«« Ïðåä. | ÎÃËÀÂËÅÍÈÅ
Agenda and Procedures
Types of Issues
Given the Herculean nature of the task of drafting a constitution, it will be natural to subdivide it. Three broad divisions come to mind: delineation of the institutions of government, definitions of rights, and procedures for amending or replacing the constitution.
Whether to create a two-party or multiparty parliament, one that is unicameral or bicameral, etc., are the types of questions that will most concern members of an elected parliament should they make up the constitutional convention. In this case, agreement will be difficult, and compromise surely necessary for a consensus to be reached.
William Riker (1987) highlighted how lucky America was that the Antifederalists chose to boycott the Philadelphia Constitutional Convention. The convention was thus composed almost entirely of Federalists with similar views as to what the structure of government for the United States should be. The Federalists were able to reach a near consensus on a coherent set of institutions that allowed the country to grow and prosper over the next 200 years. Had the Antifederalists been present, no consensus might have been reached, or if one had it might have involved compromises that produced an unworkable governmental structure.
The lesson to be learned from the American experience is not to exclude all who do not share a single view of optimum government structure. The lesson is that delegates have to choose among alternative governmental structures (two-party or multiparty, centralized or federation) and, having made these basic choices, design institutions that function well together. It would be better to pick by lot one from among three or four logically consistent government structures than to choose a compromise that is an unworkable hodgepodge.15
In Chapter 14, we analyzed the characteristics of constitutional rights under the assumption that each individual at the constitutional stage is uncertain of her future position and assumes that she has an equal probability of being all future citizens. This uncertainty-equal-probability assumption at the constitutional stage seems most justified with respect to rights. (Who can be sure that some group will not seek to ban a book that she or her granddaughter will wish to read?) Even if a parliament drafts the constitution, the kind of partisan issues that arise when the rules for electing parliament are written, are not raised regarding rights.16 Consensus on a proposed bill of rights should be much easier to obtain than on the particulars of government structure.
More room for compromise also exists in choosing a set of rights than for government structure. An article regarding slavery can be written independently of one concerning free speech. Such logical independence opens up avenues for compromise. A feels strongly about free speech, B about abortion. A gives in to B on abortion in exchange for B's support for a strong article on free speech.
Last of all, the convention must decide how the constitution is to be amended or perhaps even replaced in toto. There are several options here that we have discussed already.
1. Provision is made for the periodic convening of a constitutional assembly with the authority to amend or replace the existing constitution. This assembly could be governed by the same rules as the first, or by new rules.
2. Provision is made for calling a new convention upon some fraction of the citizens or parliament or states demanding it. Once convened it has the same powers in amend or replace the existing constitution.
3. The constitution can be amended by referendum.
4. The constitution can never be changed, and all ambiguities and boundary disputes are settled by the judiciary.
The same issues regarding selection of delegates and voting rules arise under options 1 and 2 above as for the original constitution. The case against the parliament serving as a constitutional convention is even stronger when the convention is called under an existing working constitution. Time and the threat of anarchy should not be pressing concerns, and the polity can afford the luxury of convening an assembly that functions independently of the parliament.
If the consensual nature of the constitution rests on the size of the majority that ratifies it, the same supramajority should be required to amend it as was originally required to ratify it.17 This requirement, coupled with the inability to amend a referendum proposal once made, makes amendment by referendum an unlikely occurrence.
Another potential disadvantage of amendment by referendum arises from the likelihood that the original set of rights obtained their supramajority as a package-i.e., as a result of vote trades. If B traded her vote on free speech for A's vote on abortion, and the article on abortion is subsequently amended to 5's disapproval, B will have given A her vote rather than traded it. If B anticipates such a successful referendum effort, she will not trade her vote in the constitutional convention, and the likelihood of obtaining the required majority on a package of rights falls. This consideration favors reconvening a full convention to amend the constitution rather than piecemeal amendments by referendum18, so that whole packages of changes involving new trades replace previous packages.19
Entrusting all changes to their agents in the judiciary is the simplest and potentially quickest procedure the citizens can adopt to keep the constitution up to date. It suffers from two main defects. First, the court with final authority to settle constitutional disputes must consist of but a handful of individuals, and is thus a very small sample of the population. Its judgment may diverge from that of the larger population, just as the mean of any small sample can diverge from the population mean.
