Hotline on Call has some sobering numbers on Democratic turnout:
Just 663K OH voters cast ballots in the competitive primary between LG Lee Fisher (D) and Sec/State Jennifer Brunner (D). That number is lower than the 872K voters who turned out in '06, when neither Gov. Ted Strickland (D) nor Sen. Sherrod Brown (D) faced primary opponents.
Only 425K voters turned out to pick a nominee against Sen. Richard Burr (R-NC). The 14.4% turnout was smaller than the 444K voters -- or 18% of all registered Dem voters -- who turned out in '04, when Gov. Mike Easley (D) faced only a gadfly candidate in his bid to be renominated for a second term.
And in IN, just 204K Hoosiers voted for Dem House candidates, far fewer than the 357K who turned out in '02 and the 304K who turned out in '06.
By contrast, GOP turnout was up almost across the board. 373K people voted in Burr's uncompetitive primary, nearly 9% higher than the 343K who voted in the equally non-competitive primary in '04. Turnout in House races in IN rose 14.6% from '06, fueled by the competitive Senate primary, which attracted 550K voters. And 728K voters cast ballots for a GOP Sec/State nominee in Ohio, the highest-ranking statewide election with a primary; in '06, just 444K voters cast ballots in that race.
I have repeatedly argued that Democrats would lose a net 2% of the national vote from 2008 to 2010 just because of the age demographics of the two major coalitions (Democrats are skewing younger and younger, and young people don't vote in midterms). However, these turnout figures paint a picture significantly worse than just the expected 2% drop-off. This is more than just a demographic problem based on age--there really is a meaningful enthusiasm gap.
And yet, despite this, there are still no public, national polls looking for answers on why Democratic turnout is so low. All it would take would be to ask a single, open-ended question to 500 people who voted in 2008, but self-identify as unlikely to vote in 2010, "why don't you intend on voting?" Everyone has theories, but those theories lack empirical supporting evidence and invariably little more than "I speak for all unlikely voters, and they are unhappy for the exact same reasons I am."
Per the article linked above, the DNC is promising to spend $30M on GOTV efforts this year. Surely, they could spend a little of that money on a transparent, representative, scientifically random, poll of unlikely voters of the sort I listed above. A lot of people are going to be working to try and improve turnout this year, and our jobs would be a lot easier if we actually knew what was motivating unlikely voters.