Communications of the ACM
Economic and business dimensions
A Response to Fake News as a Response to Citizens United
Credit: Adragan / Shutterstock
How do we fight fake news? Promoting free speech can, at the same time, promulgate false speech and the more vigorously we protect free expression, the more we inadvertently permit deception. The problem is hard. Legislatures across Asia, North America, Europe, and Latin America grapple with it. It touches campaign finance reform, already a thorny issue. As a computer scientist, I argued frequently with my law professor father over ways to solve it. Many of the technological methods we might use to curb false speech run afoul of current law and the very idea of free speech. Yet, the space between law and technology is where we might find better answers.
The problem runs deeper than the technical challenges of machine learning. At one level, reducing Type I errors simply invites those of Type II while training one filter to recognize fake news simply invites adversaries to train other filters to write it.3,4 No, at a different level, courts question the desire to regulate fake news at all: they bar intervention in cases of politically protected speech. If fake news is political, it should not be regulated. From a legal theory perspective, there are elements of this policy that are wise—courts should avoid judging political truth—at the same time there are elements that are unwise—courts should dismantle systems that prevent us from hearing truth. At present, court decisions stifle competing truths and it is here that an old idea from computer science, the Church-Turing thesis, suggests the ban on some forms of intervention should be lifted. Computability theory has much to add to legal practice in the design of better systems.
In 2010, U.S. law became unambiguous. The landmark case Citizens United v. Federal Election Commission resolved that the First Amendment forbids suppression of voices "the government deems suspect."a This rises to a categorical imperative for political speech. A Supreme Court majority believed that even if corporations, labor unions, and others gain harmful influence, the damage they cause by unrestricted spending is outweighed by the damage to "democratic process resulting from the restrictions upon free and full discussion."b
If fake news is political, it should not be regulated.
This followed an earlier decision in Buckley v. Valeo that struck down a "restriction on the amount of money a person or group can spend on political communication" because it "reduces the quantity of expression ... the depth of [issues explored], and the size of the audience reached."c Buckley v. Valeo committed the Court to a position that more speech is always better, a position reaffirmed in Citizens United as "there is no such thing as too much speech."d As Justice Anthony Kennedy, writing for the majority observed, any "statute which chills speech can and must be invalidated."e Together, these two decisions opened the floodgates for unlimited spending by Political Action Committees.
Kennedy's view is one of pure principle and easy deliberation. Citizens, individually and collectively, have an absolute right to spend on speech. If spending is speech, especially political speech, government regulation is limited. Constitutional law scholars describe this view as deontological: it examines only the rightness of the rule and not the nature of the consequence.f Further, the Supreme Court has long held that government has no business at all regulating the truth or falsity of speech: "under the First Amendment, there is no such thing as a false idea. However pernicious an opinion may seem, we depend for its correction not on the conscience of judges and juries but on the competition of other ideas."g
So it was that my father shot down intervention after intervention that I and others conceived. After a year of deliberation, I hit upon two ideas, one that survives strict scrutiny and lives within existing law, and another that provides a logic to overturn that law. It is the second, grounded in philosophy and computer science, that I articulate here in hopes of continuing a conversation cut prematurely short. Sadly, my father died January 29, 2019, a date late enough to have argued the first idea but too early to argue the second.6 For explication of the first idea, I refer readers with more time to a forth-coming paper, "The Problem of Fake News."h The gist there is to put friction on harmful externalities not speech, and to invoke intellectual property law so as to motivate news platforms to alter their business models. The second idea operates directly on constitutional law. The gist here is the Court's logic is self-contradictory and thus invalid. The first deals with the "fakeness" of news, the second with its volume. This column addresses solely the distortions of ideas caused by excesses of money as speech.
Let us set aside all the normal exceptions to free speech—incitement to violence, defamation, fraud, trade secret misappropriation, national defense, and so forth. These exceptions create "balancing tests" that recognize competing interests also matter. Preserving life and avoiding corruption do create limits. Set aside all these competing interests and focus only on the truest most politically protected speech. Let us accept the Court's assumption of the need to protect such speech in order to invalidate its conclusion that it actually does protect speech. This essay takes aim at the core logic supporting the Supreme Court's decision to remove contribution limits on independent expenditures. In effect, the Court holds that infinite amounts of politically protected speech, and its corollary infinite amounts of money, are valid means of protecting "free and full discussion." Their presumption is the cure for misinformation is more information, yet this turns out to be provably false. As shown in this column, unlimited spending creates a paradox as truths can cancel truths. From there, free and full discussion collapse when falsehoods cancel truths. In seeking to protect speech, the Court has devised means to obstruct speech.
