May 18, 2020 6:00 AM

Here’s What an Antitrust Case Against Google Might Look Like

Two DOJ veterans lay out a roadmap for cracking down on the company’s digital advertising juggernaut.

Over the weekend, The Wall Street Journal reported that federal and state officials are likely to file antitrust lawsuits against Google later this year, focusing on the company’s dominance of the digital advertising industry. This morning, thanks to a pair of former Obama administration officials, we got our best look yet at what the theory of the case may be. In a new paper released by the Omidyar Network, they argue that Google has harmed competition in the US, and identify 20 different ways in which it may have broken the law on its road to digital advertising domination.

Until recently, the Big Tech antitrust movement has mostly been driven by outsiders who challenged the conventional legal wisdom. The authors of the new paper, by contrast, are very much a part of the establishment: David Dinielli served as special counsel in the US Department of Justice’s antitrust division, and Fiona Scott Morton, an economics professor at the Yale School of Management, was its top economist from 2011 to 2012—during a time when the White House and Google were as chummy as can be. They’re the kind of people who could end up working on antitrust in a Joe Biden administration, should he win in November. While their paper is unconnected from any official government investigation (its timing is coincidental), it’s as good a window as we have into how current antitrust enforcers may be thinking.

Titled “Roadmap for a Digital Advertising Monopolization Case Against Google,” the paper’s big-picture argument is simple. Google might be best known for search, maps, and email, but it makes most of its money through advertising. Over the past decade or so, the company has achieved a dominant position at each step of the way between advertiser and internet user. It uses that power, the paper argues, to extract a toll from everyone who participates in that market—earning essentially free money above whatever value it creates.

The authors are careful to note that they’re not going so far as to allege specific antitrust law violations. “Without access to any confidential information, it would not be appropriate to say, ‘So-and-so violated the law,’” said Morton in an interview last week. Still, she said, “What I can see publicly gives me grave concern that there’s been a violation of the law, that’s for sure.”

Gene Kimmelman, who worked with Morton as chief counsel of the DOJ antitrust division under Obama and advised on the paper, said that some of the conduct could have innocent explanations. “But the fact that we found so many of them, and a pattern of behavior, is what is so troublesome about this,” he said.

Google rejects the idea that its ad business poses any monopoly problems. In an emailed statement, it said, “Competition is flourishing, and publishers and marketers have enormous choice . . . the price of digital advertising has fallen by more than 40% since 2010. No other medium has seen such a large drop. The result is that expenditure on advertising as a fraction of GDP has never been lower, with cost savings going directly to businesses and consumers.”

The question, however, is whether businesses and consumers would be benefiting even more in a market less dominated by Google—and whether some of those cost savings have come at the unfair expense of the publishers running the ads.

There’s a long series of steps—occurring nearly instantaneously—between a company paying to place an ad and that ad being shown to a pair of human eyeballs. Remarkably, Google dominates nearly every one of those steps. Since the findings of the state and federal investigations are still confidential, Dinielli and Morton rely on a comprehensive report by the United Kingdom’s Competition and Markets Authority last year to estimate the US digital advertising market. According to the report, the company controls 90 percent of the publisher ad server market, meaning if a publisher wants to sell ad space, 90 percent of the time it’s going through Google. It controls between 40 and 60 percent of the supply-side platform market, which is where different publishers come together to automatically bid on ads. On the other side of the auction sit demand-side platforms, where advertisers bid to place their ads. Google has 50 to 70 percent of that market.

In its statement, Google argues that the existence of many other ad-tech companies means that “publishers and marketers have enormous choice.” In particular, the company points to the meteoric rise of Amazon’s ad platform as an example of real competition. But Google doesn’t dispute the CMA’s estimates of its market share.

Antitrust law isn’t about size alone, however. Companies are allowed to be popular; it’s not a crime to build the best search engine or navigation app. The question is whether Google is using its control of the market to enrich itself and make it harder for others to compete. “It’s not achieving your market power and omnipresence and popularity through competition on the merits” that’s a problem, Morton said. Monopolization becomes illegal, she explained, “through suppressing or excluding others—using your market power to create a playing field that’s not level.”

