Choice, Harm, and Liberalism’s False Retreat from Morality

31 min readDec 30, 2019


Part 0: Introduction

As I begin this piece, I would like to point out that this argument is not meant to inflame anyone. Furthermore, when I refer to “Liberalism” I am not referring to any particular political party but rather to the centuries-old school of political philosophy that bears the name. Specifically, since Liberalism is such a broad school of thought, I will be disputing particular aspects and strains of thought within Liberalism.

I would also like to present specific thanks to John Safranek, Hugh LaFollette, David Brink, and John Stanton-Ife for their works which I pulled significantly from. I attempted to give credit everywhere that I took an idea or quote directly from them. I will also note that many of my “original” ideas I generated on my own, but these men came up with years or decades before me. Finally, the references may be somewhat ugly. I may clean them up at a future date. But for now, this paper has taken too much out of me.

Notably, I intend to first demonstrate that Liberalism’s claim to not govern moralistically (a.k.a. a rejection of Legal Moralism) is only nominal, and that Liberalism operates under its own moral forms. I will then demonstrate that Liberalism’s moral guidelines and axioms are incoherent, trivial, and/or self-defeating. By doing so, I hope to provide the impetus to generate new foundational axioms that can guide our way forward without falling into traditional axioms (such as Christian ethics).

Part 1: The Limits of Law

1A: Pragmatic Limits on the State

It seems highly likely that everyone will acknowledge there are limits to the law, if only at the level of pragmatism. We may very well believe that alcohol consumption is wrong, but we may reject prohibition as a policy due to the myriad negative consequences prohibition creates. In fact, even if prohibition could reduce alcohol consumption to zero, it is reasonable to assume that other, more dangerous drugs may fill the gap. This methodological problem is sometimes referred to as “means-ends limits.” Simply because the state is not all powerful, it must weigh the pros and cons of different policies.

1B: Liberalism vs the “Classical Thesis”

Where Liberalism breaks from other definitions of the state is in its apparent rejection of “Legal Moralism”, in favor of a philosophy of state governance guided by a different fundamental axiom (which we will get to shortly). John Stuart Mill, a prominent Liberal theorist, rejected legal moralism categorically, and other Liberals have claimed that the rejection of legal moralism is a central tenet of Liberalism. From this, Liberals attempt to generate a principled limit to the law beyond the mere ad-hoc practical limits discussed in the previous paragraph. While we will get to the axiom that guides them in this endeavor momentarily, it is important to note a distinction between what we can call strong vs weak legal moralism:

  1. Strong Moralism: an action’s wrongness is conclusive reason to regulate it.
  2. Weak Moralism: an action’s wrongness is pro tanto reason to regulate it.

(Pro Tanto roughly means “to a certain extent”)

(Stanford Encyclopedia of Philosophy ‘Mill’s Moral and Political Philosophy’ 2018)

Liberalism must, at the very least, reject strong legal moralism, although it may acknowledge weak legal moralism; however, this can lead to serious inconsistencies within Liberal theory (as we will see later).

So, if Liberal theory is attempting to demonstrate a principled limit to the law, it must operate under a new axiom with regards to the function of the state. As in, Liberalism breaks from what H.L.A. Hart called the “‘classical thesis’” of governance. Hart identifies this “classical thesis” as:

the law of the city state exists not merely to secure that men have the opportunity to lead a morally good life, but to see that they do. According to this thesis not only may the law be used to punish men for doing what morally it is wrong for them to do, but it should be so used; for the promotion of moral virtue by these means and by others is one of the ends or purposes of a society complex enough to have developed a legal system.

(Hart ‘Social Solidarity and the Enforcement of Morality’ 1983, p. 248)

1C: The Liberal Axiom: Autonomy

In rejection of this, Liberalism begins with arguably its most important (and pernicious) axiomatic belief:

The autonomy of individuals must be respected

One example of this comes from Professor Joseph Raz (who is fairly representative of other Liberal thought on the matter):

One common strand in liberal thought regards the promotion and protection of personal autonomy as the core of the liberal concern.

(Joseph Raz, The Morality of Freedom 1986, p. 204)

But, of course, “autonomy” isn’t a particularly specific term. Returning to Raz once again (because of how representative he is of this autonomy-centric strain of Liberal thought), if I were to ask him “How do you define ‘autonomy’,” his response would follow:

A person is autonomous if he can become the author of his own life.

(Joseph Raz, The Morality of Freedom 1986, p. 205)


Autonomy is an ideal of self-creation, or self-authorship

(Oxford Scholarship Online, The Morality of Freedom, 14 Autonomy and Pluralism, Abstract, 2010)

Nothing in Raz’s responses (or any other autonomy-centric-Liberal’s responses) appear capable of providing a serious definition of autonomy without jeopardizing the rejection of legal moralism. What I mean:

  1. If I say “I have the autonomy/choice to do X”, I can mean one of two things:

1a. “I am physically capable of doing X”

1b. “I am morally justified in doing X”

2. It is unlikely that any Liberal believes that we should govern under a principle centered around 1a, as we would end up with a world where anyone could do anything they wanted so long as they are capable of doing so. This would lead us to a rather absurd world where kidnapping and murdering a child isn’t wrong, since the murderer had the physical capability of performing the action (obviously).

