The Heritage Foundation, America’s most prominent conservative think tank, recently published Combating Big Tech’s Totalitarianism: A Road Map by research fellow Kara Frederick. The ambitious paper outlines complaints against “Big Tech” and proposes policy responses ranging from antitrust reform to amending Section 230. Amid ongoing debates on antitrust, free speech, and surveillance, research from a conservative perspective on these issues would be valuable. Unfortunately, rather than providing clear analysis, Frederick’s paper is marked by imprecision, factual errors, and vague political rhetoric. The most substantive recommendations ultimately represent a retreat from long-established conservative policy principles consistent with a free economy and free society.
Complaints and Definitions
Almost half of the paper is a catalog of anecdotal complaints against “Big Tech.” Although mentioned in the title and used throughout the paper, “Big Tech” is defined in a vague way, as “a loose compilation of companies that include the “Big Five” tech companies—Alphabet (Google), Amazon, Apple, Meta (Facebook), and Microsoft—along with other tech companies like Twitter, TikTok, Snap, and Netflix whose cultural and political impacts on public debate are significant.”
This vagueness makes addressing the paper a challenge. The term “Big Tech” includes Twitter, which looks like a very small fish in a sea where the Amazon, Google and Facebook leviathans swim. Only a minority of Americans are active Twitter users and, unlike Facebook and Google, Twitter is not competing in markets as diverse as those for smart speakers, e‑commerce, and cloud gaming. While it might be true that Twitter has a significant impact on culture and politics along the Acela Express route, most Americans debate these issues without Twitter accounts.
The inclusion of Netflix is likewise bizarre. The company undoubtedly affects culture, but it is not a monopoly streaming service. Disney owns the intellectual property associated with its classic animated movies as well as Star Wars and Marvel, arguably among the most culturally influential fictional universes in living memory. Disney’s broader streaming service has almost 120 million subscribers, while Hulu, also owned by Disney, has a little more than 40 million subscribers. Yet Disney is not included is Frederick’s definition, and nor is Wikipedia, one of the most visited websites in the world that has a clear influence on our contemporary discourse.
If the criteria for inclusion are unclear, what is obvious is that Frederick is unhappy with how the companies she does consider members of the “Big Tech” club handle user-generated content. She lists many anecdotes related to conservative speakers on a variety of platforms being suspended, banned, or having content removed. An examination of every content moderation complaint listed in the paper would require a standalone blog post. However, one example is worth noting.
The paper cites this article, which Frederick summarized as showing that: “Twitter suspended users in the ‘conservative’ category at a ratio of 21-to‑1 compared to ‘liberals.’ ” The author of the cited paper made the complete dataset available. It includes a list of 22 prominent Twitter accounts that expressed a preference in the 2016 presidential election and were temporarily or permanently suspended. One of the tweeters supported Hiliary Clinton. The others supported Donald Trump.
Who are among the 21 Trump supporters? The list includes the white supremacist Richard Spencer, the former KKK Grand Wizard David Duke, the anti-semite Tila Tequila, and the American Nazi Party. To see a Heritage Foundation paper citing research that includes such people and institutions as belonging in the “conservative” category because they back Donald Trump is, to put it mildly, jarring.
Most of the other claims are more marginal instances of private sector content moderation that upset conservatives. We as private individuals might have differing opinions on whether the decisions were justified or harsh according to the host’s rules, or indeed differ on what those privately-enforced rules should be. What is less clear is why Frederick thinks this content moderation requires a public policy response.
Private companies make regrettable decisions all the time in every market. For years, we would have expected for a traditional American conservative response to be one that embraces the power of market competition and rejects government intervention. Yet for reasons that are not made explicitly clear, the private content moderation decisions of technology companies are considered the kind of harm where market competition is untenable.
Perhaps to justify corrective intervention, Frederick goes on to complain about “Big Tech” already working in conjunction with the government, citing White House Press Secretary Jen Psaki’s comments about the White House flagging COVID-19 posts for Facebook to remove. But conservatives and libertarians should be reassured by the fact that those working for the most powerful person in the world can only ask Facebook to make specific content moderation decisions. The First Amendment ensures that the government cannot force Facebook to post or remove legal content (such as COVID-19 misinformation).