This column rejects the pure principle view that a right to free speech spending, even in protected cases, is absolute and therefore not subject to intervention. My claim is there exist no absolute rights in any complete system of citizens' rights. To invalidate an absolute right, simply create a contradiction by pitting that right against itself in the manner that mathematicians Church and Turing proved there exist systems with no absolute truths. Consider the statement: The second half of this sentence is absolutely true and the first half of this sentence is absolutely false. The inherent contradiction means the rule is void in at least one application. Not only must a right balance competing rights, but also that right must balance itself. It is this sense in which one aspect of the Citizens United decision is deeply and irretrievably flawed. It presupposes an absolute right. Such a right cannot exist, for any such right pitted against itself is a truth that is untrue, an absolute that is not absolute—a logical contradiction. Three illustrations prove instructive.
Consider a superordinate right, that to life relative to that of speech. We may presuppose that, given a binary and exclusive choice, if a majority of people would prefer a right to live as they choose over a right to speak as they choose, then speech is the subordinate right. Every society, including ours, recognizes at least one limit on an absolute right to life: One person taking a life in self-defense against his or her attempted murder has a reasonable expectation of absolution. The right to life is not so sacrosanct that it cannot be taken and the life that sought to extinguish life is the one more justly forfeit if only one survives. Most societies, excepting ours, recognize limits on political speech. The analogy is that we protect speech absolutely, even though it be used to deny speech to others, yet we do not protect life absolutely when it is used to deny life to others. The logic pertinent to the superordinate right ought to be no less profound in application to the subordinate right.
Consider Justice Oliver Wendell Holmes' famed example of impermissible speech: A person has no right to falsely shout fire in a theater to cause a panic. Such a miscreant cannot claim a right to speech that allows him to endanger the lives of others. We now imagine this miscreant shouting "Fire!" over the voices of any people whose ideas he does not want others to hear. His outcry is speech that cancels speech. He could even use a true statement such as "The earth is round!" or a politically protected statement such as "Liberty for all!" to achieve this effect. Indeed, knowing a mode of speech is protected absolutely, he could choose it in order to shout safely even more loudly. A group of such miscreants could shout more loudly still, amplified over the Internet, overriding voices they do not want others to hear. And a group of corporate miscreants could shout yet even more loudly amplified by infinite amounts of money. One message, excluding other messages, is precisely the opposite of "free and full discussion"i that the Supreme Court sought to protect. In such a case, truthful politically protected speech has stopped others' ears from hearing other truths.
Blind adherence to the law, without understanding the due process of law, is its own form of tyranny.
As a further example, consider an absolute right to anonymity and the inherent contradiction in a situation where a claim to that right is used by a party seeking to violate the anonymity of everyone around him. Consider an act of doxing and suppose a criminal unmasks others' private identity yet tries to avoid prosecution by using his right to anonymity as a means to hide from his crime. A reasonable conclusion is that the rights of the violator ought to be suspended precisely in order to protect the rights of everyone else. Hacking a person's private information and publishing it with intent to harm is already illegal. Failure to suspend the right to anonymity, in order to identify and prosecute the criminal, would be an illogical contradiction that would produce higher levels of doxing. Enforcing the right would negate the reason for granting the right. Lower, not higher, levels of privacy would result.
The string of decisions culminating in Citizens United has this same exact effect. Lower, not higher, levels of discourse flow through a market of ideas when special interests literally cannot spend too much.
A right to life that is used to deny a life or a right to speech that is used to deny speech or a right to anonymity that is used to deny anonymity are contradictions in terms. They cannot be enforced without invalidating the logic that led them to be enforced. Such is the effect of the Citizens United decision that holds no amount of spending is too much. The argument's logic is flawed.
If a single noisy truth can mask important quiet truths then what hope have we of hearing quiet truths when masked by myriad noisy lies? The market of ideas becomes not just chilly but frozen. To fight false ideas using "the competition of other ideas" we must hear those other ideas.
The theory has practical, not merely academic, value. Russian propaganda models operate under a strategy of "vilify and amplify."1 This strategy undermines the credibility of any message or messenger that opposes the propagandist while repeating ad infinitum the messages it wants others to hear and, by repetition through different channels, to believe. Repetition of messages is known to increase belief in false claims.7 Character assassination is proven to decrease belief in true claims.5 Willful failure to act against use of speech by one to override speech of another is not free speech protection but free speech suppression. Without irony, this is the manner of modern free speech suppression in Russia and other totalitarian regimes.i
In 2010, the top 100 individual donors gave $73 million to federal candidates, political parties, and super PACS. In 2016, following Citizens United, that number rose to more than $900 million.j Super PACs, as independent expenditure-only committees, may raise unlimited sums of money from corporations, trade unions, and individuals, then spend unlimited sums to promote their political causes.k They have no spending limits. With such resources and such a policy, what stops one partisan from buying all the ad space in swing districts a few weeks prior to an election? The role of government is not to decide political truths but rather to make room for enough sources of truth to enable a just market to decide. Under Citizens United, the alternative is a market with but few ideas. The appearance of choice is false, having been predetermined by those who speak loudly enough to set the list of choices.
