Search as Speech: Two Scenarios
Several privacy and antitrust complaints are now menacing Google. After gamely parrying these challenges, the search giant has now wheeled out its nuclear option: a First Amendment argument against any regulation of what appears in unpaid (aka “organic”) search results. A recent Google white paper by Eugene Volokh and Donald Falk has buttressed Eric Goldman‘s and Christopher Yoo‘s rationales for unfettered discretion in the exercise of search engines’ editorial judgment.
Volokh/Falk is the latest in a long string of Google filings describing search results as speech. It’s significant well beyond the search engine industry. If Google succeeds here, just about any information age company will start to make its selection and coordination of offerings “searchy” and thus “speechy” enough to avoid regulation.
Leading articles on the proper limits of the First Amendment include Fred Schauer’s The Boundaries of the First Amendment: A Preliminary Exploration of Constitutional Salience and Robert Post’s Recuperating First Amendment Doctrine. In Federal Search Commission?, Oren Bracha and I applied their arguments (among others) in the new technological contexts created by search engines (pages 1188-1201). I have also examined expressive dimensions of search in other work, in 2006, 2007, and 2008.
None of those prior efforts satisfied me as definitive. I wanted to write on the topic for years, but I couldn’t formulate a more general theory of search as speech. I now recognize the reason for my writer’s block: I was trying to impose a “one-size-fits-all” approach on multifarious phenomena. As Michael Carroll has shown, there are “uniformity costs” whenever we try to force a vast, sprawling array of human activities into Procrustean legal boxes. Those costs would be very high if courts were to accept the Volokh/Falk approach with respect to all the varied interactions between searchers and search engines. With that in mind, here are a few scenarios (or “test suites,” as Volokh might put it) to test the Volokh/Falk submission.
In the past few years, conservative groups have repeatedly criticized Google, which they perceive as a culturally liberal company. The rectitudinous Rick Santorum objected to the rectal results his last name generated. Columnist Michelle Malkin complained that websites like hers weren’t appearing in Google News results. Fox News reported conservative discontent at Google’s rapid response to manipulated search results related to Barack Obama, after its glacial efforts to defuse a “google bomb” aimed at George W. Bush (in which Bush’s biography would appear when the words “miserable failure” were searched). Although it took Google four years to address the issue with respect to Bush, Obama’s was resolved in a few days.
There may be good reasons for the difference in treatment. Google may have learned from the Bush experience and may now be capable of far faster responses to pranks. Or the company’s engineers may dismiss such manipulation as a silly prank that really should not waste their time. Nevertheless, political manipulation of Google merits some attention. Campaigns are a struggle for salience, and first impressions have considerable stakes. As more people form views of candidates from search results (or related Google properties like YouTube), we might worry that allegedly neutral, algorithmic representations of authority and popularity are really being influenced by a hidden agenda.
Anticipating disputes like this, Cory Doctorow envisioned a future where a “Googlecleaner” disappeared the most troubling incidents in Senators’ pasts from web search results. Such a scenario may seem fanciful in the current media landscape; if Google failed to present the “dirt” available on someone, for whatever reason, another media company would undoubtedly do so. However, as the company expands its partnerships with carriers and content providers, it aims to be the technical “back-end” ordering those companies’ search functions and digital data streams. Its secretive practices could easily spread throughout the media and communications landscape.
What would happen if an activist government, very unlike our own, tried to do something about this? What if there were some requirement of a right to reply, along the lines of the FCC rules unanimously vindicated in the Supreme Court’s Red Lion decision? John F. Blevins has taken a look at that problem, and concludes that different layers of the internet require different levels of scrutiny:
[C]ourts should defer to access regulations of uncompetitive network-layer platforms such as broadband access infrastructure. Fegulations of application-layer platforms such as search engines, by contrast, should be subjected to higher First Amendment scrutiny. . . . [because] different layers of modern digital networks have vastly distinct economic and technological characteristics.
