I Introduction

My concern with majority rule¹ and individual rights was spurred by recent developments in Eastern Europe and the former Soviet Union. In this region, majority rule is being adopted across the board. At the same time, individual rights have a precarious existence. To exaggerate somewhat, there has been a shift from the despotism of the Party to the despotism of the majority, both inimical to the protection of minority rights. Although there has been progress of a sort, since the Party did not care for the rights of the majority either, the achievements are decidedly limited. In most countries, constitutional democracy is still in the future.

This story has a precedent, or rather several. In England after 1648, in the United States after 1776, and in France after 1789, the abolition of a despotic regime gave rise to untrammeled majority rule, only to be followed some decades later by a regime subject to constitutional constraints². I shall not discuss the case of England, where the third stage took the form of a constitutional monarchy rather than a democracy, but focus instead on the French and the American experiences. In particular, I have found the debates at the Federal Convention in Philadelphia in 1787 and the Assemblée Constituante in Paris in 1789-91 very useful in illuminating the dangers of majority rule³. My strategy in this paper, therefore, is first to use these historical precedents to delineate the range of majoritarian problems and counter-majoritarian solutions, and then to look at some implications for Eastern Europe.

I shall proceed as follows. Section II makes a brief argument for the view that decision by majority vote is the ultimate criterion in any democracy, even in constitutional ones. Section III discusses some of the ways in which majority rule can infringe on individual rights. Section IV introduces four solutions to this conflict: constitutional entrenchment, judicial review, separation of powers, checks and balances. In Sections V through VIII, these solutions are considered separately in more detail. Their undesirable side effects are also canvassed. In these sections I draw extensively, but not exclusively, on materials from the two eighteenth-century debates. I conclude in Section IX with a survey of constitutional developments in Eastern Europe.

II Arguments for Majority Rule

At an abstract level, one can offer axiomatic arguments for majority rule⁴. Thus majority voting is the only system of preference aggregation that satisfies the conditions of anonymity (the outcome should not depend on the naming of the preference holders), neutrality (the outcome should not depend on the naming of the alternatives), positive responsiveness (a condition related to that of Pareto optimality), and universal domain (the aggregation mechanism should work for all possible combinations of individual preferences). At a deeper level, however, these conditions themselves are in need of justification. Consider in particular the crucial notion of anonymity. In predemocratic political systems, the idea that everybody's preferences are on a par as inputs to the social decision-making process would be seen as ludicrous.

How, then, can one argue for the condition of anonymity? There has never been a lack of groups claiming a privileged status. The rich, landed property owners, the old, the educated, the intelligent, the nobility, members of the Aryan race or of other ethnic groups, believers of some given religion, and the male half of society have all claimed to be inherently superior to their complements. None of these groups will accept the condition of anonymity. A key to majority rule is found, however, in the very multiplicity of these privilege-claiming groups. In the presence of many different groups who compete on the basis of their innate quality, only quantity can emerge as a peaceful focal-point solution⁵. Marx once observed that the only peaceful way to resolve the conflict between two royal pretenders is to have a republic⁶. In the struggle over which tribe in ex-colonial countries is to impose its language as the official one, the only solution acceptable to all has often been to choose the language of the former colonial power. Majority decision is similar to these formal, second-best solutions. Although people are not equal, they have to be treated as if they were.

What I have just said does not amount to an argument for anonymity or, to use the more familiar word, equality. It yields at best an explanation of why the idea of equality was irresistible or, more precisely, why from a certain time onward the only practical choice was between repression and equality⁷. Among the various positive arguments for majority rule⁸ I shall (for reasons made clear in the next paragraph) limit myself to its close link to utilitarianism. If more people prefer x to y, then the choice of x is likely to yield more aggregate welfare than would be realized if y were chosen. Although it is easy to think of counterexamples, because preferences can differ in intensity, the general correlation is not invalidated. Similarly, the effect of the Condorcet paradox of cyclical majorities is to weaken the correlation, not to eliminate it.

