Law Review
Northwestern Law
Northwestern University Law Review : Colloquy : 2006 : Bennett

Electoral College Reform Ain't Easy

By Robert W. Bennett

At the end of September California's Governor Arnold Schwartzenegger vetoed legislation that would have assigned the state's huge chunk of 55 electoral votes to the winner of the nationwide popular vote—once other states with 215 additional electoral votes were similarly committed.  The total of 270 electoral votes would represent the required electoral college majority and hence seemingly guarantee that the nationwide vote winner would become president.  This would make impossible what happened in 2000—and in two or three other presidential elections over the years—where the electoral college winner lost the nationwide popular vote.

It's not hard to understand the Governor's veto.  Under the legislation, California's electoral votes could go to a candidate who had lost the statewide vote.  A shift of those 55 electoral votes could thus award the presidency to a candidate whom California's voters had rejected, perhaps by a large margin.  Still, there are substantial arguments in favor of the legislation, even from the vantage point of California's self-interest.  California—along with lots of other states—has been quite predictable in recent presidential elections.  As a result, candidates of both major parties campaign very little in the state.  They lavish attention—including campaign promises—on "swing" states instead.  If the nationwide vote determined the outcome, California (and other politically lopsided states) could expect to receive a good deal more attention from candidates of both political parties.

In discussing this legislative route to a nationwide popular vote for president, I long ago suggested a more modest alternative that would have a much greater chance of success.[1]  Instead of requiring that 270 electoral votes be committed to the nationwide vote winner, legislation could commit states once a more modest number of electoral votes was on board.  For purposes of discussion, let's put the figure at 150 (though I think that many fewer could do the trick).  To get states to sign on, the electoral votes of participating states would have to be spread pretty evenly across the political divide.  If 150 votes were committed, candidates could no longer sensibly concentrate on the swing states.  California and other politically lopsided states would receive a great deal of attention from candidates for both major parties.  The backers of the California legislation have thus far disdained this alternative, but they might well reconsider, as the Governor's veto augurs poorly for success around the country at the 270 vote level.

At the same time I think there is a more pressing problem lurking in the shadows of the electoral college process.  Over the years, there has been a smattering of "faithless" electoral votes.  These are votes that an elector sees fit to cast contrary to his pre-election commitment to the candidate of the political party that advanced him for the office of elector.  There was a faithless vote in the 2000 election and another in 2004.  But faithless votes have never changed the outcome of a presidential election, and the possibility has not recently generated much concern.

In Taming the Electoral College, I argued that close electoral college elections may well start coming with greater frequency, and that the danger of elector faithlessness changing an outcome in a close election is quite real.[2]  Elector faithlessness that seemingly changes an election outcome could be disastrous—much more serious than an occasional election in which the electoral college winner loses the popular vote.  Electoral votes are counted in a joint session of the Congress in early January.  Where the outcome seemed to turn on elector faithlessness charges and cross-charges of illegitimacy could be expected at that meeting.  And each would have a good degree of plausibility.  Particularly if the two houses were controlled by different political parties, there could well be a prolonged stalemate.  And however and wherever the matter was resolved, the residual bitterness would make the discontent after the 2000 election look like child's play.

I have also argued that elector faithlessness can and should be forbidden by a uniform state law.[3]  The matter is receiving some attention from the Commissioners on Uniform State Laws,[4] and if a proposal is forthcoming from them, it might achieve the political visibility required to deal with the problem.

One possible stumbling block is that it is sometimes urged that elector faithlessness is constitutionally protected.  The constitutional argument is grounded in the original conception of electors as rising above politics to exercise real discretion as they debated and decided on who in the country could best fulfill the calling to be President of the United States.[5]  That vision of the role of electors bears no resemblance to the way they function today.  If we took the argument for constitutional protection seriously, we would also have to take seriously other aspects of the original conception of the presidential selection process.  For instance, neither political party designation nor the names of presidential and vice presidential "candidates" could appear on ballots, because that is a way of signaling pre-commitment of electors, rather than a process of debate and discussion that was the reason for creating the electoral college.  Protecting elector faithlessness, in other words, would work havoc with the modern way in which we choose our chief executive.  Thus, it is hardly surprising that the Supreme Court tilted against constitutional protection for elector faithlessness in the only decision that touched on the question.[6]

The nationwide vote effort, however, complicates the matter.  The legislation could simply add a strong elector faithlessness prohibition to the nationwide vote provisions, but its backers—at the 270 electoral vote level—are not much concerned about elector faithlessness.  This is understandable, because under the measure—and absent faithlessness—a candidate who won the nationwide vote would get the 270 (or more) votes that were committed plus electoral votes from any non-committed states where he had won the statewide vote.[7]  In all likelihood there would be quite a number of those additional votes.  The nationwide vote winner would thus be very likely to have a comfortable margin in the apparent electoral college count.  It would then take elector faithlessness on a large scale to change the outcome, and that is decidedly unlikely.

