Thursday, November 06, 2008

Tune on, drop in, turnout...

I hate all these contextless statistics being bandied in attempts to prove various tenuous points about the American election.

In particular, I find turnout percentage stats not very useful, given that they don't accurately reflect what I would consider to be turnout.

That's because the state-by-state turnout stats are based on the number of registered voters, not on the population of eligible voters.

Look at Virginia, for instance. 3,223,156 voted in 2004 and 3,474,202 voted in 2008. That's approximately an 8% increase in turnout, right?

Well, no, because registered voters increased from 4.5m to 5m between 2004 and 2008.

So even though 250,000 more people voted in 2008 than 2004 in Virginia, officially voter turnout went down from 71.3% to 69%.

(These numbers are all from Virginia State Board of Elections.)

As for making sweeping statements about racial division based on exit polls - exit polls, people - in a single state like Alabama, that's even less useful. I think the most telling stat that I've heard is that a higher percentage of whites across the country voted for Obama than any Democratic candidate since Carter.

Saying that racial division persists in the USA is a truism. But saying that race played a negative or positive role in this election - difficult to prove.

No comments: