The XM-Sirius Merger: Part One

For a first go I thought I would try something a bit controversial. We expect that the media reform movement, and I count myself part of that movement, would generally oppose mergers which increase media consolidation. As a general rule, that’s true. But the XM-Sirius satellite radio merger is a different case and raises questions about how we approach the issue of mergers generally. This is going to be a bit long (and I tend to be a bit longwinded in any case), so I shall be posting it in installments. Endnotes are at the bottom of the page. There will be a brief quiz…. No. Sorry, forgot where I was for a moment there.

Market Definition and the XM-Sirius Merger: An introduction

It may be heresy to write the words, but much of the media reform movement’s take on the proposed XM Radio and Sirius Satellite Radio merger is based on a flawed economic analysis which is likely to be self-defeating and which reflects a backward looking view of what the struggle for media reform is ultimately about.

Take, for example, the recent Senate Commerce Committee testimony of the Consumers Union’s Gene Kimmelman on the subject:

“The merging parties claim that the merger does not create a monopoly: the existence of cross-platform and intermodal competition means that all forms of distribution of audio content are interchangeable, even those that function merely as storage devices, and must be included in the market definition. They assert that national subscription radio service competes, directly and indirectly, with a variety of partial substitutes. Through this overbroad market definition, the merging parties claim that they represent two small fish in a large ocean, rather than the only two fish in a small pond. Such an overbroad definition would have disastrous consequence for consumers of satellite radio as well as both for antitrust and public interest oversight in all media markets generally. By allowing the only two companies selling a specific type of media product to merge on the basis of erroneous claims of cross-platform or intermodal competition, the fundamental basis on which all public interest regulation of broadcast media rests is destroyed.”

Dr. Mark Cooper of the Consumer Federation of America makes a similar point in testimony before the House Judiciary Committee:

“The claim that national subscription radio service competes, indirectly, with a variety of partial substitutes is suspect. The track record of intermodal competition disciplining anticompetitive abuse is poor at best. “Bank shot competition” – the claim that partial or poor substitutes that are fundamentally different than the target product – has failed to protect consumers in similar situations and the result of relying on such competition in both merger and regulatory reviews has been rising prices and stagnation.”

Both Kimmelman and Cooper insist on a strict neoclassical price theory approach to market definition, rejecting the broad market theory of partial substitutables which has been embraced by much of industry.(1) There are good reasons for rejecting broad market theory, but the neoclassical price theory approach is equally flawed.(2)

The Problems of the Neoclassical Price Theory Approach

The neoclassical price theory approach was first introduced into U.S. antitrust litigation in the 1940s.(3) It is ironic that this branch of economics came to predominate antitrust thinking precisely because it was less expansive and anti-corporate than the thinking which underlay the age of trust-busting. In short, it represented a retreat from an approach which took more than the narrowest possible economic view of competition.

The neoclassical price theory approach uses what is known as the “small but significant non-transitory increase in price” (SSNIP) test. This test constructs a “hypothetical monopolist” iteratively until it determines “the smallest group of products and smallest geographic area in relation to which sellers, if acting as a single firm (a ‘hypothetical monopolist’) that was the only seller of those products in that area, could profitably impose and sustain a significant and nontransitory price increase above levels that would likely exist in the absence of the merger.”(4)

There are several technical problems with this approach, most notably the narrow theoretical constraint imposed by neoclassical price theory’s insistence that market power is reflected solely in price, but the fundamental flaw is that it fails to take into account any empirical evidence for the actual demand elasticities and cross price elasticities of substitution of consumers or the real behavior of firms.(5) There are also other considerations of importance – structural barriers to entry and suppression of technological innovation – which the approach ignores entirely.

Markets as Social Constructs

There are, however, additional approaches in the industrial organization literature which show greater promise of providing a progressive theoretical framework for antitrust analysis. Christoph Engel’s seminal study provides a key insight from the Constructivist approach:

“…[I]n reality, a market is not simple . It is the result of a social struggle. It is socially constructed…. Constructivists stress that markets do not fall from the sky. The market participants create markets first. Markets are social institutions. They are communicative constructs that the market participants themselves create in order to make sense out of the behavior of other market participants. Only if one observes these communication actions, can one make analytical and practical statements on the boundaries of markets.” [The translations from Engel are mine.]

What follows from this approach is a strongly empirical attitude toward evaluating market definition:

“The decisive fact is not whether two products ‘objectively ought to be’ substitutes. What market participants see as substitutes is the only thing that ultimately matters.”

Engel has pertinent lessons for all of us: a progressive analysis ought to be grounded in the concrete realities of how firms actually behave and the choices consumers actually make. Hypothetical constructs little avail us in getting at the concrete consequences of real market structure. A progressive media reform movement which does not constantly take empirical counsel of how consumers behave and how its opponents conceptualize the competition and act on that conceptualization is doomed to be ineffective in fighting media concentration. This is not a concession to broad market theory’s contention that everything is virtually a partial substitute for everything else;(6) it is insistence that our analysis must be constantly empirically-based and that we make judgments about substitutability on the basis of how firms and consumers actually behave.


1. A classic exposition of broad market theory can be found in Andrew C. Hruska, “A Broad Market Approach to Antitrust Product Market Definition in Innovative Industries,” The Yale Law Journal, 102:1 (1992), 305-331.

2. There is also the odd fact that much of what satellite and terrestrial radio are competing over is market share rather than price, since satellite radio is a subscription service and terrestrial radio is free. What terrestrial broadcasters really fear is loss of advertising revenues if listeners desert them for satellite radio.

3. Viz., United States v. Columbia Steel Co., 334 U.S. 496 (1948).

4. Canada, Competition Bureau, Merger Enforcement Guidelines, 3.1.

5. The principal danger of ignoring such empirical realities is that of defining a market too narrowly; viz., Ulrich Stumpf, et al., “Methodologies of Market Definition and Market Analysis,” Final Report Study for ICP-ANACOM, 2003, 19.

6. Broad market theory privileges the views of firms about the market; the Constructivist approach looks equally at firm and consumer construction of a market



  1. Holy Wonkspeak, Batman!

    A first entry right out of the gate with footnotes and everything! Don’t you know that blog entries are supposed to be filled with ignorant ranting and undefined terms and premises? Sheesh! Logic? Substantiation? Where will this all end? It’s a slippery slope, I tells ya.

    Actually I do think your post could use a little dumbing down or at least de-jargonification in the intro. What you’re saying, I think, is that to most people, radio is radio & the fact that it comes from a satelite is no big whoop.

    To me, that makes sense. What’s different about satellite radio, to me, is that you pay for it and have to give your personal info to get it. I would never do either. Broadcast and Satellite are the same in that I can listen to both (theoretically) in my car, so in that sense they are obvious substitutes for each other. I’m not going to be listening to both at the same time, for Lord’s sake. I don’t know how that bears upon your argument.

  2. In economic terms they are substitutes because you can choose between them to receive approximately the same service. That is what substitutability is. As I’ll argue in later installments, they are also substitutables for terrestrial radio, although they carry transaction costs (subscribership) which are unlike other terrestrial stations. There is also an argument for regarding them as complementaries, but I think it is wrongheaded — another subject discussed in a later installment.

Comments are closed