Six Reasons Libertarians Should Reject the Non-Aggression Principle

Don’t be alarmed by the title, I haven’t changed my views.  The title is taken from a recent article written by Matt Zwolinski, who is an Associate Professor of Philosophy at the University of San Diego.  The article in question was posted on a few weeks ago.

In this article, I will summarize Prof. Zowlinski’s arguments and provide my commentary following each point that he makes.  In the article Prof. Zwolinski argues the following points:

1. Zwolinski claims that ALL forms of pollution violate the NAP; therefore, the NAP implies that all polluting activities must be prohibited, no matter how beneficial they may be in other respects.

I would argue that’s a bad interpretation of the NAP.  How does emitting pollution rise to the level of threats or physical assault?  It doesn’t.  If we take climate scientists at their word (LOL), then our very breath should be considered pollution.  By that standard, just breathing the same air as another person rises to the level of assault.  That’s obviously not a valid interpretation of the NAP because there is no aggression.  Further, it doesn’t pass a mens rea test, which is necessary for common law criminality.

Now, just because some minor kinds of pollution don’t have mens rea, that doesn’t mean all forms of pollution don’t.  For example, dumping toxic waste on someone else’s property obviously has some level of mens rea because it is willful property damage, and so we could say that violates the NAP.  It’s a willful act of damage against another person’s property.

Thus, the culpability for pollution lies in the details.  Each incident of pollution must be adjudicated individually.  Further, property damages have to be quantifiable.  If a person can’t quantify damages, how could they possibly claim a violation of property rights occurred?  Damages have to be proven, not speculated.

So right there we can see some pollution is permitted under the NAP as long as there is no quantifiable property damage and no willful intent.  It’s not a violation of the NAP if I accidentally bump into someone, likewise, it’s not a violation of the NAP if I’m engaged in some activity that emits some minor pollution that I did not intend to inflict upon my neighbor, the difference in both cases is intent.

Also, we haven’t touched on the homesteading of pollution rights.  Does a person who moves in next door to a pig farm have a right to sue the pig farmer for stinking up his home?  Obviously not.  The pig farmer was there first, and the new tenant should have known about the smell prior to making his decision to live there because it is his burden to do due diligence for his own investments.

So now we can see that for a particular pollution to violate the NAP, it must be quantifiable in terms of damages, have some level of intent, and also violate homesteading principles.  That’s a far cry from all pollution, which Zowlinski claims.  The NAP is all about aggression.  Which by default means it’s also about intent.  You can’t have aggression without intent.

This argument can be extended to accidental property damage as well.  Is it a violation of the NAP if I damage a person’s car by accident? Clearly not.  I didn’t intend to do so.  Of course, I’m still liable for my own negligence, but I didn’t violate the NAP with my actions.  I didn’t aggress against his property.  I hope you can see how this example applies to incidents of pollution as well.

For a more detailed breakdown on the libertarian environmentalism, check out Environmentalism and Economic Freedom: The Case for Private Property Rights, by Walter Block

2. Zwolinski claims the NAP prohibits small harms for large benefits.  The example he provides is taxing billionaires a small amount and then using that money to provide life-saving vaccinations for poor children.

In this case he’s right, but again, this is a bad way of interpreting things.  What exactly is a “small harm?”

In his example, he’s saying that it’s OK to hold a gun to a billionaire’s head and rob him for millions as long as a majority agrees to it.  I suppose if he’s the one holding the gun, then he might think the harm is small.  If he’s the billionaire?  Not so much.  Anytime threats are used, there is no such thing as a “small” harm to the victim. If escalating threats for compliance aren’t used, then it’s not a tax.

Further, what if that billionaire happens to be rich because he runs a pharmaceutical production facility? What if the millions taken from him by force, that are diverted into vaccine production, would have been used to conduct further research on curing cancer?  What if taking that money means a cure for cancer is not found because the money was diverted?  No one can know that ahead of time.  Markets ALWAYS beat bureaucracies when it comes to efficient allocation of resources.

Billionaires don’t sit on their heaps of money, they invest it.  And they must be good at investing it or else they wouldn’t be billionaires (better than your average bureaucrat at least).  They don’t dive into money piles like Scrooge McDuck at the end of the day.  Their money is always being put to some productive use.

