis independent networks is even stronger.41
Having NBCU’s content portfolio also would increase Comcast’s market power vis-à-vis
independent programmers. In particular, Comcast’s threat not to carry independent networks that
40. It is no accident that the initial trial of TV Everywhere in the summer of 2009 allowed viewers to tap into
programming provided by TNT and TBS, both of which are owned by Time Warner. See Revenge of the Cable Guys, supra.
41. Revenge of the Cable Guys, supra (“Back at Time Warner Center in New York and One Comcast Center in Philadelphia, the cable operators began to realize they had the studios locked down. As Frank Biondi, former president of the media giant Viacom (VIA), puts it: ‘Why would [the studios] make a deal with a competitor to their largest customer and risk angering them?’) (emphasis added).
-21-
NAVIGANT ECONOMICS
5102853
seek to post some of their content online becomes more credible as Comcast adds to its online
video content portfolio, as Comcast’s online platform (Fancast) would become the most valuable
platform for accessing online video content. Finally, Comcast’s acquisition of NBCU’s online
content portfolio would increase Comcast’s incentives to thwart set-top-box technologies that
would enable cable television customers to access the Internet from their televisions. If a
Comcast cable customer could access Hulu.com, for example, from his television without having
to authenticate his being a Comcast cable television customer, then the value of Comcast’s cable
television platform would be compromised.
* * *
22. This theory of anticompetitive harm arising from the proposed transaction is
explained in further detail here. My report is organized as follows: In Part I, I explain the threat
to Comcast’s dominance in the delivery of video programming. I define the relevant upstream
and downstream markets for assessing the merger’s likely effects. I provide two measures of
Comcast’s market power in the supply of video programming at the local (DMA) level: (a) high
market shares combined with high entry barriers, and (b) Comcast’s ability to exclude rivals.
Next, I describe how new and old delivery platforms threaten Comcast’s dominance in the
supply of MVPD service.
23. The theory of anticompetitive harm is not just theoretical; Comcast currently
engages in nearly identical vertical foreclosure strategies vis-à-vis its in-region MVPD rivals,
primarily at the local level. In Part II, I describe these strategies and the resulting anticompetitive
effects in detail. I provide an economic framework for assessing Comcast’s exclusionary
conduct. Next, I document the types of exclusionary conduct that Comcast currently engages in,
including (a) refusing to supply affiliated programming to rival distributors on reasonable terms,
-22-
NAVIGANT ECONOMICS
5102853
(b) denying independent programming networks’ access to Comcast’s subscribers, and (c) tying
affiliated Internet content to the purchase of Comcast cable television service.
24. In Part III, I explain how the proposed transaction would increase Comcast’s
ability and incentive to engage in these anticompetitive strategies. Owning NBC’s marquee
network and online programming would facilitate a host of exclusionary conduct by Comcast on
both a unilateral basis and on a coordinated basis with other cable operators. In particular, NBC’s
marquee local broadcast programming would allow Comcast to extend its market power in the
ten DMAs in which NBCU has a local affiliate. And NBCU’s marquee national sports
programming, once moved to Versus, would allow Comcast to affect downstream competitive
outcomes throughout its territories. Comcast would also have a fresh incentive to deny access to
independent networks that are similarly situated to NBCU’s affiliated networks, including
national sports, Spanish-language, women’s programming, and national news networks. Finally,
NBCU’s marquee Internet properties, including Hulu.com, MSNBC.com, and NBC.com, would
allow Comcast to restrict access to Comcast cable subscribers, further weakening Comcast’s
downstream MVPD rivals and slowing the development of the Internet as a competitive platform
for video delivery.
25. In Part IV, I propose remedies that the Commission should craft to protect
competition in the video marketplace. To address Comcast’s refusing to supply affiliated
programming, including Versus and the ten O&O NBC affiliates, to rival MVPDs on reasonable
terms, I propose, among other things, that Comcast should be subjected to an a-la-carte restraint
that would (1) require Comcast to sell its affiliated networks on an unbundled basis to rival
MVPDs, and (2) require Comcast to permit its customers to “opt out” of an affiliated network on
Comcast’s Standard Service tier for a rebate equal to the wholesale price of the network. I
-23-
NAVIGANT ECONOMICS
5102853
demonstrate that the a-la-carte restraint corrects the incentive problem that Comcast now faces—
namely, to choose a wholesale price for its affiliated networks in excess of the price that would
be chosen by an independent network owner.
26. To address Comcast’s denying access to independent programming networks, I
propose, among other things, that Comcast should be compelled to carry the independent
network upon the referral of a program carriage complaint by the Media Bureau to an
administrative law judge, at interim terms chosen by the Media Bureau. Moreover, upon an
administrative law judge’s decision that Comcast has discriminated in favor of its affiliated
network and that, as a result, the independent network has been unreasonably restrained in its
ability to compete, Comcast should be compelled to pay damages to the independent network in
the amount equal to the forgone licensing revenues since the discriminatory conduct began with
interest. The Commission should also consider modifying its current program-carriage
adjudication process to reflect a baseball-style arbitration.
27. Finally, to address Comcast’s tying affiliated Internet content (Fancast) to the
purchase of Comcast cable video and cable modem service, I propose the Commission bans all
such tying arrangements, and it compels Comcast to price access to its Fancast service to
customers of any MVPD and any broadband service provider on a standalone and non-
discriminatory basis. To ensure that the non-discriminatory rate for Fancast that Comcast sets for
non-Comcast broadband access customers is also reasonable, Comcast’s video and cable modem
customers should be able to “opt out” of Fancast at a savings from the cable-video/cable-modem
bundle equal to the standalone rate for Fancast.
-24-
NAVIGANT ECONOMICS
5102853
QUALIFICATIONS
28. My name is Hal J. Singer. I am a Managing Director of Navigant Economics. I
am also an adjunct professor at the McDonough School of Business at Georgetown University.
My areas of economic expertise are antitrust, industrial organization, finance, and regulation. I
have applied my expertise to several regulated industries, including telecommunications, video
programming, insurance, and health care.
29. I have published a book chapter in Access Pricing: Theory, Practice and
Empirical Evidence (Justus Haucap and Ralf Dewenter eds., Elsevier Press 2005) and in
Handbook of Research in Trans-Atlantic Antitrust (Philip Marsden, ed., Edward Elgar Publishing
2006). I am also the co-author of the book Broadband in Europe: How Brussels Can Wire the
Information Society (Kluwer/Springer Press 2005).
30. I have published scholarly articles in many economics and legal journals,
including American Economic Review Papers and Proceedings, Berkeley Technology Law
Review, Canadian Journal of Law and Technology, Federal Communications Law Journal,
Harvard Journal of Law and Technology, Hastings Law Journal, Journal of Business and
Finance, Journal of Business Law, Journal of Competition Law and Economics, Journal of
Financial Transformation, Journal of Industrial Economics, Journal of Insurance Regulation,
Journal of Network Industries, Journal of Regulatory Economics, Journal of
Telecommunications and High Tech Law, Review of Network Economics, Telecommunications
Policy Journal, Topics in Economics Analysis and Policy, and Yale Journal on Regulation.
31. Two of my articles are of particular relevance to this proceeding: “The
Competitive Effects of a Cable Television Operator’s Refusal to Carry DSL Advertising,”
Journal of Competition Law and Economics (Vol. 2, No. 2, pp. 301-31, 2006); and “Vertical
-25-
NAVIGANT ECONOMICS
5102853
Foreclosure in Video Programming Markets: Implications for Cable Operators,” Review of
Network Economics (Vol. 6, 2007).
32. In regulatory proceedings, I have presented economic testimony in several
forums, including the FCC, the U.S. Federal Trade Commission, the Antitrust Division of the
U.S. Department of Justice, the U.S. National Highway Traffic and Safety Administration, the
House of Commons of Canada, the Canadian Radio-television and Telecommunications
Commission, and the U.S. Congressional Budget Office. My written testimony on the effect of
telecom entry on cable television prices was cited extensively by the Department of Justice in a
November 2008 report titled Voice, Video and Broadband: The Changing Competitive
Landscape and Its Impact on Consumers.42
33. I have served as an economic expert for the NFL Network and for MASN, which
owns the television rights to live baseball games of the Baltimore Orioles and the Washington
Nationals, in several carriage disputes. On June 2, 2008, the arbitrator in MASN v. Time Warner,
Judge Daniel H. Margolis, ruled that Time Warner “did discriminate against MASN based on
affiliation in not negotiating for carriage of MASN on an analog tier.”43 In his decision, Judge
Margolis cited my analysis on behalf of MASN on several occasions44 in support of his decision
that MASN’s offer price “accurately reflects the fair market value of the rights to carry MASN in
its North Carolina television territory.”45 In its October 30, 2008 Order on Review rejecting Time
Warner’s appeal of the arbitrator’s decision, the Media Bureau cited my oral testimony during
Phase II in support of the proposition that “the carriage decisions of four of the largest MVPDs
42. Department of Justice, Voice, Video and Broadband: The Changing Competitive Landscape and Its Impact
on Consumers, Nov. 17, 2008, available at http://www.usdoj.gov/atr/public/press_releases/2008/239479.htm. 43. TCR Sports Broadcasting Holding, L.L.P, d/b/a Mid-Atlantic Sports Network v. Time Warner Cable Inc.,
Case No: 71-472-E-00697-07, June 2, 2008, at 22. 44. Id. at 19, 19 n.13, and 21. 45. Id. at 22.
-26-
NAVIGANT ECONOMICS
5102853
operating in North Carolina—that serve the overwhelming majority of non-TWC subscribers to
paid television service in North Carolina—are an appropriate reference point for assessing fair
market value.”46 A declaration that I submitted to the Media Bureau also was cited extensively in
the Bureau’s order designating a program carriage complaint by the National Football League for
hearing.47 I currently serve as an economic expert for the Tennis Channel in its carriage dispute
with Comcast.
34. In addition to these carriage disputes, I have served as a testifying expert in
several litigation matters. My experience as a testifying expert in litigation is summarized in my
Curriculum Vitae, which is attached to this report.
35. Before joining Navigant Economics, I was president of Empiris LLC, an
economic consulting firm based in Washington D.C. Prior to that, I worked as an economist at
Criterion Economics and LECG. In addition, I have worked as an economist for the Securities
and Exchange Commission and the Army Corps of Engineers.
36. I earned M.A. and Ph.D. degrees in economics from the Johns Hopkins University
and a B.S. magna cum laude in economics from Tulane University.
37. I file this report in my individual capacity. I have no financial stake in the
outcome of this case.
I. THE THREAT TO COMCAST’S DOMINANCE IN THE DELIVERY OF VIDEO PROGRAMMING
38. Comcast wields significant market power within its cable footprint in the supply
of MVPD services. New and traditional delivery platforms, including DBS, fiber networks built
by telephone carriers, and the Internet, threaten its power.
46. TCR Sports Broad. Holding, L.L.P. v. Time Warner Cable Inc., Order on Review, DA 08-2441, ¶ 47, n.186
(MB Oct. 30, 2008). 47. Herring Broad., Inc. v. Time Warner Cable Inc., et al., Mem. Op. & Hearing Designation Order, 23 FCC
Rcd 14787 ¶¶ 77, 79, 81, 82, 83, 86 (2008).
-27-
NAVIGANT ECONOMICS
5102853
A. The Relevant Upstream and Downstream Markets for Assessing the Merger’s Likely Effects
39. Before assessing Comcast’s market power, one must first define the relevant
markets that are implicated by the proposed transaction. As the Commission correctly observed
in the Hughes-News Corporation proceeding, “there is no need to engage in a rigorous market
definition in order to analyze the potential anticompetitive effects of the transaction.”48 The
reason is that market definition is useful in making inferences about anticompetitive effects
based on (indirect) evidence of high market shares and entry barriers; when other, more direct
methods are available for examining anticompetitive effects, such as extant exclusionary
strategies by the acquiring firm, market definition is not critical. For completeness, however, I
proceed to define the relevant product and geographic markets.
1. The Relevant Downstream Market
40. Multi-channel video programming distribution (MVPD) service is the relevant
downstream product market. A variety of government entities, from the Federal Communications
Commission (FCC) to the Department of Justice (DOJ) to the Government Accountability Office
(GAO), have concluded that, in addition to other smaller suppliers, cable television providers and
DBS satellite television providers directly compete with each other in the supply of MVPD
services. MVPD service distributed by these and other providers is “reasonably
interchangeable,” and thus meets the antitrust interpretation of a properly defined product
market.49 The market is generally defined with regard to demand substitution, which focuses on
48. See General Motors Corporation and Hughes Electronics Corporation, Transferors, and The News
Corporation Limited, Transferee, for Authority to Transfer Control, MB Dkt. No. 03-124, Memorandum Opinion and Order (released Jan. 14, 2004) [hereinafter Hughes-News Corp. MO&O], ¶ 61.
49. MVPD services, such as those offered by wireline cable and DBS, are substitutable. See In the Matter of Annual Assessment of the Status of Competition in the Market for the Delivery of Video Programming, MB Dkt. No. 06-189, Thirteenth Annual Report, released Jan. 16, 2009 [hereinafter Thirteenth Annual MVPD Report], ¶ 5, 6. See also In the Matter of Applications for Consent to the Transfer of Control of Licenses and Section 214
-28-
NAVIGANT ECONOMICS
5102853
buyers’ views of which products are substitutes. Numerous government and non-governmental
studies indicate that the presence of cable overbuilders or DBS providers constrain the price
charged by incumbent cable television operators. To borrow just one example, the FCC found in
its 2009 video competition report that MVPD service can be provided by satellite or by cable
networks.50 Following FCC and DOJ precedent, I define the relevant downstream product market
as all MVPD service.
41. Although it is too new to likely constrain the price of cable television service,
online video service is properly viewed as part of the MVPD market; the Internet is a platform
that competes with other platforms (DBS, cable plant, fiber plant) in the delivery of video service
to end users. As mentioned earlier, over the course of 2009, the average amount of time spent
watching videos online more than doubled, from 356 minutes per month (5.9 hours per month) to
763 minutes per month (12.7 hours per month). Indeed, in its thirteenth annual video competition
report, released in January 2009, the FCC considered “web-based Internet video” to be an entrant
in the MVPD market.51 In that proceeding, Comcast commented that Internet video applications
compete with cable television for the time and attention of viewers.52 Comcast also observed that
video web sites drew users in numbers comparable to the subscriber reach of cable and satellite
companies.53 Specifically, Comcast noted that “Many networks have jumped head-first into
Authorizations, CS Dkt. No. 98-178, Memorandum Opinion and Order, released February 18, 1999, FCC 99-24, ¶ 21.
50. Thirteenth Annual MVPD Report, ¶ 3 (“The marketplace for the delivery of video programming services is served by a number of operators using a wide range of distribution technologies. […] Specifically, we examine the cable television industry and other established multichannel video programming distributors (“MVPDs”), including direct broadcast satellite (“DBS”) providers, home satellite dishes (“HSD”), and broadband service providers (“BSPs”), as well as broadcast television licensees. We also examine other wireline video providers, including local exchange carriers (“LECs”), which have initiated commercial services using copper-based, fiber, and hybrid-fiber coaxial cable distribution technologies for video programming; open video systems (“OVS”); and electric and gas utilities.”).
51. Id. ¶¶ 150-63. 52. Id. ¶ 153 (citing Comcast Comments at 30, 34, 57-59; Comcast Reply at 20). 53. Id. ¶ 154 (citing Comcast Comments at 30).
-29-
NAVIGANT ECONOMICS
5102853
Internet video, providing consumers with an interactive alternative to traditional TV-set
viewing.”54 Based on its review of the Internet landscape, Comcast predicted that “video
delivered over the Internet is here to stay.”55 Although its economic experts now argue that
Internet video is a complement to cable television,56 in November 2006—before it sought to
acquire NBCU’s Internet properties—Comcast argued the opposite: “All of these modalities of
communications are important to younger consumers, all are part of the paradigm shift to a
‘what-you-want-when-you-want-it’ world, and all of them compete with traditional and not-so-
traditional video distribution technologies for time, attention, and dollars.”57 Comcast cannot
have it both ways.
2. The Relevant Upstream Markets
42. Because the proposed transaction would enable Comcast to engage in a host of
anticompetitive strategies, the relevant upstream markets must be defined with reference to a
particular type of conduct. The common denominator across all of the strategies, however, would
be to lessen competition in the MVPD market.
a. Denial of Regional Sports Programming
43. Regional sports programming is generally defined as the right to carry televised
professional regional sports events.58 Because fans generally follow their local team, regional
sports programming is not reasonably interchangeable with national sports programming (for
54. Comcast Comments at 29-30 (emphasis added). 55. Id. at 37. 56. See Israel & Katz Online Video, supra, at 24 (“Consumers use online viewing to supplement traditional
television viewing.”). 57. Comcast Comments at 59 (emphasis added). 58. See General Motors Corporation and Hughes Electronics Corporation, Transferors, and The News
Corporation Limited, Transferee, for Authority to Transfer Control, MB Dkt. No. 03-124, Memorandum Opinion and Order (released Jan. 14, 2004) [hereinafter Hughes-News Corp. MO&O], ¶ 133 (“The basis for the lack of adequate substitutes for regional sports programming lies in the unique nature of its core component: RSNs typically purchase exclusive rights to show sporting events, and sports fans believe that there is no good substitute for watching their local and/or favorite team play an important game.”).
-30-
NAVIGANT ECONOMICS
5102853
example, the NCAA men’s basketball tournament); nor is it interchangeable with another
market’s regional sports programming because a local subscriber’s taste for programming of a
local regional sports franchise cannot be satisfied by substituting programming of a distant
regional sports franchise.59
44. The FCC has recognized that denial of access to regional sports programming can
have important competitive implications in the provision of MVPD service (the downstream
market).60 Specifically, the FCC has recognized regional sports programming as a “must have”
input to the production of MPVD service. Moreover, as the FCC explained in its Adelphia Order,
“an MVPD’s ability to gain access to RSNs and the price and other terms [or] conditions of
access can be important factors in its ability to compete with [downstream MVPD] rivals.”61 The
FCC noted that Time Warner and Comcast “acknowledge that an MVPD that drops local sports
programming risks subscriber defections.”62 The FCC concluded that “lack of access to RSN
programming can decrease an MVPD’s market share significantly.”63
45. The FCC’s findings on the pivotal role of RSN access are supported by academic
research. Wise and Duwadi (2005) find that limited competitive access to regional sports
programming reduces DBS penetration and permits incumbent cable firms to demand inflated
59. Id. 60. See, e.g., FCC Adelphia Order, supra, at ¶ 68 (“In contrast, with respect to regional sports networks
(“RSNs”) and other regional networks, we conclude, as we did in the Comcast-AT&T and News Corp.-Hughes transactions, that the relevant geographic market is regional.”). Id. at ¶ 67 (“Nothing in the record suggests a need for us to define rigorously all possible relevant product markets for video programming networks. For purposes of our analysis, we will separate the video programming products by Comcast and Time Warner into two broad categories: (1) national cable programming networks and (2) regional cable networks, particularly regional sports networks.”) (emphasis added).
61. Id. ¶ 124 (“Hence, an MPVD’s ability to gain access to RSNs and the price and other terms of conditions of access can be important factors in its ability to compete with rivals.”).
62. Id. (“Applicants acknowledge that an MVPD that drops local sports programming risks subscriber defections and that MVPDs ‘will drive hard bargains to buy, acquire, defend or exploit regional sports programming rights.’”).
63. Id. ¶ 145 (“Lack of access to RSN programming can decrease an MVPD’s market share significantly. The Applicants [Time Warner and Comcast] have argued that DirecTV’s and EchoStar’s lack of access to CSN Philadelphia has not had a significant impact on DBS market share in Philadelphia and that DirecTV’s estimates of the effect are fatally flawed. We disagree.”).
-31-
NAVIGANT ECONOMICS
5102853
cable prices.64 According to their econometric model, DBS penetration is lower in markets where
cable operators carry regional sports channels.65 This effect, they believe, is explained by the fact
that cable operators sometimes discriminate against DBS providers in the provision of regional
sports content.66 The authors provide three ways that incumbent cable providers can impair DBS
competition: (1) by denying DBS competitors access to cable-affiliated regional sports
programming; (2) by denying DBS access to unaffiliated regional sports content through
exclusive contracts; or (3) by setting the terms of carriage so that DBS competitors find it
“uneconomical” to carry affiliated or unaffiliated regional sports programming.67
b. Denial of Local Broadcast Programming
46. The FCC has also recognized that local broadcast television programming
constitute a relevant product market. In its order approving News Corporation’s acquisition of
DirecTV, the FCC separated video programming into “three broad categories: (1) national and
local broadcast television programming.”68 The FCC found “substantial evidence in the record
that a temporary withdrawal of regional sports programming networks and local broadcast
64. Andrew Wise & Kiran Duwadi, Competition between Cable Television and Direct Broadcast Satellite—
The Importance of Switching Costs and Regional Sports Networks, 4 J. COMP. LAW & ECON. 679-705 (2005) [hereinafter Wise & Duwadi].
65. Id. at 702 (“We also find that DBS penetration is lower where cable operators carry regional sports channels. This is probably due to a combination of the factors discussed above. Two of the factors may involve cable operators limiting DBS operator access to regional sports networks.”).
66. Id. (“We also find that DBS penetration is lower where cable operators carry regional sports channels. This is probably due to a combination of the factors discussed above. Two of the factors may involve cable operators limiting DBS operator access to regional sports networks. If this is true, cable operators may be able to offset competitive pressures from DBS, and thus may be able to impose larger price increases without losing subscribers to DBS where they are able to transmit vertically integrated regional sports networks terrestrially, or are able to reach exclusive carriage agreements with non-vertically integrated regional sports networks.”).
67. Id. at 700 (“We, therefore, can think of three circumstances that may be contributing to reduced DBS penetration where cable operators carry regional sports networks. First, cable operators may be reducing DBS penetration by making unavailable to DBS providers affiliated regional sports networks transmitted terrestrially. Second, cable operators may be able to make unavailable to DBS providers non-vertically integrated regional sports networks, which are not covered by FCC program access rules, by signing exclusive carriage agreements. Third, the terms of the carriage agreements for some regional sports networks, either affiliated or unaffiliated with cable operators, may make them uneconomical for DBS providers to carry.”).
68. Hughes-News Corp. MO&O ¶ 60.
-32-
NAVIGANT ECONOMICS
5102853
television station signals would cause a significant number of customers to shift from their
current MVPD”,69 implying that local broadcast programming is another “must-have” input. It
also found that local broadcast station programming is “highly valued by consumers, and entry
into the broadcast station market is difficult.”70 DBS providers need access to local broadcast
stations to compete effectively with incumbent cable operators. In its 2002 annual video
competition report, the FCC found that DBS penetration had increased more rapidly in markets
where local-into-local service was available.71
c. Denial of National Sports Programming
47. National sports programming constitutes another relevant product market.72
Current suppliers of national sports cable networks include the family of ESPN networks, Golf
Channel, Versus, MLB Network, NBA TV, NHL Channel, Fox Soccer Channel, Fox College
Sports, Tennis Channel, CBS College Sports, GolTV, Speed Channel, and Horseracing
Television. The FCC has found regional sports networks to be a separate product from national
sports networks.73 The next closest substitute to national sports programming would be national
non-sports programming. But for the vast majority of viewers of national sports programming,
especially live-event coverage, generic national cable programming (whether it is a sit-com, a
news show, or a drama) would be an unacceptable substitute. For the same reason that the
regional sports networks are considered to be distinct from regional non-sports networks,
national sports networks should be considered distinct from national non-sports networks. The
only difference between the two is that an RSN carries sports with a local interest only (for
69. Id. ¶ 87 (emphasis added). 70. Id. ¶ 201. 71. See 2002 Video Competition Report, 17 FCC Rcd 26901, 26931-32 ¶ 61 (2002). 72. It bears noting that Comcast’s economists distinguish sports programming from non-sports programming
when assessing network profitability. See Israel & Katz Online Video, supra, ¶ 11. 73. Hughes-News Corp. MO&O ¶ 60.
-33-
NAVIGANT ECONOMICS
5102853
example, a regular season game of the Washington Wizards), whereas the national sports
network carries sports with a national interest (for example, the Masters or Tour de France or
Wimbledon).
d. Denial of Online Video Programming
48. Online video programming represents another relevant upstream market.74 Major
suppliers of online video programming include Hulu.com, CBS Interactive, Fox Interactive
Media, and YouTube.com. Online video is currently viewed via computers, handheld devices,
and televisions. Several cable networks, such as Tennis Channel or ESPN, replicate a portion of
their programming on websites. Downloadable movies on websites like Netflix.com are also
properly considered online video programming. By the end of 2009, the average online viewer
watched over nearly 13 hours per month of online video.
49. Recent developments in software and hardware are making it easier to view
online video via television, and OTT video providers are designing businesses around that
viewing method. Online video is sufficiently distinct from other types of online content, such as
news, email, and search to constitute its own product market. In particular, OTT video providers
that seek to assemble a portfolio of online video content would not perceive other types of online
content to be reasonable substitutes for online video. Moreover, consumers of OTT video
services would not be as likely to use their televisions to access email or perform Internet
searches, as those features typically require a keyboard and thus function more effectively on a
computer.
50. As is the case for cable network programming, some types of online video may be
considered “must-have” video content in the sense that, if denied to an OTT video provider or a
74. It bears noting that Comcast’s economists, Drs. Katz and Israel, analyze the proposed transaction’s likely
effects on the provision of “long-form, professional-quality video programming” via the Internet, which is largely consistent with my proposed market definition. See Israel & Katz Online Video, supra, ¶ 4.
-34-
NAVIGANT ECONOMICS
5102853
rival Internet service provider, the rival would be significantly impaired in its ability to compete
effectively in the supply of online video service. According to comScore, the two most popular
online video sites in February 2010 were YouTube.com and Hulu.com.75 With 912.5 million
videos viewed in February 2010, Hulu.com was second behind YouTube among all online video
properties. Because the average length of a Hulu.com video (for example, an episode of Modern
Family) exceeds the average length of a YouTube.com video (for example, a cat flushing a
toilet), this commonly used metric of ranking online video properties (by videos downloaded)
vastly understates the importance of Hulu.com. Other online video programming may not be
considered “must-have” on a standalone basis, but when combined with other types of online
programming, the resulting combination of programming may constitute a “must-have” input for
OTT video providers.
2. The Relevant Geographic Markets
51. The relevant geographic market also depends on the nature of the exclusionary
conduct under consideration.
a. Denial of Local Broadcast Programming or Regional Sports Programming
52. The relevant geographic market for analyzing the prospect of denial of a local
NBC affiliate or an RSN is the local level. When analyzing the competitive effects of
withholding Comcast-affiliated local programming, it is important to consider the
interdependence between upstream (input) markets and the downstream market (MVPD
services). Although the consistency of MVPD offerings (typically one cable operator and two
DBS providers) remains constant beyond the confines of a DMA, the demand for video
75. comScore Releases February 2010 U.S. Online Video Rankings, available at
programming—especially with respect to regional sports programming—varies by geography.
Because a vertically integrated cable operator like Comcast that controlled regional sports or
local broadcast programming in one region could not affect competitive outcomes by
withholding that content from MVPDs in another region, the relevant geographic market is
defined by the geographic boundary of the demand for said programming, which typically
coincides with a DMA. Accordingly, for analyzing this type of conduct, the geographic market
should be defined as the DMA.
53. A DMA-wide geographic market definition has also been widely adopted by other
parties that have analyzed the MVPD and regional sports programming markets. For example,
the FCC has conducted market analyses at the DMA level. In the Adelphia Order, the FCC
explained that DMAs are the appropriate market “to examine the geographic area in which
consumers are likely to place a similar value on the RSN programming at issue and to examine
the transactions’ impact in areas where viewers are likely to receive the same RSN
programming.”76 For this reason, the FCC concluded “we find it reasonable to define the relevant
geographic market for the analysis of harms concerning access to RSNs as any DMA that is
home to a sports team.”77 This geographic market definition is consistent with the FCC’s
analysis in the Hughes—News Corp. proceeding.78 In the same proceeding, the FCC also
76. FCC Adelphia Order, supra, ¶ 126 (“Using the DMA allows us here, as we did in News Corp.-Hughes, to
examine the geographic area in which consumers are likely to place a similar value on the RSN programming at issue and to examine the transactions’ impact in areas where viewers are likely to receive the same RSN programming.”).
77. Id. ¶ 125 (“Because individual DMAs usually are entirely encompassed within the authorized viewing zone for a team’s games and contain those fans that value its programming most highly, we find it reasonable to define the relevant geographic market for the analysis of harms concerning access to RSNs as any DMA that is home to a sports team.”).
78. Hughes-News Corp. MO&O, ¶ 66 (“In contrast, with respect to RSNs, we conclude, as we did in the Comcast-AT&T merger, that the relevant geographic market for RSNs is regional. In general, contracts between sports teams and RSNs limit the distribution of the content to a specific “distribution footprint,” usually the area in which there is significant demand for the specific teams whose games are being transmitted. MVPD subscribers outside the footprint thus are unable to view many of the sporting events that are among the most popular
-36-
NAVIGANT ECONOMICS
5102853
concluded that “in the case of broadcast television programming, it is reasonable to use DMAs to
define the relevant geographic market for each individual broadcast station.”79
54. The relevant DMAs to analyze here are the eleven local markets in which
Comcast owns a RSN plus the ten local markets in which Comcast will acquire an O&O NBC
local affiliate. As it turns out, there is significant overlap between those two areas. As of the
release of the last Video Competition Report, Comcast owned eleven RSNs: SportsNet Bay
Area, SportsNet California, SportsNet Chicago, SportsNet Mid-Atlantic, SportsNet New
Sports Southeast, SportsNet New York, and Mountain West SportsNet.80 NBCU owns and
operates ten local television stations that broadcast NBC Television Network programming in the
following markets: New York; Los Angeles; Chicago; Philadelphia; Dallas-Ft. Worth; San
Francisco; Washington, D.C.; Miami-Ft. Lauderdale; San Diego; and Hartford-New Haven.81 It
bears noting that Comcast does not own marquee regional sports programming in each DMA in
which it owns an RSN. For example, in the Baltimore DMA, the exclusive content on Comcast
SportsNet Mid-Atlantic (Washington Wizards basketball and Washington Capitals hockey) is not
likely considered to be “must-have” programming. Indeed, in recognition of this possibility,
Comcast has recently rebranded (sometime after 2009) Comcast SportsNet Mid-Atlantic to
Comcast SportsNet Washington.82
programming offered by RSNs. We thus find it reasonable to define the relevant geographic market as the “distribution footprint” established by the owner of the programming.”).
79. Id. ¶ 65. 80. Annual Assessment of the Status of Competition in the Market for the Delivery of Video Programming, 81. Merger Application, at 29. 82. Comcast SportsNet Washington, available at http://en.wikipedia.org/wiki/Comcast_SportsNet_Washington
(“formerly called Comcast SportsNet Mid-Atlantic (2001-2009)”).
-37-
NAVIGANT ECONOMICS
5102853
b. Denial of National Sports Programming
55. As in the case of regional sports programming, the relevant geographic market is
defined by the geographic boundary of the demand for national sports programming, which is the
nation. National sports networks are generally licensed to MVPDs nationwide. And the demand
for national sports programming does not vary significantly by geography. A vertically
integrated cable operator in a given region that controlled marquee national sports programming
could impair competition by withholding that content from MVPDs in another region.
Accordingly, for analyzing this type of conduct, the geographic market should be defined as the
nation.
c. Denial of Online Video Programming
56. The relevant geographic market to analyze the competitive effects of a denial of
online video programming to an OTT video provider or to a rival ISP is the nation. Online video
programming is available to Internet subscribers nationwide, and the demand for such content
does not vary significantly by geography. A vertically integrated cable operator in a given region
that controlled marquee online video programming could impair competition by withholding that
content from OTT video providers or ISPs in another region. Accordingly, for analyzing this
type of conduct, the geographic market should be defined as the nation.
B. Measures of Comcast’s Market Power in the Supply of MVPD Service and in the Purchase of Video Programming
57. Market power is defined as the ability to raise prices above competitive levels or
to exclude rivals. Market power can be proven directly, through evidence of power over price,83
83. Comcast’s monthly per subscriber cable television margins of {{$35.62}} are certainly consistent with the
notion of significant power over price. Israel & Katz Online Video, supra, at ¶ 104. Economists consider division-specific profit margins, as opposed to firm-wide accounting profits, as evidence of market power. See e.g., Kevin Kreitzman, Melanie Williams, Michael A. Williams, & William Havens, Estimating Monopoly Power with Economic Profits, UC DAVIS BUSINESS LAW JOURNAL (forthcoming 2010) (showing that the degree of a firm’s market power in a given market is embodied in its cash flows and can be measured by its economic profits and
-38-
NAVIGANT ECONOMICS
5102853
or indirectly, through high market shares in a relevant product market combined with evidence of
barriers to entry. In economic parlance, a firm with market power faces a downward-sloping
residual demand curve, which means that the firm can unilaterally affect the supply of goods in
the market by increasing or decreasing output and can, as a result, generate monopoly profits by
raising prices and reducing output.84
1. Market Shares
58. Comcast is the largest MVPD provider in United States. The relevant geographic
market over which to assess Comcast’s market power in the supply of and demand for national
sports programming or online video programming is the nation. With nearly 24 million basic
cable subscribers by the third quarter of 2009,85 Comcast provides cable television service to
approximately one quarter of all MVPD subscribers in the country.86
59. To the extent that Comcast coordinates its purchasing and supply decisions for
national sports programming and online video programming with other cable operators,
Comcast’s “foreclosure share” exceeds its nationwide MVPD market share. The foreclosure
share is the percentage of a market that can be foreclosed to rivals. There are five independent
types of evidence that suggest that Comcast is in fact coordinating its strategy vis-à-vis video
programming, including online video programming, with out-of-region cable operators.
economic rate of return). The {{$35.62}} in monthly video margin per subscriber is equal to {{$71.00}} in monthly revenue per subscriber less {{$35.37}} in monthly variable costs. The traditional Lerner index, a measure of the deviation of prices from variables costs, for Comcast’s markup is {{50.1}} percent. If Comcast’s cable video service were perfectly competitively supplied, the monthly price would be {{$35.37}}. Comcast’s ability to charge nearly {{double}} that amount indicates significant power over price.
84. CARLTON & PERLOFF, supra, at 200 (“If a firm faces a downward-sloping demand curve, it has market power. […] If consumers view brands in an industry as imperfect substitutes, a firm may raise its price above that of its rivals without losing all its customers.”). Id. at 642 (“A firm (or group of firms acting together) has market power if it is profitably able to charge a price above that which would prevail under competition, which is usually taken to be marginal cost.”).
85. Comcast Reports Third Quarter 2009 Results, Nov. 4, 2009, at 3. 86. According to the most recent data available from the Commission, there were nearly 96 million MVPD
subscribers nationwide in 2006. Thirteenth Annual Report, at Appendix B Table B-1.
-39-
NAVIGANT ECONOMICS
5102853
During the trial between Comcast and the NFL, former NFL Commission Paul Tagliabue testified that not only did Brian Roberts threaten to re-tier NFL Network if the NFL did not supply the Thursday night and Saturday night games on an exclusive basis to Versus, but that the larger “cable industry” would support Comcast’s refusal to deal.87
Comcast, Cox, and Time Warner coordinate their strategies vis-à-vis national movie studios such as DreamWorks SKG, Universal, and Columbia TriStar via iN DEMAND’s pay-per-view (“PPV”) service. PPV programming on iN DEMAND includes “movies, boxing and mixed martial arts events and the digital out-of-market subscription professional sports packages for MLB, the NBA, the NHL and MLS,” as well as “first-run movies and serves as the exclusive TV home of Howard Stern via its Howard TV On Demand offering.”88
TV Everywhere’s authentication service, which facilitates the coordination of strategies vis-à-vis national video content providers and OTT video services, was jointly conceived by Comcast and Time Warner.89 Time Warner recognized that the key to TV Everywhere’s success was persuading the rest of the cable industry to join up; in the absence of coordination, viewers could substitute to websites that did not require a cable subscription.90
Empirical research indicates that vertically integrated cable operators coordinated their carriage decisions with respect to independent programming.91 Kang’s econometric model produced empirical results that “make credible an underlying premise of a 30 percent national market share limit that the Federal Communication Commission established in 1993: namely, that MSOs may tacitly collude in their carriage decisions, having the effect of restricting market access to startup cable networks in which those MSOs have no ownership interest.”92
A U.S. district court recently certified a class of Philadelphia-based Comcast cable subscribers, who allege that Comcast entered into a series of swaps with other (out-of-region) cable operators so that each provider would have exclusive “clusters” of markets in violation of sections 1 and 2 of the Sherman Antitrust Act.93
87. See Transcript of Record, NFL Enterprises LLC v. Comcast Cable Communications LLC, File No. CSR-
7876-P, Apr. 16, 2009, 1277: 10-1279:10 (Paul Tagliabue testimony describing Comcast CEO Brian Roberts’ suggestion that the NFL’s relationship with the “cable industry” would not be “positive” on a going-forward basis.)
88. About inDEMAND, available at http://www.indemand.com/about/. 89. See, e.g., Anthony Crupi, TW, Comcast Prep ‘TV Everywhere’ Push, ADWEEK, June 24, 2009, available at
http://www.adweek.com/aw/content_display/news/media/e3i048f01beefa084a367ab3330d9e79e95 (“Time Warner today said it has partnered with Comcast to develop a cohesive strategy for its "TV Everywhere" initiative, which looks to reinforce the subscription TV model by allowing subscribers to access cable network programming on-demand, via broadband and mobile platforms.”).
