Diane Katz for The Heritage Foundation
Senator Josh Hawley’s (R–MO) Ending Support for Internet Censorship Act (S. 1914) is intended to address claims of anti-conservative bias1
S. 1914, Ending Support for Internet Censorship Act, https://www.congress.gov/116/bills/s1914/BILLS-116s1914is.pdf (accessed October 25, 2019). by Facebook, Google, and other Internet platforms. If enacted, however, the bill’s prescription for regulating online content would risk eroding free speech rather than protecting it.
Both liberals and conservatives express concerns about online bias—particularly on social media.2
For examples, see Chris White, “Liberals Cry Censorship After YouTube Demonetizing Frenzy Hits SPLC Content,” The Daily Caller, June 7, 2019, https://dailycaller.com/2019/06/07/youtube-steven-crowder-google/ (accessed July 16, 2019); Andre Damon, “Facebook Escalates Censorship of Left-Wing, Anti-War Organizations,” World Socialist Web Site, August 23, 2018, https://www.wsws.org/en/articles/2018/08/23/pers-a23.html (accessed July 16, 2019); Sanjana Varghese, “Twitter Has Purged Left-Wing Accounts with No Explanation,” Wired, October 18, 2019, https://www.wired.co.uk/article/twitter-political-account-ban-us-mid-term-elections (accessed July 16, 2019); and Fruzsina Eordogh, “Why Republicans Weren’t the Only Ones Shadow Banned on Twitter,” Forbes, July 21, 2018, https://www.forbes.com/sites/fruzsinaeordogh/2018/07/31/why-republicans-werent-the-only-ones-shadow-banned-on-twitter/#49e12912434b (accessed July 16, 2019). All of which has roused the regulatory inclinations of Congress—in contrast to the liberality toward the Internet that has prevailed since the 1990s. S. 1914 proposes a particularly aggressive regulatory regime.
Social media companies and others in the technology sector deny that they intentionally discriminate against partisan content, and systemic bias against any group has not been proven.3
See, for example, Marguerite Reardon, “House Republicans Accuse Google of Liberal Bias,” Cnet, December 11, 2018, https://www.cnet.com/news/house-republicans-accuse-google-of-liberal-bias/ (accessed July 16, 2019), and Kathleen Chaykowski, “Congressional Leaders Press Zuckerberg on Political Bias, Data Collection at Facebook,” Forbes, April 11, 2018, https://www.forbes.com/sites/kathleenchaykowski/2018/04/11/congressional-leaders-press-zuckerberg-on-political-bias-data-collection-at-facebook/#323eced11a95 (accessed July 16, 2019). Nevertheless, there are examples where actions taken against conservative content, conservative personalities, and conservative organizations appear to be arbitrary and are often reversed when protested. This perception has led many to regard the denials of those within the industry to be less than compelling. The denials are further muddled when company executives publicly express explicitly partisan opinions.4
For example, Google co-founder Sergey Brin told colleagues that he felt “offended” by the 2016 election of Donald Trump. “Let’s face it,” he reportedly said, “most people here are pretty upset and pretty sad because of the election.” During the same meeting, Alphabet CFO Ruth Porat characterized Hillary Clinton’s loss as a “massive kick in the gut…. And it was really painful.” See Richard Nieva, “Google’s Sergey Brin Calls 2016 Election ‘Offensive’ in Leaked Video,” C/Net, September 12, 2018, https://www.cnet.com/news/google-co-founder-calls-2016-election-offensive-in-leaked-video/ (accessed October 25, 2019). Jay Carney, “Amazon’s senior vice esident of global corporate affairs, questioned President Trump’s patriotism during an interview at the recent GeekWire Summit. Referring to White House officials with whom he had worked in the past, Carney said, “I never doubted that they were patriots. I never doubted that they believed they were doing things that were in the best interest of the country. I don’t feel that way now, and I worry.” Monica Nickelsburg and Todd Bishop, “Q&A: Jay Carney on the Trump White House, Amazon Antitrust Scrutiny, and Working for Bezos,” GeekWire, October 10, 2019, https://www.geekwire.com/2019/qa-jay-carney-trump-white-house-amazon-antitrust-scrutiny-working-bezos/ (accessed October 25, 2019).
