Off liberty

Why the ‘harm principle’ may do more harm than good

Why did the libertarian cross the road? ‘Because I bloody well wanted to, and it’s not your job to go and interfere in what I do, so long as it doesn’t cause anyone any trouble!’

This sentiment, beloved of the political right and free marketeers, is known as the ‘harm principle’: that we do not have the right to stop people doing something if doing it does not harm anyone else. It was expressed rather better in John Stuart Mill’s seminal political philosophy text On Liberty, which is celebrating its 150th anniversary this year. Check this out:

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. That the only purpose for which power can be rightfully exercised over any member of a civilised community, against his will, is to prevent harm to others. His own good, either physical or moral, is not a sufficient warrant. He cannot rightfully be compelled to do or forbear because it will be better for him to do so, because it will make him happier, because, in the opinions of others, to do so would be wise, or even right. These are good reasons for remonstrating with him, or reasoning with him, or persuading him, or entreating him, but not for compelling him, or visiting him with any evil, in case he do otherwise. To justify that, the conduct from which it is desired to deter him must be calculated to produce evil to someone else. The only part of the conduct of any one, for which he is amenable to society, is that which concerns others. In the part which merely concerns himself, his independence is, of right, absolute. Over himself, over his own body and mind, the individual is sovereign.

Rousing stuff, especially given the stern, paternalistic Victorian tone. (I’m not being sarcastic—I do think this kind of writing draws warmth and authority from its archaic eloquence.) However, as a manifesto for morality, this tub-thumping libertarianism is inherently, irredeemably inconsistent.

The most common problem with the libertarian outlook is that it’s very rare for any action to not have consequences for the other individuals in society. For example, even if your free choice to take up juggling electrified knives in the bath only poses a danger to yourself, and you have no friends or relatives to be left grieving, someone is going to have to come and clean up when you die in a bloodied, electrocuted pile. Worse, you’re depriving society of your contributions to it which, unless you’re either completely useless or a total bastard, we would hope will come out net positive.

Even worse still, it’s not just fabricated, exaggeratedly dangerous leisure activities which have unexpected moral knock-on effects; your decision to piddle away Sunday afternoon reading a trashy novel means that you weren’t spending the time doing something morally useful, such as trying to help the poor and deprived. Suggesting that we should be compelled to spend all our free time bettering the disadvantaged rather than engaging in hobbies makes us Westerners, as self-determined champions of the individual, squeamish; moral philosophers would bemoan that this conclusion does not accord with our moral intuitions, as though our ill-founded, socially-constrained first guesses are a yardstick by which ethical theories should be assessed. However, regardless of whether the ethical equivalence of inaction drops a moral imperative in our laps, it does illustrate that there is almost no activity which takes place in a moral vacuum where it does not affect others, rather making a mockery of the ‘harm principle’ as stated by Mill.

Even if you do regard necessitating consideration of others in these indirect ways as too prescriptive, or contrive an act in somehow insulated such that you can perform it with no impact on your peers, though, the harm principle still falls down. The reason it falls down is the less flippant punchline to the joke I opened with: imagine a libertarian (let’s call him John) and I are crossing the road.

John is an adult, quite a bright bloke, and is alert, sober, wide awake, and fully compos mentis by any measure you choose to take. In spite of this, for whatever reason he doesn’t see the car which is careering around the corner and will, in a moment, flatten his head against the tarmacadam road surface. Would he rather I put my hand in front of him and stopped him from being run over? If the answer is yes, that’s the end of libertarianism.

John would probably want me to rugby-tackle him to the ground, or even chop his legs off with a samurai sword, in extreme enough circumstances (for example, if the car was an SUV travelling at just the right speed to cause an agonising, drawn-out death): if that’s not ‘compelling him, or visiting him with any evil’, it’s not clear what is.

He made the decision to cross the road, and ‘over himself, over his own body and mind, the individual is sovereign’—but he did so with imperfect information. It seems to me abundantly clear therefore that there is some level of misjudgement of danger where it would be moral to intervene.

