Etica & Politica / Ethics & Politics, 2003, 2
Breaking the circle: The definition of individual liberty (1)
It goes without saying that the concept of individual freedom is most vital to the liberal ideal of a free society, because we do not know what is meant by a free society unless the meaning of individual freedom is sufficiently clear. Hence it does not come as a surprise that classical liberals distinguished negative freedom and positive freedom from the beginning. We find the negative concept of liberty already in a terse remark by Thomas Hobbes. “A free man is he that ... is not hindered to do what he hath the will to do.” (2)
Antony Flew has commented on this: “It is a pity that he apparently failed to notice both that, always supposing that people would in fact intervene to constrain or coerce someone, then that person must remain in that particular respect unfree; and that this is the case regardless of whether or not „he hath the will to do“ whatever it may be.” (3)
Flew brings to our attention that Hobbes’s definition is obviously too narrow for it excludes two kinds of situations which surely go for coercion. The first one is a situation in which a person is not actually hindered but would be coerced if he acted in a particular way, and the second one is a situation in which a person is hindered (actually or potentially) to do what he has not the will to do but could do if he had the will to do it. In other words, Hobbes definition is not convincing. Hence, it does not meet the clever remark by Bruno Leoni on the definition of freedom.
“Everybody can define what he thinks freedom to be, but as soon as he wants us to accept his formulation as our own, he has to produce some truly convincing argument.” (4)
Nowadays, most thinkers, liberal and others, share the conviction that individual liberty presupposes the absence of coercion. They also consent that unintentional restraint or intervention does not count as coercion. “Coercion implies the deliberate interference of other human beings within the area in which I could otherwise act. ... This is what the classical English political philosophers meant when they used this word.” (5) However clarifying this is, Berlin’s explication is incomplete. (6) It does not say what is meant by deliberate interference or by coercion. As Hayek noticed, it is the concept of coercion on which thinkers most widely differ. (7)
Friedrich A. Hayek, in his Constitution of Liberty, undertakes an exploration of what is meant by coercion, acknowledging that a clear concept of liberty presupposes a clear concept of coercion. “Our definition of liberty depends upon the meaning of the concept of coercion, and it will not be precise until we have similarly defined that term.” (8) Furthermore, Hayek pointed out that freedom “refers solely to a relation of men to other men, and the only infringement on it is coercion by men. This means, in particular, that the range of physical possibilities from which a person can choose at a given moment has no direct relevance to freedom. The rock climber on a difficult pitch who sees only one way out to save his life is unquestionably free, though we would hardly say he has any choice.” (9)
2.1 Hayek’s confusion of freedom and power
Apparently, Hayek was very much aware of the distinction between freedom and power. Nonetheless, he himself confused freedom and power on several occasions, obviously mislead by a residue of consequentialist ethics. He wrote: “So long as the services of a particular person are not crucial to my existence or the preservation of what I most value, the conditions he exacts for rendering these services cannot properly be called ‘coercion’. A monopolist could exercise true coercion, however, if he were, say, the owner of a spring in an oasis. Let us say that other persons settled there on the assumption that water would always be available at a reasonable price and then found, perhaps because a second spring dried up, that they had no choice but to do whatever the owner of the spring demanded of them if they were to survive: here would be a clear case of coercion.“ (10) And Hayek goes on, “So long as the services of a particular person are not crucial to my existence or the preservation of what I most value, the conditions he exacts for rendering these services cannot properly be called ‘coercion’”. (11)
What is “crucial to my existence or the preservation of what I value most”? What is an “unreasonable price”? Hayek gives no answer. Murray Rothbard rightly remarked that Hayek failed attempting “to distinguish, merely quantitatively, between ‘mild’ and ‘more severe’ forms of coercion”. (12) A monopolist has the right under freedom to refuse any exchange he likes to refuse, however disastrous the consequences may be. (13) Mitigating the consequences of those who suffer under the economic power of others is honourable. Doing it by confusing the terms is intellectually inappropriate.
Even more entangling than Hayek’s confusion of freedom with power is his definition of coercion. “By ‘coercion’ we mean such control of the environment or circumstances of a person by an other that, in order to avoid greater evil, he is forced to act not according to a coherent plan of his own but to serve the ends of an other.” (14) This definition creates confusions in at least two respects.
