There was a similar article in 2000 that showed this (or a similar) academic model with Gore defeating W. Then again, Gore did win the pop vote.
I posted a prediction a month ago on another arguing with some libs, based on my reading of the polls and internals and playing with my spreadsheets, I cam within .1% of the CU guys:
Here’s a hint, in the 2010 mid terms the turnout was 35-35-30, dem-pub-indy the NBC poll was weighted 46-35-19, dem-pub-indy, you’d have to be high on crack to believe those numbers. Oh yeah the poll
What they say is that Obama is not doing well among independents.
So they had to over sample dems by +11 relative to their turnout relative to 2010 and under sample indies by -11 to get Obama the incumbent to 49% and Romney to 43%.
SO in this poll the independents went for Romney 58-42, assuming each candidate got 90% of their base.
If we apply the 2010 turn out model (reasonable/conservative estimate) since party id is currently running +3 pub) as opposed to the crack turnout model with dems being 46% of the electorate and independents being 19% of voters. The result is 49-45% Romney The other 9% say 3% go to 3rd parties (based on history, is ~50% high), that leaves 6% undecided, normally 80% go to the challenger ler’s reduce that to only 66%. In that case Romney wins 53-47% No wonder the Obama campaign is spending more than it is taking in while Romney can’t spend anything prior to the convention.
I am very happy with this prediction as thing sunfold.
From the article:
Their model correctly predicted all elections since 1980, including two years when independent candidates ran strongly, 1980 and 1992. It also correctly predicted the outcome in 2000, when Al Gore received the most popular vote but George W. Bush won the election.