Proving bias is difficult because Internet activity is dynamic, the algorithms that manage content are proprietary and unavailable for public scrutiny, and the volume of data is massive. On Facebook alone, there are more than 2 billion active monthly users (1.59 billion of whom visit the site daily). Every minute, users upload 136,000 photos, 293,000 status updates, and 510,000 comments.5
Dan Noyes, “The Top 20 Valuable Facebook Statistics –Updated September 2019,” Zephoria Digital Marketing, September 2019, https://zephoria.com/top-15-valuable-facebook-statistics/ (accessed October 8, 2019). At the same time, the platform algorithms continuously adjust their specifications based on users’ “clicks,” “likes,” and “shares.”6
According to Francesca Tripodi, professor of sociology at James Madison University, there is “a deep misunderstanding” among the public about how search engines work and how content moderation is conducted. As she told The Guardian, “These algorithms are very complex and not at all intuitive. They weigh things like how many people are linking to an article, what key words appear in the headline, and what specific phrases people are using in their search.” See Oscar Schwartz, “Are Google and Facebook Really Suppressing Conservative Politics?” The Guardian, December 4, 2018, https://www.theguardian.com/technology/2018/dec/04/ (accessed October 25, 2019).
Even if systemic bias were conclusive, the fact is that these private companies have the right to moderate content according to standards of their own choosing. That does not mean that conservatives have no recourse. Indeed, conservatives have countered left-wing media privilege in the past without regulation. Deregulation, in fact, benefits the underdogs.
In the present circumstances, the better response to bias concerns is not government speech controls, but empowering consumers, investors, and competitors to create alternatives, punish misdeeds, and reward excellence.
In contrast, S. 1914, if enacted, would yield more harm than benefit, and would violate fundamental conservative principles.
Circumventing the First Amendment
The First Amendment largely precludes the government from regulating speech on social media and other interactive Web services.7
There are limited exceptions, including enforcement of federal obscenity prohibitions and the sexual exploitation of children. S. 1914 proposes to bypass that constitutional hurdle through a regulatory backdoor of sorts. The bill would effectively force platforms to either comply with a government speech code or lose statutory liability protection for their sites.
The statutory liability protection at issue is Section 230 of the Communications Decency Act,8
U.S. Code § 230, Protection for Private Blocking and Screening of Offensive Material, https://www.law.cornell.edu/uscode/text/47/230# (accessed July 16, 2019). which has been called “the most important law protecting Internet speech.”9
Electronic Frontier Foundation, “CDA 230: Section 230 of the Communications Decency Act,” https://www.eff.org/issues/cda230 (accessed July 16, 2019). The 1996 statute shields social media, blogs, streaming services, and a variety of smartphone applications from legal responsibility for much of the content posted by the public. It does not shield those who write the material (“content creators”) from claims of liable or defamation, or criminal prosecution.
Absent such legal protection, platforms could face a barrage of lawsuits over the posting, or exclusion, of controversial content. Such a threat would likely provoke some platforms to either block all but the blandest content or refrain from curation altogether—and subject the public to the extremes of human depravity. Indeed, it was that Hobbesian choice that prompted then-Representative Chris Cox (R–GA) and Senator Ron Wyden (D–OR) to introduce the Section 230 legislation in 1995.
The Genesis of Section 230
Section 230 was enacted following two lawsuits in the 1990s that distinguished online distributors from publishers in relation to liability for online content. In the first case, Cubby, Inc. v. CompuServe, Inc.,10
Justia US Law, “Cubby, Inc. v. CompuServe Inc., 776 F. Supp. 135 (S.D.N.Y. 1991), October 29, 1991, https://law.justia.com/cases/federal/district-courts/FSupp/776/135/2340509/ (accessed July 16, 2019). the co-founders of Skuttlebut (a news and gossip aggregator for the broadcast industry) sued CompuServe11
CompuServe was the first company to offer electronic mail and online chat capabilities. for libel over statements on its journalism forum (one of the company’s 150 special interest sites). The comments at issue were posted by a forum columnist.
