The idea of reforming the Electoral College recurs each time a presidential election nears. Polls show that an overwhelming majority of respondents support abolishing the Electoral College in favor of direct popular election of the President. Yet, it is doubtful whether these polls really imply that such a move would be best for the country. Despite the seeming simplicity of direct popular presidential election, its introduction in the United States-a country in which the clear separation of powers between the states and the federal government has existed for more than two centuries-would have hidden drawbacks that the media and pollsters usually fail to communicate. Further, the existing Electoral College-based system of electing a President is complicated, and the simplistic media coverage of American social and political phenomena fails to educate voters about the nuances of that system. Thus, pollsters are asking people whether they favor replacing the Electoral College, a system that many respondents don't sufficiently understand, with direct popular election, a system that many respondents also don't necessarily understand.
Since the 2000 election, a dozen electoral reform proposals have been discussed both in the scientific community and in the media. Five proposals have received particular attention. I discussed two of them-one introducing a Maine-like district scheme and the other introducing a proportional scheme of awarding state electoral votes both in some states and nationwide-in my recent editorial in the Baltimore Sun, "District Vote Proposal Falls Short." This Commentary explores the other three proposals, which are concerned with introducing the nationwide popular vote into presidential elections.
Since space limitations prevent a detailed analysis of the proposals under consideration here, this Commentary offers only a kind of a snapshot of the basic ideas underlying each proposal, along with the inevitable negative consequences of adopting two of the proposals. I don't discuss how the technical problems associated with counting every vote under direct popular elections-for instance, voter registration fraud, long lines, and voting machine glitches-might compare to those under the existing election system. Whatever these problems may be, the quality of services in administering elections shouldn't be a factor in choosing a particular scheme for electing a President.
New election rules that I first proposed in Extreme Outcomes of US Presidential Elections and (in a particular form) in How America Chooses Its Presidents would encourage the country to elect a President with a mandate from both the nation and the individual states as equal members of the Union. Under these rules, the nationwide popular vote would be a decisive factor in choosing the President, while the existing Electoral College-based system would be preserved as a backup. Also under these rules, all the states would gain the attention of presidential candidates while retaining the benefits of the Electoral College.
Though I have my own preferences regarding the three proposals I discuss in this Commentary, readers should decide for themselves which proposal merits which of the three labels that the title of this Commentary borrows from Sergio Leone's famous movie.
I. Direct Popular Election of a President
Implementing a direct popular election system-which has been proposed many times but always without sufficient support in Congress to initiate a constitutional amendment-would require revolutionary changes in the American political system.
First, American voters would acquire a constitutionally guaranteed right to vote for President. Currently, the participation of Americans in presidential elections-even voting for presidential electors-isn't constitutionally guaranteed. Appointing electors by state popular vote is no more than a particular manner of appointing state electors. Constitutionally, state legislatures can replace this particular manner with any other manner of appointing state electors, as the Supreme Court reaffirmed in Bush v. Gore in 2000.
Second, introducing this election system would likely invite multi-candidate presidential elections in the United States. Three factors-a potentially favorable electorate, an appeal to targeted voters, and financial resources-are crucial for non-major party candidates to emerge, and all of these factors look favorably upon the appearance of such candidates in direct popular presidential elections.
A sizable majority of eligible voters are, in fact, up for grabs in presidential elections, particularly, for non-major party candidates. Almost forty-five percent of all eligible voters usually do not vote in presidential elections, at least ten percent of the electorate are swing voters, and more and more registered Republicans and Democrats call themselves independent. Also, under direct popular election, the constitutionally guaranteed right to vote for President would shift the burden of voter registration to the federal government and would eliminate many currently existing obstacles to voting in presidential elections. The winner-take-all rule would not waste votes cast for the losers in state presidential contests-as it does under the existing election rules. All this would contribute to encouraging people who are currently non-voters to vote.
In 1992, Ross Perot captured almost 19% of the 55% voter turnout, i.e. more than 10% of the whole electorate (more precisely, the voting-age population), even without any special appeal to non-voters and despite the manner in which his campaign was conducted. If only half of current non-voters voted in presidential elections, they would make the group of voters potentially favorable to non-major party candidates comparable in size to the group of voters favorable to either of the major party candidates. Since the major parties do not currently seem to represent non-voters in presidential elections, there is plenty of room for non-major party candidates to appeal to both non-voters and swing voters.
Financial problems for appealing non-major party candidates also look solvable. For example, the Internet has proven effective in fundraising for presidential hopefuls who appeal to particular factions of voters and in organizing concerned voters. Also, self-financed political figures interested in running as non-major party candidates are widely perceived by Americans as independent of any particular sponsors.
If multi-candidate elections with competitive non-major party candidates become a reality, a particular scheme for conducting multi-candidate direct presidential elections in the country should be chosen. The two most-employed schemes-which are based on the simple rule "choose one candidate only"-are (a) one-round elections with a popular vote plurality winner; and (b) two-round elections with a popular majority winner and, if no candidate receives a majority of votes in the first round, a runoff between the top two recipients of votes.
