Dan’s posts this week about privacy in campaign donations and public records are related to some issues I’ll be discussing soon (at the Congressional Internet Caucus’s State of the Net West Conference, the IPSC, and the Annenberg School at UPenn). I’ve discussed the power of intermediaries to shape our reputation in a few prior projects, and now I’m concentrating on how to address that power. Here’s an extended abstract on these ideas:
Rationalizing Internet Intermediary Responsibility
Reputation regulation has become essential because traditional restrictions on data and information flows—be they in the form of privacy or intellectual property laws—inadequately constrain important intermediaries. In considering the balance of power between intermediaries and those whom their actions affect, scholars have focused on either strengthening or weakening extant doctrines of copyright, trademark, contract, antitrust, and privacy law. However, a critical mass of doctrine in these fields, established patterns of consumer behavior, and the advent of cloud computing have freed up so much information that law must now not only be concerned with information’s flow, but with what results from it: the rankings, recommendations or ratings based on it.
While the Fair Credit Reporting Act set a precedent for reputation regulation, more recently only a few powerful groups have succeeded in the US in forcing reputation-generating systems to be more accountable. For example, a medical association in Washington state has persuaded some large insurance companies that rate physicians to disclose the basis of their ratings and to permit appeals of negative ones. Concerns about stealth marketing have also led to guidance from the FTC on the separation of editorial and paid content in search engines. EU intermediaries are generally required to be more responsible than those in the US. For example, one German court has required Google to manage results associated with an actress distressed by them, while a much worse case of harassment in the US led to no requirement of corrective action from Yahoo.
The burden of this article is to make a case for extending similar responsibilities to other dominant general-purpose search engines, as well as e-commerce hubs (e.g., eBay, Amazon, and iTunes), social networks (e.g., MySpace, FaceBook, CyWorld, and 23andMe), and gossip boards (e.g., AutoAdmit and JuicyCampus). Individuals and organizations ought to have a basic right to some account of how dominant intermediaries generate associations related to them. Meaningful exercise of that right will require flexible, responsive regulation of entities which make reputation-affecting decisions.
Reputation regulation may seem like an oxymoron, given the usual associations evoked by the two concepts. Reputation is traditionally considered a most mutable intangible, existing only in the minds of individuals. It may seem like the quintessential quality beyond the reach of bureaucrats. Regulation, on the other hand, has a reputation for being rigid, ossified, cumbersome, captured, or worse.
Nevertheless, companies stake enormous sums on their goodwill, and individuals have grown accustomed to a vast network of privacy regulations over the past few decades. If this most personal of attributes can be a prerogative of the administrative state, so too can its correlates. Moreover, regulation need not be administered only or even primarily by the state—as Google’s StopBadware program has already proven, a creative intermediary can partner with NGO’s to provide “rough justice” to sites it denigrates.
Recurrent battles over “network neutrality/broadband discrimination” have focused public attention on one dominant set of intermediaries—the carriers who control the physical layer of internet transmissions. However, there are other “bottlenecks” on the internet that merit similar attention because of their parallel power to order content on the web. While many commentators assume that these sites’ innovative genius should give them a completely free hand to conduct their own affairs as they see fit, they downplay legal sources of intermediaries’ success. As the legal realists cautioned, rarely is there a clean separation between state and market actors. Intermediaries’ dependence on legal immunities both belies their moral arguments for untrammeled autonomy, and provides a legal “hook” on which to hang reformist measures.
Considered in isolation, particular legal victories won by intermediaries in the United States over the past two decades may reflect warranted extensions of precedent or sound policy judgments. However, when seen as a broad spectrum of legal development, they are aggregating toward an unfair competitive advantage unearned by commensurate public responsibility. Intermediaries only deserve immunities to the extent they realize and reflect public values.
Part III articulates these public values by describing more responsible reputation-generating entities and applying the norms apparent in their operation to dominant online intermediaries. In early work on the topic I assessed search engine quality with respect to the authoritativeness and responsibility of the metadata they provided. In the course of proposing regulatory responses to search engine manipulation, Oren Bracha and I elaborated these concepts with respect to democratic legitimacy, economic efficiency, and fairness. As this article attempts to apply some ideals of accountability to intermediaries and ranking sites generally, reputation regulation provides a more unitary umbrella concept for assessing intermediaries’ degree of responsibility.
Part IV makes the moral and economic case for complementing court-driven responses to intermediaries with regulation. Calls for increasing public responsibility for intermediaries are presently being channeled in two reformist directions: A) promoting competition among intermediaries (by lowering barriers to entry and challenging incumbents’ anticompetitive practices), and B) tinkering in particular doctrinal areas in order to promote responsible behavior by intermediaries. Even at their best, neither of these approaches can fully realize the values described in Part III.
Competition promotion is at best a partial response—while it may maximize the “consumer welfare” of users of intermediaries, it may do worse than nothing for third parties (since one competitive strategy of intermediaries like Juicy Campus is to make it easier for users to harm third parties). Moreover, in some areas, the intermediary may be a natural monopoly, and any competition in the space it occupies is bound to be contrived. Traditional litigation (or doctrinal adjustment) is more promising, but risks either over- or under-correcting intermediaries’ practices. In many areas, it may prove beyond the institutional competence of courts to deal with rapidly shifting business practices occluded by trade secret protection. An agency with teams of engineers and programmers can monitor and understand intermediaries’ practices better than nonspecialist judges. Finally, both the competition-promotion and doctrinal-adjustment schools have tended to understand the problems posed by intermediaries in purely economic terms—whereas media and credit-reporting regulation has traditionally taken into account the cultural and moral consequences of intermediary power.
Legal scholarship has traditionally focused on discrete doctrinal areas. In intellectual property law, scholars seek to rationalize copyright, trademark, patent, and related doctrines; “cyberlaw” extends to contract, property, and tort online; and privacy experts confront the welter of common law and statutory limits on the accumulation and disclosure of data. While such specialization may promise to “work the law pure” in particular doctrinal bailiwicks, lack of coordination risks reinforcing trends that few would endorse. Sectoral regulation of reputation has the potential to promote the best aspects of intermediaries, while reining in their more irresponsible actions.