Alan Abramowitz Does Not Like Rasmussen
But his statistical model does predict that Obama will lose the popular vote
Alan Abramowitz does not like Scott Rasmussen. Who is Alan Abramowitz, you ask? I have no idea either, but according to Google he is a professor of political science at Emory University in Georgia. Abramowitz published a piece at Huffington Post on Saturday that all but called Scott Rasmussen a partisan hack who has no clue what he’s doing.
So what was Abramowitz’s proof? He cited several battleground states in which Rasmussen found Romney performing better than the average of polls in the state, and that is apparently a crime. What Abramowitz did not tell his readers is that in six other battleground states — MI, MN, MO, MT, NM, and PA –Rasmussen showed Romney doing worse than the RealClearPolitics’ average. This is what passes for intellectual rigor in Abramowitz’s world.
But that’s not the best part. The best part is that in 2008 Abramowitz presented a statistical model to predict the two-party popular vote of presidential candidates. His 2008 model, which suggested that Obama would receive 54.3% of the two-party vote in 2008, has only three factors: real GDP growth in the 2nd quarter of the election year, the incumbent president’s net approval rating at the end of June of the election year, and whether a candidate is running for re-election.
Here’s the actual equation that Abramowitz published: Popular vote % for incumbent party candidate = 51.417 +(0.604*GDP) + (0.109*Net Approval) – (4.265*Re-election).
According to Abramowitz’s model, a candidate would lose nearly 4.3% just by running for re-election. Each 10 points of additional net approval would add 1% to the candidate’s total, while each percentage point of GDP growth would add 0.6% to his or her final vote share.
During the 2nd quarter of 2012, real GDP increased only 1.3 percent. Obama’s average net approval rating (total approval minus total disapproval) was only 0.4 percent at the end of June. And the variable for re-election is 1. Put all those together, and you have an Abramowitz popular vote prediction of…48% for Barack Obama.
The most recent national tracking poll from Rasmussen shows Obama at 47%. What a hack.
[Addendum: A Twitter correspondent of mine tells me that Abramowitz "tweaked" his model in July, when it became clear that his 2008 model would predict an Obama loss. The new and improved Abramowitz model predicts an Obama win, naturally.]