That gives Obama an 89,000-vote advantage in the early vote; as noted above, the Romney campaign thinks they have about 87,000 more “high-propensity voters” than the Democrats do. That looks like a really close race . . . until you get to the independents who haven’t voted early, where Romney leads by 9 in Marist (let’s say 54–45).

We don’t know how many Iowa independents will vote on Election Day, but we know 1.5 million people voted in Iowa in 2008, and 33 percent were independent, according to the exit polls, so we’re looking at roughly 500,000 independent/no party/third party voters in the state. We also know that 26.1 percent of the 675,402 early voters in 2008 were no party or other party — 176,280. In other words, in 2008, about 323,000 independents voted on Election Day instead of voting early.

If Romney has a lead of 9 points among independents, he wins. The only question is by how many votes. If independent turnout on Election Day is 50 percent of 2008, Romney wins by 14,000 votes. If it’s 70 percent of 2008, he wins by 20,000 votes. If it’s 90 percent, he wins by 26,000 votes.