3. So, we appear to be stuck with some version of 1b. as our governing principle, which demands a moral stance.

1D: The Necessity of Liberal Morality

I discussed this issue earlier in my pieces on libertarian politics and, more importantly, on neutralism. Applying the lessons from those pieces to what we just demonstrated, Liberalism cannot hold on to its apparent devotion to neutralism. It must choose a moral stance, which demands discriminating between good and bad choices. Liberalism cannot hold a neutral stance towards differing views of the Good; it must, ultimately, make a choice.

This, of course, presents an issue for Liberal theory, but not an insurmountable one (if you ask Liberals). From the beginning, there have been a variety of attempts to present a moral framework that preserves as much autonomy as possible. And, if we take the measurement of an individual’s autonomy to be the range of options they have in life, it seems, at least on the surface, that Liberal theory may be able to rescue itself.

In order to test this claim, I will begin by examining perfectionist defenses of arguably the most complete defense of an autonomy-centric moral doctrine within Liberalism: John Stuart Mill’s Harm Principle. Upon demonstrating that Mill’s Harm Principle is not capable of salvaging Liberal theory, I will move on to demonstrate why an approach centered around “harm” is doomed to fail. I will finally examine alternate defenses of autonomy-centric Liberalism, and I will demonstrate that none of those are capable of salvaging Liberal theory either.

Part 2: Mill’s “Harm Principle”

2A: An Introduction to the Harm Principle

As with any popular philosopher or philosophical doctrine, Mill’s Harm Principle has been distorted and misrepresented in the mainstream public consciousness for so long that it is difficult to pick out what he actually believed vs. what was added on to his stances by later individuals. In order to properly analyze Mill’s Harm Principle, and honestly demonstrate why it genuinely cannot save Liberalism’s commitments to autonomy and legal neutralism, we must return to Mill’s actual works and analyze what the Harm Principle actually was.

2B: The Basics and Rationale of the Harm Principle

Different Kinds of Restrictions on Liberty

In his work On Liberty, Mill begins his discussion of state coercions of individuals by distinguishing between three kinds of coercion: Paternalistic (unjustified), Moralistic (unjustified), and Harm Principle-based (justified). An example of each:

  1. Paternalistic: A restricting B’s liberty for B’s own benefit
  2. Moralistic: A restricting B’s liberty to make sure B acts morally and does not act immorally
  3. Harm Principle-based: A restricting B’s liberty to prevent harm to someone other than B

(Stanford Encyclopedia of Philosophy ‘Mill’s Moral and Political Philosophy’ 2018)

Mill’s Rationale for the Harm Principle

Mill will further distinguish (or, at least, attempt to distinguish) between genuine harm vs mere “offense”. We will reach that point after we have fully investigated the Harm Principle itself. Before we engage further, we should acknowledge the rationale Mill had which led him to his position:

  1. As noted, Mill believed in a distinction between genuine harm as opposed to mere offense. (We will get to this defense later)
  2. Mill believed the Harm Principle could be applied proactively to prevent actions that present a significant risk of harm
  3. Mill wanted the Harm Principle to have a wide scope (even including relations that had nothing to do with the government, such as parent-child and husband-wife relations)
  4. Mill is focusing on non-consensual harm and he does not believe you can complain about an action harming you if you consented to the risk of the harm occurring

(Stanford Encyclopedia of Philosophy ‘Mill’s Moral and Political Philosophy’ 2018)

Moving on from this, we have a set of considerations of Mill’s which provide greater context and specificity to the Harm Principle.

Mill’s “Free Trade” Exception and the Problems it Creates

The first caveat that Mill makes which appears to present issues for his initial claims is his “free trade” exception. While Mill acknowledges that economic losses can cause serious harm to individuals, he argues that this isn’t sufficient to justify regulating the economy; however, his claimed exception appears to rest on shaky grounds. In theory, he could invoke one of two principles in order to defend this exception:

  1. He could argue that the Harm Principle is only relevant to non-consensual harm and participating in a free market economy is freely risked and therefore consensual
  2. Mill can (and he does) argue that competitive losses are not genuine harms, because they do not deprive economic actors of something they have a right to.

-Mill appears to make the claim that economic actors do not have a right to any particular job but rather they have a right to the fair opportunity to compete for a job.

(Stanford Encyclopedia of Philosophy ‘Mill’s Moral and Political Philosophy’ 2018)

How he defends this right is unclear. He does not appear to ever defend why one only has the right to a fair opportunity as opposed to a job itself, and at no point do any of his principles (Harm Principle or otherwise) provide a conclusive reason or compelling argument for this distinction. And here we see a serious issue that will arise repeatedly for Mill, especially when he attempts to distinguish between “genuine harm” and “mere offense.”

But moving on from this caveat, we must acknowledge other aspects of Mill’s thought which provide further context and specificity for his Harm Principle.

