The Electoral College: How It Got That Way and Why We're Stuck With It

So it has happened again. A close presidential election has led to recriminations, cries of fraud, and talk of tainted mandates. Just as predictably, the 2000 election has inspired calls to reform the Electoral College—predictably, that is, because such proposals have followed every close presidential contest since the beginning of the Republic. The only difference is that this time no one asked why there’s such a long delay between election and inauguration.

The controversy goes back to America’s first contested presidential election, in 1796, when John Adams edged Thomas Jefferson by three electoral votes. On January 6, 1797—a month before the votes would officially be counted, though the results had already been leaked—Rep. William L. Smith of South Carolina introduced the first constitutional amendment to reform the Electoral College. Between Smith’s initial sally and 1889, the centennial of the Constitution’s adoption, more than 160 such amendments were introduced in Congress. From 1889 through 1946 there were 109 proposed amendments, from 1947 to 1968 there were 265, and since then, virtually every session of Congress has seen its own batch of proposals. Still, the Electoral College simply refuses to die.

More constitutional amendments have been offered to reform our procedure for electing Presidents than for any other purpose. Statesmen from James Madison, Martin Van Buren, and Andrew Jackson to Lyndon Johnson, Richard Nixon, Gerald Ford, and Hillary Clinton have endorsed an overhaul of the process. Opinion polls consistently show a large, sometimes overwhelming margin in favor of reform. Nonetheless, with the exception of a small procedural change in 1804, the Electoral College functions under the same rules today as it did in the horse-and-buggy era of 1789, when it was adopted. What accounts for the remarkable resilience of such an unloved creation? And why can’t we get rid of it?

In brief, the Electoral College works as follows: On Election Day, citizens in the 50 states and the District of Columbia go to the polls and vote for a presidential/vice-presidential ticket. Within each state, the candidate who wins the most votes gets to appoint a certain number of presidential electors, the number being equal to that state’s total seats in the Senate and House of Representatives (the District of Columbia gets three). This winner-take-all feature, which has caused most of the trouble through the years, is not mandated by the Constitution, but it is virtually universal; only Maine and Nebraska have laws that provide for their electoral votes to be split. In fact, the Constitution permits states to choose their electors by any means they want, and in the early days many of them left the choice to their legislatures. Since the 1830s, however, winner-take-all popular elections have been all but obligatory.

On a specified date in December, the electors assemble in their states and go through the formality of casting their votes for the candidates from the party that appointed them. Each state reports its totals to Congress, and in early January the Vice President opens and counts the votes in the presence of both houses. Whichever candidates receive a majority of the electoral votes are declared President- and Vice President-elect.

If no candidate for President has a majority (this can happen if there is an exact tie or if more than two candidates receive votes), the House of Representatives chooses a President from among the top three electoral vote-getters. In this process, each state’s congressmen combine to cast one vote, regardless of the state’s size, and the House keeps on voting until someone receives a majority. Meanwhile, if no candidate for Vice President has a majority of the electoral votes, the Senate chooses between the top two electoral vote-getters. That’s more important than it sounds, because if the House remains unable to make a choice from among its three candidates, the Vice President serves as President.

The first question that naturally arises when one is confronted with such a convoluted system is: Where did it come from? Most of us know that the Electoral College was adopted by the Constitutional Convention in 1787 as a compromise between large and small states. The large states wanted presidential voting to be based on population, as in the House of Representatives, while the small states wanted each state to have the same number of votes, as in the Senate (and the Constitutional Convention itself, for that matter). So they split the difference by giving each state a number of electors equal to its combined total of seats in both houses of Congress.

That was one reason for the Electoral College, but far from the only one. From the start, almost everyone favored some sort of indirect process for choosing a President. Although a few delegates suggested a direct popular election, the states had different qualifications for voting, and those with tight requirements—ownership of a certain amount of property, for example—worried that they would be shortchanging themselves in a nationwide poll. In particular, the Southern states had a large group of residents who were automatically disqualified from voting: slaves. (Something similar might be said about women, of course, but they were not concentrated in any one section.)

