Pew poll shows post-debate enthusiasm gap, but perhaps not much else
by David Atkins
The political world is buzzing over the latest Pew poll showing Romney up by four points after the debate. There will also apparently be a PPP poll coming out showing Romney with a national lead as well to confirm the trend (despite some interesting counter-evidence from Gallup showing Obama up by five, a shift back toward Obama from Rasmussen and a static nearly 4-point Obama lead from RAND.)
So the movement is real. But what does the movement tell us? Let's look at the Pew internals for a moment.
First, the bad news: the sample is weighted slightly toward women with 56% of respondents being women. Which means that from that perspective, the numbers for the President should be even worse.
But looking down at the other categories we start to see some eye-popping numbers.
- For starters, a full two-thirds of the respondents were over 50 years old. Is that likely to be the shape of the electorate? Very likely not.
- A full 77% of the respondents were white. That is almost certainly not going to reflect the final electorate.
- A large preponderance of the respondents were from the South (449), with the next highest total from the Midwest (294), and only 219 from the Northeast and 239 from the West. There will not be twice as many voters from the South in the election as from the Northeast or the West.
- Finally, more respondents claimed to be Republicans than Democrats, which would destroy the President's chances in November automatically. It's possible for the final electorate to resemble that Party ID, but unlikely.
Does any of this mean the poll is inaccurate or "skewed"? Not necessarily--and certainly not when it comes to the fluid, enthusiasm-based Party ID number.
What it does seem to mean, however, is that enthusiasm can make a big difference in the polling picture. And that goes in both directions.
Most pollsters don't get more than a 10-12% response rate on their polls, if that. Only the most interested people tend to answer the phone when a pollster calls. So the polling prior to the debate almost certainly overestimated Obama's national lead as dejected conservatives refused to answer the phone, but would likely have trudged to the polling place or sent in their mail ballot regardless.
Similarly, conservative-leaning groups (mostly older, whiter Southerners) seem very excited to answer the phone now.
What all this means is that the election is likely quite close in reality--and always has been. What seems to change more than anything is enthusiasm.
And that in turn means that what matters most is simply turnout. If you'd like to prevent Mitt Romney from taking over the White House, if you want to keep McConnell out of Senate leadership, if you want to put the gavel back in Nancy Pelosi's hands, and if you want to turn back the tide of Republican control of state legislatures across the country, there's only one thing to do: get involved, get on the phones, knock on the doors, and help turn out the vote.
Yelling at the President in online comments isn't going to do anything. Barack Obama is either going to perform well in the following debates or he isn't. But regardless, enthusiasm is still what seems to matter most.
Update from digby:
I just wanted to add this analysis from the Guardian:
Conclusion: we can't take too much from any one poll
There are reasons to believe the Pew poll is probably too pro-Romney. Still, with this normally Democratic-leaning poll, and the Daily Kos- and SEIU-sponsored Public Policy Polling national poll (coming out tomorrow) also having Romney in the lead, it's fairly safe to assume that this is not down to some conservative push to call this race tied.
This is what makes living in a polarized country so invigorating.
.