http://www.econ.vt.edu/tideman/peme5.pdf

This paper argues that a liberal theory of the resolution of disagreements about the requirements of justice must include the possibility of secession. When such a possibility is allowed, it can be predicted that there will be changes not only in the character of disputes about the requirements of justice, but also in the patterns of taxes and public expenditures. There will be a greater propensity for seeing the other side's point of view in disputes about the requirements of justice, and a greater tendency to support public activities by efficient taxes on the beneficiaries of public expenditures.

The paper begins with a discussion of the nature of moral truth, its relation to scientific truth, and the way in which moral knowledge grows. Next discussed is the difficulty of translating moral knowledge into social institutions, arising from the inevitability and impropriety of judging one's own cause. Ackerman's "neutral dialogue" is endorsed as the most acceptable way of dealing with this difficulty. But I suggest that in dialogues regarding the requirements of justice there should be an understanding that one possible outcome of the dialogue is failure to agree on mutually acceptable conditions for being part of the same society, leading to a parting of the ways. The conditions under which such a parting would occur constitute the most fundamental question of justice. I suggest that Ackerman's proposed condition of equal sharing of the providence of nature (Ackerman's initial manna) among all generations constitutes an appropriate basis for parting if agreement should be impossible.

I argue that such an understanding of the possibility of secession would provide a better framework for the growth of moral knowledge than when the politically successful are able to preclude experiments with alternative conceptions of justice. It would also reduce opportunities for the politically adroit to exploit the less adroit. If competition among societies for citizens resulted from the possibility of secession, the competitive equilibrium would include land value taxation.

The Nature of Moral Truth and Moral Knowledge

Any society embodies shared moral convictions. The dominant shared moral convictions of our own society are those of liberalism, which has been defined as the effort to achieve "a political, legal, and economic framework in which assent to one and the same set of rationally justifiable principles [enables] those who espouse widely different and incompatible conceptions of the good life for human beings to live together peaceably within the same society."

constitution, and has enjoyed growing success for more than two centuries.

Moral frameworks such as liberalism have the character of paradigms in science. A scientific paradigm is a set of axioms or presuppositions, along with a cultural tradition within a community of scientists, specifying how one identifies an interesting research problem and how one decides whether a new claim qualifies as an addition to knowledge. Similarly, liberalism embodies a set of presuppositions ("We hold these truths to be self-evident, that all men are created equal...") and a tradition regarding the way in which new moral truths are identified and incorporated into the social order.

My reference to new moral truths may seem incoherent. Isn't morality just a matter of personal view? The philosopher Renford Bambrough has offered the following proof of the possibility of moral knowledge: "We know that this child, who is about to undergo what would otherwise be painful surgery, should be given an anaesthetic before the operation. Therefore we know that at least one moral proposition is true."

Bambrough provided this example to show that a person who denies the possibility of moral truth denies platitudes that are as widely accepted as the proposition that physical objects exist. I accept Bambrough's view that moral propositions can be just as true as propositions about the physical world, but I deny his apparent assertion that we can have a basis for being completely confident that we will never revise a view that any particular proposition, of any sort, is true. Any understanding of what is true exists within a framework that entails presuppositions, which are often not even discerned until they are questioned, and when they are questioned it may seem inconceivable that they could be false.

For example, most of us, myself included, operate in our daily lives on the basis of an unprovable presupposition that space-time has a unique history, that there is always a unique answer to the question, "What really happened?" We may find it impossible to imagine having a basis for revising this presupposition, which is involved in so much of the way we organize our understanding of our existence. The degree of our commitment to this presupposition is probably no less than that of nineteenth century physicist to the (then unexamined) assumption that the mass of an object does not depend on its velocity, which twentieth century physicists reject.

We have neither the possibility nor the need for absolute certainty. Thus the only possible operational meaning of "true" is "belief that is justified within a framework shared by those engaged in dialogue." However, it is an important part of human psychology that we are consistently unaware of the presuppositions that give us less than absolute certainty. This unawareness might even be necessary to save us from doubt-induced immobility. Thus when we hear someone assert that something is true, we generally understand him to be asserting that he is offering a correct description of an aspect of objective reality. "True" conveys this understanding of "absolutely true" within a system, for persons who are not aware of their presuppositions, while those who are aware of the presuppositions of the system in which an assertion of truth is made understand "truth" to mean "belief that is justified within the system in which the asserter operates."

New moral truths, like new scientific truths, are propositions that attain the status of being regarded as fully justified within the community of those who share a given paradigm. The most far-reaching new moral truths to have emerged from our commitment to liberalism are that slavery is utterly wrong and that all persons must be accorded the same civil and political rights, irrespective of gender and race.

