If Past Polls Are Any Indication, N.H. Primary Outcome Still Subject to Change

Jan 20, 2016

Credit Kate Brindley for NHPR

The UNH Survey Center turned plenty of heads this week with its latest poll on the New Hampshire Democratic primary. The poll, conducted for WMUR and CNN, had Bernie Sanders leaping even further ahead of Hillary Clinton, now leading 60 to 33 percent.

The poll did come with a few caveats. 

For one, about a quarter of those surveyed were still undecided. Another quarter were just “leaning” toward one candidate. Still, Sanders enjoyed a lead over Clinton in every age group, across all education levels and among “undeclared” (or independent) voters as well as registered Democrats. Sanders, too, enjoyed a net favorability rating that the survey authors described as “an almost unheard of +84%.”

And while the UNH results do show a wider margin than other recent polls on the New Hampshire Democratic primary, they aren't the only ones who have found Sanders ahead of Clinton by at least several points in recent weeks. A poll conducted by ARG around the same time frame as the UNH poll had Sanders leading by 6 points; another one from Monmouth conducted the week before had Sanders leading by 14 points.

The timing of polls – out of UNH and elsewhere – has varied from election cycle to election cycle, so any side-by-side comparisons from past primaries are bound to be a little imperfect. Still, it’s worth looking back at the predictions made in past cycles — as has been pointed out repeatedly, a lot can still change.

Take the 2008 Democratic primary. About this time in 2008, Real Clear Politics showed the following polling averages in New Hampshire: 31.5 percent for Barack Obama, 28.5 percent for Hillary Clinton and 16.3 percent for John Edwards.

Credit UNH Survey Center

The UNH poll at that same point in the New Hampshire campaign had Obama leading Clinton by about 2 points — which was roughly on par with another LA Times-Bloomberg poll from around the same time. Overall, though, other polls from the same time period were kind of all over the place: USA TODAY and Gallup reported a tie, while an ARG poll had Clinton up by 14 points and Rasmussen had Clinton leading by three. Clinton, in the end, would win the New Hampshire primary with 39.1 percent of the vote, beating Obama by about 2.6 points.

Then there was 2004 — the cycle where a lot did, in fact, shift substantially in the weeks ahead of the New Hampshire primary. While there’s no publicly available UNH polling results from three weeks ahead of that primary, their polls from mid-December (a little over a month in advance of the election) and mid-January (within two weeks of the election) both showed strong leads for Howard Dean over John Kerry.

Credit UNH Survey Center

According to Real Clear Politics, ARG tracking polls were also showing Dean ahead of Kerry  three weeks out from the New Hampshire primary – with the former Vermont governor commanding about 35 percent of voters’ support, and Kerry still trailing in third place behind Gen. Wesley Clark.

The race shifted substantially after the Iowa caucuses – conventional wisdom says that the infamous “Dean scream” turned off would-be supporters, which bolstered Kerry’s chances. In New Hampshire, Kerry ended up beating Dean 39 to 26 percent, with Clark trailing at 13 percent. 

And it's not just the Democratic race where polling averages can resemble a rollercoaster. Ahead of the 2012 Republican primary, a UNH poll in December 2011 (accurately) had Mitt Romney leading the pack — but it also had Newt Gingrich in second place. Gingrich ultimately ended up coming in fourth place, behind Ron Paul and John Huntsman.

This isn’t to say that Sanders isn’t going to end up winning New Hampshire – other polls have put him ahead of Clinton here for several months, though not by the same margins seen in the latest UNH results. But it’s worth keeping in mind that, as history has shown, the polls only tell a portion of the story. Voters get to write the final page.