William D. Bramhall Jr.

J anuary 20, 1981. For the first time in history the Inaugural stand has been built on the West Front of the Capitol, facing Pennsylvania Avenue and the White House a mile away. But at noon, the time fixed for the Inauguration, a lone workman standing on the Inaugural platform sees only the normal traffic of a midwinter midday. The platform is ready, but there is no new President—none of the candidates carried a majority of the electoral vote on November 4, 1980.

Behind the West Front, the House of Representatives, as the Constitution provides, is trying to decide which of the top three candidates in the electoral vote should become the fortieth President. In fact, the House has been trying for over five weeks, without success, to break that deadlock. Ronald and Nancy Reagan are waiting across the river in Virginia in a rented house that once belonged to John and Jacqueline Kennedy. John and Keke Anderson are in their modest Bethesda home; he has been third in the House voting all along, but he still hopes to be the compromise choice. This morning, Jimmy Carter and Rosalynn moved out of the White House into the Madison Hotel and a unique political twilight: he is not quite President and not quite ex-President.

Walter Mondale today becomes Acting President. Almost immediately after the electoral deadlock, the Senate carried out its constitutional duty and named him over George Bush for Vice President. (The House must select from the top three presidential finishers in the Electoral College, the Senate from the top two vice-presidential finishers.) Now Mondale has a presidency—of sorts. He will keep it—for a day or for four years—until the House makes up its collective mind. No one thought it appropriate for Acting President Mondale to be sworn in amid ruffles and flourishes from the splendor of the Inaugural platform. Instead, at noon, he takes his oath quietly, in private, with only one television camera permitted to film this unique ritual of American public life. He gives no Inaugural address; he only hopes, he says, that the House will soon finish its work.

This scene is not idle fantasy. Two presidential elections have been decided in the House of Representatives and four others, including the elections of Abraham Lincoln in 1860 and John F. Kennedy in 1960, have come within 30,000 votes of requiring a decision by the House. Three others, in 1912, 1924, and 1968, came close. In 1980, a victory by independent candidate John Anderson in just a few key states could throw the election into the House. If Anderson wins only the thirty-nine electoral votes of Massachusetts, New Jersey, and Connecticut—three states in which Ronald Reagan's polls have shown Anderson leading—Jimmy Carter could end up with 230 votes from the urban Northeast and part of the South, forty short of the 270 needed for a majority, and Reagan could take the rest—including the West and Midwest—for a total of 268 electoral votes, two votes short of the presidency. In fact, former President Gerald Ford flatly predicted during the Republican convention in July 1980 that the Anderson candidacy will throw the election into the House of Representatives. And responsible analysts think it unlikely that any of the candidates could win the votes of a majority of the state delegations in the House. A deadlocked House election is not a sportive daydream; many consider it a nightmare.

Americans have awakened to the prospect of an election that fails to elect with a sense of fear—a feeling that a House election is proof that we, our politics, even our Constitution, have somehow failed. There is a pervasive and unquestioned reaction that it is a disaster to make a President this way, just as many feared that it was a disaster in 1974 to unmake the President by invoking the impeachment power for the first time in more than a century, after teaching generations of schoolchildren to regard impeachment as a dread instrument and a dead letter. Mr. Ford, our only appointed President, agrees with the view that an election of the President by the House would be "tragic," even though such House election is a device in which the Constitution's designers took pride. Politicians, political scientists, reporters, all haunted by the specter of an electoral deadlock, have proposed vote-switching, restricted media coverage of third-party candidates, private agreements, congressional resolutions, even constitutional amendments, to prevent a presidential election by the House of Representatives. Reaction before the 1968 election, the most recent one that seemed headed for Congress, foreshadowed today's fears: an article in the September 1968 Atlantic pleaded with voters to "Keep It Out of the House!"

That the odds of a House election seem high in 1980 is not just a function of Anderson's strong showing in the polls; those odds will remain high and grow higher as the political gaps left by our major parties continue to widen. But before we rush to dismantle the present system, it would be wise to review the history that has brought us here. Like an impeachment, an election in the House is a tempting source of terror-mongering because it seems so alien: 155 years have passed since the House last chose a President in 1825, 179 years since the election of 1800 was decided on the thirty-sixth ballot by the House of Representatives. But that is not so long ago in a constitutional lifetime, and it is by borrowing the time-view of our most enduring political document that we can absorb the lessons of its experience. That experience teaches that our fears may be more a product of reflex than reflection, and that the system itself did not cause the crises linked with earlier deadlocks or near-misses. Each such crisis resulted from a fractured political consensus which the electoral system simply mirrored. Nor does history show that our forefathers have forsaken us by leaving behind what anyone would regard as an unworkable way to select a leader: no one disagreed or was distressed when George Mason predicted to the Constitutional Convention in 1787 that nineteen out of twenty elections would be decided by the House. And there is nothing unique about a President without a majoritarian mandate; nearly half our Presidents have lacked a popular vote majority.

The Electoral College, dominated by the large states, converts popular pluralities and even minorities into clear victories. The House and Senate have the same function one step removed: through a voting mechanism dominated by the small states, they turn electoral pluralities and even minorities into Presidents and Vice Presidents.

