What’s An Internet Monopolist? A Reply to Professor Wu

Cite this Article
Geoffrey A. Manne and Joshua D. Wright, What’s An Internet Monopolist? A Reply to Professor Wu, Truth on the Market (November 22, 2010), https://truthonthemarket.com/2010/11/22/whats-an-internet-monopolist-a-reply-to-professor-wu/

We’ve been reading with interest a bit of an blog squabble between Tim Wu and Adam Thierer ( see here and here) set off by Professor Wu’s WSJ column: “In the Grip of the New Monopolists.”  Wu’s column makes some remarkable claims, and, like Adam, we find it extremely troubling.

Wu starts off with some serious teeth-gnashing concern over “The Internet Economy”:

The Internet has long been held up as a model for what the free market is supposed to look like—competition in its purest form. So why does it look increasingly like a Monopoly board? Most of the major sectors today are controlled by one dominant company or an oligopoly. Google “owns” search; Facebook, social networking; eBay rules auctions; Apple dominates online content delivery; Amazon, retail; and so on.

There are digital Kashmirs, disputed territories that remain anyone’s game, like digital publishing. But the dominions of major firms have enjoyed surprisingly secure borders over the last five years, their core markets secure. Microsoft’s Bing, launched last year by a giant with $40 billion in cash on hand, has captured a mere 3.25% of query volume (Google retains 83%). Still, no one expects Google Buzz to seriously encroach on Facebook’s market, or, for that matter, Skype to take over from Twitter. Though the border incursions do keep dominant firms on their toes, they have largely foundered as business ventures.

What struck us about Wu’s column was that there was not even a thin veil over the “big is bad” theme of the essay.  Holding aside complicated market definition questions about the markets in which Google, Twitter, Facebook, Apple, Amazon and others upon whom Wu focuses operate — that is, the question of whether these firms are actually “monopolists”  or even “near monopolists”–a question that Adam deals with masterfully in his response (in essence: There is a serious defect in an analysis of online markets in which Amazon and eBay are asserted to be non-competitors, monopolizing distinct sectors of commerce) — the most striking feature of Wu’s essay was the presumption that market concentration of this type leads to harm.

While Wu describes network externalities as one possible reason for the presence of firms with large shares in these markets, and gives some lip service to the notion that monopolists of the 2.0 variety might provide some consumer benefits, consider the following use of language to describe market outcomes:

  • “Market power is rarely seized so much as it is surrendered up, and that surrender is born less of a deliberate decision than of going with the flow.”
  • “We wouldn’t fret over monopoly so much if it came with a term limit. If Facebook’s rule over social networking were somehow restricted to, say, 10 years—or better, ended the moment the firm lost its technical superiority—the very idea of monopoly might seem almost wholesome. The problem is that dominant firms are like congressional incumbents and African dictators: They rarely give up even when they are clearly past their prime. Facing decline, they do everything possible to stay in power. And that’s when the rest of us suffer.”

Wu’s claim is that the modern “information monopolies” will be socially harmful.  Consider the first quote above.  These firms do not earn and keep their share by satisfying consumer demand from active consumers with preferences; it is “surrendered” by forces beyond the consumers’ control.  And with the second, why would we be concerned about time limits if these concentrated markets (again, lets assume arguendo the market definitions Wu has in mind for now) were generating competitive results?

One need not read the tea leaves here. Wu cuts to the chase when he writes that “info-monopolies tend to be good-to-great in the short term and bad-to-terrible in the long term.”  Benefits come early, but “the downside shows up later, as the monopolist ages and the will to innovate is replaced by mere will to power. … The costs of the monopoly are mostly borne by entrepreneurs and innovators. Over the long run, the consequences afflict the public in more subtle ways, as what were once highly dynamic parts of the economy begin to stagnate.”

Professor Wu seems to long for the good old days of antitrust — when big was bad not only as a matter of economics, but as a matter of law.  He writes, apparently assuming the economic truth of his theory of monopoly, that:

These negative effects are why people like Theodore Roosevelt, Louis Brandeis and Thurman Arnold regarded monopoly as an evil to be destroyed by the federal courts. They took a rather literal reading of the Sherman Act, which states, “Every person who shall monopolize…shall be deemed guilty of a felony.” But today we don’t have the heart to euthanize a healthy firm like Facebook just because it’s huge and happens to know more about us than the IRS.