Necessary, but Insufficient: The Harm Principle is not Alone

First, it should be noted that Mill presented the Harm Principle as a necessary but insufficient test for government regulation. As in, for the government to regulate an action, it must violate the Harm Principle; however, just because an action violates the Harm Principle does not mean the government is now justified in regulating the action. Mill believed that the final test for regulation was a utilitarian one, determining whether or not the regulation would lead to a greater common good. This is the first limit on the relevance of the Harm Principle. (Note: It is important to remember Mill was, at heart, a utilitarian. On Liberty can be understood as a way for Mill to try to reconcile Liberal rights with utilitarian ethical theory, driven by his axiomatic belief in “autonomy”)

Only Certain Kinds of “Liberty”

Second, Mill is not defending liberty per se. He is defending what we can call “basic liberties” or perhaps “foundational liberties”. His Harm Principle is meant to be a test for government regulation of these basic liberties, not of any form of liberty/action. Mill broadly identifies these basic liberties as belonging to three main categories:

  1. Liberties of conscience/thought and expression
  2. Liberties of tastes, pursuits, and life-plans
  3. Liberties of association

(Stanford Encyclopedia of Philosophy ‘Mill’s Moral and Political Philosophy’ 2018)

These basic liberties can, and do, obviously cast a wide net, but they aren’t all encompassing. This is the second limit on the relevance of the Harm Principle.

Is Harm Even Necessary to Justify Regulation? Mill is not Clear (for now)

Third, Mill actually appears to concede that harm is not necessary to justify regulation, which would entirely undermine his Harm Principle. In order to see how we might salvage Mill’s Harm Principle, let’s investigate these aspects of Mill’s thought.

Mill makes many restrictions on individual liberty that appear designed to benefit others, rather than to prevent harm. These can be grouped into two main categories:

  1. Enforceable duties to others that benefit them, such as the duty to give evidence in court and Good Samaritan duties
  2. Enforceable duties to contribute one’s fair share to the provision of various kinds of public goods (incl. common defense, community infrastructure, mandatory education, state support of the arts, etc)

(Stanford Encyclopedia of Philosophy ‘Mill’s Moral and Political Philosophy’ 2018)

Each of these present somewhat different issues for Mill’s Harm Principle, but neither presents insurmountable obstacles.

Type 1 of Enforceable Duties: Duties to Others that Benefit Them

Mill’s central claim to support the enforcement of these duties is that failing to enforce them leads to harm for others. But is this true? It does not seem entirely clear. One might appeal to a kind of demandingness problem: certainly I could be out there volunteering and donating all of my money, but is it reasonable to expect everyone to be doing that 24/7? It doesn’t appear so. While we will return to this kind of act/omission discussion later (especially since it is used in other attempted defenses of libertarianism), but what is really at stake with Mill’s claim is how to determine the baseline situation against which we can determine harm and/or benefit.

We appear to measure harm counterfactually. You could have done X and I would not have been affected in a noticeable manner but you did Y and now I am worse off. Ignoring the ambiguity with how we determine what is or is not noticeable, there is an appeal to this kind of argument. The issue of course, again, is how do we define the baseline. It would be an entirely uncompelling argument if Mill claimed that the baseline was the world in which you fulfilled your duties, since any world in which you didn’t fulfill your duties would harm others. It would appear almost circular. We need an independent rationale to determine the baseline.

To further investigate this, let’s look at an example.

The Drowning Child

Let us assume there is a child drowning and I have the opportunity to save that child. Let us further assume that for whatever reason, I decide not to save the child. I have not directly harmed the child; however, even if my failure to rescue the child does not harm him, he is harmed by drowning. Here, we see a potential way to square the Harm Principle with this type of enforceable duties. We can broaden our definition of the Harm Principle:

Even if restrictions on A’s freedom, requiring him to benefit B, cannot be justified on grounds of preventing A from harming B, they may nonetheless be justified on the grounds of preventing harm to B

(Stanford Encyclopedia of Philosophy ‘Mill’s Moral and Political Philosophy’ 2018)

The First Ambiguity and a New Harm Principle

Here we are presented with the first ambiguity in Mill’s definition of the Harm Principle: Mill talks both about preventing one from harming others and about harm prevention. And, as we saw in the drowning child example, these two things are not the same.

So, we now have to differentiate between two variants of the Harm Principle:

  1. HP1 (Anti-Harming): A can restrict B’s liberty only in order to prevent B from harming others
  2. HP2 (Harm-Prevention): A can restrict B’s liberty only in order to prevent harm to others

(Stanford Encyclopedia of Philosophy ‘Mill’s Moral and Political Philosophy’ 2018)

HP1 is narrower than HP2 (it can be understood as a subset of HP2), and considering Mill’s dedication to these enforceable duties, it appears HP2 is closer to Mill’s actual Harm Principle.

Type 2 of Enforceable Duties: Contributions to the Common Good

The issue with Mill’s belief in this duty is that public/common goods are structured such that the effect of any one individual’s contributions on the provision of said good is typically negligible (ignoring, for example, taxation on the uber-wealthy). So, the cost of an individual failing to provide for such goods does not seem to meet Mill’s criteria of “harm.” The negative impact would be small and would be spread widely amongst all of the users of the common good. It therefore would not breach “a distinct and assignable obligation to any other person or persons,” in Mill’s language.

