Vertical Integration in Multichannel Television Markets: A Study of Regional Sports Networks

Tom Van Riper, The New Money Ball, Forbes, Apr. 9, 2012, at 1.

The National Hockey League (NHL), the National Basketball Association (NBA), and Major League Baseball (MLB) offer web-based streaming options with two caveats. First, games of a given team are not available on an á-la-carte basis; instead, consumers must purchase a bundle of all games. Second, games of local teams are often blacked out, presumably so as not to upset relations with local television distributors. See, e.g., MLB.TV website, available at http://mlb.mlb.com/mlb/subscriptions/index.jsp?product=mlbtv&affiliateId=MLBTVREDIRECT.

Multichannel video programming refers to subscription-based television services. In the US, these are provided by cable television systems, direct-broadcast satellite (DBS) providers, and wireline video operators (usually incumbent telephone service providers).

Diana Moss, Regional Sports Networks, Competition and the Consumer, in The Business of Sports 287 (Scott Rosner and Kenneth Shropshire, eds. Jones and Bartlett 2011) (“RSNs are hugely profitable, with margins estimated at 30 to 40%. . . .); Frank Ahrens, Area Baseball Network Must Form Quickly, Washington Post, Sept. 30, 2004 (estimating RSN margins at 30 to 40%); Michael Yablon, Madison Square Garden, Graham and Doddsville (Fall 2011) [“MSG’s RSNs warrant a premium valuation due to their 35% EBITDA margin and top-tier market presence. By way of comparison, Liberty Sports Group’s RSNs sold for 13x EBITDA and had 22% EBITDA margins, generated just $45M (vs. MSGs $230M) and had far less iconic content.”], available at http://sportgamma.net/2012/01/02/nugget-madison-square-garden-in-graham-doddsville.

In the FCC’s order approving Comcast’s and Time Warner’s acquisition of Adelphia, an RSN was defined as “any non-broadcast video programming service that (1) provides live or same-day distribution within a limited geographic region of sporting events of a sports team that is a member of Major League Baseball, the National Basketball Association, the National Football League, the National Hockey League, NASCAR, NCAA Division I Football, NCAA Division I Basketball and (2) in any year, carries a minimum of either 100 h of programming that meets the criteria of subheading 1, or 10% of the regular season games of at least one sports team that meets the criteria of subheading 1.” Memorandum Opinion and Order, Applications for Consent to the Assignment and/or Transfer of Control of Licenses Adelphia Communications Corporation (and subsidiaries, debtors-in-possession), Assignors to Time Warner Cable Inc. (subsidiaries), Assignees, July 21, 2006, pp. 190–91 and Appendix B [hereafter Adelphia Order].

Federal Communications Commission, Annual Assessment of the Status of Competition in the Market for the Delivery of Video Programming, MB Docket No. 07-269, Fourteenth Report (July 2012), Table C-2.

Derek Baine, RSN license fee/sub up 8.9% over the last 5 years, SNL Kagan, Aug. 30, 2011.

Joe Flint, Time Warner Cable and Lakers ready launch of L.A. sports channels, Los Angeles Times, Oct. 1, 2012 at 1, available at http://www.latimes.com/entertainment/envelope/cotown/la-et-ct-twc-lakers-launch-20121001,0,7720510.story.

SNL Kagan, TV Network Summary, RSN Networks By Affiliate Revenue Per Avg Sub/Month (2012). Vertical affiliation status was determined through a review of publicly available documentation as described below.

See, e.g., Hal J. Singer and J. Gregory Sidak, Vertical Foreclosure in Video Programming Markets: Implications for Cable Operators, Review of Network Economics, 6:372, 373 (2007); Keith Klovers, Unfit for Prime Time: Why Cable Television Regulations Cannot Perform TRINKO’s ‘Antitrust Function’, Michigan Law Review, 110:489, 510 (2011). (“Because an RSN’s audience consists entirely of a sports team’s “home” television market – Philadelphia Phillies games are only important to Philadelphia fans, and by league rule cannot be shown in other teams’ markets such as New York City or Baltimore – cable operators are also the “make-or-break” customer for RSNs.”)

