What Happened with the Colorado Exit Polls?

As you saw on the 2008 Youth Vote Map, Colorado is one of the states for which we have no exit poll data on young voters. Early in the evening on election night, CNN reported exit poll data showing Obama losing the Colorado youth vote 47 - 53%. Based on that data, the NY Post wrote:

Colorado was the only state that Obama flipped from Bush's 2004 result without winning the youth vote - he lost 47-53 among that age group there, perhaps due to the large military presence in that state.

The story was also picked up by Politicker.

Unfortunately (or, fortunately), the story turned out to be false. Something was dramatically wrong with the exit polling data, and before the night was through CNN replaced the 47 - 53% youth figures with "N/A." As of this writing, the exit poll still shows "N/A" for its 18 - 29 data.

No one seems to know what went wrong (though I've got a few inquiries out at the moment), however the folks at New Era Colorado were able to get a local pollster to go on record about state polls leading up to election day. Here's what they had to say about the Colorado youth vote:

RBI conducted a number of statewide surveys among likely voters over the course of the 2008 election season and Obama was never behind in the Colorado youth vote. In fact, RBI's last statewide survey just days before the election showed Obama receiving over 60% of the 18-34 vote.

Yes, RBI's numbers are for 18 - 34, not 18 - 29, but it seems unlikely that the numbers would be substantially off considering that the Gen Xers in the 30 - 34 category tended to be more conservative, rather than more liberal. Any bias they introduced into the data would likely swing in McCain's favor, not Obama's.

It always seemed fishy that Colorado youth would swing towards McCain. The state is trending blue for a while now, it's got a great progressive infrastructure, and a growing Hispanic population - all recipes for Democratic support. It's good to know that is likely still the case. I'll post more if I get a good explanation for why the exit polling was so flawed.