The four polls taken this week in Iowa that are listed by RealClearPolitics show widely different results. NBC/WSJ/Marist shows President Obama up by 6 percentage points — 50 to 44 percent. Gravis Marketing shows Obama up 4 points — 49 to 45 percent. WeAskAmerica shows Obama up 1.5 points — 48.8 to 47.3 percent. Rasmussen Reports shows Mitt Romney up 1 point — 49 to 48 percent. What’s one to make of this?
WeAskAmerica doesn’t list its projected turnout by party, but the other three polls do. NBC/WSJ/Marist projects a 3-point edge in turnout for the Democrats over the Republicans (34 to 31 percent). Gravis projects a 6-point edge in turnout for the Democrats (41 to 35 percent). Rasmussen projects a 4-point edge in turnout the other way — for Republicans (39 to 35 percent). That’s a double-digit difference between Gravis and Rasmussen, with NBC/WSJ/Marist being more than two-thirds of the way toward Gravis.
So, who’s right? Well, no one can say with any certainty, but we can nevertheless take an educated guess. According to exit polling in 2008, when Barack Obama swept to victory over John McCain — by 7 points nationally and 10 points in Iowa— Democrats enjoyed a 1-point turnout edge over Republicans in the Hawkeye State (34 to 33 percent). Nationally, exit polling showed Democrats enjoying a 7-point turnout edge (39 to 32 percent). So, in terms of voters’ party affiliation, 2008 exit polling indicates that Iowa ran 6 points to the right of the national average.
Obama did better in Iowa than nationally in 2008 because, at least according to exit polling, he won a higher percentage of the Democratic vote (+87 points in Iowa versus +80 points nationally), and a higher percentage of the independent vote (+15 points in Iowa versus +8 points nationally).
It’s hard to imagine anyone thinking that Iowa voters’ party affiliation this time around won’t be at least somewhat to the right of where it was in 2008 — in other words, somewhere to the right of a 1-point Democratic edge. How far to the right of that is harder to guess. But if we assume that party affiliation is dead-even nationally — which is slightly to the left of what Gallup indicates (Gallup shows Republicans with a 1-point edge) — and if we assume that Iowa’s party-affiliation split will again be 6 points to the right of the nation as a whole, that would mean a GOP turnout advantage in Iowa of 6 points this time around. If, however, Iowa is only half as far to the right of the nation in 2012 as it was in 2008 (so, 3 points), or if it remains about 6 points to the right but Democrats actually have, say, a 3-point edge nationwide in party affiliation (meaning Gallup is underselling Democrats by 4 points), then the GOP edge in party affiliation in Iowa would be about 3 points.
Either way — whether Republicans will have a 3-point or 6-point edge in turnout in Iowa (or even anywhere between 2 and 7 points) — that’s pretty close to Rasmussen’s projection of a 4-point GOP edge in turnout, and it’s a long way removed from the projections from NBC/WSJ/Marist (Democrats by 3) or Gravis (Democrats by 6).
Interestingly, Rasmussen actually shows Obama doing much better among independents in Iowa than Gravis does. Among Iowa independents, Rasmussen shows Obama winning by 12 points (52 to 40 percent), while Gravis shows Romney winning by 5 points (43 to 38 percent). (NBC/WSJ/Marist doesn’t say.) If not for this disparity, which offsets some of the 10-point difference in projected party turnout between these two polls, the 5-point gap between their overall results in the Hawkeye State would be even greater.