90. See, e.g., Revenge of the Cable Guys, supra. 91. See Jun-Seok Kang, Reciprocal Carriage of Vertically Integrated Cable Networks: An Empirical Study,
Indiana University Working Paper, August 30, 2005. 92. Id. at 1. 93. See Behrend v. Comcast Corp., 2010 U.S. Dist. LEXIS 1049 (E.D. Pa. Jan. 7, 2010).
-40-
NAVIGANT ECONOMICS
5102853
To use a simple example, to the extent that Comcast and Time Warner coordinate in their refusal
to carry independent national sports networks or to threaten cable networks or movie studios that
post their content online, then the foreclosure share increases from approximately 25 to 37
percent (equal to Comcast’s 25 percent share plus Time Warner’s 12 percent share94); if the
coordination involves the entire “cable industry,” as intimated by Brian Roberts in his
conversations with Paul Tagliabue, then the foreclosure share increases to over 60 percent.95
Even if Comcast is not coordinating with other cable operators, Comcast’s share of national
MVPD subscribers (25 percent) is sufficiently large to create a presumption of anticompetitive
effects under the antitrust laws.96
60. The relevant geographic market over which to assess Comcast’s market power in
the supply of regional sports programming is the DMA in which Comcast owns an RSN.
Similarly, the relevant geographic market over which to assess Comcast’s market power in the
supply of local broadcast programming is the DMA in which NBCU owns an NBC broadcast
network. As it turns out, Comcast is the dominant MVPD provider in the majority of the DMAs
in which NBCU owns and operates an NBC affiliate. For example, Comcast serves
approximately 70 percent of MVPD subscribers in the Philadelphia DMA, and it serves
approximately 60 percent of MVPD subscribers in the Chicago, Miami, and San Francisco
DMAs. In these and other DMAs, Comcast’s market share is clearly above levels typically
associated with monopoly power (assuming high barriers to entry). Table 1 shows Comcast’s
share in the ten DMAs in which NBCU owns and operates a local NBC broadcast network.
94. Thirteenth Annual Report, at Table B-3. 95. NCTA Industry Data, available at http://www.ncta.com/Statistics.aspx (showing 62.1 million cable
subscribers as of December 2009). 96. See PHILLIP AREEDA, IX ANTITRUST LAW 375, 377, 387 (Aspen 1991) (indicating that 20 percent
foreclosure is presumptively anticompetitive); See also HERBERT HOVENKAMP, XI ANTITRUST LAW 152, 160 (indicating that 20 percent foreclosure and an HHI of 1800 is presumptively anticompetitive).
-41-
NAVIGANT ECONOMICS
5102853
TABLE 1: COMCAST’S MARKET SHARE FOR THE 10 DMAS IN WHICH NBCU OWNS AND OPERATES A LOCAL NBC AFFILIATE AS OF 2010
DMA SNL Kagan* Warren’s** TVB***
1 Philadelphia, PA 68.5% 68.9% 74.4%
2 Chicago, IL 63.2% n/a*** 62.8%
3 Miami-Fort Lauderdale, FL 61.2% 50.9% 67.7%
4 San Francisco-Oakland-San Jose CA 59.3% 73.4% 68.5%
5 Washington, DC (Hagerstown, MD) 48.0% 44.7% 71.1%
6 Hartford & New Haven, CT 40.7% 49.6% 76.8%
7 New York, NY 9.8% 11.1% n/a
8 Dallas-Fort Worth, TX 0.0% 0.0% n/a
9 Los Angeles, CA 0.0% 0.0% n/a
10 San Diego, CA 0.0% 0.0% n/a
Notes and Sources: Percentages represent Comcast share of “MVPD Subscribers,” in late 2009. * SNL Kagan/MediaCensus Competitive Intelligence as of fourth quarter 2009. ** Comcast basic subscriber data available from Advanced TVFactbook, Warren’s Communications News, May 2010. Data on DMA Cable/ADS penetration taken from TVB Local Cable Reach Guide Feb. 2010, Television Bureau of Advertising, available at: http://tvb.org/nav/build_frameset.aspx. *** Percentages represent Comcast share of “Local Market Interconnects” as a percentage of HHs reached by Cable/ADS. Data on DMA Cable/ADS penetration taken from TVB Local Cable Reach Guide Feb. 2010, Television Bureau of Advertising, available at: http://tvb.org/nav/build_frameset.aspx. *** Data for Chicago, IL DMA from the Advanced TVFactbook was incomplete. ^ Non-overlapping markets.
As Table 1 shows, Comcast serves over 40 percent of the market in six of the ten markets in
which NBCU owns and operates an NBC affiliate according to the Television Bureau of
Advertising.
2. Barriers to Entry
61. Because monopoly power is the ability to engage profitably in substantial and
sustained supra-competitive pricing, a finding of high market shares with evidence of barriers to
entry supports a conclusion of monopoly power. In the absence of such barriers, a price increase
above the competitive level may invite entry sufficient to make that price increase unprofitable.
62. In any given MVPD market, entrants face significant barriers. One barrier to entry
in the Philadelphia DMA has been a legal barrier—the franchise process. RCN spent more than
two-and-a-half years attempting to obtain a franchise agreement from the City of Philadelphia
and, tellingly, never received a franchise. Verizon also faced delays when it applied for a video
-42-
NAVIGANT ECONOMICS
5102853
franchise in the City of Philadelphia.97 Legal measures, such as patents and franchises, have been
widely recognized by economists as significant barriers to entry.98 Furthermore, as described
below, these legal barriers to entry have been bolstered by Comcast’s actual conduct in response
to its competitors’ attempts to gain local franchises.
63. Physical barriers to entry—in the form of large fixed costs—also exist in the
MVPD market in the Philadelphia DMA. These fixed costs, which are incurred for any level of
production, imply that the provision of MVPD services is characterized by economies of scale.
As the American Bar Association explained in its treatise on antitrust in telecommunications
markets:
Economies of scale also play an important role in analyzing the ease of entry into the communications industry. Barriers to entry created by large capital outlays required in many segments of the industry also create strong economic incentives to build an economy of scale. As Judge Posner has noted, the costs of building a cable television grid—e.g., laying cable on all major streets—are both huge and ‘invariant to the number of subscribers a system has.’ Any operator must build the grid, and, once that is done, the cost of adding another subscriber by connecting the grid to his home is relatively small.99
Such economies of scale are generated by the fixed costs cable operators must incur to establish
a cable system. In its Thirteenth Annual Report on MVPD competition, the FCC noted the
substantial investments cable companies must make in their systems. For instance, cable
operators report that they have invested over $100 billion to construct advanced two-way fiber
optic networks,100 which can cost from $100,000 to $300,000 per mile.101 In addition, the FCC
reports that cable companies indicate that they spent $10.6 billion on capital improvements in
97. See Bob Fernandez, Politics, Comcast slow Verizon’s pay-TV bid (Philadelphia Inquirer, Dec. 14, 2008), available at http://www.philly.com/philly/business/36120899.html.
98. See, e.g., CARLTON & PERLOFF at 77 (recognizing that “a good example of a long-run barrier to entry is a patent.”). See also Ex Parte Submission of the Department of Justice, MB Docket No. 05-311, available at http://www.usdoj.gov/atr/public/comments/216098.htm.
99. See AMERICAN BAR ASSOCIATION SECTION OF ANTITRUST LAW, TELECOM ANTITRUST HANDBOOK 97 (American Bar Association 2005).
100. Thirteenth Annual MVPD Report, supra at ¶52. 101. Jeremy Feiler, RCN Out to Block Comcast, PHILADELPHIA BUSINESS JOURNAL, Aug. 16, 2002 (“RCN’s
business is capital-intensive – installing fiber-optic or coaxial cable can cost $100,000 to $300,000 per mile – and it has halted its expansion.”).
-43-
NAVIGANT ECONOMICS
5102853
2005 and an estimated $11.1 billion in 2006.102 These reported expenditures indicate that an
entrant must undertake significant upfront investments in cable plant and facilities before
entering the MVPD market.
3. Ability to Exclude Rivals
64. Pursuant to a nationwide strategy of clustering local cable franchises, Comcast
assembled several clusters that allowed it to control entire DMAs, so that it could exercise
monopoly power over customers and other critical inputs needed by rival MVPDs. Having done
so, it was then economically and commercially feasible for Comcast to (a) require that its
contractors not work for RCN (an overbuilder, or cable firm that “overbuilds” and seeks to
compete for households already served by an incumbent MSO) as a condition of working for
Comcast; and (b) require that the regional sports programmer (in which Comcast owned a
controlling interest) not make its programming available for DirecTV and Dish Network. The
ability to exclude rivals is another way to demonstrate market power. In this section, I review
two means by which Comcast has thwarted entry by MVPD rivals.
a. Withholding Critical Local Inputs
65. Comcast has actively sought to deny its competitors—such as overbuilders like
RCN and DBS providers like DirecTV and EchoStar—access to critical local inputs such as RSN
programming and contractors.
i. Regional Sports Networks
66. The FCC has recognized Comcast’s incentive to use its dominance in the
upstream RSN market to impair competition in the downstream MVPD market. In assessing the
potential anticompetitive effects of the joint acquisition of Adelphia by Time Warner and
Comcast, the FCC found that the acquisition would increase the likelihood of harm to MVPDs in
102. Thirteenth Annual MVPD Report, supra at ¶52.
-44-
NAVIGANT ECONOMICS
5102853
markets in which Time Warner or Comcast hold, or have the potential to hold, an ownership
interest in an RSN.103 The FCC also found that Time Warner and Comcast would gain an
incentive and increased ability to deny carriage to unaffiliated RSNs.104 The FCC concluded that
even small increases in the market share of Comcast or Time Warner would increase the firm’s
incentives to increase prices for affiliated RSNs.105 In particular, the FCC found that a uniform
price increase was likely to occur in fifteen of the 39 “key” geographic markets known as
DMAs.106 The Commission also provided a recognized link between RSN ownership and MVPD
competition. Specifically, the FCC noted that DBS penetration levels are significantly lower in
Philadelphia, where DBS operators—such as DirecTV and Dish Network—cannot offer the local
RSN to their subscribers.107
67. There appears to be a direct relationship between Comcast’s share of households
and instances where Comcast discriminates against DBS providers. Specifically, Comcast
engages in discrimination against some unaffiliated MVPDs in every market in which (1) it owns
the sort of marquee sports content to make such discrimination worthwhile and (2) it supplies
cable service to at least 35 percent of the households within the DMA.108 The 35 percent
103. FCC Adelphia Order, supra ¶116. 104. Id. 105. Id. ¶141. 106. Id. ¶144 (“Key DMAs are DMAs that are home to professional sports teams that play in one of the four
major U.S. sports (football, baseball, basketball, and hockey). These DMAs are most likely to be within the “inner zone” of where sports programming is most popular. Therefore, these DMAs are the most susceptible to subscriber losses if the RSN is withheld. We find a potential for an increase in the RSN’s affiliation fee of at least five percent in 15 of the 39 key DMAs. These DMAs are Atlanta, Boston, Buffalo, Charlotte, Cincinnati, Cleveland, Columbus, Dallas, Jacksonville, Los Angeles, Miami, Minneapolis, Pittsburgh, San Diego, and Washington. In these DMAs, a uniform price increase is likely to extract at least an additional $4.2 million per market in RSN fees from unaffiliated MVPDs under conservative assumptions in our model.”).
107. Id. ¶146. In the cities where the local RSN is not available to DBS subscribers (Philadelphia, San Diego, and Charlotte), the FCC’s regression analysis shows a statistically significant drop in market share in Philadelphia and San Diego. Id. ¶149. (“We find that the percentage of television households that subscribe to DBS service in Philadelphia is 40% below what would otherwise be expected given the characteristics of the market and the cable operators in the DMA.”).
108. Hal J. Singer & J. Gregory Sidak, Vertical Foreclosure in Video Programming Markets: Implication for Cable Operators, 3 REV. NETWORK ECON. 348 (2007).
-45-
NAVIGANT ECONOMICS
5102853
threshold may represent the critical share necessary to profitably own an RSN and withhold
regional sports programming from competing MVPDs.
TABLE 2: TOP 30 LOCAL MARKETS IN WHICH COMCAST OWNS A RSN Market (DMA) Affiliated RSN Comcast Subs as %
of Total Households in DMA (Before
Adelphia Merger)
Comcast Subs as % of Total Households
in DMA (After Adelphia Merger)
Does Comcast Own Marquee Sports
Content?
Discriminate Against
Unaffiliated MVPD
Orlando Comcast/Charter Sports Southeast
5 8 No NA
Tampa Comcast/Charter Sports Southeast
10 10 No NA
Atlanta Comcast/Charter Sports Southeast; BravesVision
29 32 No NA
Washington SportsNet MidAtlantic
28 38 Yes No
Sacramento SportsNet West 35 35 Yes Yes
Miami Comcast/ Charter Sports Southeast
37 42 No NA
Philadelphia SportsNet Philadelphia
58 60 Yes Yes
Baltimore SportsNet MidAtlantic
53 56 No NA
Detroit Comcast Local 48 48 No NA
Chicago SportsNet Chicago 49 49 Yes Yes
Note: Reproduced from Singer and Sidak (2007).
Although the exact share of total households required to make discrimination vis-à-vis rival
MVPDs profitable is difficult to ascertain, based on the pattern contained in Table 2, it is
reasonable to infer that the “critical share” is somewhere between 28 percent (pre-merger
Washington DMA) and 35 percent (pre-merger Sacramento DMA). Comcast does not own
marquee sports content in six of these ten DMAs: Miami, Atlanta, Tampa, Orlando, Detroit, or
Baltimore.109 However, Comcast’s experience in the other three DMAs (not counting
Washington) in Table 2 demonstrates that Comcast would discriminate against DBS providers
109. BravesVision carries some live Atlanta Braves baseball games in high-definition that are also carried on
other RSNs (TBS and Turner South), although those RSNs do not carry the games in high-definition. See R. Thomas Umstead, Comcast, Braves Create HD Net; Regional Could Serve as Template for Other Dedicated Team Channels, MULTICHANNEL NEWS, Sept. 27, 2004, at 60; BravesVision Suits Up for Season, MULTICHANNEL NEWS, Apr. 1, 2005, at *1. Because these games are available on other RSNs not affiliated with Comcast, Atlanta is labeled “N/A” in Table 1. Even though neither RSN is carried by a DBS provider, Comcast’s content on both BravesVision and CSS is not sufficient to be labeled “discrimination” because neither RSN carries exclusive marquis content.
-46-
NAVIGANT ECONOMICS
5102853
once (1) Comcast secures the rights to marquee sports content and (2) establishes a sufficiently
large downstream footprint.
68. Multiple econometric studies have estimated the decrease in DBS penetration in
the Philadelphia DMA that can be attributed to Comcast’s exclusionary conduct. For example,
on behalf of DirecTV, Bamberger and Neumann estimate DBS penetration in each of the 210
DMAs throughout the country as a function of many demographic and economic characteristics
of the DMAs.110 Using this model, Bamberger and Neumann predict that, given its market
characteristics, DBS penetration in Philadelphia should have been 20.9 percent in 2005 (rather
than the actual 10.4 percent). The competitive effect of Comcast’s refusal to provide CSN-
Philadelphia was also documented by Robert Willig and Jonathan Orszag.111 Like Bamberger
and Neumann, Willig and Orszag estimate the extent to which Comcast’s denial of CSN-
Philadelphia reduced DBS penetration in Philadelphia. Willig and Orszag extended this
foreclosure analysis by estimating the size of the incremental profits that Comcast extracted as a
result of its RSN foreclosure strategy. Willig and Orszag calculated that Comcast’s conduct
caused Dish to win 190,000 fewer subscribers in the Philadelphia DMA than it would have
otherwise. Bamberger and Neumann refined and updated these initial findings with a second
study provided to the Commission in March 2006.112 In both March 2005 and December 2005,
110. Lexecon, Analysis of Effect of RSN Availability on DBS Penetration, at Appendix Table 1 (attached as
Appendix A to Applications of Adelphia Communications Corp., Comcast Corp., and Time Warner Cable Inc. for Authority to Assign and/or Transfer Control of Various Licenses, MB Dkt. No. 05-192, DIRECTV Surreply, Oct. 12, 2005) [hereinafter Lexecon October 2005 Analysis].
111. See Redacted Letter from David K. Moskowitz, Executive Vice President and General Counsel, EchoStar Satellite L.L.C. to Marlene H. Dortch, Secretary, FCC, MB Dkt. No. 05-192 (filed Jan. 25, 2005) (citing the redacted Willig/Orszag study) [hereinafter Willig & Orszag].
112. Gustavo Bamberger and Lynette Neumann, Updated Analysis of Effect of RSN Availability on DBS Penetration, Mar. 17, 2006 (attached as Exhibit 1 to Letter from William M. Wiltshire et al., counsel to DIRECTV, to Marlene H. Dortsch, Secretary, FCC, Re: Ex Parte Presentation in MB Dkt. No. 05-192, Mar. 17, 2006) [hereinafter Bamberger and Neumann March 2006 Analysis].
-47-
NAVIGANT ECONOMICS
5102853
their model estimates that the actual DBS penetration rate in Philadelphia is approximately 10
percentage points smaller than it should be based on Philadelphia’s characteristics.
69. On behalf of Dish Network, Willig and Orszag calculate this number by
“comparing its penetration in the Philadelphia DMA…to the average penetration in other DMAs
where Echostar has local-into-local service (as in Philadelphia) but also carries professional
sports.”113 Willig and Orszag noted that there were approximately 2.8 million MVPD subscribers
in the Philadelphia DMA at that time, implying that Dish’s penetration rate was reduced from 9.5
percent to 3 percent as a result of Comcast’s conduct.114 Note that this reduction does not
estimate the total effect of Comcast’s conduct on DBS penetration because it does not include
the effect of Comcast’s conduct on DirecTV’s penetration rate. Willig and Orszag provided a
supplementary analysis that includes both the extent to which Comcast’s conduct reduced DBS
penetration and the extent to which Comcast charged higher expanded basic cable rates as a
result. Willig and Orszag explicitly link Comcast’s refusal to provide its DBS rivals with CSN-
Philadelphia to Comcast’s exercise of market power by charging supra-competitive rates for its
expanded basic cable service.
70. The FCC produced its own econometric study to examine whether DBS
penetration was unusually low in Philadelphia.115 Consistent with the DirecTV and Dish
Network studies, the FCC’s analysis indicated that, if DBS providers in Philadelphia had access
to Comcast SportsNet, then DBS penetration would be approximately six percentage points
113. Id. at 3 (“In Philadelphia, Comcast has been able to deny Echostar (and DIRECTV) access to the regional
sports that it controls by transmitting the programming terrestrially to its own headends and thereby avoiding the exclusivity prohibition of the Communications Act. The result? Almost 190,000 subscribers lost. Echostar has estimated this loss by comparing its penetration in the Philadelphia DMA, which was extremely low at about 3% as of November 2003, to the average penetration in other DMAs where Echostar has local-into-local service (as in Philadelphia) but also carries professional sports – about 9.5% at the same point in time. The loss of about 70% of the expected penetration rate, applied to the Philadelphia DMA population of 2.8 million television households points to a loss of 188,000 subscribers.
114. Id. 115. FCC Adelphia Order, supra.
-48-
NAVIGANT ECONOMICS
5102853
greater.116 The FCC updated and revised its Adelphia regression analysis in the 2007 Program
Access Order. It again found that DBS penetration in Philadelphia was significantly below the
predicted level and pointed to “empirical evidence that [Comcast’s] withholding” of SportsNet
programming from DBS providers in Philadelphia “has had a material adverse impact on
competition in the video distribution market.”117
71. Using the regression coefficients and the demographic and economic
characteristics of Philadelphia, Singer and Sidak estimate that DBS penetration in the
Philadelphia DMA should be 15.4 percent.118 The 95 percent confidence interval around our best
prediction is 13.1 to 17.8 percent. Because the actual DBS penetration rate in the Philadelphia
DMA is outside the 95 percent confidence interval, one must reject the hypothesis that
DBS penetration rate is less than what one would expect given its characteristics, which
demonstrates that Comcast’s foreclosure strategy has reduced the DBS penetration rate in the
Philadelphia DMA.
ii. Local Contractors
72. Comcast also sought to interfere with RCN’s efforts to construct systems in the
suburbs by limiting RCN’s access to local contractors. According to an antitrust complaint filed
on behalf of Comcast’s subscribers in Philadelphia, RCN relied upon construction and
installation contractors to deploy, operate, and maintain its competing infrastructure.119 Comcast
116. Id. Appendix D ¶ 18. 117. FCC NPR ¶¶ 39; see id. ¶ 115 (noting findings in Adelphia Order that withholding of SportsNet in
Philadelphia “has had a materially adverse effect on competition in the video distribution market”). 118. J. Gregory Sidak & Hal J. Singer, Vertical Foreclosure in Video Programming Markets: Implication for
Cable Operators, 3 REVIEW OF NETWORK ECONOMICS 348 (2007). 119. Third Amended Complaint, Caroline Behrend et al. v. Comcast Corporation et al., No. 03-6604 (E.D.
Pa.), ¶91.
-49-
NAVIGANT ECONOMICS
5102853
entered into and enforced non-compete clauses with its Philadelphia-area contractors.120 Comcast
also threatened its contractors with a loss of work in the event that they performed services for
Comcast’s competitors.121 Comcast’s conduct in the Philadelphia DMA has foreclosed
overbuilders—most notably RCN—from access to contractors that can install and maintain a
competitive wireline system. Altogether, RCN indicates that Comcast has prevented or attempted
to prevent fifteen Philadelphia-area contractors from doing business with RCN.122 According to
RCN, these contractors represent “virtually all of the viable construction and installation
contractors in the area.”123
b. Control over Local Franchise Authorities
73. Comcast has successfully lobbied local franchise authorities (“LFAs”) to delay or
deny entry of MVPD rivals. Comcast’s activities vis-à-vis RCN and Verizon in the Philadelphia
DMA have been reported widely in the press. Comcast’s ability to use LFAs as an entry barrier
is evidence of Comcast’s market power. Although such activities are protected as free speech,
they reveal Comcast’s market power (the ability to exclude rivals) and its anticompetitive intent.
74. RCN first attempted to enter Philadelphia in 1998. On June 8, 1998, the FCC
announced that it was reviewing RCN’s application to provide Open Video System (OVS)
service to Philadelphia.124 Established under the Cable Act of 1996, OVS operators are exempt
from licensing and build-out requirements as long as they provide video transmission capacity to
120. Id. 121. Id. 122. See, e.g., Petition of RCN Telecom Services, Inc., To Deny Applications or Condition Consent,
Applications for Consent to the Transfer of Control of Licenses from Comcast Corporation and AT&T Corp., Transferors, to AT&T Comcast Corporation, Transferee, MB Docket No. 02-70 (FCC filed Apr. 29, 2002) at 17 (“RCN is aware of no less than fifteen (15) contractors in the Philadelphia market – representing virtually all of the viable construction and installation contractors in the area – whom Comcast or, prior to its acquisition by Comcast, Suburban Cable, have prevented or tried to prevent from doing business with RCN.”).
any requesting unaffiliated programmers.125 As an OVS applicant, RCN received similar
treatment and assumed similar obligations.126 At the conclusion of the statutory 10-day review of
RCN’s application, the FCC granted RCN approval to compete with cable operators in
Philadelphia and the surrounding counties as an OVS operator.127
75. RCN planned to invest heavily in a new cable system in the Philadelphia DMA.
On June 5, 1998, RCN filed an application with the FCC seeking to build and operate a 330-
channel OVS system in 109 Philadelphia-area communities, including the City of
Philadelphia.128 The FCC approved this request on June 15, 1998.129 On October 2, 1998, RCN
filed another document with the FCC asserting its intent to build an OVS cable system covering
the same 109 Pennsylvania communities listed in the FCC’s Order.130 In its initial build-out
design, RCN planned to construct a $250 million system in the Northeastern and Northwestern
sections of the City of Philadelphia.131 In preparation for its entry into the Philadelphia DMA,
125. See FCC, Implementation of Section 302 of the Telecommunications Act of 1996, Open Video Systems,
Order on Remand, CS Dkt. No. 96-46, rel. Nov. 19, 1999 (discussing the development of OVS licensing and adopting a modified OVS rule).
126. Id. Unaffiliated programmers would have been able to sign up for RCN’s network as OVS programmers. See also Ted Hearn, ‘Open Video Systems’ A Turn Off, MULTICHANNEL NEWS, Feb. 27, 2006.
127. FCC OKs RCN in Philly, S.F., MULTICHANNEL NEWS, Jun. 22, 1998. 128. See RCN of Philadelphia, FCC OVS Application, undated, on file with author. See also In the Matter of
RCN Telecom Services of Philadelphia, Inc, Certification to Operate an Open Vide System, Memorandum Opinion and Order, released June 15, 1998 [hereinafter FCC Philadelphia OVS Decision], at ¶ 1 (“On June 5, 1998, RCN Telecom Services of Philadelphia, Inc., a wholly owned subsidiary of RCN Corporation (“RCN”), filed an application for certification to operate an open video system pursuant to Section 653(a)(1) of the Communications Act of 1934 […] and the Commission's rules. RCN-Philadelphia seeks to operate an open video system in the City of Philadelphia and in the Counties of Buck, Chester, Delaware, and Montgomery, Pennsylvania.”).
129. FCC Philadelphia OVS Decision, supra at ¶ 12 (“Accordingly, IT IS ORDERED, that the certification application of RCN Telecom Services of Philadelphia, Inc. d/b/a RCN of Philadelphia to operate an open video system in the communities of: [109 Philadelphia-area communities listed].”).
130. See In the Matter of RCN Telecom Services of Philadelphia, Inc., Notice of Intent to Establish an Open Video System, Oct. 2, 1998, on file with author.
131. Princeton, N.J., Cable-TV Firm Withdraws Proposal for Philadelphia Network, WORLD REPORTER, Feb. 15, 2001 [hereinafter RCN Withdraws].
-51-
NAVIGANT ECONOMICS
5102853
RCN employees walked out 3,900 miles of fiber optic cable.132 RCN also mapped and digitized
more than 2,100 miles of the area and constructed a head end in neighboring Valley Forge.133
76. Despite obtaining approval from the FCC and from nine Philadelphia suburbs,134
RCN could not obtain approval from the Philadelphia City Council to begin construction in the
city proper. RCN began discussions with city authorities in approximately June 1998.135 In
contrast with the FCC, which reviewed RCN’s application in ten days, the City of Philadelphia
took an entire year just to produce an application for RCN to complete.136 With the delay, RCN’s
application process took months longer to process than other cities.137 Even after RCN obtained
the necessary application documents, it faced delays imposed by the city administration. For
example, Philadelphia Mayor Ed Rendell told RCN that approval would be unlikely before the
end of 1999, which was already one and a half years after RCN initiated contact with the City.138
77. As it turns out, Comcast played a central role in pressuring City of Philadelphia
officials to delay or deny RCN’s entry into the Philadelphia market. Comcast’s efforts are
evident from statements made by city officials. For example, in May 1999 Mayor Rendell
132. RCN Begins Construction In Philadelphia, Company's Initial Progress Far Outpaces Early Development
Of Prior Markets, PR NEWSWIRE, Jul. 15, 1999. 133. Id. (“RCN is nearing completion of the high-tech operations center that will house the hardware
necessary to sustain its bundled product offering in Philadelphia. Located in Valley Forge area, the facility will include a state-of-the-art data center, digital cable television head-end with a capacity of several hundred channels, and a Lucent 5-ESS telephone switch.”).
134. See RCN Begins Construction in Philadelphia, Company’s Initial Progress Far Outpaces Early Development of Prior Markets, July 15, 1999 (“The company has secured local licenses from nine communities in the region and has been granted federal Open Video System (OVS) approval to serve customers in the Philadelphia area. RCN's successful application to the FCC for OVS certification covers the Philadelphia metropolitan area and surrounding counties of Delaware, Bucks, Chester and Montgomery.”).
135. RCN Encounters A Tough Path, supra (“RCN, founded by Bostonian David McCourt in 1997 and backed by billions from the likes of Microsoft Corp. cofounder Paul Allen, first approached Philadelphia in June 1998.”).
136. Id. (“It took RCN a year just to get the city's negotiating team—officials from the Public Property and Law Departments—to generate and supply the elaborate application, many months longer than the process had taken in other cities.”).
137. Id. 138. Id. (“At that May 1999 meeting, the mayor grew conciliatory after [RCN official John] Estey explained
that under the federal Telecommunications Act of 1996, which encouraged cable competition, the city could not legally keep RCN out. ‘If it is the federal law, then we have to follow it,’ [mayor] Rendell said, adding that he doubted RCN had time to get approval before he would leave office seven months later.”).
-52-
NAVIGANT ECONOMICS
5102853
reportedly told RCN: “Good God! ... We have to tear up our streets so you can come in here and
compete against one of our best corporate citizens?”139 The mayor was also quoted as telling
RCN that Philadelphia was “Comcast country.”140 Comcast’s influence similarly extended to the
Philadelphia City Council. City councilman James Kenney, the chairman of the City Council’s
RCN hearings, admitted that “there were substantial pressures from Comcast.”141 For example,
chairman Kenney met Comcast CEO Brian Roberts at the Ritz Carlton Hotel in October 2000,
just days before the first public hearing on RCN’s plan.142 Although Councilman Kenney had
been in office since 1992,143 the two had never met before.144
78. Comcast also privately collaborated with the city council regarding RCN’s
application. As part of this effort, Comcast produced an “internal document” entitled
“Competition or Chasm: RCN’s Effort to Digitally Divide Our City.”145 This document, which
was allegedly provided by Comcast to the Philadelphia city council, contained a list of suggested
questions that council members should ask RCN.146 The document also implored the city to “use
139. Id. (“‘Good God!’ [mayor Ed] Rendell recalls exclaiming. ‘We have to tear up the streets so you can
come in here and compete against one of our best corporate citizens?’”). 140. Princeton, N.J., Cable-TV Firm Withdraws Proposal for Philadelphia Network, KNIGHT RIDDER
TRIBUNE BUSINESS NEWS, Feb. 15, 2001 (“The political hurdles became clear, though, when in June 1998 the company approached Mayor Edward G. Rendell about offering service in the city. Rendell has acknowledged that he told Burnside that Philadelphia was Comcast country.”).
141. RCN Encounters A Tough Path, supra (“‘There were substantial pressures’ from Comcast and Time Warner Wade Cable, which also serves part of the city, Kenney said.”).
142. Id. (“On Oct. 24, several days before he was to chair the first public hearing on RCN’s plan, Kenney arranged to meet Comcast President Brian Roberts; the two had never met before. At the Ritz Carlton, they talked family—and RCN—over sodas.”).
143. See Jim Kenney, City Councilman-at-Large, James Kenney for Council Committee, available at www.jameskenney.com (“Since taking office in January 1992, Councilman Kenney has built a well-deserved reputation for offering a commonsense approach and a willingness to tackle the tough problems confronting the City of Philadelphia.”)
144. RCN Encounters A Tough Path, supra (“On Oct. 24, several days before he was to chair the first public hearing on RCN’s plan, Kenney arranged to meet Comcast President Brian Roberts; the two had never met before.”).
145. Id. (“RCN’s executives knew they were in for a grilling: They had already seen a Comcast document called ‘Competition or Chasm: RCN’s Effort to Digitally Divide Our City.’”).
146. Id. (“Stanley Wang, Comcast executive vice president, says it was an ‘internal document, not intended to be submitted to Council.’ Regardless, copies of the document were sitting on a table in the hearing room…It
-53-
NAVIGANT ECONOMICS
5102853
its substantial regulatory powers to monitor and mitigate conditions which might otherwise allow
RCN to cherry-pick its way to profitability.”147 Aided by Comcast’s clandestine lobbying efforts
and Comcast documents, the city council asked questions that closely resembled those provided
by Comcast at RCN’s initial public hearing.148
79. Comcast also engaged in a campaign of misinformation to block the RCN
proposal. Specifically, Comcast commissioned an analysis that concluded that RCN’s proposal
“appears to deny the benefits of this new competition to those who would, in theory, stand to
gain the most.”149 Comcast later admitted that the report was based on incorrect information that
it had provided to the author of the report.
80. As a result of Comcast’s intense lobbying of city officials, the city council refused
to act on RCN’s application. At the close of the year 2000, RCN’s application had been stalled
by the City of Philadelphia for two and one half years. Disenchanted with the delay, RCN
reportedly considered filing a federal lawsuit or a complaint with the FCC.150 Upon hearing that
RCN was considering litigation, chairman Kenney scheduled another hearing for February 1,
included a list of suggested questions for Council members to ask, such as: ‘Are any RCN cable systems unionized?’ ‘Is RCN for sale?’ and ‘Would RCN be willing to move its corporate office to Philadelphia?’”).
147. Id. (“In the document, Comcast questioned RCN’s ability to pay for the system it was promising, its willingness to serve low-income households, and its commitment to the city. ‘In order to keep RCN from becoming a high-tech, profiteering interloper, the city must use its substantial regulatory powers to monitor and mitigate conditions which might otherwise allow RCN to cherry-pick its way to profitability,’ the document said.”).
148. Id. (“Council members asked RCN about its relationships with unions, its record on minority hiring, and its construction plans. Many of the questions resembled those in the Comcast document.”).
149. Id. (“Comcast had hired a Wharton professor to analyze RCN’s construction plans, and his report said that RCN would serve ‘primarily mid- to upper-income and predominantly white households in the first several years’ and ‘appears to deny the benefits of this new competition to those who would, in theory, stand to gain the most.’”).
150. RCN Wants the FCC or the Federal Courts to Speed Up the Approval of Its Cable Franchise Bid in Parts of Philadelphia, CABLE WORLD, Nov. 20, 2000 (“RCN wants the FCC or the federal courts to speed up the approval of its cable franchise bid in parts of Philadelphia. The company began exploring other options last week when it learned that the city council may postpone ruling on the proposed contract until next year.”).
-54-
NAVIGANT ECONOMICS
5102853
2001.151 At this second meeting, however, the city council again refused to vote on RCN’s
application.152 Shortly after that meeting, RCN announced that it was withdrawing its application
to build a competitive cable system in the City of Philadelphia.153
81. The municipal approach to cable franchising shielded Comcast from competition
by delaying, and sometimes preventing, potential competitors’ entry, including Verizon, into the
City of Philadelphia.154 LFAs can impose extraneous requirements on potential entrants,
prolonging negotiations and delaying the issuance of a cable franchise to the detriment of its
citizens.155 Even when a franchise is approved quickly and with little resistance, negotiating
contracts with so many regulatory bodies takes time and delays competition. Prolonged
negotiations allow incumbents, such as Comcast, to influence the franchise process.156
82. On June 6, 2006, Senators Dominic Pileggi and Anthony Williams introduced the
Cable Choice & Competition Act in Pennsylvania.157 The legislation would have created a
Pennsylvania franchise authority, allowing wire-based cable service providers to apply for a
151. RCN Encounters A Tough Path, supra (“Fed up with the continuing delays, RCN threatened a federal
lawsuit. The City Solicitor's Office quickly scheduled a meeting with Kenney in early December. That same day, Kenney scheduled a second hearing on the RCN matter for Feb. 1.”).
152. RCN Withdraws, supra (“The committee did not vote that day, nor did it vote on RCN's proposal after a second public hearing Feb. 1.”).
153. Id. (“RCN Corp. of Princeton withdrew its proposal yesterday to build a $250 million cable-TV and telephone network in the northern half of Philadelphia, blaming the move on City Council delays.”).
154. See John Thorne, Symposium Comments, 2007 Telecommunications Symposium: Voice, Video and Broadband: The Changing Competitive Landscape and Its Impact on Consumers (Nov. 29, 2007), available at http://www.usdoj.gov/atr/public/workshops/telecom2007/submissions/228010.htm (For a discussion of how LFAs can unnecessarily delay competition). See also Letter from Thomas O. Barnett, Antitrust Division, U.S. Department of Justice, to Senator Michael W. Morrissey (Apr. 30, 2007) (To view his concerns about the franchise process).
155. See Thorne, supra. See also Ex Parte Submission of the Department of Justice in MB Docket No. 05-311, at 7 (FCC filed May 10, 2006), available at http://www.usdoj.gov/atr/public/comments/216098.htm. See also Barnett, supra, at 3 (A franchise delay can hurt consumers in many ways—by delaying competition for video services, but also by delaying important system upgrades: “Some new video providers, such as the telephone companies, are providing video services over upgraded networks that support voice, video, and higher-speed broadband services. Because the revenues from offering video factor into the profitability of these upgrades, a delay in receiving a cable television franchise can cause new entrants to postpone modernizing their networks.”).
156. Id. 157. See State Senator Dominic Pileggi press release June 6, 2006, available at
state-wide license instead of negotiating individual agreements with every municipality.158 At the
time, more than a dozen states had passed or were considering such legislation.159
83. Comcast began combating support for state-wide franchise licensing in
Pennsylvania before the legislation was introduced. Although the act may have limited local
authorities’ abilities to gain concessions from cable companies over and above franchising fees,
it maintained the fee requirements of the individual municipality agreements.160 The bill
continued the standing maximum five percent franchise fee permitted by federal law, ensuring
that municipalities would not lose out on a valuable revenue stream.161 Comcast nonetheless was
able to successfully oppose Pennsylvania’s credible reform effort and defeat the Cable Choice &
Competition Act, forcing Verizon to gain approval to bring its wire-based cable service to
market one municipality at a time.
