The principle of universalizability is not only a necessary but also a sufficient condition to demonstrate that the non-aggression principle is the only logically possible non-arbitrary coercive rule.
30 OCT 2023 – by GIOVANNI BIRINDELLI
Philosophers of liberty can be seen as climbers. The level of logical coherence in their defense of non-arbitrary coercive rules can be seen as the altitude they reach. Consequently, the non-aggression principle can be seen as the peak of the highest mountain.
In the past, several climbers approached the summit1, but the first one to conquer it was Murray Rothbard. The path he chose to reach it, however, was traced on the slippery rock of ‘human nature’. As it has often been observed (Antiseri, Giorello 2008, 90; Hoppe 1988, 61), this concept is problematic: it is partly influenced by cultural factors to the point of potentially being arbitrary. To reach the summit along this path, Rothbard, for example, had to assume that, outside the field of economic theory, value can be objective, not just subjective. Specifically, he claimed that in the field of ethics it is objective (Rothbard 1998, 11-12).
After Rothbard, Hans-Hermann Hoppe (1988; see also Kinsella 2011) reached the same peak (more precisely, one which is very close to it2). However, he traced his path on the solid rock of logical-aprioristic reasoning, without ever touching the slippery one of ‘human nature’. Not only does the path traced by Hoppe reach the same summit, but it is brilliant, enriching, and beautiful. Rothbard (1988) immediately recognized the superior quality of the rock on which the new path had been traced.
The path traced by Hoppe begins with the principle of universalization as a necessary condition to demonstrate that a rule is logically justifiable. If a norm is not universalizable, that is, if it violates the logical and binary principle of equality before the law, then by definition it is arbitrary (“this is right because I say so”) and therefore unjustifiable.
Hoppe believes that universalizability is only a necessary condition to demonstrate the justifiability of a norm, not a sufficient one. Consequently, after some lengths of rope, he leaves the path of the principle of universalization, which he believes does not reach the summit, to trace a different one: that of argumentation ethics. Again, we are fortunate that Hoppe makes this choice. The new path he traces is extremely panoramic and leads to the summit: by exclusively scaling the solid rock of logical-aprioristic reasoning, Hoppe demonstrates that the non-aggression principle (which passes the test of universalization) is the only rule in favour of which one can consistently argue.
However, the assertion that universalizability is only a necessary condition to demonstrate the justifiability of a norm is incorrect: it is also a sufficient one. In other words, the path of the principle of universalization also leads to the highest summit: the non-aggression principle as the only possible non-arbitrary (and therefore justifiable) coercive rule. Furthermore, the path of universalization is probably the quickest and shortest route to that summit, although certainly less panoramic and enriching than argumentation ethics. Moreover, like the latter, it is traced exclusively on the solid rock of logical-aprioristic reasoning, without ever referring to ‘human nature’.
In the following, I will divide my reasoning into three parts.
In the first part, I will briefly clarify the concept of ‘coercive rule’ as it will be used in this article, along with some of its implications.
In the second part, I will explain why the argument put forth by Hoppe, claiming that the principle of universalizability is only a necessary condition for the non-arbitrariness of a norm and not a sufficient one, is incorrect.
In the third part, I will demonstrate that universalizability is not only a necessary condition for the logical justifiability of a coercive rule but also a sufficient one, and that the only coercive rule that meets this criterion is the non-aggression principle.
Finally, I will provide brief concluding remarks.
1. COERCIVE RULES
By the term ‘coercive rule’ here, I refer to a rule whose violation justifies resorting to physical coercion: not in the sense that in this case physical coercion would be ‘just’, but in the sense that, in a particular juridical context, the violation of this rule legally allows resorting to physical coercion. Whether the coercion legally allowed by the rule is just or not is an entirely different matter which depends on the logical consistency of the underlying theory of justice.
A coercive rule can be positive (“Alice must do X”) or negative (“Alice must not do Y”). Logically, there is no third possible way.
It is important to stress that the problem of whether a rule is just or not, and that of the criteria that have to be satisfied for it to be just, exist only for coercive rules. If a rule is not a coercive rule, then a logical test of justice does not make sense in relation to it3.
