Not with a Bang
The Laboratorium 2013-03-30
So ends the Federal Trade Commission’s long and contentious investigation into Google. Out of the four serious issues on the table, Google walks away cleanly on one (“search bias”), the FTC gets a clear victory on one (“standards-essential patents”), and Google makes mushy-mouthed “commitments” on the remaining two (“vertical opt-out” and “ad portability”). But the issue on which the FTC let Google walk—“biasing” its search results to favor its own content over competitors’—was far and away the most important. The mood over at the Googleplex has to be pretty good right now.
Full analysis below, but first, the essential documents:
- FTC statement on the settlement
- Consent order on standards-essential patents
- Plain(er) english analysis of the consent order
- Google letter making commitments on vertical opt-out and ad portability
- FTC statement on search bias
Standards-Essential Patents
The FTC didn’t start out focused on Google’s use of patents. But as the smartphone industry descended into a post-apocalyptic Hobbesian landscape in which scavengers picked over the shells of burnt-out companies for usable patents and litigators started waving around patent arsenals of fearful destructiveness, the FTC followed the smell of gunpowder to its source. Google acquired Motorola in large part for its patent portfolio.
One key piece of Google’s strategy, like that of everyone else holding smartphone patents of immense value, was to seek injunctions against other companies’ phones. These injunctions, however, undermine a key premise of the standardization process that makes these wonderful gizmos feasible: fair, reasonable, and nondiscriminatory (FRAND) licensing for patents that are “essential” to implementing a standard. FRAND licensing is in effect a pact by the companies putting a standard together that once they agree on a standard, none of them will try to hold the others for ransom by brandishing its patents like a C-list rapper. FRAND licensing doesn’t mean free licensing, but it does mean being willing to license at all, rather than asking the authorities to drop competitors’ phones into a Blendtec.
Recent judicial decisions—including this Richard Posner barn-burner involving Motorola—have been skeptical of injunctions on standards-essential patents. The FTC concluded, 4-1, that the practice is an unfair method of competition, and Google agreed not to engage in it in the future.
A cynic might say that Google threw in the towel because the winds in the land were already blowing against standard-essential patent injunctions. An even bigger cynic might add that Google and its Motorola marionette have tended to be at the wrong end of the patent stick more often than they’ve been at the right end, so Google in particular has even less to lose, and might even gain from entrenching the rule it just agreed to as an industry-wide one. A truly epic cynic might ask whether Google turned to standards-essential patents in the first place because the rest of its larder was so bare. A mere realist would look at today’s settlement as good news for Apple: one more Jersey barrier has been cleared from the path of the iPhone juggernaut.
Ad Portability
One of the defining features of the search industry is that it’s a two-sided market. In addition to the consumers who use Google, there are the advertisers who actually pay for it. Google’s first and most important money factory was its ability to match advertisements to specific search queries. In the years since, Google’s AdWords program has grown into a huge and complicated system that lets advertisers customize almost everything about their ads and when they’ll be displayed. Other search engines that compete with Google for users and for advertisers have their own similar systems; the competition on the a din vertiser side thus involves advertisers transferring campaigns from one search engine to another.
Google’s critics alleged that Google restricted the use of AdWords in one clever and crucial respect: “The AdWords API Client may not offer a functionality that copies data between Google and a Third Party.” That is, you can advertise on both Google and Bing, but you can’t use a program to copy your Google AdWords campaign over to Bing. This provision and a few related ones, it was alleged, made AdWords into the Roach Motel of search advertising. A medium-sized advertiser—one big enough to have a campaign that would be painful to replicate by hand but small enough to depend on third-party tools to manage its campaign—would have a hard time switching out.
I have seen defenses of this rule. Indeed, FTC Commissioner Rosch offers one in his dissent from the commission’s statement closing the investigation: the rule “ensur[es] that third-party intermediaries take advantage of the unique features available on AdWords.” And even if the rule has few benefits, neither Rosch nor his colleague Commissioner Ohlhausen thought that it harmed consumers in a way that brought it within the FTC’s antitrust enforcement powers.