84. Verizon initiated contact with the City of Philadelphia regarding a franchise
agreement in April 2008 and negotiations began quietly in June.162 Comcast officials voiced their
concerns over Verizon’s “rush” to push its agreement through the city council, and reviewed the
Verizon proposal “line for line to make certain Verizon’s deal [was] no more favorable than
Comcast’s own pay-TV franchise with Philadelphia.”163 Comcast prompted the council to
request detailed deployment plans from Verizon, but Verizon resisted, asserting that Comcast
would use the information to lock its customers in special deals, making it harder for Verizon to
158. Summary: Cable Choice & Competition Act, available at
http://www.pasenategop.com/news/archived/2006/0606/cablechoiceact/SummaryCableChoiceAct.pdf. 159. See Pileggi, supra. 160. See Titch, supra, at 3. See also Summary, supra, at 1. 161. See Summary, supra, at 1. 162. See Bob Fernandez & Jeff Shields, Verizon pitches FIOS (Philadelphia Inquirer, Nov. 14, 2008),
available at http://www.philly.com/philly/business/34448269.html. 163. See Bob Fernandez, Politics, Comcast slow Verizon’s pay-TV bid (Philadelphia Inquirer, Dec. 14, 2008),
available at http://www.philly.com/philly/business/36120899.html. This was a common tool used by incumbents to threaten LFAs during negotiations with potential entrants. See Thorne, supra, at 24-27.
-56-
NAVIGANT ECONOMICS
5102853
win subscribers when FiOS did become available.164 The City of Philadelphia finally granted a
franchise agreement to Verizon FiOS on February 5, 2009.165
B. New and Old Delivery Platforms Threaten Comcast’s Dominance
85. There is ample evidence that competition from cable overbuilders, DBS providers
and the recent entry of telco competitors in certain local markets exerts downward pressure on
the prices charged by cable companies for video service. In this section, I review the evidence,
by competing technology, of how MVPD competition manifests itself when not impaired by
exclusionary conduct by incumbent cable operators. The purpose of the section is to demonstrate
that Comcast has a strong incentive to impair competition through exclusionary conduct made
possible or facilitated by the proposed transaction; in the absence of such impairment, these
MVPD rivals impose significant constraint on the price of Comcast’s cable television service.
1. Fiber Networks Deployed by Telephone Companies
86. Entry by the telephone companies (“telco entry”) has exerted significant pressure
on the margins of cable operators. In particular, cable operators have responded to telco entry by
competing on price, aggressively moving customers to bundled services, upgrading the video
products they offer, and adding other services. The Boston Globe described the fierce
competition for video customers between Comcast and Verizon as follows: “Customers welcome
the competition, with many switching from Verizon to Comcast and back again in search of
164. See Bob Fernandez, Verizon presses Philadelphia for TV franchise (Philadelphia Inquirer, Dec. 4, 2008),
available at http://www.philly.com/philly/news/homepage/35525069.html (“A Verizon official said it would not disclose zip-code-level build-out of the new high-speed network because Comcast could then target those neighborhoods with special promotions and multiyear contracts.”).
165. See Dave Davies, Council OKs Verizon for cable-TV (Philadelphia Daily News, Feb. 6, 2009), available at http://www.philly.com/philly/hp/news_update/39189912.html.
-57-
NAVIGANT ECONOMICS
5102853
better service and lower prices. Customers who switch can often win discounted rates from their
new provider. What they probably won’t see is a big difference in the two systems.”166
87. Incumbent cable operators were quick to reduce prices upon telco entry.
According to a 2006 Bank of America survey, in areas where Verizon was rolling out FiOS,
cable operators responded with “not actively advertised” price cuts of 29 (by Charter) to 43
percent (by Cox).167 In the spring of 2007, AT&T’s double-play package in Dallas was priced 21
percent below a comparable package offered by Time Warner;168 Verizon’s triple-play package
in Fairfax was priced 23 percent below a comparable package offered by Cox.169 Press accounts
documented non-price responses to telco entry as well. In October 2007, Comcast was reportedly
“rushing to deliver the new features [video-on-demand, more channels] in Santa Rosa because
rival AT&T has started offering its own digital TV service in select neighborhoods.”170 In April
2007, Cable World noted that Comcast was set to “offer more linear and high-definition
channels, video-on-demand titles and digital phone features to its 700,000 basic customers in and
around Houston before the end of June” in “response to these two IPTV deployments” by
AT&T.171 In August 2007, a Time Warner cable product director said that he was launching new
166. Hiawatha Bray, Comcast, Verizon Battle it out for Market Share, BOSTON GLOBE, February 7, 2010,
available at http://www.boston.com/business/technology/articles/2010/02/07/comcast_verizon_battle_for_market_share/?page=2.
167. Bank of America Equity Research, Battle for the Bundle: Consumer Wireline Services Pricing, Jan. 23, 2006, at 10 (surveying prices in Herndon, VA, Keller, TX, and Temple Terrace, FL).
168. AT&T’s launch in bits Cable, broadband deal to come out gradually, may push prices down, Dallas Morning News, Mar. 6, 2007, at 1D. AT&T charged $59 per month for 100 cable channels and Internet speeds of 1.5/1 mbps. Time Warner charged $75.60 per month for 84 cable channels and Internet speeds of 1.5/0.384 mbps.
169. Cox Bundle Fairfax, available at http://www.cox.com/fairfax/bundle.asp (accessed Apr. 1, 2007); https://www22.verizon.com/ForYourHome/NationalBundles/NatBundlesHome.aspx# (accessed Apr. 1, 2007). Verizon charged $120 per month for a digital, sports, and news tier, unlimited calling, and Internet speeds of 15/2 mbps. Cox charged $156 per month for digital, sports, and information tier, unlimited calling, and Internet speeds of 15/2 mbps.
170. Comcast Raising Prices Again, Except In Most Of Santa Rosa: Average Cable Bill To Jump Almost 5%; Santa Rosa Rates Unchanged As System Gets Upgraded, THE PRESS DEMOCRAT, Oct. 17, 2007, at A1.
171. Comcast Promises To Fight For Every Customer As It Faces DBS And IPTV Competition In Houston, CABLE WORLD, Apr. 16, 2007.
-58-
NAVIGANT ECONOMICS
5102853
products, including a triple-play service, as a direct response to AT&T entering the market in
San Antonio.172 In May 2007, in response to AT&T’s deployment of U-Verse, the regional
manager of Cox San Diego explained how “the expansion of video-on-demand, high-definition
and local programming are … key weapons to be used against Cox’s competitors.”173
88. As the Bank of America survey made clear, the price and non-price effects
associated with telco entry are often not reflected in the advertised cable price, as discounts are
generally offered to existing customers to avoid having them defect. Thus, looking at published
cable prices for evidence of the price effects of telco entry can muddle the real story. To isolate
these price effects, it is more informative to look at the change in a cable operator’s average
revenue per unit (ARPU) upon telco entry. For Comcast, video ARPU growth as estimated by JP
Morgan has slowed considerably, and it even declined in the second and third quarters of
2009.174 According to JP Morgan, “The discounted offers [for video service] offset the benefits
of strong demand for advanced services (HD and/or DVR) and implemented price increases.”175
The authors of the JP Morgan report explain that they “expect AT&T and Verizon to continue
building out fiber facilities at a measured pace, introducing additional video competition in many
of Comcast’s markets.” As a result, the analysts “expect heavy competition for subscribers, with
a potential impact to customer adds, profitability, or both.”176 According to Credit Suisse,
Comcast’s Q4 2009 revenues were lower than expected due to “lower than expected ARPU
growth… primarily driven by softness in video ARPU.”177
172. Meet the System San Antonio: Time Warner Cable Fights AT&T in its Hometown, CABLE WORLD, Aug.
13, 2007. 173. San Diego: Meet the System, CABLE WORLD, May 7, 2007. 174. Mike McCormack, Scott Goldman, and Manish Jain, Comcast Corp: Favoring Customer Growth over
Profitability…for Now, JP Morgan Equity Research, Nov. 5, 2009, at 3. 175. Id. 176. Id. at 6. 177. Spencer Wang & Stefan Anninger, Comcast, Credit Suisse Equity Research, Feb. 3, 2010, at 2.
-59-
NAVIGANT ECONOMICS
5102853
89. Even video ARPU trends can mask the effects of price competition for a number
of different reasons. First, Comcast and the other cable companies are aggressively pushing their
bundled services, and Comcast’s attribution of revenue to the different services in the bundle is
somewhat arbitrary. According to Morgan Stanley analysts, Comcast “increas[ed] its
promotional intensity in the second half [of 2009]—by matching competitor offers already in the
marketplace and specifically by focusing on triple play bundles, which typically entail lower data
ARPUs.”178 Comcast’s CFO Michael Angelakis, in discussing Q4 2009 results, said “Total video
revenue declined slightly by 0.6% reflecting a 199,000 decline in basic video customers and
video ARPU growth was 2%. This slowdown in video ARPU growth is a result of more
moderate rate increase in the fourth quarter of 2009 compared to the increases in the fourth
quarter of 2008 and additional bundling and promotion.”179
90. Second, cable companies were aggressively transitioning their customer base from
analog to digital service over the past couple of years, with digital plans offering significantly
more channels and often improved picture quality. This shift in the product mix to higher priced
digital plans provided a temporary boost to ARPU that obscured the overall price pressure on
cable’s video service.180
91. Third, primarily in response to competition from the telco providers, Comcast and
the other cable companies have over the last few years significantly upgraded their service.
Examples include significantly more HD channels, growing video-on-demand libraries, faster
178. Benjamin Swinbourne, David Gober, Ryan Fiftal, Comcast Corporation: Raising FCF Estimates,
Reiterate Equal-weight, Morgan Stanley Research, Feb. 4, 2010, at 3. 179. Transcript of Comcast Q4 2009 Earnings Call, Feb. 3, 2010, available at
http://seekingalpha.com/article/186373-comcast-q4-2009-earnings-call-transcript?page=-1. 180 Stiff Competition and Economic Challenges Will Pressure Telecom and Cable TV Sector in 2010, Fitch,
December 3, 2009, available at http://www.businesswire.com/portal/site/home/permalink/?ndmViewId=news_view&newsId=20091203005599&newsLang=en.
-60-
NAVIGANT ECONOMICS
5102853
Internet speeds, and free Wi-Fi access for subscribers.181 These service upgrades make
comparisons of pricing and ARPU for cable video service and/or bundled services over time
somewhat misleading, as the quality of the product has improved significantly.
92. Finally, Comcast’s own executives have noted the effect of telco entry on its
business. As Steve Burke, Comcast’s COO, explained on Comcast’s Q2 2009 earnings
conference call, “The RBOCs now have over built 28% of our footprint and they and satellite
were very aggressive during the quarter. We have recently launched a number of high impact
marketing campaigns and are seeing better trends in July than we did in the second quarter.”182
2. Non-Telco Cable Overbuilders
93. Many studies have found that cable prices are lower when cable overbuilder
competition is present. Most, but not all, of these studies have been conducted by agencies of the
U.S. government. In January 2009, the FCC released its most recent version of its Report on
Cable Industry Prices (“Cable Price Report”).183 This report is the latest of several installments
in the FCC’s analysis of the effect of competition on cable prices. As part of this inquiry, the
FCC conducted a statistical analysis of how incumbent cable prices are affected by the extent of
181. Comcast CEO Brian L. Roberts Announces Project Infinity: Strategy to Deliver Exponentially More
Content Choice on TV, Comcast Press Release, January 8, 2008, available at http://www.comcast.com/About/PressRelease/PressReleaseDetail.ashx?PRID=724; Comcast Begins Rollout of Extreme 50 Mbps High-Speed Internet Service, Comcast Press Release, October 22, 2008, available at http://www.comcast.com/About/PressRelease/PressReleaseDetail.ashx?PRID=814; Dalila J. Paul, Comcast, Cablevision Turn to WiFi to Retain Customers, New Jersey Business News, February 28, 2009, available at http://www.nj.com/business/index.ssf/2009/02/comcast_turns_to_wifi_to_retai_1.html.
182. Transcript of Comcast Q2 2009 Earnings Call, Aug. 6, 2009, available at http://seekingalpha.com/article/154406-comcast-corporation-q2-2009-earnings-call-transcript?page=-1.
183. In the Matter of Implementation of Section 3 of the Cable Television Consumer Protection and Competition Act of 1992, Statistical Report on Average Rates for Basic Service, Cable Programming Service, and Equipment, MM Dkt. No. 92-266 (rel. Jan. 16, 2009)
-61-
NAVIGANT ECONOMICS
5102853
competition from overbuilders. The FCC bases its analysis on a survey of cable price and
competition data it collects annually.184
94. In every iteration of the FCC Cable Price Reports, the FCC has found that the
effect of overbuilder entry on cable prices is significant. In 2005, the Commission wrote that
incumbent cable prices “are 17 percent lower where wireline cable competition is present.”185 If
the appropriate “base” rate is the rate charged in competitive areas, then the same FCC data
indicate that prices are 20.6 percent higher in non-competitive areas than they are in competitive
areas.186 This finding is consistent with FCC cable price survey findings over the past few years,
which are presented below in Table 3.
TABLE 3: FCC-ESTIMATED EFFECT OF OVERBUILD COMPETITION ON INCUMBENT CABLE EXPANDED BASIC PRICES, 2002-2008
Difference Date Average Overbuilt Area Cable Price
Average Non-Competitive Area Cable Price Nominal Percentage
January 2002 $31.01 $36.21 $5.20 16.8% January 2003 $32.83 $39.11 $6.28 19.1% January 2004 $34.00 $41.18 $7.18 21.1% January 2005 $36.79 $43.77 $6.98 19.0% January 2006 $40.24 $45.53 $5.29 13.1% January 2007 $42.77 $47.49 $4.72 11.0% January 2008 $45.04 $49.97 $4.93 10.9% Average $37.53 $43.32 $5.80 15.4%
184. Id. ¶ 15 (“The information and analysis provided in this Report are based on the Commission’s survey of
cable industry prices (‘surveys’) that collected data as of January 1, 2005; January 1, 2006; January 1, 2007, and January 1, 2008, and also based on a supplemental survey that collected data as of July 1, 2006.”).
185. Statistical Report on Average Rates for Basic Service, Cable Programming Services, and Equipment, MM Dkt. No. 92-266, released Dec. 27, 2006, ¶ 2.
186. As presented in Table 4, the FCC reports that incumbents charge an average price for basic and expanded basic programming of $35.94 in competitive areas and $43.33 in non-competitive areas. The difference is thus $7.39. The difference in the percentage is related to the base number used. If the appropriate base is the non-competitive price ($43.33), then the calculation is: $7.39 / $43.33 = 17.1%. That is, cable prices are 17.1% lower in competitive areas than they are in non-competitive areas. If the appropriate base is the competitive price ($35.94), then the calculation is: $7.39 / $35.94 = 0.206. That is, cable prices are 20.6% higher in non-competitive areas than they are in competitive areas.
-62-
NAVIGANT ECONOMICS
5102853
The percentages in Table 3 represent the extent to which incumbents demand higher average
cable prices in non-competitive areas than they do in competitive areas. In January 2009, the
FCC noted that “cable prices decrease substantially when a second wireline cable operator enters
the market.”187 Importantly, the FCC’s analysis assigns causation as well, arguing that the mere
entry of an overbuilder reduces cable prices.
95. A regression analysis conducted by the FCC also confirms that overbuilders
significantly constrain the price an incumbent can charge for expanded basic MVPD service. In
an analysis released in January 2009 in its 2008 Cable Price Report, the FCC found that:
In markets with two competing cable operators, the results show that the incumbent operator charges 14.1 percent less, on average, all other things held constant, than operators charge in markets where a second cable operator is not present. The results also show a tendency for the incumbent operator to undercut the overbuild rival’s price rather than simply matching that price.188
The FCC regression analysis also controlled for market demographics (income), market structure
(HHI, vertical affiliation, and number of national subscribers), and DBS competition (whether
local-into-local programming was available from DBS).
96. A 2004 GAO study has also found that the presence of cable overbuilders
produced significant benefits to consumers. Specifically, the GAO published a report based on a
series of interviews with cable overbuilders, incumbent cable operators, and other parties such as
cable franchising boards.189 The GAO interviews provide a wealth of information regarding the
competitive interactions between overbuilders, incumbent cable operators, and local regulatory
authorities. The GAO documented that extended basic cable prices were as much as 41 percent
187. Report on Cable Industry Prices, MM Dkt. No. 92-266, released Jan. 16, 2009, ¶ 3 (“Cable prices
decrease substantially when a second wireline cable operator enters the market.”). 188. Id. ¶ 14. 189. Government Accountability Office, Wire-Based Competition Benefited Consumers in Selected Markets,
GAO-04-241, Feb. 2004.
-63-
NAVIGANT ECONOMICS
5102853
lower when overbuilders were present relative to markets without overbuilding.190 Based on this
analysis, the GAO found that overbuilding induced incumbent cable operators to respond to
competition “by providing more and better services and by reducing rates and offering special
deals.” 191
97. The FCC’s and GAO’s findings are corroborated by academic studies. Thomas
Hazlett analyzed the reduction in consumer surplus that results from local franchising restrictions
aimed at impeding overbuilding competition.192 Using GAO data, Hazlett estimates that entry by
an overbuilder would reduce cable prices by 15.6 percent.193 This price reduction amounts to
$7.32 per subscriber per month, and induces total cable subscribership to grow by 19 percent.194
With these numbers, he estimates that consumers would realize $8.9 billion per year in
incremental savings.195 Consumers gain, at least in part, at the expense of their incumbent cable
providers, who would lose approximately $6.0 billion per year as a result of greater cable
competition.196 Overall, Hazlett finds that greater cable competition would yield efficiency gains
of $2.9 billion per year (equal to the consumer gain of $8.9 billion less the incumbent cable loss
190. Id. at 15 (“For example, in 1 BSP market, the monthly rate for cable television service was 41 percent
lower compared with the matched market, and in 2 other BSP locations, cable rates were more than 30 percent lower when compared with their matched markets.”).
191. Id. at 12 (“In the 12 markets we reviewed, the entry of a BSP appears to induce incumbent cable operators to respond by providing more and better services and by reducing rates and offering special deals.”).
192. Thomas W. Hazlett, Cable TV Franchises as Barriers to Video Competition, 12 VIRG. J. L. & TECH.2
(2007) [hereinafter Franchises as Barriers to Competition]. 193. Id. at 66 (“The GAO model also reported a partial price elasticity of demand for cable subscribers of
about 1.5. Consequently, the results reported in Table 6 replicate the GAO analysis assuming an elasticity of demand of 1.5 and an entrant market share of between twenty and twenty-five percent. In the following benchmark analysis, twenty-five percent market share is used with an elasticity of 1.5. This implies a cable price reduction equal to 15.6%.”).
194. Id. at 67 fn. 233 (explaining that the consumer gains are based on a price reduction of $7.32 per subscriber per month and an increase in total cable subscribership of 17,904,234 (from 92,295,766 subscibers)).
195. Id. at 66-7 (“The change in consumer surplus associated with a 15.6% cable price reduction is approximately $741 million per month, or $8.9 billion per year.”).
196. Id. at 67 (“This change is negative; owing to the greater competitiveness of the market, suppliers as a whole see profits decline by about $501 million per month, or $6.0 billion per year.”).
-64-
NAVIGANT ECONOMICS
5102853
of $6.0 billion).197 Hazlett calculates that this efficiency gain represents six percent of cable
industry revenues.198
98. Karikari, Brown, and Abramowitz (2003) examine the competitive inter-
relationships between DBS providers, local exchange telephone cable entrants, cable
incumbents, and overbuilders.199 The authors report that their study is novel because it explicitly
accounts for competition from non-cable MVPDs like DBS providers, includes data from a
variety of cable systems operating under different market structures, and factors in technological
improvements and upgrades in the provision of cable services.200 To model the economic
variables that characterize the MVPD market, the authors estimate a system of equations that
predict DBS and cable penetration, cable rates, and the number of channels available to
subscribers.201 The results of the study indicate that DBS penetration is lower where overbuilders
and local exchange telephone companies have entered the market as these entrants lower cable
rates, inducing subscribers to choose cable over DBS.202
99. Savage and Wirth (2005) also examine the effects of competition on cable quality
and cable prices. Specifically, the authors quantify how incumbent cable companies respond to
197. Id. (“An estimate of the net benefits to society accruing from competitive, nationwide entry combines the
respective estimates of consumer gain and producer loss. Efficiency gains are estimated to be $241 million per month or about $2.9 billion per year. These potential gains reveal the value of the opportunity society loses for each month (or year) of delay in which nationwide cable market entry is deterred.”).
198. Id. (“With pre-entry monthly revenues in the model used for this analysis of approximately $4.3 billion ($51.6 billion annually), the efficiency gains are approximately six percent of industry revenues.”).
199. John Agyei Karikari, Stephen M. Brown & Amy D. Abramowitz, Subscriptions for Direct Broadcast Satellite and Cable Television in the US: an Empirical Analysis, 15 INFORMATION ECONOMICS AND
POLICY 1 (2003). 200. See id. at 4 (“The Study is based on a reasonable representation of developments in the subscription
television services market. First, unlike most previous studies, we consider explicitly competition from non-cable providers in the market, DBS in particular. Second, the data used include different segments of the market—competitive and non-competitive cable providers, regulated and unregulated cable providers, and cable franchises of different sizes. Third, we incorporate technological changes in the market—systems upgrades that enable cable operators to offer high quality services (digital and more channels) and integrated services (cable, telephony, and Internet).”).
201. Id. at 5. 202. Id. at 13.
-65-
NAVIGANT ECONOMICS
5102853
the threat of potential competition.203 To quantify potential competition the authors estimate the
probability that an overbuilder or local exchange telephone carrier will enter the market based on
factors that affect cost, demographic variables, and market structure indicators for a given
geographic observation.204 These probabilities are then included in a full supply-demand model
so that equilibrium price and quality effects can be determined.205 The authors find that:
Second-step estimation of the supply-demand system shows incumbent cable operators offer more channels to consumers in markets facing greater potential competition from BSP wireline overbuilders and/or ILECs. In particular, when the probability of entry rises to about 42%, the average cable system provides six more channels, and price per channel declines from US $0.77 to US $0.66.206
This conclusion suggests that consumers gain from the mere possibility that an overbuilder or
local exchange carrier will enter the MVPD market.
100. Kelly and Ying (2007) use the changing regulatory and competitive environment
in the MVPD market from 1993 to 2001 to study the effects of market structure on cable
prices.207 The authors estimate the cable rate in a given area as a function of cost, factors
affecting the elasticity of demand, regulation, and competition.208 They conclude that “Overall,
competition is effective in lowering average cable rates, from 5.6 to 8.8 percent. Moreover,
excluding low penetration systems from the competitive group generally results in larger rate
differentials than what were found by the FCC it its price surveys.”209 In addition, they find that
203. Scott J. Savage & Michael Wirth, Price, Programming, and Potential Competition in U.S. Cable
Television Markets, 27 JOURNAL OF REGULATORY ECONOMICS 27 (2005). 204. Id. 205. Id. 206. See id. at 40. 207. See Mary T. Kelly & John S. Ying, Testing the Effectiveness of Regulation and Competition on Cable
Television Rates, Department of Economic Alfred Lerner College of Business & Economics University of Delaware Working Paper No. 2007-07 (2007) at 5 (“The remarkable period of deregulation, re-regulation, and re-deregulation offers a natural experiment to test the effect of regulation and competition. Using a unique data set on cable rates from 1993 to 2001, we model and econometrically estimate the effect of regulation and competition on cable rates.”).
208. Id. at 9. 209. Id. at 19.
-66-
NAVIGANT ECONOMICS
5102853
the price-disciplining effects of competition are strongest in large markets.210 The study also
confirms the findings in previous studies that competition increases the quality of services
offered by incumbent cable companies.211
3. Direct Broadcast Satellite
101. At least four studies have concluded that effective competition from DBS
providers constrains an incumbent’s price for cable service. The studies uniformly indicate that
effective DBS competition can significantly constrain incumbent cable providers’ prices.
102. In its 2009 Report on Cable Industry Prices,212 the FCC compiled survey data of
cable and DBS prices that it used to analyze trends in the those prices. Although the FCC’s
written summary of its survey data finds that DBS does not significantly constrain the price of
basic service offered by the incumbent cable operator,213 a closer look at the FCC’s survey data
and its econometric analysis tells a different story.
103. To analyze the effect of DBS and cable overbuilders on the incumbent cable
provider’s price, the FCC disaggregated its survey data into different cuts. The first cut of data
was taken between non-competitive DMAs and DMAs that passed the FCC’s competition test
and were therefore relieved of rate regulation. Additional data segments are created by
calculating the average prices for different tiers of service, such as expanded basic and digital
service. Given the FCC’s discussion of the effects of cable overbuilding and DBS entry on
incumbent price, it would appear that the FCC based its opinion that DBS does not constrain
incumbent price solely on a comparison of average price of expanded basic service.214
210. Id. 211. Id. at 20. 212. Implementation of Section 3 of the Cable Television Consumer Protection and Competition Act of 1992,
Statistical Report on Average Rates for Basic Service, Cable Programming Service, and Equipment, MM Dkt. No. 92-266, Report on Cable Industry Prices (released Jan. 16, 2009).
213. Id at ¶4. 214. Id.
-67-
NAVIGANT ECONOMICS
5102853
104. DBS, however, is a digital product, which competes with cable on the digital tier.
And the effect of DBS on incumbent cable operator prices is evident from a review of digital tier
prices. For example, in 2008, the FCC found that incumbent cable prices for digital service in
areas that were effectively competitive supplied by virtue of high DBS penetration were
approximately seven percent below the prices for digital service offered by incumbent cable
operators in non-competitive areas.215
105. More compelling evidence of the effect of DBS on MVPD prices is present in the
econometric analysis provided in Appendix B of the FCC’s report. The FCC estimated two
separate regressions that explain MVPD prices, and both of these regressions included the HHI
(Herfindahl-Hirschmann Index) of the most popular zip code in the market area as a right-hand-
side variable. In both regressions, the estimate parameter on the HHI variable is positive and
statistically significant.216 The positive sign on the regression parameter indicates that a decrease
in concentration, indicated by a decrease in HHI, results in a decrease in the price of MVPD
service, all else the same. Therefore, an increase in penetration for either DBS or a cable
overbuilder would have the effect of suppressing the price charged by the incumbent cable
operator. Because overbuilding activity in 2007 and 2008 was limited to a few areas in the
country—telco entry was just underway and the number of U.S. homes overbuilt prior to telco
entry was small217—the majority of the variation in the FCC’s HHI variable was likely due to
changes in DBS penetration. Accordingly, the estimated coefficient on the HHI variable in the
215. Id at 13 (tbl 1-b). 216. Id at 85. 217. For example, the FCC reported that in 2008, there were a total of 3,205 communities relieved from rate
regulation (only a fraction of which is due to overbuilding), compared to 30,352 communities in the noncompetitive group. Id at 85. In 2006, the FCC reported that BSPs and OVS providers, which typically operate overbuild systems, reported no appreciable change in subscribership, maintaining total subscribership of approximately 1.4 million. Thirteenth Annual Report, supra, at ¶ 9.
-68-
NAVIGANT ECONOMICS
5102853
FCC’s regression analysis serves as a reasonable proxy for the change in cable price owing to a
change to DBS penetration.
106. Again, the findings of DBS price effects in government studies are corroborated
by academic economists. National Bureau of Economic Research (NBER) members Goolsbee
and Petrin similarly find that effective DBS competition can constrain incumbent cable prices.
The authors find that DBS is a significant demand substitute for expanded basic and premium
cable. Goolsbee and Petrin arrive at this conclusion by examining what other form of video
service DBS customers would choose if DBS were unavailable. Goolsbee and Petrin estimate
that fully 92.3 percent of DBS customers would, if DBS was unavailable, subscribe to an
incumbent cable providers’ service (either expanded basic or premium cable service).
107. Armed with evidence of substitution between DBS and incumbent cable services,
Goolsbee and Petrin then compute the cross-price elasticity of demand for DBS relative to cable.
That is, they estimate how DBS subscribership changes given a change in the price of expanded
basic or premium cable. They find cross-price elasticity of demand for DBS given a change in
cable price of approximately one.218 This finding implies that an increase in cable prices in a
given community is associated with a proportional increase in satellite subscribership in that
community.
108. The authors conclude that their results “suggest that more competition from DBS
is correlated with lower cable prices and somewhat higher quality cable.”219 Specifically,
Goolsbee and Petrin estimate that, in the absence of DBS entry, cable prices would be $4 per
218. See, e.g. Goolsbee & Petrin (showing that, for an increase in the price of expanded basic and premium
cable, satellite market share increases by 0.951 and 1.187, respectively). See also id. at 369-70 (“In terms of substitutability, the demand elasticities reflect the correlation of observed and unobserved tastes…Similarly, for premium price increases, DBS has the largest cross-price [elasticity], followed again by expanded basic.”).
219. Id. at 377(“The supply side results exploit the estimated controls from the structural demand side model and suggest that more competition from DBS is correlated with lower cable prices and somewhat higher quality cable.”).
-69-
NAVIGANT ECONOMICS
5102853
month higher.220 They then apply this estimate to the 70 million cable users, finding that the
aggregate gain to cable consumers from lower cable prices amounts to between $3 and 4
billion.221 They also estimate that DBS users—who absent DBS would use cable—enjoy an
additional $2 to 3 billion in surplus.222 When considered together, Goolsbee and Petrin estimate
that effective DBS competition has generated as much as $7 billion in consumer surplus by
limiting cable rate increases, inducing cable quality improvements, and providing a valuable
alternative to incumbent cable systems.223
109. Wise and Duwadi (2005) found that DBS can act as a constraint on cable prices
for even basic services.224 Specifically, Wise and Duwadi found evidence of significant
consumer substitution between cable and DBS. Substitution was particularly strong in areas
where quality-adjusted relative cable prices increased by ten percent or more.225 The authors
conclude that their “findings are consistent with the hypothesis that DBS providers are a
constraining factor on quality-adjusted price increases for basic cable services by cable firms.”226
110. The article also introduces the concept of “switching costs,” which limit consumer
substitution and have important competitive effects. The authors find that the cable market is
220. Id. (“For cable subscribers, our results suggest that cable prices are at least $4 per month lower than they
would have been.”). 221. Id. (“The aggregate gains to the 70 million cable users amount to between roughly $3-4 billion.”). 222. Id. (“Overall there is a significant welfare gain to the 16 million satellite buyers between $2-3 billion,
depending upon whether changes in cable prices and characteristics are added back into the calculation.”). 223. Id. at 377-78 (“In the end, our results suggest large gains from DBS entry, some of which are not
captured if the price and characteristics’ response is ignored. The overall gains from this production may be as large as $7 billion, illustrating once again the importance of understanding the impact of new goods on consumer welfare.”).
224. Wise & Duwadi, supra, 679-705. 225. Id. at 701 (“These results indicate that, as previously shown by Goolsbee and Petrin, consumers view
DBS as a substitute for cable in terms of higher-quality services offered, such as premium movie and high-definition channels. Additionally, even for basic cable services, consumers appear to turn to DBS as a substitute for cable when facing large quality-adjusted cable price increases, but may turn to cable as a substitute for DBS to a lesser extent when presented with large quality-adjusted cable price decreases.”). See also id. at 698 (“The coefficient of PLUS10RISE, representing communities with large quality-adjusted cable price increases, is positive and has a large magnitude, showing higher DBS penetration in these areas, perhaps through substitution from cable to DBS.”).
226. Id. at 698
-70-
NAVIGANT ECONOMICS
5102853
characterized by substantial “switching costs,” or, in other words, costs that consumers incur
when they seek to switch service from one provider to another.227 Wise and Duwadi find that the
presence of these switching costs may allow an incumbent cable firm to charge supra-
competitive cable prices. The authors come to this conclusion by examining the strength of
MVPD demand substitution for different levels of cable price increases. They find that the
presence of switching costs “limits substitution between cable and DBS services when quality-
adjusted price changes are small.”228 For their analysis, “small” is defined as price increases of
10 percent or less. Wise and Duwadi indicate that under certain circumstances when switching
costs are large, the incumbent cable firm “would charge supra-competitive rates to existing
subscribers.”229
111. Clements and Brown (2006) show that an increase in the quality of DBS service
offerings forces cable companies to improve the quality of their services. The authors’
conclusions are based on a study of the effects of the Satellite Home Viewer Improvement Act of
1999, which allowed DBS providers to deliver local broadcast stations to their subscribers.230
Before 1999, the Act’s effective date,231 the laws governing the provision of DBS services
prevented satellite-based carriers from providing these channels to their subscribers in most
227. See, e.g., Paul Klemperer, Entry Deterrence in Markets with Consumer Switching Costs, 97 ECON. J. 99
(1987). See also Paul Klemperer, Markets with Consumer Switching Costs, 18 RAND J. ECON. 138 (1987). 228. Wise & Duwadi at 701. 229. Id. at 702 (“In a situation where price discrimination between new customers and repeat customers is not
possible, and where the consumer switching cost is high, the incumbent would charge supra-competitive rates to existing subscribers and not compete for new subscribers….Our results point to this possibility, since it appears that consumers switch multichannel video providers only in response to relatively large price changes, not small ones.”).
230. Michael E. Clements & Stephen M. Brown, The Satellite Home Viewer Improvement Act: Price and Quality Impact of Direct Broadcast Satellite Companies’ Provision of Local Broadcast Stations, 30
TELECOMMUNICATIONS POLICY 125 (2006) [hereinafter Clements & Brown]. 231. The FCC’s Satellite Homer Viewer Improvement Act Page, available at http://www.fcc.gov/mb/shva/
(visited Apr. 8, 2009).
-71-
NAVIGANT ECONOMICS
5102853
situations.232 Meanwhile, cable companies, regulated under a different set of laws, were able to
deliver local broadcast services to their subscribers placing DBS providers at a competitive
disadvantage.233
112. The authors find that when both major DBS companies carry local broadcast
stations, cable companies are forced to offer more channels to their subscribers.234 On average,
such an increase in channels of DBS offerings induced cable companies to provide an additional
three channels to their customers.235 On the other hand, the authors find that the increase in the
number of channels available to DBS subscribers did not have a statistically significant effect on
cable prices.236
4. Internet-Only Video Distributors
113. Given its nascent state, it is difficult to measure the price-disciplining effect of
Internet video on cable television service. Accordingly, this section is more qualitative than the
prior sections. Based on the evidence presented here, it is reasonable to infer that Internet video
either mildly constrains the price of cable television services or threatens to do so in the near
future.
114. As noted above, Comcast and other cable operators perceive Internet video to be a
substitute (and not a complement) to its cable television business. In comments to the FCC in the
thirteenth annual report on video competition in November 2006, Comcast explained that “Many
networks have jumped head-first into Internet video, providing consumers with an interactive
232. See Clements & Brown, supra, at 126 (“According to the Federal Communications Commission (FCC),
the SHVA had the general effect of preventing satellite-based companies from delivering local broadcast stations to subscribers in most circumstances.”).
233. Id. 234. Id. at 132. 235. Id. 236. Id. at 133.
-72-
NAVIGANT ECONOMICS
5102853
alternative to traditional TV-set viewing.”237 Comcast provided further evidence that Internet
video is a substitute to cable television: “All of these modalities of communications are
important to younger consumers, all are part of the paradigm shift to a ‘what-you-want-when-
you-want-it’ world, and all of them compete with traditional and not-so-traditional video
distribution technologies for time, attention, and dollars.”238 Glenn Brit, CEO of Time Warner
Cable, explained the pending threat in May 2009: “The reality is, we’re starting to see the
beginnings of cord cutting where people, particularly young people, are saying all I need is
broadband.”239 In April 2009, Comcast’s president and chief operating officer, Steve Burke,
likened television viewers’ movement to online video to “wildfire.”240 According to Melinda
Witmer, Time Warner Cable’s programming chief, OTT video providers are actual (and not just
potential) competitors: “We wake up every day and there is some new competitor out there—a
Roku or a Boxee. People like to think of cable operators as monopolists, but we face a lot of
competition just to keep the business we have.”241 Despite this threat, incumbent cable operators
still maintain an advantage over OTT video providers by virtue of their control over video
content, as explained by Time Warner executive Jeffrey Bewkes: “We’re fortunately in a
position where this [Internet video] doesn’t cost us much money. We have an advantage and
we’re going to use that advantage.”242
237. Comcast Comments at 29-30 (emphasis added). 238. Comcast Comments at 59 (emphasis added). 239. Christopher Lawton, More Households Cut the Cord on Cable, WALL STREET JOURNAL, May 28, 2009,
available at http://online.wsj.com/article/SB124347195274260829.html. 240. Tom Lowry, Cable TV: Pushing to Become More Web-like, BUSINESSWEEK, Apr. 16, 2009, available at
http://www.businessweek.com/magazine/content/09_15/b4126050298367.htm [hereinafter Cable TV]. 241. Ronald Grover, Tom Lowry & Cliff Edwards, Revenge of the Cable Guys, BUSINESSWEEK, Mar. 11,
2010, available at http://www.businessweek.com/magazine/content/10_12/b4171038593210_page_2.htm [hereinafter Revenge of the Cable Guys].