In its initial conception, that system was bottomed on the shakiest of political premises. The Constitution was written as if political parties were a scourge the nation must and would avoid; but the precursors of today's parties in fact emerged as early as the first presidential election in 1789—and immediately started to unravel the Framers' design. Parties seized control of state legislatures and muted the voices of opposing minorities by adopting the winner-take-all systems now in use in every state: suddenly all of a state's electoral votes, with only a rare stray here or there (like the Ford elector who cast his lot with Reagan in 1976), went to the winner of even the narrowest popular plurality. The Framers had expected that a district system, which would allow each state to divide its electoral vote, would be used. Instead, as parties organized the way electors would vote, they eliminated the scatter of votes that the Framers expected and quickly produced a system in which everyone expected elections to be decided before they ever reached the House.

Almost as quickly, this produced a crisis that made clear the need to amend the Constitution to provide for separate elections for President and Vice President. For the original plan had a fatal flaw that emerged with the rise of the parties. Since the electors could not designate which candidate they preferred for President and which for Vice President (the runner-up in electoral votes won the vice presidency), a host of dangers was created by party voting. If the opposition gave some unwanted extra votes to the vice-presidential candidate, or if malefactors within the party withheld votes from the presidential candidate but gave them to the vice-presidential nominee, a party's presidential and vice-presidential candidates could be reversed. If the leading party threw away some of its vice-presidential votes to prevent a reversal of its candidates, the opposition could win the vice presidency. In 1796, Democratic-Republican Thomas Jefferson was in fact elected Vice President this way. And if no votes were thrown away and the opposition did not boost the votes of the vice-presidential candidate, then the two candidates' totals could tie and the election would go to the House. This is exactly what happened in 1800, when the Democratic-Republican party ran Jefferson and Aaron Burr with the clear understanding that Jefferson, not Burr, was the presidential candidate. All seventy-three of the Republican electors voted for Jefferson and Burr—and then, in a context the Framers never intended, the House had to decide. Alexander Hamilton, who opposed Burr's attempt to seize the presidency, was killed by Burr in a duel in 1804.

The operation of political parties had thus turned the game of presidential politics into a team sport. The players had to change the rules, and did so with the Twelfth Amendment, which reshaped the Electoral College in 1804: henceforth, electors would vote separately for President and Vice President. Two more recent amendments modified the system further. In 1933, the Twentieth Amendment specified that the new Congress would meet seventeen days before the new President's Inauguration. By statute, the new Senate and House, not the lame ducks, pick the President and Vice President if the Electoral College produces no majority. The second change, a relatively minor one, came in 1961: the Twenty-third Amendment granted the District of Columbia the right to cast three electoral votes, although the District still has no representatives or senators and could not vote in any House or Senate election.

Disputed elections

The Constitution's election rules may seem specific, but they leave open as many questions as they answer. And it is in the gray areas not illuminated by the Constitution that the battle for the executive branch may be fought again—and has been won before.

Some of those gray areas, potentially touching every election whether or not it goes to the House, hardly seem to be the stuff of which high political and legal drama are made. But in 1980, the election could turn on the seemingly innocuous constitutional mandate that "the President of the Senate shall, in the presence of the Senate and House of Representatives, open all the Certificates, and the Votes shall then be counted." Who counts the votes, and which votes count? Here, the Constitution is silent; yet the question could become crucial if a state's electoral votes are disputed. The margin between Anderson and Reagan in a state such as Illinois could be a matter of a few thousand votes—narrow enough to invite the tombstone voting that was reported in Chicago in 1960. Or a Republican election board in a critical state could exploit a narrow margin between Anderson and Reagan to certify the state's votes for Reagan. Congress would then have to decide somehow whether to recount the votes.

Although we have arrived at that point only once, in 1876, vote manipulation in presidential elections goes back to the very beginning. In 1800, Aaron Burr circumvented New York's requirement that voters own a minimum amount of property by persuading landless Republicans to pool their funds and purchase enough as "joint tenants" to meet the requirement. The special magic of the joint tenancy was that each tenant, no matter how large the group or how small his contribution, "owned" the entire estate. The Federalists responded by locating a loophole in New Jersey law, which did not specifically exclude women from voting. They marched their wives, daughters, and any other females they could find to the polls and buried the male Republican vote.

By 1876, Jefferson's Republicans had been transmuted into the post-bellum Democrats and the Federalists into the new Republican party, but vote fraud on both sides endured. And in 1876, the Republicans' tactics almost started a second civil war. The election returns on November 7, 1876, showed that Democrat Samuel Tilden, the governor of New York, had won a landslide victory over Republican Rutherford B. Hayes. But the Republicans charged that the Democrats had won the votes of South Carolina, Florida, and Louisiana through force and fraud to intimidate and disqualify newly enfranchised black voters. The Democrats challenged one Republican elector in Oregon, leaving Hayes with 165 undisputed electoral votes and Tilden with 184, one short of the 185 votes he needed.

The Republican-dominated governments of the disputed states sent new electoral totals to Congress, showing Hayes the winner of every disputed contest. The Democrats in those states, two of which were ruled by rival Democratic and Republican governments, submitted electoral returns of their own showing Tilden to be the victor. The Constitution did not say which votes to count. To resolve the dispute, Congress established a fifteen-member Electoral Commission. The bill creating the commission required both the House and the Senate to vote on each commission decision; either house could thus delay the verdict past Inauguration Day. This provision gave both sides an incentive to bargain—and what the Republicans promised was nothing less than an end to Reconstruction in the South. Hayes also offered one or two Cabinet posts for the South, aid for a southern railroad, and increased spending in that region for internal improvements.