For purposes of allotting seats in the House of Representatives, the framers finessed this problem by counting each slave as three-fifths of a person. To retain the same measure of influence in a nationwide popular election, though, the South would have had to let its slaves vote. That, obviously, was out of the question. But with the Electoral College acting as an intermediary, the Southern states retained these “extra” votes based on their slave population. If not for the three-fifths rule, Adams would have defeated Jefferson in their 1800 election squeaker.

Slavery aside, there were other reasons the framers settled on an indirect scheme for choosing a President. Few of them thought the general public would be competent to make such a choice. George Mason of Virginia was particularly scathing in his denunciation of popular election. As summarized in Madison’s notes, “He conceived it would be as unnatural to refer the choice of a proper character for chief Magistrate to the people, as it would, to refer a trial of colours to a blind man.” This remark sounds supercilious until you read the next sentence: “The extent of the Country renders it impossible that the people can have the requisite capacity to judge of the respective pretensions of the Candidates.”

In a country without nationwide media, where traveling 20 miles was an arduous undertaking, this concern made ample sense. Even nowadays, how many Americans can name the governors of more than two or three states besides their own? Or consider the most recent election. Without television, would you have known any more about the Vice President than you know about the Secretary of Commerce? The world of the average eighteenth-century American was parochial to an extent that is unimaginable in the information age. To most of the framers, a popular vote for President would have been about as useful as drawing names from a hat.

With this in mind, the framers thought of the Electoral College not as a formality to ratify the popular will, as it is now, but as an assembly of respected figures (not unlike themselves) who would exercise their judgment to bring forth deserving candidates for the nation’s highest office. At one point, in fact, the Constitutional Convention considered a plan to have electors from across the country meet in a single place and hash things out as a body.

Also noteworthy is that in the original version of the Electoral College, electors did not specify one candidate for President and one for Vice President, as they do today. Instead, they put on their ballots two names for President, at least one of which had to be from outside their state. In this way, the framers thought, the electors could satisfy their local loyalties with one vote and use the other to recognize a man of national prominence. Under this system, if the first-place finisher was named on a majority of ballots, he would become President, and the second-place finisher—regardless of whether he was named on a majority of ballots—would become Vice President.

But that wasn’t supposed to happen very often. The most important point to understand about the Electoral College is this: The Constitution’s framers never actually expected it to choose the President. George Mason of Virginia thought the electors would give a majority to a single candidate only once in 20 times; later he amended this figure to 1 in 50. That’s how rarely most of the framers thought anyone would be well known and well respected enough across the country.

Almost always, they expected, the Electoral College would serve as a nominating committee, winnowing a large body of candidates down to the top five vote-getters (reduced to three in 1804), from whom the House of Representatives would make the final choice. The framers, then, saw the Electoral College chiefly as a mechanism for bringing candidates to nationwide prominence. It sounds very cumbersome and inefficient until you look at how we do the same thing today.

This explains why the Constitutional Convention spent so much time debating which house of Congress would choose the President if no one had an Electoral College majority. Nowadays that’s an afterthought, something that hasn’t happened since 1824, but the framers expected it to be the normal course of events. After considerable discussion, the final choice was given to the House, rather than the presumably aristocratic Senate. To appease the small states, though, each state was given a single vote without regard to its size.

During the ratification debate, the Electoral College inspired remarkably little controversy. As Alexander Hamilton wrote in The Federalist No. 68, “The mode of appointment of the chief magistrate of the United States is almost the only part of the system [i.e., of the entire proposed Constitution], of any consequence, which has escaped without severe censure, or which has received the slightest mark of approbation from its opponents.” Sure enough, the first two presidential elections went more or less as expected. Every elector used one of his votes for a figure of national prominence (in this case George Washington, though it was not expected that there would always be such an overwhelmingly obvious choice), and the second votes were scattered among a wide variety of local and national figures. In both elections, John Adams won the second-highest number of votes and thus the dubious honor of the Vice Presidency.