2.2 Coercion people like
Hayek’s definition neglects cases in which coercion that makes a person serving another person’s ends goes along with the coherent plan of that person. An old movie, starring Henry Fonda as lieutenant and John Wayne as recruit, illustrates such cases in which men are coerced - but coherently with their own plan. The troop reaches an old abandoned fortress. Some soldiers find a left cask filled with whiskey. Asked what to do, the lieutenant gives order to annihilate the whiskey. It goes without saying that the soldiers were happy to follow that order. It also goes without saying that the soldiers had clear order to do so and, hence, were coerced to annihilate the whiskey.
It is unclear which part of the environment and of the circumstances has to be changed by another person in order to produce possible coercion. Hayek is more precise on this elsewhere. “Freedom thus presupposes that the individual has some assured private sphere, that there is some set of circumstances in his environment with which others cannot interfere”. (15)
2.3 Private sphere vs. protected sphere
However, to Hayek, not only that strictly private sphere should be protected from coercion, “reasonable expectations” should be protected too. “In determining where the boundaries of the protected sphere ought to be drawn, the important question is whether the actions of other people that we wish to see prevented would actually interfere with the reasonable expectations of the protected person.” (16)
But which expectations are reasonable in Hayek’s view? He gives two examples. “The enforcement of religious conformity, for instance, was a legitimate object of government when people believed in the collective responsibility of the community toward some deity and it was thought that the sins of any member would be visited upon all.” (17) Moreover, Hayek follows Bertrand Russell’s opinion on the treatment of homosexuality. Russell wrote: „If it were still believed, as it once was, that the toleration of such behaviour would expose the community to the fate of Sodom and Gomorrah, the community would have every right to interfere.“ Hayek adds: “But where such factual beliefs do not prevail, private practice among adults, however abhorrent it may be to the majority, is not a proper subject for coercive action for a state whose object is to minimise coercion.” (18)
Hayek seems to ignore or neglect that the private sphere of one person easily can collide with the protected sphere of another person. He does not provide an analytic criterion of reasonable expectations. Who are “the people”, especially if some members of the community do not share the view of the majority, however qualified this may be? Hence the protected sphere becomes arbitrary. Religious freedom and homosexuality do not limit the private sphere of possible and legitimate actions of others, though it might infringe with others “protected sphere”, if these fear religious freedom or homosexuality to influence their interests in a negative way. (19) A related and no less fatal consequence is that the protected spheres of all individuals cannot be upheld in a non-contradictory way, since in a world of contradictory minds the protected sphere of one necessarily overlaps with the protected spheres of others.
Thus, Hayek conception of the private sphere is self-contradictory and, hence, unacceptable. (20)
3. Rothbard’s definition of individual liberty
As the previous chapter indicates, Hayek’s definition of individual liberty includes several shortcomings, some of them noticed and mentioned by Murray Rothbard. Unfortunately, Rothbard’s definition is only in some respects better than Hayek’s. Rohtbard defines individual freedom using an expression by Herbert Spencer: “A man is free when he is not aggressed against.” (21) What Rothbard has in mind is that a person is free as long as his or her self-ownership and private property is not subject to any offence, actual or threatened. He does not make any concessions to what Hayek calls the “protected sphere” or to services which are “crucial” to another person’s existence or to the preservation of what that person “values most”. In this respect, Rothbard is much clearer and more coherent.
However, he does not define what he means by aggression. In a similar way, Rothbard’s closest ally, Hans-Hermann Hoppe, does not define what is meant by physical force when characterising coercion as “the initiation or the threat of physical violence against another person or its legitimately - via original appropriation, production or exchange - acquired property.” (22)
It goes without saying that our daily experience with physical force seems to make an explication of it redundant. However, for the purpose of our definition, we need to be more explicit when it comes to aggression and physical force for another reason. Aggression and physical force may well describe interventions in another person’s self-ownership and/or legitimately acquired private property (privacy, for short). And for both hold that they are not accepted or demanded freely by the person suffering the offense, as it is with coercion. (23) Coercion, whether actual or threatened, is seen as an encroachment to which the coerced person has not consented.
