Jeff Lax sends along this article:

Are the polls obscuring the reality that Barack Obama is beating Hillary Clinton in the race for the Democratic nomination for president? Drew Cline, the editorial page editor of New Hampshire’s Union Leader thinks so.

Based on money-raising and visible support on the streets of New Hampshire, “the evidence shows that Obama has broader support than is being picked up by the polls,” Cline writes at his Union Leader blog. “So the polls must be wrong.”
. . .
“Think of it like a House, M.D. episode. When you have a test result you know is accurate (in this case, the fund-raising numbers) that contrasts with a symptom or test result you can’t explain (the poll numbers), you go with what you know is right and keep testing the other one until they match.” . . .

Much as I hate to contradict anyone named “Drew,” I have to admit that a natural explanation for the discrepancy is that the visible support he’s seen on the “streets of New Hampshire” does not represent a random sample of primary voters. Of course, as I never tire of saying, a poll is a snapshot, not a forecast, and things can definitely change.