Even while Washington was in office, however, a change occurred that made a mockery of the framers’ vision of disinterested wise men carefully weighing the merits of the nominees. This was the development of political parties. Madison, in his classic Federalist No. 10, had praised the Constitution’s “tendency to break and control the violence of faction,” predicting that in a country as large and diverse as the United States, nationwide factions, or parties, were unlikely to form. Yet all theory went out the window almost as soon as the First Congress assembled. What Madison and his fellow framers did not realize was that the very existence of a government makes people align themselves one way or another, pro or con, like iron filings under the influence of a magnet. Any time you have ins, you will also have outs, and parties will form spontaneously around these two poles.

In recognition of this reality, the Twelfth Amendment, ratified in 1804, imposed the only major change that the Electoral College has ever seen. By then the failure of the founders’ vision was clear; in 1796 and 1800 electors had run as Adams men or Jefferson men, instead of standing on their own merits, as had been expected. Yet although the notion of a presidential/vice-presidential ticket had developed, electors still had to put two names on their ballots, both officially candidates for President.

In 1800 the duo of Jefferson and Aaron Burr won the election with 73 electoral votes against 65 for the Adams ticket. The trouble was that Jefferson and Burr each received exactly 73 votes, because every Jefferson elector had named both men on his ballot. The election went to the House of Representatives, where Jefferson’s opponents managed to forestall a majority until they finally yielded on the thirty-sixth ballot. (In this case, the House was restricted to breaking the tie between Jefferson and Burr rather than choosing from the top five vote-getters, as it would have done if no one had gotten a majority.)

To avoid a repetition of such a fiasco, the Twelfth Amendment required electors to specify separate candidates for President and Vice President. (A similar plan had been the subject of Representative Smith’s 1797 proposal.) Outside of this change, however, the rest of the Electoral College was left in place. Most Americans saw no need to open a can of worms by designing a new procedure from scratch.

After the excitement in 1800, the next five elections saw little controversy, with 1812 the only one that was at all close. Still, the inadequacies of the Electoral College—even in its new, improved form—were manifest. As Adams’s old Federalist party dissolved and new factions started to crystallize, the 1824 election promised to be splintered, and some observers wondered if the Constitution’s creaky old machinery would be up to the task. In 1823 Sen. Thomas Hart Benton of Missouri wrote: “Every reason which induced the convention to institute Electors has failed. They are no longer of any use, and may be dangerous to the liberties of the people.” That same year, James Madison, the father of the Constitution, candidly admitted the failure of his beloved progeny and suggested dividing the states into districts and having each district choose its own elector.

In fact, the 1824 election worked closest to what the framers had in mind, and it was a God-awful mess. Four candidates—Andrew Jackson, John Quincy Adams, William Crawford, and Henry Clay—received electoral votes, with none having a majority. Three New York electors who were supposedly pledged to Clay voted for other candidates, while two Clay supporters in the Louisiana legislature were unable to vote for electors after falling from their carriage on the way to the capital. This combination of treachery and bad luck bumped Clay down to fourth place, eliminating him from the balloting in the House, of which he was the Speaker.

At this point the normally fastidious Adams, who had finished second to Jackson in the electoral vote, put aside his scruples and began making deals for all he was worth. Adams won the House vote on the first ballot by a bare majority and immediately made Clay—whose support had swung Kentucky’s House delegation into the Adams column, though the citizens of that state had chosen Jackson—his Secretary of State. This led many to accuse the two men of a “corrupt bargain.”

Jackson, it is often pointed out, won the most popular votes in this election. But 1824 was the first year popular votes were widely recorded, and the figures are of questionable accuracy. The reported turnout was a derisory 27 percent nationwide and less than 15 percent in some states where the race was one-sided. On top of that, in 6 of the 24 states, the legislature chose the electors, so there was no popular vote.

The 1824 election was the last gasp for legislative selection, though. In 1828 only South Carolina and tiny Delaware still used it, and by 1836 every state except South Carolina (which would stubbornly retain legislative selection until the Civil War) had adopted the popular vote, winner-take-all method. Give or take a few small anomalies, then, the electoral system in place by the 1830s was identical to the one we are still using.