However, the fact of absent free consent (24) raises the greatest difficulty to the definition of individual liberty, because the definition becomes circular if individual freedom is defined as the absence of an intervention (actual or threatened, named aggression, physical force, or coercion) to which the suffering party has not voluntarily consented. In other words, Rothbard’s definition of individual liberty is circular, because it says nothing else but: “A man is free when he is not exposed to an unfree encroachment in his privacy.” (25)
4. Jasay’s definition of individual liberty (26)
Another potential source for a strict definition of individual liberty is the work of Anthony de Jasay. In his approach to strict liberalism, contracts as the source of rights play a central role. It goes without saying that rights deriving from a contract are worth the paper on which they are written only if all parties included voluntarily agreed upon the contract. Forced contracts are based on non-contractual obligations and thus establish no rights. Hence obviously, freedom is a necessary cornerstone in Jasay’s thinking. However, if I am not mistaken, Jasay does not explicitly define freedom, though he stresses, of course, that freedom includes the absence of coercion in a person’s set of feasible acts. The point he is making about coercion is the following:
“Coercion is an intentional act by A, whether actual or threatened, whose effect is to change B’s set of feasible alternatives as to make his chosen alternative different from what A presumed B’s preferred alternative to be. Successful coercion must make B act otherwise than A thought that he had intended to act. It achieves this by intrusion into B’s feasible set. Successful coercion alters the cost of alternatives so as to make B choose as A wishes, and not as A thought B wished.” (27) The point about this definition is that it does not lose its applicability if one replaces the term coercion either by trade or seduction. (28)
At another place, Jasay gives a different definition of coercion that avoids this ambiguity when saying that, “ ‘A coerces B’ means: ‘A has given B reason to believe that if B committed a certain act, or omitted another, A would inflict some sanctions upon him.’ ”. (29) Jasay too, as Rothbard, does not give in, as Hayek does, when it comes to “protected spheres” or “crucial services”. (30) And he generally excludes the withholding of benefits from coercion. (31) Administering a sanction means to harm someone, according to Jasay. Harms, i.e. torts, are inadmissible actions as are breaches of obligations.
As in Rothbard and Hoppe, we meet the circularity problem in Jasay, for sanctions include aggression or physical force upon the privacy of a person to which that person has not freely consented.
As we have seen, Rothbard, Hoppe, and Jasay consent that freedom is the absence of coercion in a person’s privacy or private sphere. Before we go on, let us reflect on what we necessarily assume when we talk of coercion and of a private sphere. Let us start with two very plain remarks. If coercion takes place it cannot but interfere in the sphere of possible actions of another person.
If a person acts he cannot but consume goods. Whether these goods are objects, living beings, or even the person’s body does not matter.
However, with respect to the relation between men and goods, we usually tell private from non-private goods, namely goods from which others are excluded and goods open to joined consumption. Though we could think of a world without non-private goods, we cannot imagine a world without private goods. Our conceptions and our use of speech presuppose at least some goods, e.g. human beings, to be private. Otherwise it would not make sense to say, for instance: “I am playing piano”. If I would not own my body exclusively, I would give a false statement, for it was not me, rather me and others who played piano. Despite from that, goods enable us do things we could not do if they were absent. If Andrew hinders Bruce to use a good, than Bruce lacks an advantage he otherwise had. This is so no matter whether the good is private or non-private.
However, we distinguish these two possible cases for good reasons. Recall the case of Diogenes. Alexander approached him while he laid in the sun. Being asked what Alexander could do for him, he answered: “Step out of the sun”. Alexander hindered him to use a non-private good (sun). Though cases like these can be unpleasant, we do not say that they can constitute coercion. Otherwise we constantly could claim that others coerce us as soon as they make or want to make use of any non-private good. Vice versa, they could claim the same when we opt to use that very non-private good. Hence, mutual exclusive claims of coercion were not only possible. They were both valid. Thus, we had no analytic criterion to decide on the superiority of these claims.
If, in turn, the good was private, then mutual exclusive claims of that kind could not be equally valid. Private goods have an analytic criterion by which the superiority of mutual exclusive claims can be decided, i.e. sovereignty. (32) Of private goods, we can say that they are owned exclusively by one person. Though it seems redundant to note, we should state for the sake of clarity that we assume that the use of own private goods cannot constitute coercion. Otherwise the word “coercion” similarly would become useless. All could claim that others coerce them when these use their own goods, and vice versa. For instance, I could claim Steve to coerce me to walk when he prefers to ride his car alone. He could claim, in turn, that I would coerce him to buy expensive gas, for I use my power exclusively to walk rather than to push his car.
