I have been speaking thus far of voting in its pure "one person-one vote" sense, however, and often impurities are deliberately introduced. Corporate voting is by share rather than by shareholder; votes are weighted by the voter's financial interest in the outcome of the vote. We shall see that John Stuart Mill thought that in political elections abler individuals should be given additional votes. Even in modern American political elections there are eligibility requirements that, deliberately or not, prevent or discourage the most incompetent, apathetic, and irresponsible elements of the population from voting. Voters must be adults, must register to vote, must not be in jail or prison (often, must not be ex-felons as well), must be residents of the voting district, usually must be citizens. Obviously these are coarse sieves if the objective is to confine voting to intelligent, knowledgeable, civic-minded, politically mature, responsible people with a tangible stake in the outcome of the election.

Why anyone who is eligible bothers to vote is a puzzle,3 especially to economists, who emphasize self-interest as the engine of human action. The puzzle is not why voters do not invest much time in equipping themselves to vote intelligently (something that many of them would lack the intelligence or education to be able to do, no matter how much time they spent studying the issues and the candidates); obviously they have little to gain from such an investment, since political elections, other than on the most local level, are virtually never decided by a single vote. The puzzle is that people vote at all in a society in which voting is not compulsory, given the lack of instrumental value to which I have just pointed—even the 2000 Presidential election in Florida was not decided by a single vote.

Of course, many Americans who are eligible to vote do not vote. The nationwide turnout in the 2000 Presidential election was only 50 percent of the eligible voters, and it is often much lower in more obscure elections. Yet many do vote. The costs are slight. But what are the benefits? Because a single voter cannot swing the election, voting is less plausibly regarded as an activity yielding instrumental benefits than as a form of consumption, like rooting for one's alma mater at a college football game. Consistent with this suggestion, voter turnout is largely a function of how exciting the election campaign has been; the voter chooses between the candidates much as he would between competing brands of an ordinary product or service.4 Voting is also a way of constructing and signaling the voter's solidarity with like-minded people. But that is often an element of ordinary consumption too; the choice of brand may be influenced by a desire to signal taste, values, or affluence. Because the benefits of voting are small, however, even modestly high costs of voting can depress turnout significantly. The fact that turnout in U.S. elections is low by international standards is largely explained by mundane factors affecting, however slightly, the cost of voting. For example, elections are scheduled on regular workdays in the United States, rather than on holidays or weekends as in a number of other countries. Registering to vote is also more of a bother in the United States, because registration is carried out at the state level; Americans tend to move around a great deal and so have to register more frequently than the citizens of less mobile or more centralized societies.5

The popular vote is often thought to be the defining element of political democracy, but that is a modern view. The classical view, which originates in the Athenian concept of democracy 6 and contains an insight that we moderns have largely lost sight of, is that filling political offices by popular election rather than by lot conduces to aristocratic (in the Aristotelian sense) rather than democratic government. Conduces, in other words, to government by "the best" rather than by the average (as in a system of filling offices by lot) or by "the people" as a whole.7 Only representative democracy is aristocratic in this sense, however. Direct democracy—voting on issues rather than for officials—is not. Although a hallowed part of our tradition in the form of the New England meeting and still employed to a limited extent in some of our states, primarily in the form of the referendum, direct democracy is infeasible in a large, complex polity. Switzerland employs the referendum extensively—and it is exceptional even among small countries in this respect.8 Moreover, Switzerland is also and primarily a representative democracy, like California and other U.S. states that authorize referenda. In these instances direct democracy merely supplements representative democracy, providing a means of breaking cross-party coalitions of politicians.9

The aristocratic character of representative democracy is rooted in the fact that, especially if political parties are weak or nonexistent (the significance of this qualification will become clear shortly), voters will tend to pick the best candidate for each office. The best candidate—which is to say the candidate most likely to occupy the office with distinction—is likely to be a superior person, not at all typical of the voters. And it is the officeholders, the elected officials and the officials whom they appoint, who govern—not the people who elect them.10 Of course voters may often be deceived about who is the best candidate; even so, an average Joe is unlikely to prevail in electoral competition, just as an average Joe is unlikely to win the world boxing title. The glib, the clever, the shrewd, the handsome, and the charismatic are likely to dominate the electoral competition, occupy the principal offices, and constitute, in short, a political aristocracy. Not that the people are an inert element in representative democracy. The politicians will vie for their favor by advocating policies that command broad support. But these advocates will be drawn from a narrow and unrepresentative segment of the community, and the policies they advocate, or seem to advocate, to win popular support will often not be the policies they implement.

Representative democracy becomes monarchical when the representative is a dictator and there is only one candidate, as in the plebiscites employed by Napoleon and Hitler to cement their rule, or the one-party voting in communist states. But democratic aristocracy, oxymoron as it may seem, best describes our system. The framers of the original Constitution, the Constitution of 1787, decreed that the House of Representatives would be popularly elected, but they had a distinctly aristocratic conception of the Presidency; it is reflected in the device of the Electoral College. In the wake of the 2000 election fiasco, there are renewed calls to abolish the Electoral College. We shall be considering the merits of abolition throughout this book, particularly in the last chapter. The deadlock, and the difficulties of resolving it, were inseparable from that institution. But even if the Electoral College were abolished and the President directly elected by popular vote, our political system would remain aristocratic in Aristotle's sense.

