Conflicts of Liberty and Equality
1. INTRODUCTION
The aim of this paper is to show how there can be conflicts between liberty and equality as political values. It might be thought that this undertaking was either unnecessary or impossible. On one account of liberty (or, to use the word more usual at this point of the argument, freedom) there are quite obviously conflicts between freedom and equality: this is the account by which (very roughly indeed) I am free to do X if I am able to do X (if I have the “capability” to do it). Any form of co-existence restricts freedom in this sense; so, in further ways, does co-existence under government; in yet further ways, so does co-existence under government dedicated to policies of equality.
However, it is not clear that freedom in this sense is yet any form of political value. It is unclear, for instance, what of value has been lost by a restriction of capability as such. Something has been lost by the agent whose activities have been restricted, but that fact in itself does not give him any claim on society’s concern: “It stops me doing something I want (or might want) to do” is not yet, in itself, a political argument. It has the shape of an argument, it might be thought, only if it means “It stops me doing something I have a right to do.” But if this is the necessary form of a complaint in liberty, then conflicts between liberty and equality (where equality, also, is understood as implying rights) appear to be impossible; in aiming at coherent interpretations of liberty and equality as implying rights, we must look for accounts under which they do not imply conflicting rights. There has been a long tradition of argument in this direction. Rousseau supported one version of it, and Ronald Dworkin has recently offered another.1 On these accounts, the aim of this paper is impossible.
I shall try to find a path—or rather, perhaps, a set of connected spaces—between these two positions which will enable us to accept that there are genuine conflicts between liberty and equality, while giving an account of liberty as a political value which does not make this consequence trivial. This might be thought to agree with Dworkin in his premiss, that liberty should be understood as “a distinct and compelling political ideal” (PL 6), but not in his conclusion from it, that the favoured interpretations of liberty and equality should be such that these values cannot conflict.2 However, this does not locate our disagreement exactly, because I do not agree with Dworkin on what counts as understanding liberty as a political ideal or value. Indeed, part of my case will be that his approach does not adequately acknowledge the political character of this value.
2. FLAT FREEDOM AND CAPABILITY
I introduced the minimum sense of freedom, under which freedom inevitably conflicts with equality (and with many other things) in terms of capability, or what has sometimes been called “power.” However, this is not how Dworkin draws his contrast between a sense of “liberty” in which it does not yet represent a value, and the sense in which it does. “We use ‘liberty’ in its flat sense,” he says (PL 5), “simply to indicate the absence of constraint.” He says, further, that this is a descriptive sense (cf. PL 12), whereas the other sense is normative: in the normative sense, liberty is something like “rightful freedom.”
I agree with Dworkin in identifying the flat or minimum sense of freedom in terms of the absence of constraint, but not with the claim that such an account is not normative. It seems that it must be to some extent normative because what counts as “constraint” is, beyond a certain point, a normative matter. (Some questions about constraint will be of concern later.) I do not think that it matters that flat freedom is a partly normative concept. Indeed, both Dworkin and I want flat freedom not yet to be a political value, and my own strategy will be to build up an account of liberty as a political value by starting from flat freedom; but neither this strategy nor Dworkin’s depends on supposing that flat freedom is in no way normative.
The notion of freedom as mere capability does seem untouched by that source of normativity, at any rate, since it does not import the notion of constraint. However, I do not think that it is free from normativity. Any useful notion of capability is surely going to have to construe the idea of “things I cannot do” in such a way as to include (at least) things that I cannot do except at some quite unacceptable cost, and this involves a normative notion.
In fact, I do not believe that capability is a plausible candidate for a notion of freedom at all, unless it is understood in a way that has quite elaborate normative implications. This is not the place to develop the point, but very roughly the idea is that a notion of freedom is useful only if we can deploy “more” and “less,” particularly in the context of extending or restricting someone’s freedom.3 A main reason why freedom as capability has seemed attractive to political theory is that it allows us to say that extensive formal freedoms can co-exist with very restricted capability freedoms: an increase in resources to poor people can, under this interpretation, itself be understood as offering increased freedom. However, we can say this only if we can assess greater, less, increased, and so forth, capabilities, and without heavy normative help there is no way of doing this, because there is no non-arbitrary way of identifying a capability or an increase in capabilities.4 It is only in terms of what counts as a significant or worthwhile human capability that we can make these kinds of arguments.