In a 1991 ruling, the U.S. District Court for the Southern District of New York cleared CompuServe of wrongdoing because the plaintiffs failed to show that the company directly curated the forum contents. Instead, a contractor uploaded content to the CompuServe site and, therefore, the company did not know, or have reason to know, about the columnist’s comments. According to the court,
CompuServe has no more editorial control over such a publication than does a public library, book store, or newsstand, and it would be no more feasible for CompuServe to examine every publication it carries for potentially defamatory statements than it would be for any other distributor to do so. First Amendment guarantees have long been recognized as protecting distributors of publications…. Obviously, the national distributor of hundreds of periodicals has no duty to monitor each issue of every periodical it distributes. Such a rule would be an impermissible burden on the First Amendment.
In the second case, Stratton Oakmont v. Prodigy,12
Stratton Oakmont, Inc. v. Prodigy Services Company, 1995, https://h2o.law.harvard.edu/cases/4540 (accessed July 16, 2019). an investment firm sued computer network Prodigy over a user’s posts that it claimed were defamatory. The New York Supreme Court ruled that Prodigy was liable because the company moderated its message boards and thus exerted editorial control as a publisher.13
Conversely, Prodigy would not likely have been liable if the court determined that the company was merely a “distributor” (unless Prodigy knew or had reason to know about the allegedly defamatory statements). The ruling stated:
By actively utilizing technology and manpower to delete notes from its computer bulletin boards on the basis of offensiveness and “bad taste,” for example, PRODIGY is clearly making decisions as to content, and such decisions constitute editorial control. That such control is not complete and is enforced both as early as the notes arrive and as late as a complaint is made, does not minimize or eviscerate the simple fact that PRODIGY has uniquely arrogated to itself the role of determining what is proper for its members to post and read on its bulletin boards. Based on the foregoing, this Court is compelled to conclude that for the purposes of Plaintiffs’ claims in this action, PRODIGY is a publisher rather than a distributor.
By punishing Prodigy for screening posts, critics warned that the ruling would either prompt websites to abandon content moderation altogether (lest they be deemed a publisher) or to overcompensate by scrubbing all potentially contentious content from sites—either of which would inhibit the Internet’s function as a democratizing medium of communication.
Thus, Cox and Wyden crafted the liability exemption in hopes that platform administrators would actively curate content with less risk of legal challenge. As Section 230 states,
(c) Protection for “Good Samaritan” blocking and screening of offensive material
(1) Treatment of publisher or speaker
No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.
(2) Civil liability
No provider or user of an interactive computer service shall be held liable on account of—
(A) any action voluntarily taken in good faith to restrict access to or availability of material that the provider or user considers to be obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable, whether or not such material is constitutionally protected; or
(B) any action taken to enable or make available to information content providers or others the technical means to restrict access to material described in paragraph (1).
The statute also directs service providers to notify customers that parental control protections are commercially available for limiting access to material that is harmful to minors.
Section 230 immunity is not absolute; it does not extend to federal criminal liability or intellectual property claims. Immunity is also forfeited when a website operator becomes an “information content provider,” that is, “responsible, in whole or in part, for the creation or development of information provided through the Internet or any other interactive computer service.”
Neither is Section 230 immunity exhaustive. Nothing in the law bars enforcement of federal obscenity prohibitions, the sexual exploitation of children, or any other federal criminal statute. However, states are precluded from imposing liability under inconsistent state or local statutes.
Some proponents of S. 1914 claim that Congress granted the safe harbor in exchange for platforms “providing a forum free of political censorship.”14
Office of Senator Josh Hawley, “Senator Hawley Introduces Legislation to Amend Section 230 Immunity for Big Tech Companies,” June 19, 2019, https://www.hawley.senate.gov/senator-hawley-introduces-legislation-amend-section-230-immunity-big-tech-companies (accessed July 16, 2019). Therefore, they contend, platforms should have to prove that they practice political neutrality in order to benefit from the liability shield.
Well-intended though such pronouncements may be, they are in error. Congress did not establish a neutrality quid pro quo for the safe harbor conferred by Section 230. Quite the contrary. The congressional architects sought to encourage content curation by eliminating the liability that might otherwise arise from moderating online speech.