Adelphia Order, pp. 134, 156–57, 160, 162.

Ibid. pp. 190–191.

For instance, a 2010 impasse between ABC and Cablevision caused viewers in the New York City metropolitan area to lose access (via cable) to a portion of WABC’s broadcast of the Academy Awards. Similar disputes have involved cable networks as well: In June 2012, Dish Network dropped AMC, which carries popular cable series such as Mad Men and Breaking Bad. AMC responded by running anti-Dish advertisements, and by allowing Dish subscribers to view a live online stream of the Breaking Bad Season 5 premiere. As of August 2012, AMC’s home pages for Mad Men and Breaking Bad both included prominent warnings to viewers that the series were “Not Available on DISH.” See John Eggerton, “WABC Back on Cablevision,” Broadcasting & Cable (March 8, 2010); see also Todd Spangler, “AMC To Provide Dish Subs Live Web Stream Of ‘Breaking Bad’ Premiere,” Multichannel News (July 12, 2012).

See Memorandum Opinion and Order, In the Matter of Applications of Comcast Corporation, General Electric Company and NBC Universal, Inc., For Consent to Assign Licenses and Transfer Control of Licensees, MB Docket No. 10-56 (released Jan. 20, 2011) [hereafter Comcast-NBCU Order]. See also Lacey Rose, Government Gives Comcast-NBC Universal Marriage Its Blessing, Forbes, Jan. 18, 2011.

Comcast-NBCU Order, Appendix B, p. 37. (“…vertical integration of NBCU’s programming and Comcast distribution assets would improve the bargaining position of the integrated firm when negotiating the sale of programming to one of Comcast’s video distribution rivals because failure to reach an agreement means that some of the rival’s subscribers will shift to Comcast, thus improving the integrated firm’s best alternative to reach an agreement relative to that of pre-transaction NBCU. As a result, the integrated firm improves its bargaining position, allowing it to extract higher prices from rival MVPDs than pre-transaction NBCU was able to when negotiating with Comcast’s distribution rivals. These higher programming prices to rivals would ultimately result in higher consumer prices for MVPD service unless efficiencies resulting from the transaction that lower the cost of the joint venture providing programming lead to offsetting reductions in consumer prices.”)

The integrated networks were Fox Movie Channel, Fox News, Fox Soccer Channel, Fox Sports en Espanol, Fuel TV, FX Network, National Geographic, Speed, Fox Business Network and Fox College Sports. Ibid. Appendix B, pp. 48–52.

See Comcast-NBCU Order at Appendix B. The FCC draws on analyses submitted by several economic experts. See Mark Israel and Michael L. Katz, “Application of the Commission Staff Model of Vertical Foreclosure to the Proposed Comcast-NBCU Transaction,” (Feb. 26, 2010). See also Kevin M. Murphy, “Economic Analysis of the Impact of the Proposed Comcast/NBCU Transaction on the Cost to MVPDs of Obtaining Access to NBCU Programming,” (June 21, 2010). See also William Rogerson, “Economic Analysis Of The Competitive Harms Of The Proposed Comcast-NBCU Transaction,” attached as Exhibit A to Comments by the American Cable Association In the Matter of Applications of Comcast Corporation, General Electric Company, and NBC Universal, Inc., to Assign and Transfer Control of FCC Licenses MB Docket No. 10-56 (June 21, 2010) (hereafter Rogerson). The FCC also employed a bargaining model to evaluate vertical foreclosure in the News Corp./DIRECTV transaction. See Memorandum Opinion and Order, In the Matter of General Motors Corporation and Hughes Electronics Corporation, Transferors, and The News Corporation Limited, Transferee, For Authority to Transfer Control, 190 FCC Rcd 473 (2004), at Appendix D.

