Less than a quarter-century ago, democracy appeared to be confined, with a few exceptions, to North America and Western Europe. These nations had advanced industrial economies, sizable middle classes, and high literacy rates-factors that many political scientists regarded as prerequisites for successful democracy. They were home not only to free and competitive multiparty elections but also to the rule of law and the protection of individual liberties. In short, they were what had come to be called "liberal democracies."
In the rest of the world, by contrast, most countries were neither liberal nor democratic. They were ruled by a variety of dictatorships-military, single-party, revolutionary, Marxist-Leninist-that rejected free, multiparty elections (in practice, if not always in principle). By the early 1990s, however, this situation had changed dramatically, as an astonishing number of autocratic regimes around the world fell from power. They were generally succeeded by regimes that at least aspired to be democratic, giving rise to the phenomenon that Samuel P. Huntington termed the "third wave" of democratization. Today, well over a hundred countries, in every continent in the world, can plausibly claim to have freely elected governments.
Outside of Africa, few of these aspiring new democracies have suffered outright reversions to authoritarianism. But many, even among those that hold unambiguously free and fair elections, fall short of providing the protection of individual liberties and adherence to the rule of law commonly found in the long-established democracies. As Larry Diamond has put it, many of the new regimes are "electoral democracies" but not "liberal democracies." Citing Diamond's distinction, Huntington has argued that the introduction of elections in non-Western societies may often lead to victory by antiliberal forces. And Fareed Zakaria has contended that the promotion of elections around the world has been responsible for "the rise of illiberal democracy"-that is, of freely elected governments that fail to safeguard basic liberties. "Constitutional liberalism," Zakaria argues, "is theoretically different and historically distinct from democracy . . . . Today the two strands of liberal democracy, interwoven in the Western political fabric, are coming apart in the rest of the world. Democracy is flourishing; constitutional liberalism is not." Drawing upon this distinction, Zakaria recommends that Western policymakers not only increase their efforts to foster constitutional liberalism but diminish their support for elections, and suggests that "liberal autocracies" are preferable to illiberal democracies.
The basic distinction made by all these authors is both valid and important. Liberal democracy-which is what most people mean today when they speak of democracy-is indeed an interweaving of two different elements, one democratic in a stricter sense and the other liberal. As its etymological derivation suggests, the most basic meaning of the word "democracy" is the rule of the people. As the rule of the many, it is distinguished from monarchy (the rule of one person), aristocracy (the rule of the best), and oligarchy (the rule of the few). In the modern world, where the sheer size of states has rendered impossible the direct democracy once practiced by some ancient republics, the election of legislative representatives and other public officials is the chief mechanism by which the people exercise their rule. Today it is further presumed that democracy implies virtually universal adult suffrage and eligibility to run for office. Elections, then, are regarded as embodying the popular or majoritarian aspect of contemporary liberal democracy.
The word "liberal" in the phrase liberal democracy refers not to the matter of who rules but to the matter of how that rule is exercised. Above all, it implies that government is limited in its powers and its modes of acting. It is limited first by the rule of law, and especially by a fundamental law or constitution, but ultimately it is limited by the rights of the individual. The idea of natural or inalienable rights, which today are most commonly called "human rights," originated with liberalism. The primacy of individual rights means that the protection of the private sphere, along with the plurality and diversity of ends that people seek in their pursuit of happiness, is a key element of a liberal political order.
The fact that democracy and liberalism are not inseparably linked is proven by the historical existence both of nonliberal democracies and of liberal nondemocracies. The democracies of the ancient world, although their citizens were incomparably more involved in governing themselves than we are today, did not provide freedom of speech or religion, protection of private property, or constitutional government. On the other side, the birthplace of liberalism, modern England, retained a highly restricted franchise well into the nineteenth century. As Zakaria points out, England offers the classic example of democratization by a gradual extension of suffrage well after the essential institutions of constitutional liberalism were already in place. In our own time, Zakaria offers Hong Kong under British colonial rule as an example of a flourishing of liberalism in the absence of democracy.
ALL MEN ARE CREATED EQUAL
Although "unpacking" the component elements of modern liberal democracy is a crucial first step toward comprehending its character, overstating the disjunction between liberalism and democracy can easily lead to new misunderstanding. While many new electoral democracies fall short of liberalism, on the whole, countries that hold free elections are overwhelmingly more liberal than those that do not, and countries that protect civil liberties are overwhelmingly more likely to hold free elections than those that do not. This is not simply an accident. It is the result of powerful intrinsic links between electoral democracy and a liberal order.
