I hope that our readers have read the response of NewsGuard’s Gordon Crovitz to my recent criticism of the company’s rating system for news sites. He makes important points, including the fact that the company has given high ratings to conservative sites and low ratings to some liberal sites. I have mutual friends of both Gordon and his co-founder Steve Brill, who have always sworn by their integrity and motivations. I do not question Gordon’s account of past ratings for sites.
However, I also welcome the opportunity to further this discussion over media rating systems and to explain why I remain unconvinced by his defense. It is a long overdue debate on the use and potential misuse of such systems.
As a threshold matter, I want to note that I am aware of conservative sites reviewed by NewsGuard that have been given favorable ratings. That is a valid distinction from past rating sites like the Global Disinformation Index (GDI).
Moreover, while I noted that NewsGuard has been accused of bias by conservatives and is being investigated in Congress, my primary objections are to rating systems as a concept for media sites. Before addressing that opposition, I should note that I still have concerns over bias from the email that was sent me, particularly just after a column criticizing the company.
Now to the main concern.
A Shield or a Sword?
In his response to me, Gordon argues that “I would have thought, including based on your recent book, that you’d especially welcome an accountable market alternative to censorship.”
I disagree with Gordon’s suggested dichotomy. As I argue in the column, rating systems are arguably the most effective means to silence opposing voices or sites. These systems are used to target revenue sources and have been weaponized by the current anti-free speech movement. They are used more as a sword than a shield by those who want to marginalize or demonetize a site.
We have seen such campaigns targeting various sites and individuals, led by political groups opposed to their viewpoints, including figures such as Joe Rogan. This includes Elon Musk and X after the reduction of censorship systems and the release of the “Twitter Files.” After being targeted by these campaigns for years, rating systems have been denounced by Musk as part of an “online censorship racket.”
Moreover, the use of private entities like NewsGuard is precisely what makes the current movement so insidious and dangerous. Whether by design or by default, rating systems are effective components of what I have described as a system of “censorship by surrogate.”
What NewsGuard is attempting is potentially far more impactful for the funding and viability of websites. Rather than an alternative, it can be an avenue for censorship.
I have also written about my concerns with the Global Alliance for Responsible Media and its use of rating systems to deter advertisers for targeted sites. The group states that it “unites marketers, media agencies, media platforms, industry associations, and advertising technology solutions providers to safeguard the potential of digital media by reducing the availability and monetization of harmful content online.”
As the column discusses, NewsGuard seeks to position itself as a type of Standard & Poor’s rating system for media. The role would give the company unprecedented influence over the journalistic and political speech in America. The rating can be used to discourage advertisers and revenue sources for targeted sites. Just as S&P scores can kill a business, a media rating could kill a blog or website.
That is an enormous amount of power to be wielded by any organization, let alone a for-profit enterprise started by two self-appointed monitors of media. That is not meant to disparage Gordon and Steve, but to acknowledge that this is not just a hugely profitable but a hugely powerful enterprise.
It is also not a criticism of the founding principles. We have seen many organizations that began as faithful to principles of neutrality only to see those principles corrupted with time. Indeed, as we have previously discussed, the very principles of objectivity and neutrality are now rejected in many journalism schools.
The Criteria
While NewsGuard insists that its criteria is completely objective and neutral, that does not appear to be the case. The site’s standards include key determinations on whether some sites run statements that NewsGuard considers “clearly and significantly false or egregiously misleading.” (That appears part of the most heavily weighted criteria for credibility at 22 points).
The staff will determine if it believes that a site shows a tendency to “egregiously distort or misrepresent information.”
The staff decides if information is false and, if it is considered false by NewsGuard, whether the site “identifies errors and publishes clarifications and corrections, transparently acknowledges errors, and does not regularly leave significant false content uncorrected.” Thus, if you disagree with the claims of falsity or view the statement as opinion, the failure to correct the statement will result in additional penalties.
The site will also determine if it finds the sources used by a site to be “credible” and whether “they … egregiously distort or misrepresent information to make an argument or report on a subject.”
If the site decides that there are errors, it will lower ratings if the site does not “transparently acknowledges errors, and does not regularly leave significant false content uncorrected.”
The company pledges to combat “misinformation” and “false narratives.”
We have seen mainstream media use these very terms to engage in highly biased coverages, including labeling true stories or viewpoints “disinformation.”
Given these terms and the history of their use in the media, NewsGuards assurances boil down to “trust us we’re NewsGuard.” GDI made the same assurances.
This is not to say that some of these criteria cannot be helpful for sites. However, the overall rating of media sites is different from Standard & Poor’s. Financial ratings are based on hard figures of assets, earnings, and liabilities. “Liquidity” is far more concrete and objective than “credibility.” What NewsGuard does is fraught with subjectivity regardless of the motivations or intentions of individual raters.