Blevins believes that the old “scarcity” rationale behind Red Lion is obsolete, because “unique features of the Internet’s architecture make application markets (including search engine markets) inherently competitive and contestable.” There is an admirable forthrightness behind that “inherently,” but it does not match my experience of the search market.
In his book Planet Google, Randall Stross suggested that the company was using up to a million computers to index and map the web. Beyond the infrastructural challenge, many other factors make it extremely difficult for competitors to emerge in the general purpose search space. Google’s secrecy is not only designed to keep spammers from manipulating its results; it can also prevent rival companies from copying its own methods, or building upon them. Unlike patents, which the patent holder must disclose and which eventually expire, it is possible for trade secrets to never be revealed, let alone enter the public domain.
Innovation in search is heavily dependent on a base of users that “train” algorithms to be more responsive. The more search queries an engine gets, the better able it is to sharpen and perfect its results. For example, if a search engine finds that everyone in a given area clicks on the third result instead of the first result in a given day, it can tailor results for that area to elevate what was once merely the third result. If other firms were able to observe this process, they might be able to develop rival, and better, computational strategies. Instead, the data is kept secret. The self-reinforcing “Matthew Effect” described by Robert Merton takes hold: to those who already have much, more is given. Incumbents with large numbers of users enjoy substantial advantages over smaller entrants.
Competition may not lead to less secretive search engines unless the dominant search engine—Google—becomes more open about its own data and algorithms. It is impossible to find better interpretations and applications of data without access to it. As long as Google’s search data store is secret, no would-be rival will have access to this critical “raw material” for search innovation.
Quantum leaps in technology capable of overcoming these brute disadvantages are unlikely. Search is as much about personalized service as it is about technical principles of information organization and retrieval. Current advantage in search is likely to be self-reinforcing, especially given that so many more people are using the services now than when Google overtook other search engines in the early 2000s. And even if Goliaths like Facebook and Apple manage to squeeze Google out of the burgeoning worlds of social and mobile computing, they will likely raise the same concerns that Google has.
Finally, even if you doubt everything I’ve said above (prediction is hard), the relevant issue should not necessarily be “is competition possible,” but instead “is there competition?” Some media tycoon could sweeped in to the US in the late 1960s, lobbied the right people, assembled enough capital, and built an alternative network. (Indeed, some would say that conservative media tycoons have done exactly that with the rise of News Corporation around the world and Fox News in the US in particular.) The mere fact that, in a world of vast inequalities, spare capital is always at the ready to engage in some massive project, is no reason to think that it’s also always already on the verge of becoming a deus ex machina in a Schumpeterian drama of creative destruction. A key US rival of Google, Microsoft’s Bing, is hemorrhaging money.
Jerome Barron, the author of the 1967 article that provided key support to the Red Lion decision, was right to reaffirm in 2007 that access rights remain as relevant (and as constitutional) as ever. Scholars like Jennifer Chandler and Dawn Nunziato have carefully and convincingly applied access rights theories to intermediaries like dominant search engines. Volokh/Falk would need to engage their work directly in order to convince me their approach is correct.
Imagine that you own Company A, and your main competitor is the persistent (but demonstrably worse) Company B. In searches for the products you sell, you reliably end up in the top five results in the studies you’ve commissioned; your competitors at Company B are on the fifth or sixth pages. What happens if Google purchases Company B, and immediately after the purchase, Company B appears to dominate the first page of results on Google, and your company has been relegated to later pages? Stipulate that repeated appeals to webmaster forums and other mechanisms of corporate due process fail. What if the government decides (via antitrust enforcement or regulation) that the search engine must disclose its commercial connection to Company B?
If the action is happening in the paid search listings, the answer is relatively clear. The FCC already regulates stealth marketing extensively. The FTC has insisted that “disclosures that are required to prevent an ad from being misleading, to ensure that consumers receive material information about the terms of a transaction or to further public policy goals, must be clear and conspicuous.” Even mere bloggers are not immune.