While one can certainly point to a Tragedy of the Commons situation as perhaps a pragmatic or utilitarian argument for regulation, we cannot ever reach this utilitarian calculus if we stay true to Mill’s Harm Principle because it does not violate said principle. Of course, as we have noted when we entered into this discussion of enforceable duties, Mill appears to present rationale to regulate liberties for the benefit of others, not merely to prevent their harm. We were able to salvage the Harm Principle from the first kind of these enforceable duties, perhaps we can salvage the Harm Principle from the second kind as well.

It should be noted that this type of enforceable duty presents issues for both HP1 and HP2 (as defined above). One Millian response is to return to the distinction between mere liberty and basic liberties, as I explained earlier. If we take Mill at his word that the Harm Principle is only intended to be used to protect egregious violations of basic liberties as opposed to any kind of liberty, we can present another new distinction between different kinds of Harm Principle formulations.

The New Formulas of the Harm Principle: Liberty vs Basic Liberties

  1. HP1A (liberty anti-harming) A can restrict B’s liberty only in order to prevent B from harming others
  2. HP1B (basic liberties anti-harming): A can restrict B’s basic liberties only in order to prevent B from harming others
  3. HP2A (liberty harm-prevention): A can restrict B’s liberty only in order to prevent harm to others
  4. HP2B (basic liberties harm-prevention): A can restrict B’s basic liberties only in order to prevent harm to others

(Stanford Encyclopedia of Philosophy ‘Mill’s Moral and Political Philosophy’ 2018)

Mill’s points on basic liberties and harm prevention (as we explored throughout this piece) seem to point to HP2B as the proper formulation of the Harm Principle, following Mill’s actual beliefs.

Harm Still Isn’t Necessary: Paternalistic Exceptions

One final point I would like to note is that Mill provides an exception to his categorical rejection of paternalistic regulations: he believes it is reasonable that the government prevent an individual from selling himself into slavery. His argument is still autonomy-centric, noting that one can restrict current autonomy to maintain future autonomy. This argument is, unfortunately, slippery at best, and can very well lead to a whole bunch of consequences that Mill would not look forward to. The key takeaway here is that the violation of the Harm Principle is not necessary to justify government regulation of an individual’s liberty.

Offense vs Harm

Perhaps the last place where Mill possibly has a way to salvage his Harm Principle is when it comes to his distinction between genuine harm and mere offense. In his mind, genuine harm occurs when someone violates a clear obligation to respect the basic liberties and rights of a distinct person or set of persons. The issue, of course, is that Mill is not very specific about what makes an obligation “clear” or what makes a person or set of persons affected by an action “distinct”.

Furthermore, as noted above with his “free trade” exception, Mill’s rationale for his identification of basic liberties can be used to justify regulation for a wide variety of actions. Mill actually presents an argument for the regulation of public drinking (or, more broadly, public indecency) as an exception to the Harm Principle. But his argument ends up being so vague, that this exception could be used to pretty much swallow up the entire Harm Principle. If it is the publicity of drinking that makes it a problem, Mill is appealling to offense, which threatens to undermine his entire Harm Principle.

Ultimately, we are presented with an issue of how to define harm as opposed to offense. A more recent Millian, Feinberg, attempted to distinguish between them through an analysis of the concept of “nuisance.” Feinberg actually concedes that there are certain kinds of offenses, nuisances, that can be regulated, and he introduces a kind of balancing test to differentiate between nuisances that can be regulated vs those that cannot be. Of course, his balancing test is controversial at best, and still appears to undermine Mill’s own position.

We will return to this dilemma of how to define “harm” in a further section.

(Stanford Encyclopedia of Philosophy ‘The Limits of Law’ 2006)

Harm Principle Recap and Conclusion

As noted above, the definition of the Harm Principle that remains most faithful to Mill’s actual thought is: “A can restrict B’s basic liberties only in order to prevent harm to others”

Mill does not present any bulletproof rationale as to why certain things count as basic liberties while others do not. But even if we take his rationale to be compelling, this definition of the Harm Principle is not much like modern “libertarian” positions. Its protections are far narrower (basic liberties vs liberty in general) and its justifications for government regulation far broader (preventing harm to others vs just preventing individuals from harming others). And yet, Mill’s argument remains fairly coherent throughout, once one takes his whole corpus of thought into account.

The final issue for Mill which ultimately knocks down his Harm Principle as a true savior for Liberal theory is that Mill’s own arguments either explicitly state or ultimately imply that the Harm Principle is neither a necessary nor sufficient justification for government regulation. Mill himself states it is insufficient for government regulation, providing at best a pro tanto justification for regulation, which must then pass a utilitarian calculus to justify an actual government regulation. And Mill’s arguments, as we demonstrated, clearly demonstrate the Harm Principle is not necessary to justify government regulation (see: the Paternalistic exceptions and, arguably, the enforceable duties).

So, Mill’s Harm Principle is ultimately insufficient to salvage Liberal theory. First, it must rely on other moral theories to justify government regulation (utilitarianism), violating legal neutralism. It could be argued that a government operating under Mill’s influence would still be a moralistic one, simply operating under the Harm Principle + Utilitarian Calculus to determine regulations of behavior.

