Despite the playing field pretty much agreed upon, when you get down to the last handful or so of states, the best polling firms from each party are coming up with widely disparate results, with very different turnout assumptions as the best explanation for the disparity.
With President Obama pulling 90 to 95 percent of the African-American vote, 70 percent of Latinos, and 60 percent of the 18- to 29-year-old vote, the turnout assumptions for each group make huge differences in a very close race. To the extent that any one, two or all three groups turn out in proportions similar to 2008, Obama would do significantly better than if each of the three votes in proportions more like 2000 or 2004 (adjusting for population changes in the three groups). It’s somewhat more complicated than that. For example, Obama is actually pulling larger shares of the Latino vote against Romney than against McCain in 2008, but their likelihood of voting appears less than before. Then again, there are 4 million more registered Latino voters than there were four years ago.
Obama campaign strategists argue that too much is made of the 2008 “Obama surge” in minority and younger voters, pointing to data showing that the increases were more a function of changing demographic and population trends in the country than of a one-time surge. They suggest that it’s improbable that there will be fewer Latino votes cast for Obama in 2012 than in 2008. Again, there are 4 million more registered Latino voters now.
Read more of the Charlie Cook article here: Chance of Split Electoral-Popular Vote Very Real – NationalJournal.com.
- Why Obama Is Toast (realclearpolitics.com)
- The Folly of David Axelrod’s Turnout Model (ironmill.com)
- A popular-vote/Electoral College split? (hotair.com)