Helping Truth with Its Boots: Accreditation as an Antidote to Fake News
abstract. A generally accepted, objective way to differentiate reliable generators of accurate information from purveyors of “fake news” would take significant positive steps toward combating its spread. Other spheres of our economy successfully use private accreditation systems to distinguish quality products and services from those that are inferior or fraudulent. This Essay considers online platforms’ current approaches to fake news before contemplating how a similar nongovernmental accreditation system might work to distinguish reliable journalism from disinformation.
“A Lie would travel from Maine to Georgia while Truth was getting on his boots.”1
Early in the morning on June 14, 2017, a gunman opened fire as Republican members of Congress were practicing at baseball fields in Alexandria, injuring five people.2 Within minutes, misinformation began circulating across the Internet.3 InfoWars accused the shooter of being “radicalized” by mainstream media attacks on President Trump.4 Alex Jones claimed that CNN “endorsed” the attack.5 A tweet blaming the attack on a nonexistent call by former Attorney General Loretta Lynch for “more blood on the streets” was widely retweeted and liked.6 All of this patently false information soon spread through social media like wildfire. The episode was not unique,7 and it demonstrates how the decline of the newspaper industry and the rise of social media websites as the primary distributors of news enables the rapid dissemination of false information.
The issue of “fake news” is of course multifaceted, and even its definition can be slippery.8 For purposes of this Essay, we are concerned with information that is knowingly false or intentionally misleading, presented in forms that make it difficult to distinguish truth from fiction. Most importantly, this includes fake news propagators that mimic the appearance of a journalistic endeavor.9 When purveyors of fake news imitate traditional media, they give their audience the impression that they function within the same constraints as traditional media. Some even refer to themselves as “the new yellow journalists,”10 implying that their activities are simply sensationalist departures from the core of journalistic norms. But their endeavors have no basis in the traditions of journalism.
These “news” stories are not fact-checked. Their sources (to the extent that they use any) are not vetted. No attempt is made to contextualize information accurately or to obtain comments from the subjects of their articles. Such disseminators of fake news publish without regard to truthfulness, accuracy, or fairness, but they vouch for their veracity by copying the styles and tone of traditional reporting.11 This kind of fake news is not simply false information masquerading as truth. It is false information masquerading as the verified truth—a misrepresentation that is designed to inspire trust even as it deceives.
This Essay engages in a thought experiment. Rather than focusing on ways to target false information and limit its spread, it asks whether it is possible to identify and promote the spread of truthful information. Many industries have developed private accreditation systems to establish industry standards, limit fraud, and ensure the quality of services or products. Is there a similar way to “accredit” reliable providers of original reporting? Here, we take a first stab at conceptualizing an accreditation system that would distinguish reliable sources of information based on the processes they employ to verify the accuracy of their reporting.
In Part I, we review recent efforts by social media platforms to curb the spread of fake news. Part II briefly addresses the absence of a viable solution to the problem through direct government regulation or private litigation, before turning to how a system might be developed to verify sources of reliable reporting based on accreditation methods used today in other fields. Part III then contemplates how such accreditation might help social media platforms, advertisers, and consumers to combat the impact of fake news. Part IV argues for the feasibility of this sort of accreditation system.
