Summary
When producers of TV channels, such as Time Warner, and distributors of those channels, such as AT&T, are merged, what are the impacts on consumers, rival producers and rival distributors? Because ‘vertical integration’ of this kind can have both efficiency and anti-competitive effects, competition authorities and courts evaluating a prospective merger need to assess both the potential benefits and the potential harms to determine the overall welfare effects.
A team of economists has developed a new framework for doing just that, and used it to quantify welfare effects in the context of high-value sports content in the US cable and satellite TV industry. To conduct the analysis, the researchers first construct an extensive dataset on the US multichannel TV industry for the years 2000 to 2010.
They then design a structural model of the industry that captures consumer viewership and subscription decisions, distributor pricing and carriage decisions, and bargaining between distributors and channel providers – in this case, so-called ‘regional sports networks’ (RSNs).
The results highlight the importance of ‘program access rules’ in determining the effects of vertical integration in this industry. Such rules ensure that non-integrated rival distributors have access to integrated content: if they are effectively enforced, the overall welfare gains are positive. But failure to enforce program access rules effectively for integrated RSNs leads to consumer and total welfare losses.
The researchers find that integration of a single RSN with effective program access rules would reduce average cable prices by 1.2% per subscriber per month in markets served by the RSN and increase overall carriage of the RSN by 9.4%. This would increase total welfare per household by an amount equal to approximately 17% of the average consumer’s willingness to pay for a single RSN.
But failure to enforce program access rules would lead to a reduction in both consumer and total welfare of 1-2% of the average consumer’s willingness to pay for a single RSN. The loss would be significantly larger in cases where rival distributors are completely denied access to integrated content – what is known as ‘input foreclosure’.
Aside from demonstrating a method to measure both pro- and anti-competitive effects of vertical mergers, one of the most important finding of this study is that both of these effects are strongly present. Divisions of integrated firms internalize to a significant, although incomplete, degree the effects they have on other divisions within the firm, both when integrated distributors make pricing and carriage decisions and when integrated channels consider supplying rivals of their downstream divisions.
This does not preclude the possibility that in other specific settings, either pro- or anti-competitive effects might be absent or more limited. The results of this research highlight the fact that the magnitude of both pro- and anti-competitive effects is very market-specific. The implication is that vertical merger analysis that aims to be effective on a case-by-case basis is likely to require a highly fact-specific inquiry.
Main article
When producers of TV channels, such as Time Warner, and distributors of those channels, such as AT&T, merge, what are the potential benefits and potential harms for consumers, rival producers and rival distributors? This column reports evidence on the effects of such ‘vertical integration’ in the context of high-value regional sports programming in the United States. Analyzing a rich dataset on the US cable and satellite TV industry for the period 2000-2010, the researchers find that pro- and anti-competitive effects of vertical mergers are both strongly present in this setting. Their findings indicate that vertical merger analysis that aims to be effective is likely to require evaluation on a case-by-case basis with highly market-specific inquiries.
So-called ‘vertical’ mergers between producers of TV channels and distributors of those channels are regular – and sometimes highly contested – events, both in the United States and elsewhere in the world. The attention that such mergers have attracted is partly due to the industry’s overwhelming reach and size: over 80% of the approximately 120 million households with televisions in the United States subscribe to multichannel TV; and the average person consumes about four hours of TV per day. The patterns are similar in other countries.
The ‘blockbuster’ merger between AT&T, the largest cable TV distributor in the United States, and Time Warner, the owner of high-profile TV channels including HBO and CNN, provides a recent example. The challenge by the US Department of Justice – which was the first of a vertical merger in over 40 years – was at odds with recent practice in both the United States and Europe, where such mergers have been routinely permitted as long as the parties have agreed to conditions designed to prevent abuses induced by their newly vertical structure. The result of the highly publicized trial was called a ‘historic defeat’ for the Justice Department in the Wall Street Journal (2018). (See also Crawford et al, 2019, for a discussion of the implications of AT&T-Time Warner for vertical merger policy; and Caffarra et al, 2018, for a comparison of the treatment of vertical mergers in the United States and Europe.)