However, there were doubters and dissidents in the Democratic-controlled House. They launched a filibuster to prevent Hayes's Inauguration by postponing beyond Inauguration Day any House vote on the commission's ruling, which had awarded all twenty disputed electors to Hayes. The House session of March 1, 1877, just three days before the scheduled Inauguration, was described as "probably the stormiest ever witnessed in any House of Representatives." The galleries were packed; members shrieked for attention; one representative jumped up and down on his desk in an incoherent rage; lobbyists swarmed the House floor. Representative William Levy of Louisiana then rose to announce that he had received "solemn" assurances that Hayes would withdraw federal troops from the South. After eighteen hours, the filibuster died out. At 4:10 on Tuesday morning, March 2, 1877, the count was completed: Hayes had won by 185 votes to Tilden's 184. President Hayes promptly ordered an end to the federal occupation of the southern states.

The electoral crisis of 1876 mirrored a national crisis over the legacy of the Civil War. The Republicans traded away Reconstruction. Freed of federal interference, the South took its revenge. Blacks were disenfranchised and segregated. The crisis of race relations had been postponed to the next century. Eight decades after the Great Compromise of 1877, another Republican President, Dwight Eisenhower, had to send federal troops back to the South, to Little Rock, Arkansas, to force the integration of Central High School.

Except in cases as extreme as the one in 1876, political realities generally dissuade losing candidates from potentially bitter challenges to the electoral results. Yet challenges remain a real possibility, and there is no settled procedure for disposing of them. The need for improvisation has often been avoided only narrowly: no fewer than twenty presidential elections would have come out differently if less than one percent of the vote had shifted. The margin of victory can be all but invisible. A shift of 116 votes to Tilden in South Carolina in 1876 would have given him the one electoral vote he needed to defeat Hayes. A switch of 575 votes in New York in 1884 would have made James G. Blaine President rather than Grover Cleveland. Woodrow Wilson would have lost the election of 1916 to Charles Evans Hughes if Hughes had won the votes of 2000 more Californians.

Electoral bargaining

The popular vote counts for very little as a national total, but makes all the difference in the Electoral College, because a tiny margin in a few states can deliver a decisive block of electoral votes. Fifteen Presidents have won without a popular majority, but many of them have enjoyed comfortable majorities in the Electoral College. Or apparently comfortable majorities—for the system invites political dealing and electoral bargaining.

The invitation to bargain could be irresistible in a case where three candidates split the electoral vote, with no one receiving the necessary majority. What sort of bargain might be struck in a disputed or deadlocked election in 1980? One thing is clear: the public would not necessarily regard every possible bargain as a political deal. The Hayes-Tilden result was seen as the Great Compromise of 1817, which, like the Missouri Compromise of 1850, settled sectional differences without resort to civil war. But the 1877 bargain was elevated to the status of statesmanship because it centered on a single, overriding question that divided the nation. The Republicans could gain the presidency by yielding to the Democratic assault against Reconstruction.

Such bargains could be struck in 1980 without House or Senate involvement; a candidate might simply persuade the electors chosen to support him on November 4 to cast their ballots for someone else. Indeed, electors could do so on their own, since the Constitution makes them free agents. In every state, they run on a party or independent-candidate slate; but no pledge in advance to any party, to any candidate, or to the voters binds the electors after their slate has carried the state. They could even vote for someone who had not run in the November election—perhaps Gerald Ford.

The manipulation of electoral votes is as old as the history of the presidency—despite the Framers' attempt to sequester the electors in separate state capitals, all voting on the same day. In the first four elections, electors made agreements to throw away votes to prevent the accidental election to the presidency of their vice-presidential choice. The failure of one Republican elector to do just that caused the crisis of 1800. In 1960, fourteen Democratic electors from Alabama and Mississippi and one Republican elector from Oklahoma cast their votes for Senator Harry F. Byrd of Virginia, who was not even a declared candidate for President. Democratic ballots in Alabama and Mississippi were "blind"—voters cast their votes only for parties, not for candidates. If the electoral count between Kennedy and Nixon had been closer, the Byrd electors could have negotiated their support for concessions on civil rights. Or they could have forced the election into the House, where southern delegations could have tipped the election either way.

In 1968, George Wallace publicly offered to barter with his electoral support. Wallace's electors pledged in writing to vote as he told them to. In a celebrated press conference on February 19, 1968, he named his price for delivering the presidency to someone else: the criminal indictment of anyone advocating Viet Cong victory; the elimination of the federal antipoverty program; cuts in foreign aid to any nation that refused to support the United States in Vietnam; a tough stand on law and order; the repeal of all civil rights legislation; the appointment of "differently oriented" judges to the Supreme Court; and a return to the states of all power over housing, school, and hospital integration, over reapportionment, and over congressional redistricting.

As the campaign wore on, Nixon told reporters he was sure that neither he nor Hubert Humphrey would ever make a deal with Wallace. Humphrey, whose famous speech on civil rights at the 1948 Democratic convention had provoked Wallace and the Alabama delegation to walk out of that convention, insisted that he would never bargain with Wallace, saying: "If there's any office in this country that ought to be above any kind of deal with Mr. Wallace ... it's the presidency. I'm a no-deal man." On October 14, 1968, Humphrey's campaign manager, Lawrence O'Brien, suggested that "secret negotiations" had already begun between Nixon and Wallace. Two days later, it was reported that emissaries of Nixon and Wallace were negotiating in New York—and that Nixon campaign manager John Mitchell was pushing the strategy. But because Nixon won decisively in the Electoral College, he never had to pay Wallace for the presidency.