What about organic search listings, which are the focus of the Volokh/Falk submission? At least in the antitrust context, enforcers’ prerogatives seem clear. Consider the following quote from Associated Press v. U.S. (1945):
[T]he argument is made that to apply the Sherman Act to this association of publishers constitutes an abridgement of the freedom of the press guaranteed by the First Amendment. . . . The First Amendment, far from providing an argument against application of the Sherman Act, here provides powerful reasons to the contrary. That Amendment rests on the assumption that the widest possible dissemination of information from diverse and antagonistic sources is essential to the welfare of the public, that a free press is a condition of a free society. Surely a command that the government itself shall not impede the free flow of ideas does not afford non-governmental combinations a refuge if they impose restraints upon that constitutionally guaranteed freedom. . . . Freedom to publish is guaranteed by the Constitution, but freedom to combine to keep others from publishing is not. . . . The First Amendment affords not the slightest support for the contention that a combination to restrain trade in news and views has any constitutional immunity. [internal quotes omitted]
Six years later, the Supreme Court reaffirmed that position in Lorain Journal v. U.S. (where, as Allen Grunes states, a “newspaper refused to deal with advertisers who advertised on [a] new competitor, a radio station”):
The publisher claims a right as a private business concern to select its customers and to refuse to accept advertisement from whomever it pleases. We do not dispute that general right. But the word right is one of the most deceptive of pitfalls; it is so easy to slip from a qualified meaning in the premise to an unqualified one in the conclusion. Most rights are qualified. The right claimed by the publisher is neither absolute nor exempt from regulation. Its exercise as a purposeful means of monopolizing interstate commerce is prohibited by the Sherman Act.
The operator of the radio station, equally with the publisher of the newspaper, is entitled to the protection of that Act. In the absence of any purpose to create or maintain a monopoly, the act does not restrict the long recognized right of trader or manufacturer engaged in an entirely private business, freely to exercise his own independent discretion as to parties with whom he will deal. . . . Injunctive relief under section 4 of the Sherman Act is as appropriate a means of enforcing the Act against newspapers as it is against others. [internal quotes omitted]
Volokh/Falk argue that Lorain is inapplicable to the search engine context because the newspaper in that case “was not excluding advertisements because of their content, in the exercise of some editorial judgment that its own editorial content was better than the proposed advertisements. Rather, it was excluding advertisements solely because the advertisers—–whatever the content of their ads—–were also advertising on a competing radio station.” It seems to me that several of the antitrust-based complaints against Google follow exactly that Lorain pattern, suing Google for decisions that the complainants say are made only according to an economic, and not an editorial, logic. Let me lay out the allegations in one of these situations, as told to me during an interview with the Shivaun and Adam Raff, owners of Foundem. I have not independently verified the following these allegations, but I want to give the Raffs’ version of events.
Foundem is a UK-based firm that provides specialized search for price comparisons. It is run by a husband-and-wife team of engineers with formidable CV’s and a track record of innovations. One of the UK’s leading consumer bodies, Which?, ranked Foundem third in a comparison of 10 travel price comparison sites. A leading technology television program, The Gadget Show, ranked it first when it compared several sites.
Foundem has been unable to convert critical acclaim into a mass user base. The company lays the blame squarely at the feet of Google. The Raffs say that, less than six months after Foundem launched, Google started blocking it from the front pages of most organic, or unpaid, search queries for price comparisons.
The organic search penalty was a blow to Foundem’s business. Web-based start-ups can get a large proportion of their traffic from internet giants like Google. The penalty that downranked Foundem’s site dramatically reduced its visibility in Google’s organic results. The company then tried to reach users with advertisements, but Google also effectively cut off that option. It required Foundem to offer a minimum bid of about 5 pounds—–rather than the 5 pence it had been bidding—in order to participate in auctions for search advertising terms like “product comparison.” As Foundem put it, it was “now being asked to pay a staggering and prohibitive 10,000% more than its unpenalised competitors.” This massive increase made it prohibitively expensive for Foundem to use Google’s search advertising. From August, 2006 to September, 2007, Foundem was effectively eliminated as an option for those searching for price and product comparisons on Google.