The most clear-cut examples of this exist at the nanny state level; from government health warnings on packets of cigarettes to legal minimum tyre tread depths, government has a moral duty to protect us, whose brains are so notoriously ill-equipped to comprehend statistics, from doing ourselves a nasty.

It would be wrong to take this as a vindication of any intervention intended to protect people from themselves. Firstly, freedom is clearly of some importance to human happiness: whether free will is illusory or otherwise, it seems empirical fact that people have more fun when they’re granted the illusion, and allowed to make their own decisions. Secondly, in a world largely comprising shades of grey, the lack of rigorous statistical certainty on many issues leaves us unable to work out whether intervention is right or not; a problem which is especially vivid at the individual level. It’s much trickier to know with certainty how bad an impending road traffic accident is going to be, and thus what evil you should be allowed to visit John with to prevent it, especially in the fraction of second you have to make the decision. John would probably be pretty peeved if I cut his legs off with a samurai sword, and I’d have to be able to demonstrate that I’d saved him from a near-certain, horribly unpleasant fate.

Thus, though we do not know what a perfect moral theory might look like, I cannot see how Mill’s harm principle is a useful contribution to ethics. Indeed, it seems a great shame that Mill gives a heavyweight intellectual’s rubber-stamp to irksome ‘get-your-nanny-state-off-me!’ right-wing–ness. Plus, in tedious sociological recursion, it’s not clear whether our slavish devotion to free will is at least partly a social construct, built thanks to the appeal of the libertarian narrative.


  1. You’re right to point out the limitations of the harm principle and the grey areas around it. Like every other ethical principle ever devised, it has limitations and weak points.

    You’re mistaken to dismiss it on that basis. In doing so, you miss the genuine value it has, and the contribution it makes to ethics.

    As you rightly say, no society could ever live purely by the harm principle. And indeed none does; nor does Liberal Democrat thinking or policy follow it slavishly.

    But the harm principle is valuable. It reminds us how easy it is to slip down the road of everything effects someone else somehow and so justify pretty much any action.

    It pulls us back from that and tells us to find a balance, correctly valuing our autonomy more highly than any idea of society preventing us from causing our relatives grief or the NHS expense.

    That’s it’s real value: not whether it’s absolutely right or wrong, but whether we make better ethical decisions by taking it into account; and in general, we do.

  2. Warnings on cig packets fit in with JSM’s view. Society is cajoling or entreating with someone not to smoke for their own good; but it is not compelling them not to. I have a legal limit on my tyres to prevent my car posing a harm to others.

  3. Mill’s contribution is a good starting point. Especially useful, perhaps, in a society where the police would kick people out of pubs to make them attend church.

    However, you’ve cited some interesting examples.

    I’d agree that intervention in emergency situations might not require consent – especially when time does not permit. That’s a good example you’ve given.

    Before juggling with knives, of course, we should ensure that we have a good funeral plan in place!

    But, I don’t think society has a right to measure the opportunity costs of failing to do good. The logical conclusion would be that there’s only one socially acceptable course of action at any time – that which does most net good.

  4. "This sentiment, beloved of the political right and free marketeers"

    You think so? I certainly hear a lot of rightwingers invoke it but only on a opportunistic basis; I’ve met very few (in real life or online) for whom it’s actually an ideological constant. We’re a long way from the Daily Telegraph and the Mail pressing the government to repeal the Dangerous Drugs Act.

    Also, whilst I agree wholeheartedly with the argument against the broken and woefully simplistic "harm principle" it’s worth remembering that the left can be just as guilty of evoking it. I’m thinking of all the times I’ve heard "what right does anyone have to tell me what to put in my own body?!" from self-styled leftists with regards to the aforementioned drugs laws. The thing is, of course, that the law isn’t just there for individuals…

  5. Anonymous 2:

    My article actually wasn’t written with this ’blog in mind, so it seeks neither to criticise nor endorse Lib Dem policy, so sorry for any confusion there.

