Image via Wikipedia
The recent senate hearings on Google were political theater based on the unsettling fact that Google has achieved a natural monopoly in search and possibly the brokering of Internet ads. In an almost exact replay of Microsoft's 1998 Senate hearing, senators accused Google Chief Executive Eric Schmidt of serving up "cooked" search results that hurt smaller rivals.
The truth lies locked within Google's search algorithms, which must remain secret, or else websites would clog a Google page with nonsense and destroy its utility.
One solution, proposed by Mark Patterson of Fordham University School of Law, is to require a sort of incremental disclosure similar to what the Dodd-Frank financial reform law requires of credit-rating agencies. Patterson fits this intriguing idea within the evolving world of antitrust law in the essay below:
What should we do about Google? The recent Senate hearing addressed claims that Google uses monopoly power over Internet search to disadvantage its competitors. But the participants at the hearing offered few solutions, and what few they offered were inconsistent. In fact, though, Congress has already produced a solution, albeit directed at different market actors: credit-rating agencies. The two businesses—Internet search results and credit ratings—have more in common than one might think, so Congress’s solution for rating agencies might work for Google, too.
Both Google and ratings agencies are information providers. Google provides information that consumers use to choose which web sites to visit. Credit-rating agencies provide information that investors use to decide which investments to make. In each case, there is too much information for individual consumers to search efficiently themselves, so consumers pay Google and the ratings agencies to gather information for them.The two businesses are also similar in that consumers do not pay for information directly. Google’s search results are free, but Google is supported by advertising, so consumers pay when they buy the products and services that advertise on Google. Investors who use credit ratings also do not pay directly, because ratings are generally paid for by those issuing securities. Investors pay indirectly, though, because the companies in which they invest must pay for the ratings.
This payment model matters, because it arguably makes Google and ratings agencies less accountable to consumers of information. If information is free, the recipient of that information comes out ahead so long as the information has some value. As a result, consumers of information may not scrutinize the information as carefully as they would if they paid for it.
The danger is made worse because the ways in which Google is alleged to have tilted the playing field are difficult to detect. Google is alleged to have manipulated its search results and advertising in order to disadvantage potential competitors. The goal, competitors allege, is to cause consumers to visit Google-affiliated sites rather than the competitors’ sites. It is likely that the Google sites would have to be much worse than competing sites, not just somewhat worse, before consumers would become concerned about the quality of Google’s search results.
Similarly, the ratings agencies are alleged to have allowed their ratings to be influenced by their efforts to attract ratings business. As with Google, there are few details regarding specific instances of manipulated information, in part because there is no benchmark for the accuracy of either search results or credit ratings. But an antitrust suit several years ago alleged that when a school district hired competing credit-rating agencies rather than Moody’s to rate its bonds, Moody’s retaliated by issuing negative statements about the district’s financial position.
So both Google and credit-rating agencies are alleged to have been willing to provide inaccurate information to prevent competition. If these allegations are true, what could be done to prevent such distortion of information? As noted above, there is no benchmark for accurate search results or for accurate credit ratings, so it is difficult not only to know whether information has been distorted but also to devise a rule that would prevent it.
The solution that Congress adopted for credit-rating agencies in the Dodd-Frank financial reform law has two elements. First, the reason for material changes to the rating agencies’ algorithms must be publicly disclosed. Second, such changes must be applied consistently to all ratings to which the algorithms apply. These rules could be applied equally well to Google’s search algorithms.
This solution would not be a panacea, of course. Distortion of search results would still be possible, but the rule would make that manipulation more difficult, for two reasons. First, Google would be required to articulate reasons for changes in its results, which might in itself prevent some changes directed at injuring competitors. Second, the requirement that a change be applied uniformly would mean that a change to one search ranking could require other changes that Google might prefer not to make, which would discourage some ad hoc changes.
Moreover, the solution would satisfy two desirable goals. First, it would not prevent Google from making changes to its algorithm, even if those changes harmed its competitors, so long as the changes were applied consistently. Second, it would not require Google to disclose its algorithm, only to explain changes to it. Google is rightly protective of the investment that it has made in its algorithm, and although some have called for its disclosure -- if only to the government -- such a remedy seems too intrusive. Even a partial disclosure like that required by Dodd-Frank might be viewed as intrusive, but it seems likely that Google could provide reasons for changes without compromising innovation. (The SEC’s proposed rules implementing the relevant Dodd-Frank section require disclosure not only of the reason for a material change, but also of the change itself. That may be going too far, at least for search algorithms, but the provision has yet to take effect, so its implications will become clearer over time.)
As information becomes a more central part of the economy—perhaps the central part of the economy—antitrust law must develop new techniques directed at the particular problems that it presents. The current antitrust rules for limiting anticompetitive conduct by monopolists focus on forcing monopolists to share access to their goods or services. But when the conduct does not involve denial of a good or service but distortion of information, forced access is not the answer. Antitrust needs new techniques for ensuring equal treatment and preserving a level playing field. The solution adopted in the Dodd-Frank act is a step in that direction.