Our theoretical framework is inspired in part by a similar test designed by Goolsbee (2007), who demonstrates that, if changes in the degree of favoritism (in terms of affiliated channel placement) can be explained by changes in market power, then the observed conduct is more likely to be the result of anticompetitive conduct (as opposed to procompetitive efficiencies). See Austan Goolsbee, Vertical Integration and the Market for Broadcast and Cable Television Programming, FCC Media Ownership Study (2007), p. 2. (“Looking at decisions of cable systems regarding what channels to carry shows that carriage rates for vertically integrated channels are higher on systems that own the given network but only in places where there is not much competition from DBS. This suggests, potentially, a problem for an efficiency based explanation for the behavior.”). Ibid. p. 29. [“For those nine (vertically integrated networks), the interaction of vertical integration with the DBS share has a significant negative coefficient. This evidence suggests, perhaps, an explanation rooted in competitive pressures rather than efficiencies.”] The FCC used Goolsbee’s test in its review of the Comcast-NBCU merger, determining that Comcast gives preferential treatment to its affiliated networks, Versus and the Golf Channel, relative to unaffiliated sports networks. See Comcast-NBCU Order, p. 70. (“The empirical analysis supports the conclusion that Comcast discriminates against unaffiliated programming in favor of its own.”)

As explained below, we cannot measure the effects of vertical integration on RSN fees in cases where the vertically integrated MVPD refuses to license the RSN to other MVPDs, effectively demanding a license fee at or above the choke price. In this sense, our analysis may understate the extent to which vertical integration influences the negotiating process.

Memorandum Opinion and Order, TCR Sports Broadcasting, LLP d/b/a Mid-Atlantic Sports Network v. Time Warner Cable Inc., FCC 10-202 (adopted Dec. 20, 2010).

Ibid. p. 21. (“Absent such documentation, we find it reasonable in this case to credit the testimony of the factual witnesses put forth by TWC on this issue.”)

See Carolinas Sports Entertainment Television to Launch Service in October, NBA.COM, 2004, available at http://www.nba.com/bobcats/news/cset_launch.html; see also C-SET is shuttered, Broadcasting & Cable, July 3, 2005, available at http://www.broadcastingcable.com/article/157695-Fast_Track.php#d9e326-142-span; see also Bobcats’ arena reportedly to be renamed Time Warner Cable Arena, Associated Press, April 7, 2008, available at http://sports.espn.go.com/nba/news/story?id=3335185.

Comcast, MASN Reach Settlement in Carriage Dispute, Sports Business Daily, Dec. 28, 2009, available at http://www.sportsbusinessdaily.com/Daily/Issues/2009/12/Issue-72/Sports-Media/Comcast-MASN-Reach-Settlement-In-Carriage-Dispute.aspx.

NFL, Comcast settle NFL Network carriage dispute, agree to 10-year deal, Associated Press, May 19, 2009, available at http://www.nfl.com/nflnetwork/story?confirm=true&id=09000d5d81065fa0&template=without-video.

In the Matter of Tennis Channel, Inc., v. Comcast Cable Communications, L.L.C., MB Dkt. No. 10-204, Memorandum Opinion and Order (rel. July 1, 2012).

Adelphia Order, p. 149.

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 Rules and Examination of Program Tying Arrangements, MB Dkt. Nos. 07-29, 07-198, Report and Order and Notice of Proposed Rulemaking, p. 39 (rel. Oct. 1, 2007).

Cecil Conley, Dish dispute leaves Kings fans in foul mood, Roseville Press Tribune, Dec. 9, 2010, available at http://rosevillept.com/detail/166790.html.

Eric Young, Comcast-Dish spat ends, games return to TV, San Francisco Business Times, Feb. 4, 2012, available at http://www.bizjournals.com/sanfrancisco/blog/2011/02/comcast-dish-spat-ends-games-return.html.

Joelle Tessler, FCC move to close program access loophole upheld, Associated Press, June 10, 2011.

Ibid.

Ibid.

Ibid.

See John Nash, The Bargaining Problem, Econometrica, 18: 155–162 (1950).

See Kenneth Binmore, Ariel Rubinstein and Asher Wolinsky, The Nash Bargaining Solution in Economic Modeling, Rand Journal of Economics, 17: 176–188 (1986); see also Avinash Dixit and Susan Skeath, Games of Strategy, 524–547 (1999).