Second, even if a decision made by the court is exactly the decision that a newly convened convention, or citizens' referendum, would have made, there may be an advantage in reaching this decision by one of the more circuitous routes. If the constitutional convention is properly constituted, citizens know that they have been fairly represented. They can observe and consider the arguments on all sides of the issue as it is debated; debates among members of a court are never observed. The citizen knows, and presumably accepts, that the original constitution was ratified by a substantial majority and that any changes in it require the same majority. Changes cannot be made by simple majorities as is generally true of courts of last resort. In a referendum the citizen participates directly and is exposed to arguments on all sides of the issue. The more time-consuming amendment procedures should generate a better understanding of the decisions made, and thereby conviction that the correct decision has been made. In so doing, they are more likely to maintain citizen consensus on the provisions of the constitution and compliance with its provisions.20
In the end, the convention must consider all elements of the constitution as a package. Final approval is of the total package. It has been a theme of this book that the choice of one part of the constitution depends on the other parts. Rights protecting a citizen against the state are more important in systems in which minorities are poorly represented in the parliament and the parliament uses the simple majority rule than where supramajorities are required in a fully proportional representation system. Rights can be defined more precisely the easier it is to amend the constitution. Although the initial work of a constitutional convention can be expected to involve dividing the task into subsets of issues, to obtain a coherent whole, some consideration for how each part looks must go into the Grafting of the other parts. A reconciliation of the different parts must occur before final ratification.
The Question of Timing
If the constitution provides that new conventions be called at set intervals, it must stipulate these intervals. Tension arises between the long-term nature of the constitutional contract and the desirability of its adaptation to the changing circumstances, knowledge, and interests of citizens. The only example of such a provision in a constitution of which I am aware is the stipulation in the Polish Constitution of 1791 to reconvene every 25 years. This interval would satisfy Jefferson's recommendation that each generation be allowed to write its own constitution (see quote at beginning of this book). For the constitution to be a contract among living citizens, a citizen should have the opportunity to observe and participate in the process once in his lifetime. An interval somewhere between 25 and 50 years would satisfy these requirements.
The issue of timing arises in yet another way. We have stressed how uncertainty over future position can facilitate consensus. By introducing impartiality of a sort, it can also add normative content to the constitutional contract (Rawls, 1971; Gauthier, 1986). Uncertainty is greater the further into the future an action occurs. Both the normative content of a constitution and the likelihood of its achieving a substantial majority should be enhanced if its provisions take effect sometime in the future. This option is not likely to be optimal with respect to all provisions of the original constitution, but it might be an attractive option for some of them.21 Both its feasibility and attractiveness are greater in conventions established to modify or replace an existing constitution.
For example, suppose that a newly formed country had adopted a two-party system because its founders wanted an «effective government,» given the many serious and urgent problems the country faced. Some years later, when the survival of the nation is no longer threatened, many citizens feel that the two-party system generates alienation and civil strife, owing to the lack of direct representation of minority views.22 If a constitutional convention proposes that the required institutional changes go into effect in one year, vigorous opposition from the current majority party and their supporters, and perhaps also from the second largest party, is likely. If the changes are to take place after ten years, many current members of parliament will have retired, others will have had time to adjust. Introducing delays in implementation might make agreement on change more likely, leading to a more rapid updating than would otherwise occur.
The Question of Secrecy
The final procedural issue to touch upon is whether the convention's deliberations are to be secret or open.23 This question arose at the Philadelphia Convention and it opted for secrecy. It did so to encourage frankness, to inhibit «grandstanding,» and to allow delegates to change their minds without being criticized (Bems, 1988, pp. 134-35). With the exception of grandstanding, these dangers of open discussion seem likely to be most acute when delegates to the convention are or expect to become elected representatives of one form or another. Representatives of this sort must always be concerned about how what they say «plays back home.» But a delegate chosen by lot or election solely to participate in a constitutional convention can stand to «win» as much as she «loses» by being frank or changing her mind. She need fear no retaliation at the polls. Any costs she bears from what she says are likely to be her own «psychological costs « and these might even arise in a secret session.