Finally, coercive rules can hardly be avoided: even the rule “there must be no coercive rules” is itself (if enforceable: i.e. if its violation allows physical coercion) a coercive rule. This said, whether the use of force in the defense of coercive rules can be avoided, under what conditions and to what extent, is a different subject that will not be discussed here.
2. HOPPE’S MISTAKE ON UNIVERSALIZABILITY
In order to demonstrate that the test of universalization is not a sufficient condition for a norm to be justifiable, Hoppe resorts to some examples of rules that he claims pass this test but are still arbitrary and therefore unjustifiable.
The first two examples are the following rules: “everybody must get drunk on Sundays or else he will be fined” and “anyone who drinks any alcohol will be punished”. These two opposing rules “are both rules that do not allow discrimination among groups of people and thus could both claim to satisfy the condition of universalization” (Hoppe 1988, 64). However, this statement appears to be incorrect.
The first of the two rules mentioned by Hoppe is a positive coercive rule. At first glance, it seems like a rule that passes the test of universalization. However, upon closer examination, it is clear that it fails the test. In fact, it presupposes that there is someone who has the right to perform an action that other people do not have the right to perform: specifically, to initiate the violation of others’s property (in this case, by imposing a fine4). Indeed, as we will see further, this is the specific reason why, in general, all positive coercive rules necessarily fail the test of universalization.
It should be emphasized that here we are not yet concerned with the question whether the initial violation of someone’s property is just or not (we still haven’t defined what ‘justice’ is). We are merely observing the objective fact that this particular rule logically implies that someone has the legal right to initially violate someone else’s property and that others don’t have this legal right; and therefore that, in this sense, the principle of universalization is violated in this instance.
The second rule is a negative coercive rule. Again, the use of the term “anyone” makes it appear that it passes the test of universalization. However, this rule fails the test for the same reason the previous rule failed it. It presupposes that someone has the legal right to perform a specific action that others do not have the right to perform: once again, to initiate the violation of others’s property (in this case, by imposing a punishment).
The reason why both of these rules fail the test of universalization is implicitly provided by Hoppe himself (which is why it is surprising that, in the specific case of the two rules mentioned above, he arrives at a different conclusion): “all of the actually practiced alternatives to libertarianism and most of the theoretically proposed non-libertarian ethics would not even pass the first formal universalization test, and would fail for this fact alone! . . . Such rules would not be acceptable because they would imply that one group is awarded legal privileges at the expense of complementary discriminations against another group” (Hoppe 1988, 64). The organization that awarded itself the legal right to award legal privileges to others (the state), is a group like all the others. The fact that this group is made by one single entity is irrelevant for the matter at hand.
In order to reinforce the argument that the test of universalization is not a sufficient condition for a rule to be justifiable, Hoppe (1988, 64) provides a third example: “everybody has the right to uninvitedly aggress against other people (a rule, by the way, that would formally pass the universalization test!)”. Unlike the previous two rules, this one does actually pass the universalization test. However, it is not a coercive rule. Specifically, it is not a positive coercive rule, as in this case it would be “everyone must aggress against other people”. Nor is it a negative coercive rule: in fact, a negative coercive rule establishes what a person (or everyone) should not do, not what they “have the right to do”. Given a negative coercive rule, what Bob “has the right to do” is undefined and can only be deduced by subtraction. For example, let’s consider the coercive rule that prohibits everyone from evading taxes. This rule implies that everyone “has the right” to read books (here I’m assuming that there are no other coercive rules prohibiting this specific action). While, on the one hand, the rule that prohibits Alice from evading taxes is a coercive rule, the fact that she “has the right” to read books, or to do many other things, is not a coercive rule: it is just an implication, or a consequence, of the previous coercive rule or, perhaps more precisely, of the fact that the existing coercive rules do not specifically forbid this action.
Therefore, the rule “everybody has the right to uninvitedly aggress against other people” is not a coercive rule. However, as mentioned above, if a rule is not a coercive rule, then a logical test of justice does not make sense in relation to it.
Hoppe’s claim that universalizability is not a sufficient condition for the justifiability of a norm remains unproven.
