Public Policy Polling for Daily Kos & SEIU (4/7-10, MoE: ±3.1, registered voters, Obama trendlines 3/31-4/3, all others 3/25-27):
Why the upturn for Dems and the drop for Republicans? Mostly it's because we wound up with an unusually Democratic sample this week, 43 D, 34 R. Usually the proportions are a lot closer, with Dems just a few points ahead. As you know if you follow our polling closely, PPP weights for gender, age, and race, but not for party identification. There are solid methodological reasons for this decision, chief among them that party ID can simply fluctuate too much on its own naturally. As I've pointed out before, when you wake up each morning, you typically are the same race, sex, and age you were yesterday. But depending on how you feel that day, you might change your mind about your party affiliation like you change clothes.
And again, this is why the long-term trends almost always matter more than the weekly changes. I expect the D vs. R spread in our sample to shrink back to "normal" next week. But what's interesting to me is that despite these shifts in our samples, Obama's approvals have stayed in a very tight band all along - usually somewhere around dead even, occasionally wandering above and sometimes wandering below.
Interestingly, the same has been true of most of the other figures we test regularly, including Republican Senate Minority Leader Mitch McConnell. The one guy who has really seen his numbers trend meaningfully is Speaker John Boehner, as we pointed out last week. Such are the perils of moving from relative obscurity to prominent leadership. Everyone already knows Barack Obama and has an opinion of him. But the country is still just getting its first taste of Boehner, and it looks to be an unpleasant one. And if Republicans take back the Senate next year, I'd expect McConnell's numbers to follow a similar path.