For further information on this topic, see What Is Seen and What Is Not Seen by Frederic Bastiat and Economics In One Lesson by Henry Hazlitt.

3. Zwolinski claims the NAP has an all-or-nothing attitude toward risk.  He cites an example by David Friedman of putting one round in a six shot revolver, spinning the chamber, and then pointing it at a person’s head.  He then compares that to the risks we impose on others by driving on the highway.

Again, he leaves out the intent part of the equation.  Friedman is wrong to use that comparison and so is Zwolinski.  You can’t just ignore the word aggression in the non-aggression principle. The word aggression is tightly coupled with intent and loosely coupled with negligence.

If I close my eyes and swing my fists around in a crowded room, I might not intend to hit someone, but I’m clearly acting with willful negligence. If I close my eyes in what I think is an empty field and do the same thing, and someone runs up to me without me knowing about it, and I end up hitting them, then I’m not at fault because the willful part of the equation is not satisfied.

If I drive down the highway, am I putting others at risk?  Sure.  Am I acting with harmful intent or willful negligence?  Of course not.  Since neither apply, I’m not violating the NAP with my actions. Even though I may be putting others at risk, I’m not aggressing against them.

I’m not just pulling these qualifiers to the NAP out of my ass, they are inherent to the definition of aggression.  We can’t just go around modifying the meaning of words to suit our agenda.  We can’t modify the meaning of aggression to include accidental bumps or mistakes so that we can create absolute expressions, such as Zwolinski does.

4. Zwolinski claims the NAP offers no prohibition on fraud.

Hey, he’s right again!  Fraud is a pretty deep subject, but the principled libertarians will argue that legal protections from fraud are not necessary.  Fraud can only come about if an investor failed to do his due diligence.  Basically it’s your own fault if you get scammed.

That said, a totally free and unregulated market place will automatically come up with its own checks against fraud that don’t rely on courts or police to impose them.  For example, open markets like Ebay or the former Silk Road use a system of ratings and arbitration to weed out the fraudulent vendors.  A vendor who engages in fraud isn’t going to last long in such a market place.

If a completely criminal unregulated underground market like the Silk Road can effectively deal with fraud on its own, I think it’s safe to say that an above ground legal market can also do the same.

That said, some fraud can take the form of a contract violation, and if a contract is broken, the victim has a right to make themselves whole once again at the expense of the violator.  A contract violation is tantamount to theft rather than fraud.

5. Zwolinski claims that, “even if the NAP is correct, it cannot serve as a fundamental principle of libertarian ethics, because its meaning and normative force are entirely parasitic on an underlying theory of property.”  To this end, he provides the example of someone getting attacked because they walked across someone else’s lawn. He concludes that,  “aggression,” on the libertarian view, doesn’t really mean physical violence at all. It means “violation of property rights.”

Well, he’s wrong here.  If he’s arriving at that conclusion, it’s because he’s been reading articles by libertarians who don’t grasp the principle of aggression themselves.  Again, we have to look at intent here.

Under the NAP, if I walk across someone else’s lawn, they have absolutely no right to attack me.  I have not yet demonstrated intent to cause them or their property any harm. The lawn is clearly a publicly accessible place.  However, they could ask me to leave, and I would be obligated to do so.  To establish intent, the property owner would have to take reasonable steps to make sure people didn’t walk on his lawn in the first place.

For example, if a property owner put a large fence around his lawn and posted “Trespassers will be shot” warnings all around his property, then it could be assumed from the start that I was there to cause trouble.  I would have to jump the fence and ignore the warnings to be on his lawn in the first place. Once I jump the fence, now I’m breaking and entering.  At that point there is obvious intent.

6.  Zwolinski claims the NAP implies that there is nothing wrong with allowing your three year-old son to starve to death, so long as you do not forcibly prevent him from obtaining food on his own. Zwolinski cites Rothbard’s Ethics of Liberty to support his argument.