And, ultimately, Mill’s Harm Principle does not present a solid defense of autonomy. The Harm Principle is neither necessary nor sufficient to justify the regulation of the liberties of others. Autonomy has not been seriously defended, even in a more limited sense, such as with the “basic liberties” that Mill argued for. As of now, the axiom of autonomy remains undefended. I will now move on to a further examination of harm and then I will engage with more recent liberal and libertarian defenses of autonomy, and demonstrate that they are incapable of salvaging the axiom of autonomy.

Part 3: What is “Harm?”

As noted in the section on Offense vs Harm when we were examining Mill’s Harm Principle, there appears to be a serious issue with how we define Harm. None of Mill’s own definitions appear to be concrete enough to pull out a definitive list of basic liberties. Instead, we have vague, abstract principles that sound nice but offer us little in the way of real guidance. This is perhaps a deeper issue, not only with Mill but with a huge swath of Liberal thought: how do we define Harm?

In order to demonstrate this, let us begin with the most constrained, specific definition of Harm possible. We will furthermore be operating under a more modern libertarian principle which is simply that, “I can perform any action so long as the action does not harm others.” In which case, more modern libertarian principles align closer with the HP1A formulation of Mill’s Harm Principle. By beginning with the most specific definition of harm possible and getting broader as we grapple with counterexamples, we will demonstrate that “harm” is insufficient to generate any meaningful moral imperative, and therefore much of Liberal theory loses its foundation.

The Most Specific Form of Harm: Direct, Targeted, and Physical

I think we can all agree that punching someone else in the face is a clear instance of harm. Certainly there can be caveats in our judgment of the action: was it in self-defense? are these two people in a boxing match they both consented to? But either way, there is a clear instance of harm. One individual physically hurts another individual within a particular space at a particular time, presumably without consent.

There are Other Harms!

But is this the only thing that we consider wrong in the world? Imagine an individual cyber-bullies another person to suicide. There was no direct physical contact. Perhaps these two people don’t even know each other offline. But we have seen numerous cases, unfortunately, in recent years where this has happened, and I believe it is fairly self-evident that, even without direct physical contact, serious emotional damage to another individual is wrong.

A Broader Definition of Harm: Targeted Physical, but Both Direct and Indirect

So, perhaps we can broaden the definition of harm slightly. One cannot perform actions targeted at another individual that post a direct or indirect threat to their physical well-being. So, we still have a reasonably constrained “harm principle” which would allow a fairly sturdy defense of individual liberties while still preventing serious wrongdoing.


The issue here, of course, is that there are serious problems with limiting our definition of harm to just actions targeted at another individual or distinct group of individuals. I am not going to defend Mill’s enforceable duties here, nor am I going to get into the positive vs negative rights distinction yet (that comes later). Rather, I will present a different example. Imagine I somehow am able to acquire the equipment and time to perfectly counterfeit $1 billion. It is not important how I acquired this equipment, but suffice it to say, I am counterfeiting this money. My actions not only presumably increase inflation (perhaps not noticeable on a national scale but I will almost certainly affect at least home prices in certain localized areas) which harms a broad swath of people, but I am also devaluing the hard work done by others. I am cheating. This action isn’t targeted, and the harm is spread across many people. Is this not wrong? We have laws against counterfeiting, and I find it difficult to imagine that many people will defend counterfeiting with anything resembling a sophisticated argument.

A Yet Broader Definition of Harm: Physical, Direct and Indirect, but either Targeted or Untargeted

So we now modify our definition again, incorporating our new understanding, and we reach a definition of harm that comprises any direct or indirect harm to physical well-being, whether targeted or untargeted. This, of course, is already pretty damn broad, and it seems reasonable to assume that pretty much any government regulation could at least be justified to some degree under this broad definition of harm. That provides little solace or hope to a libertarian. But it gets worse, unfortunately.

The Issue with Modern Libertarian Usage of “Harm”

Much of our common sense definitions of right/wrong have to do with emotional well-being. Here, we see a very important distinction with modern libertarian descriptions of harm as opposed to Mill’s Harm Principle:

Modern Libertarians apply their “harm principle” broadly: it is not only a justification for government action, but for individual action as well.

In other words, a modern libertarian will argue that they are justified in performing some action if it doesn’t harm others. We have stepped outside of the bounds of government regulation and moved to individual morality. Had we stayed within the bounds of government regulation, this next point might be weaker, but as a modern libertarian ties harm to individual morality, their doctrine holds an even deeper problem.

Yet Another Problem for Our “Harm” Definition

Let us assume that an individual A bullies an individual B not to the point of suicide, but rather to the point of mental illness, perhaps an eating disorder, anxiety, and/or depression. If this “harm” has to do with individual morality, it becomes intensely clear that this kind of emotional damage is wrong. We do not condone emotional bullying. There is certainly some wiggle room with regards to banter, etc, but it appears unlikely that there is a solid defense of an individual being intentionally malicious and emotionally bullying another person.

