Google Search found out

Google is liable for illegally monopolising the general online search market, according to a United States federal judge in a ruling issued on 5 August. Remedies to the behaviour of Alphabet, Google’s parent, will be identified in a further process.

On one hand, this feels unexpected. US courts, particularly higher courts which are more ideological and farther removed from the evidence, have been trending in a conservative direction on competition cases. They often seem willing to assume anticompetitive conduct is not possible, rather than assessing market realities (as seen, for example, in Federal Trade Commission v. Qualcomm Incorporated).

On the other hand, the conclusion reached by Judge Amit Mehta was easy. The court found that Google has a 94% share of the general search market today, and that it has been very high for a decade. It paid its most threatening competitor, Apple, more than $20 billion per year to stay out of the search market. Apple positioned Google Search as the exclusive default in search access positions on its iPhone operating system (iOS) and in exchange received 40% of the search revenue generated on iOS. Google’s monopoly was thus preserved even though it had to split the rents with a potential competitor. 

The many contracts that Google requires of handset makers (OEMs) that license Google Android prevent OEMs from sponsoring entry by rival search engines. OEMs are contractually required, or paid, to make Google Search the exclusive pre-installed default search engine on their devices at all search access points. 

The US Department of Justice (DOJ) litigated the case. Their approach cleverly emphasised how similar Google’s conduct was to the conduct that was condemned as anticompetitive in the turn-of-the-century US DOJ Microsoft case, which found that Microsoft illegally maintained its monopoly in operating systems. The exclusionary tactics used by the two monopolists were remarkably similar, with the main difference being the lack of internal Google documents explaining the reason for their anticompetitive strategy. But Judge Mehta noted the lengths to which Google had gone to hide or destroy internal conversations, creating a strong inference that such internal evidence, as in the Microsoft case, had existed.

Google’s argument – it is the best search engine so that is why everyone uses it – did not prevail for obvious reasons. If a product is the best, its maker does not need to deploy exclusive contracts that forbid its customers from using rivals, nor should Apple need to be paid billions of dollars per year to use it. These contracts freeze the market with Google in place as a monopolist and prevent entry by rivals, thereby shutting down competition.

Business harm from the monopolisation arose from advertising that was overpriced or lower in quality than it would have been under competition. The consumer harm from the monopolisation is primarily in the area of quality and innovation. Today Google’s search engine results page is crammed with ads, many of which are designed to look almost exactly like organic results, and which crowd the organic results off the page, reducing the consumer’s ability to get the information they are seeking.

Lost innovation is likely to be significant. Without Google’s monopoly, search engines could specialise through curation, differentiate through algorithms, offer business models that preserve privacy and innovate in new services such as voice search (Siri) or search across domains (Spotlight, Branch). When customers are locked up with exclusives on Android and iOS, no funder will back an entrepreneur with an idea for a better search engine. If the court’s remedy changes the market so that innovators have a chance to attract consumers, investment and entry will likely result.  

There are several categories of remedy to consider. The first is obvious: Google should not be permitted to contract for exclusive use of its search engine with any channel in exchange for any type of consideration. This would prevent Google from paying Apple a share of its monopoly profits and force Apple to consider entry itself to re-capture that search revenue. Under this remedy, Google would likewise not be able to compel Android OEMs to use its search engine exclusively. However, their relationships with Google are of complete dependence (what other OS can they license?) and so they would be unlikely to carry a search competitor for fear of retribution. This justifies a second remedy, divestiture. 

The simplest and cleanest remedy to create competition in search would be for the court to order Google to divest Google Android into an independent corporation (and set a regulated price for the OS for a number of years). This would give OEMs the security of licensing the Google Android OS at the regulated price. They can then bargain over search engine placement and revenue share without fear of losing access to the OS. An OEM could choose a different search engine as the default across some or all of its handsets in exchange for more search revenue. Consumers would pay less for handsets because that higher level of search revenue would be competed away in handset prices. Thus more Google profits would find their way into consumers’ hands. A third good option is mandatory licensing of the web index (like a card catalogue of what is on the internet) at regulated rates, which would lower the fixed cost of operation for competitors and therefore help stimulate entry.