It may be that there is a helpful sense of freedom to be expressed in these terms, but if there is, it is already a richly conceived ethical and political value. It is certainly no candidate for flat or minimum freedom. The only candidate for that, I think, is the traditional one that Dworkin selects, absence of constraint.
3. THE PRIMITIVE SITUATION AND ITS EXTENSIONS
The most elementary model of not having freedom (losing it, having it restricted, etc.) seems to be being in someone else’s power. The most elementary form of this, again, is that of being in a situation in which someone else (call him “A”) has the intention that I should act in certain ways and forces me to do so “against my will”—that is (in the simplest case), where I would want not to act like that if not forced. If we want to extend and better understand this primitive situation (PS), we have to consider four elements of it: (a) the intentionality of A; (b) the identity of A; (c) what counts as “forces”; (d) (relatedly) what counts as “against my will.”
(a) immediately raises the prospect of two ways in which the PS may be extended. In the PS, A wants precisely a certain thing to happen and adopts means to extract it from, specifically, me. But, first, A may want some general result and may apply pressure to me only to the extent that this is necessary or advantageous to his getting that result. Alternatively, he may have no intention directed toward forcing me to do anything. A restriction on my power may follow only as a side effect of his pursuing something else. In terms of the PS, this is a restriction on my power and not on my freedom. But it might come to be understood as a restriction on my freedom if certain other conditions were met.
One is that, although A has no relevant intentionality, and his professed purposes are such that the disadvantage imposed on me and others is simply a side effect, there might be an account of the situation in which those “side effects” were understood as functional. In such a case it might be said that while it was not an aim of A’s, it was an aim of the system, that my powers should be restricted. Such accounts are notoriously hard to make good, and I shall not pursue the issue here. The point is simply to locate this kind of problem on the map of liberty. If an account in some such terms were made good, it is fairly clear how the PS account of freedom would be extended naturally to cover such a situation.
Another possibility, of much wider significance, is that it may depend on certain normative conditions on my and A’s activities, whether what he is doing to me counts as a limitation of my freedom or not. In one direction, a restriction on my activities which is a side effect of A’s activities may be more readily identified as an affront to freedom if I for some special reason have a right to do that thing. Conversely, interference that would unqualifiedly be an offence to freedom if it had no special reason may not count as such if it has certain kinds of legitimation. In order to mark such points, we require normative supplements to the PS account. We shall have to consider at more than one point various normative supplements, and to ask about their usefulness in building up notions of freedom more complex (and more political) than the basic one displayed in the PS. I shall mark various points as ones in which the question of such a supplement will come up. This is the first: QNS 1 (activity).
(b) The agency may become more collective or, again, abstract: Dick Turpin, Don Corleone, the Mafia, the king, the monarchy, the party in power, the government, the law. This variation itself may be taken to have normative implications for ascriptions of freedom. To live under the rule of law rather than that of men is often said to be a paradigm of political freedom.5 This is QNS 2 (agency).
(c) What counts as forcing? This is a famously difficult, and in general terms perhaps insoluble, question (cf. Nozick on coercion). One point is particularly relevant here. When an agency is allowed to try to bring about results by forcing, a question will arise whether a given method of forcing is allowable: at all, or in relation to particular objectives. It is a further question whether an objection to particular means could naturally take the form of saying that these means, though not some others, make the attempt to force into a violation of freedom. It seems that it sometimes may: if so, this is QNS 3 (means).
(d) Does “against my will” mean against my actual desires? My actual and operative desires? Against the desires I would have if better informed?Against those I would have if some yet more elaborate condition were satisfied? It seems to me that both intellectual and political history suggest that only bad news follows from trying to do much of the normative work at this point of the construction. I suggest that “against the balance of my actual desires” is a sufficient condition of “against my will.”6 This is so even if my actual desires are dependent on, for example, gross and easily corrigible ignorance. Such a case might justify peremptory, even violent, paternalistic action in my interests, and it does seem a paradox to deny that that is a temporary incursion into my freedom, though it is one that may be easily justified.
If “against my will” roughly means “against the balance of my actual desires,” then there can be offences against my freedom which are not “against my will,” because there are, manifestly, ways of offending against freedom by manipulating people’s desires.7 There is a further normative question here, about what people might naturally be expected to want, but I shall not pursue it.