Yet Frederick seems to think there is a risk of government banning certain platforms entirely, writing: “Government control of the ability to exist on any private digital platform would undermine the creation of competitors such as RightForge, GETTR, Odysee, and Rumble.” However, the government does not control the ability of a social media company to exist in America, and that is not going to change absent a revolutionary revision in First Amendment jurisprudence.
The mention of Rumble and Odysee is especially notable. Frederick mentions Psaki’s comments on the podcaster Joe Rogan (who she incorrectly identifies as a Spotify employee) as an example of government pressure on private business. Yet only a few days ago Rumble and Odysee both offered to host Rogan’s podcast. Perhaps these offers are publicity stunts, but it nonetheless remains the case that Rogan has plenty of hosting options if his relationship with Spotify collapses, even if the White House does not like his content.
Among Frederick’s other government-related concerns are those associated with Big Tech surveillance, an issue Cato Institute researchers have been discussing for years. The Cato Institute archive is full of articles on the American government snooping on personal communications and social media posts. We have hosted numerous events focused on surveillance, and the Cato website also hosts a timeline of the history of American surveillance.
Given that background, we were surprised to see Frederick write, “Unfortunately, the United States is no longer immune to this impulse to restrict individual liberties through tech. Previously, governance practices in the United States provided a safeguard against the potential for technology abuse: sufficient rule-of-law protections, a culture of openness, a free press, an independent judiciary, and an engaged citizenry. But that bulwark is increasingly in jeopardy.” The history of the US government using technology to restrict liberties is as long as the history of the country and is not exclusive to the age of “Big Tech.” It is surprising to read a Heritage scholar arguing that until recently the rule-of-law protections and the free press acted as a bulwark against the government using new and emerging technologies to spy on American citizens.
As well-grounded as very real concerns with mass surveillance are, Frederick’s solution misses the mark. That law enforcement regularly spies on social media networks is not an argument for increased regulation of the networks. Those concerned about law enforcement access to social media networks ought to concentrate on 4th Amendment jurisprudence, which regrettably holds that you do not have a reasonable expectation of privacy in information you volunteer to third parties (such as Google and Facebook).
Antitrust
Despite the core gripes being about content moderation and speech issues, the paper recommends using federal antitrust policy as a bludgeon to change ‘Big Tech’s’ behavior. This is perplexing. First, because antitrust laws are ill-suited for such non-economic concerns. Second, because having complained about a supposed government-Big Tech nexus, it is counterintuitive to rely on an emboldened federal government to solve the perceived problems.
The only common thread underpinning the antitrust recommendations is that Frederick wants federal agencies to use their bully pulpits to put heat on ‘Big Tech’ companies. At no stage does Frederick define relevant markets so as to identify where monopolization by Big Tech companies has occurred or if anticompetitive conduct has even taken place. All her recommendations are narrowly targeted at ‘Big Tech’ platforms too, rather than conduct she dislikes across the whole economy. Her agenda is therefore about applying pressure on a narrow range of well-known businesses for political reasons. This focus necessarily means conservatives abandoning long-held principles about the desirability of neutrality in the application of economic regulation.
The paper demands using “aggressive antitrust investigations” to “ensure that Big Tech companies are not utilizing legitimate oligopoly power in one market to gain advantage in another market.” It also supports more scrutiny of mergers — “past, present, and future.” Conservatives used to oppose retrospective actions, and (absent time travel) it’s unclear how agencies would investigate that which has not happened yet. But this all amounts to asking committees and federal agencies to just dig around to see if they can find something that sticks against these companies. As with the left-wing Hipster Antitrust movement, Frederick suspects there is something nefarious going on and wants the feds to plough on until they find it. So much for innocent until proven guilty.
If that were all, it would be bad enough. But more substantive changes to antitrust laws are proposed. Frederick endorses a law change that would see “Big Tech platforms that leverage their market dominance to shape the public square” banned from self-preferencing their products or combining a “buy-side” and “sell-side” of the market.