The link between majority rule and utilitarianism is confirmed by the fact that they have the same opponent: the defender of individual rights. There is a large literature on the relation between utility and rights⁹. The relation between majority rule and individual rights – the topic of the present paper – has also been the topic of an extensive legal literature. (These two bodies of writing are, however, rarely related to each other.) In two closely parallel arguments, defenders of individual rights have argued that they trump, respectively, utility maximization and majority rule¹⁰. To the greatest good for the greatest number and the rule of the many over the few, they oppose respect and concern for the individual. The connection between the two doctrines is undermined, however, if one reason why majority rule has to be constrained by rights is that the majority in the heat of passion may fail to perceive what is in its true interest. In that case, rights are needed to promote aggregate welfare and majority rule becomes the enemy of utilitarianism rather than its natural ally. We shall see, however, that there are other reasons to fear majority rule that do not turn on this argument.

I conclude this section with two remarks that should be kept in mind in what follows. Although one may believe that majority rule needs to be limited and constrained in various ways, these limits and constraints can ultimately have no other normative foundation than a simple majority decision. Consider the ideal case of a constituent assembly operating in a complete historical and social vacuum, for example, a group of settlers writing a constitution for their new country. Although the assembly may decide that a qualified majority shall be required to change the constitution, that decision itself must be taken by a simple majority. If one required a qualified majority at the constitutional convention, two problems arise. First, the assembly might not be able to produce a constitution at all. In constitutional amendments, the existing document serves as the status quo that remains in force when a proposed amendment fails, but in a creation ex nihilo there is no status quo that can serve as fallback position. Second, and more important, the decision to use a qualified majority would itself have to be made by a simple majority, to avoid an infinite regress¹¹. Although the relevance of this remark is attenuated in actual instances of constitution making, which always take place in a context that imposes or suggests a structure on the process, the fundamental logic of constitution making remains that of a simple majority deciding that a simple majority may not be the best way to decide some issues.

A second, related remark concerns the effects of majoritarian decision making at a constitutional convention when that assembly also serves as an ordinary legislature, as was the case at the Assemblée Constituante. That combination obviously may be undesirable. A main task of a constituent assembly is to strike the proper balance of power between the legislative and the executive branches of government. To assign that task to an assembly that also serves as a legislative body would be to ask it to act as judge in its own cause. A constitution written by a legislative assembly might be expected to give excessive powers to the legislature. In the abstract, this problem could be solved by means similar to the ones used in legislative bodies, by checks and balances. A royal veto over the constitution might, for instance, have kept the legislative tendency to self-aggrandizement in check. However, even those who argued for extensive checks and balances in the constitution, did not believe in a similar system for deciding on the constitution. Mounier, for instance, argued that the strong unicameral assembly necessary to create the constitution, would be inappropriate for ordinary lawmaking. Similarly, Mirabeau argued that the King should have a veto in the constitution, but not over the constitution itself. Summarizing both points, Clermont-Tonnerre observed that the "three-headed hydra" – king, first chamber, and second chamber – that the constitution should create could not itself have created a constitution. To get around the problem of self-interested framers, the Assemblée Constituante adopted another solution, voting its members ineligible to the first ordinary legislature. Robespierre¹², in his first great speech, won the assembly for this "self-denying ordinance"¹³. Although sometimes viewed by posterity as a disastrous piece of populist overkill¹⁴, Robespierre's solution did correspond to a genuine difficulty.

III How Majority Rule May Infringe on Individual Rights

For the purposes of this paper, I need not discuss what rights are, nor which rights individuals have. It is sufficient to consider principles claimed as rights, and how they might be endangered by majority rule. I shall not consider the inappropriately named "positive rights" that entitle the individual to have part of the social product spent on activities that directly enhance his material welfare, such as the right to work, to welfare, or to a clean environment. Instead I shall limit myself to the traditional rights, such as civil liberties, political liberties, property rights, and the freedom of contract.

