The Incompleteness of the Harm Principle
Jason Pontin has written a perceptive analysis of a timeless question: what changes in law need to be adopted in order to account for technological advances (see “Free Speech in the Era of Its Technological Amplification”)? In answering that question, he takes the right approach by taking up John Stuart Mill’s harm principle, which at its core makes this claim:
“The sole end for which mankind are warranted, individually or collectively, in interfering with the liberty of action of any of their number is self-protection … The only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others. His own good, either physical or moral, is not a sufficient warrant.”
This principle is both profound and incomplete. The last sentence announces a strong antipaternalist manifesto that no government, individual, or group should be able to reform the private preferences of other individuals. That principle is a welcome recipe for social peace: if the rule were otherwise, it would be necessary to decide which individuals or groups occupied the dominant position with respect to other groups. Any interested group would find it most difficult to choose on neutral principles which group should have that preferred position. In contrast, a principle of parity works as well for large and fractured societies as it does for smaller and coherent ones, by making that question irrelevant. Score one for Mill.
Mill does less well, however, in defining how the harm principle works. One possible implication of this principle is, as Pontin notes:
The only principle I can imagine working is yours, where “harm” is interpreted to mean physical or commercial injury but excludes personal, religious, or ideological offense.
Pontin’s version is clearly correct insofar as it excludes “religious or ideological offense” from the category of what lawyers call “cognizable” harms. That odd term “cognizable” is meant to capture this dual understanding. The offense that people take at the conduct of others cannot be dismissed with a wave of the hand, given that these feelings are often deep and long–lasting. They are in fact real harms, subjectively experienced. So the willingness to cut them out of the harm principle cannot rest on a simple denial of the fact, but must rest on the awareness that for the long-term success of the system, each person must waive that claim against all others, no matter how acute the feeling.
Essentially, the comprehensive judgment is that we all are better off when we have to suffer the slings and arrows of this sort of abuse than we are when these attacks are subject to extensive legal control that runs the serious risk of state censorship. The claim, therefore, behind the offense principle is one of universal privilege to hurt the feelings of others, and not a factual claim that there is no harm at all. That difference matters, because when voluntary institutions put speech codes into place for their members, they are responding to a real harm that by consent can be controlled within a limited forum, even if the state cannot dictate that same relationship among strangers. Private ordering has advantages that public ordering cannot match.
Why then the difference? The simplest explanation is that any effort to transport the harm principle to larger social settings gives everyone a huge incentive to become truly offended at the speech of others so that they can now have a lever to suppress their ideas. The angrier you get, the greater your rights. Work yourself into a white heat, and the world must yield to your outrage. That dynamic will not work in a private setting where the organizer of the group has strong incentives to prevent any systematic move to extremes.
Unfortunately, Pontin’s formulation of the rule contains two serious errors that stem from his willingness to accept both “physical and commercial injury.” Physical harm certainly sets out a valid prima facie case that is subject to defenses that relate to consent and self-defense. But commercial harm is much too broad to be treated in the same fashion. The root of the difficulty is that Pontin’s formulation fails to distinguish three separate cases.
The first is commercial harm that is wrought by defamation or violation of trade secrets. Here the libertarian norms against fraud reach the defamation case, which always involves a false statement about a plaintiff that a defendant makes to at least one third party. Trade secrets are a bit trickier, but they are best understood as property claims acquired to information by self-help that people can either keep to themselves or share with any number of persons under a promise of confidentiality.
Second, the single most dangerous version of the harm principle abroad in the land is that competitive injury suffered when a rival firm sells a better good at a lower price should be damned as a form of “unfair” or “ruinous” competition, notwithstanding the well-nigh universal proposition that competitive markets lead to optimal resource allocation. This form of commercial injury (which is surely real, as is offense) will lead to massive cartelization if given any legal respect. Like mere offense, it has to be treated as a noncognizable harm.
Third, the reference to commercial injury does not adequately deal with the position of a natural or legal monopolist in a network industry, whether it be railroads or cyberspace. The correct rule in these cases does not allow the monopolist to charge whatever it wants to whomever it wants. Instead, the long common-law tradition says that the party who holds that monopoly power can never engage in an arbitrary refusal to deal, but must offer his goods and services at reasonable and nondiscriminatory rates. Here the first term is intended to squeeze out monopoly profits, and the second to make sure that the monopolist does not engage in favoritism.
There is nothing whatsoever in Mill that addresses this middle category of public utilities or network industries. Nor does Pontin focus explicitly on this middle category. Nonetheless, it is of profound importance for dealing with the Internet. There are questions as to whether some carriers exert that power, and if so, how they should be regulated, both on pricing and on their ability to turn down content. Yet by the same token, if these services are in competition with each other the case for regulation is much diminished given consumer choices. It is in this middle territory where all the complexities arise. But the Millian harm principle, articulated in 1859, supplies none of the answers on a topic that has engaged the courts in the United States as a key constitutional matter for nearly 140 years, starting with Munn v. Illinois in 1876.
There is no space here to go through that longish history. Readers with interest in it can consult my Principles for a Free Society (1998). Chapter 3 contains a general discussion of the harm principle. Chapter 10 offers a fuller account of its application to common carriers.
Richard Allen Epstein is the Laurence A. Tisch Professor of Law at the New York University School of Law and the James Parker Hall Distinguished Service Professor of Law Emeritus and a Senior Lecturer at the University of Chicago Law School.
Geoffrey Hinton tells us why he’s now scared of the tech he helped build
“I have suddenly switched my views on whether these things are going to be more intelligent than us.”
ChatGPT is going to change education, not destroy it
The narrative around cheating students doesn’t tell the whole story. Meet the teachers who think generative AI could actually make learning better.
Meet the people who use Notion to plan their whole lives
The workplace tool’s appeal extends far beyond organizing work projects. Many users find it’s just as useful for managing their free time.
Learning to code isn’t enough
Historically, learn-to-code efforts have provided opportunities for the few, but new efforts are aiming to be inclusive.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.