“Shape the public square” seems amorphous terminology to justify developing a two-tier legal system for what is acceptable business conduct. Although she presents no evidence of these companies harming consumers, Frederick gives us an example of who would be affected. Under the law, she says, Google would not be able to bundle its operating system, app store, and search engine together. And these restrictions would apply, we are told, to Alphabet, Meta, Amazon, Apple and Microsoft too, suggesting all these companies shape the “public square” per the unannounced definition.
What does this mean in practice? At face value: the law would: ban Apple iPhones coming preloaded with apps that compete with those on App Store; ban Amazon selling Basics products competing with third parties on Amazon marketplace; and ban Google preferencing its Maps higher up on results from Google search. If those sorts of crude restraints sound familiar, it’s because a version of this sort of bill is being pushed by Senate Democrats, led by Sen. Amy Klobuchar (D‑MN), who is intent on “overhauling and modernizing” antitrust law.
Economists have recognized for decades that bundling, tying, self-preferencing and other vertical restraints like this are usually benign and enhance consumer welfare. Banning this conduct through such a law therefore amounts to abandoning the very consumer welfare standard that has underpinned the application of antitrust in recent decades. Frederick not only fails to present analysis as to how these businesses have harmed consumers on conventional grounds, but overlooks an abundance of evidence that suggests cracking down on vertical restraints would significantly harm them.
This is strange, because the next major recommendation actually demands that the consumer welfare standard be formally written into law. Historically, consumer welfare has been assessed through case law by the impact of firm conduct on prices and innovation in providing goods and services to customers in well-defined product markets. But Frederick wants it now codified in statute. Why? Well, because she wants the concept of consumer welfare broadened, again in an unexplained way, to incorporate the consumer benefits of open political speech and to account for “the exploitation of user data.”
While a notional commitment to consumer welfare is welcome, it is bizarre to see conservatives desiring the federal government force businesses to host political speech that any legislation would deem protected. What happened to conservative support for free association? It is more strange still that in this and other recommendations Frederick presumes targeted advertising developed via data accumulation has only downsides for consumers.
What this antitrust agenda amounts to, then, is an attempt to use and expand laws and federal agency powers to achieve outcomes they were not designed for, while re-defining established antitrust concepts to ensnare Big Tech companies for their current conduct. The agenda starts from the political ambition of browbeating these companies into changing their content policies, and then works back to consider how antitrust policy can be weaponized to achieve that goal.
If you were in any doubt about this, the antitrust recommendations section says explicitly:
“If the enforcement of existing antitrust law does not sufficiently address Big Tech’s abuse of market dominance, business practices, and harm to consumers, then reform or modernize antitrust laws…”
Translated: if we don’t see the outcomes we want from this all-out assault from the federal government, then we’ll pass new laws to finish the job. To paraphrase Groucho Marx, “Those are my principles, and if they don’t achieve our Big Tech objectives…well, I have others.”
Speech
Frederick moves on from antitrust to address speech, arguing in favor of changes to Section 230 of the Communications Decency Act, which states that interactive computer services such as Facebook, Google, and The Wall Street Journal’s comments section cannot be considered the publisher of the vast majority of third party content. You can sue a tweeter if their tweet defames you, but you cannot sue Twitter over the same tweet.
The law is widely misunderstood. Unfortunately, the paper includes a few of the most common mistakes made by those writing about it. Frederick argues that thanks to Section 230, Big Tech companies flourished “in the late 1990s and early 2000s.” Yet by her own definition, some “Big Tech” companies were not around during this time. TikTok was founded in 2016. Snapchat appeared a little earlier in 2011. Netflix ceased to be a DVD rental company relatively recently, moving into the steaming business in 2007. Amazon, Google, Microsoft, and other “Big Tech” companies certainly benefited from Section 230, but so did every other interactive computer service.
Frederick argues that Congress should “[r]emove Section 230 liability protection if a firm is feeding certain content to users or otherwise favoring or disfavoring speech based on viewpoint (such as labels or warnings on some speech).” Depending on how this is worded, this proposal would run afoul of the First Amendment.
