A new Iowa survey from Strategic Vision, a Republican polling outfit, seems to confirm the recent pro-Obama trend in the state. The Illinois Senator leads the Democratic pack with 32%, followed by chief rivals Hillary Clinton and John Edwards at 25%. But as with all recent polls here, the numbers are within the margin of error.

Should Hillary be worried? The campaign can’t be happy with the trend lines in the horse race — especially since momentum always plays such a key role in Iowa. They also can’t be happy with the fact that she’s dropping among women, and that voters seem to care less about experience than change. But staffers constantly point out that polls don’t matter here. While that’s partially spin, since aides won’t hesitate to point out a poll that looks good for their candidate, there’s a lot of truth to the assertion.

I asked Opinion Dynamics CEO John Gorman, who runs the Fox News / Opinion Dynamics polls, why it’s so difficult to get a read on the Hawkeye state. “Polling in Iowa is very difficult and of very uncertain quality,” he wrote back. “This results from the following factors:

“1) Turnout is generally pretty low; thus extraneous factors can screw things. This is one of those few places where the ’snow on one town, no snow on another’ can actually matter.” (my note: in 2000, the last caucus that was contested in both parties, just 145,000 out of a voting-eligible population of 2.1 million showed up to caucus. That’s 6.6%. By comparison, 44.4% of eligible New Hampshire residents voted in the primary that year).

“2) Low turnout also means organization can matter a lot; in a normal election 5% is really “a lot”; in Iowa 20% is possible.” (my note: it also means the right kind of voter can matter a lot; previous caucus goers are much more likely to actually show up than first timers. This helps John Edwards, who has more of the established, older support in the state, while it’s challenged Obama and Clinton to figure out how to get younger Iowans to the caucuses)

“3) Iowans have come to believe in their own superior wisdom and influence on events; a chance to help a worthy 2nd tier candidate and keep them in the race can produce shifts.

“4) The public nature of the voting discourages oddball voting. People are less likely to vote for a Kucinich in front of their friends and neighbors (employers, customers).

“5) The [15%] threshold on the Dem side means that in the course of just the hour of voting the stronger candidates go up and the weaker ones go down. This means a pre-election poll is actually going to overestimate the bottom tier candidates.” (my note: And since candidates who don’t have the support of 15% of the voters at a caucus site are out of the running there, the second choice of all those Dodd, Biden, Kucinich and Richardson voters really does matter).