A tripartite division of rights will prove convenient. First, there are the rights that enable real and equal political participation, notably the right to vote and freedom of speech and association. Second, there are rights that promote the rule of law, such as a prohibition of bills of attainder, a ban on retroactive legislation or retroactive taxation, a guarantee of full or fair compensation for confiscation of property, a ban on arbitrary search and seizure, and the right to a fair trial. Under this category, I also include the right to be able to count on the laws being reasonably stable¹⁵. Third, there are rights that protect religious and ethnic groups, by guaranteeing freedom of worship or the right to use and be educated in one's own language. This particular way of classifying rights has no intrinsic merit, except that it is usefully correlated with ways in which – and motives for which – majority rule might possibly infringe on the exercise and the value of rights.

First, a majority government will always be tempted to manipulate political rights to increase its chances of reelection. If it is free to change the timing of the election, it may choose a moment when economic conjunctures are favorable¹⁶. If electoral district boundaries have to be redrawn because of population changes, the government may try to do so to its advantage. If the majority is free to change the electoral system – for example, proportional representation versus single-member districts – it may exploit this possibility for strategic purposes. If voters have to be registered before they can vote, the government may have an incentive to make registration more difficult to disenfranchise de facto some of those who would have voted for the opposition. In countries with state-owned radio and television, the government may give itself disproportionate time. We may note for later reference that in these cases the attack on rights comes from the majority in parliament, not from the majority in the population. The danger is precisely that the parliamentary majority may have means at its disposal to prevent the popular majority from putting a new government in place.

Second, a majority may set aside the rule of law under the sway of a standing interest or a momentary passion. This was Madison's main worry. "In all cases where a majority are united by a common interest or passion, the rights of the minority are in danger"¹⁷. This distinction between interest and passion is crucial¹⁸. If the poor or relatively propertyless form a majority, their interest might induce them to enact laws that are contrary to the rights of property, by creating paper money, legislating debtor relief, and so on. A quite different danger arises if the majority is animated by a sudden passion that makes it deaf to the demands of the rule of law. The impulse may originate either in a majority in parliament or in a popular majority that manages to impose its will on parliament by nonelectoral methods. The risk of such legislation being passed is especially great in wartime and other emergency situations, a famous case being the internment of Japanese-Americans during World War II.

Third, a majority may set aside the rights of an ethnic or religious minority under the sway of what one might call a standing passion. In earlier centuries, religious fanaticism has been the mainspring of this form of majoritarian domination. Today, ethnic hatred, sometimes combined with religious differences, is proving a horribly potent source of oppression.

Two distinctions are implicit in what I have said. On the one hand, we have to identify the relevant majority, and notably whether it is parliamentary or popular. On the other hand we have to identify the motives which move the members of the majority to infringe on the rights of the minority. Here I have discussed three cases: standing interests, standing passions, and momentary passions. Although all six combinations of actors and motives might be relevant, I shall limit myself to five.

First, there is the case of a parliamentary majority that acts to preserve itself as a majority, by the various procedural stratagems mentioned earlier, or to promote such other interests as it might have. Madison, for instance, noted that "It had often happened that men who had acquired landed property on credit, got into the Legislatures with a view of promoting an unjust protection agst. their Creditors"¹⁹.

Second, there is the case of a parliamentary majority being swayed by the (standing) passion of amour-propre, that is, vanity or self-love. Although virtually absent at the Federal Convention, the fear that political agents might act on such motives was often expressed in the Assemblée Constituante. Bergasse argued, for instance, that a suspensive veto of the King would not have the intended effect of making the assembly reconsider its vote, because its amour-propre would prevent it from backing down²⁰.

Third, there is the case of a popular majority acting (through its representatives) to further its economic interest. A special case is that in which this interest is defined in terms of present value of future income, discounted by some positive factor. If the discount rate is high, members of the majority might find it in their interest to take confiscatory measures against property owners, even if they know that in the long run they or their descendants would be better off respecting property.