Platforms such as Facebook, Google, and Twitter have attempted to address fake news by identifying and curbing its spread, increasing the prominence of reliable news, and better equipping their users to distinguish between the two. This Part discusses some major initiatives these platforms have enacted in pursuing these goals.12
Both Google and Facebook have taken steps to disadvantage purveyors of fake news by manipulating their algorithms to “demote low-quality content”13 that contains “little substantive content and is covered in disruptive, shocking or malicious ads.”14 Both platforms have also partnered with third-party fact-checkers such as The Associated Press (AP), PolitiFact, and Snopes. On Facebook, fact-checkers evaluate news stories flagged by the site’s users and label stories they deem inaccurate as “disputed content.”15 The labels then follow the stories wherever they are posted or shared.16 Google allows publishers to label their stories with a “Fact Check” tag that indicates the information has been verified by a fact-checker.17 Google has also tweaked its search algorithm to “surface more authoritative pages” by examining pages’ expertise, authoritativeness, and trustworthiness.18
Effectiveness of these efforts is limited since they inevitably provoke accusations of censorship and partisanship.19 Changes to algorithms have often been met with suspicion and criticism over lack of transparency.20 These critiques may well explain why Facebook and Google rushed to outsource fact-checking to third-parties, rather than handling the decisions in-house and risking being seen as “arbiter[s] of truth.”21 But third-party fact-checkers are not immune to attacks for political bias. Conservatives immediately attacked Facebook’s fact-checking initiative as a “disaster for news coverage,” contending that it attempts to restore “gatekeepers who have a bias as the ultimate arbiters of truth.”22 Moreover, even in a world with reliable and accepted fact checking, a fact-checked report will often not reach the same audience that received the original misinformation.23
Platforms have also tried to combat the spread of fake news by promoting the dissemination of real news. Facebook has established a Journalism Project, which seeks to collaborate with news outlets to develop new products that “better serve the needs of people on Facebook” and to help journalists build audiences online.24 It has also redoubled its efforts to incentivize the production of news content that is more compatible with the click economy by implementing programs with advantageous revenue splits. Watch, an initiative meant to spur the creation of original video content, offers producers fifty-five percent of ad revenue.25 Facebook is also revamping Instant Articles, a tool that allows Facebook to host (rather than link to) original news content, making such content faster to load and more visually appealing.26 Publishers already receive seventy percent of ad revenue from Instant Articles; Facebook is now trying to sweeten the deal by allowing them to offer subscriptions through the tool.27
Tools like Instant Articles and Watch represent a devil’s bargain. In exchange for a larger share of the advertising pie, content producers must sacrifice revenue from ads hosted on their own sites, and also lose on branding, access to user data, and paid subscriptions.28 Even increased traffic to news sites through social media does not necessarily translate to larger profits. In the past, such traffic has driven down profits from digital advertising.29 Underlying all of this is the worry that by increasing reliance on platforms like Facebook, news organizations leave themselves at the mercy of the business and editorial decisions of social media platforms.
A third response to the impact of misinformation has been to increase the news literacy of social media users. Sixty-two percent of U.S. adults get news on social media,30 where information circulates through peer-to-peer sharing,31 often making an assessment of reliability difficult. To foster greater online news literacy, Facebook released a list of ten tips to spot fake news, which appeared at the top of users’ newsfeeds in fourteen countries for a few days in April 2017.32 It has also committed to working with the News Literacy Project on a series of public service announcements designed to provide users with tools to become “informed consumers of news.”33
But fostering news literacy is not the same as changing a community policy. A few postings or videos are unlikely to substantially modify users’ news consumption habits or sustain a change in community expectations. The peer-to-peer nature of social media platforms like Twitter and Facebook means that they are not one community of users, but several, with very different news consumption habits and beliefs about the reliability of the press.34
These varied initiatives to address fake news represent a good start, but they fail to address both the underlying mistrust of the media and the destruction of the economic base that sustained legacy news organizations. These factors have hindered the spread of reliable information and opened the door for fake news to wreak havoc on our public discourse, and it is not clear that any of the current platform initiatives can successfully overcome them.
If platforms are unable or unwilling to meaningfully address the problem of fake news, what are the other options? Any direct governmental regulation must be consistent with the First Amendment, but it is a nearly impossible definitional task to distinguish information that might properly be regulated by the government without violating constitutional norms. Most fake news occurs in the realm of public discourse—the arena where our speech is most protected.35 Laws that regulate expression based on content are “presumptively invalid,”36 and even falsity does not automatically remove the constitutional protections afforded to public discourse.37 Rightly so—no one wants the government to act as a “Ministry of Truth” deciding what political speech the public should hear or believe.38 Any governmental effort to regulate the spread of fake news would therefore run headlong into nearly insurmountable First Amendment constraints, raising troubling definitional, vagueness,39 overbreadth,40 and viewpoint discrimination issues.41
Pursuing private rights of action against the purveyors of fake news, or creating new ones, also seems unlikely to provide a workable solution. The First Amendment, of course, cabins libel claims to some extent,42 and libel litigation generally is ill-suited to remedy fake news. While threats of libel suits have occasionally prompted purveyors of fake news to retract stories,43 litigation has not stopped those same outlets from continuing to produce fake news. Moreover, lawsuits can normally address misinformation only after it is published, at which point people have already consumed and internalized it. Litigation also takes time and money—placing an undue burden on those who do not have the resources to take up these fights.
