Reforming The Electoral College

Norman Williams - Willamette University

As the 2000 Presidential election reminded everyone, the President is not elected directly by the People but rather by the Electoral College.  Because each state has as many presidential electors as it has U.S. Representatives and Senators, smaller states have more electoral votes than their population warrants.  At the same time, all but two states have adopted a “winner-take-all” system in which the winning presidential candidate receives all of the state’s electors regardless of the actual vote margin in the state.  As a result, the Electoral College vote does not track precisely the national popular vote.  A candidate who wins many states by a few percentage points can achieve a dominating Electoral College vote, as Ronald Reagan did in 1980.  More rarely, the national popular vote winner can actually lose the election, as the 2000 Presidential election graphically demonstrated.

Since 2000, one of the most serious efforts to reform the Electoral College has quietly unfolded, not in Washington, D.C., but in state capitals across the nation.  Galvanized by a shared sense of outrage regarding the 2000 election, several reform-minded citizens, including Akhil Amar and his brother Vikram Amar, imagined a novel way to transform the manner in which the nation elects its President that avoids the time consuming and daunting process required for a federal constitutional amendment.  Their idea is to have a large group of states agree to appoint their presidential electors in accordance with the national popular vote rather than their respective statewide popular vote.  Memorialized in a proposed interstate compact known as the “National Popular Vote Compact” (NPVC), their proposal goes into effect once states comprising a majority of the Electoral College join it.  From that point on, the national popular vote will conclusively decide the winner of the election regardless of whether all the states agree or a constitutional amendment abolishing the Electoral College is adopted.  In essence, these reformers seek to use the coordinated action of a number of states to turn the Electoral College into the vehicle of its own reform.  To date, eight states, including the electoral behemoth of California and the District of Columbia, have formally adopted the compact, and several other states have moved toward joining it.

The notion of a direct, popular election for the Presidency has an obvious appeal to Americans, who have an abiding faith in majoritarian democracy.  On closer inspection, however, the current system does not derogate from majoritarian democracy to a substantial extent, while the NPVC would produce Presidential elections deeply violative of majoritarian democracy.  In short, the NPVC is neither a necessary nor desirable reform.

First, the Electoral College is not the threat to American democracy that its critics urge.  While the Electoral College admittedly gives some states more electoral clout than their population would otherwise require, that “malapportionment” is both modest in degree and, more importantly, merely the price paid for having a presidential election system that combines elements of majoritarianism and federalism, as other large, federal democracies do.  In fact, when the actual operation of the Electoral College is examined, it turns out that the Electoral College blends those two values in a manner heavily weighted toward majoritarianism.  In all but one election (out of 56), the Electoral College has elected the candidate who won a majority of the national popular vote.  To be sure, the Electoral College does reward candidates whose political support is spread in a more geographically broad fashion throughout the nation, but, in a federal union such as the United States, that federalism-based bias against “favorite son,” sectional candidates is a desirable feature – and one that would be lost in moving to a purely majoritarian election system as the NPVC seeks to do.

Second and relatedly, the NPVC poses a far greater threat to majoritarian democracy than does the Electoral College.  Under the express terms of the NPVC, the national popular vote winner need not win a majority of the national popular vote; rather, the candidate who receives the most votes in the nation is deemed the national popular vote winner.  Thus, a candidate who wins only 40% or 30% (or even less) of the popular vote could become President so long as every other candidate received fewer votes.  That is hardly a move toward majoritarian democracy.

To be sure, plurality Presidencies are possible under the current system, but the prevailing winner-take-all process used by most states places a floor on the level of support the winning candidate receives by favoring the current, two major parties and discouraging third-party bids for the White House.  As a result, the Presidential contest is typically a two candidate affair, with the winner usually receiving a majority of the popular vote.  In fact, since the Civil War, most Presidents have received a majority of the popular vote, and even those Presidents who have been elected with a plurality of the vote have received, on average, over 47% of the vote.  In contrast, under the NPVC, plurality Presidencies with ever lower levels of support would become more common over time.  By transforming the current, state-by-state voting process, the NPVC would erode the current two-candidate system, producing more minor party candidates, which in turn would further fragment the national popular vote and produce more plurality Presidents with ever-declining levels of support.  Indeed, that has been the experience of other countries with voting systems like that proposed by the NPVC, and, as those countries have experienced, plurality presidencies typically lack the legitimacy and political support necessary to effectively lead the nation.