Fourth, there is the case of a popular majority acting (through its representatives) under a sudden impulse, a momentary passion. The founders in Philadelphia and the constituants in Paris constantly referred to this danger. In Philadelphia, we find references to "the turbulence and follies of democracy", "the fury of democracy", "the popular passions [which] spread like wild fire, and become irresistable", "fickleness and passion", "the turbulency and violence of unruly passion", and to the "precipitation, changeableness, and excesses of the first branch". In Paris, Lally-Tollendal referred to the assembly being "entrainée par l'éloquence, séduite par des sophismes, égarée par des intrigues, enflammée par des passions qu'on lui fait partager, emportée par des mouvements soudains qu'on lui communique, arrêtée par des terreurs qu'on lui inspire". Others warned against "les prestiges de l'éloquence, l'effervescence de l'enthousiasme", "les causes d'erreur, de précipitation ou de séduction oratoire", or "l'erreur, la précipitation, l'ambition"²¹.

Fifth, there is the case of a popular majority acting (through its representatives) from a standing, permanent passion. A perusal of the quotations in the previous paragraph brings out the predominance of terms such as "sudden", "fickle", "unruly", "precipitation", "changeableness", and the like. By contrast, there were few references in the two eighteenth-century assemblies to more permanent passions and prejudices that might fashion the will of the majority. In Philadelphia, for instance, nobody mentioned racism or religious sectarianism as potential threats to individual rights²². In the late twentieth century, these problems of ethnicity and religion may prove to be the outstanding danger of majority rule. In addition to the problems in Eastern Europe that I discuss in the concluding section, the specter of Islamic majoritarianism in Algeria offers a striking example.

I should add a nuance to this somewhat mechanical presentation. Although I believe that the distinction between interest and passion is of fundamental analytical importance, they often go together in practical politics. On the one hand, passion often makes us believe that something is in our interest which really is not. What Tocqueville called "the democratic sentiment of envy" may dress itself up as a theory that the rich, if not restricted, will use their wealth to subvert the polity. On the other hand, an interest, to be effective in politics, often has to take on the garb of passion. Norms of equality and other social norms can impart a passionate tone to claims that otherwise might be seen as mere expressions of self-interest²³. In practice, therefore, people will not acknowledge a conflict between interest and passion. But this need not prevent the outside observer from being able to identify one of them as the causally efficacious motive, and the other as its dupe or handmaiden.

IV Countermajoritarian Devices: an Overview

In the following sections I discuss what I believe to be the four main countermajoritarian devices used in modern societies. In this section I offer a broader perspective, by attempting to relate the four devices to each other. First, however, I shall briefly comment on a proposal that, although not immediately relevant for current events, was an important background element for the eighteenth-century assemblies²⁴. This is the view that when the polity is too large for direct democracy one can reduce the dangers of representative democracy by a system of bound mandates, perhaps combined with the possibility of recalling delegates at any time if they exceed their briefs. In both Philadelphia and Paris, there was general agreement that this system was undesirable. It would reduce democracy to a mere system of preference aggregation, and leave no room for the transformation of preferences through rational deliberation²⁵. For many writers, from Aristotle to the present, majority rule is in fact justified by the opportunity it offers for the exchange of ideas and discussion²⁶. In a small polity, this ideal can be realized without creating a legislative body with interests of its own. In a large polity, which requires a representative democracy, the latter danger is an unavoidable concomitant of any attempt to realize the deliberative ideal. Independence inevitably cuts both ways. At the same time, direct democracy is more vulnerable to violent popular passions. Whereas both the town meeting and an assembly of (unbound) representatives allow for discussion, the former is vulnerable to the problem of passionate popular majorities and the latter to the problem of self-interested legislative majorities. A representative system constrained by bound mandates may limit both majoritarian dangers, but at the cost of giving up the benefits of deliberation²⁷. If the only purpose of representation was to protect individual rights, this system might be optimal. But an assembly is also created to get things done, to work out compromises, and to make good decisions.

The four devices I shall discuss are, to repeat, constitutionalism, judicial review, separation of powers, and checks and balances. In some political systems, these form a tightly knit whole. Judicial review, separation of powers, and checks and balances are all written into the constitution. Judicial review is a mechanism to interpret and enforce the constitution. In doing so, it also serves to prevent usurpation of power by the other organs of state. It can only perform that function, however, if it is reasonably independent of those organs. More generally, checks and balances presupposes some separation of powers: If A is to act as a check on B, it must have some degree of independence from B.

