The National Popular Vote Plan: If Ohio, Virginia and Florida Alone Sign On, They Will Change Primary Politics Profoundly, and Greatly Increase the Chance that the National Popular Vote Winner Will Prevail

In this column, I'll explain why certain key 2008 battleground states for the general election - such as Ohio, Virginia, and Florida - could eliminate the relevance of electoral college math for the current Democratic primary.

The Shadow the Electoral College Is Casting on the Democratic Primary

The Democratic contest between Senators Clinton and Obama has highlighted the issue of "electability" like never before. Granted, analysts and voters in prior presidential election years talked a fair amount about the electability of various primary contestants. In 2000, for example, George Bush's perceived "electability" advantage helped him secure the Republican nomination early, and in 1992, Bill Clinton's perceived "electability" problems dogged him throughout the primary season.

But what is different about 2008 is that the electability discussion is tightly focused not on national appeal and national poll numbers (the way it was in the Nineties and even 2000), but rather on the few key battleground states that will almost certainly dictate the result in this year's electoral college. Indeed, Senator Clinton's backers have been increasingly arguing that the main thing Democratic Party superdelegates should look to, in resolving who will be the nominee this summer, is who can beat Senator McCain in the contested swing states in the fall -- not who won more votes nationally in the Democratic primary process in the spring. As Professor Sandy Levinson from the University of Texas's law school has put it so eloquently and colorfully, "the electoral college virus is worming its way into the Democratic primary process."

Of course, Senators Obama and Clinton seem to disagree on how they stack up against Senator McCain in the key swing states. It is hard to fully assess their claims here -- especially if one looks to actual primary results, as opposed to polls. One big reason for that difficulty is this: How Clinton and Obama fare against each other in a Democratic primary race in a given state may not tell us much about how either would fare in that state in the November election against McCain. For example, Senator Obama's victory over Senator Clinton in Wyoming does not suggest any real possibility he could carry that state in November, and Senator Clinton's victory over Senator Obama in New York similarly does not indicate that he wouldn't beat McCain there in the fall.

With the Likely Battleground States Already Quite Clear, a National Popular Vote Plan May Be Even More Appealing and Easy to Achieve

As much as they disagree, Senators Obama and Clinton may well agree even now (and most analysts may concur) about which states are the key battleground arenas that are "in play" for the fall, with voting outcomes that will almost certainly dictate the electoral college result. Among those states are Ohio, Virginia, and Florida.

This effective narrowing of the Presidential election to a handful of largely-already-known states highlights the possibility for reforming (and essentially eliminating) the electoral college scheme itself. As I have written before, there is a National Popular Vote (NPV) plan movement underway. The idea is to get various states to sign onto an "interstate compact" that would require each signatory state to cast its electoral college votes not for the candidate who happened to win a plurality of popular votes in that state, but rather for the candidate who won the most popular votes nationally. This system, with enough signatories, would ensure that the winner of the Presidential contest would always be the person who had won the largest number of votes from individual voters nationally. Importantly, the compact, by its own terms, does not go into effect until enough states to comprise a majority of the electoral college - that is, states whose electoral college allotments total 270 or more -- ratify it.

There are many exciting recent developments concerning this plan. Since I last wrote about the plan, the state Senates of Washington and Vermont have passed it, as have both houses of the Hawaii legislature. But the recent Democratic party primary discussions and arguments have caused me to consider some additional questions: First, does an NPV plan really need states that can contribute a total of 270 electors -- yielding a formal, mathematical electoral college majority no matter what the election results may be -- to work? Or, could an NPV plan work if just a few battleground states -- say, Ohio, Virginia and Florida -- signed on.

The possibility of just a few states being able to effect electoral change along these lines was suggested as early as 2001-2002 by former Dean Robert Bennett of Northwestern's law school, around the same time my brother, Yale law professor Akhil Reed Amar, and I were beginning to write about possible state coordination efforts in this area.