Third, even if moving to a direct popular election for President were desirable, a sub-constitutional, interstate compact is the wrong mechanism to use to achieve that result.  Unlike a constitutional amendment abolishing the Electoral College, the NPVC does not ensure a fair and workable presidential election process.  To the contrary, as an interstate compact which governs only those states that join it, the NPVC promises a number of political and legal fights among the states that will undermine the legitimacy of presidential elections and provoke enervating constitutional litigation of the sort witnessed in 2000.  The problems with the NPVC in this respect fall into two broad categories:  problems of obstruction and problems of implementation.

As to obstruction, the NPVC cannot prevent non-signatory states from undermining the calculation of the national popular vote, nor can it ensure that even signatory states will not withdraw from the compact on the eve of, or worse, shortly after the general election in November.  The easiest (and therefore most likely) form in which such obstruction could take place would be in the calculation of popular votes in each state.  Non-signatory states could simply refuse to release their popular vote totals prior to the date on which the Electoral College votes in mid-December.  If a sufficient number of non-signatory states engaged in this tactic, it would be impossible for the signatory states to determine which candidate was in fact the national popular vote winner.  In such a circumstance, it is purely a guess as to how signatory states would respond – indeed, the NPVC fails to provide any guidance to signatory states in such a scenario, making it highly likely that different states would act in different (and therefore contradictory) manners.

Even worse is the problem of states withdrawing from the compact either right before, or worse, right after the general election in November.  In fairness, the compact itself provides that states may not withdraw after July 20th in a presidential election year, but that proviso is unenforceable both legally and practically.  If the NPVC is not ratified by Congress, it possesses no greater force than any other state law adopted by a state legislature.  As such, like any other state law, it can be repealed by a state legislature at any time in accordance with that state’s constitutionally prescribed legislative procedure.  Moreover, Article II of the U.S. Constitution delegates to the state legislatures the power to select their presidential electors in such manner as they see fit.  That power cannot be alienated by a prior state legislature.  Thus, a state legislature’s adoption of the NPVC in 2011 cannot prevent that same state’s legislature in 2016 (or any future presidential election year) from choosing to alter its presidential election process at any time up to and even after the November general election.

Nor would Congress’s ratification of the NPVC preclude the ability of signatory states to withdraw after July 20th.  Given Congress’s repeated failure to adopt a constitutional amendment abolishing the Electoral College, it is highly unlikely that Congress will approve the NPVC, but, even if it did, that would not matter.  Article II delegates the power to determine the manner in which electors are selected in each state to the state legislature alone.  In the views of some, that proviso means that Congress may not interfere with a state legislature’s choices with regard to how that state’s presidential electors are selected.  Equally importantly, even if Congress’s ratification did take place and have binding effect, it is far from clear that any court, federal or state, would be able to force a withdrawing state to appoint its presidential electors in accordance with the nationwide popular vote against its will.  In fact, any such injunction might only succeed in producing two competing slates of electors from the same state – one based on the statewide poll and one based on the nationwide poll.  In that case, the determination of which slate of electors was the “legitimate” one would fall to Congress under the Electoral Count Act, and Congress neither is nor can be bound by the NPVC in determining which slate of electors is the appropriate one.