The relevant passage from Rothbard reads as follows:

Suppose now that the baby has been born. Then what? First, we may say that the parents—or rather the mother, who is the only certain and visible parent—as the creators of the baby become its owners. A newborn baby cannot be an existent self-owner in any sense. Therefore, either the mother or some other party or parties may be the baby’s owner, but to assert that a third party can claim his “ownership” over the baby would give that person the right to seize the baby by force from its natural or “homesteading” owner, its mother. The mother, then, is the natural and rightful owner of the baby, and any attempt to seize the baby by force is an invasion of her property right.

But surely the mother or parents may not receive the ownership of the child in absolute fee simple, because that would imply the bizarre state of affairs that a fifty-year old adult would be subject to the absolute and unquestioned jurisdiction of his seventy-year-old parent. So the parental property right must be limited in time. But it also must be limited in kind, for it surely would be grotesque for a libertarian who believes in the right of self-ownership to advocate the right of a parent to murder or torture his or her children.

We must therefore state that, even from birth, the parental ownership is not absolute but of a “trustee” or guardianship kind. In short, every baby as soon as it is born and is therefore no longer contained within his mother’s body possesses the right of self-ownership by virtue of being a separate entity and a potential adult. It must therefore be illegal and a violation of the child’s rights for a parent to aggress against his person by mutilating, torturing, murdering him, etc. On the other hand, the very concept of “rights” is a “negative” one, demarcating the areas of a person’s action that no man may properly interfere with. No man can therefore have a “right” to compel someone to do a positive act, for in that case the compulsion violates the right of person or property of the individual being coerced. Thus, we may say that a man has a right to his property (i.e., a right not to have his property invaded), but we cannot say that anyone has a “right” to a “living wage,” for that would mean that someone would be coerced into providing him with such a wage, and that would violate the property rights of the people being coerced. As a corollary this means that, in the free society, no man may be saddled with the legal obligation to do anything for another, since that would invade the former’s rights; the only legal obligation one man has to another is to respect the other man’s rights.

Applying our theory to parents and children, this means that a parent does not have the right to aggress against his children, but also that the parent should not have a legal obligation to feed, clothe, or educate his children, since such obligations would entail positive acts coerced upon the parent and depriving the parent of his rights. The parent therefore may not murder or mutilate his child, and the law properly outlaws a parent from doing so. But the parent should have the legal right not to feed the child, i.e., to allow it to die.[4] The law, therefore, may not properly compel the parent to feed a child or to keep it alive.[5] (Again, whether or not a parent has a moral rather than a legally enforceable obligation to keep his child alive is a completely separate question.) This rule allows us to solve such vexing questions as: should a parent have the right to allow a deformed baby to die (e.g., by not feeding it)?[6] The answer is of course yes, following a fortiori from the larger right to allow any baby, whether deformed or not, to die. (Though, as we shall see below, in a libertarian society the existence of a free baby market will bring such “neglect” down to a minimum.)


I would argue that Rothbard doesn’t have good reasoning here. He contradicts himself, which means his logic is off at some point.

Rothbard asserts:

“A newborn baby cannot be an existent self-owner in any sense.”


“In short, every baby as soon as it is born and is therefore no longer contained within his mother’s body possesses the right of self-ownership by virtue of being a separate entity and a potential adult.”

So does the baby possess self-ownership or not?  I think it’s clear that a baby does possess self-ownership, even if he can’t act upon that.  Rothbard is right that parents act as trustees.  They voluntarily signed up for this responsibility the moment they decided to have the kid.  That responsibility exists just as firmly as if they had signed a contract with the child saying they will be responsible for his care.  To starve the kid is to violate their role as trustees.

Since the child has no choice in being born, yet he is a being possessing self-ownership, the parents actions of creating his existence implies a three way contractual obligation with the parents agreeing to act as trustees in their capacity as creators to ensure the child continues to live.

Rothbard argues it’s a violation of the NAP to impose positive acts, but the thing he forgets here is that it’s not a violation if a contract is broken that promises they will preform some positive act.  I would argue that act of having a child is the same as signing a contract saying that they will responsibly care for the child.  They voluntarily took on the responsibility, and they can’t just abdicate that responsibility for no good reason.

If there was no self-ownership, the child could assumed to be property, and therefore be destroyed at will by the parents.  But because there is self-ownership, the parents cannot destroy the child without his consent, and since he can’t consent, a trustee obligation must be assumed.

