The 2004 Presidential Election and the Electoral College:
By VIKRAM DAVID AMAR
|Friday, Nov. 12, 2004|
This week, analysts have been scrambling to understand what last Tuesday's Presidential election really teaches us. In this column, I'll argue that some of the election's lessons relate to constitutional law and structure - and, in particular, to the electoral college system we use for selecting Presidents.
Overall, the 2004 election serves as further proof that the country would gain much, and lose virtually nothing, by abandoning the electoral college system.
The Basic Critique of the Electoral College and the Relevance of the 2004 Election
In a previous column, I have explained why to my mind the electoral college model was flawed from the outset, and should be replaced with a truly national, one-person one-vote, direct election. I will quickly summarize this argument here.
In the late Eighteenth Century, the electoral college system reflected an unfortunate -- albeit perhaps necessary at the time - compromise. The compromise allowed Southern states, whose assent was needed to get the new Constitution off the ground, to count their slave populations (at the rate of 3/5) for purposes of representation in the House of Representatives -- and thus in the electoral college - even though slaves obviously were not allowed to vote. As a result, a slave state like Virginia had significantly more electoral college clout than a free state like Pennsylvania, even though Pennsylvania had more eligible voters.
Perversely enough, the more slaves that a state bought or bred, the more electors it would get. And if a Southern state were to free any of its slaves, who then immigrated North, the slave state would actually lose strength in the electoral college relative to its free state neighbors.
What was true for slaves was also true for women: A state had (and has) no incentive under the electoral college model to expand its franchise to anyone, for its voice in the college is determined simply by how many persons - voters or not - live in the state.
By contrast, a well-designed system of national direct election could (with federal safeguards in place to prevent fraud and abuse) actually create incentives for each state government to increase voter turnout within the state. Since a state would have a voice only through its voters, it might plausibly think: The more voters, the better.
Proponents of the current system often concede some of the electoral college's tainted origins, but nonetheless try to offer modern defenses to justify its continued existence. Last week's election shows, however, that many modern defenses of the electoral college are makeweight or at the very least overblown.
Myth # 1: The Electoral College Is Needed Because It Helps Small States
Take, for starters, the idea that the electoral college is a good thing to preserve because it gives a boost to small (that is, low population) states. There is a sense in which small states get an exaggerated say in the electoral college; each state, regardless of size, gets two electoral college votes for its two U.S. Senators.
But even assuming that helping small states were a good thing, the electoral college, as it has evolved, does not in fact favor the less populous states on balance.
The reason for this is the winner-take-all system that (almost) all states use for allocating electors. This winner-take-all method gives mid-sized and large states an enhanced voice in the electoral college game. Under winner-take-all, the candidate who wins a plurality of popular votes in a state gets the state's entire electoral bounty. By holding this large prize out to the candidates, a mid-sized or big state - particularly one where the swing voters are in play - is able to create tremendous incentives for the candidates to visit and to make promises to that state in particular.
And, indeed, that is exactly what happened in 2004. The three key states on which the candidates lavished attention down the stretch were all large states whose swing voters were in play -- Florida, Pennsylvania and Ohio. Most analysts (rightly) predicted that the winner of two of three of these states would win the White House, regardless of who won a state like New Mexico. And the candidates took note of this analysis; witness the fact that President Bush visited Pennsylvania a record 44 times during his first term.
Winner-take-all is unlikely to go away anytime soon. Consider another, related 2004 lesson: Colorado voters soundly rejected a statewide initiative that would have split its nine electoral votes.
As Colorado voters probably realized, winner-take-all is preferable for any given state precisely because it gives that state more clout than the state would have if it split its electors (proportionally to the state's overall popular vote, or on a Congressional district-by-district basis). And as long as winner-take-all is the rule in the overwhelming majority of states, the claim that small states are specially favored by the electoral college system will remain a myth.
The truth is that with winner-take-all favoring bigger states, and the "two Senators per state" rule favoring small states, neither large nor small states are specially favored. And that means that even those who favor the small states having disproportionate power lack a reason to prefer the electoral college over a direct national election.
Myth # 2: Inversion -- in Which a National Popular Vote Winner Loses the Electoral College - Is Statistically Unlikely To Recur
Some people resist replacing the electoral college with a national direct election on the ground that almost all popular vote winners prevail in the electoral college anyway. The idea is: No harm, no foul. The electoral college and a direct election will virtually always yield the same result, so why bother to change the system?
There are several problems with this argument, however. First, even if it were true that the two methods would virtually always generate the same result, why should we tolerate any possibility of electoral "inversion" - an electoral winner losing the popular vote?
Second, and more fundamentally, it's not true that the two methods will virtually always lead to the same result. To the contrary, for a variety of demographic reasons, the possibility of inversion is quite real today, and may remain so for years if not decades.
Consider recent history. 2004's numbers show 2000's inverted outcome was not statistically anomalous. It's true that last week George Bush won both the electoral college vote, and the nationwide popular vote - which he took by 3.5 million votes (or so). But inversion was a real possibility.