In 1980, it seems unlikely that John Anderson would or could barter his independent electors to the highest bidder. But there are other possibilities. The Anderson electors could vote to put another candidate over the top rather than let the election go to the House. Indeed, they could seek a deal with at least a tint of statesmanship. They could offer to spare the nation a House election by trading votes for President in return for all of the Democratic votes for Vice President: the bargain could result in a Carter-Anderson administration. Or the Anderson electors could move the contest into the House, but with a new choice altogether. Anderson's Massachusetts electors, for example, could vote for Edward M. Kennedy and put him in third place and the House runoff—even though he had not been a candidate in November.

Electors for any candidate could shift to prevent a deadlock as the deadline draws near. Because television admits all of them into a single, electronic Electoral College that the Framers never foresaw, electors on the West Coast have at least three hours on December 15 to decide whether to shift any of their ballots after all the electoral votes of the industrial Northeast have been cast in nominal secrecy—but almost certainly deciphered and broadcast within minutes by the media. Thomas Jefferson once argued that if the Electoral College did not reach a majority decision on the first attempt, it should be given a second try. In effect, television could give midwestern and western electors that chance—with plenty of advice on how to use it. Although few electors have been faithless in the past, few elections have been close enough in electoral votes to tempt electors to bolt.

The turbulent politics of 1980, which inspired cries for an "open" Democratic convention where delegates could vote their conscience, could lead finally to an "open" Electoral College where electors would abandon their pledges out of a higher fidelity to the national interest—or from less lofty motives. How many would resist the lure? By the time the electors meet in their state capitals in December 1980, more than a month after the November election, an uncertain America could be ready and enough electors could be willing to have an open Electoral College.

Either by betraying their trust or by keeping it, the electors may deny any candidate a majority, and the House of Representatives would ballot for a President for the third time in 180 years.

To the House

The Constitution gives each state delegation in the House one vote, with a majority of twenty-six states required for victory. But the Constitution leaves open a plethora of questions that could determine the outcome. The three-day-old Ninety-seventh Congress convenes at 1 P.M. on January 6, 1981, to count the electoral votes cast on December 15, 1980, and has just two weeks in the event of an electoral deadlock to select a President before Inauguration Day. So the lame duck Ninety-sixth Congress would probably act between this November 4 and the Christmas recess to pass the rules governing that selection. A party that is losing power from the lame duck Congress to the new one, such as the Federalists of 1800, may write the rules to make it harder for the opposition to elect its candidate when the new House of Representatives meets to pick the President. For instance, would votes be by sealed ballot or by voice roll call? Voice voting would allow strategic switching by states that come late in the alphabet. Would individual votes be secret? Such secrecy would work to the disadvantage of representatives from single-district states, whose votes would necessarily be known, but it would greatly reduce the effect of bargains: who wants to deal when there is no way to know if the other side has welshed? Would the votes of a majority of a state's representatives or a mere plurality be required to cast a state's vote? Requiring majorities may leave large states—where Carter, Reagan, and Anderson might all do well—unable to cast a vote. Plurality voting in the House, on the other hand, runs the same risk as the Electoral College itself: by eking out a victory in enough states, a minority candidate could still out-total an opponent who enjoyed overwhelming support everywhere else. Under the House rules passed in 1800 and 1824, majorities were required, individual votes were recorded, and states voted by sealed ballot. But these precedents are not binding.

The Constitution requires that in the event of an electoral deadlock, the House must choose a President "immediately." Does this mean that the House must vote continuously, or can it pause for rest and other business? The Founders required an immediate choice so that there would be no time for deals to be struck, but if no choice results on the first ballot the House proceedings may, as one observer recalled of 1801, resemble a dance marathon more than an election: "The scene was now ludicrous. Many had sent home for night caps and pillows, and wrapped in shawls and great-coats, lay about the floor of the committee rooms or sat sleeping in their seats. At one, two, and half-past two, the tellers roused the members from their slumbers, and took the same ballot as before." The House voted for the twenty-seventh time at sunrise. After this heroic attempt to follow the Constitution to the letter, the representatives slowed to the pace of about a ballot a day, and James Bayard of Delaware claimed that promises made in the pauses between ballots won his vote and the election.

Perhaps the most dangerous possibility left open by the Constitution is that the lame duck Congress could try to move the time for counting electoral votes back to a date before January 3, when the new Congress convenes, so that the old Congress could choose the new President. A shift of 12,000 popular votes in 1948 would have switched enough electoral votes to send the presidential election to the House and the vice-presidential election to the Senate. That year, the Republicans lost their majorities in both houses of Congress. The outgoing Republican Eightieth Congress, with a "do-nothing" record that had become the central issue in Harry Truman's campaign for re-election, could have responded by moving up the date for picking among Democrat Truman, Republican Thomas Dewey, and States' Rights candidate Strom Thurmond. In any year, this tactic would surely stir popular protest, but a partisan Congress could decide to take the heat: the re-elected members are likely to be from safe districts; the lame duck members have little or nothing left to lose.