The advantage of an open convention is that the entire nation can follow the discussion, learn the arguments on all sides of the issues, and understand why the sequence of circumstances led to the particular compromises reached. In following the course of a convention, the citizens may be led to the same conclusion as the convention-that the proposed constitution is the best to be gotten. With a closed session, the process of educating the citizenry must begin after the constitution is completed. In the case of the U. S. Constitution, this occurred through the published exchanges between the Federalists and Antifederalists in the months during which each state considered ratification. In the European Union new «constitutions» have been presented to voters without their participation in the drafting process or much national debate. With no Jays, Madisons, or Hamiltons to plead for their adoption they sometimes, as with the Maastricht Treaty, have met with a cool reception.
Thus, the answer to the secrecy question hinges on the nature of the ratification process. If the convention is composed of delegates elected for the sole purpose of writing the constitution, and its vote is final, an open convention is desirable so that citizens can follow the arguments and debate leading up to the document that is eventually approved by the convention. Candor and a willingness to compromise on the part of delegates can be hoped for, since their future careers do not rest directly on their actions at the convention. Should they so rest, as when the constitution is drafted by a sitting parliament, its deliberations should be secret to encourage candor and compromise. Public debate must then come after the convention has finished its work in the interval leading up to the ratification of the constitution by a referendum vote of the citizens.
When States Are Delegates
This book builds on the assumption that the constitutional contract is joined by individuals. But it sometimes happens that a state forms by several smaller states (nations) joining. The United States and, in its political attire, the European Union are examples. Ultimately, whatever constitutional contract joins a group of states joins their people, but a convention in which citizens are represented by states raises some different issues.
The Contract and the Convention
Let us suppose that there are five areas, each separated from one another by natural boundaries formed by rivers (see Figure 21.1). No government exists that includes the people from all five areas, but residents of each area feel that such a government to handle transportation and environmental issues related to the river boundaries might be
Figure 21.1. The Territorial Distribution of Five Homogeneous Groups
collectively beneficial. A convention is called to consider drawing up a constitution for a new state in which the residents of all five areas are members.
Two quite different circumstances consistent with these facts can be envisaged. First, suppose that A, B, C, D, and E together form a landmass surrounded by water. The different areas have been occupied by settlers from another country. They all speak the same language and share the same culture and religion. Mobility from one area to another is high and most people have relatives in other areas. No existing governments in A, B, C, D, and E exist. More importantly, the viability of each of the areas as a separate political unit in the long run is in doubt.
Under these assumptions it is reasonable to think of the entire landmass as the potential state, the constitution a contract joining all individuals on it. The considerations discussed above apply. An election with the entire landmass as a single district, and individuals represented proportionally, is optimal. Unanimity is the ideal voting rule, a supramajority rule the practical alternative.
Contrast the above situation with one in which A-E are autonomous states, each with its own independently functioning government. Each is inhabited by a different ethnic group speaking a different language, and religious differences also exist. Each could survive as a separate state with common issues involving rivers resolved by mutually beneficial bargaining a la Coase (1960) between the states. A superstate joining all five becomes optimal only if it reduces the costs of making their mutually beneficial collective agreements. Under these circumstances a constitutional agreement involving all states is reached only if representation is by state and the unanimity rule is used. No rule other than unanimity is an option, since each state can survive alone should agreement on a superstate not be forthcoming. Each state might as well get one vote regardless of size, since the superstate cannot form without even the smallest state's consent.
The crucial difference between these two situations is the viability of a geographic area as a fully autonomous political unit. The prior existence of an autonomous government in an area is prima facie evidence that this area can survive as an independent political unit should no superstate be created. The people from this area will not join with those of the other areas if they do not expect to be better off as a result. Their vote-i.e., the vote of their state-is essential for the superstate to form. Should no area be viable as an independent political unit, however, the constitutional convention might under a supramajority voting rule decide to form the superstate over the unanimous opposition of the representatives from one of the five areas. Because this area cannot survive as a state on its own, it would have no alternative but to join the superstate.
Once formed, both superstates might look very similar, if the institutions of government are designed in an optimal way. Only issues that affect people in all five areas would be resolved by the central government. If the rivers that create A, B, C, D, and E lead to large sets of public goods common to the people of a given area, then separate state governments will be created in the five areas. If separate cultures, languages, etc., have evolved in the five autonomous states of the second case, the natural boundaries are evidently substantial and would make the retention of many government functions in these states optimal.
