LEWISTON — With more families replacing their hard-wired telephones with cell phones, the reach of political pollsters is getting shorter, say political scientists and pollsters.
One of every four homes in the United States is now a cell-only household, according to a December 2009 National Health Interview Survey, conducted by the U.S. Centers for Disease Control and Prevention in Atlanta.
Particularly hard to track down are the nation's young voters, ages 18 to 34, because that is the group more likely to have cell phones only, said Patrick Murphy, the president of one of Maine's oldest polling companies, Portland-based Pan Atlantic SMS group.
"You've got to make sure you've got a significant sampling of these people, and increasingly they are becoming harder and harder to find," Murphy said Friday.
To compensate, polls have to be weighted to ensure sample fairness, Murphy said. So far, they feel they've been able to reach enough in that segment of the population to get accurate results, he said.
Firms that conduct only automated calling are even less likely to get an adequate survey of younger voters, prompting concern that those polls may be far less accurate or tilted toward older, and subsequently, more conservative voters, according to a recent report by the Pew Center for the People and the Press.
"Cell-only adults are demographically and politically different from those who live in land-line households; as a result, election polls that rely only on land-line samples may be biased," according to the report.
Maine has 753,000 hard-wired telephone lines and 1.01 million cell-phone lines, according to the most recent information from the Federal Communications Commission. The number of cell-only homes is unknown, according to Andrew Hagler at the Maine Public Utilities Commission.
"How many have cut the cord and wouldn't be reflected in a telephone poll? I can't tell you," Hagler said.
Recent political polls in Maine have been conducted both by people calling potential voters and by automated calls, and the results from those polls have varied.
All have shown Republican candidate Paul LePage leading, but the size of that lead has varied.
For example, a Pan Atlantic poll released on Friday showed about a 7 percent gap between LePage and independent Eliot Cutler. An automated poll released a day earlier by the North Carolina-based Public Policy Polling showed LePage with a 12-point lead over Cutler.
Democrat Libby Mitchell trailed in both polls.
Tom Jensen, director of Public Policy Polling, said reaching cell-only voters is increasingly challenging — and something his firm doesn't do — but a poll showing consistent results across age brackets, as their latest showing a LePage lead, tends to support the validity of overall results.
"LePage is winning every group by a very similar margin," Jensen said Friday.
He acknowledged that the increasing number of cell-only voters presents a growing challenge for political pollsters.
"It's certainly becoming a bigger and bigger issue by the year and telephone polling in general is on the way out," Jensen said. He predicted polling methods would change drastically in the next 10 years.
Another aspect on which Jensen and Murphy seemed to agree was that cell-only people are less likely to be interested in answering polls and may even be hostile to polling calls.
Many view their cell phones as a last bastion of privacy and calls from pollsters are seen as intrusions, Jensen said.
Curt Mildner of Market Decisions, a Portland-based company that doesn't do political polling, said cell-only households tend to be lower-income and younger — typically groups that lean more toward Democrats and independents than Republicans.
"A lot of people are very skeptical about polls," Mildner said. "They don't believe they are true, also it's very hard to predict whether somebody will do something in the future such as vote ... it is much more difficult, they don't know, they change their mind, or they don't show up. Every year you see surprises in elections."
What's surprising to Mildner in this election is it could end with the winner being elected by a minority of the vote.
"I don't think polling is lulling people to sleep," Mildner said.
"None of the candidates is likely to garner majority support," he said of the coming vote in Maine. "A candidate that has minority support is going to be elected – with what could be less than 40 percent of the vote –that would be staggering. That doesn't show the will of the people being met by any one candidate."