The Obvious Flaw in the Google “Monopoly” Decision

The U.S. District Court’s ruling that Google is a “monopoly” has been treated as a kind of bombshell that will redefine the digital economy. Most reporting takes the decision as having made some authoritative determination about Google’s market power. Having read the decision, however, we are shaking our head in disbelief. The economic analysis of monopoly power in this decision is so flawed that it’s hard to understand why anyone takes it seriously.

Market definition is always the most important element to consider in determining whether a company has a monopoly. If a market is defined extremely narrowly, then any business can be characterized as a monopoly because all businesses differentiate their products to compete. Consequently, battles over market definition are common in antitrust cases. In this case, however, the DC Circuit Court’s definition of the relevant market is not too narrow or too wide, it completely misses the mark. The DC Circuit’s definition of a market…is not even a market.

The DC Circuit defines the market as “General Search Engine.” But there is no such market. Google does not sell its search services. Search capabilities are provided at zero price. This is done in order to attract users’ attention, which facilitates the sale of advertising. Google’s customers are advertisers, not individual Internet users. Google competes in the advertising market, not in some artificial, judge-made construct called “General Search Engines.”

The problem here is that the judge has no clear conception of a two-sided market. This term is used to describe an intermediary who performs a function that connects buyers and sellers. The intermediary often offers zero-priced benefits to one side to connect them with the other. As a result, Judge Mehta’s decision, all 280 pages of it, is full of misapplied notions.

Its discussion of how antitrust law defines markets is based on a 1962 case concerning the sale of shoes. While shoe sales may be two-footed, they are not two-sided; the buyer gets a tangible object in exchange for money. The ruling refers to the “hypothetical monopolist test,” which asks whether “a company with control over a set of substitutable products could profitably raise prices on those products. If so, the products may comprise the relevant product market.” The obvious problem with the “hypothetical monopolist test” is that since Google does not charge for its search engine, it cannot use its “monopoly power” to “raise prices on those products.” Indeed, Judge Mehta openly recognizes that fact (p. 143). As he says, “None of Plaintiffs’ economics experts performed a quantitative hypothetical monopolist test. That is entirely understandable … because search is a zero-priced good to the end user..” The Judge goes on to say that “Pricing, however, is central to the advertiser-side markets.” And so it is.

So if search is not priced, maybe it is not the relevant market; and if advertising is priced, maybe it is the relevant market? Alas, the ruling fails to make this obvious connection. One gets the impression that the judge cannot bring himself to this conclusion because it would let Google off the hook. If advertising were recognized as the relevant product market, it would be impossible to hold that Google has a monopoly. “The court found [the absence of a quantitative test of monopoly] “surprising,” but went on to say that “its absence is not fatal [because] there is no legal requirement that a plaintiff supply quantitative proof to define a relevant market.” The judge tries to fill this evidentiary gap by noting that “Plaintiffs did offer proof of what they say are ‘real-world’ hypothetical monopolist inquiries conducted by Google, as the company routinely measured the effects of price increases on advertiser demand. (p. 144). Here again, the court tacitly recognizes that advertising and its prices are the relevant market, and yet, Judge Mehta blithely goes on to argue that “General Search Services” is the “Relevant Product Market.”

The court’s opinion argues that the General Search Engines (GSEs) function, which Google dominates, is not one option in a broader set of online navigation capabilities, but a singular product. Other platforms are “walled gardens,” by which he means that “their query responses are derived from structured data available only on that particular platform.” “Because a user’s search is confined to the SVP’s structured data, users cannot use an SVP to navigate beyond the platform.“ Even if we accept the premise that unpriced searches are a “market,” this argument is a very weak one. Some of these platforms, e.g., Facebook or TikTok, encompass literally a billion or more users and provide access to trillions of digital resources or services. These are more like national forests than walled gardens. But in any event the distinction the judge erects between “GSE’s” and other online platforms is totally irrelevant to the question of market definition. Neither GSEs nor “walled garden” platforms charge for their searches. Ergo, their searching functions are not a “market.” The market is defined by what you get paid for, and none of them get paid for searches. All of them get paid for advertising. As far as we can tell, there is robust competition for advertising dollars.

Consequently, all of the Judge’s discussions of Google’s attempts to gain and protect its share of what he calls “general” searches cannot be accurately portrayed as efforts to monopolize the relevant market. They are more accurately described as an attempt to compete for a larger share of users’ attention (eyeballs) by heightening the availability of its (unpriced) way of attracting users into its two-sided market. This, of course, strengthens its ability to sell ads, but does not give it a monopoly over digital advertising. SVPs, GSEs, apps, streaming services, billboards, websites – and recently, chatbots – are all striving to attract user attention. Google gets a lot more attention than most of them, but it is far from monopolizing advertising. All of the other online services also offer search for free. The actual market is for paid advertising. Google has roughly 26% of digital advertising alone, its share is declining, and search advertising constitutes roughly half of all ad spending.

The sad truth is that in this 21st century age of digitally-fueled multi-sided markets, Judge Mehta is mentally stuck in a 20th century mindset. The Brown Shoe does not fit comfortably on this case, and it’s full of holes. Mehta wants to define a zero-priced element as monopolized, when the actual market – the product or service people pay for – is advertising. The consumers at issue here are advertisers, not end users conducting searches.

For reasons we cannot understand, the idea that advertising was the relevant market was ignored by both sides! Google’s loss in this case can be partially attributed to the weakness of its own defense. Its expert witness and defense team also failed to advance a correct definition of the relevant market. Google’s defense team tried to advance an economically inept argument that the relevant market is query responses, not advertising. In other words, they correctly pointed out that people search for online resources in many different places, not just via Google search, but still misidentified the “market” (advertising, not queries) and the consumer (advertisers, not end users). One can only wonder how a company so big and wealthy, so able to afford the best experts and lawyers, muffed this case so badly.