The dismay and outrage that have greeted the 2000 election were nothing compared with the public’s reaction to the 1824 disaster. When the next Congress assembled, a flood of schemes was offered to reform America’s procedure for electing a President. None of them got anywhere. And the pattern has repeated itself through the years: After a one-sided election, everyone shrugs off the Electoral College, and after a close election, everyone makes a fuss for a year or two, and then the issue fades away.

Through the years, numerous inadequacies of the Electoral College have come to the fore: potentially fractured multi-party elections (including 1912, 1924, 1948, and 1968); contested results (Hayes-Tilden in 1876 and Bush-Gore in 2000, plus a near-miss with Nixon-Kennedy in 1960); “minority” Presidents (1824, 1876, 1888, and 2000, with near-misses in 1960 and 1976); and “faithless” electors voting for candidates other than the ones they were chosen to vote for (as some Southern electors threatened to do in 1948 and 1960).

It’s safe to say that if you were designing an election method from scratch, it wouldn’t look like the Electoral College. Yet it’s worth pointing out what’s not wrong with our current system before we think about fixing what is. The famous 1876-77 Hayes-Tilden fiasco, for example, is not a good argument for abolition; it was the result of outright fraud and corruption, which could occur under any system. Indeed, the present Electoral College decreases the possibility for vote fraud (while admittedly increasing the payoff if it’s successful) by restricting it to a few states where the vote is close. In a direct nationwide popular election, votes could be stolen anywhere, including in heavily Democratic or Republican states where no one would bother under the current rules. In this way, the Electoral College acts as a firewall to contain electoral tampering.

It is also often said that under the Electoral College a popular-vote winner can be an electoral-vote loser. But this “problem” dissolves upon closer examination. Popular-vote totals are not predetermined; if they were, there would be no use for campaign consultants and political donations. Rather, the popular vote is an artifact of the electoral system. With a winner-take-all Electoral College, candidates tailor their messages and direct their spending to swing states and ignore the others, even when there are lots of votes to be had.

In the recent election, for example, neither presidential candidate made more than a token effort in New York, which was known to be safely in Gore’s pocket. To residents, it seemed as if neither man visited the state at all except to ask for money. Gore ended up receiving around 3.7 million votes to Bush’s 2.2 million. Now suppose Bush had campaigned in New York enough to induce 170,000 of those Gore voters, or less than 5 percent, to switch. He would have made up the nationwide popular-vote gap right there. Instead, both candidates spent enormous amounts of time and money fighting over handfuls of uncommitted voters in Florida, Michigan, and a few other states. That’s why in a close election, it doesn’t make sense to compare nationwide popular-vote totals when popular votes don’t determine the winner. You might just as well point out that the losing team in a baseball game got more hits.

As for faithless electors, not since the anomalous situation of 1824 have they made a difference in a presidential election. There is some reason to believe that if an elector broke his or her trust in a close race today, the switch would be ruled invalid. In any case, this problem can easily be eliminated with state laws or an act of Congress. These laws could also be tailored to take account of what happens if a candidate dies before the Electoral College meets or if a third-party candidate wishes to give his or her votes to another candidate. Flexible electors can even sometimes be useful, as in the three-way 1912 race, when some Theodore Roosevelt electors said before the election that if Roosevelt could not win, they would switch their votes to William Howard Taft.

Nonetheless, the flaws of the Electoral College, however exaggerated they may be, are clear. It magnifies small margins in an arbitrary manner; it distorts the campaign process by giving tossup states excessive importance; it gives small states a disproportionate number of votes; and perhaps worst of all, many people don’t have a clue about how it works.

Each of these except the last can be turned around and called an advantage by traditionalists: Magnified margins yield a “mandate” (though have you ever heard anyone who wasn’t a journalist talk about presidential mandates?); the need to pander to a diverse set of constituencies makes candidates fashion platforms with broad appeal; and after all, small states deserve a break. Still, nobody really loves the Electoral College—until a specific alternative is proposed.