It is customary to define the former type of superstate a federation and the latter a confederation. But the important point to recognize is that the optimal distribution of governmental structures across the different levels of government should be essentially the same. Note, too, that for decision-cost reasons, both structures of government may choose to require less than unanimous consent to implement decisions at the central governmental level. The key differences between the two types of government are (1) that a confederation of independent states must use the unanimity rule as applied to states at the constitutional convention,24 whereas the constitutional convention in which individuals are represented geographically can use a supramajority rule applied to individual representatives, and (2) that a convention in which autonomous states are represented has more reason to adopt constitutional provisions to protect the individual states. We discuss three of them.25
A Nullification Clause
When independent states sacrifice their autonomy to form a superstate, the danger arises that the superstate undertakes actions that make the citizens of the individual states worse off. For example, suppose that the citizens in C belong to one religion whose sabbath is Tuesday, the citizens of A, B, D, and E belong to a different religion, which has a Thursday sabbath. The superstate's parliament votes to have Thursday be the sabbath in all parts of the superstate (stores and schools must close, etc.). A proposal, to declare Tuesday as a sabbath also, is defeated.
Such actions might impose obvious costs on citizens in C. They could protect themselves from such discriminatory legislation by agreeing to the superstate's constitution only if it contains a nullification clause. This clause might read as follows:
Nullification Clause: The force of any action by the Parliament of the Superstate as it pertains to the citizens of any individual state shall be nullified throughout the territory of the individual state should a motion of nullification receive at least three-fourths of the votes of the representatives to said state's parliament.
As always, the choice of a three-fourths majority as opposed to a two-thirds or four-fifths is somewhat arbitrary. But one suspects that a supramajority would be required for nullification, even if a simple majority was required for all other legislation, to ensure that the superstate's actions were not routinely nullified. When an action was clearly discriminatory with respect to all or nearly all of a state's citizens, achieving the required supramajority in its parliament would not be a major obstacle. Such a clause in the superstate's constitution would protect against the discriminatory legislation described in the example.26
State Veto Authority
A nullification clause protects a state from certain kinds of harmful legislation, but not from others. Suppose now that the religious composition of the individual states was as just described-i.e., A, B, D, and £ are of one religion, C of another, and suppose that a neighboring state, F, is of the same religion as C (see Figure 21.2). Border clashes occur between B and F and escalate to the point where the superstate declares war on F. A nullification clause would allow C to stay out of the war officially, but given its location would not be likely to keep it from being damaged by the war or perhaps even brought into it. Moreover, C might not wish merely to stay out of the war, but to
Figure 21.2. A Confederation of Groups with an Adjacent Foreign State
prevent it altogether. To allow C to accomplish this objective a different form of constitutional protection is required, a state veto clause.27
Veto Clause: Any action of the Parliament of the Superstate can be nullified across the entire area of the Superstate by a three-fourths vote of a parliament of any one of the member states.
A Right to Secede
Autonomous democratic states will voluntarily contract to form a superstate if the expected welfare of their citizens in the superstate exceeds what they expect to achieve going it alone. However, although contracts are written to reduce uncertainty, they do not eliminate it entirely. The people of some state may find themselves worse off in the superstate than they think they would have been had they stayed out. Anticipating this possibility, they may demand conditions in the contract forming the superstate stipulating how they can peacefully withdraw from it. Since all joining states may be uncertain of their long-term benefits from being in the superstate, all may have an interest in delineating procedures for the withdrawal of individual states or the complete dissolution of the whole. Such protection would be provided by a secession clause.
Secession Clause: Any state can secede from the confederation of superstates, if it satisfies the following x conditions: . . .
The conditions under which secession should be allowed take several forms. Clearly all states would wish to allow a member to exit if all were going to benefit from its doing so. Secession should be allowed when the parliaments of every state approve- by the constitutionally required majority-the secession of a particular state. Often the exit of one state will impose costs on the remaining members of the superstate. Secession can be allowed in this case if the exiting state compensates the others. For example, secession could be allowed if the exiting state assumed its share of the fixed costs that were incurred under the assumption that the exiting state would remain a part it. At least three options exist: (1) Require that the exiting state pay the present value of its share of the debt amortization when it exits, (2) assume its share of the debt when it exits, or (3) allow states to exit only after an interval has passed from the time they announce an intent to withdraw. The logic of option 3 would be, since all capital depreciates, to run down the level of superstate capital while requiring the departing state to pay its share of the costs of the investments already made.