Fortunately, the background of the Twentieth Amendment strongly suggests that the loophole left by its language is plugged tight by its history: the whole point of that amendment's provision for the new Congress to convene seventeen days before the new President takes the oath was to prevent a lame duck Congress from selecting the President or Vice President. When history and principle are this clear, the silence of the Constitution's text becomes almost irrelevant. The document should be interpreted to forbid lame duck manipulation of the presidency, although a partisan Congress might decide differently—and it is anyone's guess how far the courts would go to halt the lame ducks as they tramp across the spirit of the document for their own narrow ends.

The House decides

Because representatives would have so many options, voters in 1980 are likely to ask candidates for the House how they will vote if the choice of President goes to the House. Seeking a defensible answer, beleaguered House members will find no instruction in the Constitution and no guidance from the history of constitutional interpretation. In the long history of deadlocks and near-deadlocks, the members have invented five alternative standards for themselves.

1. Party lines. Many House members, left to their own preferences in selecting a President, would vote their party loyalty. The Democrats now hold 63 percent of the seats in the House. They control twenty-nine state delegations; the Republicans control twelve; nine others are evenly split. Whichever party controls the Ninety-seventh Congress may yet fear the voters' wrath if it selects its party's nominee despite his poor second-or third-place showing. One ABC News-Harris survey taken in August 1980 showed Reagan in first place nationally, with Anderson second and Carter third. What if the final vote shows a photo finish between Anderson and Reagan, with Carter still in third place? In 1801, the Federalists dominated the House and tried to elect the more nearly Federalist of the two Republican finishers. The Democrats in 1981 might do the same by rallying to Anderson.

2. Popular vote. In 1968, with George Wallace threatening to deny either Humphrey or Nixon an electoral majority, Representatives Morris Udall of Arizona and Charles Goodell of New York sponsored a bipartisan plan under which candidates for the House would pledge in advance that they would support the winner of the national popular vote. The Washington Post endorsed the idea, saying it would protect the presidency from charges that a corrupt bargain had delivered the nation's highest office to the highest bidder.

John Anderson was a member of the "People's Presidential Committee" formed by Udall and Goodell to promote the plan. Senator Henry Jackson seconded it. But Governors Ronald Reagan of California and Paul Laxalt of Nevada (now Reagan's campaign chairman) were both opposed. Reagan's tone was uncompromising: "I have too much faith in the people to shut someone out in the beginning. I have no intention that I would collaborate with the Democrats on anything." George Wallace called the plan "a conspiracy to circumvent the Constitution." Wallace's point had some merit: if the Constitution contemplated anything like a mechanical choice of the popular-vote winner when nobody wins and electoral majority, no House election would have been provided by the Framers. The popular-vote plan suffered most because the major party candidates, Humphrey and Nixon, would not commit themselves to it. As might be expected with any proposed pledge, neither wanted to sacrifice whatever advantages he imagined he might enjoy in an unpledged House election. On October 15, 1968, when it seemed clear to Nixon that he would carry a popular majority, he said he thought the popular-vote winner should be elected by the House. "If the man who wins the popular vote is denied the presidency," Nixon said, "the man who gets the presidency would have very great difficulty in governing." But five days later, Humphrey, who was certain the Democrats would control the House no matter how he fared, appeared on Face the Nationto say that he wanted the members of the House to be bound by no mechanical pledge and to pick the President they believed would be best for the country. He was sure such a President would "still have the capacity to govern," even if he had not won a popular plurality.

This time, the second-place finisher might bargain his way to the top of a House election. Whether he thereby purchases a crippled presidency will depend not only on his failure to obtain a popular plurality (such Presidents as Lincoln, Wilson, and Truman have governed successfully after such failure) but on whether he traded for the Oval Office as if it were a piece of local patronage.

3. District winner. To improve their chances in their own elections, many House candidates may promise in advance to support the presidential nominee who carries their district—something conservatives have demanded for years, most recently in a plan proposed in August by the American Conservative Union. The criterion could be especially appealing because the winner of the national or state popular vote may not do well in a particular representative's district, and enough voters cross party lines for it to be a political liability in many cases for a representative to vote on a party basis in a House election.

Colorado's five-member House delegation typifies the resulting problem. It is divided three-to-two for the Democrats. Democratic Representative Tim Wirth, who faces a tough re-election fight in 1980, comes from a district that Jimmy Carter lost to Gerald Ford in 1976 by 37,000 votes. For political self-preservation, Wirth might feel forced to announce that he will vote the presidential preference of his constituents, thus probably giving Colorado's vote to Reagan in a House showdown even if Carter or Anderson carries the state on November 4. In fact, if the 1976 presidential election had gone to the House and all representatives had voted for the candidate who carried their districts, Carter would have had the support of 220 representatives and Ford 215. But because presidential selection in the House is by state and not by district, Ford would have won the presidency: although outvoted both in the popular total and in a majority of districts nationwide, Ford carried a majority of the districts in twenty-seven states and Carter in only twenty-three. House balloting guided by district winners thus emerges more as a temptation than as a goal.