The lack of agreement among would-be reformers has allowed the Electoral College’s vastly outnumbered supporters to defend it successfully against all attacks for nearly two centuries. Before the Civil War, slavery, called by its polite name of States’ Rights, stymied electoral reform in the same way it stymied so many other things: The Southern states would not consider any reform that did not increase their region’s importance in national elections, Oddly enough, by losing the war, the South got the influence it had always wanted.

From the end of Reconstruction into the 1940s, Democrats could count on a sure 100 to 120 electoral votes from the Solid South—the 11 states of the old Confederacy. Though the three-fifths rule was gone with the abolition of slavery, it had been replaced by something even worse, for while blacks were effectively disenfranchised in most of the South, their states now got full credit for their black populations in the House of Representatives and thus in the Electoral College. This allowed Southern whites not only to keep blacks from voting but in effect to vote for them. For most of a century after the 1870s, then, the Electoral College was a racket for the Democratic party.

Today the Solid South is a thing of the past. Nonetheless, since 1804 no electoral reform amendment has even made it through Congress. Why not? Who benefits from the Electoral College? Briefly put, two groups benefit: big states and small states. The winner-take-all feature favors the first of these groups, while the disproportionate allotment of electors favors the second.

With their tempting heaps of electoral votes, the big states attract by far the greatest bulk of the candidates’ attention. If you consider having politicians descend upon your state a benefit, the winner-take-all feature is a big plus. In 1966, in fact, Delaware sued New York (which then had the most electoral votes) and other states in hopes of forcing them to abandon the winner-take-all policy. A dozen other states soon climbed on board. Although the suit, which was based on the novel theory that a provision of the Constitution can be unconstitutional, was summarily rejected by the Supreme Court, it revealed the frustration that the small fry have always felt. In response, the small states cling to their three or four electoral votes the way an infant clings to its blanket. Since no one pays any attention to them anyway, they feel entitled to an extra vote or two.

Partisan considerations persist as well, this time on the Republican side. Today a group of Plains and Mountain states (Kansas, Nebraska, the Dakotas, Montana, Wyoming, Idaho, and Utah) can be thought of as a Solid West, reliably delivering most or all of their 32 electoral votes (as of 2000) to the Republican ticket, though their combined population is about equal to that of Michigan, which has only 18. As we have recently seen, those few extra votes can make a big difference if the election is close; and if the election isn’t close, any electoral system will do.

It’s impossible to say definitively whether the big-state or small-state advantage predominates, though that hasn’t stopped generations of political scientists from trying. But these two opposing factors explain how the 1970s notion of “urban liberal bias” and the 1980s notion of a “Republican electoral lock” can both be correct: The former results from winner-take-all, while the latter results from disproportionality.

Through all the analysis, reform proposals keep coming. They generally fall into three classes: a straightforward nationwide popular vote; election by districts, with the Electoral College retained but each congressional district choosing its own elector (and, in most such schemes, the statewide winner getting a bonus of two); and proportional representation, with electoral votes determined by each candidate’s percentage of the popular vote in a given state. Any of these would probably be better than what we have now, but each one has imperfections. Since every change would hurt someone, the chances of getting through all the hoops needed to pass a constitutional amendment—a two-thirds vote in each house of Congress plus approval by three-quarters of the states—look dim.

Direct popular election? First of all, there’s the question of what to do if no candidate receives a majority. Would there be a runoff, which would make the campaign season last even longer and might encourage third parties? Would the top vote-getter always be the winner—a system that could elect a candidate opposed by a majority of citizens? Would we mystify voters by asking for second and third choices?

Moreover, a nationwide election—something that has never taken place in America—would require a nationwide electoral board, with all the rules, forms, and inspectors that go along with it. Would states be allowed to set different times for opening and closing their polls? Would North Dakota be allowed to continue to have no form of voter registration, as it does now? Would a state seeking more influence be allowed to lower its voting age below 18? Then there is the potential discussed above for stolen or suppressed votes. Combine all these problems with the inevitable effect of concentrating candidates’ time, resources, and money on populous areas, and the case for a small state to support direct election looks mighty shaky.