Vertical mergers can lead to both pro-competitive efficiency benefits and anti-competitive harms. Potential efficiency benefits include the elimination of ‘double marginalization’ (Spengler, 1950) and the alignment of investment incentives (Williamson, 1985; Grossman and Hart, 1986).
Potential anti-competitive harms in mergers between TV channel producers and distributors include ‘customer foreclosure’, where the integrated distributor refuses distribution of one or more TV channels that competes with its upstream channel affiliates (shifting advertising revenues to its channels); and ‘input foreclosure’, where the integrated firm denies supply of one or more integrated TV channels to a distributor that competes with its downstream distribution division (possibly causing customers to switch to the integrated distributor).
Either type of foreclosure need not be ‘complete’: instead of denying access/preventing carriage, the distribution and/or channel arms of an integrated firm can ‘partially’ foreclose rivals by negotiating for terms that are worse than they would be absent the merger. In the context of input foreclosure, this is called ‘raising rivals’ costs’ (Salop and Scheffman, 1983; Ordover et al, 1990).
Effects of vertical integration in high-value regional sports programming
Because a vertical merger can embody both efficiency and anti-competitive effects, competition authorities and courts attempting to evaluate a vertical merger need, in principle, to quantify both benefits and harms to determine the net welfare effect of the merger. Our research develops methods to do exactly that, and applies those methods to quantify the welfare effects of vertical integration in the context of high-value regional sports programming in the United States.
Our focus on the multichannel TV industry – and in particular regional sports programming – is driven by several factors. First, there is significant variation across the industry in terms of ownership of so-called ‘regional sports networks’ (RSNs) by cable and satellite distributors across both regions and years. (In the United States, cable and satellite distributors are referred to as ‘multichannel video programming distributors’.)
In addition, the industry is the subject of significant regulatory attention in addition to merger review, including the application of ‘program access rules’ (PARs) and exceptions to this rule, such as the ‘terrestrial loophole’. If PARs are effectively enforced, they ensure that non-integrated rival distributors have access to integrated content; the terrestrial loophole, by contrast, exempts certain distributors from the requirement to supply integrated content to rivals.
Finally, high-value sports content like that offered on RSNs has often been a focus of competition authority concerns about input foreclosure.
Data on the US multichannel TV industry
To conduct our analysis, we first construct an extensive dataset on the US multichannel TV industry between 2000 and 2010. We then specify and estimate a structural model of the industry that captures consumer viewership and subscription decisions, distributor pricing and carriage decisions, and bargaining between distributors and channel providers.
In a novel feature of our approach, we estimate, rather than impose, the degree to which firms internalize the profits of integrated units both when distributors make pricing and channel carriage decisions and when channels decide to supply or foreclose rival distributors.
A critical input into measuring the welfare effects of foreclosure are estimates of the change in distributor profits from the addition or removal of an RSN from any of its programming bundles. We use the relationship between distributors’ market shares and channel carriage, as well as observed viewership patterns and negotiated affiliate fees, to infer the values that consumers place on different channels (estimating different patterns for sports versus non-sports channels).
The pro-competitive effects of vertical integration are largely informed by the degree to which RSN carriage is higher for integrated distributors than would be implied by the RSN’s profitability to the distributor alone. Anti-competitive foreclosure effects are primarily informed by lower RSN supply to downstream rivals of integrated RSNs in so-called ‘loophole’ markets (Philadelphia and San Diego).
Our results indicate that integrated distributors internalize 79 cents of each dollar of profit realized by its affiliated upstream RSN when making pricing and carriage decisions. We also find that integrated RSNs fully (and perhaps more than fully) take into account the benefits reaped by their affiliated downstream distributor when a rival distributor is denied access to the RSN’s programming.
To assess the overall welfare effects of vertical integration in this market, we use our estimated parameters and structural model of the US multichannel TV industry to simulate vertical mergers and divestitures for the 26 RSNs that were active in 2007.
When doing so, we consider integration scenarios both when PARs are effectively enforced and when they are not. When PARs are perfectly enforced, our counterfactual simulations capture only the pro-competitive effects of integration from improved internalization of pricing and carriage decisions within the integrated firm.
When PARs are not enforced, our simulations allow as well for integrated (typically cable) distributors to engage in foreclosure, denying access to or charging higher prices for their owned RSN to non-integrated rival (typically satellite) distributors.
