« BACK  |  PRINT

RS

FRONT PAGE CONTRIBUTOR

Rasmussen and Gallup generic ballots diverge

By: Neil Stevens (Diary)  |  July 20th, 2010 at 11:00 PM  | 

RESIZE: AAA

House

Until now, Gallup and Rasmussen Reports have generally pointed in the same direction with their generic ballot polls. If they’ve differed, it’s been in the magnitude.

This week, that has changed. How big a difference is it, and what does the Swingometer say about it all? Let’s find out.

We’ve got two polls: Gallup’s which shows a huge swing for the Democrats giving the incumbents a 49-43 lead, mirroring the Republicans’ big lead in May.

We also have Rasmussen’s which continues to shows the Republicans ahead, giving the challengers a 45-36 lead, the biggest in a month.

If we boil these numbers down to the usual two-party splits, here’s what we see:

Category D R
2008 two party split 56 44
Rasmussen two party split July 18 44 56
Gallup two party split July 19 53 47

Enter the Swingometer: Rasmussen takes us from D+12 to R+12 for a 24 point swing to the Republicans, a 64 seat Republican gain, and a 242 R-193 D House majority for the GOP.

Gallup, meanwhile, takes us from D+12 to D+6 for a 6 point swing to the Republicans, an 18 seat Republican gain, and a 239 D-196 R House majority for the Democrats.

So the predicted results are nearly mirror images of each other. Fascinating.

Crossposted from Unlikely Voter

Tags: , , , , ,

COMMENTS

POSTING RULES  |  LOGIN TO COMMENT  |  REGISTER TODAY
  • ilgop24

    Gallup just seems a bit off everything else right now…

  • barleycorn

    I find it beyond curious that Gallup shows an 8 point swing toward the Dems (in the absence of any obvious game changers) at the same time that Ras shows a 3 point swing to the GOP.

    Gallup suggests this swing is perhaps due to the Wall Street Reform bill. I don’t buy it.

    People just don’t forget how ticked off they have been for the past 10 months simply because of a piece of legislation that doesn’t directly impact the average person.

    A 15 point difference between the two polls means somebody is not just wrong but absurdly so.

  • IJB

    Since mid-May, they’ve had, in order:

    R+1, D+1, R+6, D+0, R+5, D+0, D+1, R+2, D+1, D+6

    There’s absolutely no rhyme or consistency to these values, AFAIAC – it seems to rise above just statistical noise, IMO.

    And Gallup is the *only* organization showing this supposed “huge” shift to the D’s this week.

    In fact, I’m willing to bet that, by next Monday, Gallup will be back to D+0 or R+1 or something.

    Compare them some time to Rasmussen – now Rasmussen is polling Likely Voters, vs. Gallup’s Registered Voters, but you see a lot less “jumping around” in Rasmussen numbers. In short, I put a lot more stock in them.

  • Adjoran

    Their own recent track record is to show the generic ballot balanced, and this is more than a standard deviation towards the Democrats. Sorry, “financial reform” just doesn’t earn that sort of support swing.

    With no other apparent explanation, the ready conclusion is that this sample was an outlier even for Gallup – it doesn’t even match their other poll questions on Obama approval, etc., which didn’t show such movement left.

    It should also be noted that Gallup’s generic ballot poll is regularly off, and always to the Democrats. They even had the Democrats ahead going into the 1994 wave. Even with other polling firms, when registered voters aren’t screened down to likely voters, the results always seem to favor Democrats. My hypothesis is that Democratic-leaning sub-groups are under great pressure to register whether they care about politics or not, and many or most do. They just don’t vote – but they will answer a poll because it makes them feel cool to be asked.

  • The_Gadfly

    With no other apparent explanation, the most charitable answer is that this sample was an outlier even for Gallup

  • Tbone

    Either that or they bought their call list from the ringtone downloads of Hip Hop Times.

  • Oz

    I have to admit that the first thing I look for every day when I get on redstate is your latest run through the Swingometer.

    Will be looking forward to more and more of this in the run up to November.

  • http://aposematic.wordpress.com aposematic

    Maybe, just maybe, that Media Matters guy should check on Gallop’s polling. Gallop being the only polling group out of many polls that not only puts the Dems ahead in generic ballot, but ahead by a lot…hummmm!

  • moosebytes

    For this survey, Gallop sampled adults. All of the other generic ballot polling they’ve done sampled registered voters.
    Why the change?

    See RCP:
    http://www.realclearpolitics.com/epolls/other/generic_congressional_vote-901.html#polls