Some of these links are immediately apparent. Starting from the democratic side, elections would seem to require the guarantee of certain civil liberties-the freedoms of speech, association, and assembly-if they are to be genuinely free and fair. Thus even minimalist definitions of democracy offered by political scientists usually include a stipulation that such liberties must be maintained at least to the extent necessary to make possible open electoral competition. If we begin instead with the human rights mandated by the liberal tradition, these are generally held today to include some kind of right to electoral participation. Thus Article 21 of the U.N. Universal Declaration of Human Rights states: "Everyone has the right to take part in the government of his country, directly or through freely chosen representatives . . . The will of the people shall be the basis of the authority of the government; this will shall be expressed in periodic and genuine elections which shall be by universal and equal suffrage and shall be held by secret vote or by equivalent free voting procedures." One may regard this as a formal or even merely definitional link between liberalism and electoral democracy, but it points to a more profound kinship.
For the political doctrine at the source of liberalism also contains a deeply egalitarian and majoritarian dimension. This is the doctrine that all legitimate political power is derived from the consent of individuals, who are by nature not only free but equal. In the opening pages of his Second Treatise of Government, John Locke states that men are naturally in "a state of perfect freedom," which is "a state also of equality, wherein all the power and jurisdiction is reciprocal, no one having more than another: there being nothing more evident, than that Creatures of the same species and rank promiscuously born to all the same advantages of Nature, and the use of the same faculties, should be equal one amongst another without Subordination or Subjection." The essential point is that no man has a natural claim to rule over another, and its clear corollary is that the rule of man over man can be justified only on the basis of a mutual agreement or "compact."
Now it is true that neither Locke nor his immediate successors concluded from this that democracy was the only legitimate form of government. For while they held that the consent of all is essential to the original compact that forms a political community, they also contended that the political community is free to decide where it chooses to bestow legislative power-whether it is in a democracy, an oligarchy, a monarchy, or a mix, as it was among the King, Lords, and Commons in England. Liberalism did not originally insist on democracy as a form of government, but it unequivocally insisted upon the ultimate sovereignty of the people. Thus Locke argues that if the legislature is dissolved or violates its trust, the power to institute a new one reverts to the majority of the people.
In order to grasp the distinctive character of liberal egalitarianism, it is necessary to appreciate how different modern liberal democracy is from the premodern (and truly illiberal) democracy of the ancient city. Reliance on elected representation in the legislature, the key political institution of modern liberal democracy, was understood by its proponents as a decisive departure from ancient democracy. The authors of The Federalist frequently contrast two very different kinds of "popular government." They write in favor of a "republic" ("a government in which the scheme of representation takes place"), which they argue need not be subject to the infirmities of "a pure democracy" ("a society consisting of a small number of citizens who assemble and administer the government in person," Federalist 10). In pure or direct democracies, they contend, "there is nothing to check the inducements to sacrifice the weaker party or an obnoxious individual," and therefore they "have ever been found incompatible with personal security or the rights of property." Later, in Federalist 63, acknowledging that the principle of representation was not unknown to the ancients, Madison states: "The true distinction between [ancient democracies] and the American governments lies in the total exclusion of the people in their collective capacity, from any share in the latter, and not in the total exclusion of the representatives of the people from the administration of the former." (Italics in original). In short, modern liberal democracy from the outset was inclined to minimize the direct political role of the people. In this sense, Zakaria is on solid ground in stressing the anti-majoritarian aspects of liberalism.
In part, of course, the substitution of representative government for direct democracy was justified by the larger size of modern states, which made it impractical for the whole people to assemble. But this very fact had led thinkers like Montesquieu and Rousseau to conclude that democratic or republican government was possible only in a small state, and Rousseau to assert that "the moment that a people gives itself representatives, it is no longer free." There was, however, another ground used to justify representative government. In Madison's words, it "would refine and enlarge the public views by passing them through the medium of a chosen body of citizens, whose wisdom may best discern the true interest of their country and whose patriotism and love of justice will be least likely to sacrifice it to temporary or partial considerations." In other words, elected representatives are expected to be superior to the average citizen. In the ancient democracies, by contrast, most public officials were chosen by lot. In The Politics, Aristotle characterizes lot as the democratic mode of choosing officials, and election as the oligarchic mode. Montesquieu reiterates this judgment, adding, "The suffrage by lot is a method of electing that offends no one, but animates each citizen with the pleasing hope of serving his country." Where elections are used instead, those chosen tend to be richer, better educated, and more talented than most of their fellow citizens. In this light, representative or electoral democracy, besides largely eliminating the people from direct participation in self-government, also seems to constitute an aristocratic deviation from political equality.