The importance of program access rules
Our three main results highlight the importance of PARs in determining the effects of vertical integration in the US multichannel TV industry.
First, in counterfactual simulations that enforce them, we find that integration of a single RSN with effective PARs in place would reduce average cable prices by 1.2% per subscriber per month in markets served by the RSN and increase overall carriage of the RSN by 9.4%, increasing total welfare per household by an amount equal to approximately 17% of the average consumer’s willingness to pay for a single RSN. These benefits arise due to both lower cable prices (through the reduction of double marginalization) and greater carriage of the RSN.
Second, when PARs are not enforced, at the estimated lower bound of our rival foreclosure parameter, we find that rival distributors would be denied access to an integrated RSN in four of out of 26 cases. For the other 22 cases, rival distributors continue to have access, but pay on average 18% higher affiliate fees (‘raising rivals’ costs’).
Failure to enforce PARs leads to a reduction, relative to the case with effective enforcement, in both consumer and total welfare of 1-2% of the average consumer’s willingness to pay for a single RSN, with the loss significantly larger in cases in which the rival distributors are completely denied access.
Our analysis reveals four factors that increase the likelihood of complete foreclosure:
- Fewer non-subscribers and lower ad rates (as these lessen the market expansion benefits of serving rival distributors).
- A greater difference in margins between the integrated firm and rival distributors (as this increases the integrated firm’s benefits of shifting consumers from rival distributors to its integrated distribution division).
- A higher cross-substitutability between the integrated firm and rival distributors (as the integrated distributor would benefit from more subscribers switching to it absent the RSN’s carriage on a rival).
- A larger share of households in the ‘footprint’ served by the integrated distributor (for similar reasons). For the markets we study, our results suggest that satellite distributors are excluded from carrying an RSN when the integrated cable distributor’s footprint exceeds approximately 85%.
Our third main result is that allowing for both efficiency and foreclosure incentives, the overall effect of vertical integration in the absence of effective PARs is to increase consumer and total welfare on average, resulting in statistically significant gains of approximately 15-16% of the average consumer’s willingness to pay for an RSN.
An important caveat to this conclusion is that these are average effects: we observe considerable heterogeneity across markets in their overall welfare effects from vertical integration. Indeed, we estimate negative (but statistically insignificant) welfare effects in two of the four markets in which rival distributors would be denied access. This heterogeneity highlights the importance of evaluating closely the market conditions associated with vertical mergers.
Qualifications and implications
Despite the richness of our empirical model, the effects that we document are still subject to important qualifications. Perhaps most importantly, our model and analysis does not allow for vertical integration to influence investments made by either RSNs or distributors.
As emphasized in previous research on the investment effects of vertical integration (for example, Grossman and Hart, 1986; Bolton and Whinston, 1991; Hart, 1995), the direction of these effects on consumer and aggregate surplus are ambiguous a priori, and remain an important topic for future research. We also do not consider possible foreclosure of rival channels (that is, customer foreclosure) by integrated distributors.
Aside from demonstrating a method to measure both pro- and anti-competitive effects of vertical mergers, perhaps our most important finding is that in the setting we study both of these effects are strongly present: divisions of these integrated firms internalize to a significant, although incomplete, degree the effects they have on other divisions within the firm, both when integrated distributors make pricing and carriage decisions and when integrated channels consider supplying rivals of their downstream divisions.
Of course, this finding does not preclude the possibility that in other specific settings, either pro- or anti-competitive effects might be absent or more limited. For example, with sufficiently rich contracts, welfare losses from double marginalization need not arise under arms-length contracting.
Our results also highlight the fact that the magnitude of both pro- and anti-competitive effects is very market-specific, so that vertical merger analysis that aims to be effective on a case-by-case basis is likely to require a highly fact-specific inquiry.
——————————————————————————————————————–
This article summarizes ‘The Welfare Effects of Vertical Integration in Multichannel Television Markets’ by Gregory S. Crawford, Robin S. Lee, Michael D. Whinston, & Ali Yurukoglu, published in Econometrica 86(3) in May 2018.
Gregory S. Crawford is at the University of Zurich. Robin S. Lee is at Harvard University. Michael D. Whinston is at MIT. Ali Yurukoglu is at Stanford University.