But this was never going to be the hill on which Google made its last desperate stand. However much Google gains from encouraging advertisers to create uniquely tailored campaigns by limiting the use of third-party tools, it has more to lose from a protracted legal battle. The only question is why Google didn’t simply delete the term earlier to make the issue go away. But today’s settlement gives the answer: a bargaining chip can only be bargained away once. Google held on to the term so it would have something to give the FTC as part of a compromise that preserved the powers Google really cares about.
Vertical Search Opt-Out
Google has had a series of long-running disputes with websites over just it uses the content they put online. Google does offer an opt-out for websites. In fact, it offers three: a robots.txt file, META tags in webpages’ HTML, and its own Webmaster Tools forms. Any of these opt-outs can be used on a page-by-page basis, so I could, for example, ask Google to index this blog’s front page, but not the archives.
But what these protocols don’t do is let websites opt out of particular uses that Google might make of the pages it indexes. So if you’re Yelp, and you didn’t want Google copying short snippets of user reviews for its own local search, your only practical option was to tell Google to go away entirely. But that would mean giving up your Google traffic, and one significant way that users find sites like Yelp is through Google searches in the first place. From Yelp’s point of view, Google was using Yelp’s own content to compete with it. Newspaper publishers have regularly raised the same complaint about Google’s snippets in Google News: they like the traffic Google provides but not Google’s competing product. The publishers tried their own alternative proposal for more granular opt-outs: ACAP. But it was badly designed and never was adopted by the one group whose adoption mattered: search engines.
Under pressure from the FTC, Google has now agreed to offer an opt-out from one of its vertical search properties. The details are interesting—and, as usual, all of the interesting details are in the footnotes. First, the opt-out only applies to “Covered Webpages,” which “have the primary purpose of connecting users with merchants.” Thus, Yelp can opt out of Google Local without opting out of Google, but if you want to opt out of Google Image Search without opting out of Google, no such luck. Thus, this rule really only applies to transactional vertical searches: ones whose goal is to find and buy something. It only really protects other organizing intermediaries, like Yelp and Expedia. Google can still strong-arm content producers, like newspapers and blog empires, to the fullest extent allowed by law.
Second, the opt-out is site-wide. You can opt out example.com as a whole, but not sub.exmple.com or example.com/sub by itself. I wouldn’t be surprised if there’s a back-end reason for this related to how Google’s crawlers and search-box integration systems work. It also doesn’t seem like a dealbreaker from the websites’ side of the table.
The opt-out changes the terms of the bargain between Google and websites moderately, but not very dramatically. Websites have always had the ability to opt out entirely, so the question is what kinds of concessions they can extract from Google in exchange for not taking their ball and going home. Google has proven willing to dicker over details in the past; it has changed its crawling rules to play nicely with newspaper paywalls, for example.
Here, Google conceded the essential point long ago: it gave Yelp the opt-out it wanted. Today’s concession institutionalizes this practice and makes it broadly available to other websites. But Google was clearly prepared to make this retreat. Call it the recognition of the inevitable, call it a reasonable compromise, or call it a concession offered to help the FTC save face, something like this has been waiting in the wings for months.
Search Bias
The center of the investigation, and of the complaints against Google for the past four years, has been “search bias”: the accusation that Google deliberately slants its search results to favor its own sites (like its own local results and flight search) over its competitors (like Yelp and Expedia). Last year, FTC staff wrote a 100-page report that reportedly recommending suing Google to prohibit search bias. But today, the FTC by a 5-0 vote decided not to take any action against Google over search bias. According to the FTC, then, whatever bias Google has engaged in is fine.
This outcome is a giant middle finger extended in the direction of the Google critics, like FairSearch, who have been calling for stringent FTC action on search bias. (Just yesterday, FairSearch put up a plangent blog post telling the FTC that there was “no reason to rush” its investigation.) I cannot overstate the extent to which the anti-Google case was premised on search bias. It was the hub to which all the other issues connected. Google’s vacuum cleaner approach to personal information and its systematic manipulation of its search results were the two constant themes Google’s enemies used to explain its actions and to appeal for government intervention.