A very significant step in the incorporation of a newly discovered moral truth into our social order is its embodiment into our constitution. Mere legislation does not suffice to certify a moral truth, for a reason articulated by Bruce Ackerman in an exegesis of the views expressed by James Madison in Federalist 103. We understand that citizens have too many other interests in life to attend to all the political issues that arise. Therefore everyday politics is bound to be dominated by a few people pursuing narrow interests. In the constitution-writing process, on the other hand, the level of involvement is sufficiently extensive and the level of consensus required sufficiently great to permit those who succeed in having their views incorporated into the constitution to say that they speak for the whole nation. What is crucial, according to Madison, is not that the letter of established constitutional procedures be followed, but rather that the society be widely involved in discussions about what is done. This was Madison's reason for arguing that the constitution written in 1787 could be valid despite the fact that the group who wrote it was not supposed to be writing a constitution. Ackerman argues that the same reasoning applies to the seemingly defective procedures by which the post-Civil War constitutional amendments were forced through, and to the revision of the Supreme Court's interpretation of the Constitution in the 1930's, under Franklin Roosevelt's threat of court-packing. The standard constitutional amendment process, of course, is designed to entail the levels of involvement and consensus that permit the advocates of a successful amendment to speak for the nation.

While the constitutional process succeeds in achieving a substantial consensus about moral truths, it is not a perfect process. Our constitutional reversal with respect to the prohibition of alcohol is proof, if one were needed, that not every idea incorporated into the constitution is an advance. The real truth of any idea comes not from the fact that it is incorporated into a constitution, but rather from the consensus that supports it. And in the process by which a constitution evolves, there are often substantial groups that feel extremely ill-treated. These groups can attack the justice of any constitutional outcome on the ground that those who favor the successful outcome are judging their own cause.

The Problem of Judging One's Own Cause

Concern for the inappropriateness of judging one's own cause has a long history in political philosophy. Thus Thomas Hobbes says, "And seeing every man is presumed to do all things in order to his own benefit, no man is a fit arbitrator in his own cause."4 John Locke says, "I easily grant, that Civil Government is the proper Remedy for the Inconveniences of the State of Nature, which must certainly be Great, where Men may be Judges in their own Case, since 'tis easily to be imagined, that he who was so unjust as to do his brother an Injury, will scarce be so just as to condemn himself for it."5 Immanuel Kant says, "...the people, under an existing civil constitution, has no longer any right to judge how the constitution should be administered. For if we suppose that it does have this right to judge and that it disagrees with the judgement of the actual head of state, who is to decide which side is right? Neither can act as judge of his own cause." 6 In Federalist 10, James Madison says, "No man is allowed to be a judge in his own cause, because his interest would certainly bias his judgement, and, not improbably, corrupt his integrity."7

Among modern writers, recognition of the need to avoid the bias of self-interested judgement is reflected in Rawls' assertion that justice is described by what individuals would unanimously choose from behind a "veil of ignorance," where they did not know their particular circumstances.8 Buchanan and Tullock employed a similar concept in their assertion that desirable social institutions entail "conceptual unanimity."9 But neither Rawls nor Buchanan and Tullock address the practical question of what steps are warranted in the face of an actual absence of unanimity. The work of Buchanan and Tullock appears to be intended to be primarily descriptive of what individuals will choose in specified circumstances, rather than prescriptive of what ought to be chosen, while that of Rawls is more clearly prescriptive, and thus seems to invite the following interpretation: "This is what individuals would agree to if they were not biased by self interest and knowledge of their particular circumstances, so we are justified in imposing it on them whether they agree to it or not." I do not believe that Rawls actually thinks this; I would guess that his view is actually closer to the following: "It is inevitable that democratic majorities will impose their will on minorities. But people do have a moral sense. And the framework I offer is one that can be commended to individuals who wish to have their democratic actions informed by moral considerations."

In commenting on Rawls, Kenneth Arrow has said,

...if the specific application of the principles is judged to be different according to different life experiences (and of course different genetic experiences), even as between parent and child, then the needed concordance of views may evaporate....To the extent that individuals are really individual, each an autonomous end in himself, to that extent they must be somewhat mysterious and inaccessible to each other. There cannot be any rule that is completely acceptable to all. There must, or so it now seems to me, be the possibility of unadjudicated conflict, ...10

This unadjudicated conflict leaves any constitution with an imperfect claim for respect, as it is necessarily supported by individuals who are judging their own cause. How are conflicting claims about the constitutional requirements of justice then to be resolved?