The Res Ipsa Review
The inquiry sent to this blog reflects those concerns. The timing of the inquiry was itself chilling. I had just criticized NewsGuard roughly a week earlier. It is not known if this played any role in the sudden notice of a review of Res Ipsa.
One inquiry particularly stood out for me. The reviewer informed me:
“I cannot find any information on the site that would signal to readers that the site’s content reflects a conservative or libertarian perspective, as is evident in your articles.Why is this perspective not disclosed to give readers a sense of the site’s point of view?”
The effort of NewsGuard to label sites can have an impact on its ratings on credibility and transparency. Yet, sites may disagree with the conclusions of NewsGuard on their view of the content. What may seem conservative to a NewsGuard reviewer may be less clearly ideological to the host or blog.
Moreover, despite noting that it asked MSNBC to state its liberal bias, it is not clear if the company has suggested such a notice from many other sites from NPR to the New Republic. For example, is Above the Law supposed to warn readers that it takes a liberal perspective and regularly attacks conservatives? What about other academic blogs like Balkinization?
The point is not to say that they should be required to label their own views (though some sites choose to do so) but to ask whether all sites are asked to do so. If not, when is this demand made for sites? For some reviewers, a liberal perspective may simply seem like stating the obvious or unassailable truth.
Labeling
In fairness to NewsGuard, we all often engage in labeling as part of our discussions — both labeling ourselves and others. For example, I often acknowledge that I hold many libertarian views. However, I continue to write columns that run across the ideological spectrum and I continue to be attacked from both the right and the left for those columns.
Identifying yourself as a libertarian does not convey much information for readers. Many readers have erroneous views of libertarians as a monolithic group. (The public high school teacher of one of my kids told the class that libertarians were just conservatives who did not want to call themselves Republicans). In actuality, it is a group that runs from liberal to conservative figures who maximize individual rights. Labeling your site as libertarian is about as helpful as saying that it is utilitarian.
The suggestion in the email is that readers should be informed that anything they read is coming from a libertarian or conservative on the site. Yet, most law professor blogs are very liberal, but do not make the same type of warning.
We often discuss these labels in judging the diversity of faculties. Yet, that is based largely on surveys of professors self-identifying or the political registration of academics. It is admittedly a blunt tool, but there is little debate that faculties around the country are overwhelmingly liberal. Indeed, even sites like Above the Law have strived to defend “predominantly liberal faculties” as just reflecting the fact that most conservatives are simply wrong on the law.
There is always an overgeneralization in the use of such labels, but we try to take that into consideration in discussing the overall lack of diversity of viewpoints on campuses today.
Conclusion
Rating media sites is vastly different. You are often relying on the views of the reviewers that may be challenged by the site. Postings that challenge popular narratives are often denounced as false or disinformation by critics.
I am particularly concerned over the reported government contracts given to NewsGuard by the Biden Administration as well as agreements with teacher unions to help filter or rate sites. The Twitter Files have shown an extensive system of funding and coordination between agencies and these companies. The funding of such private rating or targeting operations is precisely what I have warned about in congressional testimony as a type of “censorship by surrogate.” The government has been attempting to achieve forms of censorship indirectly that it is barred from achieving directly under the First Amendment.
Consider those bloggers and scientists who were censored and denounced for voicing support for the lab theory on Covid 19 and other subjects from the efficacy of masks to the need to shutdown schools. They spent years having mainstream media figures denouncing them for refusing to admit that they were spreading disinformation or conforming to general views on these issues.
The Washington Post declared this a “debunked” coronavirus “conspiracy theory.” The New York Times’ Science and Health reporter Apoorva Mandavilli wascalling any mention of the lab theory “racist.”
Political and legal commentary are rife with contested opinion over the facts and their implications. Having a company sit in judgment on what is fact and what is opinion is a troubling role, particularly when the rating is used to influence advertisers and financial supporters.
Once again, there are many people on the other side of this debate who have good-faith reasons for wanting a standardized set of criteria for news sources and commentary sites. I simply believe that this is a degree of influence that is dangerously concentrated in a small number of groups like NewsGuard.
Jonathan Turleyis the Shapiro Professor of Public Interest Law at George Washington University. He is the author of “The Indispensable Right: Free Speech in an Age of Rage” (Simon & Schuster).
N.B.: After this response ran, NewsGuard wrote me that Above The Law actually was marked down for failing to clearly delineate between news and opinion. It further said that the New Republic acknowledges its liberal take, so there is no issue on labeling. What is not clear is whether every site, including academic blogs, are asked to label themselves and who makes that decision on what label should apply.
