On November 2, the oddities of our presidential electoral system produced another near miss. Although George W. Bush received 3.5 million more votes than did John Kerry, a shift of fewer than 60,000 votes in Ohio would have given Kerry enough votes in the electoral college to make him president. Had that occurred, Bush would have had a unique place in our political history, as the only man to become president in a popular election that he lost and then to lose the presidency in an election in which he received more votes than his opponent.

As George C. Edwards tells us in Why the Electoral College Is Bad for America, such near misses have been remarkably frequent. He shows that on sixteen occasions before 2004 a shift of 75,000 votes or less in key states would have given the presidency to the candidate who finished second in the popular vote; six times, a switch of less than 10,000 votes would have done the trick. After four other well-known elections (1824, 1876, 1888, and 2000), a candidate took office as president despite his failure to win the popular vote. When presidential elections are close—as they seem to be about half the time—there is always a serious chance that the candidate who wins the most votes will not end up in the White House.

That, of course, is an important reason why many Americans have long favored getting rid of the electoral college or changing its most objectionable features, such as the “unit rule,” which gives all of a state’s electoral votes to the winner of that state’s popular vote.1 Elections, after all, are mechanisms for determining the preferences of the people. But the electoral system called for in the US Constitution assembles those preferences in such a way that the will of the majority can be thwarted.2 Not surprisingly, then, more than five hundred constitutional amendments to change the system have been introduced into Congress since 1816, and they have been supported by diverse political leaders, including Andrew Jackson, Richard Nixon, and Jimmy Carter—each of whom was a candidate in a disputed or “near miss” election.

The electoral college is, in fact, our most unpopular political institution. For more than fifty years, public opinion polls have indicated that between two thirds and three quarters of the American people favored replacing the electoral college with direct, national elections: “There is little question,” reported the Gallup organization in 2001, “that the American public would prefer to dismantle the Electoral College system.” In 1976, even Wisconsin’s presidential electors made clear their unhappiness with the electoral college; after casting their state’s ballots for president, they passed a resolution calling for the abolition of the institution in which they had just participated.

George Edwards, the editor of Presidential Studies Quarterly and Distinguished Professor of Political Science at Texas A&M University, shares their sentiments. His book methodically presents the case for eliminating the electoral college and replacing it with national popular elections. With excellent descriptions of how the electoral system actually works, it is the most cogent and up-to-date criticism I have read.3 Its title, however, is somewhat misleading: the book is less about the harm that the electoral college inflicts on the nation than it is about the deficiencies of the institution itself.

The deficiencies are not hard to spot. Some are procedural, including the problem of “faithless electors,” which arose briefly this fall when Richie Robb, a Republican elector from West Virginia, let it be known that he might not vote for George Bush even if West Virginia did. Although the Founding Fathers probably did intend to permit electors to form their own judgments about presidential candidates, the electoral college, since its inception in 1787, has been a messenger system rather than a deliberative body: the only role of electors is to cast ballots for the candidates to whom they are pledged. Yet very occasionally an elector has strayed, and in 1960 there was even a screwball plot by some electors to deny the presidency to the “labor Socialist nominee” Kennedy, and to make Harry Byrd of Virginia the president instead. In the end, only six unpledged electors from Alabama and eight from Mississippi, as well as a single Republican elector from Oklahoma, cast their votes for Byrd.

No faithless elector has ever determined the outcome of an election, but it could happen: the laws that prohibit independent voting by electors, on the books in more than a dozen states, are likely to be unenforceable. Had the last election ended in an electoral college tie (which some experts were predicting) and had Richie Robb voted for someone other than George Bush (West Virginia has no law against independent voting), the legal tangle would have far exceeded Bush v. Gore in complexity.

As Edwards correctly notes, however, the procedural weaknesses of a system that depends on electors adhering to a pledge are small in comparison to more fundamental flaws. Most importantly, the electoral college violates the principle of political equality central to modern conceptions of democracy. The allocation of electoral votes to the states (giving every state one electoral vote for each of its senators as well as for each of its members in the House of Representatives) guarantees that the votes cast by people living in small states will count for more than the votes of those living in large states. Wyoming, for example, has one electoral vote for every 170,000 residents while California has one for more than 600,000 residents. This “small state advantage” is sometimes compounded by another built-in flaw: since the size of a state’s congressional delegation is adjusted to match its population only after the decennial census, relatively fast-growing states are always underrepresented in the electoral college.