4. Inspiration. There is a way for House members to vote that, in the long run, may be the safest of all—the "Oh, my God!" system. It has been used only once. In 1825, John Quincy Adams needed the votes of thirteen states to be elected by the House. He had twelve states, but some of them were shaky and seemed inclined to shift to Andrew Jackson on later ballots. Adams needed a first-ballot win. The best hope for the thirteenth state was New York, whose thirty-four congressmen were split nearly down the middle: seventeen for Adams, sixteen for Crawford, and one undecided. Representative Stephen Van Rensselaer, a kindly old aristocrat, had promised to support everyone; he held the swing ballot in the decisive state. On the morning of February 9, 1825, the day of the House vote, Daniel Webster and Henry Clay, the two most persuasive men in America, cornered Van Rensselaer in the speaker's office and plied him with threats and oratory, demanding that he vote for Adams. The old gentleman held out.

When the House balloting began that afternoon, Van Rensselaer was still in a quandary. As the rest of the New York delegation passed their ballots to the teller, he bowed his head in prayer, seeking a sign from on high. He opened his eyes, and there on the floor was a ballot marked for John Quincy Adams. "Oh, my God!" Van Rensselaer snatched up the ballot and cast it as his own. So Adams was elected President. Not even the 1980 field of three "born-again" presidential candidates is likely to embrace reliance this time on such divine clues. But whether supposedly more "rational" methods are truly wiser is beyond the capacity of political and legal theory to say.

5. Bargaining. Representatives overwhelmed by the absence or multiplicity of heavenly hints and rational solutions alike may decide that self-interest is the better part of politics and vote for the candidate making them the offer they find hardest to refuse. And many representatives may each be in a position to deliver or withhold the presidency as their half of the bargain. In 1912, Teddy Roosevelt's Bull Moose party split the Republicans and finished second. If that election had gone to the House, each of thirty-two different representatives would have held the power to change an entire state's vote by switching his own. In the 1924 election, when Robert La Follette's Progressive party won 17 percent of the popular total, each of eighteen states pivoted on one vote; in those states, forty-five Democrats and twenty-five Republicans individually held the balance of power. In 1948, a Truman-Dewey-Thurmond runoff would have been required had 12,000 votes shifted in California and Ohio, and twenty-three state delegations would have turned on one member's vote; sixty-five Republicans and thirty-three Democrats each had the power to deliver an entire state. And if the Ninety-seventh Congress duplicates the Ninety-sixth, twenty-eight state delegations this time will be held by a one-vote margin and eleven state delegations by just two votes each.

Bargaining, of course, can be portrayed either as sensible compromise or as dirty dealing. Both Presidents elected by the House, Thomas Jefferson and John Quincy Adams, passionately denied that any deals were made. After Delaware's lone representative, James Bayard, gave Jefferson his 1801 victory, the new President kept certain Federalist officeholders, as Bayard had been promised, but insisted that he had made no promise to Bayard that he had to keep. Two years later, in congressional testimony, Bayard named the four offices that Jefferson had delivered as pledged. When Adams carried the House with Henry Clay's support in 1825, the appointment of Clay as secretary of state was quickly seen as proof of a "corrupt bargain." Clay had met with Adams for three hours at his home and had emerged as Adams's ally. No one can be certain what the two men said, but both of them must have been tempted to strike a deal, explicit or implicit. Adams had only to say that Clay was the logical choice for the State Department which was true. Clay, in turn, had only to state his feelings—that he preferred Adams because he feared the alternatives: Andrew Jackson was a warmonger, and William Crawford had been crippled by a drug overdose during the campaign. Though they furiously denounced the charge that a "corrupt bargain" had been struck, Adams and Clay could never escape it.

Whether deals are made or not, a loser in the House will probably allege and the people will probably believe that the winner made one.

"Deal" is a bad word, but not all deals are created equal. Depending upon the observer's point of view, they can seem good or evil. If the 1948 election had been decided in the House, four southern states controlled by the Dixiecrats could have made a majority for either Truman or Dewey. The price they exacted would have been an immoral weakening of civil rights—which was the price Rutherford Hayes gladly paid in 1877. But the Compromise of 1877 was not everywhere regarded as a hidden and awful bargain. The New York Herald called it an "open secret." The Memphis Avalanche insisted in an editorial: "There is no bargain in this movement. It is a policy." In 1980, women in the house could agree to support Reagan if, for example, he dropped his commitment to a constitutional amendment outlawing abortion. Deals are generally unpopular—except with the people who are being dealt in. George Wallace called the deal he offered in 1968 "a solemn covenant."

A House divided

While a refusal to trade for votes may be noble, it may also leave us without an elected President. Many of the elections that nearly went to the House could have deadlocked in just this way. The 1860 election narrowly avoided a House decision among Republican Abraham Lincoln, Southern Democrat John Breckinridge, and John Bell of the Constitutional Union party. The House that year was split 17 to 15 in favor of the Democrats; one delegation was evenly divided; and the American party controlled the Tennessee vote. Without some bargain, there might have been no decision at all on the very eve of the Civil War. In 1912, when Woodrow Wilson prevailed against William Howard Taft and Theodore Roosevelt ran as an independent, the House was split 23 to 22 for the Democrats, with three delegations divided equally. In 1948, as in 1960 and 1968, the Democrats controlled the House, but only by virtue of southern delegations that could readily have voted for the more conservative candidate—the Republican.