Our political aristocracy is not hereditary, and its direct or indirect dependence on the popular vote, not only for initial appointment but also for reappointment at stated intervals, tends to align its interests more closely with the interests of the population at large than do other systems of government.11 Nevertheless, representative democracy, democratic self-government, and popular government are misleading descriptions of the system for governing modern "democracies." This is so not only because the people do not rule, and because there are no guaranties that they vote intelligently or even conscientiously, but also because the voters are not the people, but only a fraction (and not necessarily a representative fraction) of the entire population, and also because, as I have noted, voting is a crude method of aggregating preferences. In the 2000 Presidential election, little more than a third of the population actually cast a ballot that was counted as a vote for one of the Presidential candidates, and the winner of the popular vote lost the election, because the loser had more electoral votes. Winner and loser, moreover, each had the votes of only about one-sixth of the total population.

Although the Electoral College is controversial, one of the principles that it embodies, that of districted rather than at-large elections, is not; and yet districting drives a further wedge between popular majorities and electoral outcomes. If two parties compete for control of a 100-member state legislature elected by districts of equal population, a party that wins a bare majority of the votes in a bare majority of the districts will end up with control of the legislature, though that party could be the preference of only a shade over 25 percent of the state's voters, who might in turn be only a modest fraction of the population. This paradox is complicated at the Presidential election level by the facts that a state's electoral votes are the sum of the state's Senators and Representatives in Congress and that each state has two Senators regardless of its population. This might be thought to imply that the votes of the voters in the less populous states are weighted more heavily in the Electoral College than those of the voters in the more populous states. Yet we shall see in Chapter 5 that this is an oversimplification, because given the winner-take-all rule that all but two states use for allocating their electoral votes, swing voters in more populous states have more influence in the Presidential election than swing voters in less populous ones.

The only constant is that voting is indeed a crude method of mapping preferences onto policies—and for the further reason that voting for a candidate is voting both for a person and, in effect, for a package of likely policies: the policies the candidate supports, possibly the policies endorsed in his party's platform. So wholly apart from voter ignorance, policies may be adopted that do not actually command majority support, either because a majority of voters prefers the candidate whose policies they do not like (they may consider him an abler leader) or because the majority likes only a subset of those policies. In a two-party system, moreover, the parties have a strong incentive to move to the center of the distribution of political opinion. This may force voters to choose between two candidates who have largely identical views that are not widely supported but that have the support of the median voter. Making a choice between two alternatives is more difficult the harder the alternatives are to distinguish—but it is also more difficult the more alternatives there are, so that a proliferation of candidates does not cure the Tweedledum-Tweedledee problem.12 And such a proliferation is in any event unlikely because a two-party system is a natural although not inevitable corollary of a Presidential as distinct from a parliamentary system. It is difficult for a third party to mount a credible Presidential campaign—third-party Presidential candidates are invariably just "spoilers"—and so it is difficult to take the party itself seriously as a national force.13 What really does in third parties, however, is a winner-take-all voting system (as distinct from proportional representation) at the legislative level. For that is likely to prevent a third party from electing any legislators at all, making the party impotent at the legislative as well as the executive level.

When ignorance about issues and candidates,14 exploitive intentions (voting for the party that you hope will, if it takes power, redistribute wealth to you from other people), low turnout, and errors in voting and in tabulating votes are added to the brew, it becomes apparent that voting registers informed public opinion in only the loosest sense and the "popular will," Rousseau's "general will," perhaps not at all. Some of the distortions may be offsetting, but that cannot be assumed. And while it is true that low turnout would not matter if nonvoters had the same interests as voters, they do not. Turnout is disproportionately high among the elderly even after adjustment is made for other influences on it 15—a factor that, in combination with the disfranchisement of children, creates a strong public policy tilt away from children and toward the elderly.16 We shall note in Chapter 5 that the expressed political preferences of nonvoters tend to be quite similar to those of voters; but those preferences may not correspond very closely to nonvoters' interests. As nonparticipants in the electoral system, nonvoters are unlikely to have given much thought to aligning their political views with their interests.

Our actual existing democracy falls so far short of the soaring ideals of the theorists of democracy 17 that some of those theorists might be inclined to deny that our system is democratic. But that is to be unrealistic, as well as to attribute a fixed meaning to a word ("democracy") of notorious plasticity. Representative democracy has decisive pragmatic advantages over alternative systems of governing a modern society, and so we can be enthusiasts for democracy without having to prate about self-government or the popular will. Paradoxically, the advantages of representative democracy emerge most clearly if we eschew pious platitudes about civic virtue and assume that people (including those who live in democratic societies) are, whether as citizens or as officials, self-interested rather than public-interested or altruistic, or that, if they are public-interested or altruistic, the specific beliefs they hold are likely to be distorted by self-interest or ignorance. It is a realistic assumption and implies that people—whether the collective "We the People" of the Constitution's preamble or individual persons—cannot be trusted to exercise power.

That is a worrisome thought. The core function of government has always been to provide services—classically, internal and external security—that the market cannot provide efficiently because the benefits which these services generate could not be appropriated by the private individual or firm that produced them. Imagine trying to finance the national defense by means of voluntary contributions. People who did not contribute would obtain the same benefits from national defense as those who contributed, and so the incentive to contribute would be meager. The same thing is true with regard to the financing of police, prosecutors, judges, and the rest of the law enforcement apparatus.

The effective performance of the core functions of government requires a concentration of physical power—the "monopoly of force" that all governments claim. Force to maintain law and order, force to extract the revenues necessary to pay for that maintenance. As a result, control of the government creates opportunities for expropriation of the citizenry's wealth, and for other abuses, provided there are significant costs of emigrating, as there usually are. Quite apart from the danger of abuse, the mere fact that the basic government services are ones not sold in a market makes the monitoring of their efficient provision difficult. Both the loyalty and the competence of our political fiduciaries must somehow be secured by the institutions of the society.