Ibid.

See Comcast-NBCU Order in Appendix B. Under temporary foreclosure, the critical value is given by the expression below, where r is the discount rate, and c is defined as fraction of consumers that defected to the integrated firm during the foreclosure episode who subsequently return to their previous MVPD once programming is restored:

Ibid. See also Rogerson, supra.

For additional discussion of the effect of vertical integration and the diversion rate on bargaining outcomes in a more general setting, see Michael H. Riordan, Competitive Effects of Vertical Integration, in Handbook of Antitrust Economics 150 (Paolo Buccirossi, ed., MIT Press 2008). [“Vertical integration can alter bargaining power…According to the theory of bilateral bargaining, a greater cost of delay in reaching agreement weakens a party’s bargaining power, resulting in an agreement that is less favorable to the more impatient party (Rubinstein, 1982). Suppose an upstream monopolist (U) is bargaining separately with two downstream firms (D1 and D2) who compete with each other. Suppose that U makes a separate initial offer to each downstream firm. If both accept, then the downstream firms each earn some level of profit. If instead D1 were to accept and D2 reject U’s offer, then D1 would experience a gain in profit by virtue of a temporary competitive advantage over D2. If contracts are bilateral, so that terms are not contingent on acceptance by the other downstream firm, then this windfall accrues purely to D1 and does not influence continued bargaining between U and D2. Now suppose that U and D1 are integrated. If D2 were to reject U’s initial offer, then the downstream windfall accrues directly the integrated firm. Thus, the costs of delay in reaching agreement with D2 are less when U is vertically integrated, which strengthens U’s bargaining power in dealing with D2. The increased bargaining power is an effect of and possible motive for vertical integration.”] In this framework, the increase in bargaining power is also a function of the diversion rate – that is, a higher diversion rate implies a higher “downstream windfall” for D1.

Riordan (2008) also reviews the economic literature on the competitive effects of vertical integration, and concludes that various post-Chicago theories of harmful vertical integration have featured prominently in recent merger reviews and regulatory proceedings. However, the specific concern that vertical integration may confer increased bargaining leverage in negotiations with downstream distributors appears to have been addressed only in the video programming industry.

See Comcast-NBCU Order at Appendix B, pp. 13–16.

For a review of demand models in characteristics space, see Daniel Ackerberg, Lanier Benkard, Steven Berry, and Ariel Pakes, Econometric Tools for Analyzing Market Outcomes, 5 Handbook of Econometrics (Heckman and Leamer, eds., North Holland 2005). For applications of this framework to consumer demand for MVPD service, see Gregory Crawford and Ali Yurukoglu, The Welfare Effects of Bundling in Multichannel Television Markets, American Economic Review, 102: 643–685 (2012); see also Austan Goolsbee and Amil Petrin, The Consumer Gains from Direct Broadcast Satellites and the Competition with Cable Television, Econometrica, 72: 351–381 (2004).

Although diversion rates are typically applied to prices in the context of merger analysis, the same logic applies to non-price characteristics. (Note that the merger analysis literature typically refers to diversion rates as “diversion ratios.”) See Joseph Farrell and Carl Shapiro, Antitrust Evaluation of Horizontal Mergers: An Economic Alternative to Market Definition, B.E. Journal of Theoretical Economics, 1:10 (2010); see also Carl Shapiro, The 2010 Horizontal Merger Guidelines: From Hedgehog to Fox in Forty Years, Antitrust Law Journal, 77 (2010).