    I’m also not sure I believe your points: you haven’t given any actual reason why society shouldn’t try to prevent us causing our relatives grief or the NHS expense, nor any evidence that taking the harm principle into account does lead to better policy.


    Your quite right about the ciggies, that’s not a great example. However, it does open up an interesting question: is writing ‘DEATH!!!!’ all over cigarette packets going to induce fear in people, and does it thus go beyond entreating? Where does ‘cajoling’ stop and ‘compelling’ start? This is another ambiguity in the harm principle, and ultimately it depends if, and if so to what extent, we possess free will.

    However, the car tyres example is not clearly unrelated to the individual. Whilst others’ welfare is a consideration, you can’t dismiss every policy of that nature as purely applying to one’s fellow man. And, as I’ve illustrated, even if we could, a world where people would not put their hand in front of you to stop you getting run over is not a world we should want to inhabit.

  6. This sentiment, beloved of the political right and free marketeers, is known as the ‘harm principle’: that we do not have the right to stop people doing something if doing it does not harm anyone else.

    I’m a long way from "the political right", and my recently published thesis is a direct attack upon the principles of the free market. So I don’t fit your caricature of the kind of person who champions the ‘harm principle’. However, I think your attack upon it, while making some valid points also overlooks some valid points in the defence of that principle and ultimately you end up throwing the baby out with the bathwater.

    I find this is true of much philosophical and political debate, though, and stems from a natural tendency within almost all of us towards absolutism. We see most things as either being right or wrong rather than being — as I believe — right and wrong.

    The strict, absolutist view of the harm principle (as embraced by hardcore libertarians) is indeed rather flawed. Put simply, it does not seem to be a sensible guide by which individual people should live their lives. Or more accurately, it’s not a sensible guide by which individuals should be forced to live their lives (if ‘John’ consciously decides it’s what’s best for him, then I don’t see a compelling reason to prevent him from doing so. Like most of us, I suspect, I have better things to do with my time than spend it trying to change the worldview of hypothetical libertarians).

    However, I see no problem in the proposal that the harm principle should be at the centre of social policy and law-making.

    I would argue (and this is obviously a gross simplification) that the primary purpose of laws should be to prevent harm. How we define that ‘harm’ is of course another, and longer, discussion. And often actions that appear harmful in the short term can be shown to prevent far greater harm over a longer period of time, or vice versa. But I stand by the basic assertion.

    So while individual behaviour should not be regulated exclusively by the harm principle (for reasons you have ably discussed in your post), it seems clear to me that any law that actively contravenes the harm principle; i.e. that can be demonstrated to cause more harm than it prevents, is by definition a bad law.

    So I would suggest that the harm principle, in that restricted sense, should play an important — even a primary — role in legislation.

  7. Jim:

    My apologies to both you and the Matty, who have called me out on a slightly glib opening sentiment. You’re both quite correct that it’s not just the political or economic right who should be singled out for careless misapplication of this principle!

    However, I do disagree with your point about non-absolutism. The harm principle gives the right answer in some cases and is wrong in others, which is what I assume you mean when you suggest that it is both right and wrong. The way that you or I judge the principle to have worked or not is by reference to an external moral standard, which I suspect approximates to somewhere between utilitarianism and our moral intuition, and it is therefore really this latter system which you are subjecting to a number of test cases involving the harm principle. This renders the harm principle irrelevant: it can’t be a useful moral short-cut if it needs to be reassessed in every instance of its use.

    Your restatement of the harm principle in the penultimate paragraph is actually inverse utilitarianism (‘the least bad for the least number’) and, whilst I broadly agree with it, it does not support the harm principle as it is widely understood, or as originally stated by Mill.

    As I said in my response to Anonymous 2, it does not suffice to state that taking the harm principle into account will improve policy: we need evidence that it will before it should routinely be used.

Leave a Reply

Your email address will not be published. Required fields are marked *