As to implementation, the NPVC cannot guarantee a Presidential election process that accords with rudimentary notions of political equality and fairness.  Different states employ different rules regarding who may vote, how they vote, and how their ballots are counted.  Those differences are substantial, and they preclude simply summing up the popular vote on a national basis.  Indeed, the same constitutional flaw that the U.S. Supreme Court identified in Florida in 2000 – the use of divergent vote tabulation standards in different counties in Florida – would be replicated fifty-fold.  If it violates the Equal Protection Clause for Broward County in Florida to employ a different vote tabulation standard than that used in Miami-Dade County in determining who won the statewide poll, it would likewise violate Equal Protection for Oregon, in determining who won the national popular vote, to use a different vote tabulation standard than that used in California or Florida.  In essence, the NPVC’s failure to provide for a nationwide, uniform election process guarantees that the problems encountered in Florida in 2000 are replicated but, this time, on a national scale.

Moreover, even if it were constitutional to amalgamate the results of state-conducted elections on a national basis, the NPVC fails to provide for a recount if the national popular vote is close, as it was in 2000 (and in several prior elections).  On the one hand, the specter of a nationwide recount with its attendant administrative chaos, expense, and all-too-assured accusations of partisan manipulation of the process in different states should give everyone pause.  Yet, in the absence of such a recount, there would be no way to know which candidate truly won the popular vote.  The only thing worse than conducting a nationwide recount would be failing to conduct one in a close election.

Thus, far from preventing another 2000, the NPVC almost assuredly would produce a series of political and legal crises, along with the accompanying litigation that inevitably form a part of such imbroglios, that would make the 2000 election look like child’s play.  That is not reform; rather, it is an invitation to repeated electoral disasters.  Indeed, for those who prefer that the winner of the Presidency not be determined by the U.S. Supreme Court via post-election constitutional litigation, the NPVC is a dangerous step backwards.


Norman R. Williams is the Ken & Claudia Peterson Professor of Law and Director of the Center for Law and Government at Willamette University.

Copyright © 2011 Georgetown Law Journal


  • I’m confused by some of these objections.

    “In all but one election (out of 56), the Electoral College has elected the candidate who won a majority of the national popular vote.”

    It has happened four times: 1824, 1876, 1888 and 2000. That’s 1 in 14. It very nearly happened in 2004, 1976, 1968, 1960 and 1948.

    Perhaps more important, under the present system, candidates consolidate their efforts in a small group of battleground states. In the past couple of elections, candidates have ignored 12 of the 13 smallest states (New Hampshire being the lone exception) for this reason. That strikes me as a serious problem.

    “Thus, a candidate who wins only 40% or 30% (or even less) of the popular vote could become President so long as every other candidate received fewer votes. ”

    As you note, that is the case now. In fact, no candidate received an absolute majority in the 1992, 1996 and 2000 elections. This is not a problem inherent to the National Popular Vote plan.

    Under the present system, third party candidates could, theoretically, leverage regional affiliation to win a small number of states, thus throwing the election one war or the other. I see no reason why the National Popular Vote plan lowers the popular vote threshold for winning election.

    “Third, even if moving to a direct popular election for President were desirable, a sub-constitutional, interstate compact is the wrong mechanism to use to achieve that result.”

    How is it reasonable to describe interstate compact as sub-constitutional when they are permitted by the Constitution? Interstate compacts are in use in every state.

    State legislators have the power to award electoral votes as they see fit, and so it makes sense to work at a state level to enact the plan.

    “Non-signatory states could simply refuse to release their popular vote totals prior to the date on which the Electoral College votes in mid-December.”

    Even if it were politically feasible to disenfranchise voters in one’s own state, federal law forbids this. Federal law also requires the report of a national vote count. This scenario is not possible.

    “As such, like any other state law, it can be repealed by a state legislature at any time in accordance with that state’s constitutionally prescribed legislative procedure.”

    This would clearly violate the impairments clause of the Constitution if done before the election. If done after the election, it would violate federal law. If it weren’t so, what would prevent states from making such a change under the current system?

    This analysis misses the mark on other scores as well, and everything here has long since been addressed at length. I would encourage you to visit and learn a bit more about what the plan does and does not do.

Post a Comment (all fields are required)

You must be logged in to post a comment.