Instead of approaching the fake news problem from the perspective of containment, we propose focusing on facilitating the spread of accurate information in ways that extend beyond what platforms can do on their own. Could an accreditation system be developed to identify reliable generators of accurate information, promote the distribution of their work, and support them economically?
Our proposed accreditation approach would focus on the conduct and standards used to produce a story, rather than the accuracy of a given report—it would distinguish information by whether it comes from a source that employs reasonable methods to verify and contextualize information. This would not, of course, mean that misinformation would always be identified. Falsity and inaccuracy can and do coexist with even the best journalism, as evidenced by the inaccuracies that inevitably creep into reporting by the Washington Post, New York Times, or Wall Street Journal.
Creating a system that distinguishes sources of information by their standards of conduct nonetheless might provide a basis for differentiation that is less partisan, easier to monitor with objective criteria, and more readily accepted than a post-hoc system of fact checking. After all, a “bright line” distinction exists between reliable news outlets and many purveyors of misinformation, both in terms of the processes they use to generate content and in terms of the standards of accuracy they strive to achieve. And a system of voluntary accreditation based on a commitment to certain standards and practices would not be the same as licensing the press—no government approval would be involved, no regulatory restraint would be imposed on non-accredited content, and no assessment of the value of any particular content would be made.
Looking to conduct to distinguish reliable journalism is not a novel idea. In 1910, W.E. Miller proposed the nation’s first journalistic code of ethics, which has been credited with helping to eliminate local practices of “yellow journalism,” including misquotations, fabricated interviews, and fake illustrations.44 A few years later, a group of newspaper editors founded the American Society of Newspaper Editors to defend themselves against accusations of shoddy practices.45 One of the group’s first acts was to articulate a code of ethics46 that was later adopted by the Society of Professional Journalists (SPJ).47 Major news outlets across the country now abide by substantially similar ethical codes.48
It is not necessary to articulate a single set of universal best practices in order to use a content generator’s standard practices as a proxy to identify generally reliable news sources. Indeed, defining newsgathering as a fixed process would be anathema to many journalists. Not only do newsrooms have their own cultures, but each journalist has her own process, and each story may require a different approach. The same is true of colleges and universities, but a set of sufficiently specific education standards has nevertheless been developed to provide a meaningful basis for accreditation of institutions of higher education.49
Accreditation under this approach might be based on a few fixed standards:
First, a generator of original content could be required to ascribe to universal principles that define the goals and ethical practice of good journalism. These should not be difficult to establish. SPJ, for example, publishes a one-page code of ethics centered around four goals: seek truth and report it; minimize harm; act independently; and be accountable.50The Radio Television Digital News Association articulates three guiding principles for its code of ethics: truth and accuracy above all; independence and transparency; and accountability for consequences.51
Second, a content generator could be required to commit to a generally reliable discipline of verification, that is, to develop and systematically apply methods of testing information before publication that are both rigorous and transparent.52 These methods might include seeking out primary or multiple sources when feasible; disclosing as much information about sources as possible; attempting to elicit comments from subjects and relevant viewpoints; accurately contextualizing information; acknowledging what key facts are unknown or omitted; and publicly correcting errors.
Third, a content generator could be required to articulate and publish its own standard practices, which must advance the universal general principles and be considered reasonably rigorous by similarly situated news outlets. For larger national outlets, for example, these might take the form of the New York Times’ Professional Guideline Documents53 or the AP’s online statement of its “News Values and Principles.”54
Such a layered approach to accreditation would ensure enough consistency for good journalistic practices to be identifiable, while granting flexibility and independence to individual journalists.
With these broad goals and standards in mind, we can look at accreditation in other contexts to develop the parameters of a potential system for accrediting news generators. Many U.S. industries employ some form of self-regulation carried out by a private third-party entity. As of 2001, over 120 nongovernmental organizations provided some form of product certification.55 Accreditation is ubiquitous in professional fields like medicine, law, and financial management. Existing systems of accreditation fall into three general categories: institutional and programmatic; product-based; and personal. This Section will consider the benefits of and problems with each type as applied to the news industry.