Ackerman's Liberalism

One writer who has characterized liberal principles explicitly in terms of the resolution of conflicts is Bruce Ackerman, in "Social Justice in the Liberal State". Ackerman gives his formulation for the foundation of the liberal state the name "Neutrality." This is the principle that every holder of power must be prepared to answer every challenge to his power with a rational, consistent justification that does not assert either that his conception of the good is better than that asserted by any of his fellow citizens, or that he is intrinsically superior to any of his fellow citizens.11 Ackerman argues that adherence to the principle of Neutrality requires a society in which each person starts his or her adult life with the same material endowment with which others start their lives, apart from compensation for genetic domination by others, and with an education that provides a basis for self-definition. Ackerman goes on to argue, in typical liberal fashion, that in pursuing one's conception of the good life, one is free to interact with others in whatever ways are mutually desired and do not interfere with others.

The limits of individual self-definition are reached when one comes to goods such as air quality, which can only be consumed collectively.12 But since no individual can assert, within the constraints imposed by Neutrality, that his conception of the proper trade-off between air quality and other goods is superior to that proposed by any other person, liberal principles come under tension. Ackerman makes two suggestions for liberal ways of dealing with this tension. The first is what Ackerman calls federalism (which would be more familiar to economists under the name of the Tiebout solution): Let people sort themselves out into groups that favor the same level of air quality and each have the level they prefer. Ackerman recognizes that this solution is not perfect, both because it is not possible for every option with respect to provision of public goods to be offered, and because costs of moving are significant. "Despite these serious imperfections," Ackerman says, "a federal regime is far better than some all-or-nothing solution that would force all deviants to pay for a collectively specified level of good air or force all environmentalists to suffer with bad air merely because they cannot transcend the organizational hurdles imposed upon them by an individualistic property-holding system."13

The other suggestion that Ackerman offers for dealing with differing tastes for public goods is "Lindahl solutions," in which people are called upon to pay for public goods in proportion to the benefits they receive, though he recognizes that we do not possess the means to implement this concept perfectly. Two chapters later, Ackerman addresses the issue of how to resolve disagreements regarding the best approximation to the liberal ideal when its complete attainment is not possible.14 Here he offers a different pair of suggestions: majority rule (with any majority-rule cycles that arise resolved by a random process), and the use of a lottery to pick one of the suggested options.

Secession as the Resolution of Moral Disagreement

One area in which Ackerman does not entertain the possibility of sustained, good-faith disagreement is in regard to what the liberal ideal requires, though he does repeatedly contrast the results of his analytical approach with those of contractarianism and utilitarianism. Ackerman is disregarding a very significant question here. Political dialogue presupposes some shared understandings, but there is no way to predict in advance the places where the shared understandings will end.

Suppose, for example, that a child is born with a disorder that will almost certainly kill him after a few very painful years. Should the child's parents be allowed to subject the child at birth to the least painful death that technology permits? I expect that there will be people with different beliefs, each feeling very strongly that his or her answer to this question is correct. Is there any reason for believing that persons of goodwill will all reach the same conclusion as to what the right answer to every such question is? And if consensus cannot be attained, should the disputants be obliged to resolve the question by majority rule or by lottery? There is another way of dealing with disagreements about what justice requires, and that is to allow for the possibility that the persons involved will not in the end be parts of the same sovereign state.

The possibility that those who disagree about what justice requires will not be part of the same sovereign state is already allowed for to some extent in the liberal consensus that no state may justifiably prevent its citizens from emigrating. (There is some sentiment that young men subject to a military draft can be prevented from emigrating, but even in this domain the U.S. was unwilling to make a serious effort to prevent emigration during the Viet Nam war.)

But why should the opportunity to register one's unwillingness to be part of the same sovereign state be limited to emigration? Suppose that, like Ackerman in his initial discussion, we were talking about setting up rules for the settlement of a previously unsettled planet, or suppose that we were talking about a fully claimed planet like our own, and that there was no nation that was willing to tolerate the behavior proposed by a minority-- perhaps the use of a drug such as cocaine or caffeine. Why shouldn't the majority and the minority each have their own sovereign state?

To some extent our federal system allows for just such a possibility, in that such issues as whether public places must be free of tobacco smoke are decided differently by different jurisdictions. But when an issue arises that is so important to one faction's sense of what justice requires that they cannot tolerate being part of a sovereign state in which their view does not prevail everywhere, then their only options are either emigration or an unending effort to impose their view on the entire nation through constitutional amendment. Is there never a time for an amicable parting of ways?

There are several practical difficulties with secession as the resolution of moral disagreement that need to be mentioned.

• First, as long as the world seems so hostile that defense against foreign invasion is required, there will be efficiencies in defending larger areas.
• Second, when sovereignty is fragmented, there will be a temptation to levy taxes that will be paid predominantly by foreigners. It was in response to the recognition of this that the U.S. Constitution included a ban on taxes in interstate commerce. The same impulse can be seen today in the desire of localities to impose taxes on restaurant meals, backed up by the argument that a substantial fraction of such a tax is paid by people from out of town.
• Third, if any partition were undertaken, there would be likely to be numerous people who found themselves on the wrong side of the dividing line, as happened so tragically when Pakistan was partitioned from India.