If no part of a federation could ever stand alone as a viable state, then no provision for secession need be included in its constitution, because no group of individuals occupying a particular area would ever choose to secede. A threat of secession would not be credible.
But viability is probably not a 0/1 variable. If Hong Kong can survive, indeed thrive, as an independent quasi-state, almost any territory large enough to become a city could conceivably survive as a separate state at some level of welfare. If the constitution allows collective decisions to be made for the whole state by a rule that requires less than a full consensus, the danger always exists that some permanent coalition forms and tyrannizes over a permanent minority. Geography is one basis upon which coalitions form. By allowing the people in a contiguous area to secede, under some conditions, the constitution could protect against such future tyranny. A secession clause is desirable even in a constitution written by citizens if they fear the formation of permanent geographically based coalitions.28
Thus, a secession clause might be an optimal part of any constitution. Nevertheless, it seems more likely to be valuable in constitutions that join states than in those that join citizens. Ethnicity, religion, language, and culture have historically been powerful catalysts for coalition formation, which results in one group tyrannizing another. When these characteristics are related to geography as they typically are when states are the building blocks of a superstate, the danger of geographically based tyrannous majorities is much greater, and so too therefore the advantage of a secession clause to protect previously autonomous states from tyrannization.
Although a secession clause may be desirable in a constitution written by citizens, the nature of any geographically based coalitions will be harder to predict and thus the wording for the clause more difficult to agree upon. Most countries are criss-crossed by rivers, mountain ranges, lakes, deserts, and the like. Which geographic features can over time isolate a minority, and lead it to be tyrannized by the rest of the country, will be unpredictable when the constitution is first written. If someone from Mars were to look at a topographical map of the United States and be asked to predict which two regions would have engaged in a Civil War, the visitor would probably predict the area east of the Mississippi against the area west of it, or the areas east of the Rocky (Appalachian) Mountains against the areas west of them. That the Potomac River would have divided the country into warring factions would be hard to predict.
Strategic Veto
A veto clause would extend the state-unanimity rule that effectively operates at the constitutional stage to the parliamentary stage. No legislation enacted by the superstate's parliament could go into effect if a single state was opposed. All of the potential disadvantages of the unanimity rule would thus exist with respect to a veto clause. In particular, the danger of a strategic veto would arise. For example, state C threatens to veto the defense appropriations bill unless its share of the costs are reduced. State B does likewise with respect to expenditures to protect the environment, and so on. These potential problems might be mitigated by delimiting the types of issues for which the veto could be invoked. But such restrictions would also reduce the potential for the clause to protect states from discriminatory actions by the other states through the super-state's parliament. The potential decision-making and strategic bargaining costs induced by a state veto clause must be weighed against the benefits of protecting individual states from discriminatory actions considering a state veto clause.
Strategic Secession
Consider an area composed of a people with a common culture. A river runs through the area and one can define an A and B side, but no one thinks in those terms. The people decide to form a government. In its absence the welfare levels (utilities) each A and B obtain are assumed to be zero. If all of the gains from cooperation go to the A's, their utilities would be raised to a level of one; the same is true for the B's (see Figure 21.3). The people choose political institutions that lead to an equal sharing of the gains from government (the Nash [1950] bargaining solution), and move from point S without government to point X with it.
Figure 21.3. Possible Outcomes with Constitutional Bargaining
Now suppose that oil is discovered under the area occupied by the B's. The income from this oil allows the utilities of everyone in the country to double. The governmental expenditure and tax structure that existed prior to oil's discovery results in a shift from X to Y. The discovery of oil awakens a sense of «Bness» among those on the side of the river where the oil was discovered, however. They point out that the sun sets on their side of the country and other important characteristics that make 5's different from A's. They announce the intention to exercise their inalienable right to secede. Should the B's secede, the social outcome would shift from Y to Z as the gains from cooperation are lost. Realizing this, the A's offer to renegotiate the constitutional contract assuming that Z rather than S is the status quo. The B's agree, the union is saved, and the new social outcome is at W. The 5's have gotten what they wanted.29
We shall not delve into the normative issue of whether the B's ought to be allowed to secede.30 Our concern is with the ramifications of the possibility of strategic secession for the original contract.