According to Google’s search team, Foundem was a “low-quality” site, composed mainly of links to other sites. Foundem disagrees, alleging that Google was effectively excluding it because Google wanted to promote its own subsidiaries, such as Google Product Search. To put this in classic antitrust terms, consider that either Foundem or Google Product Search could be conceived of as tools to find products. Is Google’s effort to elevate its offering above that of Foundem like Microsoft’s favoritism toward, say, Internet Explorer? Allen Grunes has suggested some analogies:
Consider the Microsoft decision. Favoring your own products or services, as Microsoft did with Explorer, is not enough to show monopoly maintenance. After all, Internet Explorer users could still download Netscape, so in that sense competition was just “one click away.”
The necessary next step has to be some affirmative conduct to shrink your would-be rival’s share. That was a critical element to the D.C. Circuit in Microsoft. We know how Microsoft did it, according to the court: by conduct including restrictive agreements with third parties and deception of developers. What are some of the possible parallels?
In a two-sided market, scale comes from two related things: audience and ads. In the Internet world, that means hits and ads or other revenue sources. Stop an emerging competitor from getting enough hits, and you make it less valuable to advertisers or business partners. Stop it from getting ads (or other revenue) and it won’t grow. That is the newspaper “death spiral.”
So think Lorain Journal, updated. Recall that in Lorain Journal, the newspaper refused to deal with advertisers who advertised on the new competitor, a radio station. Deliberately demoting search results is akin to a refusal to deal, since a significant loss of ranking amounts to the same thing. But notice something else: as in Lorain Journal, the conduct is aimed not just at the competitor, but also – and in fact primarily – at the competitor’s advertisers or other revenue sources. Cut off its air supply, so to speak, so it won’t grow into a competitive threat.
To complete the inquiry, I think you need to ask: What are the threats to Google’s dominance over search? Or, for that matter, to its dominance over online advertising? To the extent that vertical search services or social media may be such threats, I think you have the ingredients for a case very similar to the one DOJ brought against Microsoft. But the key is in the documents. Few would have guessed that Microsoft viewed Netscape as a threat to its operating system monopoly.
Notice something interesting here. When you recast the question, the focus is on competitive entry or expansion. It is perfectly consistent with dynamic competition. In Microsoft, the focus was on whether the defendant was taking affirmative steps to prevent the emergence of something new, something that could displace its dominance in the operating system market. When you recast the question, whether or not Google has a “duty to deal” is not the central issue. And there is no need to assume or prove that there is such a thing as a “neutral” search result.
The Volokh/Falk submission can only declare Lorain Journal inapposite by assuming what needs to proven; namely, that there is no distinct anticompetitive conduct motivating the expressive display of search results. As Mark Patterson has argued, “given how little we know about what goes on behind those closed doors at Google and given that recent events give us little reason to trust it, it is worth considering how antitrust might respond to anticompetitive conduct. It’s remarkable to see scholars dismissing antitrust concerns without even knowing what facts the investigations will reveal [emphasis added].” It would be deeply ironic if the First Amendment could be deployed to limit, rather than promote, inquiry devoted to advance public understanding of critical internet decisionmaking.
Reporter Noam Cohen recently asked:
Is Google search an intermediary like the phone company — simply connecting people with the information they seek? Or is Google search a publisher, like a newspaper, which provides only the information that it sees fit and is protected by the First Amendment?
Everything the New York Times covers, we talked about five years ago. But the question is newly urgent, and needs a fuller airing. Volokh and Falk have renewed a conversation about the nature and limits of “free expression” claims that may prove as critical to the fate of the information economy as Lochner was to decades of US economic developments.