Sign In | Create an Account | Welcome, . My Account | Logout | Subscribe | Submit News | Facebook | Twitter | All Access E-Edition | Home RSS

Presidential polls a fair prediction?

October 6, 2012
By Evan Bevins ( , The Marietta Times

Recent polls showed President Barack Obama leading Republican challenger Mitt Romney nationally and building a lead in swing states, including Ohio.

The impact on voters' opinions of what many viewed as Romney's victory in the first presidential debate Wednesday and the dropping of the national unemployment rate below 8 percent have not been reflected yet, but when they are, it won't make much difference to Devola resident Christina Frasher.

"Absolutely not. Every vote counts," she said.

Frasher, 29, said she would vote for a candidate she supports whether that individual was leading handily in the polls or trailing by a significant margin. But she admits she does pay attention to the polls, if for no other reason than that they seem to be everywhere.

Four different polls of Ohio voters in the last week of September showed anywhere from 49 percent to 53 percent favoring Obama. But there's no telling what those numbers will look like the closer it gets to Nov. 6.

"It's a long time 'til Election Day in terms of the political cycle," said McKinzie Craig, assistant professor of political science at Marietta College. "We still have a lot of movement that can happen, particularly with the debates coming up."

Fact Box

Most recent Ohio polls


Barack Obama - 51 percent.

Mitt Romney - 42 percent.

Undecided - 5 percent.

Other - 3 percent.

Columbus Dispatch (Sept. 30)

Barack Obama - 49 percent.

Mitt Romney - 45 percent.

Undecided - 7 percent.

Public Policy Polling (Sept. 30)

Barack Obama - 52 percent.

Mitt Romney - 44 percent.

No opinion - 8 percent.

Washington Post (Sept. 27)

Barack Obama - 53 percent.

Mitt Romney - 43 percent.

DK/NA - 3 percent.

Someone else - 1 percent.

Quinnipiac University/CBS News/New York Times (Sept. 26)


Sherrod Brown - 49 percent.

Josh Mandel - 41 percent.

Undecided - 11 percent.

Public Policy Polling (Oct. 2)

Sherrod Brown - 49 percent.

Josh Mandel - 39 percent.

Undecided - 9 percent.

Scott A. Rupert - 3 percent.

Columbus Dispatch (Oct. 1)

Sherrod Brown - 51 percent.

Josh Mandel - 39 percent.

No opinion - 10 percent.

Washington Post (Sept. 27)

Source: Times research

That doesn't mean current polls have no impact. Candidates being in the lead can energize their supporters, Craig said. So can trailing.

"For the one that's behind, it can be a way to motivate your base and say, 'You need to be out knocking on doors every day,'" Craig said.

But a poll wouldn't determine whether or how Cutler resident Marilyn Crace votes.

"My vote counts, regardless which way it's going," she said.

Still, Crace said she pays attention to polls, even if she doesn't always believe them.

"I get excited, sort of, when I see mine (candidate) have higher ratings, but then at the end, according to Fox News and the people on it, it's calculated in a way that can be misleading," she said.

Craig said that when looking at polls, people should consider the methodology as well as the margin of error in the poll. For example if a poll shows a candidate leading 51 to 49 percent, and the margin of error is 3 percent, it's possible the other candidate could actually be out in front 52 to 48 percent or trailing by an even larger margin. Craig said the smaller the range, the better.

"The actual numerical estimate isn't nearly as important as the range," she said. "A slim lead, often that's not statistically a lead at all."

Other factors including whether a poll is conducted using cell phones as well as landlines, whether the sample consists of likely voters or simply registered voters, how many of the randomly sampled respondents identify with a particular party and even in-house bias by a polling firm can influence the outcome, according to So different polls can show different results even for the same race in a similar period of time.

Recent polls by the Columbus Dispatch, the Washington Post and the Public Policy Polling, a Democratic-leaning organization, show Ohio Treasurer Josh Mandel trailing incumbent Sen. Sherrod Brown by anywhere from 8 to 12 percent. But while in Marietta Thursday, Mandel pointed to an Oct. 1 poll by Public Opinion Strategy, a national Republican-leaning political and public affairs research firm, that shows him trailing his Democratic opponent by just 47 to 44 percent among all likely voters and leading 49 to 44 percent among those most likely to vote.

"The reality is the only poll that matters is the actual election," said Mandel, a Republican.

Justin Barasky, communications director for Brown's campaign, said he doesn't put a lot of stock in polls either.

"Certainly it's nicer to see us winning than it would be to see us losing," he said. "We know that they've been moving up or down."

Polls are probably more useful to campaigns than individual voters, Craig said.

Former Congressman Charlie Wilson, a Democrat running to recapture the seat for Ohio's 6th District, said a good showing in polls can draw financial support to a candidate and get volunteers to pound the pavement and man the phone banks.

"The polling is important to who supports your team," he said.

If a poll paints a less optimistic picture, that can also affect support, Wilson said.

"This is terrible to say, but a lot of people like to go with the winner," he said.

There haven't been many external polls focusing on the race between Wilson and the Republican incumbent, Congressman Bill Johnson, but Wilson's campaign released its own internal poll in mid-September that showed the race tied with 46 percent of those surveyed supporting Wilson and the same number backing Johnson, with 8 percent undecided.

"We wanted people to know ... it's really close," Wilson said.

The margin of error on Wilson's poll is 4.4 percent.

Johnson said he takes any polls "with a grain of salt unless I know the pollster and the method used." He said many polls are skewed.

"Polls don't win elections; votes win elections," he said.

Johnson said polls prior to Election Day in 1980 would have led many to believe Jimmy Carter was getting a second term as president, but he was defeated by Ronald Reagan.

Craig said polls are more likely to affect turnout as Election Day approaches, but not necessarily voters' decision-making.

Marietta resident Jeff Johnson, 37, said he wouldn't be surprised if some people who don't follow the election very closely back a candidate who's leading in the polls, but it wouldn't influence him. Likewise, he wouldn't avoid casting a ballot because a particular candidate is leading or trailing.

"I vote regardless," he said.

Sharon Bopp contributed.



I am looking for:
News, Blogs & Events Web