3. THE UNIVERSALIZATION TEST IS A SUFFICIENT CONDITION
On the other hand, it can easily be demonstrated that: (a) the test of universalization is a sufficient condition for a coercive rule to be non-arbitrary and therefore justifiable, and (b) the non-aggression principle is the only coercive rule that can successfully pass this test.
As argued above, in order to demonstrate this, we must apply the test of universalization to coercive rules only.
As we have seen, all positive coercive rules fail the test of universalization. Indeed, they presuppose that someone has the right to perform a specific action that others do not have the right to perform: to initiate the violation others’s property. For example, the positive coercive rule “everyone must serve in the military” fails the test of universalization because it assumes that someone (e.g., the state) has the right to initiate the violation of others’s property (e.g., by imposing fines, imprisonment, death sentences etc., in case of refusal to serve) while others do not have that right.
We have also seen that negative coercive rules can fail the test of universalization for the exact same reason. For example, the negative coercive rule “everyone is prohibited from reading books” again assumes that someone (e.g., the state), unlike anyone else, has the right to initiate the violation of others’s property, e.g. by enforcing a sanction.
However, there is one specific (yet at the same time general) case, and by construction only one, in which a negative coercive rule does pass the test of universalizability. This specific case is when the action that is prohibited is precisely the one that logically triggers the violation of the principle of universalization in all other cases: initiating the violation of others’s property. Unlike the previous negative coercive rule (“it is prohibited for everyone to read books”), the rule “it is prohibited for everyone to initiate the violation of others’s property” does not presuppose that someone has the right to do something that others do not have the right to do.
For example, let’s suppose that Bob initiates the violation of Alice’s property by stealing her book. Based on the negative coercive rule “it is prohibited for everyone to initiate the violation of others’s property” Bob could be sanctioned and consequently experience a violation of his own property (such as a fine), carried out by Alice or by someone (e.g., a company) to whom Alice has voluntarily delegated the defense of her property rights (de Molinari  2009; Rothbard 1978).
Now, since this violation of Bob’s property would not be initial but rather a response to Bob’s initial violation of Alice’s property, it would not constitute a violation of the negative coercive rule “it is prohibited for everyone to initiate the violation of others’s property.”
If we call the initial violation of someone’s property (e.g., through violence, credible threat thereof, and other means5) ‘aggression’, then we can refer to the aforementioned negative coercive rule as the ‘non-aggression principle’.
Thus, it is demonstrated that the principle of universalization is sufficient to establish the non-arbitrariness of a coercive rule. Furthermore, given that all other negative coercive rules and all positive coercive rules (without exception) fail the test of universalization, it is shown that the non-aggression principle is the only non-arbitrary (justifiable) coercive rule that is logically possible.
If we define ‘legitimacy’ as the respect of non-arbitrary coercive rules (i.e. of the non-aggression principle) and ‘legality’ as the respect of arbitrary coercive rules (e.g. positive rules), it follows that the only legitimate property is the one that each person holds over her own body, the one derived from goods that previously belonged to no one and with which the individual has mixed his labor (homesteading), the one resulting from voluntary exchanges and voluntary donations. Any other form of property, no matter how legal, is necessarily illegitimate.
Hoppe’s argumentation ethics is a solid and rigorous path to the summit of the non-aggression principle as the only demonstrably justifiable coercive rule. His assumption that the path of universalizability does not also lead to the same summit is, however, wrong. In other words, his assumption that the universalization test is only a necessary and formal condition for a coercive rule to be non-arbitrary (justifiable), but not a sufficient and substantial one, is incorrect. In fact, it is a sufficient condition as well, and the non-aggression principle is the only coercive rule that can meet this condition. The path of universalization also reaches the summit. This route is carved on the same solid bedrock of logical-aprioristic reasoning. It may be a less scenic and enriching journey, but it is considerably easier and faster.
In addition, its implications (which here will not be discussed) are significant not only on a theoretical level but also on a strategic and communication level. In fact, especially today, most enemies of freedom revere the principle of equality before the law, which is usually one of the first articles of their constitutions. They revere it because they don’t know its logical implications. In particular, they don’t know that it logically implies the non-aggression principle as the only justifiable (non-arbitrary) coercive rule, and therefore the illegitimacy of the state.