4. NORMATIVE SUPPLEMENTS
It may be said that it is misleading to use the phrase “normative supplement” to what was offered in the PS, since even in the PS there are normative assumptions, at least to the effect that A and I are not in legitimate competition, or that my activities do not constitute a deliberate and sustained threat to A. I think we can avoid controversy here. Take “supplement” to mean whatever it is thought appropriate to add to whatever are the minimal normative assumptions appropriate to the identification of a limitation of freedom in the PS.
The question now is, how the normative supplements should be set to yield, from the concept of freedom illustrated by the PS, a concept which may be said to express freedom as a political value. This, political, concept—whatever exactly it will turn out to be—we can mark with the original political expression, “liberty.” Just as we saw at the beginning that if we took minimum freedom as capability, it would not yet in itself constitute a political value, so equally the freedom displayed in the PS is not in itself such a value. The mere fact that I am being required to do something against my will by a government is a limitation of freedom at the level of the PS account, but this cannot, in itself, count as a violation of liberty, if there is to be a concept of liberty at all. At the very least, liberty, a political value, has to be able to co-exist with the political.
To mark this point, let us say that a certain claim of a loss of liberty is (minimally) socially presentable, if it can be urged consistently with accepting a legitimate political order for the general regulation of the society. 8 (It follows that the anarchist objection is not a socially presentable claim to loss of liberty.) An objection merely to the fact that I am prevented by the police from removing my neighbour’s property is not a socially presentable claim. An objection to the operations of Franco or James II was a socially presentable claim: one could, and most objectors did, accept that these rulers should be replaced by some other rulers, and more generally they accepted a state system.
Being socially presentable of course does not entail that the claim is correct, in the sense that the activity complained of should be stopped or discouraged in the name of liberty. Nor does it follow, necessarily, that it is correctly claimed that there is a loss of liberty. It is merely that these claims can be coherently discussed within the assumption that there can be legitimate regulation of the society by a political order. We still have to consider how much has to be added, besides social presentability, to yield a concept of freedom which is liberty.
5. ROUSSEAU AND DWORKIN
Rousseau believed that there were no socially presentable claims against the state in a just society. This was because, for him, the only legitimatepolitical order was just. To get these results, he operated with some strong assumptions in the areas of the various normative supplements of activity, agent, and means:
(1) anything a just state stopped me doing I had no right to do anyway;
(2) a legitimate state is simply an expression of the General Will, which is my own, and this means that its coercive activities are not even against my will;9
(3) the question of means (more or less) falls away, since they are being applied ex hypothesi to traitors or outlaws.
Dworkin does not agree with Rousseau that the only legitimate order is just, and hence is not committed by Rousseau’s reasons to thinking that there are no socially presentable claims in a just society. He does, however, more or less, accept that there are no correct claims in liberty against the state in a just society. This is because a just society is one that is properly equal, and liberty and equality are defined in an interrelated way.
Dworkin’s view, too, relies on some strong normative supplements. In the matter of (3) (means), he holds of course, contrary to Rousseau, that the matter of what means the state employs even toward defaulters is itself important both for liberty and for equality. His views about area (2), the identity of the state as agent, equally embody liberal assumptions, about the need for impersonal agencies, the rule of law, and the like, if interventions are not to count as violations of liberty. Like liberal opinion in general, he will also think, unlike Rousseau, that it is possible to be on the losing side in a just society without being alienated from the body politic.
So much is common ground among liberals. Dworkin’s distinctive conclusion that there can be no correct claims in liberty in a just society follows from an assumption in the area of (1), the supplement concerning the normative status of the complainant’s activity. This denies that it is, other things being equal, a loss of liberty if I am stopped from doing something that I would not be allowed to do if certain ideal arrangements obtained. On Dworkin’s view, a person who claims that his liberty has been violated, or that he has suffered a loss of liberty, may make this claim entirely sincerely and while thinking, not merely that some political order could be legitimate, but that this very political order is legitimate; and yet this claim may not be correct, because the claimant’s activity does not satisfy the normative demands in area (1). His activity is not rightful, for instance, under the preferred interpretation of equality.