The key institution in our society is, precisely, representative democracy, and its defining "democratic" feature is that the representatives are elected by the public at large, with each eligible member of the public having a single, nonsalable vote. But why voting? Why majority voting? Why equal voting? And why are votes nonsalable? These turn out to be related questions, and the answers will point us toward the central concerns of this book.

Voting in effect constitutes citizens 18 the "owners" of the government, in much the same way that shareholders are the owners of corporations. But whereas shareholders vote by shares rather than per capita, because their stakes in the corporation may be very unequal, citizens have more or less equal stakes. Not entirely so, of course; people are very differently related to government so far as the costs they bear and the benefits they reap. But because of the great power of government, almost everyone has a significant stake in the government's operation.

Too great a departure from the "one person-one vote" principle (as opposed to corporate democracy's "one share-one vote" principle, under which, were it transposed to the political setting, a share might be a specified percentage of the total wealth of the electoral unit) would make acute the danger of the government's being captured by a minority (I mean an electoral, not an ethnic, minority—a cabal, an oligarchy) bent on expropriating the property or extinguishing the freedom of the majority or of another minority. The reason is that the fewer people who have to be organized in order to achieve a common end, the more likely they are to succeed: transaction costs are lower, a proposition familiar to students of price fixing and other conspiracies.

For the same reason, votes must not be salable,19 since that would enable the concentration of voting power in the hands of a relative handful of people or corporations. A market in votes presents a classic free rider problem. If a wealthy individual offered to buy the first million voting rights tendered to him at $25 a vote, he would have little difficulty closing the deal, especially if many voters assumed that others would sell their right to vote for that price.20 Suppose he wanted to buy a really commanding number of votes, such as 10 million—only a modest percentage, however, of the total number of votes cast in the 2000 election, and an even smaller percentage of the total number of votes potentially buyable, since the half of the electorate that did not bother to vote would be especially eager to sell their voting rights. He might have to pay a much higher price per vote. Suppose it would be $100. Still, many firms and even individuals can plunk down $1 billion in cash for an asset that is worth that much to them. Allowing votes to be bought and sold would, therefore, be practically as well as theoretically inconsistent with the principle of equal voting. Majority voting also follows from that principle, because it is the only method of vote counting that weights each voter's vote equally.

In short, although the people do not rule in a representative democracy—that would be infeasible—they pick the rulers, kick them out when necessary, and provide for an orderly, peaceful succession (although the one in 2000 turned out not to be so orderly) when a vacancy occurs. The last point deserves particular emphasis given the subject of this book. Hereditary monarchy solved the succession problem, but at the sacrifice of both quality assurance and democratic control, sacrifices that representative democracy avoids having to make. Taken all in all, representative democracy not only honors the democratic principle but also controls the rulers, and solves the succession problem, better than any alternative system, provided—a qualification to be borne in mind throughout this book—that the procedures for determining the succession are fixed in advance, objective, administrable, and clear, so that succession is according to rules rather than to a power play by the loser.

With such advantages, one may wonder why representative democracy is not the universal system of government, and particularly why it seems to flourish only in wealthy modern countries, with a few exceptions, such as Switzerland even before modernity and a poor India today. (The Roman Republic was an interesting mixture of democracy and oligarchy.) Notice that I did not number the ancient Greek city-states among the exceptions. Direct democracy, the ancient Greek form of democracy, is feasible only in a very small, simple polity, because in a large or complex one the information costs—not to mention the costs of underspecialization and of time—of citizen government are prohibitive. Representative democracy reduces all these costs by enabling a political division of labor between governed and governing without relinquishing popular control over the governors.

But unlike Greek-style democracy, representative democracy requires an elaborate institutional framework to avoid degenerating into oligarchy or dictatorship. Remember that in such a democracy it is officials who rule, not the people. Officials have their own interests, which may not coincide with those of the people. If able to monopolize the information relating to the administration of government, these imperfect agents may be able to perpetuate themselves in office while adopting policies that disserve their constituents. So a reasonably well-informed citizenry—a citizenry that even if not highly educated, or even highly literate, makes at least approximately accurate judgments about candidates and policies—is a prerequisite of democracy. Another prerequisite is a system of property and personal rights, including the right of free speech, enforced by judges who are not beholden to or intimidated by the representatives. For without such a system—that is, without real liberty—the representatives will be able to cow political opponents, steal elections through fraud and through intimidation of voters, and, by thus making reelection an empty formality, perpetuate their rule indefinitely.

So democracy depends on liberty, as well as being in tension with it because liberty curtails the power of the majority to impose its will. The right to vote is not enough; it is really just the tip of an institutional-cultural-juridical iceberg; a competent and independent electorate, which implies a liberty-securing institutional infrastructure, is indispensable to the operation of representative democracy. Since people are self-interested and for that and other reasons untrustworthy, pure democracy—democracy that places no limits on what the majority or its representatives can do 21—is an extremely dangerous system of government, viable if at all only in tiny polities, where ties of family and friendship may enable voluntary cooperation to be substituted for coercion. Pure democracy is unstable and likely to degenerate into oligarchy or autocracy. Limited democracy is best. Representative democracy is limited democracy, even without judicially protected liberty, because it puts up a screen between citizens and the application to them of state power by other citizens.