The principle of “one person, one vote” thus does not govern in presidential elections. Yet most Americans believe that principle to be both fair and integral to democracy; as Edwards shrewdly points out, no contemporary politician would publicly claim that the votes of some citizens ought to be worth more than the votes of others. The principle of “one person, one vote,” moreover, has been part of US constitutional law since the 1960s, when it was endorsed by the Supreme Court as a guide for drawing the boundaries of congressional and state legislative districts. “The weight of a citizen’s vote cannot be made to depend on where he lives,” declared Chief Justice Earl Warren. Yet that is precisely what happens in our most important elections.

The departure from democratic principles can become even more extreme in the event of a “contingent” election, which takes place when no candidate for president has a majority of the electoral vote. If that occurs, the Constitution requires that the president be selected from among the top three vote-getters by the House of Representatives, where each state delegation will have one vote. (This procedure was last invoked in 1824, when the House of Representatives ignored Andrew Jackson’s plurality in the popular vote and made John Quincy Adams president.) Had there been a need for a contingent election this past November, Wyoming, with half a million people, would have had the same influence as California, with 35 million; the president could have been chosen with the support of states representing less than a third of the nation’s population. In modern times, we have come closest to contingent elections in 1948 and 1968 when regionally based third-party candidates, Strom Thurmond and George Wallace respectively, won almost enough electoral votes to prevent any candidate from gaining a majority. Throwing the race into the House of Representatives was, in fact, Thurmond’s express aim, since he hoped that by doing so he could cut a deal that would slow the desegregation of the South and the enfranchisement of African-Americans.

The “contingent elections” mechanism has had few defenders (even James Madison, one of its authors, thought it was a failure), but the electoral college itself, over the years, has been defended by many politicians, pundits, and political theorists. As Edwards rightly observes, however, modern supporters of the electoral college have generally chosen not to directly engage with the criticism that the institution violates our conception of political equality. They make their case instead by insisting that the electoral college has other distinctive virtues and provides significant benefits to the country. Edwards is at his best when he is dissecting those claims and demonstrating how little solid evidence lies behind them.

He begins by rejecting the notion that in conceiving our electoral system, the Founding Fathers were following a “coherent design based on clear political principles.” In fact, the Founding Fathers had much difficulty deciding how the president should be chosen, and throughout the summer of 1787, they kept changing their minds. (A useful table in Edwards’s book succinctly details the weekly back and forth on this issue.) Many believed that Congress should choose the president, but others feared the president would then be beholden to Congress and to the cabals that would inevitably form to promote presidential candidacies. Not only Madison but James Wilson and Gouverneur Morris (both from Pennsylvania) favored direct, popular elections, presumably by the same electorate that voted for other offices. But opponents worried that the country was too large and its regions too various for the people to actually become familiar with different candidates.

Then too there was the omnipresent shadow of slavery. As Madison noted,

There was one difficulty…of a serious nature attending an immediate choice by the people. The right of suffrage was much more diffusive in the Northern than the Southern States; and the latter could have no influence in the election on the score of the Negroes.

Madison’s language was as elegant as the truth was harsh: national popular elections were not going to happen because they would not give the slave states as much power as they sought in selecting presidents. White Southerners wanted the electoral strength of their states to reflect their slave populations—which would not have been possible with direct elections, since slaves obviously would never be given the vote. It was, however, possible in an electoral system that linked a state’s voting strength to its representation in Congress—since slaves counted as three fifths of a person in apportioning seats to the House of Representatives.


The primary advantage of the electoral college, in the eyes of the Founding Fathers, was that it lacked the defects of either popular elections or selection by Congress. James Wilson indicated that the electoral college was “the second choice of many delegates though it was the first choice of few.” Remarkably little was said in the Constitutional Convention about the positive virtues of the electoral college or even about how the complicated mechanism would actually work; the framers made no claims about the electoral college protecting the small states or promoting federalism. The report of the final drafting committee, presented during the last days of the convention in September 1787, urged its adoption entirely because of the dangers of other methods. Edwards approvingly cites the conclusion of the constitutional historian Jack Rakove that the electoral college “was cobbled together nearly at the last minute and adopted not because the framers believed it would work, but because it was less objectionable” than the alternatives. The electoral college was a rough compromise among tired men who needed to bring their work to a close and who knew that George Washington would be the first president anyway.