If we are continuing with this “harm principle” as a justification for individual morality as opposed to mere state regulation, as modern libertarians do (and, to be fair, Mill wanted to, if he could’ve gotten his principle to be meaningful), we have to add yet another caveat to our definition, arguing about emotional damage. And said emotional damage can either come in a direct (a.k.a. physical and temporal proximity) or indirect (ex: online comments from hours ago) manner.

The Broad, but Most Accurate, Definition of “Harm”

So, we now have a definition of harm, informed by our everyday morality, in which any kind of physical or emotional damage, in either a direct or indirect manner, either targeted or untargeted, counts as harm. And if we take the modern libertarian at his word, all of these actions are impermissible not only for the government to allow, but also for him to perform as an individual. They are, according to the libertarian using “harm” to guide her personal morality, wrong for her to do.

The Ultimate Issue with “Harm” as a Guiding Moral Principle

And the problem becomes, of course, that any plausible action will conceivably harm someone, either physically or emotionally. Without some other principle differentiating between kinds of harm, the mental discomfort felt from being informed your child has been killed is not categorically distinct from the mental discomfort a conservative feels when seeing a gay couple. As I noted in a previous piece, any decision the government takes, too, will harm someone, even solely at the level of limiting their options. Liberals attempt to avoid this by proclaiming legal neutralism, rejecting legal moralism, and embracing axiomatic autonomy. But this, of course, is nonsense. None of these principles hold up to investigation.

“Harm” is not sufficient for use in any individual morality or governing state philosophy. No form of the harm principle can defend the axiom of autonomy. Mill’s extra principles provided a better argument, but utimately undermined themselves to the point that Mill’s own Harm Principle was incapable of salvaging Liberal theory.

Are there any other defenses of autonomy as an axiom that might salvage Liberal theory? There are certainly attempts. In the following section, I aim to investigate these alternative defenses and demonstrate why they are insufficient to salvage Liberal theory.

Part 4: Other Liberal Defenses

A New Defense of Autonomy: A New Argument for a Limited State

It should be noted: libertarianism and much of progressivism and conservatism (especially in their modern incarnations) are driven by Liberal axioms and thought. Libertarianism is usually the most open about this, making it the best place to investigate whether Liberal axioms and theory can hold up to scrutiny.

Throughout the 20th century, libertarianism gained considerable favor, with greater concerns for personal autonomy combined with the forceful arguments of individuals such as Hayek, Friedman, and Nozick. In particular, all three of them presented new arguments for a limited state. More specifically, a minimal state. In its most vulgar expression, this argument can be summarized as “taxation is theft!” Of course, this is an unfair depiction of the nuance of the argument; however, the argument still does not hold up to scrutiny. I intend to demonstrate two things:

  1. The Libertarian arguments for autonomy are self-defeating. They contradict themselves to the point where they cannot be saved.
  2. The Libertarian arguments for a limited state lead to an acceptance of a state greater than the minimal state, and by extension provide no defenses against an even broader state. In other words, the Libertarian’s own arguments allow for a state far larger than a mere minimal one.

We will begin by briefly describing modern advances in libertarian theory and demonstrating how they differ from Mill. Then, I will demonstrate how these new principles are self-defeating and leave libertarian and, by extension, Liberal theory without a saviour.

A Description of the “New” Libertarianism (Taken Largely from Hugh LaFollette’s Work)

It should be noted, first, that “new” libertarianism holds the importance of personal autonomy to be its central commitment, in the same way that Mill and other autonomy-centric Liberals believe. All of these thinkers believe that the individual should be as free from the interference of others as possible. This is generally described in one of two manners:

  1. In the language of liberty: Personal liberty is the supreme moral good. One’s liberty can only be justifiably restricted if that person consents to the restriction
  2. In the language of rights: Each individual has a set of negative natural rights, most commonly life, liberty, and property. No one may perform an action that violates his rights without his own consent. And since, as I pointed out in my piece on rights, every right contains an equivalent duty, every other person must respect everyone else’s negative rights

These two description ultimately end up at the same point: individual liberty/personal autonomy is paramount. Whether in the bare language of liberty or in the language of rights, this point shines through.

It is important to note that libertarians acknowledge “positive” rights, or claims to something of another person, whether that be property or time, etc. But these positive rights can only emerge out of consensual agreements, such as contracts. For the libertarian, there are no general positive rights and therefore no general positive duties. This is notable because general positive rights would require the violation of general negative rights and negative rights are the supreme good for libertarians.

For the libertarian, an individual can never be morally forced to give up their life, liberty, or property. One’s own negative rights can only be overridden with his own consent. So, this moral/legal prohibition makes sure that an individual’s liberty cannot be restricted in any way without his consent. This leads to us being able to illuminate two important features of libertarianism (and of much autonomy-centric liberal theory):

  1. The primary purpose of negative general rights is the protection of individual liberty, to insure that no one’s life is restricted without his consent. This largely draws from a Kantian perspective of individuals as ends, not mere means, with Nozick arguing “they cannot be sacrificed or used for the achieving of other ends without their consent.”
  2. The libertarian, exemplified by Hayek and Nozick, holds that a sufficient reason to reject any alleged moral rule or principle of distributive justice is that rule or principle restricts someone’s freedom without their consent. This is the reason Nozick rejects Rawlsian liberalism and why Hayek rejects distributive economic policies.