It is also a far more efficient and economical system of governance than direct democracy. Not only does it enable specialization in the provision of government services and political goods; in addition, because governance is delegated to the specialists, the elected and other officials, the people at large do not have to spend all or even much of their time worrying about politics. Some theorists—Hannah Arendt is perhaps the outstanding modern exemplar—are distressed that modern democracy enables the mass of people to redirect their energies from the public to the private sphere.22 Bonnie Honig would go further, radicalizing Arendt in an effort to restore "politics as a disruptive practice that resists the consolidations and closures of administrative and juridical settlement for the sake of the perpetuity of political contest."23 For most of us, it is a relief not to have to be jawing all the time in the agora.24 Not for us exhaustive deliberation, life modeled on a faculty seminar. Not only is there much more to life than politics, but a preoccupation with politics is likely to exacerbate social conflict. Political conflicts are not intellectual disagreements, resolvable by deliberation or debate. They are clashes of interests and values. The political class in this country—the class that dwells obsessively on political issues—is more contentious, radical, dogmatic, and polarized than the country as a whole.

Representative democracy is historically and today associated with the market economy, though the correlation is one-sided. Nondemocratic nations often have market economies, but nations that do not have market economies are rarely democratic. A market economy seems therefore to be a necessary but not a sufficient condition for democracy. It is easy to see why it is not sufficient. Free markets are much more effective at generating wealth than non- market economic systems, and an autocrat generally wants his country to be wealthy in order to keep his subjects happy, finance an effective security apparatus, and become wealthy himself without taking such a large proportion of national wealth that he causes serious unrest. (The qualification "generally" is important, however, as wealth may create pressure for democracy, as we are about to see.)

Three things make a market economy a prerequisite for representative democracy. First, such economies depend on respect for property rights, and the effective enforcement of property rights requires a competent and impartial judiciary, which is also essential to ensure that elections are honest and to protect the personal liberties that democracy, the (potential) "tyranny of the majority," threatens. Judicial protection of property rights is the forerunner of judicial protection of political liberty. Second, market economies generate wealth, and wealth increases the demand for and the supply of education, communications, and leisure. These goods, along with the financial security of living in a prosperous society, create a citizenry that not only is reasonably well informed about political issues and candidates, but also is sufficiently independent economically not to be the pawn of the mighty; relations of patronage and dependence undermine the power-diffusing objective of equal voting. Third, market economies reward and thus encourage commercial values, which are more hospitable than aristocratic or religious ones to the political equality that undergirds a democratic system. Like theocrats, aristocrats (not in the Aristotelian sense, the sense I used earlier, in which aristocracy is rule by the best, but in the more familiar sense of a hereditary caste preoccupied with honor and status and disdainful of commercial pursuits) think the issues involved in government too important to be left to the people They also (the extreme example is Coriolanus) disdain the dependence on the goodwill of hoi polloi that a democratic system imposes on officials.25 Persons engaged in market activities disdain others at their peril, since success in the market involves catering to the preferences of others, namely one's customers and to a lesser extent one's employees and other suppliers.

There are countercurrents. The ideology of the market can foster contempt for voting because of its lack of instrumental value and its failure to weight preferences by willingness to pay, which is how the market weights preferences. The sacralizing of property rights in the name of economic efficiency curtails democratic governance by disqualifying government from intervening in the market. And markets can foster inequalities in income and wealth, which can in turn foster selfish, exploitive voting as the electorate splits into classes that have little in common with each other. But these effects are outweighed by the support that a market economy offers to representative democracy; otherwise having a market economy would not be a necessary condition of a society's being a representative democracy.

The foregoing introduction to the practice of representative democracy, brief as it has been, suggests several points germane to the argument of this book. Representative democracy is a pragmatic institution rather than the instantiation of a theorist's ideal state. Voting is a method of control, not of administration. The people do not rule in a representative democracy; they control the rulers, their delegates. For voting to perform its function of control, voters must have some minimum of political sophistication, along with a measure of independence from other people. Voting is central to the orderly succession of democratic "rulers." "Orderly" implies ordered, and delegation of governance to specialists implies realism and practicality. American democracy is structured, formal, practical, realistic, and both supportive of and supported by commercial values. It is not starry-eyed, carnivalesque, or insurrectionary. It is not pure or participatory democracy, and it does not consider political chaos a price worth paying to actualize the popular will. Its spirit is closer to that of Burke than to that of Rousseau. The populism of a Jefferson or a Jackson remains a part of our democratic ideology, but a smaller part than in days of yore. These summary reflections, too, will turn out to be relevant to evaluating the Supreme Court's performance in Bush v. Gore.

A History of the Suffrage

The entwinement of democracy with economy will also play a role in our analysis of the Florida election. As we shall see in the next chapter, the deadlock and ome of the bitterness sparked by its resolution had economic roots, such as illiteracy, which is linked to poverty, and the financial costs and benefits associated with different systems of voting. There was a subterranean issue of whether literacy should be a voter qualification. Although federal law forbids the use of literacy as a voter qualification in federal elections,26 the punchcard ballot used in a number of Florida counties (indeed used widely throughout the United States) requires a higher degree of literacy than other common voting technologies.

Poverty fosters ignorance, and also dependence. These are age-old worries about political elections. By the eve of the American Revolution, the idea that an elected legislature was an element of civilized government had taken firm hold in Great Britain and its American colonies. But neither the principle of universal suffrage so familiar to us (though not fully actualized even today), nor the idea that all legislators (and the executive as well, and maybe even judges) should be elected by the people, had yet taken hold. The suffrage was limited as a matter of course to free adult males who owned property, specifically land. Property ownership was much more widespread in the colonies than in the mother country, with the result that something like two-thirds of free American adult males could vote. But they could vote only for the members of the lower house of the state legislatures. The governor was appointed by the Crown, and the members of the governor's council, corresponding to the Senate in the federal government ordained by the Constitution of 1787, were appointed by the governor.27

This basic structure was retained in the Constitution.28 Article I, section 2, provided that the members of the House of Representatives were to be elected by the people of each state. No effort was made to eliminate property qualifications or other limitations on the suffrage; the only stipulation (also in section 2) was that the voters have the same qualifications that the state required of people voting for members of its lower house. Article I, section 3, provided that the Senators from each state would be appointed by the state's legislature. This method of appointing Senators reflected (at a time when states were much more important than they have become) a kind of ambassadorial conception of a Senator. Indeed, state legislatures sometimes instructed "their" Senators with regard to how to vote on specific issues.29 The state legislatures were authorized to fix the time, place, and manner of choosing Senators and Representatives, though Congress was authorized to alter those regulations—except for the place of choosing Senators, which would be the state legislature itself. Each house of Congress was to be "the Judge of the Elections, Returns and Qualifications of its own Members."