This is likely, in fact, to be a fairly common phenomenon. Let E(I) be the favoured Dworkin interpretation of equality. In political fact, if there is a properly elected government engaged in applying E(I) in social reform, there will be many who reject the interpretation E(I), while there may be others who are, for instance, suspicious of the political mandate for some particular proposals (though not doubtful that they have duly become law). All this is consistent, of course, with its being, and their thinking that it is, a legitimate political order. Indeed, they may well share a lot more than this with those who, against these people’s wishes, are enacting E(I).
6. REASONABLE RESENTMENT
This very familiar type of situation should encourage us to look again at how we should set the normative supplement in area (1). Let us say that a claim to the effect that one’s liberty is being infringed is responsible if the claim-maker makes it sincerely, is convinced that the political order is legitimate, and moreover remains persuaded of his claim despite attending to serious argument, and so forth. The maker of a responsible claim will typically feel resentful at his activities’ being curtailed. The suggestion I want to make is essentially that we should understand the normative supplement in area (1) in such a way that a responsible claim to a loss of liberty means that there is a loss of liberty.
Even if a person in this situation will feel resentful, it may be asked whether he rightly feels resentful. The force of this question reveals why Dworkin’s stronger normative supplement in this area (if the government’s activity is in fact rightful, there is no complaint in liberty) looks tempting. One way of setting out the temptation is by way of the suggestion that the following are (roughly) equivalent:
(i) A ought to be allowed to do X (e.g., retain certain resources which the government proposes to take from him);
(ii) A ought not to be prevented from doing X;
(iii) it is an affront to A’s liberty to prevent his doing X.
Moreover, (iii) may be thought to imply
(iv) A would be rightly resentful about being prevented from doing X.
Suppose policies directed to increasing equality under the interpretation E(I) imply that A shall not do X (e.g., retain the resources). Grant that we accept E(I). Then, by these supposed equivalences, we get denials of (i) and (ii); denying (ii), it seems plausible to deny (iv); if we deny (iv), we deny (iii). This is enough to motivate Dworkin’s strong normative supplement.
In terms of this structure, I think that we should resist the middle step, to the effect that if we deny (ii), we must deny (iv); or alternatively, if we wish to reserve “rightly resentful” for this purpose, then we must admit that you can be reasonably resentful when you are not rightly so (and not through the ignorance of any facts). You may be reasonably resentful about a move against your activities even if you think it ought to be accepted, not stopped, and so forth: in particular, you may think this if you accept the system that generated the result, but think that the system ought not to have generated it. (Cf. the so-called Paradox of Democracy.)
I suggest that it is sensible to connect the account of liberty, and of what counts as a violation of it, with what people may reasonably resent in this sense. People’s sense of freedom is given to them through experiences such as those described in the PS, and it is through them that its value, too, is grasped. It is a requirement on them that they should move beyond the limits of the PS, and not complain of arrangements without which the society of which they are members could not exist or would not provide a legitimate order. Normative supplements related to that are entirely appropriate to a conception of liberty. They might, for various reasons, be taken further; a sense of shared citizenship might support a conception of liberty that was more heavily normative still. But it is still true that there will be people who do not subscribe to particular political and ethical programmes, such as the extension of equality interpreted in terms of E(I), and if they do not, they may be reasonably resentful of the effects of such policies on them.
It is offensive to these people, in a way rather reminiscent of Rousseau, to suggest that they are as much in error in resenting and complaining, in the name of liberty, about the restrictions on their activities that are imposed in the name of equality, as people are certainly in error who complain, in the name of liberty, of restrictions simply because they are made by the state. It is better to restrain the demands of the normative supplements in area (1) and accept that if a restriction on activities is reasonably resented in the kind of circumstances described, then it is a restriction on liberty. This is a conclusion that can be accepted by everyone, including those who favour the interpretation E(I) and who favoured its being enacted. They will see the loss of liberty of these people as a price that has to be paid for the extension of equality. They will have accepted, that is to say, that liberty and equality can conflict.
The relevant conditions can helpfully be put in terms of “reasonable resentment,” under the explanations I have sketched, though the fact of resentment is not itself the ground of the political judgement in such cases:it is, rather, a helpful reminder or indicator of what values are involved. But resentment can occur with regard to both liberty and equality. Equally, someone may be resentful who feels that his claims in equality are improperly being neglected. If we examine this more closely, we shall see why the interpretation of liberty that is being suggested, and an interpretation of equality such as Dworkin’s, stand in different relations to politics.
