Did polls tilt the governor's race?

LEWISTON — The morning before Election Day, EqualityMaine, a group promoting gay rights, sent an e-mail to supporters urging them to cast a strategic vote for governor.

Unofficial results of the governor's race:

Paul LePage, R, 38%

Eliot Cutler, I, 37%

Libby Mitchell, D, 19%

Shawn Moody, I, 5%

Kevin Scott, I, 1%

Polls released the week before the Nov. 2 election:

Critical Insights/Maine Today Media

Governor's race:

LePage, 40%

Cutler, 21%

Mitchell, 21%

Moody, 4%

Scott, 1%

Pan Atlantic SMS

LePage, 37%

Cutler, 31%

Mitchell, 22%

Moody, 3%

Scott, 1%


LePage, 40%

Cutler, 26%

Mitchell, 26%

Other, 4%

* - Independents Kevin Scott and Shawn Moody were excluded from this poll.

Public Policy Polling

LePage, 40%

Cutler, 28%

Mitchell, 24%

Moody, 4%

Scott, 1%

Unofficial results of the congressional races:

Congressional District 1

U.S. Rep. Chellie Pingree, D, 57%

Dean Scontras, R, 43%

Congressional District 2

U.S. Rep. Mike Michaud, D, 55%

Jason Levesque, R, 45%

Poll results from a Critical Insights/Maine Today Media poll, the only survey released last week on the two congressional races:

Congressional District 1

Pingree, 41%

Scontras, 45%

Congressional District 2

Michaud, 44%

Levesque, 40%

Yes on 1 at Daddy O's
Russ Dillingham/Sun Journal

Dan and Mary Wilson of South Paris are glad the casino bill passed and hope their business will pick up as a result. They were having lunch at Daddy O's in Oxford on Wednesday afternoon when they heard the news break on the close vote.

The e-mail included a link to the latest polls, which recipients were urged to consider before voting for independent Eliot Cutler or Democrat Libby Mitchell, the two pro-gay marriage candidates on the ballot who were trailing Republican Paul LePage.

EqualityMaine characterized LePage as an "ultra-conservative, tea party zealot" who had to be defeated, lest he stymie same-sex marriage efforts and repeal gay rights protections. 

"Tomorrow, please vote for the pro-marriage candidate who you believe has the best chance of beating LePage," the e-mail said.

Some political analysts believe the message highlights the strategic voting that took place Tuesday. It also shows the role polling — some of it inaccurate — may have played in the election.

"In a lot of ways, they have more of an opportunity to shape things," said Jim Melcher, a political science professor at the University of Maine at Farmington.

But some say polls should be a snapshot in time, not a determinant for voting.

"Polls are supposed to identify how people are thinking and why," said MaryEllen FitzGerald, a pollster for Critical Insights in Portland. "They're not supposed to influence voter behavior."

But e-mails like the one from EqualityMaine suggest polls had the opposite effect this year, prompting questions about accuracy, transparency and timing.

Mike Tipping, a political scientist and pollster for the Maine People's Resource Center and the Maine People's Alliance, said polls shouldn't be released toward the end of an election "for the good of democracy."

Tipping, who blogs for Down East magazine and who jointly commissioned a poll with Down East the week before the election, said polls are playing a larger role in elections and policy discussions.

"Unfortunately, a lot of the political debate in the country has been brought down to the level of just a couple numbers arranged against each other," he said. "So that's why I think it's important to have (transparency), to know where those numbers are coming from."

Melcher believes polls played a significant roll determining the gubernatorial race. He specifically referred to a Rasmussen poll released Oct. 12 that he believes changed the dynamic of the contest.

For months, Democrats and Republicans said the race was between Mitchell and LePage. Previous surveys appeared to back that claim. Cutler, an independent, consistently polled between 10 and 15 percent, well outside the viability threshold that political scientists say gives third-party candidates a chance to win.

But on Oct. 12, Rasmussen, considered to be a right-leaning polling firm, released a poll showing Cutler with 20 percent support. Suddenly, Cutler was viable.

The survey set off a chain of events that Melcher believes influenced the race.

Days later, political action committee financial filings with the state ethics commission by the Republican Governors Association and the Maine Democratic Party showed nearly $500,000 in ads opposing Cutler. The attack ads released by state Democrats elicited an outpouring of controversy, which observers such as Melcher believe contributed to Cutler's surge and Mitchell's plummeting support.

"I think (the Maine Democratic Party) probably panicked a bit," Melcher said. "They felt like they needed to do something about Cutler. But it was more the way they did it, going out of bounds, playing on (Chinese) stereotypes. It wasn't totally a mistake to be critical of Cutler at that point, but it was the way they did it."

"Cutler's rise in the polls made all the stuff possible," Melcher said.