There was little support at the constitutional convention for popular election of the President. Most of the delegates held the aristocratic conception of the Presidency and were dubious about the capacity of the public at large, with the limited filtering provided by gender and property qualifications, to pick the best candidate. There were also concerns about the logistics of conducting a nationwide popular election in a large country with poor transportation and communications. The alternative of having the President elected by Congress was unattractive, as it would make the President unduly dependent on the legislative branch. The ingenious expedient hit upon by the delegates was the Electoral College. Each state would have as many electors, and hence electoral votes, as it had Representatives plus Senators. The Electoral College was thus to be a kind of ad hoc Congress, its sole function being to elect the President. Its members were to be appointed by each state "in such Manner as the Legislature thereof shall direct" (Article II, section 1, clause 2) and were to meet in their state and vote for the President rather than assemble in one place and confer before voting. Members of Congress, and other high federal officials, were ineligible to be electors. Congress was to fix the time at which the electors would be chosen and the day on which they would cast their votes, provided the day was the same throughout the nation. The state was to transmit its electoral votes to the seat of the federal government, where they were to be counted in the presence of both houses of Congress. If no one received a majority of the votes of the appointed electors, the House of Representatives was to elect the President from among the leading candidates. But in that election, unlike the ordinary procedure of the House, each state delegation would have a single vote.

The framers' insouciance about voting qualifications for the members of the only directly elected branch of the new government, the House of Representatives, was matched by their insouciance about the method of appointing Presidential electors, which was left to the state legislatures to decide. The framers were unperturbed by the possibility that the legislatures might (as most soon did) decide that the electors would be chosen by popular vote. Apparently they thought that the people could be trusted to pick electors good enough to pick the best person as President. Moreover, they expected the contingent election procedure ordained by the Constitution—election of the President by the House of Representatives if no candidate received a majority of electoral votes—to be used frequently, and the House was the most democratic component of the governmental structure created at Philadelphia in 1787, though, since each state's delegation would have only one vote to cast for President, the contingent procedure was not actually very democratic. That was discovered in 1824, the only time the President has been picked by that method. Andrew Jackson lost to John Quincy Adams in the House even though Jack-son had the most popular votes 30 and the most electoral votes, though not a majority of the latter. The framers expressed no concern that a popular-vote loser might be an electoral-vote winner, since they had no reason to expect all states to select their Presidential electors by popular election. That such a discrepancy might be anomalous could not even be perceived until it was customary to select Presidential electors by popular vote, as it was, however, by 1824.

The expectation that the President would often be selected by the House of Representatives was related to the framers' failure to foresee the rise of political parties.31 (Parties in the modern sense were unknown in the eighteenth century.) Without parties to winnow the candidates, electoral votes were likely to be scattered among numerous candidates, reducing the likelihood that anyone would receive a majority, especially as it was doubted that many men had sufficient national reputations to garner a majority of electoral votes. In a two-party system, in contrast, electors would be choosing between just the two candidates chosen by the parties.

The Electoral College was not created to be a deliberative body, since its members would not meet but would, as expressly stated in Article II and the Twelfth Amendment, vote separately in each state.32 This decision was made partly because of the difficulties of travel, but more because of concern lest the choice of the President be influenced by cabals, agents of foreign powers, other intriguers, or corrupt deals. It was believed that requiring that the electors vote in their home states rather than congregating to vote, that the electors not be federal officials, and that all electoral votes be cast on the same day would minimize these dangers.33

Such was the scheme of representative democracy created by the Constitution of 1787 for the federal government. The history of U.S. representative democracy since that date is a vast subject, but fortunately only three facets of that history are important for my purposes. One is the general though irregular movement to broaden the suffrage. Another is the trend away from indirect election. The third is the evolution of election administration, with particular reference to efforts to avoid (or if necessary resolve) deadlocks and to count votes accurately, which turn out to be related desiderata.

The movement to broaden the suffrage antedates 1787. On the eve of the Revolution, the electorate for the colonial assemblies, following the British model, was generally limited to adult Protestant males who had freeholds (that is, land ownership) of some specified minimum value. The theory behind the limitations was that only propertied men would have sufficient economic independence and political knowledge to be independent and competent voters rather than pawns of the wealthy and knowledgeable, or, in the case of a woman, of her husband, brothers, or father. Though made at least somewhat plausible by the economic and social conditions of the time and by the fact that the ballot generally was not secret, the theory was at best only partial. It did not explain the religious exclusions from the suffrage or the preferred position of owners of real estate compared with owners of other forms of property that might be equally valuable. Religious hostilities, distrust of city folk, and fear of debtors ganging up on creditors and, more generally, of the poor expropriating the rich played a larger role in the restrictions on the suffrage than the incomplete theoretical justifications based on notions of independence and competence. (The most plausible justification for the preference accorded owners of real estate was that they were more vulnerable to expropriation because their wealth was immobile.)34 Lack of independence was decisive, however, against permitting slaves to vote, since their votes would be controlled by their masters and thus would magnify the political power of slaveholders.35