242. Knowledge at Wharton, Cable TV Follows Its Subscribers to the Internet, July 22, 2009, available at http://knowledge.wharton.upenn.edu/article.cfm?articleid=2295. In the same article, the former chief economist of the FCC, Gerald Faulhaber, described Internet video as “very disruptive,” a technology that “attacks the model of cable television.” Id.
-73-
NAVIGANT ECONOMICS
5102853
115. The recent increase in the quantity of video programming available online along
with a burgeoning ecosystem of software and hardware that enable customers to view video
delivered over the Internet on their televisions is making this threat increasingly severe. Several
analysts have noted the threat posed by online video service to traditional MVPDs.243 The
transition of Internet video from a complement to a substitute for cable television can be seen in
recent Internet user data:
According to the Pew Internet and American Life Project, from 2007 to 2009, the number of adults who have watched movies or television shows on the Internet doubled from 16 to 32 percent; the number who watched news and sports videos increased to 43 percent and 21 percent, respectively.244 Because that type of programming traditionally has been offered by MVPDs, it is reasonable to infer that such activity represents a displacement of time that would otherwise be spent watching television. (In contrast, the viewing of a home-made video on the Internet or a YouTube short clip likely represents an addition of time spent watching video on television.) Indeed, among online video watchers surveyed in June 2009, 8 percent connected their computer to their television so they can watch online video on a television screen, and 10 percent have paid to watch a video.245 Nearly seven in ten adult internet users (69 percent) have used the Internet to watch or download video.246
According to comScore, the average American web user spent about ten minutes a day in early 2009 (slightly under six hours per month) viewing online video, compared with roughly 300 minutes spent watching live television (slightly over 150 hours per month),247 suggesting that the two platforms were mild substitutes not long ago. Because the audience for online video is young and growing, and because the broadcast networks are replicating their content online, however, the Internet has quickly emerged as a serious threat to cable television. Over the course of 2009, the average amount of time among web users spent watching videos online more than doubled to nearly thirteen hours per month (from slightly under six hours per month in the beginning of the year).248
243. See, e.g., Piper Jaffray, Internet Video: Field of Dreams or Nightmare on Elm Street?, Nov. 2009, at 5
(“In 3-5 years we expect internet delivery will start to rival the physical distribution models. … We see this as a sign that online video is in its infancy, and that its growth is unlikely to slow over the next several years as the quality and consistency of the online video experience improves.”); UBS Investment Research, Can Pay TV Benefit from Online Video?, at 9 (“Online video could become a game-changer for the entire media value chain and should be one of the core industry drivers for media investors. … Online video could affect every aspect of the media space. Content production and traditional distribution platforms (broadcast and pay TV) would have to adapt. We are already seeing Internet companies expanding into studio content.”).
244. Pew Internet and American Life Project, The State of Online Video, June 3, 2010, at 2. 245. Id. 246. Id. 247. Time Warner and the internet: After the divorce, THE ECONOMIST, May 7, 2009. 248. comScore Data Shows 2009 Was a Blistering Year for Online Video, VIDEO NUZE, available at
Parks Associates reported that the number of U.S. broadband households watching premium online content doubled in 2009; as of April 2010, over 25 million U.S. broadband households regularly watched full-length television shows online, and over 20 million watched movies online.249 Parks estimates that some 900,000 U.S. homes did not pay for television and relied solely on Internet-based television in 2008.250
The Convergence Consulting Group estimated that, from 2008 to 2010, 800,000 U.S. households disconnected their cable television service and watched their television online; that number was also expected to double by 2011.251
According to a Yankee Group survey, one in eight consumers will eliminate or scale back their cable, satellite or other pay television service in 2010.252
A survey released in September 2009 by the Conference Board estimated that nearly one quarter of U.S. households have watched television online, and that 20 percent of respondents said they were watching less television delivered through traditional broadcast or paid cable-type providers.253 Another survey by the Consumer Electronics Association trade group found that 15 percent of viewers would consider cutting out traditional means of watching television altogether.
These data are consistent with the notion that online video is a substitute and not a complement
for cable television service.
116. Video content traditionally delivered via the cable and satellite providers that is
now also available online includes content available from Hulu.com (reviewed in detail below),
programmers’ websites (such as Comedy Central), and the growing libraries of services like
Apple’s iTunes, Amazon’s Unbox, and Netflix. By accessing comedycentral.com, users can
watch episodes of many of that network’s series including The Daily Show, The Colbert Report,
South Park, and many more. This content, while also available on the Comedy Central network,
is also aggregated in different ways on the Comedy Central website. An example is the four-to-
249. Parks Associates finds over 25 million U.S. broadband households regularly watch full-length TV shows
online, Apr. 20, 2010, available at http://www.fiercetelecom.com/press_releases/parks-associates-finds-over-25-million-u-s-broadband-households-regularly-watch-full.
250. Christopher Lawton, More Households Cut the Cord on Cable, WALL STREET JOURNAL, May 28, 2009, available at http://online.wsj.com/article/SB124347195274260829.html.
251. Ryan Fleming, New Report Shows More People Dropping Cable TV for Web Broadcasts, Apr. 16, 2010, available at http://www.digitaltrends.com/computing/new-report-shows-that-more-and-more-people-are-dropping-cable-tv-in-favor-of-web-broadcasts.
252. David Goldman, One in eight to cut cable or satellite TV in 2010, CNN MONEY, Apr. 30, 2010, available at http://dailycaller.com/2010/04/30/one-in-eight-to-cut-cable-or-satellite-tv-in-2010.
253. David Colker, Pulling the plug on television: More people are turning off the TV and turning on their computers to watch their favorite programs via the Internet, LOS ANGELES TIMES, Oct. 31, 2009.
-75-
NAVIGANT ECONOMICS
5102853
five minute weekly recap of The Daily Show that provides the highlights of the approximately 60
minutes worth of programming from the preceding week. There are a number of other services
that deliver video content over the internet, including, but not limited to, Amazon’s Video on
users to purchase and watch television programming on an a-la-carte basis from the comfort of
their couch. TiVo’s latest DVR, the Premiere, integrates Internet-delivered video content from
Amazon, Netflix, and Blockbuster with traditional television video content. Like Boxee, the
Premiere does not have access to Hulu.com content. There is a whole host of other software
available that runs on third-party set-top boxes, such as Nintendo’s Wii and Sony’s Playstation,
that enable streaming of internet-delivered video directly to the television. For example, with a
WiFi-capable Nintendo Wii and a Netflix account, users can stream movies from Netflix directly
to their television, choosing from a library of titles that often far exceeds what is available from a
cable providers’ on-demand service for a flat monthly fee instead of a per-movie charge.
Alternatively, a broadband customer with a Wii can download software on her computer from
Playon.tv, which replicates a streaming video from Hulu.com and other websites to the user’s
television.
120. Comcast, along with some of the other cable providers, have addressed this threat
by making some of the video content they distribute available online, but only to paying
customers of both their cable television and cable modem services. These strategies are reviewed
in Part II below.
II. COMCAST CURRENTLY ENGAGES IN A HOST OF ANTICOMPETITIVE STRATEGIES TO
PROTECT ITS MARKET POWER
121. In this section, I provide an economic framework that can be used to explain
Comcast’s exclusionary conduct, and I summarize how this conduct has manifested itself. There
is no question whether Comcast discriminates to impair its downstream MVPD rivals. The only
remaining issue (addressed in Part III below) is whether the proposed transaction will enhance
Comcast’s ability and incentive to extend their market power.
-77-
NAVIGANT ECONOMICS
5102853
A. The Economic Framework for Assessing Comcast’s Exclusionary Conduct
122. Antitrust economics considers exclusionary conduct to be anticompetitive if it
impairs a rival’s efficiency or its ability to impose price discipline. Such conduct can occur by
raising a rival’s costs, by degrading a rival’s quality of service,256 or by depriving rivals of
economies of scale.257 Anticompetitive effects may be realized regardless of whether the rival is
driven out of the market entirely;258 they require that the rival faces increasing marginal costs or
large upfront costs or both. If the conduct forces the rival to operate on a higher portion of its
marginal cost curve, then such conduct would achieve partial foreclosure, and the rival would not
be able to constrain prices as effectively. In the extreme case, when the conduct prevents the
rival from covering its average variable costs, the conduct would induce exit and thereby achieve
complete foreclosure.
123. Exclusionary strategies can take several forms, and such strategies can be used to
extend monopoly power from one market into another, or to maintain monopoly power in a
given market. When the markets in question are vertically aligned, exclusionary conduct is
referred to as “vertical restraints,” and the effect of such conduct is called “vertical
foreclosure.”259 Because regional sports programming, national sports programming, local
broadcast affiliates, and online programming are inputs in the production process of video
256. See Steven C. Salop & David T. Scheffman, Cost-Raising Strategies, 36 J. IND. ECON. 19 (1987). 257. See, e.g., Richard A. Posner, Vertical Restraints and Antitrust Policy, 72 U. CHI. L. REV. 229, 239 (2005)
(“Economies of scale are a market rather than a firm attribute. To the extent that the loyalty rebates raised LePage’s average costs by shrinking its output and thus preventing it from achieving the available economies of scale, this was not a consequence of 3M being a more efficient company in a sense relevant to antitrust policy.”).
258. See, e.g., Einer Elhauge, Defining Better Monopolization Standards, 56 STAN. L. REV. 253 (2003); Dennis W. Carlton & Michael Waldman, The Strategic Use of Tying to Preserve and Create Market Power in Evolving Industries, 33 RAND J. ECON. 194 (2002); Michael D. Whinston, Tying, Foreclosure and Exclusion, 80 AM. ECON. REV. 837 (1990); Thomas G. Krattenmaker & Stephen C. Salop, Anticompetitive Exclusion: Raising Rivals’ Costs to Achieve Power Over Price, 96 YALE L.J. 209, 234-45 (1986); Stephen C. Salop & David T. Scheffman, Raising Rivals’ Costs, 73 AM. ECON. REV. 267 (1983) (Special Issue) [hereinafter Raising Rivals Costs].
259. See Patrick Rey & Jean Tirole, A Primer on Foreclosure (reprinted in III HANDBOOK OF INDUSTRIAL
ORGANIZATION, Mark Armstrong & Rob Porter, eds., 2145-2220, 2007), at *8 (“Vertical foreclosure may arise when a firm controls an input that is essential for a potentially competitive industry. The bottleneck owner can then alter competition by denying or limiting access to the input.”) [hereinafter Foreclosure Primer].
-78-
NAVIGANT ECONOMICS
5102853
distribution, it is appropriate to analyze Comcast’s conduct through the lens of vertical
foreclosure. The FCC has found that significant incentives exist for vertically integrated
MVPDs—and Comcast in particular—to use vertical foreclosure strategies to raise rivals’
(competitors’) costs and thereby achieve partial or total foreclosure.260
124. Anticompetitive strategies that are designed to increase a rival’s costs are known
as “raising-rival-cost” strategies.261 A basic tenet of economics is that it is always better to
compete against a rival with higher costs.262 A firm can raise a rival’s costs in many ways,
including by interfering with its rival’s production or selling methods, interfering through
government regulation263 (see discussion of Comcast’s influence over LFA’s approval process
above), using tie-ins with other products, raising switching costs, or raising an input price (see
discussion of Comcast’s refusal to supply its affiliated RSNs to DBS rivals below).264 An
equivalent strategy that can be used to disadvantage a rival—and artificially inflate prices—is to
degrade a rival’s quality of service.265 For example, by denying DBS rivals’ access to CSN-
260. See In the Matter of Applications for Consent to the Assignment and/or Transfer of Control of Licenses,
MB Dkt. No. 05-192, Memorandum Opinion and Order, released July 21, 2006, FCC 06-105 [hereinafter FCC Adelphia Order] at ¶ 118 (“One way by which vertically integrated firms can raise their rivals’ costs is to charge higher programming prices to competing MVPDs than to their affiliated MVPDs.”). See also id. at ¶ 123 (“We find that the transactions [the Adelphia purchase and related cluster-driven swaps] would enable Comcast and Time Warner to raise the price of access to RSNs by imposing uniform price increases applicable to all MVPDs, including their own systems, by engaging in so-called ‘stealth discrimination,’ or by permanently or temporarily withholding programming. As commenters contend, such strategies are likely to result in increased retail rates and fewer choices for consumers seeking competitive alternatives to Comcast and Time Warner.”).
261. See Salop & Scheffman, supra, at 19 (“In this paper, we show that strategies designed to raise rivals’ costs have a number of advantages over predatory pricing. First, cost-raising strategies do not have an inherent problem of credibility. Such strategies may be profitable whether or not the rivals exit, since higher cost rivals have an incentive to cut back output and raise prices immediately, which may make it possible for the predator to reap gains even in the short run.”).
262. See, e.g., DENNIS CARLTON & JEFFREY PERLOFF, MODERN INDUSTRIAL ORGANIZATION 371(Addison Wesley 4th ed. 2005) (“A firm clearly benefits if it can only raise its rivals’ costs.”).
263. Id. at 372 (“By supporting government regulation so that a new rival cannot adopt their production techniques, incumbent firms can preserve and protect their market position and make it more costly for entrants to compete.”).
264. Id. at 371-73. A refusal to supply an input is tantamount to setting an infinite price for that input. 265. See, e.g., Nicholas Economides, The Incentive for Non-Price Discrimination by an Input Monopolist, 16
INT’L J. IND. ORG. 271 (1998) (“This paper considers the incentive for non-price discrimination of a monopolist in an input market who sells in an oligopoly downstream market through a subsidiary. Such a monopolist can raise the
-79-
NAVIGANT ECONOMICS
5102853
Philadelphia, Comcast effectively degraded their quality of service (for any given price of DBS
service, the failure to offer the Phillies/Flyers/76ers to a Philadelphia resident is a less-
compelling offer), which impaired DBS firms’ ability to constrain Comcast’s cable television
prices. These strategies represent a particularly effective means of monopolizing a downstream
market (in this case, the MVPD market) because they can engender anticompetitive harm
without forcing rivals to exit the market completely and do not require any short-term profit
sacrifice.
125. Over the course of the last few decades, defendants accused of using vertical
restraints for anticompetitive purposes have invoked the Chicago School’s “One-Monopoly-
Profit Theorem,” associated with economists who taught at the University of Chicago in the
1950s, as proof that such conduct could be motivated by efficiency reasons only. Consider a
situation similar to the fact pattern here in which a firm has monopoly power over one product
but faces competitors in the supply of some complementary good. The Chicago argument is
essentially that because there is only one source of monopoly profit—the critical input—any
additional monopoly profits a firm made by leveraging its power in the downstream market
would come at the direct cost of the upstream market.266 However, economic research has proven
costs of the rivals to its subsidiary through discriminatory quality degradation.”); see also Nicholas Economides, Raising Rivals’ Costs in Complementary Goods Markets: LECs Entering into Long Distance and Microsoft Bundling Internet Explorer, New York University Center for Law and Business Working Paper #CLB-98-004 at 1 (“Suppose that a firm monopolizes a market that provides a required input for a complementary market. Assume further that the monopolist also provides the complementary product through a subsidiary. Besides the monopolist’s subsidiary, a number of other independent companies participate in the complementary good market. Under these conditions, we show that the monopolist has an incentive to (i) raise the costs of the rivals to its downstream subsidiary; and (ii) degrade the quality of the monopolized good offered to the downstream rivals of its subsidiary. These actions of the monopolist reduce competition and social welfare.”).
266. See Rey & Tirole, supra, at 11 (“It [the Chicago School theorists] argued that there is a single source of monopoly profit, and that a bottleneck monopolist can already earn the entire monopoly profit without extending its market power to related segments, and so in the absence of efficiency gains, vertical integration cannot increase the profitability of the merging firms.”).
-80-
NAVIGANT ECONOMICS
5102853
that anticompetitive foreclosure can be profitable and effective even when the assumptions of the
one-monopoly profit theorem hold.267
126. With respect to exclusive dealing, Chicago-school economists argued that buyers
would demand to be fully compensated by the seller before entering into an arrangement
subjecting them to future monopoly power.268 But economists now recognize that when buyers
are unable to coordinate their responses, no single buyer can stop the exclusion by itself.269
Moreover, if those buyers are distributors, the upstream supplier can secure their acquiescence
by sharing with them some of its expected monopoly profits.270 As the former chief economist of
the Antitrust Division of the DOJ Dennis Carlton has explained, exclusive dealing may be
particularly anticompetitive when it deprives rivals of the scale necessary to achieve
efficiencies.271 When considered in this light, exclusive dealing may allow a monopolist to
maintain its power by impairing the ability of rivals to grow into effective competitors.
127. One could analyze Comcast’s refusal to supply inputs to downstream rivals
through the lens of a tie-in. In this case, the tying product is input or programming (for example,
a local NBC affiliate or NBCU online content) and the tied product is Comcast’s cable television
service. Comcast forces consumers in the Philadelphia DMA to purchase Comcast cable
television service as a condition of getting CSN-Philadelphia—that is, Comcast ties the purchase
of CSN-Philadelphia to a customer’s purchase of Comcast cable television service. Similarly,
267. See Salop & Scheffman, supra, at 32 (“Proposition 9 shows that a fixed coefficient technology is not a
sufficient condition for the absence of anticompetitive impact of a vertical merger (cf. Bork [1978]).”); see also Rey & Tirole, supra, at 12 (“The reconciliation of the foreclosure doctrine and the Chicago School critique is based on the observation that an upstream monopolist in general cannot fully exert its monopoly power without engaging in exclusionary practices.”).
268. See, e.g., ROBERT H. BORK, THE ANTITRUST PARADOX 304–09 (1978). 269. See, e.g., Ilya R. Segal & Michael D. Whinston, Naked Exclusion: Comment, 90 AM. ECON. REV. 296,
307 (2000). 270. See, e.g., A. Douglas Melamed, Exclusive Dealing Agreements and Other Exclusionary Conduct—Are
There Unifying Principles?, 73 ANTITRUST L.J. 375, 404 (2006). 271. Dennis W. Carlton, A General Analysis of Exclusionary Conduct and Refusal to Deal—Why Aspen and
Kodak Are Misguided, 68 ANTITRUST L.J. 659, 663 (2001).
-81-
NAVIGANT ECONOMICS
5102853
Comcast ties the purchase of its online content portfolio (via Fancast) to a customer’s purchase
of Comcast cable television service (and its cable modem service). According to the economic
literature, a monopolist may have an incentive to use a tie to monopolize a second market to (1)
deprive rivals of sales sufficient to achieve scale efficiency in the tied-product market;272 or (2)
deter entry into the tied-product market;273 or (3) earn monopoly profits in the tied-good market
that are not currently available but will be in the future.274 Applied here, Comcast is motivated by
a desire to weaken extant MVPD rivals (DirecTV, Dish Network, and Verizon FiOS) and to
deter entry of nascent rivals (OTT video providers).
128. Finally, one could analyze Comcast’s refusal to supply inputs to downstream
rivals through the lens of exclusive dealing. Exclusive dealing often requires a buyer to deal
exclusively with a seller, a seller to deal exclusively with a buyer, or a seller to enact policies
requiring customers to deal exclusively with it. Exclusive dealing can be anticompetitive
whenever it allows one manufacturer to monopolize the most efficient distribution channel and
thereby prevent its rivals from competing effectively, or when it thwarts entry or inhibits the
growth of existing rivals. Justice Breyer, among others, has recognized that exclusive dealing can
be used to foreclose “enough outlets, or sources of supply, in the hands of a single firm (or small
group of firms) to make it difficult for new, potentially competing firms to penetrate the
market.”275 An important inquiry is whether the exclusive arrangement forecloses competition in
a substantial share of the line of commerce affected.276
272. Id. at 667–68. 273. See, e.g., Barry Nalebuff, Bundling as an Entry Barrier, 119 Q. J. ECON. 159, 183 (2004). 274. Dennis W. Carlton & Michael Waldman, Tying, Upgrades, and Switching Costs in Durable-Goods
Markets, NBER Working Paper No. 11407, at 3 (2005), available at http://www.nber.org/papers/w11407. 275. Interface Group, Inc. v. Mass. Port Auth., 816 F.2d 9, 11 (1st Cir. 1987) (Breyer, J.) (emphasis in
original) (citations omitted). 276. In his antitrust treatise, Professor Hovenkamp explains that exclusive dealing requires “a plaintiff to
show that the defendant has significant market power, that the exclusivity agreement serves to deny market access to
-82-
NAVIGANT ECONOMICS
5102853
129. It bears emphasis that the economic literature does not distinguish between the
exact form of the vertical restraint (exclusive dealing versus a tie-in). A showing of scale
economies in the tied product is important for assessing both tie-ins and exclusive deals from an
economic perspective. In my opinion, Comcast has abused its downstream market power by
requiring that key inputs were exclusive to Comcast as a condition of doing business.
B. Types of Exclusionary Conduct
130. Comcast’s exclusionary strategies can be grouped according to three major types:
(a) refusal to supply affiliated programming to rival MVPDs, (b) discrimination vis-à-vis
independent programming networks, and (c) tying affiliated online content with the purchase of
cable television and cable modem service.
1. Refusing to Supply Affiliated Programming to Rival Distributors on Reasonable Terms
131. Comcast’s exclusionary conduct vis-à-vis MVPD rivals have taken several forms,
ranging from outright denial of access to unreasonable terms of access. Those strategies are
briefly reviewed here.
a. Outright Denial
132. By exploiting the “terrestrial loophole” in the program access protections of the
Cable Act, Comcast has denied its DBS rivals in Philadelphia and in Portland access to
Comcast’s affiliated RSN, leading the Commission to conclude in the case of Philadelphia that
such exclusionary conduct “has had a material adverse impact on competition in the video
distribution market.”277
one or more significant rivals, and that market output to consumers is lower (or prices higher) as a result.” HERBERT
HOVENKAMP, THE ANTITRUST ENTERPRISE 206 (2005). 277. In the Matter of Implementation of the Cable Television Consumer Protection and Competition Act of
1992, Development of Competition and Diversity in Video Programming Distribution: Section 628(c)(5) of the Communications Act, Sunset of Exclusive Contract Prohibition, Review of the Commission’s Program Access
-83-
NAVIGANT ECONOMICS
5102853
i. Comcast SportsNet Northwest
133. For years, Portland Trail Blazer games were available to cable and satellite
subscribers via the BlazerVision Pay Per View package.278 In 2001, Paul Allen, owner of the
Trail Blazers, abolished the pay-per-view system when he founded the Action Sports Cable
Network (ASCN) to broadcast Trail Blazer games and other valuable local sports programming,
including Seattle Seahawks and Portland Fire games.279 ASCN planned to team with ESPN
Network and KGW-TV to maximize fan access to local teams. The Blazers increased broadcast
production by 30 percent to meet the demand for increased coverage.280 For sixteen months,
ASCN struggled to secure a distribution deal with Portland’s “primary cable television
provider,” AT&T Broadband, which was purchased by Comcast in the midst of the negotiations
over ASCN carriage in 2002.281 A “shift in management” at AT&T Broadband in late 2001
caused the cable company to renege on its “verbal agreement” to carry ASCN in Oregon and
southwest Washington.282 ASCN was forced to fold after failing to secure a distribution deal
with the cable company, leaving the Blazers to scramble to find a home in time for the 2002-
2003 NBA season.283
134. After conducting discussions with cable providers, satellite companies, Fox
Sports Northwest, and local stations, Allen sold the rights to broadcast Trail Blazer games to Fox
Rules and Examination of Program Tying Arrangements, MB Dkt. Nos. 07-29, 07-198, Report and Order and Notice of Proposed Rulemaking ¶ 39 (rel. Oct. 1, 2007)
278. R. Thomas Umstead, DirecTV Fishes for Subs with Trail Blazers, MULTICHANNEL NEWS (Nov. 4, 1996) [hereinafter DirecTV Fishes for Subs].
279. Kristina Brenneman, Paul Allen’s Sporting Plans, PORTLAND BUS. J. (Feb. 2, 2001) [hereinafter Paul Allen’s Sporting Plans].
280. Paul Allen’s Sporting Plans, supra. 281. Andy Giegerich, Fox Hunts for Local Space to Increase Sports Coverage, PORTLAND BUS. J. (Mar. 26,
2004) [hereinafter Fox Hunts for Local Space]. For a discussion of the AT&T Broadband-Comcast acquisition, see Bob Liu, Comcast Offers to Acquire AT&T Broadband, INTERNETNEWS.COM (Jul. 8, 2001); David Lieberman, Comcast to Buy AT&T Broadband, USA TODAY (Dec. 20, 2001).
282. Action Sports Cable Network Blames AT&T for Failed Negotiations, PORTLAND BUS. J. (Oct. 31, 2001). 283. Fox Hunts for Local Space, supra.
-84-
NAVIGANT ECONOMICS
5102853
Sports Northwest in November 2002.284 The network signed a five-year deal and carried 30
Blazers games during the first year of its contract.285 Fox Sports Northwest reached more than
3.4 million homes in the Northwest in 2006,286 giving the team broad distribution; however, the
network was known to have a Seattle focus.287 The contract with Fox Sports Northwest expired
at the close of the 2006-2007 season and the parties failed to come to a renewal agreement.288
135. Comcast began negotiations for broadcasting rights with the Portland Trail
Blazers in the spring of 2007, soon after renewal talks with Fox Sports Net collapsed.289 Comcast
won the rights to the games after agreeing to a 10-year, $120 million contract.290 In May 2007,
Comcast and the Trail Blazers announced the creation of a new regional sports network,
Comcast SportsNet Northwest (CSNW), the core programming of which would be live coverage
of the Portland Trail Blazers.291 Trail Blazers Executive Vice President of Business Operations,
Mike Golub, proclaimed that the partnership would “serve sports fans in ways never before
possible and deliver an unprecedented amount of Trail Blazer games, programming and
access.”292 CSNW launched shortly before the 2007-2008 NBA season and in its first year
carried at least 55 of the 81 regular season games broadcast on television and made more of
284. Notebook: Agent Alleges Kings Reneged on Secret Deal, SEATTLE TIMES (Nov. 14, 2002) [hereinafter
Notebook 2002], available at http://community.seattletimes.nwsource.com/archive/?date=20021114&slug=nbanotes14.
285. Notebook 2002, supra; Jason Vondersmith, Local teams, local TV, PORTLAND TRIBUNE (Jul. 15, 2008) [hereinafter Local teams, local TV].
286. Select FSN Trail Blazers Games To Be Televised in HD, NBA.COM, available at http://www.nba.com/blazers/news/FSN_Trail_Blazers_Games_In_HD-201079-1218.html.
287 Local teams, local TV, supra. 288 Mike Rogoway, Comcast, Blazers to announce cable deal (updated), THE OREGONIAN (May 21, 2007)
[hereinafter Comcast, Blazers to announce cable deal], available at http://blog.oregonlive.com/business/2007/05/comcast_blazers_to_announce_ca.html.
289. Comcast, Blazers to announce cable deal, supra. 290. John Canzano, Trail Blazers start another season with many fans held hostage by Comcast SportsNet,
THE OREGONIAN (Oct. 10, 2009), available at http://www.oregonlive.com/sports/oregonian/john_canzano/index.ssf/2009/10/canzano_trail_blazers_start_an.html.
291. Comcast, Blazers to announce cable deal, supra. 292. Comcast Sports Net, Portland Trail Blazers Announce a New Regional Sports Network, NBA.COM,
available at http://www.nba.com/blazers/news/Comcast_Sports_Net_Portland_T-225869-1218.html.
-85-
NAVIGANT ECONOMICS
5102853
those games available in HDTV than ever before. Those who did not subscribe to Comcast,
however, were able to access fewer than 30 regular season games during the 2007-2008 season,
compared with 61 during the 2006-2007 season.293 As of September 2007, CSNW reached just
590,000 subscribers in Oregon and southwest Washington.294 Although a number of small cable
carriers and Verizon FiOS carried the sports network, many local fans who subscribed to
DirecTV and Dish Network were still unable to receive Trail Blazer programming.295 CSNW
began offering its Trail Blazers coverage live over the Internet for a fee to local subscribers
during the 2009-2010 season, but refused to offer the package to Blazers fans (and fans of other
NBA franchises carried on Comcast-owned regional sports networks) who could not watch the
games through their cable providers.296 Tim Fitzpatrick, vice president of communications for
CSN Philadelphia, acknowledged that Comcast was only interested in serving its “existing
audience” with the convenience of Internet streaming.297 In January 2010, three years into the
ten-year contract, and frustrated by Comcast’s failed negotiations with the two major satellite TV
carriers and some local cable operators, the Blazers announced that they would “pursue all of
[their] rights under [the] contract.”298 As of April 2010, DISH Network and DirecTV were still
without access to CSNW.
293. Comcast, Blazers to announce cable deal, supra. 294. Brent Hunsberger, Comcast-Trail Blazers Deal Still Leaving Some Viewers in the Dark, THE OREGONIAN
(Sept. 26, 2007), available at http://blog.oregonlive.com/playbooksandprofits/2007/09/comcasttrail_blazers_deal_stil.html.
295. John Canzano, Trail Blazers Start Another Season with Many Fans Held Hostage by Comcast SportsNet, THE OREGONIAN (Oct. 10, 2009); Jon Hemingway, Portland FiOS TV Subscribers Get Comcast SportsNet Northwest, BROADCASTING & CABLE (Jan. 17, 2008).
296. Mike Rogoway, Disappointed Blazers “Pursuing All of Our Rights” in Comcast Deal, THE OREGONIAN (Jan. 7, 2010) [hereinafter Disappointed Blazers], available at http://www.oregonlive.com/business/index.ssf/2010/01/blazers_say_they_are_pursuing.html; Tom Lowry, Comcast Targets ESPN by Streaming Pro Sports Games, BUSINESS WEEK (Dec. 16, 2009), available at http://www.businessweek.com/technology/content/dec2009/tc20091216_396786.htm.
136. In Philadelphia, Comcast denies access to CSN-Philadelphia to DBS providers
through the so-called “terrestrial delivery” loophole, which allows programmers to deny
terrestrially-delivered programming to any MVPD.299 Comcast is the majority owner of CSN-
Philadelphia, with approximately an 80 percent equity share in the RSN.300 Ever since Comcast
acquired the rights from SportsChannel Philadelphia and PRISM in August 1997, Comcast has
refused even to negotiate with DBS providers regarding carriage of CSN-Philadelphia.301
Comcast also controls future access to the 76ers’ and Flyers’ carriage rights. Specifically, in
1996 Comcast acquired a controlling interest in Spectacor (now “Comcast Spectacor”), a holding
company that owns the 76ers, the Flyers, and the Wachovia Center stadium that hosts the
teams.302
137. By August 1997, Comcast acquired all the local telecasting rights of Philadelphia
Flyers hockey games, Philadelphia 76ers basketball games, and Philadelphia Phillies baseball
games previously held by Rainbow Sports, the owner of SportsChannel.303 Thereafter,
SportsChannel announced that it would cease to operate as of September 30, 1997.304 On
October 1, 1997, CSN-Philadelphia debuted as a new channel on Comcast’s basic service tier in
the Philadelphia area, and it was distributed only through terrestrial microwave and fiber
299. See, e.g., In the Matter of Applications of Adelphia Communications Corporation, Comcast Corporation,
and Time Warner Cable Inc., for Authority to Assign and/or Transfer Control of Various Licenses, MB Docket No. 05-192, Comments of DirecTV, Inc., July 21, 2005, at 16-17 [hereinafter DirecTV Comments].
300. See Eleventh Annual MVPD Report, supra, at 141. 301. In the Matter of DIRECTV, Inc. v. Comcast Corporation, Comcast-Spectacor, L.P., Comcast SportsNet,
302. See Mark Robichaux, Comcast to Buy Stake in Pro Teams in Philadelphia, WALL STREET JOURNAL, Mar. 20, 1996, at B7 (“Cable giant Comcast Corp. jumped into the game of pro sports, saying it agreed to buy a majority stake in a venture that will own the Philadelphia 76ers basketball team, the Flyers hockey team, as well as two sports arenas.”).
303. SportsNet MO&O, supra, at 21,834. 304. Id.
-87-
NAVIGANT ECONOMICS
5102853
technology.305 Before introducing CSN-Philadelphia as a new channel, Comcast indicated that
CSN-Philadelphia’s programming would not be available to any national DBS provider.306 As I
demonstrate in detail below, Comcast’s discriminatory and continuing refusal to provide CSN-
Philadelphia to its primary downstream competitors caused DBS providers to experience
significantly lower-than-expected penetration rates in the Philadelphia DMA.
138. In January 2010, the Commission voted to close the “terrestrial loophole,” and it
permitted a DBS provider to file a program access complaint with the Commission to show it has
been harmed by Comcast’s withholding of CSN-Philadelphia. After a successful review, the
FCC could force Comcast to reach a carriage deal for Comcast SportsNet. One day after that
decision, Comcast announced that it would challenge the FCC action in an administrative
process at the federal agency.307 Despite the growing market share of DBS providers in
Philadelphia, Craig Moffett of Sanford C. Bernstein & Co. estimated that Comcast’s refusal to
license CSN-Philadelphia has yielded an additional 450,000 subscribers for Comcast.308
Accordingly, DBS market share in Philadelphia would have increased even faster but for
Comcast’s exclusionary practice.
b. Excessive Fees and Tying Carriage of RSNs to Less Popular Networks
139. Comcast launched Comcast SportsNet Chicago (CSN-Chicago)—which features
the Chicago Bulls (NBA), Blackhawks (NHL), Cubs (MLB), and White Sox (MLB) in 2003.309
These teams previously were carried on FSN Chicago, an unaffiliated RSN. Once Comcast’s
RSN acquired the rights to these teams, Comcast demanded that DirecTV pay a rate for CSN-
305. Id. 306. Id. 307. Bob Fernandez, Comcast to fight FCC ruling on sports telecasts, PHILADELPHIA INQUIRER, Jan. 22, 2010,
available at http://www.philly.com/philly/business/82341347.html#axzz0orclBgQ9. 308. Bob Fernandez, FCC: Comcast must share Phila. sports coverage, PHILADELPHIA INQUIRER, Jan. 20,
2010, available at http://www.philly.com/philly/business/82230427.html#axzz0orqf07iX. 309. COMCAST CORP. SEC FORM 10-K (filed February 23, 2005), at 50.
-88-
NAVIGANT ECONOMICS
5102853
Chicago that was roughly 100 percent more than what DirecTV had been paying FSN Chicago
for the same content.310 Even if Comcast charged all MVPDs in Chicago this higher rate, the
increased rates asymmetrically affect Comcast’s MVPD rivals in Chicago because Comcast sees
much of the rate increase as an intra-company transfer due to its 30 percent stake in CSN-
Chicago.311 Comcast’s pricing of CSN-Chicago has also prompted a program access complaint
by Dish Network.
c. Manipulating Zones or “Stealth Discrimination”
140. Similarly, Comcast discriminates against DirecTV in Sacramento by effectively
raising the price DirecTV pays for Sacramento Kings (NBA) games carried on Comcast
SportsNet West (CSN-West). CSN-West’s marquee programming—Sacramento Kings
basketball games—must be blacked out in large parts of its service area, including the San
Francisco DMA. As a condition of permitting DirecTV to access CSN-West, Comcast forced
DirecTV to carry the network in three “zones”—an inner zone consisting of areas in and around
Sacramento, an outer zone consisting of areas within 150 miles of Sacramento, and an “outer-
outer” zone consisting of the San Francisco DMA. Comcast charges the highest rates per
subscriber for the inner zone, and charges lower rates for zones further out. Although Comcast
charges the lowest per-subscriber rate for the outer-outer zone, the cost to DirecTV to carry
CSN-West in the outer-outer zone is especially burdensome because that zone has twice as many
subscribers as the inner and outer zones combined. As a result, the effective per-subscriber rate
for subscribers who can view Kings games (those in Sacramento but not San Francisco) is much
higher than the rates DirecTV is required to pay for comparable marquee RSN programming it
obtains from Comcast. DirecTV has stated, for example, that the per-subscriber rates it is
310. DirecTV Comments, supra, at 20-21. 311. Timothy Dwyer, Nats Caught in a TV Rundown, WASHINGTON POST, June 28, 2005, at A1.
-89-
NAVIGANT ECONOMICS
5102853
required to pay for CSN-West are higher than the rates it pays for FSN Bay Area—an RSN that
carries live games for four men’s professional sports teams.312
2. Denying Independent Programming Networks’ Access to Comcast’s Subscribers
141. Comcast gives preferential treatment to its affiliated, national sports networks,
Versus and the Golf Channel, relative to unaffiliated sports networks. As of the third quarter of
2009, Comcast had approximately 23.8 million total subscribers,313 nearly all of whom have
access to Comcast’s affiliated sports networks on Comcast’s “Standard Service.” Comcast
carries its affiliated national sports networks, the Golf Channel and Versus, on a tier that reaches
nearly 100 percent of its subscribers. In contrast, Comcast’s Sports Tier reaches only 11.3
percent of its subscribers (equal to 2.7 million divided by 23.8 million). It is there that Comcast
places unaffiliated sports networks.
142. Table 4 shows the tier on which sports programming appears in Comcast’s
channel lineup in the Washington, D.C. area in January 2010, which is generally representative
of its carriage decisions in other parts of the country.
TABLE 4: SPORTS PROGRAMMING ON COMCAST BY TIER AS OF JANUARY 2010 (WASHINGTON, D.C.)