A contract is implied by acts, rather than by spoken word or writing.  The parents trustee obligation last as long as it takes for the child to mature to the point where he is capable of caring for himself.

  • SNW

    The contract with the child begins at conception. The “location” of the child is not relevant. This is another inconsistency with Rothbard’s approach. In other words, is my right to not be attacked eliminated because I am in someone else’s home where I was invited to be?

    • I tend to agree. Conception is only point in the creation process where both parents are involved, so if the trustee obligation is assumed to be split between both parents, it only makes sense that it starts at the point where both parents made a decision together to bring the child into existence.

  • Mike Ellis

    Why does it seem that college professors try to bend over backwards to try to invalidate the NAP? Are they so in favor of violence that they just cannot help themselves but to try to argue its merits? Some of Zwolinski’s points may even be valid but for the most part are moot. My reasoning for that comment is that, even if the NAP is not 100% logically consistent, it is a hell of a lot more logically consistent than the logic of statism. Even if all of his criticisms are correct (which I do not believe) it would still be like living in a utopia by comparison with what we have now. One thing he forgets in his analysis is that it is a Principle, not a law or rule. Principles have at their core a philosophical intent. In other words, if we based all our social interactions on the NAP, then in those cases where strict adherence to the NAP could result in greater violence (if such a situation ever emerged) an alternate solution which resulted in avoiding violence would be found. Strict adherence to any principal to the point of it violating its own philosophical basis makes no rational or logical sense. Only a statist would think that you must follow a principal to its most ridiculous extreme. Rational people, who are dedicated to living in peace and without unnecessary violence will always find ways to settle any dispute in a mutually agreeable way.

    • He doesn’t like the NAP because he’s living in fear. Just look at the arguments he’s making.

      The NAP precludes preemptive regulations and theft funded “social services.” He doesn’t like the personal responsibility that comes with living in a stateless society.

      Any time someone is arguing in favor of initiating violence to achieve some objective, it’s because they are scared.

  • Leon Louw

    There is a much more basic reason why the non-aggression axiom (no “force or fraud”) should be struck from definitions of libertarianism: it’s wrong!

    There are countless examples of the initiation of aggression that are entirely libertarian, such as contact sport, surgery, BDSM, bumper car racing etc. There are also countless examples of rights violations without force or fraud, such as getting into an unlocked car with the keys in and driving off, sticking an arm through a window and taking a wallet off a table, shop-lifting etc.

    What legitimizes aggression and delegitimizes theft and fraud is very simple, namely consent. The negative (in both senses of the word) non-aggression axiom should be replaced by the positive “consent axiom”. According to the libertarian paradigm “no one (including people called government) may do anything anyone else or their property without their consent.”

    (PS: How precisely the elements of formulations of rights are defined is, as with all rights formulations, a jurisprudential question.)

    • If someone consents to physical contact, such as boxing, then it’s not aggression. That’s the entire point of my article. Did you read the article before commenting?

      • Leon Louw

        Of course I read your article. If what legitimises what’s called “aggression” (by standard definition) is consent, why not call it that? Why resort to convoluted linguistics when the consent axiom/principle suffices? I read your piece; did you read my examples of needless complexity and confusion presented by the NAP? Football players plainly doing their best to escape plainly aggressive “hits” are the victims of manifestly deliberate aggression. What legitimises it is consent — as you say in your rejoinder. We can redefine words to suit us, but defining the libertarianism paradigm does not require it.

        • What convoluted linguistics? If someone acts with harmful intent or willful negligence, they are violating the NAP.

          Football players, wrestlers or whatever else are not acting with harmful intent or willful negligence by playing in a consensual sport. Unless a person is retarded, I’m pretty sure they can comprehend that argument.

          The reason why consent isn’t enough is because it doesn’t cover accidental bumps, homesteading rights, or other actions that don’t amount to aggression. Consent is one aspect that’s looked at when considering what constitutes harmful intent or willful negligence.

          There is a reason mens rea has been the standard for common law crimes since early history.

