I never, in my whole life, thought I would say this, but I need to give a hat tip to Art Woolf at Vermont Tiger for this one.  He notes that the poll done by VPR significantly oversamples people over 50.

Here’s the breakdown from VPR:

(Admittedly, it was just 65+, but I couldn’t help myself).

Since this is an off year election without Obama driving the youth vote, I’m not inclined to compare this to 2008.  I’d be much more interested in seeing how this distribution stacks up against 2006.

Unfortunately, the statistics I could find about 2006 didn’t break down the ages in the same fashion as this survey did, so this makes it all extremely complicated to do a proper analysis.  I.e., 18-24 is one Census block, 25-44 another.  You can’t draw an inference from one to the other since they don’t map along the same lines.

But