Thus, we may conclude that the term “coercion” presupposes not only the assumption of private goods. It also presupposes that someone is in certain way obstructed by another person or by other people to use one or more of his private goods. Of course, saying that “someone is obstructed by another or by others in a certain way” needs some specification.
It goes without saying that an obstruction constitutes coercion only, if it is done intentionally, and if it sufficiently causes the objection. Saying so, in turn presupposes some other assumptions.
It presupposes that man can act. Otherwise we would not talk of coercion and connote this word with moral assumption, e.g. that coercion is morally bad.
It also presupposes also that the “coercer” as well as the “coerced” can act. If an act, though intended to be coercion, forces another person to react like a physical object, then we do not talk of coercion at all. To keep the definition “freedom is the absence of coercion” free from self-contradictions, we have to conclude the following: Coercion is possible only if it is an intentional act that sufficiently causes an interference in an other’s private sphere. Of course, from what we said the private sphere of a person is nothing else than the sum of all possible ways of consuming the goods owned by that person. Otherwise the private sphere of Andrew could overlap with the private sphere of Bruce. This possibility is excluded by the definition of a private good (good used by one person exclusively).
As we said in the beginning, freedom is the absence of coercion. Taking into consideration the above made explications, we can specify that definition as follows. Freedom is the absence of an intentional interference in the private sphere of another person, sufficient to produce the intended effect. Though this definition is much clearer than that of Hayek, it is certainly not the final word on that subject. With the help of that very definition we cannot distinguish between interventions in a private sphere to which the owner of that sphere agrees and those to which he does not. It goes without saying that only the latter can entail coercion whereas the former cannot. Suppose, a husband asks his wife to take away his cookies for he feels to weak to resist the smell. Doing so, the wife interferes in her husband’s private sphere (for the cookies are his). Her action is sufficient to cause what she intended. However, we would not call her action coercion, simply because the husband wanted her to do what she did. The same analogously holds for Odysseus sailing through the Sirens. (33) To eliminate these cases, we have to specify the definition of freedom a little bit more.
If freedom is the absence of an intentional interference in the private sphere of another person, sufficient to produce the intended effect, than why not simply add: to which that person does not freely agree? The reason for this is quite obvious. The italicised supplement introduces the definiendum (free, freedom) into the definiens and, hence, makes the definition worthless. The hitherto developed definition needs another supplement that avoids that disastrous consequence.
The crucial task is to express the same without smuggling in the definiendum into the definiens. How can this be achieved? How can the useless phrase “to which that person does not freely agree” be substituted by a fruitful one? Certainly, it would be of no help to use instead the formula, “which means no costs to that person”, because a person might freely agree to that another person interferes into his private sphere and creates costs, e.g., a husband agrees to be deprived of his cookies. Hence, what we are looking for is a criterion that distinguishes two types of costs of intervention in private spheres by others; one that involves coercion and one that does not.
In search for that criterion one ought to notice a very simple fact, i.e., that the perception of a new information creates a new decision. As soon as we perceive new information we cannot but decide whether or not to change our plans because of the new data. This happens every day hundreds of times. These situations constitute what we might call “either-or-choices”. Of course, these “either-or-choices” of our daily life are often routinized. Nonetheless, they call for decisions caused by new information. For instance, when we start to cross a street and perceive an approaching car, then we have to decide either to pursue the existing plan or to change it. When we read a sales offer, then we cannot but decide either to react to it or to stay to the status. When we on our way to our home cross the market and being asked by the merchant to buy either apples, oranges, or bananas, then again we cannot other than decide whether to stay to our original plan or to change it, i.e., to “react” to that offer. Of course, this decision we will make will be somehow influenced by the offer itself, e.g. by the price and quality of the fruits. Nonetheless, the decision either to react to the offer or not is not to be confused with the choice between apples, oranges, or bananas. It was even there if there was no choice between apples, oranges, and bananas, for instance, if the merchant spoke to us in a foreign language or too faint to be clearly understood. If it was the case that the merchant conveyed no choice between apples, oranges, or bananas, then we still had to decide whether we deviated from our original plan (i.e., react to the foreign speaking merchant) or stayed to our original intention (i.e., going home unflustered). The first decision precedes the second analytically, although it might coincide with it chronologically.