I will respond to these in order. As a general rule, apparent efficiency is not a sufficient reason for imposing on people what they do not want. While the efficiency of defending a larger area is real, is not a sufficient reason for insisting that people be part of the same sovereign state when some of them do not want that. The temptation to levy taxes that will be paid by foreigners is somewhat different. This would be an adequate reason for resisting the secession of a group that occupied a crossroads in a sovereign nation, but it would not be an adequate reason for resisting the secession of a group that occupied some reasonably convex region on the periphery of nation, such as the State of Maine.

Suppose then that it were proposed that Maine secede from the nation. On what liberal ground might this be opposed? If it was only a bare majority of the residents of Maine that wanted to secede, then one could raise the objection that those who wanted to secede had no right to impose costs of adjustment on those who did not want to secede. While this argument carries some weight, it is not decisive, because neither would the unionists have a right to impose their view on the secessionists. The argument is a stand-off. What can be said is that the greater the degree of consensus within a reasonably convex, peripheral region that is contemplating secession, the harder it is to justify opposition on grounds of adjustment costs. In the limit of complete unanimity, the argument would carry no weight at all. Thus liberal principles require that there be some conditions under which secession is permitted.

Still, there are some secessions that could be opposed by a liberal, even without adjustment costs. Suppose, for example, that those who wanted to secede were a few persons who owned tens of thousands of acres of virgin forest. A consistent liberal could argue that there was something faulty about the process by which individuals came to be regarded as the owners of a disproportionate amount of what had been provided not by the effort of any person, but by nature. Those who wish to secede can be obliged, in a liberal framework, to insure that they do not claim a disproportionate share of land and natural resources.

So suppose that those who wished to secede claimed no more land and natural resources per capita than were available to others, but they constituted a group that was distinctly more capable than average. Could those who were left behind argue that this departure of talent was unfair to them? They could not consistently oppose the departure per se, as those of unusual talent would certainly be permitted to emigrate rather than secede. But Ackerman argues that genetic disadvantages must be offset by extra resources devoted to education and by greater assigned shares of land and natural resources. If this argument is accepted, it means only that an unusually talented group that proposed to secede could be obliged to accept smaller-than-average allotment of land and natural resources. They could not be required to refrain entirely from seceding. At worst, they might be so talented that awarding them even the smallest portion of land and natural resources would be unfair to those left behind. In that case the secessionists would still have a reasonable claim to the chance to the use of some territory for their own sovereign state if they were prepared to pay the less talented a fair rent for its use. The superior talent of those who wish to secede cannot bar their secession.

The result of the argument that secession must be possible in some circumstances is that dialogue regarding the requirements of justice takes on a new character. Instead of being limited to endless dialogue, majority rule and random process as ways of resolving disagreements about the requirements of justice, disputants have the additional option of agreeing to disagree and go their separate ways. This means first that a bound is placed on the harm that can be done to those who might be dissatisfied with an outcome that emerges from the process of constitutional evolution. Second, those who are able to secure an outcome they desire within established procedures can justly be somewhat less concerned about their inability to achieve complete consensus, since those who are dissatisfied could secede if they wished (though the substantial moving costs that would be entailed imply that impositions would not be entirely eliminated).

This also means that greater authority is given to whatever emerges from the constitutional process. The effort to achieve a consensus on the requirements of justice, and to forge a correspondence between that consensus and a constitution, does not terminate arbitrarily at the adoption of a constitution or amendment, with a protesting minority completely vanquished. Instead, those who are disappointed in the constitution-writing process have the opportunity to give expression to their conception of justice in their own sovereign state, to set an example or to learn over time from the example of others.

Finally, the possibility of secession means that constitutional dialogue would be reshaped. In view of the costs to all parties that would be entailed by secession, the factions involved in a dispute would have a greater incentive to strive for a general agreement than when all the costs of disagreement can be predicted to fall on the faction that loses in the constitution-writing process.

With respect to tax and expenditure decisions, politically active special interests would face an important additional limitation in their efforts to exploit the rest of the public: An exploited group could secede. While the costs of secession would limit the use of secession as a device for avoiding exploitation, the need to be sensitive to this possibility would create pressure for benefit-related taxation and for an efficient public sector. If alternative states are viewed as competing for citizens, one can ask what offers potential citizens would receive from states. The answer from economic theory is that each citizen would be offered his marginal product, plus the rent on the share of land and natural resources to which the state would have a claim by virtue of his citizenship. Competitive pressures would induce states to refrain from taxing labor incomes, unless there were social costs associated with the generation of income. Future capital accumulations would similarly tend to escape taxation. Thus land, natural resources and previously accumulated capital would be the only tax bases not subject to erosion by competitive pressure.

Footnotes