Cross posted from Future Majority.
Once again, Gallup is trumpeting data claiming that youth interest in the election is the same as it was on 2004. The headline on their latest piece is "Little Evidence of Youth Surge." I've already seen this picked up on half a dozen blogs, including on Digby.
Here's what everyone should know about this Gallup poll:
- The margin of error for the youth sample is huge: +/- 7%.
So while Gallup's poll may show that young people are no more engaged than they were in 2004, and are still less likely to vote than older demographics . . .
. . . the margin of error is quite large. In addition, it may be hard to remember now, but there was a LOT of excitement in 2004, and it was excitement we saw on both sides of the aisle during a very close contest. By contrast, this election has all the signs of a blowout, and McCain's campaign has zero youth outreach and is actually kicking their own young supporters out of their events. There's a decent possibility that young McCain supporters are acting as a drag on these numbers. Look at the disproportionate amount of time and resources each campaign is devoting to energizing its young supporters (from Gallup's own data):
Regardless of self-reported measures of interest, the most interesting youth-vote statistic to watch on Tuesday won't be turnout, it will be the margin of victory that Obama enjoys over McCain. As I wrote yesterday in my post about Tips for Reporting the Youth Vote:
5. The margin of victory among young voters may be just as important as the overall increase in youth turnout. In 2004, 20 million young voters cast a ballot, with 54% selecting John Kerry. That gave Kerry an advantage of 1.6 million votes over President Bush among young voters. This year, if 22 million young voters cast ballots and 62% choosing Obama vs. 38% for McCain (numbers roughly found in most polling), that would give Senator Obama an advantage of 5.28 million votes.
One more thing about the Gallup poll - they are consistently underestimating youth share of the electorate, even in their "expanded likely voter model."
As a result, 18- to 29-year-olds now constitute 12% of Gallup's traditional likely voter sample, basically the same as the estimate in the final 2004 pre-election poll (13%). Gallup's expanded likely voter model, which defines likely voters differently (on the basis of current voting intentions only), estimates a slightly higher proportion of young voters in the electorate (14%). However, even if the share of the youth vote were adjusted upward, doing so has little or no impact on the overall Obama-McCain horse-race numbers using either likely voter model.
I'm loathe to look into a crystal ball and predict youth turnout, but young voters made up 16 - 17% of the electorate in 2004 (depending in the source of the data, Current Population Survey and Exit Polling, respectively). Gallup is setting the youth share in their models at 12 - 14%, at least two points lower.
For anyone taking the Gallup poll - or any other poll prematurely calling the youth vote - seriously, I recommend reading and spreading my Tips for Reporting on the Youth Vote. It's a good way to avoid the hysteria and common mistakes that frequently surround youth vote reporting.