I do not think Professor Wu could have made it any more clear that his view is that market concentration in internet markets is a net loss proposition for consumers.  If not an explicit endorsement of “big is bad” antitrust policy, it’s at least a rosy-eyed ode to that golden era when enforcers had more “heart.”   His stance is crystal clear.  Now, we think that view is wrong.  And we were just getting ready to write a short blog post responding to Wu’s column when things got a bit more interesting.

Thierer’s excellent response makes a few of the points above, but mostly focuses on the definitional question of “what’s a monopoly.”  Adam’s main allegation is that Wu uses the term “monopoly” inappropriately by merely focusing on market structure — whereas Adam would prefer that it is used only when competitive harm accompanies the concentrated market.  That is, Adam would not define a single seller of product X who sells at the competitive price because of the threat of entry as a monopolist, but Wu would.  It is Wu’s use of a “welfare-neutral” term that gets Thierer going:

Which gets to perhaps most stunning thing about Wu’s editorial today: He never even posits a “harm” that might be coming from the rise of these “new monopolists.”  That’s not surprising, of course, since he’d be hard-pressed to make the case that the sky is falling or that consumers are somehow the victims of some horrendous plight.  Hell, most of these services don’t cost consumers a dime!  And they are constantly innovating and offering an improved consumer experience.  If this is “monopoly,” then give us more!

What Tim Wu is really doing is propagating the simplistic old saw that “Big Is Bad.”  We could argue about how big is too big, but we shouldn’t confuse that debate with Wu’s mistaken redefinition of the term “monopoly.”  He has intentionally watered down the term “monopolist” such that it now means any combination of big firms he personally doesn’t approve of in markets that he has defined far too narrowly.   That’s not a proper understanding of the term “monopoly” and it most certainly isn’t an accurate representation of the real world of exciting digital innovation and ingenuity that we live in today.

What’s odd about the entire debate is —as mentioned above — Wu’s WSJ piece seems entirely clear where he stands on the welfare issue: the presence of a dominant firm in internet markets is bad for consumers.  The column expresses a confidence in the “badness” of concentration that, in our view, just isn’t warranted by the current theoretical and empirical evidence in the industrial organization literature.  But rather than debate that (more important) point, we’ve got what we think is a bit of a sideshow on whether Thierer and Wu are talking about textbook definitions of monopoly, the legal definition, or something else.

What’s more — after Wu’s original WSJ piece ringing the alarm, or at least going to level orange, on Web 2.0 Robber Barons, his reply to Thierer takes the incredibly odd tack of offering a “corrective” to Thierer.  After all, Wu’s new post points out, all he was saying was that the presence of a dominant firm creates a monopoly, and he isn’t saying anything about welfare.  It’s all market structure — there is nothing to see here, move along.  Thierer’s claims that Wu said anything about redefining monopolist are “sowing confusion and misleading the public” and are owed to his incorrect substitution of the “legal” definition of monopoly rather than a pure economic definition.

Now, Wu knows full well that he has constructed a long meditation in the Saturday Wall Street Journal fretting about Internet monopolies.  I’m sure he can point to all kinds of qualifying words in his piece that permit plausible deniability of his apparent claim that big is, indeed, bad.  But the effect (and, presumably, the intent) of Wu’s piece is, as we have said, quite clear:  We all suffer from these monopolies and they should be destroyed.  For Wu now to claim that he’s merely pointing out that, good or bad, there are some firms that have large shares of certain online markets (as he casually defines them) is preposterous.  For him to hide behind an allegedly-economic definition of monopoly (which, by the way, requires an economic definition of the relevant market–an analysis that is wholly absent from Wu’s piece) is even more preposterous, not least for his apparent complete lack of understanding of the economic definition of monopoly.

Here is Wu’s final answer on the true, “economic” definition of monopoly:

A monopoly is any firm that has a dominant share in the market for a given good or service (legal definitions range between 40% – 70%) resulting in power over that market.   That is the beginning and the end of the definition.   There is no further requirement that the firm be evil, gigantic, have caused consumer harm, be long-lasting, or anything else.