BY AND FOR THE PEOPLE
Yet there is another sense in which modern liberal, representative democracy is much more egalitarian than was ancient democracy. In the latter, the citizens entitled to participate in public affairs invariably represented a relatively small percentage of the overall population. Not only were large numbers of slaves and resident aliens excluded, but women had no role in political affairs. Preliberal democracy, the direct democracy of the ancient city, was not based on any concept of the fundamental, natural equality of all human beings. It is true, of course, that modern representative government for a long time excluded the poor and all women from political participation, and in the United States even coexisted with slavery. But it is no less true that these kinds of exclusions were always in tension with the underlying principle of liberalism-namely, that all human beings are by nature free and equal. The historical development of this principle inevitably transformed liberalism into liberal democracy.
It is one thing to claim that the majority of people in a traditional and hierarchical society have somehow given their tacit consent to a political arrangement in which they are excluded from having any voice. Popular sentiment in seventeenth-century England, if there had been a way of measuring it, might well have approved of a monarchical political system. But as the principle that all men are created equal gained currency, and as the educational and economic situation of the common people continued to improve, it was only to be expected that some of them would begin to demand the vote. And once they began to do so, how could it any longer be claimed that they consented to a political order in which they had no say? Popular sovereignty without popular government may be coherent in theory and even sustainable in practice for a time. Over the long run, however, popular sovereignty can hardly fail to lead to popular government.
Thus it is not surprising that throughout the Western world, liberal, constitutional regimes became more and more democratic during the nineteenth and twentieth centuries. The share of legislative power wielded by monarchs or unelected bodies receded until it had virtually disappeared. At the same time, suffrage was gradually broadened. Property qualifications and exclusions on the basis of race or sex were eliminated, to the point where "universal and equal suffrage" was endorsed by the world community in 1948 as a human right.
The moral grounds for extending suffrage are succinctly stated by John Stuart Mill in his Considerations on Representative Government, published in 1861. "It is a personal injustice," Mill argues, "to withhold from anyone, unless for the prevention of greater evils, the ordinary privilege of having his voice reckoned in the disposal of affairs in which he has the same interest as other people . . . No arrangement of the suffrage, therefore, can be permanently satisfactory in which any person or class is peremptorily excluded, in which the electoral privilege is not open to all persons of full age who desire to obtain it." On these grounds Mill also argues for the extension of the franchise to women. Yet this does not prevent him from arguing against granting the vote to illiterates and to recipients of parish relief (i.e., welfare); he also proposes that multiple votes be allotted to the educated and professional classes. Today, such departures from universality and equality in the allocation of the franchise seem shockingly "elitist." No arguments for "the prevention of greater evils" are reckoned as sufficiently powerful to overbalance the injustice of denying any citizen an equal vote.
MAKING DEMOCRACY WORK
There is another respect in which Mill's Representative Government is repugnant to contemporary sensibilities-namely, its justification of colonialism. For Mill, representative government "is the ideal type of a perfect government," but it is not applicable under all social conditions. In particular, it is ill suited to "barbarous" or "backward" peoples, who are likely to need some form of monarchical or (preferably) external rule to bring them toward the state of civilization in which they might become fit for representative government.
In part, Mill's argument in favor of colonialism is grounded in a dubious doctrine of historical progress (or of "modernization," as we would say today). Yet there is another basis for Mill's contention that representative government is not applicable under all conditions that is not easily dismissed. As he puts it, "representative, like any other government, must be unsuitable in any case in which it cannot permanently subsist." If people do not value representative government, if they are unwilling to defend it, if they are unable to do what it requires, then they will not be able to maintain it. Thus it would be vain to expect that it would serve them well.