Consider the absurdity of a law that forbids dissemination of the phrase "Tiananmen Square Uprising." A just society requires the law be broadly posted so that citizens, aware of the law, take care not to break it. Yet dissemination of the law is a violation of the law, a contradiction. Self-contradictory laws are unjust. And lest those in the West point smug fingers at those in the East, multiple instances of banning Nazi slogans in Germany or hateful ideas on American college campuses exhibit the same contradictory character under the guise of political correctness. We all have this problem. A just and better society requires just and better laws, those that censor harms rather than content, those that balance consequences of over and under reach, and those that make room for multiple ideas, even ones we do not like.
Rules whose enforcement in the extreme yield their automatic repeal cannot logically support any goal offered to justify their application. A contradiction ensues. A logic is corrupt whose extreme application leads to its own negation. The deontological view that holds pure principle to be the standard regardless of consequence cannot be correct. Having recognized the problem, the only remaining question is where to draw the boundary on consequence, not how to deny that the boundary exists.
Not only does the line exist, speech crosses that line when it suppresses others' speech. Rights are violated not only in the recognized cases of unlawful violence,l defamation,"m and fraud,n but also in the unrecognized cases of "vilify and amplify." In the limit, one voice dominates another, outspends another, and monopolizes an idea market. Then, as with antitrust, strict non-intervention is an abdication of a governing duty to ensure a fair fight in the market of ideas.
One thing my father taught me is that blind adherence to the law, without understanding the due process of law, is its own form of tyranny. We have mechanisms to change bad law. From 1928, when Hilbert posed his problem of identifying consistent systems of absolute truths, it took eight years, until 1936, for mathematicians and philosophers to solve it and recognize its implications. We may thank Church and Turing for correcting our misconceptions of truths as absolute. It has been nine years since Citizens United. Now aware of the Church-Turing thesis, legislators and Supreme Court justices ought not take longer to correct their misconceptions of rights as absolute.
References
1. Academic Outreach (AO) program of the Canadian Security Intelligence Service (CSIS). Who Said What: Security Challenges of Modern Disinformation. February 2018.
2. Fish, S. What is the First Amendment for? New York Times (Feb. 10, 2010); https://nyti.ms/2WREPGy
3. Graham-Cumming, J. How to beat a Bayesian spam filter. 2004; https://bit.ly/2x7acgZ.
4. Loder, T., Van Alstyne, M., and Wash, R. An economic response to unsolicited communication. Advances in Economic Analysis & Policy 6, 1 (Jan. 2006).
5. Oreskes, N. and Conway, E.M. Merchants of doubt: How a handful of scientists obscured the truth on issues from tobacco smoke to global warming. Bloomsbury Publishing USA, 2001.
6. Roberts, S. William Van Alstyne Dies, 84; Often-cited Constitutional scholar. New York Times (Feb. 4, 2019); https://nyti.ms/2KXFAXt
7. Tucker, J. et al. Social media, political polarization, and political disinformation: A review of the scientific literature, 2018; https://bit.ly/2ZCehWZ
Footnotes
a. Citizens United v. Federal Election Commission, 558 U.S. 310 (2010).
b. Ibid. Kennedy writing for the majority.
c. Buckley v. Valeo, 424 U.S. 1 (1976).
d. Justice Stevens, dissenting. Syllabus: Citizens United v. Federal Election Commission <https://bit.ly/2suTKTQ>, Supreme Court of the United States. p. 83.
e. Op. cit. Citizens United v. Federal Election Commission.
f. Gertz v. Robert Welch, Inc. (1974) 418 U.S. 323, 340.
g. See ssrn.com for an early draft.
h. Op. Cit. Citizens United v. FEC.
i. Op. Cit. Canadian Security Intelligence Service.
j. Brennan Center for Justice. https://bit.ly/31Hr3p1
l. Virginia v. Black, 538 U.S. 343, 359 (2003).
m. R.A.V. v. City of St. Paul, 505 U.S. 377, 3399–406 (1992).
n. Ohralik v. Ohio State Bar Association, 436 U.S. 447, 456 (1978).
The author thanks Communications' Legally Speaking columnist Pamela Samuelson the Viewpoints co-chairs for their review comments provided during development of this column.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2019 ACM, Inc.
No entries found