There's a very curious thing going on in our national polling right now: voters are down on Barack Obama and it's looking more likely that there could be a Republican in the White House in 2013. But they're also way down on the Republicans in Congress and Democrats have now led on 11 consecutive PPP generic House ballot polls dating back to February. Could Democrats gain a lot of seats in the House, possibly even a majority, while simultaneously voting out Barack Obama? It would certainly run counter to all historical precedent but these are weird times politically.
PPP's newest national poll finds that Americans think even less of the new Republican majority in the House than they thought of the Nancy Pelosi led Democratic majority. 42% of voters think the GOP is doing a worse job than the Democrats did while in power to only 39% who believe they're an improvement. 18% say the new majority's performance is 'about the same' as the old one's.
Democrats lead this month 46-44 on the generic House ballot. That's a 9 point shift from the Republicans' 7 point margin of victory on that measure in November's election and although it might not be enough to give them back control that margin would almost definitely translate to Democrats winning back a lot of the seats that they lost last fall. The GOP does maintain an 8 point advantage with independents at 41-33, but that represents a significant decline from their 19 point victory with them according to last year's exit polls.
John Boehner is not proving to be popular with the American public. 33% of voters approve of the job he's doing to 37% who disapprove, with independents splitting against him by a 34/37 margin almost identical to the overall numbers. He does have one thing going for him though- with Republicans he's at 55/15, suggesting there's not too much opposition to him within the party. We also found on our GOP primary poll earlier this week that 41% of Republican voters have a higher opinion of Boehner to just 18% who picked Eric Cantor so there's not any appetite for a change in leadership at least with the public at large.
The House landscape has shifted dramatically in the last 9 months.
Full results here
A relatively small change in the generic ballot for the House ensures big House gains for Democrats. Hard Right Republicans in districts best described as D+5 to R+2 can lose seats in large numbers. All House seats will be up for election, and there is no getting around that.
ReplyDeleteAs for President Obama losing -- think of all the Governors having trouble during the budgetary processes. In hard times and polarized politics, few people are going to get what they want. Such is the President's difficulty in the polls right now. Mitt Romney can safely remain silent through this mess; President Obama can't. But gues who gets to run against an extremist Congress that achieves little? You guessed it.
The election is more than a year so i think your Obama polls are useless.
ReplyDeleteI'm relatively confident that Obama will win reelection (quite easily) and the Democrats will win back the House.
ReplyDeleteI feel like if the Republicans nominate a true nut like Bachmann it could be a huge wave election. Her crazy would drip all over the rest of the GOP.
Obama has already won reelection and it's not based on silly polling.The 13 keys to the White House has picked the winner for the popular vote in every election. http://ww2.gazette.net/stories/05272011/policol193154_32542.php
ReplyDeleteI'd be interested in hearing the justification for using a 42D/33R/25I party split sample.
ReplyDeleteIn exit-polling for the 2008 election with record Democratic turnout, the split was 39/32/29.
We all know what happened in 2010, and although it was an off-year election, the splits were far lower for Democrats.
So these splits obviously aren't reflective of real-world electoral results.
The latest polling on party registration shows a party split of 29/30/38 or 42/47 if you include leaners.
So by any measure I can find, the split you're using *vastly* oversamples Democrats. Why do your samples skew so far from the "real world?"
"I'd be interested in hearing the justification for using a 42D/33R/25I party split sample.
ReplyDeleteIn exit-polling for the 2008 election with record Democratic turnout, the split was 39/32/29."
First, we don't weight for party, so it just falls where voters say they are at the time, and obviously it's self-identified, not related to registration. Second, +7 and +9 is essentially the same thing, especially when you consider the greater faultiness with exit polling (and different methodology) than with telephone polling; they don't do a true random sample, but rather cluster sampling, trying to get representative precincts and people.
Third, this is not going to be 2010. The electorate will be much closer to 2008 in a presidential year. Midterms are always more Republican, but 2010 was unprecedentedly so in terms of the turnout shift and enthusiasm gap from one election to another. The electorate is really volatile right now.
If anything, Rick Scott secures the Presidential election for Obama. If not, then John Kasich.
ReplyDelete