7. LIBERTY, EQUALITY, AND POLITICS
Consider two copybook cases. A has more than B
(i) because he has stolen it from him;
(ii) because he drew the lucky straw in a randomized allocation to which they both uncoercedly agreed.
Presumably everyone thinks that B may have reasonable resentment in (i) but not (ii); and that A would not have reasonable resentment in (i) if there were an intervention to change the situation, but that his resentment in (ii) would be reasonable. Resentment from a disadvantaged party is more reasonable in circumstances that approximate to (i), circumstances that call for rectification.
How do the hypothetical scenarios that Dworkin uses in giving his interpretation of equality, E(I), stand in relation to this? In their light, the actual social situation lies between (i) and (ii): nearer (i), since it demands rectification, but not so near that an item or sum that belongs to a given A should be uniquely reallocated to a given B. So what rights does B have under such an interpretation? As has just been said, he does not have
(a) a right to what A distinctly has.
This comes out also in the consideration that a jacquerie to secure it would (except in unusual circumstances) be criticized for more than, so to speak, impatience. Even in (i), B would be open to criticism, if there is effective legal redress, for taking the law into his own hands; in this case, it is not even the law he would be taking. B certainly does have, on the other hand,
(b) a right to whatever redistributed resources he will get when redistribution is enacted;
but this is uncontroversial. He has this right under E(I), so it looks as though B has
(c) a right to what he would get under the application of E(I).
So is it then true that he has
(d) a right that (the application of) E(I) be enacted?
(d) is quite hard to defend. Certainly it is not like (b), which appeals to what everyone would concede of a well-ordered and legitimate state, and would be accepted both by those who accept and those who reject E(I).But (c), and hence (if at all) (d), are accepted only by those who accept E(I); and (d) commits them to thinking not only that there is good reason for others to agree with them, but that the disadvantaged have the right to demand that they agree with them. This is very strong; it is in fact another Rousseau-like tack, to suppose that all the urgency and dignity of justice applies to one’s own political interpretation of justice.
Faced with this, there are two opposite temptations, both well-known from history. One is to weaken (c) and (d) into something like
(e) a right to a political system in which E(I) and its rivals are fairly discussed, and so forth.
This (it might be called, Popperian) line is politically conservative, and it also leaves a mystery about what is said about E(I) and its rivals in the course of the discussions it recommends: can a holder of E(I) not express him/herself by talking in terms such as (c)? The alternative (Saint-Just) line is to accept (c) and (d) and suspend the courtesies of the liberal democratic state. This is objectionable and, for Dworkin, incoherent, since among the things people have a right to under E(I) are the courtesies of the democratic state. So the supporter of E(I) has to accept, it seems, a state of mind which rejects (d) while accepting democracy and (c). This is not necessarily confused or inconsistent; rather, it is double-minded in a way that is perhaps necessary in pluralistic free societies.
Even if we accept (d), we need some double-mindedness in seeing E(I) in relation to other interpretations. For the holder of E(I), what is wrong with the situation of the disadvantaged is what is set out by E(I); it does not depend on the political outlook of those who reject E(I). But our attitude to E(I)’s not being enacted, and hence the use we make of (d), do depend on what those outlooks are. It makes some difference, for instance, whether those who vote against the policies flowing from E(I) do so because they have some other responsible interpretation of equality.
This kind of consideration helps to motivate constructing liberty in the way I am suggesting. The resentment of B with regard to the substance of (c), if E(I) is not enacted, co-exists with a sense (if not in B himself, in his political advocates) that others need to be persuaded, have other views, and so forth; just as the resentment of the person who does not accept E(I) but is being required to pay co-exists, if E(I) is enacted, with the acceptance that it has been enacted. There is no incoherence in this—merely the containment within the law of, and a shared political system of, conflicting interests, passions, and interpretations. But it is not merely that for each interpretation of equality there will be a corresponding interpretation of liberty, such that each interpretation adjusts its liberty and its equality to each other without loss. Rather, the proposed interpretationof liberty is what we need in order to live in society with others who have different interpretations of equality.
