But neither Clinton nor Sanders did so well as to make me think either candidate will gain momentum heading into the New Hampshire primary next week. Sanders is likely to win in the Granite State — he has either an 89 percent chance or a 96 percent chance of winning there, depending on whether you look at FiveThirtyEight’s polls-plus forecast or polls-only forecast. Still, the results in Iowa suggest that polls in New Hampshire may tighten. That’s because the states look similar demographically. Even taking into account that Sanders lives next door in Vermont, Clinton probably shouldn’t be behind by 17 percentage points in the New Hampshire polling average right now.

Assuming Sanders holds on to win in New Hampshire, would that be bad news for Clinton? Although I’m not sure that anyone wants to lose a primary, how she frames a loss in New Hampshire will matter a lot. If she is seen as doing better than expected, she could get a bump (something she is now unlikely to get coming out of Iowa). If, however, Sanders gets good press, he may improve his chances despite the demographic challenges facing him in other states.

We’ve said for months that Iowa and New Hampshire are two of the best states for Sanders demographically. You can see why in the entrance poll taken in Iowa. Sanders won very liberal voters over Clinton by 19 percentage points, but he lost self-identified somewhat liberals and moderates to Clinton by 6 percentage points and 23 percentage points, respectively. That’s bad for Sanders because even though 68 percent of Iowa Democratic caucus-goers identified as liberal this year, only 47 percent of Democratic primary voters nationwide did so in 2008. We’ll need to see if Sanders can do better in a state that is more moderate than Iowa before thinking he can win the nomination.