The concern with making democracy able to maintain itself, with training and spurring the people to do what is needed to make democracy work, is certainly not outdated. It is at the heart of most programs of "democracy assistance" now being provided to new democracies by Western governments, international and regional organizations, and nongovernmental organizations alike. It is at the root of the central concern today of political scientists who study new democracies-the problem of consolidation, or how to bring a democratic regime to the point where its breakdown becomes extremely unlikely. And it explains the widespread attention to issues of citizenship and civil society today, not only in new democracies but in long-established ones as well. These concerns reflect the irreducible fact that making self-government work is not easy. A democratic government can be given to any people, but not every people can maintain it. But what is to be done in the case of a people that is not, at least for the time being, capable of making democracy work? Mill's answer to this question was colonial rule. What is ours? That is the question implicitly raised by Zakaria's article.
The difficulty in answering it points to an acute tension within the modern democratic tradition between the liberal doctrine of just or legitimate government and the practical requirements of popular government. (In The Social Contract, Rousseau says that "all legitimate government is republican." But later in the same work he says that "freedom is not a fruit of every climate, and it is not therefore within the capacity of every people.") The principle that all men are born free and equal, and that no one has a right to rule them without their consent, has now swept the world. As I have argued above, this has inevitably come to be understood as meaning that they cannot be ruled without their clearly expressed consent, in the form of an election. Yet the experience of past ages and of many lands suggests that this principle cannot be effectively put into practice everywhere and immediately. The failure in the 1960s of so many of the democracies bequeathed by the departing colonial powers once again demonstrated the fact that under certain conditions democracy is unlikely to endure. But if democratic government is required everywhere in principle, what course can a good liberal democrat follow where it appears unable to work in practice? This conundrum largely accounts for the alternating cycles of euphoria and despair about the prospects for the spread of liberal democracy.
How does Zakaria suggest that this dilemma be resolved? He contends, first, that constitutionalism, the rule of law, and the protection of individual liberty are more essential than representative government. Accordingly, he recommends that, rather than encouraging the introduction of elections in many developing countries, Western policy should favor the establishment of "liberal autocracy." As noted above, the prime example of liberal autocracy that he presents is nineteenth-century Europe, where the introduction of constitutional liberalism by monarchical governments preceded democratization. It has often been remarked that the sequence of first liberal constitutionalism, then gradual democratization, can have advantages in accustoming people to the requirements of self-government. But is this a practical strategy today?
During the nineteenth and early twentieth centuries, democratization proceeded in a context in which more traditional principles of social hierarchy still had a considerable hold over the popular imagination. The idea of equality had not been fully accepted as the preeminent principle of political legitimacy. Monarchy and aristocracy still prevailed in most of Europe, so that even a limited legislative role for an assembly elected with a restricted suffrage could seem like progress toward popular government. Today the situation is dramatically different. There are only a few countries-principally Islamic monarchies-in which anything like traditional rule still holds sway. In these cases, perhaps the nineteenth-century European model can to some extent be emulated. Elsewhere, existing autocracies-or the regimes that aspiring democracies have replaced-are generally ideological rather than traditional regimes and espouse some kind of egalitarian doctrine of their own. In a post-communist or formerly one-party socialist regime, what principle could be accepted as a basis for restricting suffrage? And what legitimate mechanism other than election could be used for deciding who will rule?
The only example in the contemporary world of liberal autocracy that Zakaria explicitly cites is British-ruled Hong Kong. Yet he certainly does not seem prepared to recommend a revival of colonialism. Earlier in this decade, there was a flurry of discussion of the problem of "failed states"-former client states of the superpowers during the Cold War that threatened to collapse once the support of their patron had been withdrawn. Amid the talk of a new world order, there seemed to be some inclination to have the "international community" intervene in such cases, in effect reviving something like colonial rule under the aegis of the United Nations. Whatever the merits or the feasibility of this idea, the fiasco of the U.S. attempt at political (as opposed to humanitarian) intervention in Somalia, along with the proliferation of states that might have been candidates for such costly international reconstruction operations, quickly made it clear that the political will for this kind of policy was lacking.
