UPDATED 9:40 a.m. Aug. 6, 2012

Tulsi Gabbard is on the verge of completing an epic out-of-nowhere comeback to topple Mufi Hannemann and become the Democratic Party nominee for Hawaii’s vacant seat in Congress.

The 31-year-old first-term Honolulu City Council member has opened up a decisive 49 percent to 29 percent lead over the more experienced Hannemann with many early votes already cast and the Aug. 11 primary just days away, according to The Civil Beat Poll.

Six months ago, the Honolulu Star-Advertiser– and Hawaii News Now-sponsored Hawaii Poll showed Hannemann with an early 65-20 lead. Even recent Hawaii Poll results released just last week still said Hannemann had a “healthy margin,” leading Gabbard 43-33.

Civil Beat surveyed 577 likely 2nd Congressional District Democratic primary voters — nearly all of whom had either already voted or said they were definitely going to vote — between Tuesday (July 31) and Thursday (Aug. 2).1 The results have a margin of error of +/- 4.1 percent. Former Office of Hawaiian Affairs chief advocate Esther Kiaaina secured 8 percent and Big Island lawyer Bob Marx 7 percent. That leaves 7 percent who said they wouldn’t vote for any of the candidates or were undecided.

Either Gabbard or Hannemann would be a clear favorite to best the Republican torchbearer in November and replace Democrat Mazie Hirono, who is not seeking re-election in the House and instead running for U.S. Senate. If The Civil Beat Poll results hold for Gabbard, people will soon be calling her “congresswoman.”

What Happened?

Gabbard’s 20-point advantage is a shocker, but she had been making significant strides in fundraising and has used that money — along with some major help from mainland political action committees — to maintain a near-constant presence on the airwaves.

From the beginning of April through July, Gabbard raised $421,000 to Hannemann’s $312,000, according to Federal Election Commission filings. Compounding the advantage was the outside money flowing into the race on Gabbard’s side.

VoteVets.org has spent $270,000 on television advertisements and mailers supporting Gabbard, who was deployed to Iraq as a member of the Hawaii National Guard. The Sierra Club has spent $138,000 on Gabbard’s behalf, as has Women Vote, a campaign of the pro-choice EMILY’s List. In all, that’s more than a half-million dollars pumping up Gabbard’s campaign versus just $18,000 for Hannemann from the University of Hawaii faculty union.

Hannemann’s campaign cited that outside spending as the reason he needed to plug $150,000 of his own money into the race two weeks ago, signaling he knew he might be in trouble. Gabbard injected $30,000 of her personal funds into her campaign on Friday, federal records show.

While money doesn’t necessarily equal votes, The Civil Beat Poll shows the spending has helped Gabbard.

Of those who said they got “a lot” of helpful information from advertising about candidates across all races this year, Gabbard nearly doubled Hannemann’s support, 58 percent to 30 percent. Her advantage was far narrower — just 39 percent to 32 percent — among those who said they got no helpful information from ads.

Hannemann came closest to Gabbard — 42 percent to her 46 percent — among the respondents who said endorsements by political groups and leaders provided a lot of helpful information. Hannemann racked up two dozen union endorsements — including the Hawaii Government Employees Association, United Public Workers and Hawaii State Teachers Association — and got a stamp of approval from both the Maui News and the Star-Advertiser.

But the endorsements weren’t enough.

When Did It Happen?

Gabbard pulled away from Hannemann in the final weeks leading up to The Civil Beat Poll.

Of the 30 percent of respondents who said they made up their mind in the two weeks before answering the phone, Gabbard led Hannemann 52 percent to 26 percent. Of another 30 percent who made up their minds in the past month, the gap was even greater at 63 percent to 21 percent. Only among voters who made up their minds more than a month ago was Hannemann close, at 44 percent to Gabbard’s 45 percent.

That finding lines up with previous Civil Beat Poll results showing a dead heat between Gabbard and Hannemann in two surveys conducted in quick succession more than a month ago.

The recent surge also explains why Gabbard’s edge is marginally smaller (50 percent to 35 percent) among the 27 percent of respondents who said they’d already voted, either by mail-in absentee or early walk-in voting. Her lead among the 71 percent who said they’ll definitely vote was 48 percent to 27 percent.

In addition to capitalizing on Hannemann’s eroding support, Gabbard appears to be picking up the voters abandoning the second-tier candidates as election day approaches.