In 1981, it is entirely possible that the House could deadlock. Some strongly Democratic states, such as Massachusetts and New Jersey, could vote for John Anderson on November 4 and their representatives might feel compelled to go along. At the same time, the national shift to the right favors Republican House candidates this year. Republican leaders hope to gain at least thirty House seats in November. The July 1980 ABC News-Harris poll showed the Republicans leading the Democrats by 47 to 43 percent in the race for the House—the first time since 1952 that the Republicans have been ahead. In six southern states, a gain of one Republican seat would give the Republicans a majority or divide the delegations evenly. As matters stand, with twenty-six states needed to elect, only twenty-one are now clearly Democratic, eight are narrowly Democratic, nine divide evenly, and twelve are Republican. These numbers show that 1981 could be the year when democracy deadlocks.

In 1801, the House voted thirty-five times without breaking the stalemate. In 1877, the debate in the House came within twenty-four hours of delaying the selection of a President until after Inauguration Day. If the House does not choose a President by noon on January 20, 1981, then the Vice President becomes Acting President. But if there is no electoral majority for President, there will probably be none for Vice President either, and the Senate will then have to choose the Vice President from among the top two finishers. As with the choice of a President by the House, all states would have the same number of votes. Indeed, voting in the Senate—one vote per senator—may be even less democratic than in the House, since two thirds of the Senate will have been elected in prior years.

The Senate has chosen a Vice President only once, in 1837. Virginia's twenty-three Democratic electors refused to vote for the Democratic vice-presidential candidate, Richard M. Johnson, because he had a longstanding romantic involvement with a black woman. Justice Catron wrote to Andrew Jackson that "the idea of our voting for him is loathed beyond anything that has occurred with us since we have begun to act in concert with our sister states on national policy." The Senate elected Johnson, but a President was already in office. If the presidency had gone to the House, and if there had been a deadlock, our Acting President in 1861 would have been Joseph Lane; in 1913, Thomas R. Marshall; in 1925, Charles W. Bryan; in 1949, Alben W. Barkley; in 1961, Lyndon Baines Johnson; in 1969, Edmund Muskie.

In 1981, the Acting President could be Walter Mondale—or George Bush, or John Anderson's running mate. Indeed, Bush's chances in the Senate could be better than Reagan's in the House, since the Republicans need to win only nine seats from vulnerable Democratic incumbents to control the Senate. It is not likely, but there is more than an outside chance of a Republican Senate.

President for a day—or for four years

One combined effect of the Twelfth and Twentieth Amendments is that the House could go on voting, with interruptions for other business and indeed with an infusion of new members in midterm, for four full years. Imagine an acting presidency subject to termination at any time until the House deadlock is finally broken. That would transform the American government into a quasi-parliamentary system. As matters stand in 1980, any disgruntled House member from any of twenty-eight states could threaten to switch his or her vote—and thereby turn the Acting President back into a lowly Vice President again. The desire to be more than the spare tire of the executive branch could leave an Acting President open to constitutional blackmail. This seems at best a freakish approximation of parliamentary rule, and it would give unprecedented weight to the demands of the narrowest possible interests. But it is conceivable that even such a system could muddle through—at least until the 1982 congressional elections, which would become hybrid parliamentary votes. Thus, a vote for Republican representatives would be a vote for Reagan; for Democratic representatives, a vote for Carter. Anderson might sponsor his own slate of independent candidates. Thus the deadlock could be broken—or prolonged.

The succession

Because senators vote for Vice President individually rather than by state, the odds of deadlock in the Senate are slim, but it could happen there, too. The majority party in the Senate might want someone other than the top two vice-presidential finishers to act as President. To achieve that goal, thirty-four senators need only absent themselves so as to block a quorum. And even fewer abstentions, of course, could deprive both leading candidates of the 51 votes needed for a majority. At that point, the succession statutes would be triggered. House Speaker Tip O'Neill would be first in line if he was willing to resign from Congress, followed by the seventy-five-year-old Warren Magnuson, president pro tempore of the Senate, and then Edmund Muskie, Jimmy Carter's secretary of state.

In 1800, the Federalists considered a plan to refuse to vote at all for Jefferson or Burr and to appoint a Federalist president pro tempore of the Senate who would act as President indefinitely. But if Congress tries to change the succession statutes in this election season, the members will face the same two hurdles that frustrated the Federalists in 1800. In his role as president of the Senate until January 20, 1981, Vice President Mondale could rule any such scheme out of order, as Vice President Thomas Jefferson vowed he would do in 1800. And Carter, still President until noon on January 20, could threaten to veto any such statute, as John Adams would not pledge to do in 1800. The two-thirds vote needed to override him might prove impossible to muster.

The common sense of uncommon elections

In 1823, two decades after he was chosen President by the House, Thomas Jefferson wrote: "I have ever considered the constitutional mode of election ultimately by the legislature voting by states as the most dangerous blot on our Constitution, and one which by some unlucky chance will some day hit." That "unlucky chance" hit just two years later—when John Adams was accused of buying the presidency from Henry Clay. And surely there are good reasons to question the wisdom of a procedure through which a candidate can win the presidency by carrying the votes of fifty-nine representatives from the twenty-six smallest states while losing the votes of the other 376 members—fully 86 percent of the House. The five smallest states, with a combined population of just over 2 million and with only five representatives, would enjoy as much influence in a House election as the five largest states, with a combined population of over 72 million and 153 representatives.

But the conjuring of such horrors reflects theoretical possibilities, not political realities. Why would the small states band together? They have widely divergent interests. They span the continent from Alaska to South Carolina, from Maine to Arizona. Freakish coalitions cannot be ruled out; but constitutions designed to be fail-safe against the remotest contingencies are unlikely to prove fit for the workaday world.