More sophisticated “random-coefficient” demand systems lack closed-form solutions, but also relax some of the restrictive assumptions of the logit functional form, allowing consumers to substitute among products with similar characteristics. See, e.g., Steven Berry, James Levinsohn, and Ariel Pakes, Automobile Prices in Market Equilibrium, Econometrica, 63:841–890 (1995); see also Ackerberg et al. (2005), supra. For example, in the auto market, foreclosure of BMWs in a given market would be expected to disproportionately increase the market share of other luxury cars (with smaller shares), as opposed to inducing proportional substitution towards cars with higher market shares (e.g., Hondas). But in the case of RSN foreclosure, the vertically integrated RSN is, by definition, offering a product with an identical characteristic (the RSN programming). If the absence of RSN programming is sufficient to induce a consumer to switch MVPDs, this indicates a disproportionately high taste for RSN programming, which implies disproportionate substitution towards MVPDs that continue to carry the RSN (such as the vertically integrated firm). Furthermore, consumers are, all else equal, more likely to switch to the vertically integrated firm when its downstream market share is higher, since high market shares are ultimately the result of a more favorable vector of product characteristics. (As noted above, one key characteristic in the MVPD market is the technical feasibility of switching from one provider to another.) Both Crawford and Yurukoglu (2010) and Goolsbee and Petrin (2004) have applied such frameworks to model consumer demand for MVPD service.

See Comcast-NBCU Order at pp. 159–161.

SNL Kagan, TV Network Summary, RSN Networks By Affiliate Revenue Per Avg Sub/Month (2012). We exclude Comcast/Charter Sports Southeast because it does not offer NBA, NHL, or MLB programming. We also exclude the network NewsSports, for which insufficient historical ownership records were available, and Victory Sports One, for which SNL reported zero average subscribers during its 2 years of existence.

The controlling MVPD for a given RSN is the cable or satellite operator that owns a controlling share in the network. An MVPD with an ownership share above 50% is assumed to hold a controlling interest. An MVPD is also counted as controlling an RSN if corporate documentation indicated a controlling interest.

See Tasneem Chipty, Vertical Integration, Market Foreclosure, and Consumer Welfare in the Cable Television Industry, American Economic Review, 91(3):428–453 (2001) at p. 432. (“A system owner is considered vertically integrated if it owns any portion of a program service that serves the system’s franchise area.”) Unlike Chipty (2001), we classify an RSN as affiliated with an MVPD even if the MVPD does not have a presence in the local market. As explained below, this allows us to separately identify the effect of the local downstream footprint on the vertical affiliation premium. Similar to Chipty (2001), our classification methodology “errs on the side of labeling too many operators as vertically integrated and may, if anything, underestimate the effects of ownership structure.” Ibid.

See Nielsen, Local Television Market Universe Estimates (2003–2010). See also SNL Kagan, U.S. Multichannel Operator Comparison by Market, Multichannel Video Subscribers (2009Q4–2010Q4). See also Warren Communications, Television and Cable Factbook (2003–2010).

Matching of RSNs with local markets was performed using Forbes’ Business of Baseball, Basketball, and Hockey annual reports. See generally Kurt Badenhausen, Mike Ozanian, and Christina Settimi, The Business of Baseball 2012, Forbes, March 21, 2012; Kurt Badenhausen and Mike Ozanian, The Business of Basketball, Forbes, Jan. 25, 2012; Kurt Badenhausen and Mike Ozanian, The Business of Hockey, Forbes, Nov. 30, 2011. These data were supplemented as needed for historical changes in telecast rights ownership using team websites, news articles, and press releases.

For example, Comcast SportsNet Philadelphia refused to grant DISH Network and DirecTV access to its programming prior to the FCC’s closure of the “terrestrial loophole.” See FCC to Force Cable Operators to Share Sports Channels, Wall Street Journal MarketWatch, Jan. 20, 2010, available at http://articles.marketwatch.com/2010-01-20/industries/30716517_1_sports-channels-terrestrial-loophole-hd.

Using a 5-year window, the FCC estimated statistically and economically significant price effects associated with the vertical integration of Fox programming with the DIRECTV distribution platform. In the FCC’s regression models, the independent variable of interest was the percentage of the previous 5 years that the programming was integrated with DIRECTV. See Comcast-NBCU Order at Appendix B, p. 51.

We use a restricted cubic spline to capture nonlinear effects of network age. See F.E. Harrell, Regression Modeling Strategies with Applications to Linear Models, Logistic Regression And Survival Analysis, 19–26 (Springer-Verlag 2001).