Indeed, it was specifically crafted to keep government out of content moderation. As then-Representative Cox said during a floor debate on the legislation,
Our amendment will do two basic things: First, it will protect computer Good Samaritans, online service providers, anyone who provides a front end to the Internet, let us say, who takes steps to screen indecency and offensive material for their customers. It will protect them from taking on liability such as occurred in the Prodigy case in New York that they should not face for helping us and for helping us solve this problem. Second, it will establish as the policy of the United States that we do not wish to have content regulation by the Federal Government of what is on the Internet, that we do not wish to have a Federal Computer Commission with an army of bureaucrats regulating the Internet because frankly the Internet has grown up to be what it is without that kind of help from the Government.15
Christopher Cox, “Personal Explanation,” Congressional Record, Vol. 141, No. 129 (August 4, 1995), https://www.congress.gov/congressional-record/1995/8/4/house-section/article/h8460-1 (accessed July 16, 2019). (Emphasis added.)
A Closer Look at the Proposed Bill
S. 1914 proposes de facto content regulation by the federal government. If enacted, large Internet platforms would require an “immunity certification” from the Federal Trade Commission (FTC).16
If enacted, the law would apply to a provider of “interactive computer service” that, in the most recent 12-month period, had more than 30 million active monthly users in the United States; more than 300 million active monthly users worldwide; or more than $500 million in global annual revenue. To obtain certification, platforms would have to prove to the FTC “by clear and convincing evidence” that they do not “moderate” information provided by other content providers “in a manner that is biased against a political party, political candidate, or political viewpoint”—and that they have not done so in the preceding two years.
Supporters of the bill claim that the legislation offers platforms a choice between obtaining a neutrality certification in exchange for liability immunity and forgoing the certification and the immunity. But the lack of a legal safe harbor could substantially hinder the functionality of social media and other platforms.
The bill defines the term “moderate” as influencing or altering content “in a manner that is designed to negatively affect a political party, political candidate, or political viewpoint” or “disproportionately restricts or promotes access to, or the availability of, information from a political party, political candidate, or political viewpoint.” Restrictions on content moderation apply to the use of algorithms (or other automated processes), as well as officers and employees “motivated by an intent to negatively affect a political party, political candidate, or political viewpoint.”
Imposing the burden of proof on the platform requires companies to prove an absence of bias. From a practical standpoint alone, that is an overwhelming task, given the billions of daily postings, the countless viewpoints expressed, and the virtually infinite political sensitivities across the ideological spectrum. Requiring tech companies to convey their proprietary intellectual property to government (to prove they are not acting in a discriminatory fashion) is an insupportable violation of property rights.
The threshold for immunity certification is also problematic because an inability to prove neutrality is not proof of bias. Moreover, what constitutes a “disproportionate” restriction of political discourse—and compared to what? How are the five FTC commissioners supposed to determine whether algorithmic outcomes are intentional when they result from a multitude of factors in play at any given moment?
The degree of arbitrary power that the bill would confer on the FTC is unprecedented. The five political appointees (who do not answer to the White House or to Congress) could effectively control civic discourse online—and just two could block immunity certification. There is nothing to prevent the parameters of “neutrality” from shifting as often as commission membership.
Even if immunity certification were granted, S. 1914 would subject platforms to endless oversight by inviting complaints of discrimination and granting a complainant a right to participate in any hearings on applications for certification. Given that certification must be renewed every two years, the application process would essentially be continuous.
The bill also authorizes the FTC to hire “sufficient staff” to process what would effectively become a federal department of online speech regulation. However, S. 1914 does not provide any new money.17
In testimony before the House Committee on Appropriations Subcommittee on Financial Services and General Government on September 25, 2019, FTC Chairman Joseph Simons mentioned that the agency’s Technology Task Force plans to “examine industry practices, conduct law enforcement investigations and coordinate and consult with staff throughout the FTC on technology related matters, including reviews of consummated technology mergers.” Simons also confirmed an antitrust investigation of Facebook. See Joseph Simons, “Federal Trade Commission: Protecting Consumers and Fostering Competition in the 21st Century,” testimony before the Subcommittee on Financial Services and General Government of the House Committee on Appropriations, September 25, 2019, https://www.ftc.gov/system/files/documents/public_statements/1545285/appropriations_committee_testimony_092519.pdf (accessed October 2, 2019). The cost of monitoring—whether via human reviewers or algorithms—would be too burdensome for all but the biggest platform operators, conferring additional advantage on the incumbent dominant companies and further stifling competition.