To distinguish these two types of decision a terminological distinction is useful. It is also useful to start with the description of the second decision of the example mentioned above. The second decision (choice between three different fruits) is characterised by the fact that it constitutes a decision among different objects. I propose to call that second decision an object-decision. The number of objects does not influence the character of an object-decision. With respect to the fact that the content (objects) does not determine the decision as such and in order to distinguish the two types of decisions clearly, I propose to call the first decision meta-decision. The first decision (meta-decision) is a decision between staying to the original plan or deviating from it, hence an “either-or-choice”. Having these two and only these two alternatives is a constitutional character of a meta-decision. The decision is necessarily no other than a decision between “either” “or”, independent of the content of the plan to which one either stays or from which one deviates.
However, what is the distinction between object-decision and meta-decision good for, beside from the fact that it might serve as an analytical insight of decision processes? The decision helps us out of our circular definition. To use that help, let us look, firstly, at a typical case of coercion and, secondly, at that very case transformed into a typical case of free choice.
(A) Suppose, you liked to keep your money and your life. Suppose also, an armed robber asked for your “Money or life!”. That would be a clear case of what we use to call coercion.
(B) Suppose now, that, ceteris paribus, the same person would be unarmed and obviously unable to threat or extort you in any possible way, and suppose, that very person would kindly ask you to give him either your money or your life. Then we would not say that he coerces you. We would classify its saying as a case in which he offers you a free choice. (Although it does not matter analytically, we also would classify his saying as silly.)
Now, let us separate the meta-decision from the object-decision. With regard to the object-decision, the two cases (A and B) do not differ. In both cases you have the choice either to give your money or your life. Hence, in both cases the costs of your decision will be the same.
With respect to the possible meta-decisions, we notice a difference. The costs of a positive meta-decision (i.e., to consider the offer of the person), however huge, are the same in both cases. Nonetheless, the costs of a negative meta-decision (i.e., to ignore the offer of the person) differ decisively. In the first case you have to expect additional costs by the person (e.g., being shot or hurt). In the second case you do not have to expect such costs. Hence, under the above mentioned preconditions the difference between a case of coercion and a case of free choice is in the artificial costs (34) which are to be expected in case of a negative meta-decision.
We now have found a differentia specifica that moves our provisional definition out of the vicious circle. Freedom, we defined provisionally, is the absence of an intentional interference in the private sphere of another person, sufficient to produce the intended effect. To abridge this definition, I propose to speak of “artificial interference” instead of “intentional interference, sufficient to produce the intended effect”. Hence, the short version of our previous provisional definition would be: Freedom is the absence of artificial interference in the private sphere of another person.
Considering our final reflections on non-circular definitions we can resume: Freedom is the absence of artificial interference in the private sphere of another person that would produce artificial costs for that person if she/he opted for a negative meta-decision.
(1) Throughout this text, we use individual liberty and individual freedom synonymously.
(2) Thomas Hobbes, Leviathan, 1651, II (xxi).
(3) Antony Flew, Equality in Liberty and Justice, London: Routledge 1989, p. 5.
(4) Bruno Leoni, Freedom and the Law, expanded 3. edition, reprinted by Liberty Fund, Indianapolis 1991, p. 4.
(5) Isaah Berlin, Two Concepts of Liberty (1958), cited after Liberty, ed. by David Miller, Oxford: OUP 1991, p. 35.
(6) Berlin later said what he had meant. Unfortunately he introduced more vagueness by saying that freedom is "absence of obstacles to possible choices and activities" put by "alterable human practices." (See Isaah Berlin, Four Essays on Liberty, Oxford: OUP 1969, pp. xxxixf.) This comes close to confusing "freedom" with "opportunity". On this see William A. Parent, "Some recent work on the concept of liberty", in: American Philosophical Quarterly, July 1974, pp. 149-153 and Murray Rothbard, The Ethics of Liberty, Atlantic Highlands, N.J.: Humanities Press 1982, pp. 216ff.