For example, suppose that there are two types of RSNs in the dataset, A and B. Suppose that independent RSN A was acquired by an MVPD in 2005, and that the license fees for RSN A were locked into a 5-year contract, such that the vertically integrated entity was unable to negotiate higher license fees with downstream distributors until 2010. Suppose further that independent RSN B was acquired by a different MVPD in 2005, and that license fees for RSN B were governed by a contract that had just expired, enabling the vertically integrated entity to re-negotiate higher license fees immediately. The coefficient estimates for each lag of Affil_it-k will reflect an average across these two contractual structures. The greater the proportion of networks that resemble RSN A, the lower will be the magnitude and significance of the estimated coefficients on the shorter lags (Affil_it, Affil_it-1, etc.).

For an overview of two-sided markets, see Jean-Charles Rochet and Jean Tirole, “Two-sided markets: a progress report,” RAND Journal of Economics, 35:645–667 (2006); see also Roberto Roson, “Two-Sided Markets: A Tentative Survey,” Review of Network Economics, 4:142–160 (2005); see also Julian Wright, “One-sided Logic in Two-sided Markets,” Review of Network Economics, 3:44–64 (2004).

The introduction of the downstream market share variable could create simultaneity bias if the econometric error term is correlated with the market share. In particular, large MVPDs may be more inclined to acquire high-quality RSNs to avoid exposure to potential bargaining breakdowns. This would allow vertically integrated entities with large downstream shares to negotiate higher license fees with downstream distribution rivals than they would otherwise, leading to a positive correlation between market share and prices. However, note that each of the regression models controls for network quality in three different ways, through the inclusion of programming costs per subscriber, advertising revenues per subscriber, and network fixed effects (which control for all time-invariant quality differences). Accordingly, the coefficients would be tainted by simultaneity bias only if network quality were changing over time in a manner (a) not captured by programming costs or advertising revenues, but (b) still correlated with downstream market share. Finally, as shown in Table 3, the average license fee across affiliated RSNs is actually lower than the average license fee across independent RSNs, which is inconsistent with the notion that MVPDs generally tend to acquire RSNs of higher quality.

We also tested the sensitivity of these results using specifications in which the programming cost and advertising revenue variables were also lagged, along with affiliation and the interaction term. The results of these specifications do not qualitatively alter our findings with respect to the effect of vertical integration and the downstream footprint. However, in these specifications, the signs and significance of the coefficients on programming cost and advertising revenue were more likely to deviate from their expected values. This suggests that parties negotiating multi-year RSN contracts take expected future programming costs and advertising revenues per subscriber into account.

In the Adelphia Order, the FCC explained the type of data that could be used to inform fair market value:

• e To determine fair market value, the arbitrator may consider any relevant evidence (and may require the parties to submit such evidence to the extent it is in their possession), including, but not limited to:

  • current or previous contracts between MVPDs and RSNs in which Comcast or Time Warner do not have an interest as well as offers made in such negotiations (which may provide evidence of either a floor or a ceiling of fair market value);

  • evidence of the relative value of such programming compared to the Covered RSN programming at issue (e.g., advertising rates, ratings);

  • contracts between MVPDs and RSNs on whose behalf Comcast or Time Warner have negotiated, made before Comcast or Time Warner acquired control of the systems swapped and acquired in the Adelphia transactions.

• See Adelphia Order, Appendix B, p. 3 (emphasis added).

Ibid. The Adelphia Order goes on to explain that the “arbitrator may not consider offers prior to the arbitration made by the MVPD and the Covered RSN for the programming at issue in determining the fair market value.”Ibid. at p. 3.

The Department of Justice has found that vertical mergers in the video programming industry are less likely to yield efficiencies from the reduction or elimination of double marginalization, because “much, if not all, of any potential double marginalization is reduced, if not completely eliminated, through the course of contract negotiations between programmers and distributors over quantity and penetration discounts, tiering requirements, and other explicit and verifiable conditions.” See Competitive Impact Statement, United States of America et al. v. Comcast Corporation, General Electric Company and NBC Universal, Inc., Case 1:11-cv-00106 (Jan. 18, 2011) pp. 29–30.