Even though you and I share a certain conception of equality, and are happy to see its policies being enacted, we can and should use a political concept of liberty in terms of which we can not only sympathize but agree with our fellow citizen who does not share this conception of equality, resents what is being done to him in its name, and says that he has lost some of his liberty. He is reasonably resentful of what is happening to him, because he is being coerced against his will; and these are the complaints of someone who accepts, not only some political order, but this one.10 Can we really tell him that if he only understood liberty better, he would see that he was deluded in thinking that he had lost some?
This is a thoroughly political concept of liberty, because it acknowledges in its construction the on-going existence of political conflict. Dworkin’s picture, on the other hand, assimilates these questions to the interpretation of terms in a constitutional text. But governments, parties, political actors, and complainants are not justices or advocates contributing their various inputs to the unfolding interpretation of what they all agree to be a unitary text. If a given conception of equality is politically dominant at a given time, this is because of a certain conjunction of political forces, aspirations, and opportunities; other conceptions continue to move around society in their variously resentful or hopeful ways.
For these various forces and passions to co-exist in some semblance of a stable political order under democratic forms requires a good deal of what I called “double-mindedness.” We need it, in part, because the on-going political framework that contains all this conflict is not given to us, as, for instance, the institutional protocols of the Supreme Court supply an on-going framework for its decisions. We have constantly to reinvent the political framework—in part, through our attitudes to our fellow citizens. It is a contribution to this process that we bear in mind that when their activities are restricted in the name of objectives which they seriously do not accept, they are indeed being coerced against their will, and that when they describe that as a loss of liberty, we should not simply tell them that they are wrong. Our relation to them is not that of offering them instruction in reading a document which we believe we can read better than they can. It is that of sharing a society with them under some degree of liberty, and an expression of that is our sharing with them a concept of liberty which allows us to say that there has been a cost when (at the least) what we believe is right has to be imposed against the will of people who do not think it is right and who are adversely affected by it. We should agree that liberty is normatively richer than primitive freedom, but at the same time not forget that there are many cases in which the fact that people are coerced against their will represents a loss, even though the coercion is done in the name of right.
1 “What Is Equality? Part 3: The Place of Liberty,” Iowa Law Review 73 (1987). Henceforth “PL.”
2 Dworkin holds in fact that the consequence may break down under various extreme or primitive circumstances; he has room for what Rawls calls “imperfect compliance theory.” I shall ignore the exceptional cases here.
3 The so-called freedom of the will would presumably be an exception: one that proves the rule.
4 If I introduce a new commodity, do I increase your capabilities by making it possible for the first time for you to choose this one? Or decrease them, by making it impossible for you to make an informed choice without assessing this one? Or both? Again, theorists might think of a capability as the possibility of one’s making a sentence true. Then if you can make it true that Q, and I now for the first time make it true that P, then I have given you a new capability, that of making Q true in conjunction with P. And so on. I have argued that there are related difficulties for Amartya Sen’s construction of equality on a space of capabilities, in a review of his Inequality Re-examined, London Review of Books, November 18, 1993.
5 General Galtieri to a political prisoner in jaiD:\ebook\pdf\output “Your name is Rosaria? It is my daughter’s name. You may live.”
6 In considering this proposal, one must bear in mind that “against my will” may not be a sufficient condition of a limitation, violation, or the like, of freedom in some more complex or political sense: cf. QNS 1 and 2. Rousseau, of course, thought that it was always a sufficient condition; hence he concluded that legitimate political authority could never go against my will (really).
7 This is related to a well-known paradox in J. S. Mill’s view that the measure of freedom is the ratio between one’s actual desires and one’s ability to satisfy them: the reformer who gives the apathetic slaves for the first time a desire for freedom makes them less free.
8 “The” rather than “a” society: a degree of relativization to historical circumstances is acceptable.
9 I ignore the point, important in Rousseau, of the distinction between executive and legislative activities.
10 The case of someone who satisfies all these conditions is offered as a sufficient condition of a claim to loss of liberty that we should accept while agreeing that what is being done to him is rightly done. It represents a high setting of the normative supplement about activity. I think that settings lower than this might well be appropriate—basically because I doubt that the question whether we can identify with someone’s reasonable resentment depends very directly on what we see as the virtue of his opinions. But the sufficient condition is enough for now.