The Fatal Conceit of the “Right to be Forgotten”

Intelligence Squared hosted a lively debate last week over the so-called “Right to be Forgotten” embraced by European courts—which, as tech executive Andrew McLaughlin aptly noted, would be more honestly described as a “right to force others to forget.”  A primary consequence of this “right” thus far has been that citizens are entitled to demand that search engines like Google censor the results that are returned for a search on the person’s name, provided those results are “inadequate, irrelevant, or no longer relevant.”  In other words, if you’re unhappy that an unflattering item—such as a news story—shows up as a prominent result for your name, you can declare it “irrelevant” even if entirely truthful and ask Google to stop showing it as a result for such searches, with ultimate recourse to the courts if the company refuses.  Within two months of the ruling establishing the “right,” the company received more than 70,000 such requests.

Hearteningly, the opponents of importing this “right” to the United States won the debate by a large margin, but it occurred to me that one absolutely essential reason for rejecting this kind of censorship process was only indirectly and obliquely invoked.  As even the defenders of the Right to be Forgotten conceded, it would be inappropriate to allow a person to suppress search results that were of some legitimate public value: Search engines are obligated to honor suppression requests only when linking some piece of truthful information to a person’s name would be embarrassing or harmful to that person without some compensating benefit to those who would recieve the information.  Frequent comparison was made to the familiar legal standards that have been applied to newspapers publishing (lawfully obtained) private information about non-public figures. In those cases, of course, the person seeking to suppress the information is typically opposed in court by the entity publishing the information—such as a newspaper—which is at least in a position to articulate why it believes there is some public interest in that information at the time of publication. 

Google, of course, is not a newspaper or originating publisher, but a conduit between hundreds of millions of users seeking information and the billions upon billions of Web pages containing it. Unlike a newspaper, they do not index information with any concrete preconcieved ideas about why a particular Web page is likely to be interesting, relevant, or valuable to a specific audience.  Rather, their software crawls the Internet and ingests information broadly, on the premise that their hundreds of millions of users will have their own myriad reasons for finding that information valuable. They could not, of course, possibly know why each of those millions will find any particular result relevant or interesting in the future—and the great utility of the service they provide is precisely that they do not need to. This creates a profound asymmetry, however, that inevitably biases the “Right to be Forgotten” process in favor of suppression.  If I am seeking to block Google from displaying some piece of unflattering information in connection with my name, obviously I have some articulable—perhaps even facially legitimate—reason for wanting it suppressed. But neither Google nor any court or “information commissioner” considering my argument can possibly take into account the myriad reasons unknown persons in the future who conduct searches on my name might want that information available.

Consider one of the most famous Right to be Forgotten cases—that of Mario Costeja Gonzáles, who successfully forced Google to exclude from searches on his name an item in the Spanish newspaper La Vanguardia, published in 1998, concerning the forced sale of some of Costeja’s properties to satisfy debts.  More than 15 years later, Costeja convinced the Court of Justice of the European Union that information about his old financial woes was “no longer relevant” and should not be displayed in response to Google searches on his name, rendering them invisible for pratcical purposes even though the original item remained in the newspaper’s archives.  (In a classic example of the Streisand Effect, the enormous attention generated by the case itself guaranteed that his efforts would be in vain as countless new articles appeared recounting those facts—but the point and intent of the RtbF is obviously that this would not be the norm.)    The problem, of course, is that while Costeja was present to make his case that this information was now “irrelevant,” the particular individuals to whom it might be relevant in the future were not, and inherently could not, be there to make the counterargument.  The most the Court could do was imagine who those future people might be—potential employers? business partners? lenders? lovers? journalists?—and attempt to weigh their hypothetical future interests against Costeja’s. Google, of course, is in precisely the same position—and faced with tens of thousands of requests over time, cannot possibly devote intensive scrutiny to each one. Now that the court has confirmed their obligation to entertain such requests, they have every financial incentive to err on the side of complying in most cases, as refusal may require them to devote time and financial resources to fighting an appeal governed by a vague and highly subjective standard.

The core fallacy underlying the Right to be Forgotten, then, is a species of the “Fatal Conceit” that F.A. Hayek saw at the heart of the case for centralized economic planning.  The “conceit” was the beguiling belief that some body of experts could calculate the best uses of an entire society’s productive resources, allocating them more rationally than messy, wasteful competitive markets.   This was a mistake, Hayek saw, because production and distribution in those messy markets was coordinated by a system of prices that harnessed and transmitted widely dispersed information stored in the brains of millions of market actors—information about individual preferences and desires, about local surplus and scarcity, and about the probability of those conditions changing in the future. History has, of course, amply borne out that pessimism about the ability of central authorities to solve the economic “knowledge problem” and rationally calculate the optimal use of a society’s resources.

Lying behind the Right to be Forgotten is a similarly grandiose—and similarly erroneous—conceit: The assumption that Google or a court or an “information commissioner” will be able to reliably determine, in tens of thousnds of cases per year, what factual information will be “adequate” or “relevant” to the users  who are searching for it, not just today, but in potentially quite different circumstances tomorrow. Those motivated to search, for reasons today’s arbiters cannot predict, may well make unwise or hasty or frivolous judgments about the adequacy or relevance of those facts—but surely they are still in a far better position to make that determination, bringing to bear their own local knowledge of their specific facts and circumstances, than a body of experts relying exclusively on the perspective of the person petitioning for censorship.