The fantasy of closure

There is an old joke that comes to mind when reading about the Washington governer's race. A tour guide in the Museum of Natural History takes his group past an exhibit of dinosaur bones. One of the children ask him how old the bones are. He replies "sixty million and two years old". The child's father ask him how he knows that, and the guide replies "Well, when I started working here two years ago, they told me that they were sixty million years old".

Recent events have thrown the issue of accurate voting into the public spotlight. From the Florida in 2000, to Ohio, Washington State and Ukraine in 2004, fierce partisan voices have called for recounts and revotes when the outcome is close. This is an entirely different issue than that of election reform, in which systematic sources of error and disenfranchisment are hopefully purged from the voting process.

In a winner-take-all system, tiny margins are of huge significance. A baseball team can lose the World Series becauses of a single error taking two seconds after seven championship games and a long season stretching over a year. And in sports, that's the outcome. Final. Winner and loser. Wait 'till next year.

But in politics, there are very different implications of a close election. First of all, nearly half of the electorate wanted someone else to win. And as citizens, they have just as much right to the services that we hire our president to perform as do those who voted for him. Unfortunately, the way that politics is played in this country, the winner doesn't see it that way. The concepts of a "mandate" and "political captial" speak to just the opposite conclusion- the winner is primarily the president of his supporters.

All this becomes magnified when the election is too close to call, and that is where an understanding of mathematics and the scientific method is important. Leaving aside issues such as systematic or malicious disinfranchisement, many people think that if the votes could just be counted correctly, the true winner would emerge. Unfortunately, this is not possible.

Any scientific test has a margin of error. The joke above is funny (well, I thought it was funny) because we recognize how rediculous it would be to make such an accurate measurement of such a large number. Voting is like that- no matter how carefully you look at hanging chads, almost any workable system that allows you to poll millions of people on a single day is always going to have some degree of inaccuracy. The Washington Governer's race was initially decided by a margin of 261 out of 2.8 million votes (giving the win to Republican Dino Rossi) and then 42 votes after a recount (again giving the win to Rossi), and finally after a third and certified recount, giving the victory to the Democrat Christine Gregoire by 129 votes.

Regular readers of this column (Hi, mom!) will know what a partisan I am. However, the comment of the Democratic spokesman Morton Brilliant- "It's irresponsible to spend $4 million in taxpayer money on a new election just because you don't like losing this one."- smacks of the same kind of scientific misunderstanding that was thrown around by Republicans during the 2000 Florida recount (remember "Sore Loserman"..?).

However- and I say this in the spirit of bipartisanship- Gregoire did NOT win the election. And Rossi did not win the election. It is a tie. A statistical dead heat. If you recounted 50 times, the results would probably go one way half the time and the other way the other half of the time. This is because of the inherent inaccuracy of the process. This is different than a revote, where you are measuring a different thing, as opposed to remeasuring the same thing. By the time a vote is repeated (as in Ukraine), many factors have been in play which may change the actual outcome.

But we can't govern if we allow ties, so there has to be some process for picking a winner in a close election. This is where the state constitution and election lawyers come in. Brilliant's underlying point is true- at some point, you have to declare a winner and stop spending resources on an inherently flawed process. And those who voted for the declared loser will just have to move on.