In view of this history, the case for preserving the electoral college rests not on its origins but on claims about the benefits it provides in the modern era. Edwards takes up those claims, one after another, poking holes in each of them. In response to the often repeated assertion that the electoral college protects the interests of the small states, he points out that most states are too diverse internally to possess coherent, unified interests and that small states, in fact, have little in common with one another. Rhode Island and North Dakota share few characteristics. Moreover, most farmers (prominent among the imagined beneficiaries of the current system) live in large states like California and Texas. Edwards, citing campaign visits, speeches, and advertising expenditures, also dem- onstrates that small states, even now, attract little attention from presidential candidates. Campaigns, not surprisingly, concentrate on closely contested states such as Ohio or New Hampshire, not on big or small states per se.

Nor does the electoral college necessarily protect the interests of minorities who might be ignored in a direct, national election. This is an important issue, both because defending minority interests is a legitimate concern of any constitution and because some African-American leaders in the 1970s defended the electoral college on the grounds that it encouraged both parties to seek African-American votes. The historical record, however, makes plain that ethnic or racial minorities stand to benefit from the electoral college only if they are potential swing voters who are concentrated in competitive states: Cubans in Florida are the most obvious current example. The system may once have worked to the advantage of African-Americans when they were swing voters in some Northern states, but the unit rule appears to be having the reverse effect now. Since most African-Americans vote Democratic and live either in the solidly “red” South or in safely Democratic urban centers in the North, they are not strenuously courted by either party. Whether the nation’s growing Hispanic minority will become a potentially decisive bloc of swing voters that will benefit from the electoral college—and potentially resist reform—remains to be seen.

Edwards also considers other standard arguments for keeping the college: that the institution is critical to maintaining federalism, preserving the two-party system, and limiting fraud. On close inspection, there is little evidence to support the first two of these claims,4 and the third seems to be false: the unit rule appears more likely to encourage fraud than to discourage it. As Chicago’s Mayor Richard Daley may have demonstrated in 1960—when the delayed and disputed vote tally in Cook County, Illinois, provided the margin for John Kennedy’s victory—stealing a few thousand well-located votes can give you all of a state’s electoral votes and potentially the presidency.

The case that Edwards builds against the electoral college is strong, even overwhelming. Why the Electoral College Is Bad for America will be widely cited in debates about our method of electing presidents. Yet his arguments are so forceful that they pose sharply a question that Edwards never addresses: Why do we still have the electoral college? If the institution is flawed and the reasons for keeping it so feeble, why haven’t we junked this “jerry-rigged…Rube Goldberg mechanism”?5

Stated most broadly, the answer is that any change in our electoral system (or any electoral system for that matter) will mean short-run disadvantages for some political factions; and, in view of the large majorities required to amend the Constitution (two thirds of each house of Con- gress and three quarters of all state legislatures), those factions are likely to be able to block reform. Notably, and contrary to much popular punditry, small states have not been the main defenders of the college. In the late 1940s, for example, Republican Senator Henry Cabot Lodge of Massachusetts and Democratic Representative Ed Gossett of Texas sponsored an amendment calling for a proportional distribution of each state’s electoral votes (thus abolishing the unit rule). In 1950, the Senate passed the amendment by the requisite two-thirds majority. It was defeated, however, in the House, by a coalition of Republicans and Northern, liberal Democrats who had come to recognize that the immediate beneficiaries of the amendment would be Democrats in the one-party South. Although the more farsighted Lodge argued that the proportional plan might, eventually, help to build the Republican Party in the South, it was clear that, in the short run, the South’s electoral votes would remain solidly Democratic while the electoral votes of most Northern states would be split.6

Twenty years later, in 1969 and 1970, the US came very close to replacing the electoral college with direct, popular elections. An amendment with impressively broad support—from the American Bar Association, the Chamber of Commerce, the AFL-CIO, and the League of Women Voters, among others—was passed overwhelmingly by the House in 1969 and supported by the recently elected president, Richard Nixon. After being considered for over a year by the Senate (where the specter of George Wallace’s 1968 run for the presidency hung heavy over the debate), the amendment was blocked in 1970 by a filibuster led by Strom Thurmond of South Carolina, James Eastland of Mississippi, and Sam Ervin of North Carolina.7

This was the closest we have ever come to getting rid of the electoral college; and the amendment’s defeat owed much to the regional interests that have guaranteed the college’s survival for most of our history. To put the problem bluntly, it was never in the interest of the white South to agree to have direct popular elections so long as the region possessed a huge population of disfranchised African-Americans. That this was true under slavery was obvious and is well known: as noted above, thanks to the barbaric arrangement known as the three-fifths compromise, slaves counted as three fifths of a person toward the South’s representation in Congress and thus toward its weight in the electoral college. Had there been popular elections, the influence of the South in presidential elections would have dropped sharply, limited to the number of whites who were enfranchised and actually voted.