The limitations on the franchise eroded significantly between 1775 and 1787. The causes of this erosion were ideological and practical. The slogan "no taxation without representation" made it difficult to justify the denial of the franchise to people who paid taxes yet happened not to own land, especially since the limitation of the franchise to freeholders had been justified in part by the argument that they paid most of the taxes.36 The equally influential slogan "all men are created equal" pointed toward universal adult male suffrage. And it was difficult to ask people to join the Continental Army without giving them the rights of a citizen.37

Yet when the Constitution was adopted, only Vermont had abandoned all property qualifications for voting,38 and the framers decided to allow each state to set the qualifications for voters in federal elections conducted in that state. As I have already noted, the qualifications to vote for members of the House of Representatives, the only form of direct election ordained by the Constitution, were to be the same as those for voting for the members of the lower house of the legislature of the voter's state, while Senators were to be appointed by the legislature of each state and Presidential electors by each state in the manner directed by the state legislature.

The trend toward broadening the franchise continued after 1787, powered by democratic sentiment,39 by (what is not the same thing) a decline of deference, by agitation of the disfranchised for the vote,40 and by the inherent ratchet effect of changes in the franchise. The franchise is likely to be enlarged whenever the currently dominant political forces in the society believe that the newly franchised voters will support them rather than their opponents,41 and once a new group is enfranchised it becomes difficult later to withdraw the franchise from it if the political winds change, because the members of the group will vote solidly against the change. By the eve of the Civil War, universal adult (age 21 or older) male suffrage had been achieved in the Northern states,42 and reinstatement of property qualifications would have been politically infeasible, quite apart from ideological considerations.

But it would be wrong to infer from the ratchet effect that the expansion of the franchise is monotonic. The effect is real, but other forces are also at work. After Reconstruction ended following the election of Rutherford Hayes to the Presidency in 1876, the Southern states, forced by the Fifteenth Amendment to extend the franchise to blacks, nevertheless were able through a variety of devices, ranging from poll taxes and literacy tests to outright intimidation, to disenfranchise most blacks. Moreover, the scope of the franchise can contract merely because of demographic changes occurring against a background of unchanged rules. For example, if aliens are forbidden to vote and they become a larger fraction of the population, the fraction of eligible voters will, if nothing else changes, fall automatically. The same is true if the birth rate increases and as a result a larger fraction of the population is below the voting age. The effect of expansion of the suffrage on actual voting can also be—and in fact to a considerable extent has been—offset by a decline in the turnout of eligible voters.43

Aliens are an example of a group that, being unorganized and unpopular, may not be able to take advantage of the ratchet effect. After the Civil War, the flood of immigrants created anxieties about the voting power of these new citizens, and literacy tests were instituted for the first time 44—along with an effective de facto literacy test that bears a distant resemblance to the punchcard ballot. Until late in the nineteenth century, the government did not supply the ballots for voting in elections. Instead each political party supplied ballots containing a party-line vote for the party's candidates. The voter would simply select one party's ballot and drop it into the ballot box. The party ballot was replaced, largely in the last decade of the century, by the "Australian" ballot: a paper ballot, supplied by the voting authorities, that contained a list of the candidates, on which the voter would mark his preference. (Such a ballot had first been used in Australia; hence the name.) A voter who was not literate would have difficulty using the Australian ballot,45 a result welcomed by those who feared the voting power of immigrants. At the same time, the heretofore rather casual attitude toward voting by aliens 46—the distinction between citizens and aliens not being strongly marked—gave way to rules excluding aliens from the franchise.47

The practical and ideological forces that, despite some backsliding, led to the enlargement of the franchise in the nineteenth century also led to expanding the categories of officials elected by the people. In the eighteenth century the only popularly elected officials had been the members of the lower houses of the federal and state (or colonial) legislatures. Judges became elected rather than appointed officials in most states in the nineteenth century, as did state governors, members of the upper houses of the state legislatures, and Presidential electors. It has thus become anomalous that the President of the United States is not elected by popular vote. Yet there are only two certain instances in which the popular-vote winner failed to win the electoral vote: the election of 1888, in which Benjamin Harrison lost the popular vote to Grover Cleveland but won the electoral vote and so became President, and, of course, the election of 2000. The controversial elections of 1800, 1824, 1876, and 1960 are distinguishable—some of them, however, only narrowly. Let's consider them briefly.

In 1800 Jefferson and his running mate, Aaron Burr, received the same number of electoral votes, and Jefferson was elected President by the House. The deadlock in the Electoral College had been caused by the fact that Article II of the Constitution, while giving each elector two votes, did not provide for the electors to vote separately for President and Vice President. The theory was that the best man would come in first and become President and the second-best man would come in second and become Vice President. But if the electors thought alike on who should be President and who Vice President, and accordingly cast one of their two votes for their preferred Presidential candidate and the other for their preferred Vice Presidential candidate, the two—though candidates for different offices—would end up with the same number of electoral votes, and the designated Vice Presidential candidate might not gracefully withdraw (Aaron Burr did not, and Alexander Hamilton's decision to throw his support to Jefferson was one of the events that lay behind the famous, fatal duel).48 For that matter, if there was more agreement on who should be Vice President than on who should be President, the Vice Presidential candidate might receive more electoral votes (as almost happened in 1800), and so become President. These sources of deadlock (or worse) were removed by the Twelfth Amendment, adopted in 1804. The amendment required the members of the Electoral College to vote separately for President and for Vice President.

In 1824 Andrew Jackson won both the popular and the electoral vote, but because he did not have a majority of the electoral vote, but only a plurality, the choice of President was made by the House of Representatives, which picked John Quincy Adams.

