|1. Isn't the problem worse than 2-3%?
Two recent polls suggest that Democrats are losing more than just 2-3% from a lack of enthusiasm among the base.
First, Daily Kos polling currently shows that 54% of Democratic self-identifiers say they will "definitely" or "probably" vote in 2010, compared to 75% of Republican self-identifiers. When this enthusiasm gap is factored into the Daily Kos generic congressional ballot, a Democratic advantage of 38%-35% becomes a Republican advantage of 43%-38%, for a net shift of 8%.
Second, Bloomberg polling (PDF, page 5) currently shows that Democrats lead 36%-35% among all adults, but that Republicans lead 42%-38% among those most likely to vote. For this poll, Republicans gain 5% due to relatively greater enthusiasm.
However, while the Daily Kos and Bloomberg polls are useful data points when measuring of the general voter turnout problem Democrats face, neither are good measures of the specific enthusiasm problems they face in 2010. This is because Democratic voting groups are always less likely to vote than Republican groups, even in Presidential elections, and even in 2008. For example, African-Americans, the strongest Democratic demographic of all, have always turned out at lower levels than whites, even in 2008 with Barack Obama on the ballot.
Demographics that vote Democratic tend to be relatively young, relatively poor, and relatively less registered to vote. All of this translates into lower turnout for Democratic voting groups. This is a problem Democrats always face, and is more endemic to the state of our public sphere and civic space than to the actions of any political party. Properly measuring the 2010 enthusiasm gap requires determining how much worse than normal the turnout problem is for Democrats, rather than how large the problem is overall (which is what Daily Kos and Bloomberg are doing).
2. No, it really is 2-3%
Democracy Corps is the only polling outfit currently measuring the specific 2010 Democratic enthusiasm gap. In their most recent, mid-November poll (PDF), DCorps provided a subset of "drop off" voters who actually voted in 2008, but are not likely to vote in 2010. When it comes to determining the specific 2010 enthusiasm gap, this is more accurate than measuring the difference between Democratic performance among all registered voters (or all adults) and among likely voters.
What Democracy Corps found is that "drop-off" voters favor Democrats by a whopping 53%-36%. This compares to a narrow, 47%-45% Democratic advantage of likely voters. If drop-off voters were included in the overall sample, Democrats would lead 48%-43%. While this is only one data point, it means that the enthusiasm gap is currently costing Democrats about 3% nationally.
3. Midterms electorates are worse for Democrats than presidential electorates.
There is a further caveat to these numbers. Long-term data from the census bureau indicates that the turnout gap between Americans above and below the age of 45 widens significantly in mid-term elections. For example, over the last nine Presidential elections, Americans aged 45-64 turned out, on average, at a rate 12.7% higher than Americans aged 25-44. However, in mid-term elections, the average gap over the last nine cycles has been 17.1%. Given that Obama won 55% of the vote among Americans aged 25-44, but only 50% of the vote among Americans aged 45-64, this "natural" turnout problem facing Democrats in mid-term elections also makes the specific problems they face in 2010 appear more pronounced than it actually is. Midterms electorates are worse for Democrats than presidential electorates.
4. Mind the gap
One more quick note--quite a few bloggers are citing the Daily Kos poll showing that only 54% of self-identified Democrats who are registered to vote indicate that they are likely to vote in 2010. While that sounds pretty bad, the truth is that less than 40% of all voters have turned out in every midterm election since 1974. If 54% of self-identified Democrats who are registered to vote actually turned out in 2010, then Democrats would end up with about 64 seats in the Senate and close to 300 in the House. A 54% turnout rate among Democrats who are registered to vote would be fucking enormous for a midterm election, and as such is a terrible data point to use when arguing that the Democratic base is depressed. What matters in the Daily Kos poll is the gap between Democrats and Republicans, not the overall numbers.
5. Who are the drop-off voters?
Still, no matter which way the numbers are sliced, Democrats are facing an enthusiasm gap that is worse than normal. It would be helpful if there were more polls like Democracy Corps measuring this group, so that we had more data to determine who they are and what makes them tick. What little data we have from Democracy Corps indicates that they tend to be young and unemployed, and that they are more likely to be female and non-white than other voters. None of this is very surprising--these voters are not seeing any improvement in their lives, and so perhaps they are growing more cynical about participating at all.
The Democracy Corps poll also does not indicate that this group is particularly left-wing. Among the drop-off voters, 22% self-identify as liberal, compared to 19% of the sample as a whole. As such, what little data we have indicates that Democrats are facing a stronger than usual enthusiasm gap because they aren't delivering for the base that votes Dem due to their economic fragility, rather than for the part of the base that votes for more ideologically oriented reasons.
That isn't to say that ideology isn't actually a problem for Democrats in D.C. If, as a group, they had been more left-wing, they might very well have passed legislation and policy that would have done better for the economically fragile. The problem isn't passing centrist, neoliberal policy that will anger ideologically oriented activists, but in passing centrist, neoliberal policy that won't make anyone's lives better.