Once again, I became nitpicky and wanted to look at the poll's internal demographics. Here's what I found: The Daily Kos poll demographics look more like the electorate in 2010 than in 2008. Actually, the Kos sample is more conservative than even the ultraconservative 2010 electorate. And yes, if you believe the makeup of the electorate in 2012 will be the same as or more conservative than the makeup in 2010, Obama will lose. Also, I have a bridge in Brooklyn to sell you.
I would like you to notice a couple of things here:
- On ideology, Daily Kos "likely voter" model assumes a higher portion of the electorate to be conservative this year (44%) than was in 2010 (42%), and they assume a much lower portion of the electorate to be liberal this year (16%) than even in 2010 (20%). Compared to the actual vote of 2008, the conservative advantage is more than double in the Kos poll (+28 vs +12).
- On party ID, the Daily Kos "likely voter" model assumes the 2012 electorate to be roughly even between Democrats and Republicans (D +2 with 39%-37%), essentially the same as the 2010 electorate (even at 35% each). Relative to the actual vote in 2008, the Kos model is oversampling Republicans by 5 points.
The internals of this poll make about as much sense as the usual pathetic hair-on-fire Obama-sucks rants that cover much of the "Left" media. And that's the rub. The prognosticators aren't panicking because of the polls. They are commissioning really weird polls because they want you to panic. Because if you panic, you will give them more attention, more click, and perhaps more money.
So, don't panic. We're going to win this thing, and we are going to win it decisively. Keep getting out the vote. Keep working hard. Keep moving forward.