The biggest question about Iowa, with only a few days to go before the caucuses: Will either Newt Gingrich’s support or Michele Bachmann’s support melt away, and if so, where will it go?

First, Newt. He’s still probably running just a bit ahead of Bachmann, Rick Perry and Rick Santorum in the race for third place. But his favorable/unfavorable ratings are now awful; there’s no sign of any campaign organization in place in Iowa; and every poll brings a new headline about how he’s losing ground.

Bachmann? A super PAC that planned to support her has defected to Mitt Romney. Her Iowa state chair last night defected to Ron Paul. Even worse, in typical Bachmann faction, she immediately claimed that he was bribed to do so, a claim presumably so outrageous that her own Iowa political director is publicly siding with the turncoat and against his own candidate.

In both cases, it wouldn’t be surprising at all for a complete meltdown in support. There’s reason to believe that Gingrich’s support — pegged at about 15 percent in polling averages — is very soft; there’s reason to believe that Bachmann’s support, which is around 10 percent, might shift away if her supporters are convinced that she’s going to lose but another similar candidate is more viable.

Now, that might now happen. It’s not clear how many Iowa voters will really be aware of these late developments. They are relatively attentive but still (mostly) not die-hard political junkies. It’s also not clear how much Gingrich’s lack of organization will matter in the end. To me, it’s easy to imagine a voter walking into a caucus, seeing evidence of four or five of the other candidates but nothing for Newt and flipping away from the disgraced former House speaker … but that doesn’t mean it will really happen. Similarly, it’s easy to imagine an anti-Romney Bachmann supporter realizing she’s toast and switching to Santorum or Perry, but it’s also easy to imagine a caucus voter who is only vaguely aware of the vacillating fortunes of the six candidates active in Iowa. Not to mention the normal dynamics in which voters who decide in favor of one candidate, even after being on the fence between two apparently similar choices, wind up seeing the world through the spin of the one they picked. So there’s no guarantee that any of this will happen.

But if it does? That’s up to about 20 percent of the total vote that might be redistributed to the other four candidates by the time votes are cast. Would it go to “surging” Santorum? To high-spending Perry? To well-organized Paul, who certainly has the most enthusiastic supporters? To front-running Romney? I don’t think we really know, and I’m not even sure the polling data can help us all that much; the whole point here is that people will change their minds from the time they talked to pollsters early this week, so what they told them then might be moot. Not to mention that other dynamic effects are possible … for example, if Santorum does really get to 20 percent in the final polls, could Perry’s 10 percent wind up added to the pot on caucus evening? What I do know, from basic math, is that adding 20 percent to anyone’s total is enough to boost them to a first-place finish. Not that I’m predicting that the defections will be that high, or that they’ll all go to the same place. I have no idea! But neither, I think, does anyone else.