In truth, what we have learned to live with in the presidential selection system is a complicated process, rich in possibilities for high statesmanship and low politics alike. Do the voters necessarily lose in this byzantine game? Must democracy get lost in it? Should we heed the demands to amend the Constitution to prevent House elections—replacing them with popular-vote runoffs, or even replacing the Electoral College itself with some other system? Whatever the answer, the question is moot this election year. Realistically, no amendment could be ratified by November 4. But even if it could, experience counsels caution in changing the Constitution's fundamental design, moved by what are at worst hypothetical fears and contingent anxieties. The Electoral College persists despite repeated challenge both because we know how it works and because we know how it distributes power. That it is not congruent with pure democracy or majority rule is true enough, but certainly is not decisive. Majority rule was not an absolute principle of self-evident wisdom to the Founders, nor should it be for us. Apart from the Constitution's many substantive constraints upon majorities—constraints such as those in the Bill of Rights—it is worth recalling, with Walter Lippmann, that the Founders thought of the people "as having many dimensions in time, space, and quality ... The Founders sought to approximate a true representation of the people by providing many different ways of counting heads."

Choosing Presidents by popular vote, either initially or to break ties when no electoral majority has been won, would shift power away from urban states, with their concentrations of minority groups, at the same time that such states remain permanently under-represented, on a per capita basis, in the Senate. As John Kennedy argued in 1956 on the Senate floor: "It is not only the unit vote for the presidency we are talking about, but a whole solar system of governmental power. If it is proposed to change the balance of power of one of the elements of the solar system, it is necessary to consider all the others." Beyond that, any system that made presidential choices depend on national popular vote totals would put a premium on the size of a candidate's plurality in every city and town, encouraging both voter fraud and morning-after challenge. And holding a national runoff election between the top two finishers if no candidate achieves a minimum of 40 percent of the popular vote, as Senator Birch Bayh has proposed, would invite a proliferation of fourth and fifth parties, each trying to bootstrap its base of support into a berth in the two-way presidential sweepstakes. American politics could thus become a grotesque parody of European pluralism.

Nor is it clear that the House election procedure, rickety and peculiar as it seems, must ill serve the public will. In three-or four-way elections in which no candidate wins an electoral or popular majority, how can we know who the voters' second choices would have been? The House may be the best place for such second preferences to be determined. Moreover, in a year when the voters are divided and no one candidate can claim to "represent" most of the country, a House election can serve as a forum for compromise—as a force for unity when an unqualified triumph for a minority faction would only increase the divisions that made a majority victory impossible in the first place. "Deals" that make national unity possible can be matters of policy and not personal advantage.

The experience of openly recognized division followed by deliberation and eventual reconciliation can itself be a source of release, catharsis, and reunion. Jefferson wrote after the crisis of 1800: "I am persuaded that weeks of ill-judged conduct here has strengthened us more than years of prudent and conciliatory administration could have done." Close presidential elections do not cause the tensions that divide us; they reflect these tensions and provide both a ritual and a forum through which the underlying conflicts can be mediated—and even resolved.

In any case, it is pointless and counterproductive to declaim in advance that a House election would represent a constitutional crisis that voters should seek to avoid by casting their lot with the front-runner. Prophecies of crisis may easily become self-fulfilling, and no strategy to avoid such crisis is likely to be very secure anyway. Besides, campaigns have played on such fears too often. In 1924, editor George Harvey wrote an influential article in the North American Review arguing that Robert La Follette's third-party candidacy would throw the election into the House, which would deadlock, resulting in the election by the Democratic Senate of Charles W. Bryan, the younger brother of the "populistic and pacifistic" William Jennings Bryan, "who unquestionably would continue to act as his guide and counsellor." The Republicans developed a bizarre campaign cry: "A vote for La Follette is a vote for Bryan and a vote for Davis is a vote for Bryan. A vote for Coolidge is a vote for Coolidge." Coolidge won. In 1968, Richard Nixon deployed the same argument in the southern states—a vote for Wallace was a wasted vote and, worse, a vote for Wallace might elect Humphrey by denying Nixon an electoral majority and leaving the election to the House.

The point is hardly to extol the virtues of presidential election by the House; any such election has its costs and risks. The point is simply to regain a measure of perspective. The strategy of winning votes by generating fear about the process itself not only makes the tragic view of a House election a self-executing perception; it could also scare a majority of voters out of casting ballots for the third-party or independent candidate who might represent their first choice for the presidency, and the best chance to revitalize the major parties and restore legitimacy and unified leadership to the national government. Afraid that their votes for the independent candidate they favor would actually be votes for the party candidate they like least, a majority may vote for the other party's nominee—as a second best compromise. And afraid that votes for anyone but the front-runner would actually be votes for a House election, many may even vote for their third choice rather than their first or second—to avoid getting the wild card of an Acting President.

In his 1933 presidential address to the Organization of American Historians, Professor John D. Hicks argued that "in a remarkable number of instances, third parties marked out in advance the course that later on the nation was to follow. The supporter of third-party tickets need not worry, therefore, when he is told, as he surely will be told, that he is 'throwing his vote away.' [A] backward glance through American history would seem to indicate that his kind of vote is after all probably the most powerful vote that has ever been cast." How ironic it would be if the exaggerated fear of an election by the House robbed us, in the end, of an election by the people.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.