But today’s settlement directly repudiates the search bias claims. It doesn’t say, “Google might be doing something bad but we didn’t find a smoking gun,” or “Google is doing something bad but not something the FTC can prevent.” No, the settlement says, “We looked, and what Google is doing is good.” If the final FTC statement had been any more favorable to Google, I’d be checking the file metadata to see whether Google wrote it. Just look what the FTC concluded:
- Google was pure of heart: “The totality of the evidence indicates that, in the main, Google adopted the design changes that the Commission investigated to improve the quality of its search results, and that any negative impact on actual or potential competitors was incidental to that purpose.”
- Google helped users when it helped itself: “Notably, the documents, testimony and quantitative evidence the Commission examined are largely consistent with the conclusion that Google likely benefited consumers by prominently displaying its vertical content on its search results page.”
- The data agree with Google: “Analyses of ‘click through’ data showing how consumers reacted to the proprietary content displayed by Google also suggest that users benefited from these changes to Google’s search results.”
- Everyone else did it, too: “We also note that other competing general search engines adopted many similar design changes, suggesting that these changes are a quality improvement with no necessary connection to the anticompetitive exclusion of rivals.”
- Users win when (some) websites lose: “For example, for shopping queries, Google demoted all but one or two comparison shopping properties from the first page of Google’s search results to a later page. … These changes resulted in significant traffic loss to the demoted comparison shopping properties, arguably weakening those websites as rivals to Google’s own shopping vertical. On the other hand, these changes to Google’s search algorithm could reasonably be viewed as improving the overall quality of Google’s search results because the first search page now presented the user with a greater diversity of websites.”
- Relevance is subjective: “Reasonable minds may differ as to the best way to design a search results page and the best way to allocate space among organic links, paid advertisements, and other features. And reasonable search algorithms may differ as to how best to rank any given website.”
- Google needs a free hand: “Challenging Google’s product design decisions in this case would require the Commission – or a court – to second-guess a firm’s product design decisions where plausible procompetitive justifications have been offered, and where those justifications are supported by ample evidence.”
The decision to drop the search bias investigation, and this statement all but lauding Google, were made by 5-0 votes. That’s three Democrats and two Republicans, at the end of an investigation the FTC fought hard to get, for which it hired a veteran outside litigator, and for which its own staff was raring to go. If they just wanted to make the case go away, it could have been dropped more quietly, without the encomia to Google’s search-engine design. No, the end result of the FTC investigation is a prominent public vindication for Google.
The public case against Google—and hence also the legal case against Google in these uncharted antitrust waters—is now immeasurably weaker than it was before Google’s critics pushed so hard for an investigation. This is one of the biggest lobbying backfires I have ever seen. In fact, an ironist might argue that the lobbying backfired because it was so aggressive. Certainly it put Google on the defensive, hiring lobbyists and proxies. But it also seems to have raised hackles inside the FTC, undercutting critics’ credibility.
The FTC will “continue to monitor” Google on search bias, but further action seems unlikely unless something especially dramatic happens illustrate bad bias in action. Google is still in negotiations with the EU over the same slate of issues, and there have been reports that the EU was preparing to take a tougher line than the FTC. (State attorneys general, also on the case, may well also try to bring their own actions against Google.) So Google might now take this outcome across the pond to push for a similar resolution there. Or, if the EU is able to extract bigger concessions, the FTC could end up looking gun-shy.
One Last Thought
One interesting piece of timing is that yesterday, the Senate confirmed legal scholar Josh Wright to be an FTC commissioner. Wright, who has written Google-funded papers defending Google from antitrust arguments over search bias, has recused himself from all Google-related matters for two years. So if the FTC had waited until Wright took his seat, it would have faced the risk that it would end up deadlocked 2-2 on this high-profile case. On the main issue—search bias—this wouldn’t have been a problem, but the FTC was fragmented over whether a voluntary “commitment” by Google was an appropriate remedy on vertical opt-out. Being down a Commissioner would have increased the risks of deadlock somewhere in the case.