The 20 examples of Google’s behavior that Morton and Dinielli flag fall into several broad categories. The first is simple: acquisitions. Over the years, Google has entrenched itself in each part of the ad tech chain by buying up existing companies like DoubleClick and AdMob. Acquisitions aren’t illegal on their own, but they are when the goal is to stifle competition. “Together,” the authors write, “these acquisitions reveal a sustained effort to occupy the entire ad tech stack as well as the related analytics market through mergers.” (Critics of antitrust enforcement in the US have argued that the government was derelict in allowing mergers like these to go through in the first place. The Federal Trade Commission recently announced that it would be reviewing the past 10 years of acquisitions by the biggest tech companies. Google argues that in the years since these acquisitions, competition in ads has gone up while prices have declined.)

Another major category: the way the various steps in Google’s ad system are designed. “Although an auction can be designed to drive prices to competitive levels," Morton and Dinielli write, "Google’s role in running the auctions on behalf of both buyers and sellers (including when Google itself is the seller, as it is for its Google search supply and for YouTube and its other properties) gives it the incentive and ability to bias auction prices." That means ad prices go up, compared to a competitive market, and publisher revenues go down. Ordinary people may end up paying at both ends: Higher advertising costs get passed along to consumers, while lower revenues for publishers ultimately reduce the amount and quality of news and entertainment.

Google disputes this claim, noting that it doesn’t directly set auction prices. But Morton and Dinielli aren’t saying that it does; they argue instead that, because Google both runs and participates in the ad buying process, it can’t resist the juicy opportunity to design the system to advantage its own offerings. Microsoft complained to the CMA that one of Google’s products, for example, feeds real-time data back to advertisers for Google ads but not for Bing, even though there’s no apparent technical reason to treat the search engines differently. The Google spokesperson said that the product does not in fact treat different engines differently, but in its responses to the CMA report, Google didn’t challenge this finding.

Or take YouTube. It used to be that any demand-side platform (representing advertisers) could compete to buy ads on YouTube videos. But, in 2016, Google made a change: to advertise on YouTube, you generally now had to use Google’s own DSP. Google argues that the shift was made to protect user privacy. “Third-party DSPs with access to YouTube inventory could build profiles of users based on their viewing history, which would be a data protection risk,” the company told the CMA. But, as Morton and Dinielli note, the change might have given Google an unfair advantage, not just on YouTube but across the whole internet. Why would a brand pick a DSP that couldn’t place ads on one of the biggest platforms in the world?

This gets to one of the core issues: Google doesn’t just run the market; it participates in it, as a publisher that wants to maximize eyeballs. YouTube, the CMA found, accounts for 20 percent of the market for video ads (and as much as 50 percent when you exclude Facebook, which is its own closed system). Between that and the ads that appear above Google search results, it’s competing for revenue against the same publishers who rely on its services.

The tech antitrust movement has always been an uphill battle. Since the late 1970s, Supreme Court doctrine has had a strongly libertarian bent when it comes to monopolies, basically holding that if it doesn’t raise prices for consumers, it’s fair game. Since the most visible offerings of the likes of Facebook and Google are free (or, in the case of Amazon, insanely cheap), that standard has helped bulletproof Silicon Valley from regulatory jeopardy. A great deal of the recent antitrust backlash has focused on the need to revise antitrust law to account for the realities of the digital economy. Lina Khan’s influential 2017 law review article, “Amazon’s Antitrust Paradox,” for example, made waves for suggesting that Amazon was harming competition in ways not captured by current doctrine.

The looming case against Google, however, could simply skirt those theoretical disputes. Morton and Dinielli aren’t calling for a new paradigm; their argument operates within established standards. If it turns out to be a preview of the government’s case against Google—if US authorities find that the company’s dominance of the digital ad industry is leading to higher prices or inferior consumer offerings—then the upcoming litigation may fit squarely under existing antitrust law, no doctrinal revolution necessary.