In 1876 Samuel Tilden was declared the winner of the popular vote, but competing slates of electors in several Southern states (including Florida!) made the outcome of the electoral vote uncertain. An ad hoc commission created by Congress in January 1877 to resolve the dispute over the electors awarded the disputed votes to Hayes, who thus obtained, when objections in Congress to the commission's award failed, a majority in the Electoral College, and so became President.49 Because of extensive vote fraud by Democrats as well as by Republicans, it is uncertain whether Tilden really did win the popular vote.50

In 1960 John F. Kennedy won the popular vote by only a shade over 100,000 votes. There were serious allegations of fraud by Democratic election officials in Illinois and Texas, but the number of votes affected was not enough either to change the result in the Electoral College vote or to deprive Kennedy of his plurality of the popular vote nationwide. Deciding who won the popular vote in 1960 is clouded, however, by a question unrelated to fraud, namely how to classify the popular votes for the 6 unpledged Democratic electors in Alabama, who, out of a total of 11 Democratic electors in that state, ended up voting for Harry Byrd rather than for Kennedy. If Kennedy is allocated 5/11 of the popular vote in Alabama, then Nixon, not Kennedy, had a plurality of the nationwide popular vote.51

In 2000 Al Gore won the popular vote, but George W. Bush obtained a bare electoral-vote majority—thanks to Florida, which many people believe Bush "stole" just as Hayes may have done 124 years earlier. We shall examine that contention in later chapters. Whether it is well founded or not, there is no question that Gore won the nationwide popular vote fair and square.

In short, in five elections (1824, 1876, 1888, 1960, and 2000) the winner of the popular vote for President either was not or may not have been (1876 and 1960) elected President. But the wedge that the Electoral College drives between popular democracy and election to the Presidency is larger. Because Presidential candidates campaign to win the electoral vote rather than the popular vote, they allocate their efforts, their appeals, their choice of running mate, and their policies and appointments when elected differently than they would if the President were chosen by popular vote rather than by the vote of the Electoral College. We shall return to this issue in the last chapter.

The twentieth century saw further expansions both in the suffrage and in the categories of official appointed by popular vote. The Seventeenth Amendment, ratified in 1913, required the direct election of Senators. The reasons were, in part, intensely practical, such as that elections for state legislators were sometimes overshadowed by concerns over whom a legislator might support for U.S. Senator and that legislatures frequently deadlocked over the choice of Senators.52 In 1920, the Nineteenth Amendment was adopted, guaranteeing the right to vote to women. Subsequent amendments abolished poll taxes, entitled the District of Columbia to appoint Presidential electors, and guaranteed the right to vote to 18-year-olds. More and more states adopted the primary election as the method of choosing the candidates of the major parties. And legislation designed to make the Fifteenth Amendment a meaningful protection of the franchise of black citizens 53 culminated in the abolition of literacy tests for voters in all federal elections. With universal adult suffrage within reach, attention shifted to subtle issues of voting power, including malapportionment (a conspicuous feature of the U.S. Senate and therefore of the Electoral College as well), gerrymandering, the creation of districts in which minority groups would have a voting majority, restrictions on candidacy, lengthy residency requirements, and campaign financing.

Because it became an issue in the 2000 election deadlock and its aftermath, we should consider the merit of the goal of universal suffrage—especially when it is pressed to the point of insisting that people who cannot read well enough to follow voting instructions should be permitted, perhaps even assisted, to vote. One way to put the question is to ask whether it would be desirable, were it politically feasible (which it is not), to confine the vote to people who are well informed about the public issues—including issues of leadership, competence, and probity—that are relevant to picking the best representatives. It would be undesirable. The politically well informed are an unrepresentative slice of the population. They differ systematically from the poorly informed along the dimensions of race, income, and education.54 They have interests as well as opinions, and, since they are not a random draw from the population, their interests differ systematically from those of other groups. If the politically savvy monopolized the vote, their interests would have more weight than those of other people, the politically apathetic, immature, or alienated, unless we indulge the unrealistic assumption that well-informed people can be depended upon to be altruistic rather than self-interested voters. Once that assumption is abandoned, it becomes clear that the interests of any group that lacks the franchise are likely to be undervalued in the political process.55 It is on this basis that I have urged consideration of giving each parent an extra one-half vote for each of his or her minor children.56

Some people, such as small children and the severely retarded, cannot form a competent conception of their own interests. But this is not the case for illiterates, especially in an era in which people get most of their information about public affairs from radio and television rather than from the print media.57 Literacy tests for voting have now been outlawed, in major part because of their historical association with the pertinacious efforts of the Southern states to deny the vote to blacks. But a voting technology that as a practical matter requires literacy to cast a vote that will be counted operates in the same direction. Another difference between children and the severely retarded, on the one hand, and illiterates, on the other, is that children and the severely retarded usually have someone—a parent or guardian—to look after them, and so are less needful of the vote to protect their interests; illiterates do not.

There is also a psychological factor to be considered. The right to vote is a symbol of equality. It dramatizes the principle that every person is to count for one and no one for more than one, at least in the political sphere. This may be hokum or sentimentality, or even a mask for the inequalities of circumstance and opportunity that pervade our (as every) society; but it is a brute fact about the American political culture. It is one reason why giving additional votes to members of particular groups (say, to blacks, or to veterans) to reward past services or to compensate for past or to offset present discrimination is out of the question, and why denial of the suffrage is taken so much more seriously than denial of giving votes equal weight (as in the Senate and the Electoral College); and it is why ex-felons are denied the vote in some states—as a symbol of disapprobation. People are not illiterate by choice, and to deny them the vote would therefore be a gratuitous insult, as is not the case with denying the vote to ex-felons.58

The psychological effect of denying the vote to people may have practical consequences. In Mill's words, "whoever, in an otherwise popular government, has no vote, and no prospect of obtaining it, will either be a permanent malcontent, or will feel as one whom the general affairs of society do not concern."59 But not too much weight should be placed on the symbolic and psychological dimensions of the right to vote. They are extremely speculative. The important thing is that even people who cross only a very modest threshold of competence and independence ought to be allowed to vote in order to protect their interests, and that illiterates have interests and, in our radio- and television-saturated society, generally cross that threshold.