What is less widely recognized is that the white South benefited even more from the electoral college after slavery, particularly between the 1890s (when blacks were again disfranchised) and 1965, when the Voting Rights Act was passed. During those years, the region enjoyed what might be called the “five-fifths” clause: the African-American population counted fully toward congressional representation and thus electoral votes, but most blacks still could not cast ballots, thanks to devices such as poll taxes and literacy tests, which were expressly aimed at disenfranchising the black population. Had the electoral college been abolished in 1900, Southern states would have had far less power in presidential elections during the first two thirds of the twentieth century. In 1920, for example, there were 200,000 more votes cast in New York than there were in the entire South; but New York had only 45 electoral votes, while the states of the old Confederacy had 126.

With African-Americans now enfranchised in the South, Southern politics have changed substantially since 1965. But the South, once again dominated by a single party, remains a potentially critical roadblock in the path of electoral college reform. No matter how persuasive the arguments against the electoral college may be, Republican leaders may see little advantage in supporting popular elections or even a Colorado-like proportional plan: secure in their Southern majorities in the short run (except perhaps in Florida), they are likely to be deeply reluctant to let Democratic votes in the South, cast largely by African-Americans, contribute to a potential national Democratic majority.

It was partly for this reason that the possibility of President Bush winning the popular vote but losing the electoral vote was much discussed even before the vote count in Ohio.8 Such an outcome would not only have been revenge for the Democrats; it would also have tossed into a cocked hat all claims about the partisan advantages of the electoral college. Neither party could count on a system that would skew the vote one way in 2000 and the opposite way in 2004. The uncertainty of the partisan calculus might thus have given rise to a bipartisan effort to replace the system with the popular elections that the American people have long preferred. Presumably that is still a possibility; a political analyst as shrewd as Karl Rove might be expected to recognize that the Republican victory margin in Ohio was far too narrow for comfort. Yet rarely do victors dwell on the sources of near defeats.

Following the 2000 election, former President Jimmy Carter observed that it was “a waste of time to talk about changing the Electoral College…200 years from now, we will still have the Electoral College.”9 Carter’s pessimism was certainly understandable: in the 1970s he supported an amendment calling for direct elections that was badly beaten.

Yet we can hope that Carter’s prognosis was too bleak. The electoral college is a strange, archaic system possessing all of the flaws that George Edwards attributes to it. In addition, thanks to the unit rule, it deforms presidential election campaigns, saturating battleground states with nonstop advertising while leaving the rest of the nation watching a faraway spectator sport.

The problem with this is not simply, or even primarily, that campaign visits and advertising are restricted to a handful of states: few residents of uncompetitive states like New York or Georgia want to have their airwaves filled with attack ads. The issue instead is the future of civic and political participation in a nation where both are declining. The 2004 election may well have been, as many analysts claimed, the most important presidential election in decades. Yet there was no way for most citizens to engage in the campaign, other than by writing checks or moving to a swing state. In at least thirty-five states, with a majority of the US population, there was no reason for residents to argue with neighbors, go to public meetings, march in the streets, or drive people to the polls. For the young, there was not much chance to experience the mundane, yet instructive and even uplifting, work of getting out the vote. In the end, voter turnout was up, but at 61 percent it remained unimpres-sive for an election in which passions were so high: more people did not vote than voted for either Bush or Kerry. Still, it testifies to the democratic impulses of Americans that so many millions did show up at the polls, participating in an electoral system that they did not like, casting ballots that, in most cases, did not really count.10

It is not impossible that those same democratic impulses could prompt a surge of sentiment in favor of popular elections, leading to the passage of a new amendment. Many activist groups across the country are calling for the reform of the electoral college; and mainstream newspapers such as The New York Times have now come out in favor of direct elections. But it is difficult to maintain popular enthusiasm to change an institution that is used only once every four years; and there is little reason to believe that the parties and politicians now in power will promote reforms simply because the people want them and because there are strong arguments that they would benefit the nation.

This Issue

March 24, 2005