In other systems, these elements are decoupled from each other to a larger extent. England, for instance, does not even have a written constitution, and yet there are limits on majority rule that form part of what has been called the unwritten constitution²⁸. Judicial review may go beyond the constitution and consider rights not specifically enumerated in that document. Also, the institution of judicial review may not itself be specifically mentioned in the constitution. This is the case in the United States. Although checks and balances presupposes a separation of powers, the converse is not true. The French constitution of 1789, which was "based on an extreme version of the doctrine of the separation of powers"²⁰, had no checks and balances beyond the suspensive veto of the King. In the constitution of the Fifth French Republic, the Conseil Constitutionnel was originally created to strengthen the executive against the legislature. It was only fifteen years later that it was made into (and made itself into) an institution for independent judicial review³⁰.

V Constitutionalism as a Constraint on Majority Rule

Majority rule can be restrained by the constitution, both directly and indirectly. In this section I discuss the direct influence that derives from the combination of the constitutionalization of certain laws and procedures that make it difficult to amend the constitution. In later sections I discuss other restraining devices that may or may not be explicitly mentioned in the constitution but that, if they are, operate in a different manner.

A constitution can affect behavior by acting on the desire of the majority to change the law or on its opportunities to do so³¹. The first mechanism operates by making the process of constitutional amendment very slow and time-consuming, so that impulsive passions can cool down and reason (or interest!) reinstate itself. The second operates by requiring qualified majorities for changing the constitution or, at the limit, declaring some clauses unamendable. Some constitutions (such as the Norwegian one) impose both qualified majorities and delays; others (such as the Swedish one) require only delays³². Still others (such as the Hungarian one) require only qualified majorities; New Zealand appears to be unique in that "only ordinary legislative efforts are required to supplement, modify or repeal the Constitution"³³.

Delaying devices are designed to counteract sudden impulses and momentary passions among the majority. Qualified majorities are intended to protect individual rights against the standing interests and passions of the majority. It is easily seen that these two mechanisms must differ fundamentally in their mode of adoption. If we focus on delays, the following description is indeed appropriate: "Constitutions are chains with which men bind themselves in their sane moments that they may not die by a suicidal hand in the day of their frenzy"³⁴. It is in the straightforward interest of the majority to prevent itself from making rash decisions under the sway of passion. However, the use of qualified majorities cannot be explained or justified by the idea that Peter when sober acts to bind Peter when drunk³⁵. If a majority among the founders has a standing interest on some particular issue, that interest will not induce them to set it aside. If they are moved by religious fanaticism, this is a passion they embrace rather than fear. In the eighteenth century the question did not have the importance it assumes today. The founders as a whole, and a fortiori a majority among them, represented a minority elite within the population that could impose its views on the rest. In the United States, for instance, today's majority is bound by a founding minority.

As with the other countermajoritarian devices discussed below, constitutionalism has a potential for creating problems as well as solving them. One should keep in mind a dictum of constitutional lawyers, due to Justice Robert Jackson: The constitution is not a suicide pact. It must be possible to unbind oneself in an emergency. Society must not be confined too tightly³⁶. In the debates over the constitutional ban on paper money at the Federal Convention, George Mason said that "Though he had a mortal hatred to paper money, yet as he could not foresee all emergencies, he was unwilling to tie the hands of the Legislature. He observed that the late war could not have been carried on, had such a prohibition existed"³⁷. The ensuing dilemma is very tight. On the one hand, one might wish for the constitution to allow for unforeseen and unforeseeable emergencies. On the other hand, some of the occasions that will be claimed to have emergency status will be the very situations in which the constitution was supposed to act as a protection. An alcoholic will always be able to specify some way in which today is special and exceptional³⁸.

[ Continued in Issue No. 3-4/1996 • 1-2/1997 ]

Webmaster: Slobodan Markovic (twiddle@eunet.yu)

All rights reserved. Belgrade Circle Journal encourages the reproduction of material appearing on its pages, provided that the source and the author are cited, except in cases where this would constitute violation of copyright held by other organisations or individuals.