This should be obvious to anyone watching polls for this election, but I just wanted to point it out for the benefit of any Democrats who are in a position of decision-making power for this election. There is no way we can win this election unless we turn out Democrats at high levels. Here are the last five
generic ballot polls of registered voters:
- USA Today / Gallup (9/17): 51% D--42% R
- AP-Ipsos (9/13): 51% D--39% R
- Pew (9/10): 50% D--39% R
- ABC (9/7): 50% D--42% R
- Gallup (9/7): 53% D--41% R
The five-poll registered voter average from these polls is 51.0% D, 40.6% R, with very little variation among the five polls. An advantage of 10.4% is excellent for Democrats, no matter how you look at it. However, in contrast to the situation among registered voters, here the are the last five generic ballot polls for Democrats among likely voters:
- USA Today / Gallup (9/17): 48% D--48% R
- Zogby (9/14): 37% D--34% R
- Fox, (9/13): 41% D, 38% R
- AP-Ipsos (9/13): 53% D--39% R
- CNN (9/2): 53% D--43% R
That comes out to an average of 46.4% D, 40.4% R, for a margin of 6.0%. That is 4.4% closer than the registered voter generic ballots. Further, in most of the likely voter polls, the Democratic advantage disappears almost entirely. Fox / Opinion Dynamics goes from a 16-point Democratic lead among RV's, to a 3-point Democratic lead among LV's. Gallup goes from a nine-point Democratic among RV's lead to a tie among LV's. Back in February,
Gallup predicted as much when they wrote (emphasis mine):
Typically, this early in the election year Gallup does not ask the "likely voter" questions that are designed to identify the subset of registered voters (RV) who are most likely to vote. Still, our experience over the past two mid-term elections, in 1998 and 2002, suggests that the RV numbers tend to overstate the Democratic margin by about ten and a half percentage points.
Now, turnout in primaries have consistently shown advantages for Democrats, and when it comes to likely voters,
polling firms such as Ipsos and Gallup are clearly taking polls of different countries. Further, there are serious problems with conducting likely voter polls this early in the campaign, since,
as Robert Erikson, Costas Panagopoulosm and Christopher Wlezien put it, "likely voters early in the campaign do not necessarily represent likely voters on Election Day." That is why I have consistently stated my opposition to using likely voter models until mid-October when early voting actually begins.
However, I simply cannot imagine any way for Democrats to win this election if Republicans turn out at higher rates than Democrats, as the Gallup poll LV model must undoubtedly show them doing. If we hold a nine-point edge among registered voters, than we should also hold a nine-point advantage among likely voters. If we fail to do so, then we have utterly failed in our field operations, our message, and our media strategy. You cannot find a nine-point nation swing within Independents. According to 2004 turnout levels, that would require moving 17% of Independents out of the Republican column and into the Democratic column. According to 2006 turnout levels, where Independent turnout will probably be even lower, an even larger swing will probably be required.
Democrats must turn out in higher rates than Republicans in order for us to win this election. This is requires a media strategy that targets African-Americans, since African-Americans are the ultimate Democratic base demographic. This requires
a Latino GOTV program that is not vaporware. This is going to require a field strategy that targets Democrats,
not just trying to increase the vote. This requires candidates
to actually say they are Democrats in their commercials. Above all else,
this is going to require not ducking Iraq, since nothing is on the mind of voters, especially Democratic voters, more than Iraq. If we lose on November 7th, and wake up the next day with Republicans still in charge of both branches of Congress, it will be because we did not do a good enough job getting Democratic voters to the polls. We have made real gains in partisan self-identification over the past two years relative to Republicans (see
here,
here and
here). If we end up getting beat in turnout anyway, as the Gallup and Fox polls suggest we will, then our leadership will have squandered the best chance to retake Congress in a generation because they simply failed to see their own base as a meaningful part of the overall campaign equation.