As emphasized in Chapters 5 and 16, the constitutional contract may have attributes that resemble an insurance contract. Risk-averse individuals may include provisions in the constitution that mitigate the effects of commonly shared risks. Provisions for sharing the costs of a natural disaster that strikes a particular area, or the gains from a windfall to an area, have this property. Implicitly, such provisions do exist in most if not all constitutional contracts. An earthquake, flood, or avalanche that devastates an area brings an immediate reaction from the president or the parliament to send relief, paid for by the rest of the country. Such responses do not seem to engender much if any opposition. This behavior may reflect altruism in the other citizens, or the existence of an implicit insurance component in the social-constitutional contract that joins them. People in each pan of the country realize that disaster could strike their area and expect their fellow citizens to respond as they do should it happen.
To reinforce the insurance provisions of the constitutional contract, the framers could make strategic secession of the type described more difficult-i.e., to make it difficult for an area that has experienced a large positive windfall to secede from the country, and for the rest of the country to secede from an area struck by disaster.
At the same time, the constitution framers may wish to establish institutions to reward and encourage effort, investment, risk-taking, and other market activities with positive spillovers. To the extent that finding oil requires investments in education, capital equipment, and risk-taking, those finding it should be rewarded. Once the efforts and investments have been made and the oil discovered, self-interested individuals in other areas have an incentive to change tax laws and property rights to obtain a bigger share of the income generated from the oil. The use of a qualified majority voting rule that falls short of requiring a consensus may allow them to do so. The option of seceding can protect an area from tyrannous majorities seeking to engage in such ex post redefinitions of property rights (Buchanan & Faith, 1987).
Ex post it may be difficult to distinguish strategic threats of secession stemming from a chance change in the status quo, from threats to protect the legitimate gains from effort from an ex post recontracting by a tyrannous majority. Suppose the income of the B's is higher because they work harder and longer hours than do the A's. The A's attribute their disadvantage to its being hotter and more humid on their side of the river. The B's complain that the A's are lazy and have too many children. The A's use their numerical advantage to vote for measures that redistribute income from B to A. Is this the kind of chance outcome that the constitutional contract should ensure against, or is it an example of a tyrannous majority?
General answers to questions such as this are difficult if not impossible to deduce. What one can say is that the case for including a secession clause in a constitution joined by autonomous states, and the nature of such a clause, is far stronger and easier to conceptualize than for a constitution joining a community with a common culture. If A and B are autonomous states before they join, and oil is subsequently found in B, the B's may argue (1) that the oil would have been theirs alone without question had the superstate with A not been formed, (2) that the superstate was formed to engage in a limited set of collective activities implicit if not explicit in the constitution, and (3) that insurance-type risk and windfall spreading was not one of them. And the A's might reluctantly agree. Certainly, the kind of put-oneself-in-the-other-fellow's-shoes experiments that lead to impartiality and insurance-like social contracts are more readily accepted by individuals who share a common culture than by those who do not. But in both situations the potential for desired redistribution for insurance purposes and undesired redistribution is present. When deciding whether to include a secession clause, the constitution framers need to consider the functions that government is intended to serve, and how best to ensure that it is these functions, and only these, that it undertakes.
The Secession of the American Southern States: An Aside
Nothing in the U.S. Constitution specifies an unconditional right to secede held by each state, or a right to secede under certain conditions. In this context the Articles of Confederation were much more precise than the Constitution that replaced them in stating that «each state retains its sovereignty, freedom and independence,» which might be interpreted as allowing immediate secession from the union or the existence of no real union whatsoever.32 If a right to secession is in the U.S. Constitution, it appears as one of those implicit rights that owes its presence to the absence of any explicit statement to the contrary.
The above discussion suggests that the citizens of each American state would have felt that their state and every other had a right to secede to the extent that they were autonomous and viable political units at the time they joined and could remain so had no superstate formed. In favor of their being a perception of such a right are the facts that (1) the states had functioned as essentially autonomous political units since the Revolutionary War's end, (2) voting was by state and not by individual representatives at the Constitutional Convention (a proposal for the latter made by Benjamin Franklin was defeated), and (3) language in the Constitution and the Declaration of Independence referred to sovereign states and the union thereof.
Against this perception are the facts that (1) the states shared a common history as colonies of Great Britain; (2) they thus also shared a common culture;33 (3) they had seceded collectively from Britain and had been joined in battle against the British, actions which must have engendered feelings of common bonds as Americans; (4) although the delegations voted as states rather than as individuals, they did not employ the unanimity rule (the final vote by the dozen states present was unanimous, however); and (5) despite strong opposition to the federal structure created in the Constitution, no state chose to stand outside the union, suggesting that no state felt sufficiently viable as an independent political unit to go it alone, as, for example, Switzerland has in Europe.