This is wrong, or at least woefully misleading.  As every Industrial Organization and Intermediate Microeconomics textbook makes clear, other than in the literal sense (as in “a single seller in a market” — which is the definition in Carlton & Perloff’s Modern Industrial Organization), a monopoly is a firm or firms that are not price-takers–in other words, that can control prices and maintain supra-competitive returns by charging an equilibrium price above marginal cost.  As Carlton & Perloff describe it “in contrast to a price-taking competitive firm, a monopoly knows that it can set its own price and that the price chosen affects the quantity it sells.  Whenever a firm can influence the the price it receives for its product, the firm is said to have monopoly power or market power.  The terms monopoly power and market power typically are used interchangeably to mean the ability to set price above competitive levels.”   Wu’s attempt to distinguish from Thierer on the grounds that he is using a truly economic definition paints him into a bit of a corner here.  He surely does not mean the literal “single seller” definition.  But then, what is left?  If Wu is asserting the proposition that Google and EBay and Facebook all face downward sloping demand curves and thus have economic market power, i.e. the ability to price over marginal cost, well, so do nearly all firms in the economy.  On the other hand, if he’s saying that these firms have the power to control market prices and conditions (and not just their own prices), then this is indeed (by definition) a concession that the definition turns on the ability to engage in conduct that is socially harmful.  We think that Wu’s position — in the WSJ piece — is the latter.  But to the extent it is, it is a position that is incomplete from an economic perspective.

Monopoly as an economic concept — once used outside the strict “single seller” context — is not simply about the number of firms competing but, rather, in Irving Fisher’s phrase, “the absence of  competition.”  The fact that competition is extraordinarily complex means that facile definitions like “having a dominant share” or “one firm in a market” are inadequate and it is well-understood that markets with only one seller can be competitive.  Wu’s effort to bring in market power is closer, but now we’re talking about market definition and a host of other concepts that are, in fact, completely absent from Wu’s assessment.

As Adam points out, there is a lot of competition among Wu’s purported monopolists, but the failure to see this (or to point it out) is a function of Wu’s crabbed market definitions.  But on Wu’s own terms, it’s really hard to see the above-marginal-cost pricing in these markets.  From the point of view of the buyers whom Wu is using to define his markets, these monopolists are really pathetic at extracting profits, as most of them give away their products for free and/or have brought so much efficient competition that they have dramatically lowered prices overall (think Amazon).  I guess once Amazon puts all other retailers–including Wal-Mart, Target, Best Buy, Barnes & Noble, etc., etc.–out of business we’ll all suffer mightily.  But that simply isn’t going to happen (and, meanwhile, the presence of all these retailers (all with substantial online presences, as well as off) puts paid to Wu’s claims about single firm dominance.

By the way–also well-worth a read on this is the DOJ’s Section 2 Report, especially Chapter 2 on monopoly power, or Josh’s piece on the distinction between economic market power and market power in antitrust law.  In a nutshell:  market share is, at best, the beginning, not the end, of the monopoly power analysis.

But that’s all beside the real point, isn’t it?

The real debate is over whether concentration in internet industries systematically leads to reduced innovation, consumer welfare, and economic growth.  We thought perhaps We’d chime in with a post and attempt to tempt Wu to engage in that debate.  After all, the WSJ article was filled with assertions and anecdotes about the certainty of future harm following from a so-called “age of monopolists.”  That debate is where the action is.  Indeed, one of the primary changes in antitrust economics over the past 40 years has been a shift from reliance on market concentration as a predictor of competitive effects to other, more reliable measures.   For example, the changes to the most recent Horizontal Merger Guidelines.  But lets not get distracted.  It would be a great debate to have.

But it now appears that Professor Wu — at least in his blog posts — has stood down on that front.  Forget the odes to 1960 antitrust and forget the promises of economic stagnation and consumer harm to come.  The point was merely, we are told, to point out that sometimes (if one assumes a particular market definition) one can find firms with large shares in internet markets and this might be good, might be bad, or none of the above.   Well, that’s not saying much about internet markets and frankly, its not saying much of anything at all, is it?  We can all certainly agree that the ubiquity of two-sided markets, network externalities, and other features of internet markets can sometimes lead to relatively concentrated markets or the presence of a dominant firm.  We can probably also all agree that these market structures are often efficient, involve serious competition for the market, and lead to boons for consumer welfare.

We could all agree on that, right?  Maybe.  Maybe not.  But let us be frank about this: the policy preferences in Wu’s WSJ piece are not well-camouflaged and they should be of concern to those who actually care about consumer welfare arising from these important–and oft-demonized–firms.  Does anybody who read the WSJ piece really believe that the point was that Internet markets get concentrated sometimes?   We don’t.  Thierer didn’t — and he got a “correction” on the definition of monopolization for his efforts.

We’re not so interested in the corrective on the definition of monopoly.  But, Professor Wu, how about a debate on the merits of the substantive claims that concentration in Internet markets leads to consumer harm?  Or the merits of the good old days of antitrust?