“Standard Service” Affiliation “Digital Classic” Affiliation “Sports Entertainment” Affiliation ESPN No ESPN Classic No Fox Soccer Channel No ESPN2 No ESPN U No Fox College Sports No Golf Channel Yes MLB Network Yes* Tennis Channel No Versus Yes NBA TV Yes** CBS College Sports No SportsNet MA Yes NHL Channel Yes*** GolTV No MASN No^ Speed Channel No Big Ten Network No Horseracing Television No TV Games No NFL Red Zone No^^ Sources: Comcast Channel Lineup, available at http://www.comcast.com/Customers/Clu/ChannelLineup.ashx (accessed on Jan. 4, 2010); affiliation is from 13th Annual Report, Appendix C, Table C-1; Comcast 8-K, filed 12/04/09 for the Period Ending 12/03/09, at 6. Notes: * Comcast owns 8.3 percent of MLB Network. ** Comcast holds equity in NBA TV through its ownership in the National Basketball Association. *** Comcast owns 15.6 percent of the NHL Channel, and the League provides anchor programming for Versus. ^ MASN is carried subsequent to a settlement of a carriage complaint, as is the NFL Network, which is carried on
312. See DirecTV Comments, supra at 23-25. 313. Comcast Reports Third Quarter 2009 Results, Nov. 4, 2009, at 3.
-90-
NAVIGANT ECONOMICS
5102853
Comcast’s “Digital Starter” tier, which is Comcast’s most broadly distributed level of digital service. ^^ Comcast also sells the HD version of the NFL Red Zone as part of its extra-charge HD package. As Table 4 shows, none of the sports networks carried on Comcast’s “Sports Entertainment” tier
is affiliated with (or owned by) Comcast. In contrast, with the exception of ESPN channels—
which have sufficient countervailing market power vis-à-vis Comcast by virtue of their
significant sports holdings to obtain broad access for their networks—all of the sports networks
that are carried on Comcast’s “Standard Service” tier are either affiliated with (and owned by)
Comcast (Versus, the Golf Channel, SportsNet Mid-Atlantic), or are carried by Comcast
subsequent to the settlement of an FCC program-carriage complaint (MASN). For completeness,
Table 4 also shows sports networks carried on Comcast’s “Digital Classic” tier in Washington,
D.C., which achieves greater distribution than its Sports Entertainment tier but less distribution
that its Standard Service tier. On its Digital Classic tier, Comcast owns a minority equity stake in
the MLB Network (8.3 percent), the NHL Network (15.6 percent), and NBA TV (through its
equity stake in the National Basketball Association).314 Moreover, the National Hockey League
provides Versus its anchor programming (live professional hockey games). With the exception
of the two ESPN networks on the Digital Classic tier, which again have countervailing market
power, it appears that a sports network can avoid being relegated to Comcast’s Sports
Entertainment tier so long as Comcast is at least modestly involved in its success; significant
involvement leading to outright ownership yields access to Comcast’s Standard Service tier and
all the associated benefits, including exposure to a much larger audience and a more desirable
channel number.
314. Comcast Spectacor owns the Philadelphia 76ers, which jointly owns the National Basketball Association
along with the other teams, and thereby owns part of the equity in NBA TV. See Comcast Corp. SEC Form 10-K for fiscal year ended Dec. 31, 2007, at 1.
-91-
NAVIGANT ECONOMICS
5102853
143. As a consequence of Comcast’s discriminatory tiering policy, an unaffiliated
sports network is restrained in its ability to compete effectively for viewers, advertisers, and
programmers. Comcast is the largest MVPD in the United States, with roughly 24 million
MVPD subscribers. Competition scholars have concluded that 20 percent of a market constitutes
a significant foreclosure share.315 The reason why 20 percent is considered critical is that, in the
presence of economies of scale, missing out on such a large portion of the market can inflate a
rival’s average costs. Because Comcast’s market share of roughly 25 percent of U.S. MVPD
subscribers exceeds that 20 percent standard, economists typically would consider Comcast’s
exclusionary conduct here to be presumptively anticompetitive. Moreover, the actual foreclosure
share may exceed Comcast’s market share to the extent Comcast is acting in coordination with
other vertically integrated MSOs as part of a reciprocal compensation strategy, or other MSOs
are following Comcast’s tiering strategy, or both.316
144. Economists have derived market conditions under which exclusionary conduct
can harm competition. In particular, when markets exhibit economies of scale or when markets
display network effects, exclusionary conduct can impose barriers to entry and expansion that
make rivals smaller, causing them to be less efficient and therefore less capable of restraining the
incumbent’s prices.317 This market condition appears to be satisfied here. By refusing or
315. See PHILLIP AREEDA, IX ANTITRUST LAW 375, 377, 387 (Aspen 1991) (indicating that 20 percent
foreclosure is presumptively anticompetitive); See also HERBERT HOVENKAMP, XI ANTITRUST LAW 152, 160 (indicating that 20 percent foreclosure and an HHI of 1800 is presumptively anticompetitive).
316. See Jun-Seok Kang, Reciprocal Carriage of Vertically Integrated Cable Networks: An Empirical Study, Indiana University Working Paper, August 30, 2005, at 1 (“These [empirical] results make credible an underlying premise of a 30 percent national market share limit that the Federal Communication Commission established in 1993: namely, that MSOs may tacitly collude in their carriage decisions, having the effect of restricting market access to startup cable networks in which those MSOs have no ownership interest.”). Moreover, Comcast appeared to act in concert with other cable operators in its dealings with the NFL Network. See Transcript of Record, NFL Enterprises LLC v. Comcast Cable Communications LLC, File No. CSR-7876-P, Apr. 16, 2009, 1277: 10-1279:10 (Paul Tagliabue testimony describing Comcast CEO Brian Roberts’ suggestion that the NFL’s relationship with the “cable industry” would not be “positive” on a going-forward basis.)
317. See, e.g., Einer Elhauge, Defining Better Monopolization Standards, 56 STANFORD L. REV. 253 (2003).
-92-
NAVIGANT ECONOMICS
5102853
conditioning a programmer’s access to its highly penetrated tiers, Comcast deprives rival sports
networks of critical economies of scale.318 Because many costs of the cable network (including
program acquisition costs) are invariant to the number of subscribers, increasing a network’s
number of subscribers (and therefore increasing advertising and license revenues) reduces the
cost of providing service on a per-subscriber basis. A review of the economic literature suggests
that the scale economies associated with national television advertising are significant.
Advertisers can receive better returns by advertising with larger audiences, and as a result,
advertising rates generally increase with audience size.319 Accordingly, the ads that smaller
networks sell are sold at a significant discount disproportionate to the rates charged by their
larger and more widely distributed competitors. National sports networks are highly dependent
on advertising revenue. According to SNL Kagan, nearly half of total revenue for a national
network is derived from advertising revenue; in contrast, advertising revenues comprise only 15
to 20 percent of total revenues for regional sports networks.320
145. A second potential motivation for Comcast’s discriminatory conduct is that
Comcast seeks to expand its footprint from golf, hockey, and bull riding—the sports content
carried on its affiliated national sports networks—into complementary sports programming. That
motivation is particularly salient here because Comcast’s objective according to its 2008 Annual
Report is to expand its reach into sports programming: “We have invested and expect to continue
to invest in new and live-event programming that will cause our programming expenses to
318. See, e.g., Dennis W. Carlton, A General Analysis of Exclusionary Conduct and Refusal to Deal—Why
Aspen and Kodak Are Misguided, 68 ANTITRUST L. J. 659 (2001) 319. See, e.g., Johan Arndt & Julian L. Simon, Advertising and Economics of Scale: Critical Comments on the
Evidence, 32 J. IND. ECON. 229, 231-2 (1983); Dong Chen & David Waterman, Vertical Foreclosure in the U.S. Cable Television Market: An Empirical Study of Program Network Carriage and Positioning, Oct. 2005, at 7. Advertisers may also consider factors such as the season and time of day. But these factors are not affected by Comcast’s tiering decision.
320. Derek Baine, Comcast sports networks: Opening the kimono on $2 billion in hidden value, SNL Kagan, Oct. 8, 2009.
-93-
NAVIGANT ECONOMICS
5102853
increase in the future.”321 Comcast’s “Programming segment,” which “consists primarily of [its]
consolidated national programming networks, including E!, Golf Channel, VERSUS, G4 and
Style,” earned revenues of $1.4 billion in 2008.322 The (upstream) programming division’s
operating cash flow grew at 28.3 percent in the second quarter of 2009, whereas its
(downstream) cable division grew by only 4.1 percent.323 With those rights on an exclusive basis,
Comcast could seek higher carriage fees from its downstream rivals (DirecTV, Dish, and
Verizon) as a means of raising rivals’ costs. Alternatively, as the Commission has concluded in
other contexts,324 Comcast could deny that exclusive programming to its downstream rivals as a
means of degrading their quality of service.
a. Failed Attempts to Extract Equity in Exchange for Carriage
146. Comcast has a long history of denying carriage for discriminatory reasons to
unaffiliated sports networks such as MASN and the NFL Network. In both cases, Comcast
sought to acquire the underlying programming of the unaffiliated sports networks on an
exclusive basis. With respect to MASN, Comcast sought the rights to the Washington Nationals
games from Major League Baseball. With respect to the NFL Network, Comcast sought the
321. Comcast SEC Form 10-K, for the fiscal year ended Dec. 31, 2008, at 29 (emphasis added). 322. Id. 323. Comcast 2nd Quarter 2009 Results, Aug. 6, 2009, at 4. 324. See In the Matter of Applications for Consent to the Assignment and/or Transfer of Control of Licenses,
MB Dkt. No. 05-192, Memorandum Opinion and Order, released July 21, 2006, FCC 06-105, ¶ 118 (“One way by which vertically integrated firms can raise their rivals’ costs is to charge higher programming prices to competing MVPDs than to their affiliated MVPDs.”). See also id. ¶ 123 (“We find that the transactions [the Adelphia purchase and related cluster-driven swaps] would enable Comcast and Time Warner to raise the price of access to RSNs by imposing uniform price increases applicable to all MVPDs, including their own systems, by engaging in so-called ‘stealth discrimination,’ or by permanently or temporarily withholding programming. As commenters contend, such strategies are likely to result in increased retail rates and fewer choices for consumers seeking competitive alternatives to Comcast and Time Warner.”); Federal Communications Commission, Sunset of Exclusive Contract Provisions, Review of the Commission’s Program Access Rules and Examination of Programming Tying Arrangements, CS Dkt. Nos. 07-29, 07-198, Report and Order, rel. Oct. 1, 2007, ¶ 53 (“We also find that three additional developments since 2002 provide cable-affiliated programmers with an even greater economic incentive to withhold programming from competitive MVPDs: (i) the increase in horizontal consolidation in the cable industry; (ii) the increase in clustering of cable systems; and (iii) the recent emergence of new entrants in the video market place, such as telephone companies.”).
-94-
NAVIGANT ECONOMICS
5102853
rights to eight live (Thursday- and Saturday-night) games from the NFL. When it failed to obtain
what it sought, Comcast retaliated by refusing to carry MASN and the NFL Network on
Comcast’s digital tier. As noted in Table 4, Comcast ultimately settled these carriage disputes.
i. MASN
147. In 1981, the Baltimore Orioles were granted the television rights to the
Washington area by Major League Baseball (MLB). When the Nationals (formerly the Expos)
moved to Washington from Montreal for the 2005 baseball season, the Orioles agreed to share
the territory and dropped their opposition to the franchise’s relocation in return for control of
local television broadcasts of the National’s games. Comcast Executive Vice President David L.
Cohen described the arrangement between the Orioles and MLB as “original sin.”325 The Orioles
launched a new RSN called Mid-Atlantic Sports Network (MASN) to carry both the Orioles
games (formerly carried on Comcast SportsNet Mid-Atlantic) and the Nationals games. Comcast
refused to carry MASN in the Washington area. In addition, Comcast filed a complaint against
MASN and the Orioles in Maryland Circuit Court to maintain the rights to Orioles games on
Comcast SportsNet Mid-Atlantic. (Comcast’s complaint was dismissed twice.) Asserting that
Comcast’s refusal to carry MASN was based on discrimination in favor of its own RSN, in June
2005, MASN filed a program-carriage complaint with the Commission.
148. When approving the acquisition of Adelphia by Comcast and Time Warner in
2006, the Commission confirmed that dominant MVPDs still had the ability to jeopardize
competition in upstream programming markets, especially with regard to the foreclosure of
unaffiliated RSNs. In particular, the Commission found that the (shared) acquisition of Adelphia
325. Arshad Mohammed & Thomas Heath, Comcast Will Air Nats Games, Ending Dispute, WASHINGTON
POST, Aug. 5, 2006, at A1.
-95-
NAVIGANT ECONOMICS
5102853
would increase Comcast’s incentive and ability to discriminate against unaffiliated RSNs.326 To
protect against this type of discrimination, the Commission imposed non-discrimination
remedies in the form of commercial arbitration on Comcast (and Time Warner) that prevented
Comcast from discriminating against independent RSNs.327
149. Later that month, the FCC found that MASN had made a “prima facie showing”
that Comcast had discriminated against the network, and that Comcast had “indirectly and
improperly demanded a financial interest” in the network in exchange for carrying it.328 The FCC
gave MASN the right to send the dispute to an arbitrator or an administrative judge.329
Recognizing that it stood little chance to prevail in that setting, in August 2006, Comcast
withdrew its case in the Maryland Circuit Court, and it agreed to carry the Nationals broadcasts
for some 1.6 million Comcast customers in the Washington area.330 As part of the agreement,
MASN withdrew its carriage complaint at the FCC. One week later, Comcast announced that it
would raise cable television rates by $2 a month for its 1.6 million Washington-Baltimore area
customers because of the cost of carrying MASN.331 MASN added the Orioles games to its
network for the 2007 season. In 2009, MASN brought another carriage complaint to the FCC,
326. Adelphia Order ¶¶ 114, 116, 189. 327. Id. (“We find that this strategy would be made less likely by the arbitration and program access
conditions that we adopt but recognize that Comcast and Time Warner nevertheless may be more likely to succeed in foreclosing an unaffiliated RSN as a result of the transactions. As a result, consumers could be unable to view the RSN’s programming or could have to pay higher costs for the programming. Accordingly, to prevent such behavior, we adopt a further condition requiring Comcast and Time Warner to engage in commercial arbitration with any unaffiliated RSN that is unable to reach a carriage agreement with either firm, should the RSN elect to use the arbitration remedy.”).
328. Arshad Mohammed, FCC Finds Possible Bias Against MASN by Comcast, WASHINGTON POST, Aug. 1, 2006, at D1.
329. Arshad Mohammed & Thomas Heath, Comcast Will Air Nats Games, Ending Dispute, WASHINGTON
POST, Aug. 5, 2006, at A1. 330. Late Innings; For Nats' fans, a breakthrough: baseball on TV! , WASHINGTON POST, Aug. 9, 2006, at
A16. 331. Arshad Mohammed, Comcast To Raise Rates for Nationals; Subscribers to Pay $2 a Month More,
WASHINGTON POST, Aug. 12, 2006, at D1.
-96-
NAVIGANT ECONOMICS
5102853
alleging that Comcast discriminated against it by not carrying MASN in the Harrisburg and Tri-
Cities DMAs. The dispute was also settled.
ii. NFL Network
150. In 2006, the NFL elected to televise eight live NFL games per year on its
network, NFL Network. In doing so, it declined the opportunity to assign those games to a rival
sports network, including Comcast-affiliated Versus. In an apparent reaction to its failure to
secure exclusive distribution rights for those games, Comcast moved the NFL Network from a
digital tier to its Sports Entertainment Package shortly thereafter. The NFL Network initiated a
carriage complaint before the Commission; separate contract litigation was initiated in a New
York state court. In August 2009, the NFL Network and Comcast reached a settlement.
Subsequent to that agreement, Comcast carried the NFL Network on Comcast’s more popular
Digital Classic or Digital Starter tiers,332 increasing the NFL Network’s subscribers on Comcast
from two to eleven million.333 In sum, given Comcast’s prior discriminatory conduct vis-à-vis
The Big Ten Network, MASN, and the NFL Network, and given the pressures applied by cable
operators to the MLB Network, NBA TV, and the NHL Network to exchange equity for broader
carriage, it is reasonable to infer that Comcast’s discrimination here is motivated by a desire to
extend its sports programming footprint into football.
iii. Big Ten Network
151. Comcast’s dealings with the Big Ten Network, another unaffiliated sports
network, reveal the same exclusionary pattern. In June 2008, a spokesperson for the Big Ten
Network, an unaffiliated regional sports network, told the Philadelphia Inquirer that “Comcast
wouldn’t sign a deal because the Philadelphia company [that is, Comcast] didn’t own at least
332. Brian Mackey, Comcast settles dispute with NFL Network, News-Leader, May 20, 2009. 333. Joe Flint, DirecTV, Comcast fight over Versus distribution, LOS ANGELES TIMES, Sept. 2, 2009.
-97-
NAVIGANT ECONOMICS
5102853
part of the new network, and it was treating the new network differently than Comcast’s own
sports networks, Versus and the Golf Channel, which have limited audiences and low ratings.”334
b. Successful Attempts to Extract Equity in Exchange for Carriage
152. I briefly review the pressures placed on unaffiliated sports networks to assign
equity to vertically integrated cable operators in exchange for broader carriage. My review is not
meant to be exhaustive; for example, although Comcast disclosed its direct ownership interest335
in the NHL Network around the time336 that it moved the network from the Sports Tier to a more
broadly penetrated digital tier, the NHL anecdote is not reviewed here.
i. NBA TV
153. Comcast changed its tiering decision vis-à-vis NBA TV following a deal between
the NBA and Turner, which gave Turner, a division of Time Warner, a share of NBA TV’s
profits. Comcast has an incentive to carry Time Warner’s affiliated programming broadly to the
extent that Time Warner would reciprocate by carrying Comcast’s affiliated programming
broadly.337 In 1999, the NBA launched NBA TV, the league’s television network.338 As of the
2006-07 basketball season, NBA TV had only 12 million subscribers, as most cable operators,
including Comcast and Time Warner, carried the network on a sports tier.339 In November 2006,
Reuters reported that “Time Warner owns a 2 percent stake in NBA TV and [NBA
334. Bob Fernandez, Comcast, Big Ten reach pay-TV deal, PHILADELPHIA INQUIRER, June 20, 2008 (emphasis
added). 335. Comcast SEC Form 8-K, filed 12/04/09 for the Period Ending 12/03/09, at 6 (showing ownership of 15.6
percent of NHL Network). 336. The disclosure of Comcast’s direct ownership was made in December 2009. Id. The NHL Network was
re-tiered in the Washington, D.C. area around July 30, 2009. See Important News for Comcast Customers, June 2009.
337. Vertically integrated cable operators have been recognized to enter into reciprocal carriage agreements. See Jun-Seok Kang, Reciprocal Carriage of Vertically Integrated Cable Networks, Indiana University Working Paper (Aug. 30, 2005) at i (“The research supports the reciprocal carriage hypothesis by finding that: (1) A vertically integrated MSO is more likely than a non-vertically integrated MSO to carry the start-up basic cable networks of other MSOs; and, (2) a vertically integrated MSO is no more likely than a non-vertically integrated MSO to carry independent start-up basic cable networks.”).
338. Barry Jackson, MLB Network Read to Launch, MIAMI HERALD, July 18, 2008. 339. NBA sees bigger TV deal with partners, REUTERS, Nov. 28, 2006.
-98-
NAVIGANT ECONOMICS
5102853
Commissioner David] Stern said the cable company and the league are discussing an increase in
that stake.”340 In August 2007, Variety reported that NBA TV was “relegated to digital sports
tiers ... If Time Warner, which owns cable systems reaching more than 14 million subscribers,
agreed to shift NBA TV from sports tiers to digital basic, the network would add millions of new
customers.”341 In January 2008, while in the midst of renewal discussions with existing
distributors, the NBA made a deal with Turner Sports, an affiliate of Time Warner Cable, that
passed operations of NBA Digital, including NBA TV and NBA.com, to Turner Sports in an
effort to revamp the entities’ marketing and programming.342 The deal granted Turner an
undisclosed share of the profits from the NBA TV and the NBA’s other digital services.343 The
NBA, which handled distribution negotiations for NBA TV, also planned to offer equity in the
network to other MVPDs to secure adequate carriage.344 Following these transactions, the NBA
announced that it had secured—at great cost—distribution on Time Warner’s digital basic tier.345
In June 2009, NBA TV secured distribution to Comcast’s 11 million “Digital Classic”
subscribers, up from the 2 million Comcast subscribers it reached before the NBA deal with
Turner Sports.346 In the span of one year, in which the NBA surrendered equity in its network to
a vertically integrated cable operator (Time Warner), NBA TV’s distribution skyrocketed from
12 million subscribers in the 2008-09 season to 45 million subscribers in the 2009-10 season.347
340. Id. 341. John Dempsey, NBA TV may bounce to Time Warner Media conglom could take over hoops network,
VARIETY, Aug. 21, 2007. 342. NBA.com Press Release, Turner Broadcasting and NBA Broaden Partnership with Digital Rights
Agreement (Jan. 17, 2008) [hereinafter NBA Press Release]. 343. Daniel Frankel, NBA TV makes novel partnership; Turner Sports in charge of day-to-day programming.
DAILY VARIETY, Oct. 24, 2008. 344. Jon Lafayette, NBA Drives to Hole; New Programs, Marketing Set for Network, TELEVISION WEEK, Oct.
6, 2008. 345. NBA TV reaches 45 million U.S. homes with carriage deals, REUTERS, Oct. 22, 2009. 346. Deborah Yao, Comcast Reaches Deal to Add NBA TV to Popular Digital TV Tier, ASSOCIATED PRESS,
Jun. 3, 2009; Bob Fernandez, Comcast to Put NBA TV on ‘Digital Classic’ Tier, PHILADELPHIA INQUIRER, Jun. 3, 2009.
347. NBA TV reaches 45 million U.S. homes with carriage deals, REUTERS, Oct. 22, 2009.
-99-
NAVIGANT ECONOMICS
5102853
ii. MLB Network
154. Major League Baseball’s television network, MLB Network, debuted in January
2009 with 50 million subscribers. The network “got the distribution [it] needed,” according to
president and CEO Tony Petitti.348 MLB “learned from mistakes made in launching other single-
sport networks” and the resulting distribution difficulties those networks faced.349 The network
“avoided such distribution problems”350 by partnering with DirecTV and three leading cable
companies, Time Warner, Comcast, and Cox. According to one sports analyst, “the reason that
the MLBN has been able to enjoy a compatible arrangement with cable broadcasters is that it
gave up a share of its equity in order to reach that goal.”351 DirecTV, Time Warner, Comcast,
and Cox together acquired a third of the MLB Network. Unsurprisingly, each offers the network
as part of its digital basic package.352
3. Tying Affiliated Internet Content to the Purchase of Cable Television Service
155. In this section, I briefly review Comcast’s exclusionary conduct regarding online
properties. Although the conduct relating to the Vancouver Olympics (reviewed below) was
ostensibly committed by NBCU, it provides a harbinger of exclusionary strategies available to
Comcast.
a. Fancast
156. Comcast has made some of the video content it distributes available online, but
only to paying customers of its cable television services through its TV Everywhere model,
348. Bill Doyle, Tardy MLB Finally Debuts; New Network Largest Launch, WORCESTER TELEGRAM &
GAZETTE, Jan. 9, 2009 [hereinafter Tardy MLB Finally Debuts]. 349. Tardy MLB Finally Debuts, supra. 350. Id. 351. Dianne M. Grasse, MLB Network Rolls Out with Bait and Switch, SPORTS CENTER, Jan. 9, 2009,
available at http://www.sports-central.org/sports/2009/01/09/mlb_network_rolls_out (emphasis added). 352. Tardy MLB Finally Debuts, supra.
-100-
NAVIGANT ECONOMICS
5102853
which is branded in Comcast markets as “Fancast.”353 Fancast provides access to network
television shows and movies integrated with television-related news and a viewing guide for its
video service.354 During 2009, Fancast has had between two and five million unique visitors per
month, or approximately 10 percent of the size of Hulu’s audience.355 As of March 2010, Fancast
offered nearly 20,000 television shows and movies from two dozen channels, including HBO,
Discovery, AMC, FX, and NBC.356 Subscribers can use the service on three personal computers,
but not on their televisions.357
157. The proper lens to view this conduct is a tie-in, with Comcast’s cable television
service serving as the tying product and the online content serving as the tied product. By tying
access to its online content portfolio to the purchase of Comcast cable television service,
Comcast may impair the ability of in-region MVPD rivals, including OTT providers, to compete
effectively; by tying access to its online content portfolio to the purchase of Comcast cable
modem service, Comcast may impair the ability of in-region rival ISPs to compete effectively. In
particular, Comcast and other members of TV Everywhere can induce substitution away from
fiber/DSL connections offered by telephone companies.
158. BusinessWeek explained how Time Warner and Comcast coordinated in the
development of the TV Everywhere model. As Time Warner CEO Jeffrey Bewkes “watched one
entertainment company after another put their TV shows and movies online for free . . . [he]
began to fear that the pay TV industry would eventually find itself in the same untenable position
353. Fancast provides access to network shows and movies integrated with television-related news and a
viewing guide for their video service. Information about Comcast’s Fancast is available at http://www.fancast.com. 354. http://www.fancast.com/ 355. Site statistics for Fancast available at http://siteanalytics.compete.com/fancast.com/?metric=uv. 356. TV Everywhere, BUSINESSWEEK, Mar. 10, 2010, available at
as newspapers.”358 To exert the maximum pressure on content providers, Time Warner needed an
MVPD partner; Bewkes reportedly floated his proposal to Brian Roberts, CEO of Comcast, in
early 2009.359 Bewkes prevailed on Roberts to allow customers to “tap into any cable channel’s
Web site as long as it was part of the TV Everywhere ecosystem.”360
159. Collectively, Comcast and Time Warner possess significant buying power vis-à-
vis independent cable networks. BusinessWeek explained that the cable operators were “using the
$32 billion they pay content providers each year as leverage” to ensure that cable networks do
not grant OTT providers such as Boxee and Roku access to their content.361 Not all content
providers are willing to go along with Comcast’s design; in April 2009, Disney CEO Bob Iger
explained that “preventing people from watching any shows online unless they subscribe to some
multichannel service could be viewed as both anti-consumer and anti-technology, and would be
something we would find difficult to embrace.”362
b. 2010 Olympics
160. NBCU held the rights to the 2010 Winter Olympics in Vancouver, and it owns the
rights to the 2012 Summer Games in London.363 According to the New York Times, NBC’s
online Olympic portal, NBCOlympics.com, streamed fewer sports in the Vancouver games in
real-time than it did in the Beijing games, “marking a step backward in online access to marquee
358. Ronald Grover, Tom Lowry & Cliff Edwards, Revenge of the Cable Guys, BUSINESSWEEK, Mar. 11,
2010, available at http://www.businessweek.com/magazine/content/10_12/b4171038593210.htm?chan=magazine+channel_top+stories.
359. Id. 360. Id. 361. Id. 362. Chris Albrecht, Get ready to prove you paid for that video, BUSINESSWEEK, Apr. 24, 2009, available at
http://www.businessweek.com/technology/content/apr2009/tc20090424_766570.htm. 363. Cecilia Kang, Merger plans for Comcast, NBC ignite battle over television access, WASHINGTON POST,
Dec. 4, 2009, at A1.
-102-
NAVIGANT ECONOMICS
5102853
events.”364 NBC’s online portal aired only hockey and curling in real-time; every other event on
the Internet was shown on delay, sometimes of more than a day.365 Moreover,
NBCOlympics.com streamed Olympic events on delay provided that Internet users could verify
a subscription to participating cable or satellite providers.366
161. The reduction in real-time, online sports programming was significant. NBC
provided 2,200 hours of live streaming for the Beijing Olympics (61 percent of its 3,600 total
hours of event coverage); in contrast, NBC provided only 400 hours of live streaming for the
Vancouver Olympics (48 percent of its 830 total hours of coverage).367 According to Gary
Zenkel, the president of NBC Olympics, NBC’s decision turned on the relative importance of
television revenues to Internet revenues: “When we roll out our digital coverage, there are some
financial considerations to take into account. The lions’ share of advertising revenue continues to
be generated by our television coverage.”368 It bears noting that, at the time of this decision,
NBCU had agreed to be purchased by Comcast. Accordingly, as in the case of Hulu’s decision to
block access to Boxee, the impact on Comcast’s profits may have played a role in NBC’s
decision. Given NBC’s bundled sale of television and Internet advertisements for the
Olympics,369 it is not clear how NBC suffered a revenue loss if viewers watched some live
events via the web portal, as NBC’s allocation of the bundled price across its television and
Internet divisions was inherently arbitrary.
162. NBC owns the rights to the 2012 Summer Games in London. In 2010, the
International Olympic Committee will sell the rights to the 2014 Winter Games in Sochi, Russia,
364. Brian Stelter, A trickle of life streams on the web, NEW YORK TIMES, Feb. 18, 2010, at 15. 365. Id. 366. Id. 367. Id. 368. Id. 369. Id.
-103-
NAVIGANT ECONOMICS
5102853
and the 2016 Summer Games in Rio de Janeiro.370 In April 2010, the United States Olympic
Committee and Comcast ended their plans to create an Olympic channel.371 Richard Carrion, an
executive board member of the International Olympic Committee, told the New York Times that
Comcast’s acquisition of majority control of NBCU “probably played a part” in the end of the
U.S. Olympic Committee’s network.372
III. THE PROPOSED TRANSACTION WOULD INCREASE COMCAST’S ABILITY AND INCENTIVE TO
ENGAGE IN THESE ANTICOMPETITIVE STRATEGIES
163. NBCU’s portfolio of programming is vast and impressive. In this section, I
identify NBCU’s marquee programming that Comcast would acquire, and I explain why the
acquisition of that programming would increase Comcast’s incentive and ability to engage in
exclusionary conduct vis-à-vis independent networks, downstream MVPD and broadband access
rivals.
A. NBC’s Marquee Network Programming
164. NBCU’s marquee network programming consists of the ten O&O local NBC
affiliates and national sports programming.
1. Local NBC Affiliates
165. NBCU owns local NBC affiliates in ten major DMAs: Philadelphia, PA; Chicago,
IL; Miami-Fort Lauderdale, FL; San Francisco-Oakland-San Jose CA; Washington, DC
(Hagerstown, MD); Hartford & New Haven, CT; New York, NY; Dallas-Fort Worth, TX; Los
Angeles, CA; San Diego, CA. Collectively, these DMAs cover 30.6 million television
370. Richard Sandomir, ESPN sets sights on different way to show Olympics, NEW YORK TIMES, Feb. 25,
2010, at 13. 371. Richard Sandomir, U.S.O.C. Ends Plans for Its Own Olympic Channel, NEW YORK TIMES , Apr. 22,
2010, at 13. 372. Id.
-104-
NAVIGANT ECONOMICS
5102853
households or 26.7 percent of all U.S. television households.373 In addition to covering local
news, these affiliates carry the most popular network programming from 2002 to 2009, including
Friends (2001-02, 24.5 million total viewers), ER (2000-01, 22.4 million total viewers), The
Apprentice (2003-04, 20.7 million total viewers), Law & Order (2001-02, 18.7 million total
viewers), Will & Grace (2000-01, 17.3 million total viewers), The West Wing (2001-02, 17.2
million total viewers), Leap of Faith (2001-02, 16.5 million total viewers), The Apprentice 2
(2004-05, NBC 16.1 million total viewers), Frasier (2000-01, 15.9 million total viewers), Scrubs
(2002-03, 15.9 million total viewers), Just Shoot Me (2000-01, 15.6 million total viewers), and
Deal or No Deal (2005-06, 15.6 million total viewers).374 In the 2009-10 season, NBC’s Sunday
Night Football (described below) was the second most popular network show (19.6 million total
viewers).375
2. National Sports Programming
166. NBC Universal Sports & Olympics is home to many of today’s top sporting
events, including the U.S. Open Championship, The Ryder Cup, Presidents Cup, Kentucky
Derby, Preakness Stakes, Notre Dame football, Wimbledon, French Open, Stanley Cup Finals,
Summer and Winter Olympic Games through 2012, and NBC Sunday Night Football. Table 5
lists the top sports assets owned by NBCU, and includes contract length and terms where
available.
373. U.S. TV Household Estimates Designated Market Area (DMA) — Ranked by Households, available at
http://www.tvb.org/rcentral/markettrack/us_hh_by_dma.asp. 374. Decade’s Top 125 Most-Watched TV Shows (2000-2009), available at
http://blog.newsok.com/television/2010/01/12/decades-top-125-most-watched-tv-shows-2000-2009/. 375. Top 20 Most-Watched TV Programs in 2009-10 (through Dec. 20), available at
TABLE 5: MARQUEE NATIONAL SPORTS PROGRAMMING OWNED BY NBCU
Sports asset Length of contract Financial terms U.S. Open Championship 2008-20141 Financial terms of renewal unknown. NBC
acquired the rights to the U.S. Open from ABC in 1995 with a three-year $40 million contract.2
The Ryder Cup 2008-20143 Financial terms unknown. Presidents Cup Through 20114 Financial terms unknown. Kentucky Derby & Preakness Stakes
2006-20105 $43 million contract. NBC is in negotiations to renew its contract for the Kentucky Derby and the Preakness Stakes and hopes to add the Belmont Stakes to its lineup in 2011.5
Notre Dame Football 2010-20156 Not disclosed, but previous contract, which expires in 2010, is reported to be worth $9 million per year.6
Wimbledon Through 20117 “[E]stimated to be worth $12 million to $13 million per year through 2011.” 7
French Open 2006-20133 Financial terms unknown. NHL, Stanley Cup Finals - 2011 Stanley Cup Finals8 The extension is a continuation of their revenue-
sharing partnership. 8 Summer and Winter Olympics Through 20129 $2.2 billion (for 2010 and 2012 games). 9 NBC Sunday Night Football Through 201310 Financial terms of renewal unknown. The original
six-year deal, which was signed in the spring of 2005, was worth $3.6 billion.11
U.S. Figure Skating Through 2014 Winter Games12 Financial terms unknown.
Sources: 1 Press Release, USGA, NBC Extend Broadcast Contract Through 2014, USGA (Jan. 24, 2005); 2 Milton Kent, Enberg, NBC are up to par in U.S. Open, BALTIMORE SUN (June 19, 1995); 3 Attachment 53-1 on NBCU production; 4 Press Release, PGA TOUR reaches television agreements, PGA (Jan. 11, 2006); 5 NBC To Keep Kentucky Derby, Preakness Stakes Broadcast Rights, BLOODHORSE.COM (May 4, 2005); Matt Hegarty, Derby handle, ratings post gains, ESPN.COM (May 3, 2010); 6 Notre Dame agrees to five-year extension with NBC, ASSOCIATED PRESS (June 19, 2008); 7 Chris Pursell, NBC, ESPN Close in on Wimbledon Renewals, TV WEEK (2007); 8 NHL, NBC Extend Revenue-Sharing Agreement Through ’10-11 Season, SPORTS
BUSINESS DAILY (July 16, 2009); 9 NBC lands U.S. Olympic broadcast rights through 2012, CBC SPORTS (Jun. 7, 2003); 10 NFL & NBC Announce Two-Year Contract Extension, NBC.COM (Aug. 19, 2009); Joanna Weiss, NBC: Sunday night’s all right for football, BOSTON GLOBE (July 24, 2006); 11 Joanna Weiss, NBC: Sunday night’s all right for football, BOSTON GLOBE (July 24, 2006); 12 U.S. Figure Skating, NBC Sports Announce Broadcast Contract Extension, SPORTS MEDIA NEWS (Aug. 27, 2009).
B. NBC’s Marquee Online Programming
167. NBCU owns or partially owns several marquee online properties. Hulu, a joint
venture of NBC Universal (GE), Fox (News Corp), and ABC Networks (Disney) launched in
early 2008, offers video content from those programmers as well as others for free typically on a
delayed basis (that is, some number of days after the programming is broadcast). According to
comScore, in the month of March 2009, 380 million videos were viewed on Hulu.com by over
41 million unique viewers.376 By February of 2010 the number of videos watched was up to 912
376. Hulu Continues Ascent in U.S. Online Video Market, Breaking Into Top 3 Properties by Videos Viewed
for First Time in March, comScore press release April 28, 2009, available at http://www.comscore.com/Press_Events/Press_Releases/2009/4/Hulu_Breaks_Into_Top_3_Video_Properties.
-106-
NAVIGANT ECONOMICS
5102853
million.377 Among the most popular shows watched on Hulu.com on April 16, 2010 were ABC’s
Modern Family, Fox’s Glee, and Fox’s Family Guy. In March of 2010, Viacom removed some
of its very popular Comedy Central shows from Hulu, including The Daily Show and The
Colbert Report.378 According to Laura Martin, an analyst at Needham & Company, “the content
companies continue to experiment with how to monetize their premium content on digital
platforms. Hulu.com is a failed experiment for Viacom.”379 In addition, NBCU owns NBC.com,
MSNBC.com, CNBC.com, and iVillage.com.
C. Ownership of These “Must-Have” Inputs Would Facilitate Comcast’s Anticompetitive Schemes vis-à-vis Independent Networks and MVPD Rivals
168. If the proposed transaction were approved, Comcast would acquire significant
must-have content that could be withheld from downstream MVPD rivals at both the local level
(with NBCU O&O networks) and at the national level (with NBCU national programming and
online programming); it would also increase Comcast’s market power vis-à-vis independent
networks, OTT video providers, and broadband access providers.