Suppose the weather in Ohio had been better, or Kerry's campaign had been ever so slightly more effective there. Then Kerry might have picked up an extra 125,000+ votes in that state, edged out President Bush in the electoral college, and won the election. Yet he still would have been behind by over 3 million votes in the nationwide popular vote tally.
Myth # 3: The Electoral College Cannot be Scrapped Because Eliminating It Would Benefit One Party
This very possible inversion in favor of a Democrat in 2004 shows not only that inversion is no anomaly, but that there ought to be bipartisan support for getting rid of the electoral college. Just as the electoral college doesn't really help small states versus large states or vice versa, neither does it help the Republicans versus the Democrats or vice versa. The Republicans benefit from the small-state skew discussed above -- Republicans tend to do well these days among rural whites. But the electoral college also exaggerates the power of big states, via the winner-take-all rules analyzed earlier. And that tends to help Democrats, who win among urban minority voters.
On balance, these two opposing forces largely negate each other. Republicans win more states, but Democrats win more big states. The net effect is to disfavor neither side, but rather only democratic principle. So when inversion happens - and it has happened, and will happen in the future - it could randomly victimize either party. The snake that bit Democrats in 2000 could easily have turned around and bitten Republicans in 2004. (Or in 2000; recall that before the election, many pundits predicted that Gore would lose the popular vote and win the electoral college.)
The solution is to kill the snake: The President should be the popular vote winner, period.
Myth # 4 - A Direct National Election Would Lead to More Recounts and Voter Fraud
Some electoral college fans fear the specter of nationwide recounts if direct national election were adopted. The idea is that if all that mattered were the national vote total, then recounts in a number of places would often be necessary before Presidential election disputes could be resolved. Worse yet, because votes all across the country would be essentially interchangeable, the incentive to steal and manufacture votes anywhere and everywhere would go up tremendously.
But 2004 demonstrates how just the opposite can often be true. George Bush won the national popular vote by over three million votes, and yet 12 hours after the polls closed, John Kerry had not yet conceded because there was a possibility of a recount battle in Ohio.
The truth was that whatever shenanigans and/or improprieties might have existed in Ohio, no one was going to find a whopping three million uncounted Kerry votes lying around across the nation. Thus, if we'd had direct elections, Kerry would doubtless have conceded sooner, and the chance of litigation would have been greatly lessened.
It was because of the electoral college that Kerry supporters would not have needed to find millions of votes, but rather about 125,000 in Ohio, to cast the entire election's result into doubt. With that number comparatively small, a few corrupt Ohio officials favoring Kerry might have thrown the nation into chaos by fraudulently producing the required number of "overlooked" Kerry votes.
The lesson of the Ohio experience is this: A thin electoral college victory may occasion recounts even when there is a dominant national vote winner.
Myth # 5 - The Electoral College Reduces the Chance that the Victor Will Be a Candidate Popular in Some, But Not All, Regions of the Country
The final myth that our recent election debunks holds that the electoral college makes it unlikely that an election victor will enjoy popularity that is not national, but simply regional. 2004 was but one of a number of elections in our history that show that an electoral college winner can, indeed, have mainly regional appeal.
The red/blue map for 2004 shows that the red and blue states are hardly distributed randomly throughout the country's geography. Every red state is bordered by at least one other red state. Every blue state is neighbor to at least one other blue state. (The 2004 state bloc contiguity is even more pronounced than it was in 2000, where New Mexico was a blue island and New Hampshire was a red island.) Bush had regional appeal; so did Kerry. The electoral college system did nothing to curb the chance that a President would be elected whom some regions loved, and others disliked.
There's no doubt about it: The country's Presidential preferences definitely break down on regional lines. (And as striking as the red/blue state maps are, the red/blue county-by-county map is even more intriguing. It shows how people living near major bodies of cold water vote Democrat, and those who do not tend to vote Republican, a phenomenon discussed by John Tierney in the New York Times.)
2004's results confirm that a person with much, much stronger appeal in some parts than others can win the White House. Yet the 2004 map is hardly unique in American history. The maps of the elections of 1860, 1986 and 1924 all demonstrate nicely how the electoral college does not guarantee that the country will enjoy the benefit of having a President with strong voting support in all regions of the country. (The 1896 map is particularly interesting, showing an almost identical state-by-state breakdown as 2004, but with the Democrat and Republican states inverted; the parties' regional strengths have flipped almost completely over the course of a century!)
So contrary to myth, an electoral-college-picked President won't necessarily garner national - as opposed to regional -- support.
Indeed, a direct national election might tend to make the red states less red and the blue states less blue.
Because Kerry knew that he couldn't win over a plurality of voters in any southern state, he simply stopped trying to woo any southern voters. The same is true for President Bush in New York and California. The lopsided margins of victory in all those states was in part of function of one side giving up because winning over more voters in a state does not matter unless a candidate wins the state overall.
Granted, a direct national election might not result in a breakup of the Confederacy or West Coast voting blocs. But if campaigning - like the election - went truly national, the margins of victory in many states might be reduced. The result would be a President whose support is slightly more geographically balanced.