Applicants for certification would be allowed to seek an exception from the neutrality requirements on the basis of “business necessity,” although the criteria would be difficult to fulfill. The bill defines “business necessity” as an action that
is necessary for business or [if] the information involved is not speech that would be protected under the First Amendment of the United States Constitution, [if] there is no available alternative that has a less disproportionate effect, and [if] the provider does not act with the intent to discriminate based on political affiliation, political party, or political viewpoint.
Business necessity does not include “any action designed to appeal to, or gain favor from, persons or groups because of their political beliefs, political party membership, or support for political candidates.”
The other exception allows an applicant to obtain certification even if an employee engages in a non-neutral action, if the company,
[i]mmediately upon learning of the actions of the employee publicly discloses in a conspicuous manner that an employee of the provider acted in a politically biased manner with respect to moderating information content; and terminates or otherwise disciplines the employee.
A government directive to practice public shaming as a form of social control is most common in communist regimes.
The Failures of Government “Fairness”
Government attempts to legislate “fairness” have a record of failure. For example, the imposition of the Fairness Doctrine in 1949 by the Federal Communications Commission (FCC) yielded perverse unintended consequences.
The FCC required broadcasters to “afford reasonable opportunity for the discussion of conflicting views of public importance.” But broadcasters quickly realized that “fairness” was an imprecise concept that could not be proved, and thus opted (in large measure) to avoid controversial topics altogether. As a result, the public was left less informed.
In similar fashion, the FCC in 1971 declared that radio broadcast licenses could be withdrawn if stations aired music associated with drug use. Failure to monitor lyrics would constitute “a violation of the basic principle of the licensee’s responsibility for…the broadcast material presented over his station. It raises serious questions as to whether continued operation of the station is in the public interest.”18
Federal Communications Commission, “Licensee Responsibility to Review Records Before Their Broadcast,” March 5, 1971, https://www.fcc.gov/document/licensee-responsibility-review-records-their-broadcast (accessed July 16, 2019).
As a result, one station owner eliminated all Bob Dylan songs because the management could not interpret the lyrics. Others chose to avoid all mention of drug use—even anti-drug messages—lest the FCC revoke their license.19
Tom Wheeler, “Drug Lyrics, the FCC and the First Amendment,” Loyola of Los Angeles Law Review, Vol. 5 (April 1972), pp. 329–367, https://digitalcommons.lmu.edu/cgi/viewcontent.cgi?referer=https://www.google.com/&httpsredir=1&article=1097&context=llr (accessed July 16, 2019).
Those are lessons best remembered by proponents of government-imposed neutrality edicts.
The progressive ideology pervading American media is not a new phenomenon. More uncommon are the current demands of some conservatives for a government response. When marginalized in the past, conservatives pursued deregulation to ease entry into radio and television markets and secure a competitive foothold. Under current circumstances, reliance on principles of free enterprise will prove more effective than government interference in elevating conservatives’ online presence.
It is important to recognize, however, that even left-leaning platforms, such as Google and Facebook, have dramatically increased the reach of conservative voices, to great benefit. The platforms offer ready-made and immediate access to (potentially) billions of people worldwide. Indeed, many of the celebrity critics featured in congressional hearings on platform bias owe their popularity to the targets of their fury.
Regulation tends to reinforce the status quo rather than promote reform. Dominant players in the market often heavily influence rulemaking, and they can easily afford the costs of compliance compared to nascent firms. The inflexibility of regulation inhibits investors from customizing competitive strategies for their circumstances and instead forces them to act on standards based on political expediency.
The successes of conservatives in challenging liberal domination of media has largely resulted from deregulation. For example, conservatives’ command of talk radio originated after the FCC ended enforcement of its so-called Fairness Doctrine in 1987.
Just a year later, Rush Limbaugh’s Sacramento-based radio show was syndicated, and by 1991 became the top show in syndication.
Concurrently, the AM dial was suffering audience loss in competition with FM stations that provided high fidelity sound quality. Conservative investors were able to increase their station holdings after Congress in 1992 relaxed ownership rules to allow a single company to acquire two stations per market. Their broadcast holdings again increased after the Telecommunications Act of 1996 further eased the ownership restrictions. The affiliation of stations across markets fostered syndication of conservative programming—and thus audience growth.