This kind of new libertarianism has a degree of plausibility to it, and it is certainly an emotionally compelling thesis. Most of us want to live our own lives, find our own destiny, etc. We don’t want to be beholden to others. But, just because a theory is plausible and emotionally compelling, doesn’t make it true. What I will attempt to demonstrate is that negative general rights do not protect autonomy the way libertarians claim they do, and actually end up defeating themselves. I will furthermore attempt to demonstrate that even the most austere libertarian must acknowledge the moral justification of non-consensual limitations on freedom with regards to moral rules and principles of distributive justice.

(Hugh LaFollette ‘Why Libertarianism Is Mistaken’ 1979)

Libertarianism Defeats Itself

I’ve already noted how there is no real distinction between positive and negative general rights. Putting that aside for now and beginning simply with the new libertarian’s principles, I will demonstrate how negative general rights necessarily infringe on the liberty of others. I will furthermore note that the issues with the definition of harm that we explored in the previous part will ultimately make the new libertarian’s principles incoherent.

As noted in a couple of my previous posts, the mere existence of other persons in one’s proximity limits your own freedom without your consent. There are certain actions I can no longer take that I might want to, simply because others are around. I did not consent to this. And yet, a system of negative general rights limits the freedom of everyone within the society without anyone’s consent. The libertarian’s own principles designed to protect personal autonomy end up limiting personal autonomy.

As Hugh LaFollette put it, with a historical example:

This is even more vividly seen when we look at an actual historical occurrence. In the nineteenth century American slaveholders were finally legally coerced into doing what they were already morally required to do: free their slaves. In many cases this led to the slave owners’ financial and social ruin: they lost their farms, their money, and their power. Of course they didn’t agree to their personal ruin; they didn’t agree to this restriction on their freedom. Morally they didn’t have to consent; it was a remedy long overdue. Even the libertarian would agree. The slave­ holders’ freedom was justifiably restricted by the presence of other peo­ple; the fact that there were other persons limited their acceptable alternatives. But that is exactly what the libertarian denies. Freedom, he claims, cannot be justifiably restricted without consent. In short, the difficulty in this: the libertarian talks as if there can be no legitimate non‑consensual limitations on freedom, yet his very theory involves just such limitations. Not only does this appear to be blatantly inconsistent, but even if he could avoid this inconsistency, there appears to be no principled way in which he can justify only his theory’s non-consensual limitations on freedom.

(Hugh LaFollette ‘Why Libertarianism Is Mistaken’ 1979)

Combine this example with the actual expansiveness of “harm” as informed by our everyday conscience, and we see serious problems for this new libertarianism’s own principles.

First, there is no clear limit to the harm to an individual’s life when it comes to that negative right, so there is no clearly defined “private sphere” which the new libertarian’s arguments can defend against interference from government and others. Furthermore, as we described in that earlier piece, there is no real distinction between positive and negative rights. And, when we combine those two points with the fact that justice sometimes commands non-consensual limits on people’s freedom, we have even more issues. The final nail in the coffin is realizing that negative general rights can restrict the autonomy of people far more than positive general rights, eliminating any further defence of this position from a new libertarian perspective. Hugh lafollette again:

For example, forcing a slaveholder to free his slaves would limit his freedom more than would a law forcing him to pay ten percent of his salary to educate and provide health care for his slaves. Or forcing Hitler to not take over the world (in other words, forcing him to recognize others’ negative rights) would limit his freedom more extensively than would forcing him to support, by his taxes, some governmental welfare program. Yet the libertarian concludes that redistribution of income is unjust since it limits the taxed person’s liberty without his consent. If redistribution is unjust for that reason, then so are libertarian constraints. Libertarian constraints also limit personal liberty without consent.

(Hugh LaFollette ‘Why Libertarianism Is Mistaken’ 1979)

Second, this furthermore goes to show that the positive rights and duties of a non-libertarian scheme of redistributive justice, such as a welfare state, cannot be knocked down by new libertarian principles. The new libertarian must acknowledge that since libertarian constraints themselves demand limitations on personal autonomy, they can’t accept those constraints and then reject redistributive justice schemes like the welfare state for demanding limitations on personal autonomy. Accepting the libertarian constraints removes the new libertarian’s ability to reject non-libertarian redistributive justice and removes their ability to argue against general positive rights. Ultimately, the new libertarian arguments cannot accomplish any of their goals, as they defeat themselves.

The First Potential Response by a New Libertarian

A New Libertarian might attempt to respond by claiming that there is some kind of significant difference between redistributive policies and the constraints that the libertarian proposes to protect personal autonomy. This difference makes it such that the libertarian constraints are always justified while the redistributive policies are never justified. An argument might go that even though both limit personal liberty without consent, only the new libertarian constraints have the goal of protecting individual rights, which is the new libertarian’s ultimate goal.