Even if one accepts the position that the states, at the time the Constitution was written, thought of themselves (and in fact were) as viable, separate entities, one might argue that by 1860 the conception that the United States was a single unified country had so taken root that the nation could not be redivided into its constituent parts. James Wilson of Pennsylvania suggested that this meltdown of state sovereignty occurred as soon as the Constitution came into being. «Among the first sentiments expressed in the first Congress was that Virginia is no more. That Massachusetts is no [more], that Pennsylvania is no more, etc. We are now one nation of brethren» (quoted by Berns, 1988, p. 140).
Merit exists on both sides of the issue. If there is one simple lesson to learn from the unfortunate experience of the United States regarding secession, it is that any autonomous nation joining in constitutional contract with other nations should, if it has any reason to fear that this contract may at some point prove to be to its disadvantage, require the precise stipulation in the contract of the conditions under which secession is possible.
Conclusions
A constitutional convention is an episodic event that typically occurs following a revolution, defeat in a war, collapse of a dictatorial regime, or paralysis of a democratically elected one. It takes place immediately following, during, or in anticipation of a national crisis. It is designed by the leader(s) of the revolution, remnants of the collapsed or collapsing regime, a parliament elected under the rules of the old constitution or an improvised new one. «The contract that joins all citizens» is often handed down to them by a group whose legitimacy as representatives of the people may be questionable, or whose motives may be suspect, or whose attention was occupied by pressing short-term matters. It is no wonder that so few constitutions seem to play an important role in the political life of the societies they supposedly shape.
For a constitution to play this role its rules must be honored by citizens and public officials, its values shared by the community. An article in the constitution protecting «free speech» implies more than a legal ground upon which to challenge a town ordinance banning a book. It implies a shared value in the community for toleration with respect to the views, writings, films, etc., of those whose opinions and tastes are not shared. Such shared values and willingness to abide by a common set of rules is more likely to be found among those who share a common culture. But over time it can also become a part of the culture. Constitutional process and constitutional protection have become a part of the political culture of the United States, of Switzerland, increasingly so of Germany, and in a few other countries.
There is no better time to start the development of citizen awareness and allegiance to a constitution than at the time it is drafted. The more citizens are involved in the selection of delegates to the constitutional convention, as observers to its proceedings, as actual participants, as ratifies of the final product, the more the constitution will reflect their ideas and values, and the more likely it is to be a positive and effective political institution. There is no single optimal blueprint for a constitutional convention, just as there is no optimal blueprint for the constitution itself. But there are better and worse alternatives.
Notes
1. The question arises here, of course, as to how the rules governing the constitutional convention are chosen. Similarly, if the constitution will define citizenship, how can one know before it is written who is eligible to vote for delegates? Infinite regress problems of this type are inherently insoluble, and I shall not take space grappling with them. The presumption is that the likely composition of the citizenry is known to all prior to the writing of the constitution. The first item on the convention's agenda must be to determine the rules under which it will operate, and at this time it can add experts to its membership, conceivably even giving them a vote. Bruce Ackerman (1991, Chap. 7) makes the point that constitutional conventions are almost inherently illegitimate in that they occur in environments in which legitimate political institutions do not exist, or the legitimacy of extant institutions is at question.
2. These people would be associated with particular parties and ideologies. The fact that they were no longer holding office, suggesting that they may be past their active professional life, would tend to produce a more detached perspective on constitutional issues, the kind of perspective needed to reach a consensus.
3. Thus, a person randomly chosen as a delegate would be asked to take a full year out of her life to participate in the constitutional convention. The hardship of this could be compensated for financially. In a country in which democracy and citizenship were prized, most would presumably relish the opportunity to participate in a national event that occurs only once, twice, or at most three times in their lifetime-the chartering of their country's political future.
4. The original selection of parliaments in Eastern Europe after 1989 did not always result in all parliamentary seats being filled by popular vote.
5. In Philadelphia, Delegate Gorham of Massachusetts, arguing for ratification of the U.S. Constitution by separately elected state conventions, clearly recognized the relative advantage of
«« Ïðåä. | ÎÃËÀÂËÅÍÈÅ
| 