Kiaaina and Marx each sat at 10 percent when Civil Beat last polled the race in June. Now, they’re at 8 percent and 7 percent respectively, and could drop further — Kiaaina has only 4 percent support among those who have already cast a vote. Given a choice between only the top two candidates, those who still support Kiaaina and Marx said they preferred Gabbard to Hannemann, 40 percent to 17 percent with 43 percent either saying neither or unsure.2

A 30-Point Difference?

Some readers might wonder how the Star-Advertiser could have Hannemann up by 10 points and Civil Beat could have Gabbard up by 20 points. There are a few possible explanations:

Methodology: Civil Beat uses interactive voice response technology, also known as touch-tone polling or, critics say, “robo-polling.” The Star-Advertiser uses live telephone interviews. Though there are downsides to touch-tone polling, Civil Beat believes talking to a live person can be intimidating and could drive some voters to say they support the better-known Hannemann even if they really don’t.

Neighbor Islands: Though the Star-Advertiser did not make its full results available for the CD2 race, its story on the survey says neighbor island voters make up “about 65 percent” of the voters in the 2nd Congressional District. The Civil Beat Poll has the split at 56 percent for the neighbor islands. Because Hannemann did better in both polls on the neighbor islands than he did on Oahu, this difference could be significant. In the 2010 general election, 61 percent of votes cast in the CD2 race came from the neighbor islands. Will the heated Honolulu mayoral fight drive up turnout on Oahu in the 2012 primary?

Republicans and Independents: Again, full results for the Star-Advertiser poll are not publicly available, but a look at its cross-tabulations for the U.S. Senate race survey conducted on the same dates shows that 72 percent of the likely voters included in the Democratic primary sample usually vote Democratic versus 9 percent Republican and 18 percent independent. The split in The Civil Beat Poll of likely CD2 Democratic primary voters is 58 percent Democrats, 6 percent Republicans and 30 percent independents. Because Civil Beat found Gabbard dominated (54 percent to 25 percent) among independents, this difference could be significant. Will Linda Lingle‘s advertising campaign convince voters to pick up a Republican primary ballot, or will the heated races for Senate and Congress on the Democratic side draw more independents?

Survey Timing: Every survey is a snapshot in time. The Star-Advertiser’s poll was conducted from July 12 to July 21, while Civil Beat’s calls were made July 31 to Aug. 2. This is possible because touch-tone polling also allows Civil Beat to survey more people in a shorter amount of time. Considering Civil Beat found Gabbard pulling away in recent weeks, the difference of two or three weeks between the surveys could be significant.

Sample Size: The Star-Advertiser surveyed 343 people over 10 nights for a margin of error of +/- 5.3 percent. Civil Beat surveyed 577 people over three nights for a margin of error of +/- 4.1 percent. This difference alone is significant, but the larger sample made possible by touch-tone polling means we can slice and dice the data in different ways because the subsets used in cross-tabulation analysis have larger sample sizes and smaller margins of error.

Some Historical Context

Hannemann’s campaign will likely dismiss the results of this poll out of hand. (His comments are not included in this article because Civil Beat does not make its results available to the campaigns before making them available to our readers.)

He might raise questions about the sample Civil Beat used. In particular, he might point out that 80 percent of respondents were 50 years old or older or that 51 percent of respondents identified themselves as Caucasian. He also might criticize the poll’s touch-tone methodology.

That’s what happened during the 2010 primary for governor, when The Civil Beat Poll projected Hannemann would lose badly to Neil Abercrombie in the Democratic Primary.

The final Star-Advertiser/Hawaii News Now poll of that race had Abercrombie up 49-44. Civil Beat pegged it at 48-31 among all primary voters and about 55-35 among Democratic primary voters. Hannemann called the Civil Beat survey “suspect.”

Abercrombie won that race, 59 percent to 38 percent.


Hannemann sent an email to supporters early Monday after this article was published with criticisms of the “robo-poll” methodology and the demographics of the sample.

All of the latest polls conducted by local firms, which are familiar with Hawaii’s unique and diverse electorate, show Mufi Hannemann with a healthy lead. The Hawaii News Now/Honolulu Star-Advertiser “Hawaii Poll” showed Mufi Hannemann with a 10 point margin, a fact mirrored by Qmark Research in its latest poll.

Follow Civil Beat on Facebook and Twitter. You can also sign up for Civil Beat’s free daily newsletter.