In contrast, at the more modest—and more attainable—150 vote level, elector faithlessness that could change an election outcome might become more likely.[8]  The nationwide vote winner would have 150 votes plus the votes of non-committed states where he had won the statewide vote, subject again to the possibility of losing some to faithlessness.  But about half of those 150 votes would come from states where the apparent winner would likely have lost the statewide vote.  Electors in those states would presumably be party loyalists and hence unlikely to succumb to arguments for faithlessness.  But in a reasonably close election, it is a good bet that they would come under a lot of pressure to accede to the choice of their states' voters.

In addition, the target states for the two efforts are different.  Some states have faithless elector measures on the books now, but those laws vary a great deal.  The variability could itself be problematic. What is needed is a clear prohibition of elector faithlessness in every state.  The nationwide vote effort, in contrast, does not require universal adoption.  Indeed that is its great strength, for if even widespread adoption were in prospect, then constitutional amendment would be attainable and the legislative route would be unnecessary.  This difference in target states, and the insistence in the nationwide vote effort on pursuing 270 electoral votes, means that the two efforts do not easily go hand in hand.

As I suggested, I view faithless elector action as the more urgent of the two efforts.  Now that the California measure has gone down to defeat, it is time to proceed as expeditiously as possible with a uniform law forbidding elector faithlessness.


1. See Robert W. Bennett, State Coordination in Popular Election of the President Without a Constitutional Amendment, 5 Green Bag 2d 141, 142 (2002) (suggesting a 100-vote coalition); Robert W. Bennett, Popular Election of the President Without a Constitutional Amendment, 4 Green Bag 2d 241, 244 (2001) (suggesting that a coalition between California and Texas could prove effective); see generally Robert W. Bennett, Taming the Electoral College 166-174 (2006) (link) (discussing the merits of a state-driven legislative approach).

2. See Bennett, Taming the Electoral College, supra note 1, at 97-102; see also Robert W. Bennett, The Problem of the Faithless Elector: Trouble Aplenty Brewing Just Below the Surface in Choosing the President, 100 Nw. U. L. Rev. 121 (2006) (link).

3. For a possible elector faithlessness provision, see Robert W. Bennet, Electoral College Reform Is Heating Up, and Posing Some Tough Choices 17 n.4 (Northwestern Public Law Research Paper No. 886777, 2006), available at (link).

4. See National Conference of Commissioners on Uniform State Laws, Drafting Committees, available at (last viewed Oct. 30, 2006) (link).

5. See, e.g., Akhil Reed Amar, Presidents, Vice Presidents, and Death: Closing the Constitution's Succession Gap, 48 Ark. L. Rev. 215, 219, 230 (1995) (link); see generally Bennett, Taming the Electoral College, supra note 1, at 230, n. 20 (collecting commentary).

6. See Ray v. Blair, 343 U.S. 214, 229-231 (1952) (link) (holding that states could empower political parties to require pledges of faithfulness from electors, without reaching the question of whether such pledges could be enforced with punitive provisions). I treat the question at length in Taming the Electoral College. See Bennett, Taming the Electoral College, supra note 1, at 104-114.

7. Two states—Maine and Nebraska—award two of their electors to the winner of the statewide popular vote, and the remaining electors to the winner in each congressional district. See Me. Rev. Stat. Ann. 21-A, § 802 (2005) (link); Neb. Rev. Stat. § 32-710 (2005) (link). In all other states (and the District of Columbia) the statewide winner takes all the state's electoral votes, subject to the possibility of faithlessness.

8. Michael Herz emphasized this important point in an email exchange several months ago.

Copyright 2006 Northwestern University

Cite As: 101 Nw. U. L. Rev. Colloquy 1 (2006),

Persistent URL:

(Download PDF )


(back to top)