NB: This post is part of the “Skepticism About Information Fiduciaries” symposium. Other contributions can be found here.
I was delighted to read Lina M. Khan and David E. Pozen’s recent article, A Skeptical View of Information Fiduciaries describing the reasons to be skeptical of the “information fiduciary” concept as a promising one for solving the problems posed by giant companies like Facebook. As Khan and Pozen point out, its proponents are a bit fuzzy about the details of how to reconcile these for-profit companies’ existing duties to shareholders, with some kind of fiduciary duty to users. The users’ attention is what these companies are selling. Khan and Pozen are skeptical that such a fundamental conflict can be resolved and I agree.
I only have a couple of additional points:
(1) Before looking to import the concept of a “fiduciary” to this new application, we might ask how well that concept has worked, as a means to check anti-social behavior, in the areas in which it has traditionally applied. If that area is corporate law where officers and directors are said to have fiduciary duties to the corporation and its shareholders, the answer is “Not very well.” It does not seem to have deterred much corporate misconduct.
(2) Although Khan and Pozen rightly observe that Facebook does more than sell passive viewers to its advertisers, it uses the data it collects from those users to construct or identify vulnerabilities that go far beyond the information asymmetry as it conventionally is understood in the fiduciary concept, in their discussion of the problems that the information fiduciary concept is meant to solve, they note (18-19) that many of these problems are already “proscribed by existing consumer protection laws,” they may not be confronting the degree to which Balkin, et al. may be attempting to offer alternative rationales for that existing consumer protection law given that it is no longer resting on as firm a foundation as in the past. However, the Supreme Court has been increasingly hostile to the government’s attempt to regulate any speech at all and increasingly willing to use the First Amendment as a weapon of deregulation. As Khan and Pozen note, to the extent that other arguments for special status or duties as a way to end run the Court’s more aggressive, the Supreme Court has not signaled much receptiveness to this approach.
1. The emptiness of fiduciary duties in the corporate context
Although I suspect there are many areas where fiduciary law has real bite, when it comes to preventing corporate misconduct, the letter of the law sounds better than it has proven to be in practice.
Many forms of conduct permissible in a workaday world for those acting at arm’s length, are forbidden to those bound by fiduciary ties. A trustee is held to something stricter than the morals of the market place. Not honesty alone, but the punctilio of an honor the most sensitive, is then the standard of behavior.
(Meinhard v. Salmon at 464)
If the directors of a company are fiduciaries for the company or for the shareholders of a company, then it would seem they should avoid behavior that would not be in a company’s long-term best interest, such as concealing misconduct. Yet, they regularly do just that. For example, although perhaps the directors of The Weinstein Companies were thinking of the best interests of the company when they authorized Harvey Weinstein’s rather peculiar employment contract that provided a series of escalating fines for instances of “misconduct” because they thought his talents would make more money for the company than his behavior would cost it, that calculation seems to have been based on the assumption that Weinstein’s misconduct would not ever be widely known, as once it was widely known it essentially destroyed the company. Some similar miscalculation of the potential for discovery presumably occurred at Volkswagen where engineers created software that allowed diesel vehicles to evade accurate detection of their emissions. Volkswagen not only created the defeat software, it marketed its vehicles for their supposedly superior performance from an environmental perspective, thereby exposing the company to criminal and civil liability. Some fiduciaries!
Although directors are said to owe the company fiduciary duties of care, loyalty, and sometimes good faith, that can, in theory, be enforced by shareholders, in reality it is difficult to do so. First, with respect to the duty of care, courts will not impose liability for a breach of duty of care so long as directors take relatively minimal steps to inform themselves of the condition of the company. This flows from the “business judgment rule,” which discourages courts from second guessing directors’ or managers’ decisions merely because they turned out badly. That leaves breach of loyalty. Breach of loyalty typically involves conflicts of interest, which are rarely deemed so overwhelming as to be prima facie instances of a breach of the duty of loyalty. So long as such conflicts are disclosed, and some prescribed number of directors not implicated in the conflict waive the conflict, it will not be considered a breach of duty. Disclosure typically insulates all but the grossest, most pervasive corruption.