169. The proper economic lens through which to analyze Comcast’s refusing to supply
affiliated cable network programming (from NBCU) to downstream MVPD rivals is exclusive
dealing. Not all exclusive dealing is anticompetitive. To reduce welfare, the exclusive
arrangement must impair the ability of downstream rivals’ to compete effectively in the MVPD
market and there must be no redeeming efficiencies.380 This outcome can be achieved by either
377. comScore Releases February 2010 U.S. Online Video Rankings, comScore press release April 13, 2010,
availabel at http://www.comscore.com/Press_Events/Press_Releases/2010/4/comScore_Releases_February_2010_U.S._Online_Video_Rankings.
378. Brian Stelter, Viacom and Hulu Part Ways, NEW YORK TIMES, March 2, 2010, available at http://www.nytimes.com/2010/03/03/business/media/03hulu.html.
379. Id. 380. Patrick Rey & Jean Tirole, A Primer on Foreclosure (reprinted in III HANDBOOK OF INDUSTRIAL
ORGANIZATION, Mark Armstrong & Rob Porter, eds., 2145-2220, 2007), at *1 (“An input produced by a dominant firm is essential if it cannot be cheaply duplicated by users who are denied access to it.”).
-107-
NAVIGANT ECONOMICS
5102853
(a) raising rivals’ costs or (b) reducing rivals’ quality.381 Both conditions are satisfied here.
Comcast can directly raise a rival’s costs by increasing the wholesale price of Comcast’s
affiliated networks. And to the extent the withheld network constitutes “must-have”
programming, Comcast can also reduce the quality of the rival MVPD’s offering, thereby
weakening its ability to constrain Comcast’s price for MVPD service.
170. The proper economic lens through which to analyze Comcast’s requiring
verification of a subscription to both its cable television and cable modem service to access its
online content portfolio (Fancast) is tying. Not all tie-ins are anticompetitive. The Chicago
School one-monopoly-profit critique showed that if (1) the tying and tied products are consumed
in fixed proportions, (2) the tied product has no separate utility for consumers, and (3) if
competition in the tied and tying markets are fixed, then a tie-in is likely motivated for
procompetitive reasons. However, if the products are not used or bundled in a fixed ratio, then a
tie-in can be used to create intra-product price discrimination or to create interproduct price
discrimination.382 If the products are used or bundled in a fixed ratio, but if the tied product also
has separate utility when not used with the tying product, and if competition in the tied market is
not fixed (due to, for example, economies of scale in the supply of the tied product), then tying
can diminish rival competitiveness due to significant foreclosure in the tied market.383
Alternatively, even if the products are used or bundled in fixed ratios and lack separate utility,
foreclosing the tied market might still create anticompetitive effects if it alters the degree of tying
381. Thomas G. Krattenmaker & Steven C. Salop, Anticompetitive Exclusion: Raising Rivals' Costs to
Achieve Power over Price, 96 YALE L. J. 234 (1986) (“The simplest and most obvious method by which foreclosure of supply can raise rivals’ costs is that purchaser's obtaining exclusionary rights from all (or a sufficient number of) the lowest-cost suppliers, where those suppliers determine the input's market price. Competitors of the purchaser experience a cost increase as they necessarily shift to higher cost suppliers or less efficient inputs. Antitrust literati know this as the ‘Bottleneck’ or ‘essential facilities’ problem.”).
382. Einer Elhauge, Tying, Bundled Discounts, and the Death of the Single Monopoly Profit Theory, 123 HARVARD LAW REVIEW 399, 404-05(2009).
383. Id. at 416.
-108-
NAVIGANT ECONOMICS
5102853
market power (due to, for example, the tied market competitor evolving into a rival in the tying
market).384 Once again, the necessary conditions for Comcast’s tie-in to be anticompetitive are
satisfied here. With regard to the first theory, consumers clearly do not consume online video
and MVPD service in fixed proportions; some consumers will consume ten hours of online video
per month (alongside a single MVPD subscription) while others will consumers 100 hours per
month. With regard to the second theory, customers have a separate utility for online content (the
tied product) when not used with MVPD services (the tying product); and there are fixed costs to
supplying online video and large economies of scale, as the marginal cost of producing the last
download is close to zero. With regard to the third theory, competition in the supply of MVPD
services is not fixed. In particular, OTT video providers could evolve into significant MVPD
suppliers, and thereby erode Comcast’s tying market power. With respect to the requirement that
Comcast have significant tying market power, Comcast’s share of nationwide MVPD subscribers
alone constitutes significant foreclosure. Because Comcast has partnered with Time Warner and
other MVPD suppliers, however, the resulting foreclosure share faced by OTT video providers is
even larger. Even if Comcast perceives online video to be a complement to its cable television
service today, as its economists unconvincingly argue, to the extent that a vibrant, online-video
offering could one day induce Comcast’s subscribers to disconnect their cable television
subscriptions, Comcast’s tie-in of its affiliated online content portfolio to its cable television
service could prevent that evolution from occurring.
384. Id. at 417.
-109-
NAVIGANT ECONOMICS
5102853
1. Refusing to Supply Affiliated Programming to Rival Distributors on Reasonable Terms
171. A standard foreclosure calculus reveals that Comcast would likely withhold (a)
local NBC affiliates from MVPD rivals in markets where Comcast has significant downstream
market power (shares in excess of 40 percent) and (b) the new-and-improved Versus from
MVPD rivals nationwide.
a. NBC Local Affiliates
172. Comcast is acquiring local must-have programming via NBCU’s ten owned-and-
operated (O&O) NBC affiliates. These properties will also increase Comcast’s market power vis-
à-vis MVPD rivals, but only in the DMAs in which NBCU owns a local NBC affiliate. In the
seven local markets in which NBCU owns a broadcast affiliate and Comcast owns an RSN (the
“overlapping markets”)—Chicago, Philadelphia, Dallas-Ft. Worth, San Francisco-Oakland-San
Jose, Washington, DC (Hagerstown), Miami-Ft. Lauderdale, and Hartford-New Haven—
Comcast will realize a substantial increase in market power vis-à-vis its MVPD rivals. In
particular, Comcast can now target customers of rival MVPDs who were unwilling to switch on
account of access to local sports but would switch on account of access to NBC. Comcast serves
approximately 70 percent of MVPD subscribers in the Philadelphia DMA, and it serves
approximately 60 percent of MVPD subscribers in the Chicago, Miami, and San Francisco
DMAs. With such significant downstream market shares, Comcast can count on recouping any
upstream losses associated with the withholding of the NBC affiliate with ill-gotten downstream
gains. And given the fact that Comcast is often the largest provider of local advertising in these
local markets, the acquisition of an NBC affiliate would increase Comcast’s ability to raise local
advertising rates.
-110-
NAVIGANT ECONOMICS
5102853
173. Rather than presenting an original foreclosure calculus here, I generally embrace
the calculus used by Comcast’s economists with the caveat that several parameter values must be
revised to reflect realistic assumptions. My critique of their parameters is presented below. Even
without making any adjustments to their parameters, the foreclosure calculus reveals that, in
local markets in which Comcast has a significant market share (defined as 40 percent or greater),
a very small amount of customer departure to Comcast (to follow the withheld NBC content)
would be needed to render the foreclosure profitable—that is, the “critical departure share” is
very small. In particular, according to Comcast’s own economists, the low-end of the critical
departure shares under the “permanent foreclosure” simulation for six DMAs (Chicago, Hartford
& New Haven, Miami-Ft. Lauderdale, Philadelphia, San Francisco-Oakland-San Jose, and
Washington-Hagerstown) ranges from {{6.0}} (Philadelphia) to {{10.5}} percent (Hartford &
New Haven).385 A reasonable proxy for the actual departure share is the loss in DBS share in the
Philadelphia DMA for failing to secure Comcast SportsNet Philadelphia. In this event, the actual
departure share would likely exceed the critical level—several economists have estimated the
share shift in the Philadelphia DMA to be on the order of ten percentage points.386 Accordingly,
Comcast would find it profitable to withhold the local NBC affiliate in each of those six DMAs,
as the actual departure share would likely exceed the critical departure share.
174. It bears noting, however, that the foreclosure calculus used by Drs. Katz and
Israel does not consider the possibility that Comcast raise its wholesale price. Relative to a
standalone NBC broadcast affiliate, Comcast’s ownership reduces the cost to the affiliate of
385. Israel & Katz NBCU, supra, 46 (Table 2). 386. Lexecon October 2005 Analysis, supra (predicting that, given its market characteristics, DBS penetration
in Philadelphia should have been approximately 21 percent in 2005); Willig & Orzag, supra (finding that Dish Network’s penetration rate in the Philadelphia DMA was reduced from 9.5 percent to 3 percent as a result of Comcast’s conduct); Bamberger and Neumann March 2006 Analysis, supra (estimating that the actual DBS penetration rate in Philadelphia is approximately 10 percentage points smaller than it should be based on Philadelphia’s characteristics); Singer & Sidak, supra (estimating that DBS penetration in the Philadelphia DMA should be 15.4 percent compared to the then-actual level of 9.4 percent).
-111-
NAVIGANT ECONOMICS
5102853
withholding programming from Comcast’s MVPD rivals. When faced with the prospect of not
watching NBC programming long-term, some MVPD customers in the relevant DMAs will
switch to Comcast; as a result, Comcast will earn incremental (downstream) profits. These
profits will offset the cost to Comcast of withholding the NBC affiliate. Recognizing these
tradeoffs, Comcast will be able to bargain for higher programming prices for programming
previously owned by NBC than NBC would otherwise be able to negotiate on its own. This is
true even if the loss in upstream programming profit from withholding the NBC affiliate exceeds
the gain in downstream MVPD profit due to customer switching. And MVPD consumers will be
harmed because these higher programming prices will be passed through to them in the form of
higher cable bills.
b. National Sports Programming (Versus)
175. By transferring some of NBCU’s national sporting events from NBC’s networks
to Versus, Comcast would form a national sports network that would rival ESPN. But unlike
ESPN, which is owned by Disney, Versus will be owned by the largest MVPD, serving roughly
one-quarter of all U.S. MVPD subscribers. Consideration of Comcast’s downstream profits
would severely distort the joint venture’s pricing incentives for Versus. In particular, Comcast
would be willing to incur losses (relative to the standalone profit-maximizing price for those
rights) in its upstream content division (Versus) in exchange for ill-gotten gains in its
downstream distribution division, even if some of those downstream “benefits” were to spill into
non-Comcast territories. And MVPD consumers will be harmed in either of two possible
contingencies: (a) if Comcast merely raises the price of Versus to rival MVPDs, these higher
prices will be passed through to MVPD subscribers in the form of higher cable bills; or (b) if
Comcast outright refuses to supply this programming to rival MVPDs—for example, by moving
-112-
NAVIGANT ECONOMICS
5102853
the future marquee Versus programming online to escape the program access rules—then non-
Comcast customers will be forced to incur switching costs to follow this programming (as they
switch back to Comcast) and higher cable bills due to the reduction in MVPD competition.
176. In addition to seeking control of NBCU’s national sports programming, Comcast
has recently acquired equity interests in several national sports networks, including MLB
Network, NHL Network, and NBA TV. (Comcast also owns the Golf Channel, a national sports
network.) Consistent with the discrimination hypothesis outlined here, Comcast re-tiered those
three national sports networks from its lowly penetrated sports tier to its highly penetrated
“Digital Classic Tier” shortly after it acquired equity in those networks. Indeed, this is precisely
how Comcast extracts equity from rights holders of sports programming; refusal to grant equity
lands an independent programmer on the sports tier, which severely limits advertising
opportunities. And the best way out of this predicament is to offer equity to Comcast. These
acquisitions of national sports networks alongside NBCU’s national sports-programming
portfolio will vastly increase Comcast’s market power vis-à-vis its MVPD rivals throughout
Comcast’s service area.
177. In the Internet Age, sports programming is critical to an MVPD because viewers
demand to see it in real-time, often on their big-screen televisions. This “must-have” nature of
sports programming is recognized by advertisers, which explains why such programming
commands high advertising rates. By acquiring NBCU’s national sports content, Comcast seeks
to exploit this unique opportunity to protect its cable television profits. In particular, Comcast
could withhold affiliated national sports programming from downstream rivals to impair MVPD
competition. Non-sports or non-event programming that can be watched on delay without any
significant diminution in viewer utility is relatively less valuable, as MVPD subscribers
-113-
NAVIGANT ECONOMICS
5102853
increasingly have alternative means for accessing that content after it originally airs, including
via the Internet.
178. To compute the “critical departure share” under which Comcast would find it
profitable to withhold Versus from downstream MVPD rivals, I use the Commission’s
foreclosure calculus from the News Corporation/DirecTV proceeding. For the purpose of this
modeling exercise, I assume that Comcast would withhold Versus from all MVPDs on a
permanent basis. Assuming that 100 percent of Comcast’s MVPD profits accrue to the integrated
firm, the critical departure share for a permanent foreclosure strategy is simply:
/ ,critd A A M (1.1)
where A is Comcast’s forgone licensing revenue per subscriber per month associated with
Versus, M is Comcast’s MVPD margins per newfound subscriber per month, and is the share
of those customers who depart the foreclosed MVPD who specifically choose Comcast as their
new MVPD.387 Assuming that is Comcast’s national MVPD market share of 23.4 percent,388 M
is Comcast’s monthly profit margin per MVPD subscriber (on video services only) of
{{$35.62}},389 and A is Versus’s projected 2010 average monthly license fee per subscriber of
$0.33,390 the critical departure share is 3.8 percent. If Versus were to carry all of NBC’s national
sports programming, then the network would command higher rates. Assuming A is $2.20, or
387. The incremental loss is (1 – d)A. The incremental gain is dMα. Setting the incremental loss equal to the
incremental gain and solving for d yields A / [A + Mα]. 388. Israel & Katz Online Content, supra, ¶ 107 (citing MediaBusiness Corporation). 389. Id. ¶ 104. This is a conservative assumption, as Comcast would likely capture the margins on cable
modem and voice services for some portion of switching customers. Indeed, Drs. Katz and Israel made this assumption when modeling the foreclosure calculus for the NBC affiliates. Israel & Katz NBCU, supra, ¶ 36.
390. Kagan, The Economics of Basic Cable Networks (2009), at 52.
-114-
NAVIGANT ECONOMICS
5102853
approximately half the license fee for ESPN in 2010,391 then the critical departure share is 20.8
percent.
179. It is difficult to measure with precision the likely departure rate here, as Comcast
has not withheld Versus from MVPD rivals to my knowledge. One proxy is the loss in an
MVPD’s national share for failing to secure a license agreement with ESPN. Another proxy is
the loss in DBS share in the Philadelphia DMA for failing to secure Comcast SportsNet
Philadelphia. To the extent that the latter is a reasonable proxy, the actual departure share would
likely exceed the critical level, especially if it were closer to the low-end of my estimate (3.8
percent). In any event, because Comcast’s per subscriber margins are so much larger than
Versus’s license fee, the foreclosure calculus suggests that the Commission should be concerned
about this prospect.
2. Denying Access to Independent Programming Networks
180. By expanding its broadcast and cable network holdings, Comcast increases its
market power vis-à-vis independent cable networks. Comcast is sending a chilling signal to
independent networks—in order to be carried on Comcast’s cable system, a network must
acquiesce to Comcast’s demand for equity. As the NFL Network, MASN, and Tennis Channel
program-carriage complaints against Comcast make clear, Comcast has discriminated against
independent sports networks (and in favor of its affiliated networks like Versus and CSN Mid-
Atlantic) in its tiering decisions. With the NBCU networks, Comcast would have a fresh
incentive to advantage its affiliated networks vis-à-vis independent, similarly-situated networks.
For example, Comcast could disadvantage an independent news network by relegating it to an
inferior tier relative to the tier on which it carries CNBC and MSNBC; on an inferior tier, the
independent news network would be impaired in its ability to compete for advertisers and for
391. Id. at 52 (projecting ESPN’s license fee to be $4.41 in 2010).
-115-
NAVIGANT ECONOMICS
5102853
programming rights. In the absence of the affiliation made possible by the proposed transaction,
no such incentive to discriminate in favor of NBCU cable networks exists. The same incentives
for discriminatory carriage would arise with respect to the new-and-improved Versus (for
independent sports networks), Telemundo (for independent Spanish-language networks), and
Oxygen (for independent women’s programming networks). Moreover, Comcast’s threat not to
carry independent networks that seek to post some of their content online becomes more credible
as Comcast expands its own online content portfolio; if an independent network refuses to place
its online content behind Comcast’s walled garden (and not on the network’s website), then
Comcast can simply deny carriage on its cable system. As the walled garden grows, the
incremental value of the next independent network to Comcast declines.
3. Tying Affiliated Internet Content to the Purchase of Cable Television Service
181. Comcast has addressed this increasing threat of Internet-based television
primarily in three ways. First, Comcast has made some of the video content it distributes
available online, but only to paying customers of both its video and Internet services through its
TV Everywhere service, which is branded in Comcast markets as “Fancast.”392 The proper lens
to view this conduct is a tie-in, with Comcast’s cable television service serving as the tying
product and the online content serving as the tied product. Second, Comcast requires that
unaffiliated cable networks not make their content available online as a condition of being
carried on Comcast’s cable television systems.393 The proper lens to view this conduct is an
exclusive deal; Comcast insists that a cable network not license its content to OTT video
392. Fancast provides access to network shows and movies integrated with television-related news and a
viewing guide for their video service. Information about Comcast’s Fancast is available at http://www.fancast.com. 393. Cable TV, supra, at *1 (“As the cable companies scramble to get these technologies out of the lab, they
are trying to prevent cable networks from putting their shows online. In recent months, Big Cable has reminded USA, Bravo, TNT (TNT), and hundreds of other channels that the cable companies provide about half their revenues, in the form of fees to carry their shows. The implicit warning: put your content online and forfeit nearly $30 billion. ‘If I don’t have a customer,’ says Time Warner Cable’s Britt, ‘the programmers aren’t going to have a customer.’”).
-116-
NAVIGANT ECONOMICS
5102853
providers or post content to its own website as a condition of carriage on Comcast’s cable
system. Third, Comcast maintains complete control of the set-top box used to access its cable
television service by requiring its cable television customers to purchase the set-top box from
Comcast to enjoy the full suite of interactive services. The proper lens to view this conduct is
another tie-in; in this case, the tied product is the set-top box.
182. These three, exclusionary strategies will slow the growth of Internet-only
television and thereby protect Comcast’s market power in the supply of cable television service.
Because the marginal (or imputed) price of accessing Comcast’s online portfolio is zero as a
result of the tie-in (and Comcast’s pricing structure), providers of OTT video seeking to compete
against Comcast will be forced to raise revenues from advertisers only, which could undermine
their business model. By preventing unaffiliated cable networks from posting their content online
as a condition of carriage, and by denying access to its affiliated online content, Comcast can
deny a critical input to OTT video providers, further impairing Internet-only television. When
Comcast’s affiliated online properties are combined with those of Time Warner, which also
participates in TV Everywhere, the resulting collection of withheld content could significantly
impair OTT providers. By denying its online portfolio to in-region rival Internet service
providers (“ISPs”)—that is, by tying Fancast to Comcast’s cable modem service—Comcast and
other members of TV Everywhere can induce substitution away from fiber/DSL connections
offered by telephone companies. And by controlling the set-top box, Comcast can ensure that its
video subscribers cannot access the Internet from the comfort of their televisions. An
independent set-top-box maker would be more inclined to add features, including the ability to
connect to and download video from the Internet via a Wi-Fi connection, that could threaten a
cable provider’s cable television revenues.
-117-
NAVIGANT ECONOMICS
5102853
183. Comcast’s acquisition of NBCU’s online content portfolio would facilitate these
exclusionary strategies, both on a unilateral basis and on a coordinated basis with other cable
operators. Denying access to NBCU’s online content portfolio, especially Hulu.com, would
significantly impair the ability of OTT providers to compete effectively for online video
customers. And to the extent that TV Everywhere does not provide access to rival ISPs on
reasonable terms, denying access to NBCU’s online content portfolio would also impair the
ability of in-region ISP rivals to compete effectively for broadband subscribers. Although it may
be possible to achieve these outcomes through (exclusive) contracting, having these must-have
assets under ownership greatly facilitates the exclusionary objective, as the possibility of
defection by the affiliated, online content provider (Hulu.com and others) decreases.394 When
done in conjunction with Time Warner and other cable operators that join TV Everywhere, the
foreclosure share associated with this exclusionary strategy increases, which in turn would
increase Comcast’s market power in the supply of online video. Independent cable networks and
movie studios naturally fear upsetting Comcast by posting their video content online or making it
available to OTT video providers; by including other cable operators, including Time Warner, in
TV Everywhere, the combined leverage vis-à-vis independent networks is even stronger.395
Having NBCU’s content portfolio also would increase Comcast’s market power vis-à-vis
independent programmers. In particular, Comcast’s threat not to carry independent networks that
seek to post some of their content online becomes more credible as Comcast adds to its online
video content portfolio, as Comcast’s online platform (Fancast) would become the most valuable
394. It is no accident that the initial trial of TV Everywhere in the summer of 2009 allowed viewers to tap into
programming provided by TNT and TBS, both of which are owned by Time Warner. See Revenge of the Cable Guys, supra.
395. Revenge of the Cable Guys, supra (“Back at Time Warner Center in New York and One Comcast Center in Philadelphia, the cable operators began to realize they had the studios locked down. As Frank Biondi, former president of the media giant Viacom (VIA), puts it: ‘Why would [the studios] make a deal with a competitor to their largest customer and risk angering them?’) (emphasis added).
-118-
NAVIGANT ECONOMICS
5102853
platform for accessing online video content. Finally, Comcast’s acquisition of NBCU’s online
content portfolio would increase Comcast’s incentives to thwart set-top-box technologies that
would enable cable television customers to access the Internet from their televisions. If a
Comcast cable customer could access Hulu.com, for example, from his television without having
to authenticate his being a Comcast cable television customer, then the value of Comcast’s cable
television platform would be compromised.
D. Comcast’s Economists Do Not Effectively Rebut the Presumption That Comcast Would Withhold NBCU’s Programming from MVPD Rivals
184. In this section, I review the reports by Drs. Katz and Israel (“Comcast’s
economists”) that analyze the foreclosure calculus for both NBCU’s O&O broadcast affiliates
and NBCU’s online content.
1. Katz’s and Israel’s Claim That Comcast Would Not Withhold NBC Affiliates
185. In their first report, Drs. Katz and Israel employ the Commission’s foreclosure
model from the News Corp.-DirecTV proceeding to determine whether Comcast would have an
incentive to withhold, on either a permanent or temporary basis, NBCU’s O&O affiliates from
rival MVPDs.396 They proceed in two steps: First, using a theoretical model, they compute the
“critical departure shares”—that is, the fraction of rival MVPD subscribers who must depart in
response to the denial of access to an NBC affiliate such that the foreclosure would be profitable.
Second, using an econometric model of Comcast’s share gains in response to Dish Network’s
temporary loss of a local affiliate, they attempt to estimate the actual departure shares. They then
propose a simple test that involves the comparison of the critical departure shares to the
estimated departure shares. If the estimated departure shares exceed the critical departure shares,
then they would infer that foreclosure would be profitable; else, foreclosure would be
396. Israel & Katz NBCU, supra, ¶¶ 21-23.
-119-
NAVIGANT ECONOMICS
5102853
unprofitable. Upon implementing their test, Katz and Israel find that the estimated “actual
departure shares” are generally less than the “critical departure shares.” Their test therefore finds
that Comcast would not have an incentive to withhold NBCU’s O&O affiliates from MVPDs
that rival Comcast.
186. Below I explain that although the proposed framework that Katz and Israel have
suggested is acceptable, their implementation of the test is poor. The end result is to conduct a
very “low-power” test. In the economics parlance, this means the test is unacceptably likely to
find that Comcast would not have incentive to withhold NBCU’s O&O affiliates from rival
MVPDs in instances when such incentive does indeed exist. Therefore, the results presented in
Drs. Katz’s and Israel’s first report should be discarded as they are uninformative.
a. Drs. Katz and Israel’s Critical Departure Shares Are Inaccurately High
187. Assuming that 100 percent of Comcast’s MVPD profits accrue to the integrated
firm, the critical departure share for a permanent foreclosure strategy is simply:
(1 )
,crit a Bd
B M
(1.2)
where a is the share of MVPD subscribers who remain with the foreclosed MVPD but obtain
access to the NBC affiliate through alternative means (for example, by receiving the signal over
the air). The variable B is the NBC affiliate’s forgone advertising revenue per subscriber per
month, M is Comcast’s MVPD margins per newfound subscriber per month, and is the share
of those customers who depart the foreclosed MVPD who specifically choose Comcast as their
new MVPD. As a matter of economic theory, I agree with this general framework.397
397. Comcast’s economists argue that the Commission’s model likely overstates the critical departure share
for a few, non-compelling reasons. For example, they argue that the model “fails to account for the possibility that attempted foreclosure would trigger a Commission enforcement action, which would be costly to the firm.” Israel &
-120-
NAVIGANT ECONOMICS
5102853
188. The values of the critical shares are ultimately determined by the parameters of
the model, which are estimated. In estimating these parameters, Drs. Katz and Israel make
several aggressive assumptions, which have the effect of artificially inflating the critical
departure shares. For example, they argue that the Commission’s suggested value for a—the
share of MVPD subscribers who remain with the foreclosed MVPD but obtain access to the
NBC affiliate through alternative means—should be increased (from 22 percent or twice the
fraction of television households that receive video programming only via broadcast reception)
to account for the possibility that a subscriber to a foreclosed MVPD could access NBC content
online.398 Setting aside the notion that the Commission’s default assumption is too generous to
begin with (most MVPD subscribers who do not use bunny ears would not find bunny ears to be
an acceptable substitute to traditional MVPD service), this adjustment presumes incorrectly that
a non-Comcast MVPD subscriber would have access to NBC’s online properties. Under
Comcast’s current tying arrangement (Fancast), an online customer cannot access Comcast’s
affiliated online portfolio unless he can authenticate that he is a Comcast cable television
subscriber. It follows that a non-Comcast MVPD subscriber could not access Comcast-affiliated
online content via Fancast. Presumably, Comcast would treat NBCU’s online properties the same
as it treats its other affiliated online content—that is, tie access to those properties to Comcast’s
cable television service. Moreover, Hulu.com currently blocks OTT providers from accessing its
website, which implies that subscribers of OTT video would not have alternative paths to NBC
Katz NBCU, supra, ¶ 88. Clearly, these considerations have not tempered Comcast’s inclination to deny access to its affiliated regional sports programming in the past. Several program access complaints brought by rival MVPDs against Comcast are underway. Drs. Katz and Israel also argue that the Commission’s model fails to account for the possibility that denial of NBC programming in a given DMA “could hurt NBC’s ability to attract the best local talent and content in that DMA.” Id. ¶ 89. It is not clear what type of local “talent” Comcast’s economists have in mind.
398. Israel & Katz NBCU, supra, ¶ 37.
-121-
NAVIGANT ECONOMICS
5102853
content. Accordingly, their scenarios in which the variable a exceeds 22 percent should be
discounted.
189. With regard to estimating the parameter , the share of those customers who
depart the foreclosed MVPD and choose Comcast as their new MVPD, Comcast’s economist
make at least three aggressive assumptions that also serve to inflate the critical departure share.
First, they assume that the likelihood that a departing customer would choose Comcast is
proportional to Comcast’s MVPD share in the DMA. But if a customer is seeking out Comcast-
affiliated content, then the likelihood that he turns to Comcast exceeds Comcast’s MVPD market
share—that is, the customer would not be indifferent between Comcast and some other MVPD
supplier that might also be foreclosed from NBC programming.
190. Second, the MVPD shares used by Comcast are generally less than those reported
by SNL Kagan, the Television Bureau of Advertising (TVB), and Warren’s Advanced
TVFactbook. For example, Comcast’s economists assert that Comcast’s MVPD share in the
Hartford-New Haven DMA is {{39.6%}}. According to TVB, Comcast sells local advertising to
751,320 households (“local market interconnects”) in the DMA or 76.8 percent of television
households.399 According to Warren’s, Comcast supplies 49.6 percent of the television
households in the DMA. According to SNL Kagan, Comcast serves 40.7 percent of MVPD
subscribers in the DMA. Table 6 compares the market shares across the three data sources.
399. U.S. TV Household Estimates Designated Market Area (DMA)—Ranked by Households, available at
http://www.tvb.org/nav/build_frameset.asp?url=/arc/politicaldatabank/politicaldatabank.asp. Local Cable Reach Guide DMA Interconnect UEs Feb.10, at 5 (showing “COMCAST SPOTLIGHT – HARTFOR” as the largest MVPD in the DMA with 73.7% of television households). TVB defines an interconnect as “a group of two or more wired-cable systems within a DMA handled by the same ad sales entity.”
4 San Francisco-Oakland-San Jose, CA 58.2% 59.3% 73.4% 68.5%
5 Washington, DC (Hagerstown, MD) 45.7% 48.0% 44.7% 71.1%
6 Hartford & New Haven, CT 39.6% 40.7% 49.6% 76.8%
7 New York, NY 9.7% 9.8% 11.1% n/a
8 Dallas-Fort Worth, TX 0.0% 0.0% 0.0% n/a
9 Los Angeles, CA 0.0% 0.0% 0.0% n/a
10 San Diego, CA 0.0%}} 0.0% 0.0% n/a
Notes and Sources: * Israel & Katz NBCU, supra, Table 1, at 29. Percentages represent Comcast share of “MVPD Subscribers,” in late 2009. *** Comcast basic subscriber data available from Advanced TVFactbook, Warren's Communications News, May 2010. Data on DMA Cable/ADS penetration taken from TVB Local Cable Reach Guide Feb. 2010, Television Bureau of Advertising, available at: http://tvb.org/nav/build_frameset.aspx. **** Percentages represent Comcast share of “Local Market Interconnects” as a percentage of HHs reached by Cable/ADS. Data on DMA Cable/ADS penetration taken from TVB Local Cable Reach Guide Feb. 2010, Television Bureau of Advertising, available at: http://tvb.org/nav/build_frameset.aspx. ^ Data for Chicago, IL DMA from the Advanced TVFactbook was incomplete. As Table 6 shows, similar discrepancies appear in the nine other DMAs, with Comcast’s share
estimates generally being lower than SNL Kagan’s, TVB’s, and Warren’s estimates.400
191. Third, Comcast’s economists assume that a substantial share of Verizon’s, Dish
Network’s, and DirecTV’s customers could not switch due to “long-term subscriber contracts”
for at least a year.401 Setting aside the reliability of their estimates of the customers under long-
term contracts,402 at least some customers (for example, those with a strong affinity for NBC
programming and high incomes) would be willing to incur the penalty for early termination.
Verizon’s early-termination fee for FiOS customers who joined in 2009 or earlier is $179.403 Yet
Comcast’s economists claim that customers “under long-term contracts cannot, in fact,
switch.”404
400. Compare Table 1, infra, with Table 1 in Israel & Katz NBCU. 401. Israel & Katz NBCU, supra, ¶ 57. 402. Id. Comcast’s economists cite a DBS survey conducted by Comcast and an estimate for Verizon
customers supplied by the director of competitive analysis at Comcast. 403. Fee information is available at http://www.dslreports.com/shownews/Verizons-FiOS-ETF-360-Starting-
January-17-106253. Although the fee was increased in 2010, the fee is reduced by $20 per month for each month into the contract.
404. Israel & Katz NBCU, supra, ¶ 57.
-123-
NAVIGANT ECONOMICS
5102853
192. With regard to estimating B, the NBC affiliate’s forgone advertising revenue per
subscriber per month, Comcast’s economists assert that the “trend” toward retransmission
consent fees dictates that those fees should be added to NBC’s forgone advertising revenues.405
In a recent paper for the National Cable & Telecommunications Association, Dr. Katz argued
that the nascent retransmission consent “system significantly harms consumer welfare through
higher subscription fees and the periodic (and to consumers, unpredictable) loss of access to
retransmitted broadcast signals.”406 He reports that only 72 percent of telco video customers
subscribed to an MVPD that paid retransmission fees in 2009;407 yet for the purposes of the
foreclosure calculus, he implicitly assumes that 100 percent of those telco video subscribers (and
all other foreclosed MVPD subscribers) belong to an MVPD that paid those fees. Dr. Katz
concludes his NCTA report with the suggestion that the entire system “be reviewed to determine
the consumer benefits of restoring the balance between the parties in retransmission consent
negotiations,” and suggests that MVPDs should be allowed “to import broadcast signals from
neighboring areas,” presumably for free.408 Hence, if the Commission followed Dr. Katz’s policy
prescription in the retransmission consent proceeding, then those fees should not be included in
the foreclosure calculus.
193. Despite these aggressive assumptions, given Comcast’ high MVPD market shares
in many of the relevant DMAs, Drs. Katz and Israel estimate that the low-end of the critical
departure shares under the “permanent foreclosure” simulation for six DMAs (Chicago, Hartford
& New Haven, Miami-Ft. Lauderdale, Philadelphia, San Francisco-Oakland-San Jose, and
Washington-Hagerstown) ranges from {{6.0}} (Philadelphia) to {{10.5}} percent (Hartford &
405. Id. ¶ 66. 406. Michael Katz, Jonathan Orszag and Theresa Sullivan, An Economic Analysis of Consumer Harm from
the Current Retransmission Consent Regime, Nov. 12, 2009, at 1. 407. Id. at 34 (Table 3). That share is estimated to increase to 87 percent in 2011. 408. Id. at 41.
-124-
NAVIGANT ECONOMICS
5102853
New Haven).409 Stated differently, by their own model, if just six-to-ten percent of MVPD
subscribers in the relevant DMAs departed their MVPD in response to the denial of access to
NBC programming, then Comcast’s foreclosure would be profitable in those DMAs! (The low-
end of the critical departure share in the New York DMA is 29.1 percent, largely due to
Comcast’s small share of MVPD subscribers.) These modest shares, which are likely inflated for
the reasons described above, would be trivially satisfied to the extent that Comcast’s permanent
withholding of RSN programming from DBS operators in the Philadelphia DMA is a reasonable
proxy for the predicted departure level here. Recall that several studies,410 including two by the
Commission,411 have estimated the effect of that foreclosure on DBS shares, concluding that
Comcast’s denial of RSN programming led to a decrease in DBS market share well in excess of
10 percent.
b. The Katz-Israel Empirical Methodology of Actual Departure Share Is Conducted in a Manner that Likely Understates the True Departure Shares
194. Setting aside the implementation of the FCC’s foreclosure model, a key
component of the Katz-Israel test for foreclosure is their econometric model of foreclosure. With
that model they seek to empirically estimate the market share that Comcast could expect to gain
from a foreclosure strategy. Matched against a critical departure share, if the expected increase in
market share is less than the critical value, then, presumably, Comcast would not have incentive
to withhold content from a rival MVPD. Consequently, the accuracy of the test hinges not only
on the parameterization of the critical departure share (discussed above), but also on the accuracy
409. Israel & Katz NBCU, supra, 46 (Table 2). 410. FCC Adelphia Order, supra, ¶ 149 (“We find that the percentage of television households that subscribe
to DBS service in Philadelphia is 40% below what would otherwise be expected given the characteristics of the market and the cable operators in the DMA.”); FCC 2007 Program Access Order, supra, ¶ 39 (finding that DBS penetration in Philadelphia was significantly below the predicted level).
411. Israel & Katz NBCU, supra, ¶ 77 (Table 2).
-125-
NAVIGANT ECONOMICS
5102853
of the econometric model that estimates the actual departure share. If the econometric model
produces estimates of the actual departure share that are inaccurately low, then the Katz-Israel
test could conclude incorrectly that Comcast had no incentive to engage in foreclosure when in
fact it did. Unfortunately, it is just this type of error that is embedded in the authors’ econometric
estimates. Therefore, the Katz-Israel test vastly understates Comcast’s likely incentive to engage
in the foreclosure of NBC affiliates.
195. The Katz-Israel econometric analysis understates actual departure share for two
main reasons. First, it analyzes the departure share from temporary loss of broadcast stations by
only one MVPD. Specifically, the underlying data upon which the Katz-Israel estimates are
based involve retransmission disputes that tended to last only a matter of days and only involved
DISH Network. Second, the data compiled by Drs. Katz and Israel only measure actual departure
to Comcast. There are, however, multiple MVPDs that could have captured any share that
departed DISH Network during its disputes with the broadcasters in question. As a result, the
Katz-Israel analysis does not accurately measure the actual departure share that Comcast could
capture by sequentially excluding rival MVPDs from affiliate O&O content.
196. The first problem with the empirical model that Drs. Katz and Israel estimated is
that the majority of the “experiments” that comprise their dataset involve retransmission disputes
that last only a handful of days. The Katz-Israel study relies on four recent retransmission
disputes involving DISH Network. Three of those disputes lasted for between 46 hours and three
days, which is an unreasonably short time period in which to measure the “actual departure
share” that could take were Comcast to foreclose NBCU’s O&O affiliates from rival MVPDs.
This set of short retransmission disputes cannot serve to reasonably estimate the true departure
share that would occur were Comcast to withhold NBCU’s O&O affiliates from rivals, and,
-126-
NAVIGANT ECONOMICS
5102853
consequently, the analysis that Drs. Katz and Israel draw from their analysis of these disputes is
unreliable.