Institutions of higher education are largely accredited by private regional and national accrediting bodies that have developed evaluation criteria and standards in consultation with representatives from educational institutions.56 Accreditation can happen on the institutional or program level.57 In order to become accredited, an institution prepares an in-depth self-evaluation to measure itself against the standards established by the accrediting body.58 The accrediting body then sends a team of peer evaluators to conduct an on-site evaluation.59 If it is satisfied that the institution or program meets its standards, it grants accreditation.60 Throughout the period of accreditation, the accrediting body monitors and reevaluates the institution or program periodically for continued compliance with the accreditation criteria.61
This type of accreditation system would seemingly translate well to the news industry. A third-party entity working with representatives from across the news industry or within specialized fields of news coverage could develop an appropriate set of principles and basic best practices, as well as other requirements for certification.62 Following the higher-education model, a news organization applying for certification might submit a self-evaluation and a copy of its written policies. The accreditor would then arrange for a committee of peer journalists to evaluate the news outlet and decide whether to grant accreditation. Where news organizations are part of larger media conglomerates, accreditation might be available at the newsroom or news division level. After accreditation, a news organization would be eligible to designate its original fact-based reporting with an accreditation watermark or label indicating that the reporting originated from an accredited source using best practices. This accreditation label could be easily recognizable and follow an article across social media.
A third-party accreditor would have several avenues for monitoring and enforcing certification. As with educational institutions, periodic reaccreditation could be required.63 The accreditor might also review a news organization’s practices periodically to ensure they are consistently compliant with accreditation standards. In case of a major breach of best practices or a large number of complaints, the accreditor might conduct follow-up site visits or require new self-evaluations, in which the news organization would be obliged to address the issues raised. Penalties for noncompliance by an accredited organization could also be developed, such as fines, probation, or removal of accreditation. Public pressure could also be used to encourage compliance by transparent disclosure of investigations into complaints.
Many news organizations have arrangements with “wholesale” generators of news content such as the AP, Reuters, or ProPublica.64 In addition, many online news sites aggregate articles reported elsewhere, but often with unique headlines or graphics.65 A third-party accreditor would need to decide how to treat information from an accredited source when original material is added or content is altered by others. An AP article whose text or headline has been altered with misinformation could be considered fake news, even if the original article conformed to the highest ethical standards.66 Any accreditation designation would need to be removed from altered content, unless the modifying site were itself independently accredited.67
Institutional or programmatic accreditation presupposes a level of organizational structure that is not universal among news startups. Startups can begin as loose associations of independent journalists sharing a platform,68 and many independent journalists simply publish original pieces on their own blogs.69 Accreditation should not disadvantage these types of news generators; doing so would only stifle innovation. One way to address this problem would be to offer certification for individual journalists who adhere to defined standards and practices.
There are many examples of voluntary professional certification programs, some offered by universities or trade schools70 and others by private certifiers.71 Individual certification would necessarily differ from institutional accreditation in both procedure and requirements. Where institutions would be required to develop written policies, individual certification might require completion of online tutorials or examinations and a pledge to follow certain practices. To maintain certification, an individual might be required to generate a specified amount of original reporting, periodically verify adherence to the standards and practices, and perhaps undertake a CLE-type continuing education process or periodic reexamination on standards and procedures.72 A mechanism could also be developed to monitor compliance and to investigate shortcomings and complaints, analogous to disciplinary committees that enforce ethical practice among attorneys and other professionals.
Again, with such a certification program, the work product of a certified journalist could be marked technologically in a way that alerts a reader that the article was prepared by a journalist who adheres to strict standards in gathering and reporting the news.
While institutional accreditation is paramount as a way of comprehensively certifying reliable news sources, a labeling system could also operate to distinguish original fact-based reporting from other types of content on an article-by-article basis. Almost all news organizations publish material beyond fact-based news content. Opinion and commentary are nearly ubiquitous in newsrooms, but they are not products of the same type of journalistic processes as original, fact-based reporting. Employing a system whereby an accredited institution certified only its original, fact-based reporting would offer another, perhaps more refined and useful, approach.