In the television market, the FCC for years effectively prohibited development of cable TV competition, to the advantage of the three national networks (all of which reflected a decidedly liberal bent). But restrictions eased over time as the technology evolved and municipalities clamored for lucrative cable franchise deals.
Home Box Office (HBO) became the nation’s first pay-TV network in 1972, triggering rapid development of satellite transmissions. Investment in both cable hardware and programming surged after passage of the deregulatory Cable Act of 1984. Indeed, from 1984 through 1992, the industry spent more than $15 billion on wiring America—the largest private construction project since World War II.20
California Cable and Telecommunications Association, “History of Cable,” https://www.calcable.org/learn/history-of-cable/ (accessed October 15, 2019).
The Telecommunications Act of 1996 also spurred greater choice for consumers. Continued deregulation led investors to deploy broadband networks through which cable companies could offer higher-speed Internet and competitive telephone services.
The Fox News Channel launched in 1996 as a right-leaning alternative to the decidedly leftist CNN. By 2002, Fox News topped the cable news ratings, which it has dominated since then.
The advent of MP3 players, Internet radio, smartphone applications, and music subscription services has compelled radio stations to convert from music to the talk formats in which conservatives dominate. For example, the unapologetically conservative Sinclair Broadcast Company, established in 1986, grew from just three stations to 191 in 89 markets, becoming one of the largest broadcasters in the nation.21
Sinclair Broadcast Group, “About,” http://sbgi.net/#About (accessed October 15, 2019).
Technological advances will continue to provide conservatives with new options to reach the public. As President Trump tweeted on June 12: “Thank goodness we can fight back on Social Media.”22
@realDonaldTrump, Twitter, June 12, 2019, https://twitter.com/realDonaldTrump/status/1138789726771732480 (accessed October 15, 2019).
Recommendations for Policymakers
- First, do no harm. The proper role of government in technology policy should be limited to securing the national defense, protecting and defending voluntary transactions (such as contracts, investments, and acquisitions), and settling disputes in matters of law.
- Protect Internet freedom. Congress and the Administration must embrace policies that protect Internet freedom and reject retributive regulation. The nation’s technological progress largely depends on upholding the fundamental principles of free enterprise, limited government, and property rights.
- Reject reliance on government solutions. Policymakers must recognize that government bureaucracy is neither impartial nor altruistic. Free enterprise allows consumers and investors to act in the present based on their own information and preferences. Penalties for errors in judgment are swift and strict. In many instances, consumers and investors can exact more effective punishment than can regulators.
Recommendations for the Public
- Take private action. The public should support deregulation of telecommunications, and advocate for change through private action—such as by petitioning elected representatives, communicating with company executives, opting for competing services, and joining advocacy groups. Decades of experience have demonstrated that conservative media have benefited far more from deregulation than regulation. The likelihood of regulatory blunders by government is very high and the consequences grave, including stifling innovation, suppressing free speech, and solidifying the dominance of incumbents for years to come.
- Establish alternatives. The Internet offers virtually unlimited opportunities to create new channels for conservative content. Members of the public can put principles of free enterprise into action—by marshaling public support and private resources for alternatives. The point is not to overcome the popularity of Google or Facebook, but to secure independent means to communicate directly with target audiences.
- Pursue remedial action for excluded content, as appropriate. It is the prerogative of private platforms to moderate content based on their own community standards (or other criteria). But most sites also provide appeal procedures for instances in which excluded content appears to be arbitrary.
- Hold platforms accountable for their content management. The most constructive means of increasing accountability for platforms’ conduct is to make use of and support responsible methods of monitoring and fact-checking information on websites.
S. 1914 is crafted to punish platforms accused of online bias. Aside from the threat to the First Amendment, the government licensing of online speech would increase censorship, not prevent it. Whether the liability protection of Section 230 remains a sound policy is a legitimate topic of discussion; the online landscape today bears little resemblance to the pre-Google and pre-Facebook era when Congress enacted the provision. But that is not the issue addressed by the proposed Ending Support for Internet Censorship Act.
Diane Katz is Senior Research Fellow in Regulatory Policy in the Thomas A. Roe Institute for Economic Policy Studies, of the Institute for Economic Freedom, at The Heritage Foundation.