Antiseri, Dario, and Giulio Giorello. 2008. Libertà. Milano: Bompiani.
De Molinari, Gustave.  2009. The Production of Security. Auburn: Ludwig von Mises Institute.
Hoppe, Hans-Hermann. 1988. “From the Economics of Laissez gaire to The Ethics of Libertarianism.” In Man, Economy and Liberty. Essays in Honor of Murray N. Rothbard, edited by Walter Block, and Lew H. Rockwell, 56–76. Auburn: Ludwig von Mises Institute.
Kinsella, Anthony. 2011. “Argumentation Ethics and Liberty: A Concise Guide.” Mises Daily Articles. May 27, 2011. https://mises.org/library/argumentation-ethics-and-liberty-concise-guide.
Rothbard, Murray N. 1978. For a New Liberty. The Libertarian Manifesto. New York: Collier Books.
Rothbard, Murray N.  2010. “Beyond Is and Ought.” Mises Daily Articles. August 24, 2010. https://mises.org/library/beyond-and-ought.
Rothbard, Murray N. 1995. Economic Thought Before Adam Smith. An Austrian Perspective on the History of Economic Thought. Volume I. Auburn: Ludwig von Mises Institute.
Rothbard, Murray N. 1998. The Ethics of Liberty. New York: New York University Press.
Spooner, Lysander. 1870. The Constitution of No Authority. Boston: Published by the Author.
Tucker, Benjamin R. 1897. Instead of a Book. By a Man Too Busy to Write One. A Fragmentary Exposition of Philosophical Anarchism. New York: Benjamin R. Tucker, Publisher.
1The Taoists in ancient China were “the first libertarians, who believed in virtually no interference by the state in economy or society” (Rothbard 1995, 23). Lao Tzu (late sixth century or early fifth century BC), in particular, “worked out the view that the individual and his happiness was the key unit of society. If social institutions hampered the individual’s flowering and his happiness, then those institutionsshould be reduced or abolished altogether” (Ibid.). Chuang Tzu (369-c.286 BC), likely the most gifted among Lao Tzu’s followers, was the first person to advocate for individualist anarchism. Rothbard (1995, 24) describes him as “the first anarchist in the history of human thought.” Fast forward to the nineteenth-century United States, where we find Lysander Spooner (1808-1887) and Benjamin Tucker (1854-1939). Both regarded the state as a criminal organization based on the legalization of aggression (Spooner 1870, 13-14; Tucker 1897, 13-14).
2Problematic as it was, Rothbard’s approach aimed at demostrating the existence of the non-aggression principle as the only justifiable moral rule. Hoppe’s approach, however, logically demonstrates that it is not possible to argue in favor of a different moral rule without being logically inconsistent. The object of the two demonstrations is not the same.
3Positivists and consequentialists tend to believe that justice is in end-states, rather than in rules. Therefore, they apply a (necessarily arbitrary, not logical) test of justice on end-states, rather than on rules, which are mere tools to obtain them. Here, however, I’m concerned with the justice of rules, not of end-states. In other words, I’m concerned with a non-arbitrary theory of justice, not with an arbitrary one.
4The fact that someone can impose fines (or, more in general, violate someone else’s property) while others cannot does not inherently imply a violation of the principle of equality before the law. In other words, the absence of a state does not necessarily imply the absence of government. For example, if Alice smokes in Bob’s cinema, thereby violating the rule set by Bob and agreed upon by Alice, which prohibits smoking in the theater and imposes sanctions on smokers, the fact that Bob, unlike another customer, can enforce the payment of the fine does not constitute a violation of the principle of equality before the law. In fact, the cinema is Bob’s property. Nothing prevents Alice from constructing and managing her own cinema with different rules regarding smoking, which she will have the exclusive right to enforce within her property rights and which others will always have the option to accept or not. However, I am considering a different scenario here: one in which a particular organization (e.g., the state) unilaterally assumes the right to initiate the violation of someone’s property, such as through taxation or fines, without this person having violated anyone’s property rights.
5These may include, for example, fraud, breach of free contracts, failure to reimburse in cases of proven fault, and others.