Over the summer, as conservative energy came to its anti-tax, anti-spending, anti-stimulus crescendo amid a wave of town hall and tea party-style protests, Republicans started overtaking Democrats in generic House balloting--polls that ask people, regardless of the candidates they may get to choose from, which party they're likely to vote for in House of Representatives races in the 2010 midterms.
No less than 10 polls in August and September had Republicans leading, and one Rasmussen survey had Republicans lup by 7 percentage points. Lots of polls had Democrats ahead, too, and the average, according to Pollster.com, never quite swung into Republicans' favor--but, for a moment in August, it was very close.
Now things appear to have swung back into Democrats' favor, at least slightly.
Rasmussen (which is typically more friendly to the GOP) still has Republicans
ahead, but polls released in the past few days by Ipsos/McClatchy and CNN
have Democrats leading 48-41 and 51-41, respectively, though, when CNN
restricted its sample to registered voters, Dems were only ahead 50-44.
See Pollster.com's track of the progression (which does not yet include the Ipsos/McClatchy poll) here:
Conventional wisdom says Republicans will take a few seats back in 2010. There's a natural ebb of Democratic success after that they took control of Congress in 2006 and the White House in 2008, an eventual 60-seat majority in the Senate. They're also defending more seats than Republicans (CQ recently rated the top ten most vulnerable House seats, and nine of them are held by Democrats).
And, for all we know, they will. The question, right now, is how many they'll take--how big a dent they can make in the Democratic majorities--and that's expected to depend on big factors like the economy, whether Democrats can pass health care and energy reform, what those reforms look like, what voters in conservative Democrat-held districts think of those bills, whether populist conservative energy is maintained, and whether the conservative grassroots have enough ground presence to affect races across the country.
All those are still very much up in the air, but, for the time being, things are looking a little bit better for Democrats since the end of the summer.
Passing legislation is the best way for Dems to retain their seats. Lead.
Seconded. Or, as I prefer to put it: Do your damn job.
I wonder is this National generic ballot question is very useful any more. The regional differences are pretty stark. The Republican Party has great strength in the South and almost no strength in the Northeast. These two regions skew the results a bit as most of the competitive seats will be in other regions of the country. Perhaps somebody should start filtering (or asking) the generic ballot question region by region. And perhaps that is the best way to report these numbers in the future.
Cheers
Rasmussen was, indeed, one of three polls which nailed the popular vote in 2008, which is a number which has never held any meaning in presidential elections in this country. So, yes: They really nailed a meaningless number.
This doesn't mean that Rasmussen is biased or not (their party weights make them less likely to respond to changes in party affiliation, which is partly balanced by their huge sample sizes). It does mean that talking about "generic 2010 ballots" and "popular vote predictions" is just a space-filler for political news junkies. We might as well start triple-spacing the blog entries and increasing the margins to make it seem like there is more there...
I guess characterizing Rasmussen as "typically more friendly to the GOP" is a way for a liberal to fight off the bad news. Gee, I seem to to recall that it was Rasmussen--not the others you cite--that nailed last year's popular vote. Gee, doesn't Rasmussen do likely voters?
Likely voters for a race a few weeks out makes sense. Likely voters for 2010 in 2009 doesn't.
I don't think the generic ballot question is useful. The polls the liberal have commissioning have been showing bad news. Rasmussen questioning has bias and it needs to be accounted for and after that's done it is useful polling firm.