Gallup’s Terrible Track Record

Yes, the most recent Gallup tracker has Romney up by 5% points (despite the fact that they have the incumbent President at 51% approval) but is this really something that Romney supporters should latch onto while ignoring the other polling evidence that suggests that Obama is clearly in the lead when it comes to the race to 270?

I used web.archive.org to find out what Gallup has said about the past four presidential races at about this time:

1996: On November 2, 1996 they had President Clinton ahead of Bob Dole by 16% points. A few days later Clinton won by 8.5% points.

2000: I already posted about how way off the mark Gallup was in 2000. Less than two weeks before the election they had Bush winning by 13% points, but ended up losing the popular vote by 0.5% points.

2004: On this day in 2004 (October 31, 2004) I found Gallup had Bush ahead by a 5% margin. Bush won the popular vote by 2.4% points.

2008: On this date in ’08 they had Obama winning by 3% in the traditional likely voter model (to be fair they did have Obama up by the eventual 7% margin in the “expanded” model.)

Conclusion

This is a truly terrible track record. You could have done better by guessing. To take Gallup’s national tracker as the gospel while ignoring 538/Nate Silver & PEC/Sam Wang is absurd.

AddThis Social Bookmark Button


2 Comment(s)

  1. misses:

    7.5
    13.5
    2.6
    4

    ThePresidentialCandidates.US | Oct 31, 2012 | Reply

  2. now we can add another big miss to Gallup’s record … even after they moved their # back in Obama’s direction (how convenient?) their final number still missed by at least 3 points (probably over 4 points once the election result is finalized.)

    ThePresidentialCandidates.US | Nov 9, 2012 | Reply

Post a Comment



Plugin from the creators of Brindes :: More at Plulz Wordpress Plugins