When you add to this analysis the widespread practice of indemnifying officers and directors, of offering officers and directors’ liability insurance, and the many procedural and practical difficulties for shareholders in bringing shareholders’ derivative actions to enforce those duties, the reality is the officers and directors of most corporations can feel confident that they will rarely be asked to account for themselves in the exercise of those duties. This does not bode well for the effectiveness of the information fiduciary concept.
2. Manipulation, Regulation and the First Amendment
One of the troubling aspects of the behavior of companies like Facebook is the degree to which its users may be unknowingly manipulated. As Khan and Pozen observe: “By and large, addictive user behavior is good for business. Divisive and inflammatory posts are good for business. Deterioration of privacy and confidentiality norms is good for business.” However, according to Khan and Pozen, one proponent of the information fiduciary concept proposes mitigating this conflict by distinguishing between “predatory” and “non-predatory” advertising for purposes of regulating the use of data. This proposal seems to be directed though not at predatory advertising or data collection, but rather, at businesses which are arguably inherently predatory like payday lending. Yet Facebook itself is an example of a predatory business model. As Khan and Pozen observe, “the typical user does not even understand—much less approve of” the “basic contours” of Facebook’s business model.
Many scholars have raised concerns about how data mining can be used to manipulate those whose data is collected. Advertising and marketing are at base experimental processes, that is, its practitioners try something and then try (sometimes unsuccessfully) to measure whether it has “worked.” If the technique is not successful (i.e., if it does not cause consumers to buy the product), the technique is dropped; if it is successful, they continue to use the technique, even if the ad or practice is intensely disliked. This trial and error aspect of marketing, as well as its relative invisibility, is why I have described advertising as unregulated experiment on human subjects.
What is true of advertising generally, seems just as likely for any social media platform that relies on advertising and data mining. Thus, line drawing between advertising that is “abusive” or “manipulative,” versus that which is not, or between advertising and content when that content is on the Internet, will not be easy: it is virtually all manipulative. That was true before data-mining became ubiquitous. Data-mining and the possibility of comparing various databases makes it possible to engage in targeted advertising that gets around legal privacy protections. For example, despite assurance that data mining of pharmacy prescribing practices does not infringe on patient privacy, it is not clear this is true.
Moreover, it is not merely a hypothetical proposition that platforms profit from addictive behaviors and make money from encouraging them. For example, game developers use all sort of information to make their games more addictive, and since they know that children are more susceptible audiences, they, with Facebook’s cooperation, allowed developers to create conditions in which children might run up large bills on their parents’ credit cards playing games. “Facebook employees referred to these children as “whales” – a term borrowed from the casino industry to describe profligate spenders.”
The information fiduciary proposal appears to be motivated by a desire to create some sort of role-based status that would require platforms to protect the very people its business model uses as a resource from its own tactics. In short, it strains credulity that for-profit corporations like Facebook can resolve the inherent conflict between its business model and this fiduciary role.
Moreover, even if they adopted such a role, it would seem to require that the fiduciary engage in content filtering, particularly filtering for truth. Of course, truth is exactly what the government is not supposed to regulate under the First Amendment. Until fairly recently, there was an “out” for the regulation of truth in commerce. Indeed, the commercial speech doctrine is a species of First Amendment protection that is not even be triggered unless speech is truthful. This is the mirror image of what was required in New York Times v. Sullivan. There the Court posited that the First Amendment required “breathing space.”
Over time, though, the Supreme Court has moved to an ever more “speech protective” stance towards even commercial speech. It is requiring strict content neutrality that seems at odds with all sorts of existing regulation, let alone with the commercial speech doctrine itself. So I am not as confident as Khan and Pozen appear to be that existing consumer protection law is sufficient, and I think the information fiduciary concept may represent an attempt to preserve the government’s ability to ensure their continued viability.
However, my sense is that the problem the fiduciary concept is attempting to address will fail for the same reasons that most of our tools for corporate governance have failed. In this economy it is not just greed that is good, it is fantasy. Fantasy sells. Addiction sells. Fraud sells. The current business and media environment seems to require us to make substantive, content-based assessments that are contrary to the Supreme Court’s current interpretation of what content neutrality requires in the First Amendment context. It is unclear how the law can extricate us from this condition, but Khan and Pozen do a good job at showing why the information fiduciary concept is offers false hope.