          • Leon Louw

            I’ve known the standard formulation for 50 years, so there’s no point in restating it for me, especially not in its crude form. If you’re wedded to it, so be it. Before leaving the dialogue, I’ll have one more shot at it.

            Commercial competitors can have “harmful intent” and “willful neglect”, and go to great lengths to maximize harm without violating anyone’s rights. Such negative formulations do not define libertarianism.

            In your response, you again do what all libertarians must eventually do when you rely on consent (eg “consensual sport”). The libertarian idea cannot be coherently explained without it. Conversely, such negative and ambiguous terms as “force”, “fraud” and “coercion” serve no useful purpose.

            The core issue to which all coherent explanations of liberty eventually revert is consent. That single word says it all. You never need to ask where libertarians stand on issues. If people to whom things are done directly consent, libertarianism condones; if not it condemns. The same cannot be said for negative formulations.

            Furthermore, libertarianism is not per se opposed to coercion, aggression, harm, fraud etc; only in the absence of consent. Equally, libertarianism opposes countless rights violations that entail none of those (theft, breach of contract, pollution etc).

            Negative formulations are inevitably convoluted and needlessly ambiguous. “Accidental bumps”, for instance, entail no aggression, fraud or mens rea, They are rights violations in their absence … if there is no consent.

            Homesteading does not do anything to anyone or their property, so is a red herring. Incidentally, mens rea is a requirement of some, not all, common law systems, but is not libertarian for the obvious reason that crashing into someone without their consent violates rights regardless of whether perpetrators have guilty minds.

            I’m not one for debating fellow libertarians — more inclined to persuading non-libertarians, hence my interest in how best to do it, so I’ll leave it there and move on. Please do not take offence at me ending a dialogue I hadn’t intended starting.

          • That’s such a perverse view of the markets that I’m hard pressed to know where to begin. If you think market competition is full of harmful intent, I’m not sure how you could in good conscience call yourself a libertarian.

            As for accidents that don’t entail damage, I don’t see how you could construe them as violations of the NAP. I shouldn’t be able to sue someone for accidentally bumping into me because they didn’t have any malicious intent or cause any measurable damage. Consent has nothing to do with that.

  • Jonathan Jaech

    Overall good points, Michael. I would add that fraud, defined as a deliberate act of deception directed towards causing another to act against rational self-interest, is certainly a form of aggression, and cannot be initiated without violating the NAP. For example, deliberately telling another that their brakes are fine, when you know they are not working, thereby causing the other to crash violently (or be at risk of crashing), is clearly aggression. Mens rea, causation, all there.

    Zwolinski is right about the NAP being dependent on the definition of property. It is also dependent on the definition of “person” and “aggression.” So what? Definition of justly property rights and personhood are critical dividing issues in libertarian philosophy based on the NAP. Base it on something else, and there will be other dividing lines.

  • Klondike75

    To answer your question on how pollution rises to the level of threats or physical assault, consider the following. Love Canal is one of almost uncountable examples of pollution rising to the level of assault. It’s a particularly insidious assault since it remains undetected till long after the harm has been done. A community was destroyed by toxins dumped into their waterway, it was so bad that they had to abandon their homes and move away. If you refuse to accept responsibility for the toxins you produce as a byproduct of other activity and just dump them in a river based on the absurd assumption that when you throw it away it goes away, and the next town down stream does the same after Incuring increased costs for water purification since they draw water from that same river, and the increased costs lead them to allow dumping as the first town allowed, at a certain point the river, a common resource, becomes unusable for any purpose by those downstream who will have to move or develop other sources of water at ever greater expense. The same goes for the dumping of raw sewage which causes disease to those downstream who also would not detect the problem till harm had occured, and even then may never know the true cause or source of said harm.
    It takes a special kind of stupid to not see the blatantly obvious as you have not seen here.

    • To quote myself in the article, ” For example, dumping toxic waste on someone else’s property obviously has some level of mens rea because it is willful property damage, and so we could say that violates the NAP. It’s a willful act of damage against another person’s property.”

      I also mentioned negligence in the article, which is also something that should be considered a violation of the NAP when it comes to pollution.

      It’s always best to read the article before calling people stupid, lest you end up looking stupid yourself.