In such a politically divided country as the United States, under the first scheme multi-candidate elections with several candidates appealing to large factions of voters may eventually produce an election winner with less than 30% of all votes. Would the United States accept a President who more than 70% of the electorate had not chosen? In many countries employing the second scheme, it is common for no candidate to receive a majority of votes in the first round and for the turnout for runoff elections to be much less than 50% of the eligible voters. If, say, less than 40% of all eligible voters cast ballots in a U.S. runoff, would Americans accept a President finally elected by, say, only 20% of all eligible voters?
Though there exist schemes of choosing an election winner that overcome many deficiencies of the "choose one candidate only" rule in reflecting the wishes of voters, their introduction into possible multi-candidate direct popular presidential elections in the United States (at least currently) could be problematic. This is especially true taking into account that Americans have never had either a right or a chance to choose a President directly, even under this simple rule.
Third, many Americans may feel that the introduction of direct popular presidential election would weaken the federal system of government, because an elected President may not have a mandate from the states, even in the form currently provided by the Electoral College. Moreover, movements supporting non-major party presidential candidates at times may result in the emergence of new strong political parties, and many Americans believe that a multi-party political system would destabilize the country.
In the United States today, the majority of the population resides in the largest eleven states. As a result, candidates in direct popular presidential elections would likely campaign mainly in large, urban areas. Sparsely populated rural states would likely be ignored by major party candidates. Admittedly, many of these states are ignored by candidates even under the existing election system; but sparsely populated states may eventually have a say in the current Electoral College, especially in close elections. Currently, if no candidate receives a majority of electoral votes, the election will be thrown into the House of Representatives, where each of the fifty states casts an equal, singular vote-regardless of the size of its population-due to the one state, one vote constitutional principle. Nothing even close to this would exist for small states in any direct popular presidential elections.
As a result, each of the seventeen states with five or fewer electoral votes is unlikely to surrender what it is currently entitled to under the existing election system by supporting a constitutional amendment introducing direct popular elections. Nor would this amendment likely garner support from current medium-sized battleground states-which draw a great deal of attention in election campaigns under the Electoral College. In any direct popular election, these medium-sized states would have to compete with large, densely popular states for even less attention of the candidates than they currently enjoy.
II. The National Popular Vote Plan
In the aftermath of the 2000 election, Robert Bennett, a prominent constitutional lawyer, proposed an idea designed to effectuate the constitutionally guaranteed right of state legislatures to choose presidential electors in any manner they want, making it theoretically possible to circumvent the resistance of small states to direct popular presidential elections. This idea, described in Bennett's book Taming the Electoral College, was later reinvented by John Koza, a prominent computer scientist, and developed into the National Popular Vote plan described in the book Every Vote Equal. In its current form, the National Popular Vote plan involves assembling a compact of states that together control at least 270 electoral votes. The signatory states' legislatures would agree to award state electoral votes to the winner of the nationwide popular vote plurality, no matter how each of their own states voted.
If a sufficient number of states adopt the interstate compact, the plan is likely to face a constitutional challenge. The Supreme Court may conclude that while each state legislature is free to choose a manner of appointing state electors, a group of state legislatures cannot de facto introduce a new system of electing a President without an amendment to the Constitution-particularly over the objection of at least one-fourth of the members of the Union.
The plan strikes at the heart of the 1787 Great Compromise, which the Founding Fathers reached to keep the states together as a nation. The words of Delaware's Gunning Bedford, Jr., to delegates from large states at the 1787 Constitutional Convention-"I do not, gentlemen, trust you"-in discussing principles of state representation in Congress serve as a reminder about tension between the small and large states in reaching the compromise. At the Convention, large states pledged to honor (a) the Electoral College, an "intermediate, independent Congress" with numbers of electoral votes for small states disproportionate to the size of their population, and (b) the one state, one vote principle both in electing a President in the House of Representatives and in amending the Constitution.
Today, however, the fifty states and the District of Columbia (D.C.)-rather than a college of presidential electors-choose the President. This happens due to the winner-take-all rule coupled with the widely implied (though not constitutionally required) obligation of presidential electors to follow the will of the appointing power (based on the state popular vote in forty-eight states and in D.C. and on the popular vote within each congressional district in Maine and Nebraska). This appears to violate the one state, one vote principle of electing a President by states, since a state's quota of electoral votes is based on the size of its population. Moreover, constitutionally, the states are to elect a President through the House of Representatives only when the Electoral College fails to do so, i.e., when no candidate receives a majority of votes from all the appointed electors. In addition, adopting the winner-take-all rule of awarding state electoral votes, currently employed by forty-eight states and by D.C., has contributed to dividing the country into two parts during election campaigns. Candidates from the two major parties feel "safe" in a majority of states while the electoral battles take place in a remaining minority of states. As happened in 1992, 1996, and 2000, a third-party candidate may affect the election outcome by influencing electoral votes in key battleground states, and in close elections the existing election rules may produce an outcome contrary to the nationwide popular will.