Dean Bennett argued in 2002 that a state would doubtless be more inclined to sign onto a plan when it knows other states are committed as well: "Whatever the possibilities for one state going it alone, or some combination taking the action independently, it seems clear that one state would be more likely to make the move if it could be assured that others would join in." But is it rational for states to need (or at least want) company in order to act here? And if there are rational explanations for states' wanting company, are they powerful enough to discourage states like Ohio, Virginia and Florida from forging ahead on their own? These are key questions, as to which I offer some preliminary thoughts.

One Reason Why a Given State Might Want A Lot of Company Before Committing to the NPV Plan - and Why This Reason is Inapplicable With Respect to 2008's Battleground States

One reason a given state might insist on involvement by a substantial number of other states before committing is that the particular state in question wants to be sure that its actions will accomplish the desired effect of moving to a national popular election - and thus forcing Presidential candidates to campaign for all votes nationally, regardless of how "in play" any given state is.

The virtue of the NPV plan is that if states tallying 270 electoral college votes sign on, then the signatory states (and the presidential candidates) can be mathematically assured that the national popular vote winner will prevail in the electoral college. Candidates would then know for sure that they must pursue a national vote-getting strategy, and there would be no uncertainty plaguing the system or the participants (save for the ever-present uncertainty caused by possible litigation challenges to the NPV statutes.)

But given the (seemingly undisputed) demographic/partisan reality in 2008, as to which states are overwhelmingly likely to be determinative battleground states, a move by Florida, Ohio and Virginia together to give their electoral college votes to the national popular vote winner would create almost the same kind of complete certainty: No candidate could hope to win the White House without getting the electoral college votes of at least one of those states. So coordinated action by these three states alone (and perhaps even a two-state subset of these) would be sufficient.

Another Reason Why a Given State Might Balk Before Endorsing the NPV Plan - and Why It Is Also Inapplicable to 2008's Battleground States

There is also a second reason a state might rationally be reluctant to allocate its electoral college votes to the national popular vote winner. The reason is that the national vote winner may be unpopular in that state (and lose there by a significant margin).

Dean Bennett invoked this possible explanation for why Governor Schwarzenegger vetoed the NPV plan in California over a year ago: "Under the legislation," he wrote, "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." Understandably, Governor Schwarzenegger may not have wanted to defy his own state's voters' strong will to defer to the national will - a move that might have sunk his own popularity in California.

But my present focus on present-day Ohio, Virginia and Florida answers that problem. These three states are, as noted above, considered by virtually everyone to be "in play" for Election Day 2008 - that is, their voters are relatively equally divided between the two major parties and their candidates. So even if, say, Ohio (by pledging to cast its electoral college votes for the national vote winner) ended up giving the Presidency to someone who lost in the state, any loss there would have been by a pretty small margin. For all practical purposes, these battleground states are close to being "ties," in which both candidates are equally attractive to the state's electorate. Thus, while a state's opting for NPV might alienate half its voters, it will curry favor with the other half, and it will be difficult to make a strong argument that the will of the state as a whole has been unfairly overridden by the NPV plan.

The Third Reason Why a State Might Balk At Endorsing the NPV Plan

A third and related reason why a state might be reluctant to act is that acting could involve non-trivial sacrifice. Take Ohio, for example. If Ohio were to allocate its electors for the national popular vote winner, Ohio would be making the question of who wins the most votes in Ohio, in particular, less relevant than it is under the current scheme. In turn, the presidential candidates in the fall would have less incentive to win over the median Ohio voter, and more incentive to win over the median national voter. That might translate into less attention paid to - and fewer or less generous promises concerning - Ohio-specific issues.

This might be an impediment to change in most swing states. Yet it would remain a problem whether one state moved ahead on its own with the NPV plan, or whether three states moved, or whether eleven or even twenty states moved together. Either way, every significant swing state - that is, every state of decent size in play - would be giving something up under a national popular vote approach.

Now, perhaps swing states like Ohio and Florida that currently hold clout will never want to participate in any national popular vote movement out of selfishness. Perhaps the NPV plan is possible only if to the extent that there are enough states who are not in play who are have reasons, and are willing, to sign on.

But if, somehow, people in just a small number of states like Ohio could be convinced that their sacrifice, while historically momentous, would not be so great in tangible terms, or is worth the cost, then meaningful reform could be attained with much less national energy that might be supposed.