197. Drs. Katz and Israel do analyze a retransmission dispute that was longer in
duration than those mentioned above. A dispute between Fisher Broadcasting and DISH
Network lasted approximately six months. In applying their regression methodology to Comcast
market share data in geographies affected by this dispute, Drs. Katz and Israel again find no
significant effect on Comcast’s market share. Again, however, their test is set up to fall in
Comcast’s favor, as they fail to test whether market share for another MVPD, such as DirecTV,
increased during this time and in these media markets. If a customer has already shown a
preference for DBS service by choosing DISH, then that customer would be likely to switch
from DISH to DirecTV, which had the Fisher Broadcasting content, were the absence of that
content burdensome to the customer.
198. The failure to account for the possible movement of customers in the event of a
retransmission dispute to an MVPD other than Comcast is highlighted when one considers that
Drs. Katz and Israel calculated their critical departure shares by grouping DISH Network with
DirecTV. Because both DBS providers will be negotiating over rates for access to NBCU’s
O&O affiliates at roughly the same time, however, the effect of withholding that content from
both DBS providers would be more beneficial to Comcast’s market share than the gains accrued
from withholding it from only one MPVD. Consequently, the timing of the expiration of existing
NBCU O&O agreements would give Comcast an opportunity to engage in withholding that is
significantly more extensive than any positive effect that might have been experienced during
retransmission disputes involving DISH and various broadcast companies.
-127-
NAVIGANT ECONOMICS
5102853
199. For the foregoing reasons, the Commission should not credit Drs. Katz’s and
Israel’s NBCU O&O broadcast affiliate study when predicting the anticompetitive effects of the
proposed transaction.
2. Drs. Katz’s and Israel’s Claim That Comcast Would Not Withhold Affiliated Online Content
200. In their second report for Comcast, Drs. Katz and Israel conclude that the
proposed transaction would not reduce competition in the supply of “long-form, professional
quality video programming” via the Internet.412 They define “long-form, professional quality
video programming” via the Internet as online video that exceeds five minutes in length, such as
full episodes of a television show or a movie.413 This conclusion is based on several “findings,”
including (a) online video distribution is currently a complement (and not a substitute) to
Comcast’s cable television service; (b) online video distribution will be a complement to the
extent that online video distributors seek to differentiate themselves from traditional MVPDs in
the future; (c) online video distribution is currently a complement to Comcast’s cable modem
service; and (d) application of the Commission’s foreclosure model indicates that foreclosure
would be unprofitable to Comcast. In my opinion, these “findings” do not withstand economic
scrutiny; and even if they did, they would not cast doubt on Comcast’s anticompetitive
motivation to foreclose OTT providers.
201. Drs. Katz’s and Israel’s “finding” that online video distribution and cable
television service are complements and not substitutes is based on two (time-series) trends. First,
they note that the average number of television minutes watched from “traditional” MVPDs has
increased while online video consumption has increased.414 Second, they note that the number of
MVPD subscribers has increased while online video consumption has increased.415
Unfortunately, neither metric is a valid test of whether online video consumption and television
viewing watched from “traditional” MVPDs are economic substitutes. Economists define two
goods, A and B, as substitutes if the demand for A increases as the relative price of B
increases.416 Contrary to the test proposed by Drs. Katz and Israel, the change in quantity
demanded for B does not enter into the definition of economic substitution. Side-stepping this
critical issue, Drs. Katz and Israel appear to use the increase in the consumption of online video
as a proxy for a decrease in the price of online video. This is plainly incorrect. In reality, the
price of online video has not changed over the course of their experiment; if anything, the price
of online video has increased with the introduction of new, online subscription services such as
Playon.tv and Boxee. Comcast’s economists are completely silent on the direction of the relative
price change. In their flawed test for substitutability, online video is a substitute for television
viewing from “traditional” MVPDs if and only if incremental minutes of online video
414. Israel & Katz Online Video, supra, ¶¶ 4, 23-24. Comcast’s economists also argue that the (large) ratio of
television watching to online video consumption implies the two services are substitutes. But again, substitution does turn on the relative volume of consumption across two goods. Id. ¶24.
415. Id. ¶ 4. They also argue that differences in viewing patterns imply complementarity. Notwithstanding the problems in using this approach to test for economic substitution, their claim that television viewing is “much steadier” than online viewing is likely driven by their combination of short-form and long-form online video. See id. ¶ 25. Stated differently, to the extent that online viewers watch YouTube clips by day and Hulu.com by night, consumption of long-form online video likely tracks traditional television viewing patterns more closely. To further support their hypothesis of complementarity. Comcast’s economists also claim that traditional television viewing is “more linear” than online viewing. See id. ¶ 25. But the growing popularity of DVR use and video-on-demand services, as acknowledged by Drs. Katz and Israel, suggests that traditional television viewing is moving in the direction of non-linear programming. Id. ¶ 27 (noting that traditional television viewing has experienced a “recent growth in video-on-demand services”). They also cite surveys showing that television viewers frequently use online viewing to “supplement” their television viewing. See id. ¶ 29. This reminds one of the ways in which landline telephone consumers initially used wireless telephones to “supplement” their landline calls. As users became more comfortable with wireless telephones, they increasingly substituted away from landline calls, and some even “cut the cord” entirely. Indeed, some analysts in the mid-1990s mistakenly considered wireless telephony to be a rich-man’s complement to landline telephony. Similarly, Drs. Katz and Israel claim that the lack of “cord cutting” as of 2009 implies lack of substitutability. See id. ¶ 37.
416. See, e.g., HAL VARIAN, MICROECONOMIC ANALYSIS 60 (W.W. Norton & Co. 3rd ed. 1992) (explaining that for two factors to be substitutes, “the sign of the cross-price effect [elasiticity of demand] must be positive”).
-129-
NAVIGANT ECONOMICS
5102853
consumption come at the expense of television viewing minutes from “traditional” MVPDs. But
this test has nothing to do with the definition of economic substitutes, which turns on whether the
cross-price elasticity of demand is positive or negative.
202. Even if their trend analysis was the correct test for economic substitutes, Drs.
Katz and Israel fail to consider whether the recent doubling of online video consumption,
including watching online video via televisions, has come at the expense of minutes that would
otherwise be spent watching television. Stated differently, in the absence of the incremental six
hours of online video consumption per month in 2009 (from roughly six to twelve hours per
month), the relevant questions is: Would those same viewers have watched even more television
through “traditional” MVPDs? The answer is likely “yes,” and none of Katz’s and Israel’s data
refute this reasonable inference. The fact that television minutes through “traditional” MVPDs
has also increased over the past year says little about how many hours have been displaced by
online video consumption.
203. Moreover, the survey data on which Drs. Katz and Israel rely are based on
averages across all television or online viewers, and therefore do not convey reliable information
about specific consumer groups. To make matters concrete, suppose there are two groups of
viewers in the survey, and that each group is equally represented: (a) young viewers who
perceive online video to be a close substitute for watching television through “traditional”
MVPDs, and (b) elderly viewers who are unlikely to perceive the two as strong substitutes.
Suppose that in 2009 the first group converted twelve hours of watching television per month
through “traditional” MVPDs to online video consumption, reducing their average consumption
of television watching through “traditional” MVPDs from 150 to 138 hours. Suppose that the
second group increased their television hours through “traditional” MVPDs per month from 150
-130-
NAVIGANT ECONOMICS
5102853
to 170 hours. Total online video consumption has increased across both groups by six hours on
average (equal to 12 plus 0 divided by 2). Total television consumption through “traditional”
MVPDs has increased across both groups by four hours on average (equal to -12 plus 20 divided
by 2). Drs. Katz and Israel would infer incorrectly from this experiment that online video
consumption and watching television through “traditional” MVPDs are complements because the
average number of both traditional television hours and online video hours increased at the same
time. But the averages mask the substitution that is likely occurring among the younger group of
viewers; a group whose preferences will better represent the preferences of the larger television
audience in the coming decades.
204. Moreover, the notion that online video and television watching via “traditional”
MVPDs are complements cannot be squared with the testimony and strategies of Comcast and
other cable television providers. If Comcast believed that online video was a complement, as
Drs. Katz and Israel claim, then why would it submit comments to the Commission in November
2006 that Internet video is “providing consumers with an interactive alternative to traditional
TV-set viewing,”417 which “compete[s] with traditional and not-so-traditional video distribution
technologies for time, attention, and dollars.”418 Indeed, in their companion report filed in this
very proceeding, Drs. Katz and Israel suggest that online video is in fact a substitute for
traditional MVPD service.419 And why would Comcast engage in exclusionary conduct, such as
417. Comcast Comments at 29-30 (emphasis added). 418. Id. at 59 (emphasis added). 419. See Israel & Katz NBCU, supra, ¶ 61 (“The rise of online video has been one of the major developments
since the time the Commission staff model was first developed. If Comcast were to pursue a foreclosure strategy following consummation of the proposed transaction, some consumers who remained with their current MVPD could be expected to switch to watching at least some NBC programs online.”) (emphasis added). This observation undercuts their subsequent argument that online video is used merely to “supplement” traditional MVPD viewing, as the online NBC video content would replace the same content that was previously accessed via a traditional MVPD. See also id. ¶ 37 (“However, as discussed more fully below, in the current environment, online alternatives, such as Hulu.com and NBC.com, may have (at least partially) replaced over-the-air viewing as alternatives for those not watching on an MVPD.”) (emphasis added). If over-the-air viewing is an alternative for cable television, and if
-131-
NAVIGANT ECONOMICS
5102853
tying its online content portfolio to the purchase of its cable television service and cable modem
service, if it did not perceive online video to be a threat to its cable television business?
205. Finally, even if online video is currently a complementary service to television
watching through “traditional” MVPD service, as Drs. Katz and Israel claim, the economics
literature recognizes that a firm could engage in exclusionary conduct, including the use of a tie-
in,420 to prevent a rival in the tied (and complementary) market from evolving into a competitor
in the tying market in future periods. These economic models were originally designed to capture
Microsoft’s incentives to thwart the development of “middleware,” which in the near term was
arguably a complement to Microsoft’s Windows operating system, but in the long term could
become a threat to Microsoft’s dominance to the extent that software applications could run on
middleware in the future. The result would have been to render Microsoft’s operating system
irrelevant. (The fact that Microsoft’s operating system is still relevant implies that either the
theory was wrong or Microsoft’s conduct delayed the evolution of middleware.) Applied here,
even if Comcast perceives online video to be a complement to its cable television service today
(a dubious proposition in light of Comcast’s November 2006 testimony), to the extent that a
vibrant, online-video offering could one day induce Comcast’s subscribers to disconnect their
cable television subscriptions, Comcast’s tie-in of its affiliated online content portfolio to its
cable television service could prevent that evolution from occurring. And the acquisition of
NBCU’s online content portfolio greatly facilitates that exclusionary strategy. In sum, even if
Drs. Katz and Israel are correct that online video is currently a complement to television
online viewing is an alternative to over-the-air viewing, then online viewing is an alternative for cable television by transitivity.
420. See Dennis W. Carlton & Michael Waldman, The Strategic Use of Tying to Preserve and Create Market Power in Evolving Industries, 33 Rand J. Econ. 194, 198–212 (2002). See also Testimony of Professor Robin Cooper Feldman, Department of Justice Tying Seminar, Hearing Transcript, 65-66 (Nov. 1, 2006) (noting that tying can involve the monopolist “trying to protect its original monopoly from the next generation of products”).
-132-
NAVIGANT ECONOMICS
5102853
watching through “traditional” MVPD service, Comcast’s tie-in still could be motivated for
anticompetitive reasons, rendering the entire discussion about the proper categorization of online
content (complement versus substitute) academic.
206. With regard to their second “finding”—namely, that online video distribution will
be a complement to the extent that online video distributors seek to differentiate themselves from
traditional MVPDs—Comcast’s economists offer nothing but speculation. For example, they
suggest that OTT providers will “incorporat[e] some of the unique capabilities of online
platforms that allow them to supplement and complement traditional MVPD offerings.”421 It is
not clear what they mean by “unique capabilities;” it is hard to differentiate an online episode of
Modern Family from the original airing on ABC. The most likely way in which OTT providers
will “distinguish” themselves from cable television providers is by lowering the monthly
subscription fee and by offering a-la-carte services. Indeed, Comcast’s economists cite an
industry analyst making the same point.422 Given that all subscribers prefer spending less for the
same amount of video programming, and given that at least some prefer consuming video in
smaller bundles, this type of product differentiation will likely accelerate the erosion of
Comcast’s dominance. In any event, mere product differentiation among suppliers does not
imply complementarity, as suggested by Drs. Katz and Israel. Entrants in most product-
differentiated industries, including restaurants and clothing stores, still compete on price (albeit
to a lesser degree than firms in homogenous industries) and still constitute a threat to
incumbents.
421. Israel & Katz Online Video, supra, ¶ 4 (emphasis added). 422. Id. ¶ 20 (“We expect the [OTT] model will have to evolve by factoring direct spending from the
consumer, either subscription or a la carte.”) (citing UBS analyst Matthieu Copper). See also id. at ¶ 20 (arguing that a move by OTT providers to subscription models, even though closer in product space to traditional MVPDs, would still “not imply they have become substitutes for traditional MVPDs”). It is not clear what would satisfy Comcast’s economists. Their standard appears to be that OTT providers must embrace the identical business model to be considered a substitute for traditional MVPDs—that is, any amount of differentiation implies complementarity.
-133-
NAVIGANT ECONOMICS
5102853
207. With regard to their third “finding”—namely, that online video distribution is a
complement to Comcast’s cable modem service—Comcast’s economists confuse the demise of
OTT providers (a good thing for Comcast) with the demise of online video viewing (a bad thing
for Comcast). In particular, Drs. Katz and Israel argue that an increase in online video viewing
would stimulate the demand for cable modem service; hence, impairing the development of
online viewing would reduce Comcast’s cable modem profits. But Comcast’s tie-in of its online
content portfolio to both its cable television and cable modem service is not designed to
eliminate online viewing. Indeed, Comcast is charging an imputed price of zero for online video
viewing (the same price that Microsoft charged for Explorer conditional on buying Windows),
which should encourage the activity at the margin (for the same reason that subsidized buffets in
Las Vegas encourage overeating). Comcast prefers more online viewing to less for the reasons
articulated by its economists with one important caveat; online viewers must also subscribe to
Comcast’s cable television and cable modem service. Accordingly, Comcast would not sacrifice
its cable modem profits by foreclosing OTT providers, as Drs. Katz and Israel suggest. This is a
false choice. In sum, Comcast’s tie-in is aimed at thwarting OTT providers, not online video
viewing in general.
208. With regard to their fourth “finding”—namely, that application of the
Commission’s foreclosure model indicates that foreclosure would be unprofitable to Comcast—
Comcast’s economists’ mathematical proof is not convincing. Before criticizing their approach,
it bears noting that Comcast already ties its online video content portfolio to its cable television
and cable modem service. Accordingly, importing and calibrating a theoretical model to assess
whether Comcast would foreclose OTT providers is a curious exercise (and a moot point). The
relevant question is whether the proposed acquisition would enhance Comcast’s current
-134-
NAVIGANT ECONOMICS
5102853
exclusionary strategy (Fancast), which subsumes the refusal to supply OTT providers with
affiliated content, and not whether it would permit Comcast to engage in some hypothetical
foreclosure strategy. Setting aside the fact that Comcast is currently engaging in the foreclosure
strategy, and setting aside the fact that several OTT providers, including Boxee and Playon.tv,
are “direct competitors for traditional MVPD services,”423 Drs. Katz and Israel base their
conclusion that Comcast would incur losses from such a strategy on four faulty premises: (a)
NBCU would lose significant advertising revenues and affiliate fees were the combined entity to
deny its affiliated programming to online competitors; (b) withholding current NBCU online
content from OTT providers could not significantly harm the ability of an online provider to
attract or retain subscribers; (c) Comcast has a limited geographic footprint and would gain only
a small share of any benefits associated with foreclosure; (d) an OTT provider that was a direct
competitor to Comcast’s cable television service would be complementary to Comcast’s cable
modem service.
209. With regard to the claim that NBCU would lose significant advertising revenues,
Comcast’s economists rely on unrealistic assumptions to inflate these losses. For example, they
assume that NBCU would lose advertising and license fees that it would otherwise obtain from
the foreclosed MVPD.424 But these losses are reduced (and approach zero) to the extent that
Comcast’s foreclosure strategy induces its cable television customers not to cut the cord; if the
customer remains with Comcast cable television service, then Comcast would continue to pay
(via an internal transfer) license fees to NBCU. And NBCU would continue to enjoy the
associated revenue. Moreover, by using current license fees (per subscriber per month) paid by
423. Their model assumes “the emergence of one or more hypothetical online distributors that are direct
competitors for traditional MVPD services”—as if this is stretch of the imagination. Israel Katz Online Video, supra, ¶ 4 (emphasis added).
424. Id. ¶ 66.
-135-
NAVIGANT ECONOMICS
5102853
traditional MVPDs to NBCU’s cable networks, Comcast’s economists likely inflate NBCU’s
forgone revenue at issue here to the extent that OTT providers negotiate smaller license fees with
cable networks. The rationale for smaller fees is that OTT providers might secure a subset of the
programming that traditional MVPDs obtain; for example, OTT providers might refrain from
purchasing live-event programming or news.425 Another rationale for smaller fees is that online
rights are likely less valuable than the rights associated with traditional delivery channels.
210. With regard to the claim that withholding NBCU’s online properties would not be
an effective exclusionary strategy, Comcast’s economists rely on (a) the remote possibility that
OTT subscribers who lose access to NBCU’s online content would access that content with
bunny ears;426 and (b) their prior “finding” that Comcast did not enjoy a windfall in local markets
where Dish Network temporarily lost access to a local broadcast affiliate. Setting aside the fact
that not all of NBCU’s online content is aired on a local broadcast network (for example, some
of NBCU’s Olympic programming is aired on MSNBC), the notion that OTT subscribers would
be willing to watch NBC content via bunny ears does not pass the laugh test.427 Rather than
suffering from the reduced quality associated with bunny ears, the OTT subscriber with a strong
interest in NBC programming would most likely switch back to an MVPD that had access to the
local NBC affiliate (or never leave)—the very outcome that Comcast is counting on by tying its
online portfolio to its cable television service. And for the reasons discussed in the prior section,
Drs. Katz’s and Israel’s “finding” that losing access to a local affiliate does not impair an
425. Indeed, Comcast’s economists admit as much. Id. ¶ 89 (“For example, an online MVPD might take
advantage of the technological flexibility of the Internet to offer programming packages that are smaller than those offered by traditional MVPDs and are aimed at specific demographic groups.”) (emphasis added).
426. Id. ¶ 90 (“However, two facts indicate that such a suggestion would be unwarranted. One is that NBCU cannot deny consumers access to NBC’s signal via over-the-air reception.”).
427. The authors cite to one OTT provider, Sezmi, that has integrated an antenna into its hardware. Id. ¶ 91. The fact that one provider has done so (and even got a review in PC World) does not imply that MVPD subscribers would tolerate that solution is sufficient quantities to defeat Comcast’s foreclosure strategy. Comcast’s economists acknowledge that NBCU does not permit Sezmi to obtain NBC broadcast programming online. Id.
-136-
NAVIGANT ECONOMICS
5102853
MVPD’s ability to compete effectively is fatally flawed. Accordingly, withholding NBCU’s
online properties from OTT providers would likely be an effective exclusionary strategy.
211. With regard to the claim that Comcast has a limited geographic footprint and
would gain only a small share of any benefits associated with foreclosure, Comcast’s economists
fail to consider that Comcast’s current exclusionary conduct is being carried out jointly with
Time Warner and other out-of-region cable operators via TV Everywhere. The effect of this
collaboration is two-fold. First, the coordinated refusal to deal with OTT rivals expands the
associated foreclosure share, and thereby enables the participants of TV Everywhere to capture
the benefits from the foreclosure strategy, even outside Comcast’s territory. Second, by
coordinating with Time Warner, Comcast can deny OTT providers the online portfolios of three
content providers—NBCU, Time Warner, and Comcast itself. The effect of denying all three
portfolios is more devastating than withholding just one; because Comcast’s economists fail to
consider the coordination of exclusionary strategies between Comcast and Time Warner, their
model vastly understates the profitability of a foreclosure strategy. As it turns out, Comcast’s
“limited” downstream footprint—one quarter of all MVPD households—is nearly sufficient to
make foreclosure profitable on a unilateral basis according to their own model in some
specification. Therefore, it is inappropriate to use Comcast’s MVPD share for the likelihood that
foreclosed OTT customers switch back to Comcast (their “Diversion to Comcast Cable”
variable);428 the appropriate input is the sum of market shares of all MVPD members of TV
Everywhere, including, at a minimum, the combined shares of Comcast and Time Warner.
428. There is yet another reason why Comcast’s economists likely understate the Diversion to Comcast Cable
variable. In particular, they assume that an OTT subscriber who is upset about lack of NBCU or Comcast programming—the relevant withheld programming here—is likely to switch to Comcast in proportion to Comcast’s MVPD share, as if the selection decision among traditional MVPDs is independent of the cause of the defection. Of course, if the cause of the switch is lack of access to Comcast-affiliated programming, then the probability of
-137-
NAVIGANT ECONOMICS
5102853
212. Finally, with regard to the claim that an OTT provider that was a direct
competitor to Comcast’s cable television service would be complementary to Comcast’s cable
modem service, Comcast’s economists appear to be backtracking on the fundamental assumption
of the FCC’s model. Either one must assume online video is a substitute to cable television
service and implement the FCC’s foreclosure model, or one must assume the two services are
complements and abandon the modeling exercise. But Comcast’s economists pursue a “third
way” that involves modeling foreclosure and rejecting the fundamental assumption of
substitutability; again, they cannot have it both ways.
213. In addition to these fundamental modeling errors, Drs. Katz and Israel commit
other errors that render their conclusions unreliable:
Drs. Katz and Israel are pessimistic that OTT providers could thrive even in the absence of a foreclosure strategy. For example, they suggest that OTT providers may not have access to online content for reasons unrelated to Comcast’s foreclosure strategy: “[E]ven in the common case in which a broadcast or cable network has obtained rights from the studio that include the rights for some on-demand airings of the programming, these rights may not extend to the particular business model that an online distributor has adopted.”429 The implication is that it is the studios and not Comcast/Time Warner that may be responsible for the denial of access to OTT providers. While the studios may demand compensation for online rights, their general lack of vertical integration into video distribution ensures that they will not consider illicit downstream gains to withholding that content when formulating prices. In the same spirit, Comcast’s economists argue that current prices charged by content delivery networks (CDNs), which is a proxy for online transport costs, make it uneconomic for a company to deliver over the Internet video programming that would have viewing patterns similar to those of traditional television.”430 Setting aside the technical issue of whether CDN prices are a good proxy for the relevant costs, this calculation presumes that CDN prices per GB would not fall with more output—that is, they assume no economies of scale and no technological improvements.
Drs. Katz and Israel also argue that a “marginally profitable” OTT provider “would pose little competitive threat to Comcast and offer little expected benefits
diversion back to Comcast vastly exceeds Comcast’s market share. That is, if the same programming is also unavailable from a DBS provider, then the customer will most likely find his way back to Comcast.
to consumers because the firm would be unlikely to survive and/or develop into a significant rival.”431 Of course, price-discipline does not turn entirely on the profitability of a rival—cross-price elasticity of demand is what matters. Indeed, even a less efficient rival who can cover its average variable costs can impose some price discipline on an incumbent. And an entrant like Google, which seems content to disrupt existing technologies for little-or-no short-run incremental profits, could significantly threaten Comcast’s dominance.
In support of their thesis that online viewing is used to “supplement” traditional television viewing, Drs. Katz and Israel cite Bernstein Research showing that online viewing is generally done from a computer.432 Yet the data they cite suggests that a full 26 percent of online viewers consume online video from something other than their computers.433 To the extent that this consumption is largely occurring via their televisions, the Bernstein survey suggests that online video is not merely supplementing traditional television viewing but replacing it.
In their foreclosure model, Drs. Katz and Israel do not consider the possibility that foreclosure of NBCU’s online video content could spare Comcast from “cord shaving”—that is, a reduction in some but not all of the cable video services offered by Comcast.434 This assumption, which cuts in Comcast’s favor, is premised on the notion that most of NBCU’s valuable online content appears on analog cable tiers. Setting aside the fact that over {{ten percent}} of NBCU’s networks are not carried on analog tiers,435 because Comcast’s foreclosure strategy is being implemented in conjunction with Time Warner, one must consider access to Time Warner’s online video portfolio in the foreclosure calculus. Specifically, to the extent that any of the marquee programming in Comcast’s, NBCU’s, and Time Warner’s combined online video portfolio is not carried on analog tiers, then Comcast’s video subscribers could save money (and still access that content on an a-la-carte basis online) by opting for basic cable only. Independent of where the affiliated programming is tiered, Comcast’s subscriber may also continue their video subscription albeit at lower levels to access other programming (including live-event programming) and to avoid paying a “penalty price” for Comcast’s cable modem service when purchased on a standalone basis.
Drs. Katz and Israel use market share metrics to assess the “must-have” nature of NBCU’s online content. For example, they estimate that Universal Studios share of box office receipts was only 10.1 percent from 2005 to 2009.436 Similarly, they estimate that NBCU’s online content accounted for only {{17.9}} percent of
431. Id. ¶ 52 n. 74. 432. Id. ¶ 31. 433. Id. (“74% of online viewers do so on their computer monitors—without connecting to the TV”) (citing
October 2009 Bernstein Research). 434. Id. ¶ 39 n. 73. 435. Id. 436. Id. ¶ 50 n. 73.
-139-
NAVIGANT ECONOMICS
5102853
hours viewed of professional long-form video (excluding YouTube) in 2009.437 Setting aside the fact that {{18}} percent may constitute a significant share of long-form online video, market shares do not capture the “must-have” nature of a given program. For example, NBC’s share of ratings on a given night might be low, but no one would argue that the local NBC affiliate is not “must-have” programming. RSNs are also considered “must-have” programming, but not on account of their shares of advertising revenues or TV ratings. Accordingly, the “low” shares of NBCU’s box office and online content does not imply that an OTT provider would not be impaired without access to those online properties.
Drs. Katz and Israel only consider parameter values for the most important variable in the foreclosure calculus, Change in Online-MVPD Subscriptions, which ranges from 0 to 33 percent.438 This variable, which is completely speculative, captures the percent of online MVPD subscribers who would obtain subscriptions from traditional MVPDs, including Comcast, if online MVPDs lost access to NBCU programming. Recall that under Comcast’s current foreclosure strategy, OTT providers are denied access to Comcast’s programming and Time Warner’s programming; adding NBCU’s programming to the mix would make the foreclosed input that much more valuable.439 As it turns out, when Change in Online-MVPD Subscriptions reaches just {{37.3 percent}}, Comcast’s economists’ model predicts that, under certain parameterizations, foreclosing OTT providers would be profitable for Comcast.440
A factor completely absent from their calculus is the “penalty” price that Comcast charges for standalone cable modem service. Faced with this penalty for “breaking the bundle”—which effectively causes the customer to pay $15 per month for cable television service without getting it441—most customers would continue to purchase Comcast’s cable television service. Accordingly, the combination of the withholding of Comcast, NBCU, and Time Warner content plus a penalty price for standalone cable modem service suggests that Comcast should induce a significant percentage of OTT video customers to switch back to Comcast’s cable television service (or never leave in the first place).
437. Id. ¶ 87 (Table 1). 438. Id. ¶ 97. 439. Their failure to consider parameter values above 33 percent reminds one of Morgan Stanley’s refusal to
consider default probabilities among subprime borrowers above 10 percent before the financial meltdown, as anything over that level would imply “like, a million homeless people.” See MICHAEL LEWIS, THE BIG SHORT 211 (W.W. Norton & Co. 2010).
440. Israel & Katz Online Video, supra, 70 (Table 3). 441. Shop Comcast, available at https://www.comcast.com/shop/buyflow2/customize.cspx (“Performance for
Comcast Cable and/or CDV Customers (Add $44.95 per month) Select this option if one of the following applies: You are currently subscribed to Comcast Cable and/or Comcast Digital Voice services and purchasing or upgrading to Performance You are purchasing or upgrading to Performance and also purchasing Comcast Cable and/or Comcast Digital Voice at this time. Performance for Comcast Internet Only Customers (Add $59.95 per month)”) (accessed May 16, 2010). The difference between the in-bundle monthly price for cable modem service and the standalone monthly “penalty” price is $15.
-140-
NAVIGANT ECONOMICS
5102853
To reduce the profitability of the contemplated foreclosure strategy to Comcast, Drs. Katz and Israel implausibly assume (under one scenario) that Comcast’s broadband revenues decrease when an OTT customer is anticompetitively induced to switch back to Comcast’s cable television service. Their explanation, which strains logic to the breaking point, reads as follows: “Some of those users [switching back from an online MVPD to a traditional MVPD] would likely ‘downgrade’ to a lower-volume broadband Internet access tier.”442 Presumably, they assume that the consumer ramps up his broadband package to support online video and then ramps back down upon his decision to switch back to a traditional MVPD. But in reality, if a consumer has “cut the cord” to experiment with OTT video service, then that consumer has revealed a strong preference for Internet services. Presumably, that consumer was already subscribing to a very high-end broadband package before “cutting the cord,” and thus has little chance to reduce his consumption. Setting aside the faulty rationale for even modeling such an effect, Comcast’s economists use an inflated value to capture the purported decrease in Comcast’s monthly broadband profit margins (equal to {{$14.49 per subscriber per month}}).443 Given the speculative nature of their empirical estimate (which was based on interviews with Comcast personnel about a “hypothetical scenario”),444 the best estimate of Comcast’s decrease in broadband revenues associated with successful foreclosure is zero.
214. For the foregoing reasons, the Commission should not credit Drs. Katz’s and
Israel’s online video study when predicting the anticompetitive effects of the proposed
transaction.
IV. THE COMMISSION SHOULD CRAFT REMEDIES TO PROTECT COMPETITION IN THE VIDEO
MARKETPLACE
215. For the reasons articulated above, the Commission should deny the Applicants’
request to transfer NBCU licenses from GE to Comcast. In the alternative, should the
Commission support the proposed transaction, it should apply the following conditions to its
approval.
442. Id. ¶ 109. 443. Id. ¶ 120. 444. Id. ¶ 118.
-141-
NAVIGANT ECONOMICS
5102853
A. Remedies to Address Comcast’s Refusing to Supply Affiliated Programming to Rival Distributors on Reasonable Terms
216. To ensure that the program access provisions are not evaded through terrestrial
delivery of its affiliated programming, the Commission should compel Comcast to make its
RSNs, Versus, and NBC-affiliated networks available to all downstream MVPD rivals, including
DBS rivals, at reasonable and non-discriminatory rates regardless of the method of delivery. In
other words, the Commission should close off the terrestrial loophole as a condition of merger
approval. And to remove any short-term benefits from strategically pricing its affiliate network,
Comcast should be compelled to make the affiliate network available to the downstream rival
upon the initiation of a program access filing at the previously agreed-upon rates. With respect to
the ten NBCU O&O affiliates, Comcast should be compelled to enter binding commercial
arbitration for disputes over retransmission consent; rival MVPDs should be allowed to carry the
NBC affiliate while the dispute is being arbitrated. Finally, Comcast should be compelled to sell
its affiliate networks to MVPDs on an a-la-carte basis—that is, Comcast should be barred from
tying its marquee networks (an NBC affiliate, an RSN, or Versus) to its lesser programming.
217. A major weakness of the program access rules, however, is that they are satisfied
by charging one’s affiliated cable distributor the same, inflated wholesale price for the network.
Depending on the must-have nature of the affiliated network, a vertically integrated cable
operator may choose a price for its affiliated cable network in excess of the price that would be
charged by an independent cable network provider or what economists call the “independent
monopoly price” (IMP). The goal of regulation should be to steer the wholesale price of the
affiliated network back to the IMP, as wholesale prices in excess of the IMP constitute a
reduction in social welfare.
-142-
NAVIGANT ECONOMICS
5102853
218. The reason for the divergence between the wholesale price and the IMP for a
vertically integrated cable operator is simple: The vertically integrated cable operator realizes a
gain from the elevated costs incurred by its downstream MVPD rivals. It is always better to
compete against a high-cost rival than a low-cost rival. And by controlling the access price of a
key input for a rival, the vertically integrated cable operator can directly raise its rival’s costs.
The problem confronting regulators (or in the case of program access disputes, arbitrators) is to
distinguish the elevated offer price from the IMP. Program access disputes often focus on the
prices charged for similarly situated networks. But to the extent that those networks are also
owned by vertically integrated cable operators, the benchmark is itself tainted.
1. The Prospect of A La Carte Pricing
219. One approach to adjudicating a program-access dispute is to estimate the IMP of
the cable network based on the classic Lerner index, which sets the price-cost margin equal to
the inverse of the own-price elasticity of demand. The regulator would then impose a restraint on
the vertically integrated cable operator not to exceed the estimated IMP. The problem with this
approach, however, is that econometric estimation of the demand curve imposes significant data
requirements on the analyst. Although there are non-econometric methods, these also require
granular data on the network’s margins. At the end of the day, the dispute would likely turn into
a “battle of the experts,” with the arbitrator having to decide which economic model best
explained the data.
220. An alternative approach is to allow Comcast’s customers to opt out of the retail
bundle that includes the affiliated network in question at a price equal to the bundled retail price
less the wholesale price of the network. To implement this approach, Comcast should be barred
from selling its affiliated networks on a bundled basis to MVPDs—that is, Comcast could not tie
-143-
NAVIGANT ECONOMICS
5102853
its marquee programming (an NBC affiliate or an RSN or Versus) to its lesser programming.
This a-la-carte pricing rule would apply, at a minimum, to any Comcast RSN, Versus, and the
ten NBCU O&O local affiliates. To make matters concrete, assume that Comcast charges $40
per subscriber per month for an expanded digital basic package that includes Versus. Assume
further that, after transferring NBC sports programming to Versus, Comcast seeks $6 per
subscriber per month from its MVPD rivals for a license to Versus. At that pricing pair, a
Comcast subscriber, under this regime, could elect to subscribe to the expanded digital basic
package for $34 per month (equal to $40 less $6) by opting out of Versus. Comcast subscribers
that do not opt out of Versus would continue to pay $40 per month. (Comcast would still expose
itself to a program access complaint brought by a rival MVPD under this alternative.)
221. How would this regime affect Comcast’s incentives to price Versus to its
downstream rivals? Suppose that the IMP of Versus (with NBC’s sports programming) is $3 per
month—that is, an independent owner of Versus would chose $3 per month as the profit-
maximizing price—but that Comcast seeks an additional $3 per month to disadvantage its
downstream MVPD rivals. At a $6 wholesale price for Versus, Comcast would expose itself to a
certain level of opt-outs for all Comcast customers who value Versus at less than $6 per month.
In particular, for each Comcast subscriber that opted out of Versus, Comcast would lose $6 in
revenues on net; the fact that it no longer pays its upstream cable network $6 for the opt-out
customers is an internal transfer that can be ignored.
222. The elegance of this approach is that Comcast’s exposure to revenue loss at the
retail level is an increasing function of the wholesale price it seeks for Versus; the higher the
wholesale price it seeks for its affiliate networks, the higher the opt-outs. In the following
section, I demonstrate that, while this remedy does not ensure that Comcast selects the IMP for
-144-
NAVIGANT ECONOMICS
5102853
its affiliated networks, it does temper Comcast’s unbounded incentive to raise its wholesale
price, as it has with CSN Chicago (a near doubling of the wholesale price) and with its RSNs in
Portland and in Philadelphia (effectively an infinite wholesale price). Comcast might argue that
even if it were to charge the IMP of $3 month, it would still expose itself to opt-outs and the
associated lost revenue under this regulatory regime. But this exposure can be thought of as a
fresh cost of vertical integration—a cost that counteracts a strategic benefit of vertical integration
that Comcast has until now exploited to impair MVPD competition.445 If Comcast wants to avoid
this exposure, it can sell off its Versus network and free itself of the a-la-carte restriction.
2. Economic Proof of the Benefits of the A La Carte Restraint
223. An independent network provider maximizes the following profit function:
,w c Q
(1.3)
where w is the wholesale price, c is the marginal cost, and Q is the number of video subscribers
who belong to an MVPD (including Comcast) that purchases the network. The independent
network will choose w by setting the derivative of (1.3) equal to zero:
0Q
w c Qw
(1.4)
Rearranging terms and multiplying both sides by w/Q yields
1
,
Q ww c w
w Q
w c w
w c
w
(1.5)
445. As part of the a-la-carte restraint, Comcast should be barred from imposing restrictions that prevent rival
MVPDs from allowing the rival’s MVPD customers from opting out of Comcast’s affiliated networks under a similar a-la-carte option.
