Executive Summary

• Next March, the Department of Justice’s (DOJ) advertising technology (ad tech) antitrust case against Google will begin, and the U.S. District Court for the Eastern District of Virginia will examine how Google has acquired and operated different ad tech tools to become a dominant online advertiser.
• To win at trial, the DOJ will need to show that Google has both monopoly power in a relevant ad tech market and used anticompetitive conduct to either gain or maintain that monopoly power.
• Google’s widespread integration of different ad tech tools into its offerings and vigorous competition will not give rise to antitrust liability in isolation if Google can show that the conduct was pro-competitive or that it lacked the ability to extract monopoly rents.

Introduction

Next March, the Department of Justice’s (DOJ’s) second major antitrust case against Google will begin, with the agency in this case arguing that Google has illegally monopolized different portions of the advertising technology (ad tech) market. Google currently operates website publisher ad servers, advertiser ad networks, and critically, the ad exchange in between them. In practice, this means that Google can offer advertisers impressions on its affiliated publishers and offer publishers exclusive access to its advertisers. While both publishers and advertisers can go outside of Google’s networks, the DOJ will argue that Google has used monopoly power and anticompetitive conduct to effectively forego such competition.

To win at trial, DOJ will need to prove two elements for its primary claims. First, it must show that Google has monopoly power in a relevant ad-tech market. Second, it must show that Google achieved this monopoly power using anticompetitive conduct designed to exclude rivals and not simply through vigorous and effective competition. As Google will argue, it is unclear whether the firm has monopoly power in any relevant market due to competitive pressures from rivals and other services like video and application advertisements. And even if it does have monopoly power, the conduct at question may not actually harm competition in the aggregate.

This insight breaks down the major arguments from the DOJ and Google on both elements of the case and provides insights into their relative merits. Before Congress begins to author antitrust legislation targeting large technology firms, it should carefully consider how the courts interpret and apply current law in this case.

Background: Google and Ad Tech

Advertising plays a critical role in the Internet economy, as most websites do not charge a use fee, but rather raise revenue via selling advertising space. Advertisers buy what are known as impressions, which are essentially a display of their ads to a website’s users. The American Action Forum has previously published an insight on how digital advertising works.

Relevant to this case, Google currently competes in all three layers of the ad tech stack. First, for websites trying to sell advertising space and viewer impressions, Google offers its publisher ad server through DoubleClick, which issues requests for advertiser bids for publisher ad space. Second, for advertisers, Google offers both Google Ads and Display & Video 360, which provide bids to the publisher ad servers for advertising on a website. In between it all, Google Ad Exchange connects publishers and advertisers, including Google’s own services.

Figure 1: An outline of the ad tech market in the DOJ’s complaint.

Under this arrangement, Google can offer firms on both sides of the exchange access to a wide range of partners and better target the needs of its clients by providing an all-in-one offering, so to speak. Yet this offering also raises competitive concerns, for if Google has monopoly power in any one market, it can leverage that monopoly power to harm competition in other parts of the ad tech market.

Element 1: Monopoly Power

Because monopoly power could be used to exclude rivals, the U.S. District Court for the Eastern District of Virginia will first examine whether Google has monopoly power in a relevant market. While courts often use size and concentration as a proxy for monopoly power, monopoly power refers to the ability of a firm to price above competitive levels. Even if a firm has a 90 percent market share, if no barriers to entry or network effects exist, raising prices would lead to new entrants and maverick firms taking a large portion of those sales. Therefore, courts will look at such factors, and whether alternative substitutes exist that will ultimately restrict a firm’s ability to raise prices over competitive levels.

Publisher Ad Servers Market: First, the DOJ alleges that Google has monopoly power in the publisher ad servers market. Through its acquisition of DoubleClick, Google’s share of this market has remained above 90 percent for years, and barriers to entry – such as the cost to build such a network, high switching costs, and the difficulty to achieve scale that would make it valuable to advertisers – could give Google the ability to raise the fees it takes from the exchange (and away from the publisher) above competitive levels. Nevertheless, the DOJ complaint largely ignores rivals for website advertising space, including in-house publisher ad servers, as companies such as Facebook and Snapchat operate their own publisher servers. The complaint also largely excludes video and app ad publisher servers, which allow websites with advertising space to sell video ads. If Google raised prices on image ads, publishers could simply sell more video advertising space, limiting Google’s ability to extract monopoly rents.