But all this doesn't mean that the circumvention of a substantial number of states-signaling that the value of the Great Compromise with respect to electing a President is no longer honored by the signatory states-can be justified.
Under the compact, one can easily imagine a multi-candidate race in which a candidate would win, say, a thirty-four percent plurality of the popular vote nationwide while losing in every state and D.C. If all of the states and D.C. were signatories to the compact, all the electoral votes in such a hypothetical race would be awarded contrary to the will of voters choosing electors (still not voting directly for President under this plan). Would the United States accept a President who wasn't the choice of sixty-six percent of those voting, nor even the choice of a single state? Also, this proposal could create the appearance of "faithless" electors in the states forming the compact. These electors may decide not to follow the nationwide popular vote if it goes against the will of the voters of their states. Finally, implementing this plan would still not provide the electorate with any constitutionally guaranteed right to vote in presidential elections, even for presidential electors.
III. Election Rules Making the Nationwide Popular Vote a Decisive Factor in Electing a President but Retaining the Electoral College
The new election rules I have proposed would name as President the recipient of a majority of the nationwide popular vote and of the popular vote majorities in at least twenty-six states (or in twenty-five states and D.C.) as long as a majority of all eligible voters cast ballots in the election. This would be true even if another candidate won the Electoral College. Only if no candidate achieved the required majorities would the winner of at least 270 electoral votes-automatically awarded by the states and D.C. in the manners chosen by their legislatures-become the next President. If no candidate won at least 270 electoral votes either, then the proposed election rules would require the House of Representatives to choose a President, as the Twelfth Amendment directs.
When more than fifty percent of all eligible voters don't vote, choosing a President by the nationwide popular vote seems illogical. In such cases, either a majority of voters don't care, or they believe that the candidates do not deserve their votes. It this were the case, as described above, the Electoral College and, if necessary, the House of Representatives, should step in as protective mechanisms-backups guaranteeing that a President is elected (or selected) as a result of the election (though, currently, the existence of such a guarantee can be questioned, as How America Chooses Its Presidents argues). The Founding Fathers might have seen these two election mechanisms as protecting against a failure to elect a President, even if the popular will would not be expressed or a popular consensus could not be reached.
Of course, any state may decide that a plurality of the statewide popular vote is sufficient to carry the state. Also, if the number of voters casting ballots in a state is too small to award state electoral votes based on the statewide popular vote, the state legislature should retain the right to appoint electors to the Electoral College.
These new election rules, first proposed in my books Extreme Outcomes of US Presidential Elections and (in a particular form) How America Chooses Its Presidents, are easy to understand by using three conceptions of the U.S. Presidency. First, a candidate who wins a nationwide popular vote majority is a "President of the people." Second, a candidate who achieves a majority in each of at least 26 states (or in 25 states and D.C.) is a "President of the states." Third, a candidate who achieves a majority of votes from all the appointed electors is a "President of an electoral majority in the Electoral College"-a compromise candidate in the sense of the Great Compromise.
A candidate who garners two kinds of the voter majorities-both the nationwide and the statewide majorities in 26 states or in 25 states and D.C.-is both a "President of the people" and a "President of the states." This candidate would become the next President, even if another candidate became the "President of an electoral majority in the Electoral College." If no candidate is both a "President of the people" and a "President of the states," then the existing rules would apply, such that the Electoral College or, if necessary, the House of Representatives would have to choose the next President.
These rules would not destroy the existing system or any of its parts. Rather, they would build on the current system by potentially offering up a candidate whom society would perceive as better than the "compromise" candidate. These rules substantially differ from the Federal System Plan of 1970, and they address federalist concerns in the strongest form, since the one state, one vote principle matters in an attempt to directly elect a President by the nation as a whole.
Certainly, only a national dialogue may detect whether a compromise candidate-a "President of an electoral majority in the Electoral College"-is perceived by Americans as a better choice for the country than a candidate who is both a President of the people and a President of the states. If this were the case, there would be no need to implement the proposed new rules, despite all the well-known deficiencies of the existing election system.
The proposed rules would encourage major party candidates to actively campaign in all states, regardless of size. These candidates are likely to compete in all large states to win a nationwide popular majority and in small states to seek to win in at least twenty-six states. Both candidates are likely to compete in medium-sized states as well, especially in the "battleground" ones, since the Electoral College might eventually decide the election outcome.
These new rules would also encourage voter turnout by affording Americans the right to vote for President while leaving state legislatures with the right to appoint electors if their state's voter turnout is too small.
The circumstances surrounding the creation of the Electoral College suggest that the Founding Fathers might have believed that new generations of Americans would propose a better election system or at least a better compromise in electing a President as the country developed. Such a compromise may be found in the proposed election rules under which nobody seems to lose while all voters and states gain.