Product-based labeling has been used to advertise reliability and weed out fraudulent products in many industries—think of the “Good Housekeeping Seal of Approval.”73 One successful example of this approach is kosher food labeling. Kosher labeling arose in the early twentieth century in response to rampant fraud and ineffective government oversight.74 Before the rise of kosher certifiers, forty to sixty-five percent of meat advertised as kosher in New York City was not.75 As with fake news, purveyors of fraudulent kosher meat took advantage of the fact that non-kosher meat was cheap to produce relative to kosher meat, as well as difficult or impossible for consumers to distinguish. Kosher certifiers marketed themselves as a relatively “low-cost, highly effective marketing strategy.”76
The initial success of the kosher labeling system was due mainly to a captive and vigilant core consumer base. But today, over forty percent of new food products carry the kosher label, even though less than two percent of Americans are Jewish and only a fraction of Jews keep kosher.77 The kosher label has come to be associated with health and cleanliness among many consumers.78 The power of the kosher label is such that the kosher food industry was worth over seventeen billion dollars as of 2009, and the business of the largest kosher certifier is growing by ten percent each year.79
The news industry has not historically had a vigilant core consumer base in the same sense as the kosher food industry. After the advent of the Internet, news consumers slowly stopped subscribing to print publications when they were able to get their news for free on line.80 But after the 2016 election and the proliferation of fake news, more consumers are buying subscriptions to national newspapers with reputations for good journalism, notwithstanding their access to free news sources.81 Fact-checking and the ethics of anonymous sourcing have become regular topics of public discourse.82 News consumers are engaging with journalists so extensively on social media that New York Times publisher Arthur Sulzberger proclaimed them the “modern watchdog” when he eliminated the paper’s public editor position.83
But the voices of news consumers who are actually familiar with the intricacies of responsible newsgathering and reporting remain few and disparate, even as interest in receiving reliable information grows. A labeling system based on an articulated set of principles and best practices could provide diligent news consumers with a basis for differentiating reliable information. This, in turn, could perpetuate the reputation of the label among less attentive consumers, as the kosher food labeling process has done.
Any of these approaches to accreditation—institutional or programmatic, individual, or product-based—would require commitment to adhere to formal ethical standards, something that some media lawyers would advise against on the grounds that formalized standards “might be held against [journalists] in court, a stance that could gain no public trust or credibility.”84 It is true that articulating a code of ethics in some cases could aid plaintiffs asserting claims for defamation, invasion of privacy, and related torts. Any aberration from an outlet’s stated practices could be offered as proof of negligence. Politicians and media critics might likewise point to deviations from stated standards to challenge the legitimacy of stories of which they disapprove. But these concerns presuppose a level of credibility that much of the press does not currently enjoy.85 Any advantage formalized standards might give adversaries in litigating grievances would be isolated and specific, which likely would constitute an improvement over the broad and vague attacks on the news industry that currently exist.
As this brief review of possibilities suggests, it is certainly possible to conceive of an accreditation process that might provide an effective way to differentiate reliable sources of news from purveyors of misinformation. What benefits such an accreditation system might yield, and how such a system could be economically sustained, pose equally thorny questions. We address these issues in the next Part.
An effective mechanism to identify reliable generators of original content might be used in several ways to minimize the impact of fake news, from promoting the dissemination of reliable information and connecting accredited sources to funding, to educating news consumers to differentiate information from reliable and unreliable sources.
As discussed above, Facebook, Google, and other social media platforms have already confirmed that they can, and are willing to, preference specific news sources through their algorithms. This capability could prove a more powerful tool against fake news with a viable system of accreditation. Platforms could “uprank” stories from accredited outlets without endorsing a particular news organization or story. Post-hoc fact-checking might then become less central to the fight against fake news. This approach would especially be useful in scenarios where platforms enjoy limited or no immunity from intermediary liability.86
Upranking based on accreditation could also aid news organizations. As they become increasingly reliant on traffic from social media sites, they become more vulnerable to unpredictable algorithmic changes. Earlier this year, several small and midrange publishers reported seeing a sharp decline in the reach of their posts on Facebook even as their pages gained more followers.87 Articles posted by the Chicago Tribune, for example, went from a reach of twenty-five to fifty thousand people in the first quarter of 2016 to reach ranges as low as four thousand to six thousand people during the first quarter of 2017.88 Accreditation could provide platforms with a content-neutral way to uprank the stories of financially-stressed smaller and start-up publications, sites that might particularly benefit from algorithm bumps they would not receive without accreditation.