(7) See Friedrich A. Hayek, The Constitution of Liberty, Chicago: University of Chicago Press, 1960, p. 19.
(8) Friedrich A. Hayek, The Constitution of Liberty, p. 20.
(9) Ivi, p. 12.
(10) Ivi, p. 136.
(12) Murray Rothbard, The Ethics of Liberty, p. 223.
(13) I abstain from intruding into this argument any deeper. For further criticism see Ronald Hamowys book review of The Constitution of Liberty in: New Individualist Review, 1961, 1,1, pp. 28ff.; and William Meckling, Michael Jensen, “Human rights and the meaning of freedom”, unpublished manuscript 1985, S. 19, and David Miller, “Introduction” in the anthology Liberty, p. 15: “Finally, Hayek appears to put the cat among the pigeons when he concedes that in certain circumstances economic power might be used in a coercive manner. Once the possibility has been conceded, why restrict the circumstances as narrowly as Hayek does, confining them to extreme cases where an individual enjoys a monopoly of a vital resource? Why not admit that the distribution of resources is always going to be relevant to the distribution of negative liberty in a society?”.
(14) Friedrich A. Hayek, The Constitution of Liberty, pp. 20ff.
(15) Ivi, p. 13.
(16) Ivi, p. 145.
(18) Ivi, p. 451, note 18.
(19) Obviously, you cannot have both. Forbidding homosexuality or religious freedom means interfering with other’s privacy. It reduces the possible and legitimate actions of others. Vice versa, respecting their privacy leaves the domain of interests of those who have reasonable expectations - whatever these may be - of the “right” behaviour unprotected.
(20) For an extended version of my critique see my Freiheit, Liberalismus und Wohlfahrtsstaat, Baden-Baden: Nomos 1996, chapter two.
(21) Murray N. Rothbard, For a New Liberty (New York: Macmillan, 1973), p. 8.
(22) Hans-Hermann Hoppe, “F.A. Hayek on Government and Social Evolution: a Critique”, in: Christoph Frei and Robert Nef (eds.), Contending with Hayek, Bern, Berlin: Lang, 1994, p. 131.
(23) Of course, there are many encroachments in a person’s privacy to which this person freely gives its consent. The agreement to an operation might serve as an example.
(24) Of course, consent, by definition, is free. However, we stress this here to illustrate the circularity problem that is decisive for our concern.
(25) In other words, a man is free as long as he freely agreed to all encroachments in his privacy.
(26) It should be noted that Anthony de Jasay sharply distinguishes between freedom and liberty. To him the feasible is free, whereas a liberty is any feasible action that is neither a tort, nor a breach of an obligation, nor the exercise of a right.
(27) See Anthony de Jasay, Before Resorting to Politics, Cheltenham: Elgar 1996, p. 26.
(28) “Trade offer and seduction are intentional acts by A, whether actual or threatened, whose effect is to change B’s set of feasible alternatives as to make his chosen alternative different from what A presumed B’s preferred alternative to be. Successful trade offers and seductions must make B act otherwise than A thought that he had intended to act. It achieves this by intrusion into B’s feasible set. Successful trade offers and seductions alter the cost of alternatives so as to make B choose as A wishes, and not as A thought B wished.”
(29) Anthony de Jasay, Against Politics. On Government, Anarchy and Order, London: Routledge, 1997, p. 163.
(30) Ivi, p. 167.
(31) Ivi, p.163.
(32) See Anthony de Jasay, Choice, Contract, Consent: A Restatement of Liberalism, London: Institute of Economic Affairs, 1991, p. 75.: “... collective ownership defeats the very purpose of property, which is to vest individuals the sovereignty over employment of scarce resources. Sovereignty over certain types of decisions may be delegated revocably, or transferred for good, but it cannot be shared ...”.
(33) As we know from Homer, Odysseus asked his men to tie him at the mast of his ship. With their help he was able to listen to the seductive singing of the Sirens without becoming a victim of their tempting melodies. Of course, although they hindered him to move, they did not coerce him, because he agreed to what they did.
(34) It goes without saying that these costs must be intended and sufficiently caused by that person in question. Therefore, I call them artificial costs, for short.