Ad Exchange Market: Second, the DOJ alleges that Google has monopoly power in an “ad exchange for indirect open web display advertising” market. According to the complaint, Google AdX is the largest ad exchange in the market, and the direct purchaser of more than 50 percent of all ad impressions and revenue. Google also operates an Open Bidding system through which other ad exchanges may purchase its publisher ad inventory, but this only represents 7 percent of impressions won. Here, Google has a strong defense. The DOJ’s market definition again excludes app, social, and video ad exchanges, but also excludes direct sales between an advertiser and a website publisher, as well as limited auctions (i.e., auctions limited to a subset of advertisers). By excluding so many alternatives from consideration, the DOJ largely ignores the competitive pressures on Google’s ad exchange business.

Advertiser Ad Networks Market: Finally, the DOJ alleges that Google has monopoly power in “advertiser ad networks for open web display advertising.” This proposed market definition is the most tenuous claim for the DOJ, as it alleges Google’s monopoly is 70 percent but largely ignores that advertisers do not need to bid for web space. There are many other options for advertisers to reach consumers, such as social media and video apps. Even if the complaint’s 70 percent market share figure is accurate, these substitutable options for advertisers drastically limit Google’s ability to raise prices above competitive prices.

Figure 2: Google’s relevant businesses in each stack, as well as the relative market share according to the DOJ complaint

As displayed in the figure above, the DOJ complaint alleges that Google has more than a 50 percent market share in all three proposed markets, but the definition of the market in each case could potentially exclude rivals that prevent Google from raising prices above competitive levels. The DOJ will need to prove to the court that in fact these alternatives do not limit Google’s market power in one of these markets to proceed to the second element.

Element 2: Anticompetitive Conduct

Even if a firm has or may obtain monopoly power, firms can achieve that power through healthy competition. In these cases, courts will not declare the conduct illegal: A healthy market should encourage firms to improve their products and succeed on the merits. Therefore, courts generally look to whether conduct that leads to, or could lead to, monopoly power is anticompetitive.

Yet, as the American Bar Association of Antitrust Law Section of Antitrust Law explains, defining anticompetitive conduct “has been one of the most vexing questions in antitrust law.” Generally speaking, conduct that stems from procompetitive reasoning such as improving a product, generating efficiencies, or otherwise better competing with rivals does not violate the law. At the same time, some cases have found conduct as benign as purchasing an input as anticompetitive because it harmed a rival’s ability to buy that input. While many types of anticompetitive conduct such as tying two services together have been examined by the Supreme Court, almost no bright-line rules exist. The court here will likely look at both the effect and the intent to decide ultimately if the conduct satisfies this element.

The DOJ’s complaint cites a wide range of conduct it alleges Google engaged in to effectively prohibit competitors in each stack by tying its products and services together. For example, Google restricted advertiser demands to its ad exchange, and its ad exchange to its ad servers. And when Google introduced its open bidding option in response to a competitive threat in header bidding (a system of bidding that bypassed its ad exchange entirely), it would subsidize competitive bids to beat rivals, while raising costs on uncompetitive bids to offset the losses (known as Project Bernanke, as it resembled the “quantitative easing on the Ad Exchange”).

But, as is well-established in antitrust law, a firm need not deal with a rival. Google creating its own suite of products could have significant procompetitive justifications, such as the operation of an efficient, one-stop shop for advertisers and publishers. Such conduct could also help Google better compete with different ad exchanges, including those for videos or applications. Google will likely raise these, and a variety of other procompetitive justifications, at trial.

Importance for Congress

A variety of bipartisan legislation targeting large technology firms, and the ad tech space generally, has been introduced this Congress after stalling in the last one. These bills take a more progressive view of antitrust enforcement holding that size alone, and harm to competitors, should give rise to antitrust scrutiny. This case will be a good opportunity for legislators to take stock of current antitrust standards and the logic behind them. The decision, regardless of the outcome, will provide valuable insights to Congress about the need for reform.