Accreditation might also be used to starve out fake news sites by directing more funding to accredited news outlets. Advertisers might refuse to advertise on unaccredited websites, and social media platforms might agree to increase an accredited publisher’s share of revenue for ads or to release more user data on labeled articles.
Accreditation would not solve the problem of how Facebook and Google should fund original news content, but it could provide a mechanism to ease a solution.
Accreditation could also signal to news consumers a way to judge the source of information they receive on social media, making it more difficult for fake news to pass the “smell test.” As discussed, the educational task is daunting.89 To develop a shared concept on social media of what is reliable, it will be necessary to foster a shared understanding of how reliable information is generated and how it can be identified. An accreditation system would identify content from a wide variety of sources with an easy-to-understand label. These labels would allow the concept of accreditation to follow wherever the article is shared, penetrating most peer-to-peer networks.
Developing the full contours of an accreditation system is beyond the scope of this Essay, but even a cursory review suggests numerous ways that a reliable, content-neutral system of accreditation could improve the flow of information that actually reaches an audience and the financial fortunes of those who generate that content.
Under any scenario, the potential of an accreditation process can never be achieved if no viable economic base exists to support the necessary accreditation infrastructure. Thinking through possible funding systems in a comprehensive way requires further research on news organizations’ revenue flows and technological constraints. If an accreditation system could come to carry meaning in the market for news, however, there should be a way to capture the economic value of accreditation, both to benefit those who generate original content using accredited methods and to support the costs of the accreditation itself. Several forces of self-interest might be harnessed to produce a viable accreditation system in the news industry.
First, it is in the platforms’ economic interest to promote the distribution of reliable information over misinformation and propaganda. Users object to being constantly duped, and advertisers have little interest in being tied to misinformation. Regularly providing reliable information may even save platforms money by bolstering the effectiveness of their other initiatives. Platforms should be willing to accept accreditation as a factor to be weighted in their algorithms to steer readers to reliable news sources. Their interest in delivering reliable information could also induce the platforms to adjust advertising revenue splits to direct a greater share of advertising dollars to accredited generators of reliable information.90
Second, many advertisers have a direct interest in not being seen as supporting the spread of misinformation and propaganda.91 Their reputational concerns could similarly incentivize them to direct their online ads dollars to accredited content.
Third, generators of information would have an interest in obtaining and maintaining accreditation if such tangible economic benefits from accreditation were to manifest. The voluntary system of receiving a rating from the Motion Picture Association of America represents one model of how economic self-interest can support the costs of administering a process of accreditation.92 Just as a movie rating provides a reliable indicator of content that translates into increased profits at the box office, an increase in web traffic, a more favorable revenue split, and a more dedicated pool of potential advertisers should serve to increase revenue for the generators of accredited material. Similarly, just as universities underwrite the costs of their accrediting agencies because accreditation is essential to their continued viability,93 news organizations benefiting from increased revenues derived from accredited content would have a clear incentive to underwrite the costs of accreditation.
These preliminary musings raise more questions than they answer, and leave much to be explored. Even a fully functioning system of accredited sites would not eliminate fake news. Many means of online manipulation can shift opinions, purchases, and voting preferences without people’s knowledge. For example, it has been estimated that biased search results can shift voting preferences as much as eighty percent in some groups,94 and autofill search suggestions can dramatically impact the nature of people’s searches and the information delivered to them.95 Nevertheless, the concept of accreditation deserves consideration as one meaningful way to address a problem that appears difficult to solve through other means.
Far from being a journalist’s “license,” a system of voluntary accreditation could provide a readily identifiable and meaningful indicator of the reliability of the source. One might think of it as the press pass for the digital age.
Anna Gonzalez is a member of the Yale Law School J.D. Class of 2018. David Schulz is the Floyd Abrams Clinical Lecturer in Law and Senior Research Scholar in Law at Yale Law School. He is also Senior Counsel in the law firm Ballard Spahr LLP. We are grateful to Hannah Bloch-Wehba for her very helpful comments as this Essay took shape. We would also like to thank Allison Douglis and Meenu Krishnan for their thoughtful edits. All errors are our own.
Preferred Citation: Anna Gonzalez & David Schulz, Helping Truth with Its Boots: Accreditation as an Antidote to Fake News, 127 Yale L.J. F. 315 (2017), http://www.yalelawjournal.org/forum/helping-truth-with-its-boots.