The dramatic and ultimately successful struggles to broaden the legal franchise, plus the increasing preoccupation with subtle issues of voting power (even if everyone has one vote, voting power—that is, the power through voting to influence the outcome of an election—may be unequally distributed, as when legislatures are malapportioned), have obscured the mundane issues of election administration that the Florida 2000 Presidential election brought to light. With the successful enforcement of the Voting Rights Act, the abolition of poll taxes and literacy tests, simplified registration, the curtailment of durational residency requirements, and the reduction in the amount of corruption and the hold of machine politics in big cities, the subject of election administration became practical (rather than a matter of theoretical speculation or moral urgency), fragmented (because election administration—as distinct from federal constitutional and statutory regulation of voting—is decentralized to states, counties, and even precincts), and technological. It dipped below the radar screen of constitutional lawyers, and most political scientists as well. Not only the elections of 1824, 1876, and 1888 but also the close and problematic elections of 1960 and 1968 (the former rife with fraud and the latter almost undone by runaway electors), became "history." Even the much more recent, exceptionally bitter, and absurdly protracted congressional battle over the deadlocked 1984 election in Indiana's Eighth Congressional District was forgotten.60 Forgotten too was the push in the wake of the 1968 Presidential election to abolish the Electoral College by constitutional amendment, a proposal approved by the House by the requisite two-thirds margin in 1969.

The problematic Presidential elections that I have listed involved or exposed a variety of problems. The 1824 election taught that a President who had lost both the popular and the electoral vote—and who owed his election only to the House of Representatives, because the electoral-vote winner had had only a plurality, and not a majority, of the electoral votes—might not be accepted by the nation as fully legitimate. Adams was defensive about his victory and was soundly trounced by Jackson in their rematch in 1828, though in part this was due to the suspicion that Adams had owed his election to a "corrupt bargain" with Henry Clay, the Speaker of the House, who indeed became Adams's secretary of state 61—but such suspicions are likely when the election is thrown into the Congress. A similar danger, that the winner of both the popular and the electoral vote might nevertheless not be elected President, loomed in the 1968 Presidential election, because of George Wallace's strong third-party candidacy, since a third-party candidate who won some electoral votes might, by throwing them to the losing major-party candidate, give that candidate a majority in the Electoral College. The problem in the 1960 election was not only fraud but also, and more seriously, unpledged electors, and it loomed briefly again in 2000 when some Democrats mounted a campaign to persuade Bush electors to switch to Gore. The election of 1888 was a reminder that the popular-vote winner can lose the electoral vote, which of course happened again in 2000, though so far without untoward results. The problem in 2000 was deadlock, and it may actually have had the effect of distracting people from the anomaly of the popular-vote winner's losing the election, though Gore's lack of personal popularity was also a factor.

The 1876 election fiasco was the most ominous portent for 2000. It exposed, as had the 1800 election, an embarrassing gap in the Constitution. The framers had foreseen and provided for the case in which no candidate wins a majority of the electoral votes (though the relevant provision is ambiguous, as we shall see), but not for the case in which there is a dispute over whom a state has appointed to be its Presidential electors. An ad hoc resolution of the dispute over the 1876 election was not achieved until the eve of the inauguration, and it involved Congress's appointing a committee whose deciding member was a Republican Supreme Court Justice who, along with the other Republicans on the committee, voted for the Republican candidate.62 These events led to the enactment a decade later of the Electoral Count Act, now Title III of the U.S. Code, which specified a procedure for resolving a future such deadlock. We shall see in Chapter 3, however, that the procedure is both incomplete and of uncertain constitutionality.

Completely unforeseen was the possibility that a deadlock in the Presidential election might arise not from skullduggery (as in the 1876 election), not from a failure of any candidate to obtain a majority of the electoral votes, not from runaway electors, and not from an actual tie in the popular vote in a key state, but from innocent defects in electoral mechanics, whether defects in the design of a ballot, the staffing of the polling places, or the design, maintenance, or operation of voting and vote-tabulating machines, or gaps and ambiguities in the state statutes regulating the administration of elections and the resolution of election disputes. Nor was it foreseen that such shortcomings not only might make it difficult to determine who had won an election, and thus increase the probability of having to activate some postelection mechanism to decide who the winner was, but also might selectively disenfranchise voters. It had always been understood that the methods for counting votes are imperfect; challenges to the outcome of close elections have not been uncommon and have sometimes resulted in recounts that changed the outcome. But that this might happen in a Presidential election, that procedures that function properly in less momentous and less time-sensitive elections might not serve for a deadlocked Presidential election, and that the federal constitutional and statutory provisions relating to a contested Presidential election are rickety, were lost to view. These problems achieved visibility, not only to the public at large but also to scholars of the electoral process, only with the 2000 election. The gravity of such problems, and the difficulty of solving them with our existing laws and institutions, are the themes of the subsequent chapters.

---

Excerpted from Breaking the Deadlock by Richard A. Posner. Copyright © 2001 by Princeton University Press. Excerpted by permission. All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.

---