The practical model that Zakaria seems to have in mind is the economically successful (at least until recently) autocracies of East Asia. Yet it would surely be questionable to assert that these autocracies are genuinely constitutional or liberal, a fact that Zakaria himself seems to recognize by characterizing Indonesia, Singapore, and Malaysia not as "liberal" but only as "liberalizing" autocracies. It would be implausible indeed to claim that these states more reliably protect individual rights or have more independent and impartial judiciaries than the Latin American democracies that Zakaria describes as "illiberal." Even the Singaporeans themselves, while claiming to practice democracy, acknowledge that their regime, to quote Singapore's U.N. Ambassador, Bilahari Kausikan, "has never pretended or aspired to be liberal." Thus, despite Zakaria's talk of constitutionalism and individual rights, he seems to wind up taking the much more familiar view that authoritarian capitalist development is the most reliable road to eventual liberal democracy.
The economic achievements of these East Asian autocracies have certainly been impressive, but so have been the economic achievements of East Asian democracies, beginning with Japan. This is not the place to enter into the complex and hotly contested argument about to what extent, if at all, authoritarian rule has been responsible for Asian economic development. What is clear, however, is that in the rest of the world the overall record of autocracies in promoting economic development, let alone the growth of constitutional liberalism, has been poor. As Mill noted, the same shortcomings that make a people poorly prepared for representative government are also likely to be found in its unelected rulers. Wise and benevolent despots are the exception, not the rule.
A LOOK INSIDE THE BALLOT BOX
It was only to be expected that, as countries around the world replaced their autocratic regimes with freely elected ones, they would encounter serious difficulties in making democracy work. Self-government is indeed difficult, and holding elections is merely one step in a long and arduous process that, in the best case, will culminate in a consolidated liberal democracy. Electorates can make bad choices as well as good or (most often) mediocre ones. Demagogues can use electoral campaigns to appeal to voters' worst instincts, including ethnic or religious intolerance (although the number of new democracies in which candidates have succeeded on the basis of such appeals is far fewer than might have been expected). But in any case, how often can elections themselves be plausibly cited as the cause of problems that would not have been just as likely to persist or arise under a nonelected government? African voters, for example, may often cast their ballots along ethnic or tribal lines, but in how many African countries have dictatorial governments achieved real ethnic accommodation, rather than merely the domination of some groups by others? Most new democracies are undoubtedly confronting severe challenges, but almost none of these would be overcome by abolishing elections.
It is also true that, beyond peacefully getting rid of a bad and unpopular government (which is no small accomplishment), elections by themselves do not solve most other political problems. For this and other reasons, prudence counsels against hastily pushing elections on a fairly stable, decent, and moderate nondemocratic regime, especially in a country where the strongest opposition forces are not themselves well disposed toward liberal democracy. This, however, is a lesson that most Western governments, inherently inclined toward diplomatic caution, hardly need to be taught. In fact, their adherence to such a policy is a frequent complaint of those who accuse Western governments of being too friendly with nondemocratic governments, especially in the Arab world.
There are arguably cases where elections have made things worse, as in Angola in 1992, where Jonas Savimbi's refusal to accept his defeat in a U.N.-supervised election led to a violent escalation of that country's civil war. Yet despite some serious setbacks, most recently in Cambodia, the overall record of attempts to use internationally supervised elections as a method of conflict resolution for countries embroiled in civil strife has been surprisingly positive. This relatively recent innovation, first attempted in Nicaragua in 1990, combines peacemaking with democracy-building, but is driven primarily by the former goal. Thus elections are often held under extraordinarily difficult circumstances and at times that would not have been chosen if democracy-building were the only goal. Nonetheless, such elections have not only brought a number of bloody civil wars to a halt, but in countries like Mozambique and El Salvador have had positive political results as well. Even if such countries today are merely illiberal democracies, they are manifestly much better off than if they were still racked by civil war. Afghanistan, a country that did not undergo an electoral process and faces continuing civil war and the rule of an extremist and intolerant Islamist government, does not present a very attractive alternative model.
In more typical cases of democratic transition, where an authoritarian government either is overthrown or negotiates an agreement with domestic opposition forces on the creation of a new regime, the timing of "founding elections" can be a matter of critical importance for the success of an emerging democracy. In such cases there is room for reasonable disagreement about how soon to hold elections. Amid the devastated political landscape of the post-Mobutu Congo, for example, even those committed to trying to move the country in a democratic direction are divided about both the practicability and the desirability of conducting early elections. At the same time, it is difficult to see how dispensing with elections would lead the Kabila government to move toward "constitutional liberalism," or how such unaccountable rule would be preferable to "illiberal democracy."
IF AT FIRST YOU DON'T SUCCEED