Last week, Rasmussen released another survey showing Mitchell and Cutler tied, while LePage sat with a commanding 14-point lead. Tipping, blogging for Down East, noted Rasmussen's reputation for favoring Republican candidates. He didn't accuse the firm of tilting the survey toward GOP interests, but he noted that Scott Rasmussen, the firm's president, was scheduled to speak at a Nov. 10 event sponsored by the Maine Heritage Policy Center, a conservative think tank that state Democrats have accused of tilting research to support Republican candidates and causes.

Tipping, it should be noted, works for the Maine People's Alliance, whose PAC donated money to Mitchell's Blaine House bid.

Rasmussen isn't the only polling firm to come under scrutiny. FitzGerald, of Critical Insights, has fielded questions about her company's polling work for the group that successfully won a ballot measure to build a casino in Oxford County. 

FitzGerald said that casino work had been disclosed with its survey results.

But Critical Insights has also received criticism for the accuracy of the polls it commissioned with Maine Today Media. Its most recent poll, released the Friday night before the election, included three notable results.

Unlike three other polls released that week showing Cutler surging past Mitchell, Critical Insights showed Mitchell and Cutler running neck and neck behind LePage.

It also showed GOP challengers in the state's two congressional races either in contention or winning. In the 1st Congressional District, Critical Insights showed Republican Dean Scontras leading U.S. Rep. Chellie Pingree and GOP hopeful Jason Levesque within striking distance of U.S. Rep. Mike Michaud. 

On election night, Pingree trounced Scontras, 57 percent to 43 percent. The Associated Press called the race 90 minutes after the polls closed. Michaud defeated Levesque, 55-45. The AP and CNN called the race at 10:30 p.m.

FitzGerald said the GOP challengers were well within the poll's margin of error and that results could have been skewed by a high number of undecided respondents and a small sample size.

She said the poll also didn't capture the full "wave" of Cutler's support, citing Cutler's late endorsement by former Gov. Angus King and strategic voters who were waiting to see if the independent or Mitchell was their best bet to beat LePage.

"They were flocking to Cutler because he was appealing to the not-LePage demographic," FitzGerald said.

But Patrick Murphy, a pollster for Pan Atlantic SMS Group, another Portland polling firm, wondered whether the Critical Insights results influenced the election. Murphy, who also commissioned a poll last week showing Cutler within six points of LePage, said strategic voters like EqualityMaine supporters might have been dissuaded from voting for Cutler because the Critical Insights poll showed that the independent and Mitchell were running even.

"On the other hand, if they saw the last poll that we did, some of those Democratic voters could have given Cutler another five points and we'd be talking about him as the elected governor," Murphy said.

Murphy's wife, Victoria, was the former chairwoman of the Maine Democratic Party. Patrick Murphy said his surveys were independently commissioned.

FitzGerald acknowledged that some polls "more captured the Cutler wave than others." In the future, she said pollsters should consider when they release their surveys.

Murphy said voters and the media should pay careful attention to pollsters by examining their track records and credentials.

"They should learn to question the polls," he said. "Ask the question: Does the pollster have a vested interested?"

Tipping agreed. Although polls don't play an exclusive role in helping voters decide, he said they played a significant factor in the gubernatorial election.

He referred to the EqualityMaine e-mail to supporters.

"I mean, there was a direct message from some people that said, 'Look at the polls and vote strategically,'" Tipping said.

"People need to know that polls are both art and science, and that transparency in public opinion (research) is an important aspect of our democracy," he added.


What do you think of this story?

Login to post comments

In order to make comments, you must create a subscription.

In order to comment on SunJournal.com, you must hold a valid subscription allowing access to this website. You must use your real name and include the town in which you live in your SunJournal.com profile. To subscribe or link your existing subscription click here.

Login or create an account here.

Our policy prohibits comments that are:

  • Defamatory, abusive, obscene, racist, or otherwise hateful
  • Excessively foul and/or vulgar
  • Inappropriately sexual
  • Baseless personal attacks or otherwise threatening
  • Contain illegal material, or material that infringes on the rights of others
  • Commercial postings attempting to sell a product/item
If you violate this policy, your comment will be removed and your account may be banned from posting comments.



 's picture

Be very careful about polls

The sad thing is that in this election season, all these pollsters were given carte blance by the statewide media. The vast majority of the polls were automated phone style, performed during a 3 hour period on one night, using only fixed home phone lines. This style of poll has been shown repeatedly to not represent actual trends. In other words they are not worth the cost of the ink to print. Unfortunately our statewide journalists and editorial boards chose to publish them, just because they were available, rather than question their value and make editorial decisions based on that. Hopefully in the next campaign season you folks think harder. You cheapen yourselves by reporting cheaply.


Stay informed — Get the news delivered for free in your inbox.

I'm interested in ...