Election by districts sounds appealing, but it would replace 51 separate races with about 480. Swing states would lose their all-or-nothing leverage, so candidates might concentrate on major population centers even more than they do now. (Under the present system, each new election gives a different group of swing states their moment in the spotlight, whereas with any other system, the big states would always get the bulk of the attention.) The effects of gerrymandering would be amplified, and third-party candidates would find it easier to win a single district than an entire state. Also, the small-state advantage would remain (and in fact be reinforced, since in most cases—all the time for the three-vote minnows—they would continue to function as units) while the big-state advantage from winner-take-all would vanish. In fact, if the 1960 election had been contested by districts and the popular vote had been exactly the same (a questionable assumption, to be sure), Richard Nixon would have won.

Proportional division of electors would be even worse, combining all the disadvantages of a direct popular vote with none of the advantages. Under this method, if a state has 10 electoral votes and Candidate A wins 53.7 percent of the popular vote in that state, then Candidate A is credited with 5.37 electoral votes. In essence, proportional division amounts to a direct popular vote, except that the votes of small-state residents are given added weight. And that’s the problem: By stripping the veil of illusion and ceremony and tradition from the Electoral College, this extra weighting makes the small-state advantage nakedly apparent, which infuriates one-person-one-vote fundamentalists.

But from the small-state point of view, proportional division would dilute the already tiny influence that goes with controlling three or four votes in a single lump. Also, there is a significant element of the public that views anything involving decimals as un-American—except baseball statistics, of course. Yet restricting the division of electors to whole numbers would be far more confusing, with different mathematical rules and minimum requirements in each state and often arbitrary results (if your state has four votes and the popular margin is 55-45, how do you divide them?). Proportional division would be fine for student-council elections at MIT, but to most American voters, it would amount to a mystifying black box.

To be fair, much worse ideas have been proposed. In the mist beyond proportional representation lies the wreckage of dozens of too-clever schemes, such as one cooked up in 1970 by Sen. Thomas Eagleton and Sen. Robert Dole (each of whom would within a few years take a personal interest in presidential elections). According to The New Republic, this plan provided that “a President would be elected if he (1) won a plurality of the national vote and (2) won either pluralities in more than 50 percent of the states and the District of Columbia, or pluralities in states with 50 percent of the voters in the election. . . .” And it went on from there.

In reviewing the history of the Electoral College, it quickly becomes clear how little anybody has to offer that is new. All the plausible reform ideas, and all the arguments for and against them, have been debated and rehashed for well over a century, in terms that have remained virtually unchanged. What has killed all the reform efforts has been the lack of a single alternative that all the reformers can agree on. As the politicians say, you can’t beat somebody with nobody, and you can’t beat one plan with three.

Moreover, the present system at least has the benefit of familiarity. Any change would be attended with an element of uncertainty, and politicians don’t like that. Opinions differ widely about who would gain or lose from electoral reform, but too many states and interest groups think they would lose and too few are sure that they would gain. After all, as we have seen, the original Electoral College functioned nothing like what its designers had expected.

In the end, Americans are likely to do what they have always done about the Electoral College: nothing. Every reform or abolition scheme works to the disadvantage (or possible disadvantage) of some special interest, and when a good-government issue collides with special interests, you know who’s going to win. Outside of academia and government, there is no obvious constituency for reform; since most people don’t understand how the Electoral College works, most of them don’t understand the case for changing it. The lack of exact numerical equality and other supposed biases have always bothered political scientists much more than the average citizen, who may endorse reform when questioned by a pollster but will hardly ever feel strongly about the issue.

So we’re probably stuck with the Electoral College until the next close election, when reformers and abolitionists of various stripes will once again surge forth, only to end up annihilating each other. To break this pattern, someone will have to either find a novel and compelling set of arguments for reform and waste enormous amounts of political capital to pass a measure that arouses no public passion and has no clear-cut beneficiary, or else devise a new scheme that is simple enough to be grasped by the average citizen yet has never been advanced before. Good luck.