(Hugh LaFollette ‘Why Libertarianism Is Mistaken’ 1979)

The issue that emerges, of course, is that this is ultimately circular and just as self defeating. New Libertarian conceptions of rights are grounded in personal liberty/autonomy. In other words, they are grounded in the right to not be coerced when it comes to life, liberty, and property (subject to some version of the “harm principle”, which we already demonstrated is incoherent). From this, we can determine that there is no principled way that a concern for personal autonomy can only generate negative rights, especially since, as noted before, negative general rights can restrict autonomy more than positive general rights.

So no distinction based on “negative vs positive general rights” is capable of defending the New Libertarian position and no mythical “significant difference” between negative rights and positive redistributive justice holds while not contradicting the New Libertarian’s own principles. So, it appears the New Libertarian’s principles are self-defeating.

Other Potential Replies

A Full Embrace of Moralism

Libertarians may argue that “Liberty” is a normative concept instead of a descriptive one. I have already discussed this in a previous post, and even if we take this to be true, the libertarian has not presented any new defense of his own paramount belief in personal autonomy. Arguments around negative general rights and coercion fail. And furthermore, we would then have to determine whether or not a particular coercion was justified by whether or not the restricted action is just. This requires a new comprehensive moral theory, which new libertarianism fails to present.

Liberty Should be Maximized

This is, of course, subject to lots of problems.

First is a purely methodological one: how do we measure whether or not the actions of A actually reduce or increase the liberty of B? Must we count every possible action of B in any situation? Where is the division between different actions? As in, if I set my foot one inch to the right of where I might have, does that count as a new choice/possible action? These questions lead us down rabbit holes no new libertarian (or anyone else) appear to have answered.

Second returns to Mill’s paternalistic exception to constain personal liberty. Mill’s exception aims to preserve future autonomy by limiting present autonomy. And yet, as we noted with Mill earlier, this argument threatens to subsume Mill’s own principles and could theoretically justify a far greater set of government regulations.

Third is Nozick’s response. Nozick noted that maximizing liberty applies a kind of utilitarian calculus where I might justifiably limit the liberty of one person if it increased the liberty of many others, and the net increase in liberty within the society is increased.

“This. . . would require us to violate someone’s rights when doing so minimizes the total (weighted) amount of violations of rights in the society.”

(From Hugh LaFollette’s ‘Why Libertarianism Is Mistaken’ 1979)

Nozick rejects this proposal on this ground, as it would violate the general negative rights of individuals. If there is a categorical committment to personal autonomy, this view cannot be held.

Finally, if a libertarian attempted to reconcile this by arguing that every person should also have equal liberty, this could be used to justify incredibly repressive governments, bringing everyone down to the lowest common denominator.

So, this response appears to fail as well.

A Recapitulation of New Libertarianism’s Self-Defeat

There are other potential responses from New Libertarianism to these criticisms, although they are largely dealt with by Hugh LaFollette and John Safranek. At this point, I believe I have presented a comprehensive demonstration that New Libertarianism defeats itself, and that none of these new arguments can salvage Liberal theory. The axiom of autonomy appears to remain incoherent. It seems highly improbable that any political theory beginning with a committment to personal autonomy as a central good can be coherent and logically compelling, even if these Liberal theories may be very emotionally compelling.

Part 5: Conclusion

We have seen throughout this piece that Liberal theory is built upon an indefensible axiom: the central committment to personal autonomy. Autonomy itself lacks a moral character, requiring Liberal theory to abandon its rejection of legal moralism even if they claim they still reject it and try to redefine their principles as being “non-moral” when they are moral in character. This acceptance of legal moralism requires Liberal theory to abandon its adherence to legal neutralism, as it must ultimately choose a particular moral way to govern society. Furthermore, no Liberal moral principle appears capable of meaningfully saving these central committments. Mill’s Harm Principle ends as being far narrower in its protection of individual liberties while being far broader in its justification of government regulation. Furthermore, his principle ends as being neither necessary nor sufficient for government regulation of liberties, meaning it’s incapable of defending Liberal principles. No feasible definition of harm in line with our everyday morality appears capable of defining and defending a clear private sphere of individual liberty, separate from the public sphere subject to government regulation. And none of the new libertarian defenses of personal autonomy, whether through general negative rights or other appeals to liberty, are capable of salvaging Liberal theory and committments. Autonomy-centric Liberalism has failed.

I do not intend for this essay to be a call for a return to old justifications for state intervention and personal morality. I do not advocate for the Divine Right of Kings and traditional Christian morality. Instead, I intend to use this essay as an impetus to develop new comprehensive theories of the state and of personal morality, that are in line with our everyday consciences and a coherent understanding and metaphysics of the world. I consider that to be the defining challenge of the next decade in political theory. And I am looking forward to seeing lots of interesting responses.


Safranek, John, The Myth of Liberalism, The Catholic University of America Press, 2015

LaFollette, Hugh, ‘Why Libertarianism Is Mistaken’, Justice and Economic Distribution, 1979, p.194–206

Stanton-Ife, John, ‘The Limits of Law’, Stanford Encyclopedia of Philosophy, 2006

Brink, David, ‘Mill’s Moral and Political Philosophy’, Stanford Encyclopedia of Philosophy, 2018