It would be entirely reasonable to presume that Bernie Sanders has momentum in Iowa. He’s gained on Hillary Clinton in national polls. He keeps pulling further ahead of Clinton in New Hampshire. And he’s made substantial gains in Iowa relative to his position late last year. December polls of Iowa showed Sanders behind by an average of 16 percentage points; the race is much closer now.
There’s just one problem: Sanders’s momentum may have stalled right when it counts the most.
The Des Moines Register’s Iowa poll released Saturday, for example, had Clinton leading Sanders by 3 percentage points. That means Iowa is close and winnable for Sanders; polling errors of 5 or even 10 percentage points are not uncommon in the caucuses. But it also means that Sanders hasn’t gained on Clinton. The previous Des Moines Register poll, released earlier in January, showed Clinton up by 2 percentage points instead.
The same story holds for other polling companies that have surveyed Iowa twice in January. A couple of these pollsters — American Research Group and Quinnipiac University — show Sanders leading. But they don’t show him gaining; Sanders also led in the previous edition of the ARG and Quinnipiac surveys.
| POLLSTER | MOST RECENT POLL | PREVIOUS POLL | CHANGE |
|---|---|---|---|
| American Research Group | Sanders +3 | Sanders +3 | — |
| Gravis Marketing | Clinton +11 | Clinton +21 | Sanders +10 |
| Marist College | Clinton +3 | Clinton +3 | — |
| Public Policy Polling | Clinton +8 | Clinton +6 | Clinton +2 |
| Quinnipiac University | Sanders +4 | Sanders +5 | Clinton +1 |
| Des Moines Register | Clinton +3 | Clinton +2 | Clinton +1 |
The only polling firm to show Sanders making substantial gains in January is Gravis Marketing. It had him trailing Clinton by an implausible 21 points earlier this month. Now, it has him down by 11. That may reflect reversion to the mean (or herding) after a seeming outlier more than any genuine movement toward Sanders.
Overall, Clinton is ahead by 5.3 percentage points in FiveThirtyEight’s weighted polling average in Iowa. She led Sanders by exactly the same margin, 5.3 percentage points, on Jan. 15.
What’s blunting Sanders’s progress? Actually, it might not have a lot to do with Sanders. Clinton isn’t an easy mark. She remains extremely popular with Democrats, including in Iowa, where her favorable rating was 81 percent in the Des Moines Register poll. (Sanders’s favorable rating was 82 percent in the same poll.)
And Clinton has an impressive ground game in Iowa, where she has field offices throughout the state and voter outreach tactics modeled more on President Obama’s successful 2008 campaign than her own failed one. Clinton’s campaign takes a highly localized approach to Iowa, I learned on a visit to her Davenport field office earlier this month, with volunteers responsible for contacting a parcel of voters in a particular neighborhood again and again until the caucuses.
Sanders has an impressive ground game also. In fact, it’s not always easy to tell the campaigns apart. His campaign and Clinton’s use many of the same tools and strategies, and there’s a lot of cross-pollination into both camps from former members of the Obama 2008 campaign team.
On a visit to Sanders’s Des Moines headquarters this month, I discussed with his staffers an unusual occurrence in the polls: Some Iowa surveys, like the previous version of the Des Moines Register poll, showed the number of undecided voters increasing as the vote approached.
The Sanders staffers had a convincing-sounding explanation. Since most Democrats like both candidates, it wasn’t just a matter of flipping Clinton voters into Sanders voters. The voters had to proceed through a liminal stage first: the state of being undecided.
But it may be that some of those voters, after flirting with the idea of voting for Sanders, wind up sticking with Clinton in the end.
Read more:
Four Roads Out Of Iowa For Republicans