-145-
NAVIGANT ECONOMICS
5102853
where is the own-price elasticity of demand for the network. Define the w that solves (1.5) as
the independent monopoly price, or .IMPw
224. Now assume that Comcast acquires the network. With no restraints on its pricing,
Comcast chooses a wholesale price for its affiliate network to maximize the following profit
function:
Mq w c Q (1.6)
where M is the margin on its video subscribers and q is the number of Comcast video
subscribers. Comcast will choose w by setting the derivative of (0.4) equal to zero:
0
Mq w c Q q QM w c Q
w w w
(1.7)
Define the w that solves (1.7) as *.w The derivative of Comcast’s video subscribers with respect
to an increase in the wholesale price, ,qw
is positive because Comcast can impair its MVPD
rivals’ efficiency by raising their costs. Comparing (1.7) with (1.4) reveals that the added
positive term serves to increase the wholesale price relative to .IMPw In other words, vertical
affiliation with no restraints leads to wholesale price in excess of IMPw —that is, * .IMPw w
225. Now suppose that, under an a-la-carte restraint, Comcast were required to allow
its video subscribers to opt out of its affiliated network and receive a discount equal to the
wholesale price, w. Comcast’s profits under this regime would be:
(1 ) ( )
,
Mq M w q w c Q
Mq wq w c Q
(1.8)
-146-
NAVIGANT ECONOMICS
5102853
where [ ]w is the fraction of Comcast customers who opt out of the network at a wholesale price
of w, and '[ ] 0w . Under this a-la-carte restraint, Comcast will now choose w by setting the
derivative of (1.8) equal to zero:
0
0
Mq wq w c Q q q QM q w w c Q
w w w wq q Q
M q wq w w c Qw w w w
(1.9)
Define the w that solves (1.9) as **.w Comparing (1.9) with (1.7) reveals three new terms. The
first new term, q , is negative. The second new term is negative because the number of
Comcast’s customers who opt out of the affiliated network, , is increasing in w. The third new
term is also negative because the number of Comcast subscribers, q, is increasing in w.
226. Equation (1.9) reveals that, under the a-la-carte restraint, Comcast’s incentives to
raise w are diminished relative to no restraint—that is, ** *.w w Stated differently, Comcast’s
incentives to raise w above IMPw are diminished when any number of its customers may opt out
of its affiliated network for a savings of w. Indeed, under the a-la-carte restraint, Comcast will
choose IMPw whenever
q q
M q wq ww w w
(1.10)
227. Finally, Comcast can evade the a-la-carte restraint by divesting itself of the
network, in which case it will earn profits of
,IMPM w q (1.11)
as Comcast would pay wIMP to the independent network.
-147-
NAVIGANT ECONOMICS
5102853
B. Remedies to Address Comcast’s Denying Access to Independent Programming Networks
228. Under the current program-carriage adjudication process, a complaining
independent network must prove that Comcast has discriminated against its network for reasons
relating to affiliation, and that as a result of the discrimination, the independent network has been
“unreasonably restrained” in its ability to compete.446 To prove discrimination, the independent
network must show that it is “substantially similar” to the affiliated network with respect to
viewers and advertisers;447 it must also show that the affiliated cable operator gives preferential
treatment to its affiliated network in terms of tiering (for example, digital basic versus sports
tier), pricing, or some other condition of carriage. The complaining network must make a prima
facie case of discrimination to the Media Bureau; if convinced of the merits, the Media Bureau
then refers the case to an administrative law judge.
229. To date, there have been four program-carriage complaints filed against Comcast:
NFL Network, WealthTV, MASN, and Tennis Channel.448 The NFL and MASN cases have
settled; Comcast now carries NFL Network on its Digital Starter tier nationwide, and Comcast
now carries MASN in Harrisburg, Lykens, Lebanon, Hershey, Millersburg and Elizabethtown
Pennsylvania. In October 2009, Judge Richard Sippel ruled that WealthTV had failed to prove
that any of the defendants, including Comcast, (1) engaged in discrimination in the selection,
terms or conditions of carriage on the basis of WealthTV’s non-affiliation, and (2) any of the
defendants had unreasonably restrained WealthTV’s ability to compete fairly. The Tennis
Channel’s case is pending before the Media Bureau.
446. 47 U.S.C. § 536(a)(3). See also Implementation of Sections 12 and 19 of the Cable Television Consumer
Protection Act of 1992, Second Report and Order, 9 FCC Rcd 2642, 2643, ¶ 2 (1993) [hereinafter 1993 Program Carriage Order].
447. 1993 Program Carriage Order, 9 FCC Rcd at 2645 n.6. 448. I currently serve as Tennis Channel’s expert in that matter. I also served as the expert for MASN and for
NFL.
-148-
NAVIGANT ECONOMICS
5102853
230. By increasing its sports portfolio, the proposed transaction would increase
Comcast’s incentive and ability to extend and increase its market power in the supply of cable
programming, particularly sports programming. To discourage Comcast from discriminating in
its carriage decisions on the basis of affiliation, the Commission should consider the following
refinements to its current program-carriage adjudication process:
231. First, upon the Media Bureau’s referral of the case to an administrative law judge,
Comcast should be compelled to carry the independent network on interim terms selected by the
Media Bureau. In contrast, compelling carriage at the initiation of a program-carriage complaint
could encourage opportunistic behavior among networks; the Media Bureau could be used as a
filter to weed out meritless cases. With this remedy, Comcast could not further advantage its
affiliated networks during the period between the Media Bureau’s referral of the case and the
administrative law judge’s ultimate decision.
232. Second, the Commission should provide guidance on what constitutes
discrimination in the context of a carriage complaint.449 In prior proceedings, Comcast has
argued as an efficiency defense that carriage of an independent sports network is more profitable
on its sports tier than on its digital standard tier.450 To an economist, however the application of a
different (or lesser) standard to affiliated networks (relative to unaffiliated networks) constitutes
discrimination on the basis of affiliation. If Comcast were required to prove that it applied the
same standard to its affiliated network as it applied to the independent network, then certain
efficiency defenses would be denied. In general, the Commission should offer types of efficiency
449. See, e.g., Enforcement Bureau Comments in Herring Broadcasting, Inc. d/b/a/ WealthTV, Complainant
v. Time Warner Cable, Inc. et al. MB Dkt. No. 08-214, July 8, 2009, ¶ 9 (“Although Section 76.1301(c) was adopted in 1993, there is a scarcity of guidance and case law on the specific subject of program carriage discrimination.”).
450. See, e.g., Answer of Comcast Cable Communications LLC, Tennis Channel, Inc. v. Comcast Cable Communications, LLC, File No. CSR-8258-P, Feb. 11, 2010
-149-
NAVIGANT ECONOMICS
5102853
defenses that Comcast may or may not use to rebut a claim of discrimination on the basis of
affiliation.
233. Third, the Commission should provide guidance on the meaning of the term
“unreasonable.” To an economist, conduct that impairs a network’s ability to constrain the
pricing of Comcast’s affiliated network (in terms of license fee or advertising rate) is
appropriately considered “unreasonable.”451 Given the large economies of scale in the supply of
national or regional programming,452 discriminatory conduct that forecloses an independent
network from 20 percent of its programming footprint should be presumed to “unreasonably
restrain” the network’s ability to compete; this threshold is analogous to the presumption of
anticompetitive effects in antitrust jurisprudence triggered by a 20 percent foreclosure share.453
Comcast has argued in several proceedings that its roughly 24 percent national MVPD share was
too small to constitute competitive injury to the independent network; if Comcast is correct, then
no programming carriage complaint brought by a national network could succeed. In cases where
the foreclosure share is less than 20 percent, an independent network could still satisfy the
“unreasonably restrained” prong so long as it could demonstrate that its inability to serve the
disputed territory has significantly impaired its ability to compete for advertisers or programming
rights. For example, coverage gaps created by the discriminatory conduct that amounted to less
than 20 percent of a network’s footprint could significantly impair the network to compete for
advertising and programming, depending on the nature of competition.
451. Steven C. Salop & David T. Scheffman, Raising Rivals’ Costs, 73 AMERICAN ECONOMIC REVIEW 270
(1983). 452. See, e.g, Keith Brown & Roberto Cavazos, Why Is This Show So Dumb? Advertising Revenue and
Program Content of Network Television, 27 REVIEW OF INDUSTRIAL ORGANIZATION 17-34 (2005) (showing a 0.39 percent reduction in the advertising price per viewer given a one percent decrease in share).
453. See PHILLIP AREEDA, IX ANTITRUST LAW 375, 377, 387 (Aspen 1991) (indicating that 20 percent foreclosure is presumptively anticompetitive); See also HERBERT HOVENKAMP, XI ANTITRUST LAW 152, 160 (indicating that 20 percent foreclosure and an HHI of 1800 is presumptively anticompetitive).
-150-
NAVIGANT ECONOMICS
5102853
234. Fourth, given the mounting evidence of Comcast’s successful extraction of equity
from upstart networks,454 independent networks that are negotiating with the combined company
for carriage should be encouraged to report to the Enforcement Bureau any demands by Comcast
for equity as a condition of carriage. Reports of equity demands should be swiftly met with fines,
followed by automatic carriage of the complaining network while the carriage negotiations
proceed under the auspices of the Enforcement Bureau. This carriage requirement would not
require the Media Bureau to determine whether the complaining network made a prima facie
case. It bears noting that this requirement, while seemingly blunt, only seeks to enforce a policy
that is currently mandated by the Cable Act, but is not policed effectively by the Commission.
235. Fifth, upon an administrative law judge’s decision that Comcast-NBCU has
discriminated in favor of its affiliated network and that, as a result, the independent network has
been unreasonably restrained in its ability to compete, Comcast-NBCU should be compelled to
pay damages to the independent network in the amount equal to the forgone licensing revenues
since the discriminatory conduct began with interest. I understand that under the current regime,
the independent network’s right to recover forgone licensing revenues has not been affirmed.
236. Finally, as an alternative to the current program-carriage adjudication process, the
Commission should consider employing a baseball-style arbitration process similar to the
method used to adjudicate program-access complaints. Applied here, the independent network
and Comcast-NBCU would proffer carriage proposals, including the tier and the price per
subscriber, and an arbitrator would decide which of the two proposals was more appropriate. In
addition to expediting the process, the baseball-style rules would likely induce Comcast-NBCU
to make a more reasonable carriage offer from the onset of the dispute.
454. See part II.B.1, infra.
-151-
NAVIGANT ECONOMICS
5102853
C. Remedies to Address Comcast’s Tying of Affiliated Internet Content to the Purchase of Cable Television Service
237. The evidence presented above suggests that a primary motivation of the proposed
transaction is to extend Comcast’s market power into online content and to impair the ability of
OTT providers to compete for Comcast’s cable video subscribers. The Commission should
consider the following remedies to address this potential abuse of market power:
238. First, Comcast-NBCU should be barred from tying access to online content on its
Fancast portal (or to an NBC portal) to the purchase of a Comcast-NBCU cable video
subscription or a Comcast-NBCU cable modem subscription. Online users who access the
Internet via an alternative broadband access provider should be permitted to purchase access to
Fancast Xfinity TV content on a standalone basis. Access to premium cable content (such as
HBO) that is replicated on Fancast would require verification of the user’s subscription to that
content, presumably from a different MVPD. To steer online users to its bundle (Fancast, cable
television, and cable modem service), however, Comcast-NBCU could impose a “penalty price”
for Fancast when purchased on a standalone basis by subscribers of rival broadband access
providers. Similar to the a-la-carte restraint suggested above, the Commission could compel
Comcast to offer its MVPD subscribers a rebate from their cable television-cable modem bundle
equal to the retail price for standalone Fancast for those customers who opt out of Fancast.
239. Second, Comcast should be barred from conditioning carriage on an independent
network’s agreement not to replicate video programming on the network’s online portal.
Similarly, Comcast-NBCU should be barred from conditioning carriage on an independent
network’s agreement not to license its programming to an OTT video provider.455 The ability of
Comcast to do so shrinks the online portfolio of content, thereby reducing the private returns to
455. Comcast should also be barred from offering a differential pricing scheme that induces an independent
network not to (a) post its content online or (b) contract with an OTT video provider.
-152-
NAVIGANT ECONOMICS
5102853
investing in broadband access. Replication of a sports event by an independent sports network on
its portal 24 hours after the event originally airs does not eviscerate the value of airing the event
in real-time on Comcast’s cable system; it does create value, however, for the Internet ecosystem
generally, which Comcast perceives to be a long-run threat to its cable television network.
240. Third, within one year of the acquisition, the combined company should be
compelled to divest NBCU’s partial ownership in Hulu.com. Given the pivotal role that
Hulu.com plays as an aggregator of network television programming on the Internet, and given
the strong likelihood that Comcast would restrict access to Hulu.com, vertical integration here is
simply unwise. Rather than placing (and then policing) access restrictions on Hulu.com for OTT
providers, a better course is divestiture. Obvious candidates for acquiring NBCU’s shares in
Hulu.com are the other owners of Hulu.com: Fox (News Corp) and ABC Networks (Disney).
Free of any MVPD ownership, if Hulu.com decides to block access to OTT providers on a
going-forward basis, then the reasons would be unrelated to impairing competition in the supply
of MVPD services.
241. Fourth, the Commission should apply the program access protections to OTT
video providers, and it should extend those protections in the event that Comcast-NBCU’s
affiliated programming is ported or replicated online. For example, the program access
provisions currently prevent Comcast from denying its MVPD rivals access to Versus so long as
Versus is delivered via satellite. (The weaknesses of the program access provisions and a
potential cure are discussed above.) To evade those restrictions, the new entity could transfer
certain must-have content from Versus to the Internet. To make matters concrete, if Comcast
were to broadcast Sunday Night Football games (currently aired on NBC) on Versus, and if
Comcast were to move those games to its Versus Website, then rival MVPDs would not have
-153-
NAVIGANT ECONOMICS
5102853
access to that must-have programming. Extension of the program access provisions to the
Internet would ensure that rival MVPDs (and their customers) would have access to (transferred)
online content so long as the rival MVPD had secured a license to the cable television feed of an
affiliated network. As an alternative to extending the program access restrictions to the combined
company’s Internet properties, the Commission could simply prevent the new entity from
transferring NBC’s affiliated programming to either its affiliated cable networks or to its
affiliated online portals.
242. Fifth, Comcast-NBCU should be barred from tying the purchase of the new
entity’s cable television service to its set-top box. Comcast currently allows its cable television
customers to access Fancast online, but only from a computer. As described above, Comcast
recognizes the threat that Internet access via the television poses to its cable television system.
Because Comcast cannot be counted on to integrate features such as Wi-Fi access into its set-top
box,456 Comcast should be compelled to permit its cable television customers to purchase a set-
top box from an independent provider.
CONCLUSION
243. For the foregoing reasons, the proposed transaction presents significant antitrust
concerns. In the absence of any merger-specific conditions, the transaction would likely (a)
impair competition in several MVPD markets, (b) weaken independent networks that compete
with Comcast’s or NBCU’s affiliated networks, (c) retard the development of online video, and
(d) undermine the incentive of rival broadband Internet access providers to invest in network
infrastructure. Given its anticompetitive effects, the Commission should deny the proposed
456. In contrast, Verizon FiOS, an entrant in the MVPD industry, allows its television customers to access the
Internet via their television by downloading “TV Widgets.” See Download TV Widgets, available at http://www22.verizon.com/Residential/aboutFiOS/Overview.htm#widgets.
-154-
NAVIGANT ECONOMICS
5102853
transaction. In the alternative, should the Commission approve the joint venture, it should
condition its approval on the remedies advocated here.
* * *
I declare under penalty of perjury that, to the best of my knowledge and belief, the foregoing is true and correct. Executed on June 21, 2010.
-155-
NAVIGANT ECONOMICS
5102853
APPENDIX 1: CURRICULUM VITAE
HAL J. SINGER NAVIGANT ECONOMICS
1801 K Street, N.W., Suite 500, Washington, D.C. 20006 Phone: (202) 747-3520
EDUCATION
THE JOHNS HOPKINS UNIVERSITY, Ph.D., 1999; M.A. (Economics), 1996. TULANE UNIVERSITY, B.S. magna cum laude (Economics), 1994. Dean’s Honor Scholar (full academic scholarship). Senior Scholar Prize in Economics, 1994.
CURRENT EMPLOYMENT
NAVIGANT ECONOMICS, Washington, D.C.: Managing Director, 2010-present. GEORGETOWN UNIVERSITY, MCDONOUGH SCHOOL OF BUSINESS, Washington, D.C.: Adjunct Professor, 2010-present.
Valuing Life Settlements as a Real Option, co-authored with Joseph R. Mason, in LONGEVITY TRADING AND LIFE
SETTLEMENTS (Vishaal Bhuyan ed., John Wiley & Sons 2009). An Antitrust Analysis of the World Trade Organization’s Decision in the U.S.-Mexico Arbitration on Telecommunications Services, co-authored with J. Gregory Sidak, in HANDBOOK OF TRANS-ATLANTIC ANTITRUST (Philip Marsden, ed. Edward Elgar 2006). BROADBAND IN EUROPE: HOW BRUSSELS CAN WIRE THE INFORMATION SOCIETY, co-authored with Dan Maldoom, Richard Marsden, and J. Gregory Sidak (Kluwer/Springer Press 2005). Are Vertically Integrated DSL Providers Squeezing Unaffiliated ISPs (and Should We Care)?, co-authored with Robert W. Crandall, in ACCESS PRICING: THEORY, PRACTICE AND EMPIRICAL EVIDENCE (Justus Haucap and Ralf Dewenter eds., Elsevier Press 2005).
-156-
NAVIGANT ECONOMICS
5102853
JOURNAL ARTICLES
1. The Economic Impact of Eliminating Preemption of State Consumer Protection Laws, UNIVERSITY OF PENNSYLVANIA
JOURNAL OF BUSINESS LAW (forthcoming 2010), co-authored with Joseph R. Mason and Robert B. Kulick.
2. Why the iPhone Won’t Last Forever and What the Government Should Do to Promote its Successor, 8 JOURNAL ON
TELECOMMUNICATIONS AND HIGH TECHNOLOGY LAW (2010), co-authored with Robert W. Hahn.
3. What Does an Economist Have to Say About the Calculation of Reasonable Royalties?, 14 INTELLECTUAL PROPERTY LAW
BULLETIN (2010), co-authored with Kyle Smith.
4. Evaluating Market Power with Two-Sided Demand and Preemptive Offers to Dissipate Monopoly Rent: Lessons for High-Technology Industries from the Antitrust Division’s Approval of the XM-Sirius Satellite Radio Merger, 4 JOURNAL OF COMPETITION LAW
& ECONOMICS (2008), co-authored with J. Gregory Sidak. 5. Assessing Bias in Patent Infringement Cases: A Review of International Trade Commission Decisions, 21 HARVARD JOURNAL OF
LAW AND TECHNOLOGY (2008), co-authored with Robert W. Hahn. 6. The Effect of Incumbent Bidding in Set-Aside Auctions: An Analysis of Prices in the Closed and Open Segments of FCC Auction 35,
32 TELECOMMUNICATIONS POLICY JOURNAL (2008), co-authored with Peter Cramton and Allan Ingraham. 7. A Real-Option Approach to Valuing Life Settlement Transactions, 23 JOURNAL OF FINANCIAL TRANSFORMATION (2008),
co-authored with Joseph R. Mason. 8. The Economics of Wireless Net Neutrality, 3 JOURNAL OF COMPETITION LAW AND ECONOMICS 399 (2007), co-authored
with Robert W. Hahn and Robert E Litan. 9. Vertical Foreclosure in Video Programming Markets: Implication for Cable Operators, 3 REVIEW OF NETWORK ECONOMICS
348 (2007), co-authored with J. Gregory Sidak. 10. The Unintended Consequences of Net Neutrality, 5 JOURNAL ON TELECOMMUNICATIONS AND HIGH TECH LAW 533
(2007), co-authored with Robert E. Litan. 11. Does Video Delivered Over a Telephone Network Require a Cable Franchise?, 59 FEDERAL COMMUNICATIONS LAW
JOURNAL 251 (2007), co-authored with Robert W. Crandall and J. Gregory Sidak. 12. The Competitive Effects of a Cable Television Operator’s Refusal to Carry DSL Advertising, 2 JOURNAL OF COMPETITION LAW
AND ECONOMICS 301 (2006). 13. Uberregulation without Economics: The World Trade Organization’s Decision in the U.S.-Mexico Arbitration on Telecommunications
Services, 57 FEDERAL COMMUNICATIONS LAW JOURNAL 1 (2004), co-authored with J. Gregory Sidak. 14. The Secondary Market for Life Insurance Policies: Uncovering Life Insurance’s “Hidden” Value, 6 MARQUETTE ELDER’S
ADVISOR 95 (2004), co-authored with Neil A. Doherty and Brian A. O’Dea. 15. Do Unbundling Policies Discourage CLEC Facilities-Based Investment?, 4 TOPICS IN ECONOMIC ANALYSIS AND POLICY
(2004), co-authored with Robert W. Crandall and Allan T. Ingraham. 16. Foreign Investment Restrictions as Industrial Policy, 3 CANADIAN JOURNAL OF LAW AND TECHNOLOGY 19 (2004), co-
authored with Robert W. Crandall. 17. Regulating the Secondary Market for Life Insurance Policies, 21 JOURNAL OF INSURANCE REGULATION 63 (2003), co-
authored with Neil A. Doherty. 18. Interim Pricing of Local Loop Unbundling in Ireland: Epilogue, 4 JOURNAL OF NETWORK INDUSTRIES 119 (2003), co-
authored with J. Gregory Sidak.
-157-
NAVIGANT ECONOMICS
5102853
19. The Benefits of a Secondary Market for Life Insurance, 38 REAL PROPERTY, PROBATE AND TRUST JOURNAL 449 (2003), co-authored with Neil A. Doherty.
20. The Empirical Case Against Asymmetric Regulation of Broadband Internet Access, 17 BERKELEY TECHNOLOGY LAW
JOURNAL 954 (2002), co-authored with Robert W. Crandall and J. Gregory Sidak. 21. How Can Regulators Set Nonarbitrary Interim Rates? The Case of Local Loop Unbundling in Ireland, 3 JOURNAL OF NETWORK
INDUSTRIES 273 (2002), co-authored with J. Gregory Sidak. 22. Vertical Foreclosure in Broadband Access, 49 JOURNAL OF INDUSTRIAL ECONOMICS (2001) 299, co-authored with Daniel
L. Rubinfeld. 23. Open Access to Broadband Networks: A Case Study of the AOL/Time Warner Merger, 16 BERKELEY TECHNOLOGY LAW
JOURNAL 640 (2001), co-authored with Daniel L. Rubinfeld. 24. Cable Modems and DSL: Broadband Internet Access for Residential Customers, 91 AMERICAN ECONOMICS ASSOCIATION
PAPERS AND PROCEEDINGS 302 (2001), co-authored with Jerry A. Hausman and J. Gregory Sidak. 25. Residential Demand for Broadband Telecommunications and Consumer Access to Unaffiliated Internet Content Providers, 18 YALE
JOURNAL ON REGULATION 1 (2001), co-authored with Jerry A. Hausman and J. Gregory Sidak. 26. Determining the Source of Inter-License Synergies in Two-Way Paging Networks, 18 JOURNAL OF REGULATORY ECONOMICS
59 (2000). 27. A General Framework for Competitive Analysis in the Wireless Industry, 50 HASTINGS LAW REVIEW 1639 (2000), co-
authored with J. Gregory Sidak and David Teece. 28. Capital Raising in Offshore Markets, 23 JOURNAL OF BUSINESS AND FINANCE 1181 (1999), co-authored with Ian Gray
and Reena Aggarwal.
EXPERT TESTIMONY SINCE 2005
1. T-Mobile West Corporation v. Michael M. Crow, et al., Case No: 2:08-cv-1337 PHX-NVW (D. Az.).
2. Metlife Insurance Company of Connecticut and General American Life Insurance Company v. Thomas Petracek, Minnesota Estate Service, Inc., Michael J. Antonello, and Michael J. Antonenllo & Associates, Ltd., No. 08-CV-06095 DSD-FLN (D. Minn).
3. Tennis Channel, Inc. v. Comcast Cable Communications, LLC, File No. CSR-8258 (Federal Communications Commission).
4. MEMdata, LLC, v. Intermountain Healthcare, Inc., and IHC Health Services, Inc., Civil No. 2:08-cv-190 (C.D. Utah).
5. Caroline Behrend, et al. v. Comcast Corporation, Civil Action No. 03-6604 (E.D. Pa.).
6. In the Matter of Distribution of the 2000-03 Copyright Royalty Funds, Dkt. No. 2008-2 CRB CD 2000-03 (Copyright Royalty Judges).
7. Cindy Johnson et al. v. Arizona Hospital and Health Care Association et al., Case No. 07-01292 (SRB) (D. Az.).
8. Southeast Missouri Hospital, et al. v. C.R. Bard, Inc., Case No. 1:07CV0031 TCM (E.D. Mo.).
10. Meijer, Inc. & Meijer Distribution, Inc., et al. v. Abbott Laboratories, Case No. C 07-5985 CW (N.D. Ca.). 11. NFL Enterprises LLC. v. Comcast Cable Communications, LLC, File No. CSR-7876-P (Federal Communications
Commission). 12. Medical Mutual of Ohio Inc. v. GlaxoSmithKline PLC and SmithKline Beecham Corporation, Civil Action No. 05-
396 (E.D. Pa.). 13. TCR Sports Broadcasting Holding, L.L.P., D/B/A Mid-Atlantic Sports Network v. Time Warner Cable Inc.,
American Arbitration Association, Case No. 12 494 E 000326 07 (Federal Communications Commission). 14. Coors Brewing Company v. Hipal Partners Ltd., Case No. 71 155 00358 07 (American Arbitration Association). 15. Natchitoches Parish Hosp. Serv. Dist., et al. v. Tyco Int’l, Ltd. et al., Civil Action No. 05-12024 (D. Mass.). 16. Ralph O. Stalsberg, et al. v. New York Life Insurance Company, New York Life Insurance and Annuity
Corporation, Case No. 2:07-Cv-29-Bj (D. Utah). 17. In Re Tricor Indirect Purchaser Antitrust Litigation, C.A. No. 05-360 (KAJ) (D. Del.).
COMMISSIONED WHITE PAPERS
1. The Economic Impact of Broadband Investment (prepared for Broadband for America), co-authored with Robert
Crandall (Feb. 23, 2010).
2. Why the iPhone Won’t Last Forever and What the Government Should Do to Promote Its Successor (prepared for Mobile Future), co-authored with Robert Hahn (Sept. 21, 2009).
3. The Economic Impact of Eliminating Preemption of State Consumer Protection Laws (prepared for the American Bankers’ Association), co-authored with Joseph R. Mason (Aug. 21, 2009).
4. Economic Effects of Tax Incentives for Broadband Infrastructure Deployment (prepared for the Fiber to the Home Council), co-authored with Jeffrey Eisenach and Jeffrey West (Dec. 23, 2008).
5. The Effect of Brokered Deposits and Asset Growth on the Likelihood of Failure (prepared for Morgan Stanley, Citigroup, and UBS), co-authored with Joseph Mason and Jeffrey West (Dec. 17, 2008).
6. Estimating the Benefits and Costs of M2Z’s Proposal: Reply to Wilkie’s Spectrum Auctions Are Not a Panacea (prepared for CTIA), co-authored with Robert W. Hahn, Allan T. Ingraham and J. Gregory Sidak (July 23, 2008).
7. Is Greater Price Transparency Needed in The Medical Device Industry? (prepared for Advanced Medical Technology Association), co-authored with Robert W. Hahn (Nov. 30, 2007).
8. Should the FCC Depart from More than a Decade of Market-Oriented Spectrum Policy? Reply to Skrzypacz and Wilson (prepared for CTIA), co-authored with Gerald Faulhaber and Robert W. Hahn (Jun. 18, 2007).
9. Improving Public Safety Communications: An Analysis of Alternative Approaches (prepared for the Consumer Electronics Association and the High Tech DTV Coalition), co-authored with Peter Cramton, Thomas S. Dombrowsky, Jr., Jeffrey A. Eisenach, and Allan Ingraham (Feb. 6, 2007).
-159-
NAVIGANT ECONOMICS
5102853
10. The Budgetary Impact of Eliminating the GPOs’ Safe Harbor Exemption from the Anti-Kickback Statute of the Social Security Act (prepared for the Medical Device Manufacturers Association) (Dec. 20, 2005).
11. Reply to “The Life Settlements Market: An Actuarial Perspective on Consumer Economic Value” (prepared for Coventry First), co-authored with Eric Stallard (Nov. 15, 2005).
12. The Competitive Effects of Telephone Entry into Video Markets (prepared for the Internet Innovation Alliance), co-authored with Robert W. Crandall and J. Gregory Sidak (Nov. 9, 2005).
13. How Do Consumers and the Auto Industry Respond to Changes in Exhaust Emission and Fuel Economy
Standards? A Critique of Burke, Abeles, and Chen (prepared for General Motors Corp.), co-authored with Robert W. Crandall and Allan T. Ingraham (Sept. 21, 2004).
14. Inter-City Competition for Retail Trade in North Texas: Can a TIF Generate Incremental Tax Receipts for the City of Dallas? (prepared for Harvest Partners), co-authored with Thomas G. Thibodeau and Allan T. Ingraham (July 16, 2004).
15. An Accurate Scorecard of the Telecommunications Act of 1996: Rejoinder to the Phoenix Center Study No. 7 (prepared for BellSouth), co-authored with Robert Crandall (Jan. 6, 2004).
16. Competition in Broadband Provision and Implications for Regulatory Policy (prepared for the Alcatel, British Telecom, Deutsche Telekom, Ericsson, France Telecom, Siemens, Telefónica de España, and Telecom Italia), co-authored with Dan Maldoom, Richard Marsden, and Gregory Sidak (Oct. 15, 2003).
17. The Effect of Ubiquitous Broadband Adoption on Investment, Jobs, and the U.S. Economy (prepared for Verizon), co-authored with Robert W. Crandall (Sept. 17, 2003).
18. The Deleterious Effect of Extending the Unbundling Regime on Telecommunications Investment (prepared for BellSouth), co-authored with Robert W. Crandall (July 10, 2003).
19. Letter Concerning Spectrum Auction 35 to the Honorable Michael K. Powell, Chairman, Federal Communications Commission, from Peter C. Cramton, Robert W. Crandall, Robert W. Hahn, Robert G. Harris, Jerry A. Hausman, Thomas W. Hazlett, Douglas G. Lichtman, Paul W. MacAvoy, Paul R. Milgrom, Richard Schmalensee, J. Gregory Sidak, Hal J. Singer, Vernon L. Smith, William Taylor, and David J. Teece (Aug. 16, 2002).
WORKING PAPERS
1. Class Certification in Antitrust Cases: An Economic Framework, George Mason Law Review Antitrust Symposium, co-authored with Robert Kulick (Feb. 2010).
2. Addressing the Next Wave of Internet Regulation: The Case For Equal Opportunity, Georgetown Center for Business and Policy, co authored with Robert Litan and Robert Hahn (Jan. 2010).
3. Irrational Expectations: Can a Regulator Credibly Commit to Removing an Unbundling Obligation?, AEI-Brookings Joint Center Related Publication No. 07-28, co-authored with Jeffrey A. Eisenach (Dec. 2007).
4. An Antitrust Analysis of Google's Proposed Acquisition of DoubleClick, AEI-Brookings Joint Center Related Publication
No. 07-24, co-authored with Robert W. Hahn (Sept. 2007). 5. Inter-City Competition for Retail Trade in North Texas: Can a TIF Generate Incremental Tax Receipts for the City of Dallas?, co-
authored with Thomas G. Thibodeau and Allan T. Ingraham (revise and resubmit to JOURNAL OF REAL ESTATE
RESEARCH) (July 16, 2004). 6. An Economic Assessment of the Weight-Based CAFE Standard Proposed by the National Highway Traffic Safety Administration,
co-authored with Robert W. Crandall and Allan T. Ingraham (Apr. 2004).
-160-
NAVIGANT ECONOMICS
5102853
7. How Common Are “Conflicts of Interests” in the Investment Banking Industry?, co-authored with Charles W. Calomiris (Dec.
2003). 8. Does Clustering by Incumbent MSOs Deter Entry by Overbuilders? (July 2002).
SPEAKING ENGAGEMENTS
1. 13th Annual Symposium on Antitrust, GEORGE MASON LAW REVIEW, Washington, D.C., Feb. 4, 2010.
2. Broadband Infrastructure and Net Neutrality, ADVISORY COMMITTEE TO THE CONGRESSIONAL INTERNET CAUCUS’
STATE OF THE NET, Washington, D.C., Jan. 22, 2010.
3. The Consequences of Net Neutrality Regulations, AMERICAN CONSUMER INSTITUTE CENTER FOR CITIZEN RESEARCH, Washington, D.C., Nov. 19, 2009.
5. Second Life Settlements & Longevity Summit, INSURANCE-LINKED SECURITIES & LIFE SETTLEMENTS, New York, N.Y, Sept. 30, 2009.
6. Perspectives on Investment and a National Broadband Plan, AMERICAN CONSUMER INSTITUTE, Washington, D.C., Sept. 4, 2009.
7. Markets and Regulation: How Do We Best Serve Customers?, Wireless U. Communications Policy Seminar, UNIVERSITY
OF FLORIDA PUBLIC UTILITY RESEARCH CENTER, Tampa, FL., Nov. 13, 2008.
8. The Price Of Medical Technology: Are We Getting What We Pay For? HEALTH AFFAIRS BRIEFING, Washington, D.C., Nov. 10, 2008.
9. Standard Setting and Patent Pools, LAW SEMINARS INTERNATIONAL, Arlington, VA., Oct. 3, 2008.
10. The Changing Structure of the Telecommunications Industry and the New Role of Regulation, INTERNATIONAL
TELECOMMUNICATIONS SOCIETY BIENNIAL CONFERENCE, Montreal, Canada, June 26, 2008. 11. The Debate Over Network Management: An Economic Perspective, AMERICAN ENTERPRISE INSTITUTE CENTER FOR
REGULATORY AND MARKET STUDIES, Washington, D.C., Apr. 2, 2008. 12. Merger Policy in High-Tech Industries, GEORGE MASON UNIVERSITY SCHOOL OF LAW, Washington, D.C., Feb. 1, 2008. 13. Telecommunications Symposium, U.S. DEPARTMENT OF JUSTICE ANTITRUST DIVISION, Washington, D.C., Nov. 29,
2007. 14. Wireless Practice Luncheon, FEDERAL COMMUNICATIONS BAR ASSOCIATION, Washington, D.C., Nov. 29, 2007. 15. Association for Computing Machinery’s Net Neutrality Symposium, GEORGE WASHINGTON UNIVERSITY, Washington,
D.C., Nov. 12, 2007. 16. Regulators’ AdvanceComm Summit, NEW YORK LAW SCHOOL, New York, N.Y., Oct. 14, 2007. 17. Annual Conference, CAPACITY USA 2007, New York, N.Y., Jun. 26, 2007. 18. William Pitt Debating Union, UNIVERSITY OF PITTSBURGH, SCHOOL OF ARTS & SCIENCES, Pittsburgh, PA., Feb. 23,
2007. 19. Annual Conference, WIRELESS COMMUNICATIONS ASSOCIATION INTERNATIONAL, Washington, D.C., June 27, 2006.
-161-
NAVIGANT ECONOMICS
5102853
20. Annual Conference, MEDICAL DEVICE MANUFACTURERS ASSOCIATION, Washington, D.C., June 14, 2006. 21. Annual Conference, ASSOCIATION FOR ADVANCED LIFE UNDERWRITING, Washington, D.C., May 1, 2006. 22. Entrepreneur Lecture Series, LAFAYETTE COLLEGE, Easton, PA., Nov. 14, 2005.
EDITORIALS AND MAGAZINE ARTICLES
1. Why the iPhone Won’t Last Forever and What the Government Should (or Shouldn’t) Do to Promote Its Successor, MILKEN
INSTITUTE REVIEW (2010), co-authored with Robert W. Hahn.
2. Streamlining Consumer Financial Protection, THE HILL, Oct. 13, 2009, co-authored with Joseph R. Mason.
3. Is Greater Price Transparency Needed in the Medical Device Industry?, HEALTH AFFAIRS (2008), co-authored with Robert W. Hahn and Keith Klovers.
4. Foxes in the Henhouse: FCC Regulation through Merger Review, MILKEN INSTITUTE REVIEW (First Quarter 2008), co-
authored with J. Gregory Sidak. 5. Don’t Drink the CAFE Kool-Aid, WALL STREET JOURNAL, Sept. 6, 2007, at A17, co-authored with Robert W.
Crandall. 6. The Knee-Jerk Reaction: Misunderstanding the XM/Sirius Merger, WASHINGTON TIMES, Aug. 24, 2007, at A19, co-
authored with J. Gregory Sidak. 7. Net Neutrality: A Radical Form of Non-Discrimination, REGULATION, Summer 2007. 8. Telecom Time Warp, WALL STREET JOURNAL, July 11, 2007, at A15, co-authored with Robert W. Crandall. 9. Earmarked Airwaves, WASHINGTON POST, June 27, 2007, at A19, co-authored with Robert W. Hahn. 10. Not Neutrality, NATIONAL POST, Mar. 29, 2007, at FP19. 11. Should ATM Fees Be Regulated?, NATIONAL POST, Mar. 8, 2007, at FP17, co-authored with Robert W. Crandall. 12. Life Support for ISPs, REGULATION, Fall 2005, co-authored with Robert W. Crandall. 13. No Two-Tier Telecommunications, NATIONAL POST, Mar. 7, 2003, at FP15, co-authored with Robert W. Crandall.
MEMBERSHIPS American Economics Association American Bar Association Section of Antitrust Law
REVIEWER JOURNAL OF COMPETITION LAW AND ECONOMICS JOURNAL OF RISK MANAGEMENT AND INSURANCE REVIEW JOURNAL OF REGULATORY ECONOMICS
-162-
NAVIGANT ECONOMICS
5102853
MANAGERIAL AND DECISION ECONOMICS TELECOMMUNICATIONS POLICY