Such a proposal would see social media sites and others removing explicit content labels, which is a strange recommendation coming from the same paper that argues social media companies are not doing enough to protect children. The proposal would also likely spell the end of social media newsfeed algorithms, which certainly favor some speech over others because of viewpoint.
Conservatives concerned about what they consider to be biased social media fact-checking and warning labels are free to sue. Section 230 has never covered speech by interactive computer services. Twitter, Facebook, and YouTube are already liable for their own speech. Section 230 is not a blanket liability shield. It only protects interactive computer services from liability associated with most third party content. A Twitter fact-checking label is Twitter’s speech. If someone believes that such a label is defamatory, they can sue Twitter.
The paper includes a footnote that claims (among other things) that “Yet by censoring content that is not objectionable or otherwise labeling and editorializing, Big Tech companies have moved far afield from Congress’s original intent and are acting as publishers of information.” This is not true. The authors of Section 230, Sen. Ron Wyden (D‑OR) and former Rep. Chris Cox (R‑CA), are alive and well and have not been shy about dispelling this myth. In 2020 they wrote Section 230’s intent was “to protect and encourage content moderation.” Section 230 was never intended to facilitate an online speech environment where platforms would be expected to host all legal speech or to act as mere neutral conduits of information.
Frederick mentions that her Section 230 recommendation “applies to all ‘interactive computer services.’ ” Given that Frederick wants Section 230 protection stripped from firms that remove legal content, this is a sweeping suggestion. Such a proposal would put every interactive computer service — from a personal blog with twelve readers and a comments section to Facebook and Google — in an unenviable position. They would have to choose between enjoying Section 230 protections while putting up with pornography, beheading videos, spam, and other legal content, or screening such content at the risk of being held liable for any illegal content that slips through the screening net. Such a system would significantly limit the amount of speech online. It would also plunge online platforms back into the “Moderator’s Dilemma” Section 230 was written to solve.
Perhaps the most interesting section of Frederick’s paper is her discussion of free speech as a value. She writes, “Governments are not the only actors capable of encroaching on Americans’ individual liberties. Private, monopolistic corporations should be held accountable if they violate these liberties to the degree that Big Tech has violated them just during the past year. These companies are not sacrosanct: Sovereign citizens of the United States do not exist solely to serve the economy or maximize gross domestic product.”
This represents a significant change in the traditional conservative conception of rights. “The freedom of speech” in the American context as envisioned by the Founding Fathers and codified in the First Amendment is a protection from being silenced by the government and also a protection from being compelled to speak or publish. The First Amendment prevents the government from banning a white supremacist from writing a racist letter to the editor of The New York Times. It also prevents the government from forcing The New York Times to publish the letter. As a restraint on government, the First Amendment is great. As a private firm’s content moderation decision, it would be a disaster given the plethora of “legal but awful” speech.
Frederick is correct that private companies can infringe on your liberties. A private company could kill you or steal your property. These actions violate your rights to life and property. But if a newspaper or social media platform denies you a platform you can continue to speak elsewhere. The right to “free speech” does not entail an entitlement to platforms owned by others. Atheists are free to hand out copies of The God Delusion. They are not free to interfere with a Catholic Mass to do so and it would be absurd for them to claim that the church was infringing on their liberty by ejecting them. Similarly, your freedom of speech is not affected by Facebook removing your content.
Conclusion
Although not persuasive as policy analysis, Frederick’s paper is of use to those interested in the history and anthropology of American conservatism. The paper embraces the use of political buzzwords such as “woke,” and “left-wing orthodoxy” while using words such as “monopoly” and “oligopoly” in a manner that suggest they have suddenly changed meaning. This kind of rhetoric will be familiar to those who have been paying attention to “Big Tech” policy debates over the last couple of years. Much of it is not a surprise. What may well be a surprise to many readers is that America’s leading conservative think tank has released a paper outlining a roadmap for interfering in First Amendment-protected content moderation, empowering federal bureaucrats in alphabet soup agencies, and hampering innovation and growth among America’s most famous and successful companies.
