Arizona Reporter - Arizona News - 14/10
Hello? Hello? Are You in a Good Place to Answer a Few Questions?
Pollsters pushed to call cell phones, to avoid under counting younger voters
Contributed by NYU Livewire.
By Andrew S. Garib - Before the last presidential election, Newsday columnist Jimmy Breslin chastised the polling industry for failing to call potential voters who owned only cell phones. The omission, in Breslin’s view, probably shaved support numbers for Democratic presidential candidate John Kerry. "This makes the poll nothing more than a fake and a fraud, a shill and a sham," Breslin argued in 2004 column.
Pollsters retort that they managed to correct for that, since only 4% to 8% of Americans were cell phone-only users in 2004. But they acknowledge that land-line polling will eventually go the way of the telegraph and the rotary phone.
"It will be tough to leave out 40 percent of the population one day," conceded Gallop Polls spokesman Eric Nielsen. "It’s a train that’s coming, and there’s no way around it."
Pollsters retort that they managed to correct for that, since only 4% to 8% of Americans were cell phone-only users in 2004. But they acknowledge that land-line polling will eventually go the way of the telegraph and the rotary phone.
"It will be tough to leave out 40 percent of the population one day," conceded Gallop Polls spokesman Eric Nielsen. "It’s a train that’s coming, and there’s no way around it."
Twitter.com/ArizonaReporter
"I can imagine that my children’s children would wonder why anyone would talk on a phone that’s plugged into a wire," said Pollster.com editor Mark Blumenthal, who is also a polling analyst for the National Journal. "The wired telephone’s days are numbered. It’s just not clear how numbered."
This November, young voters could have the biggest impact on an election in decades. Cell-only voters now comprise at least 15 percent of the population, and disproportionately support Democrat Barack Obama, according to Pew Research Center survey research director Scott Keeter. That skews national polls by as much as 2 percent - within the standard margin of error, but "worth reporting," he noted.
"For the first time, we’re actually seeing a difference between cell-only voters and land line voters when you take into account age, - Keeter said. Three Pew polls in the summer of 2008 showed a much stronger Obama lean among cell-only electors under 30 (59 percent, vs. 32 percent for Republican candidate John McCain) than in the land line group (53 percent Obama, 38 percent McCain).
It’s a sensitive issue for pollsters, who stake their reputations on getting the numbers right.
"On November 4, we have our Super Bowl," Nielsen said. "One or two points is a big issue. You get that wrong, and you lose all credibility."
Pew and Gallup have started folding cell phone data into their results. Gallup is using cell phone polling to track the habits of 1,000 Americans, as part of an ambitious 100-year project to provide a comprehensive look into the lives of Americans. "If you’re going to do a mammoth undertaking, you’re going to do it right," Nielsen said.
At least four other polling efforts, including Time/SRDI and ABC polls, are calling cell phones in presidential polls.
But most pollsters say a 2 percent bias isn’t enough to justify the extra cost of cell phone polling.
"Every survey is a compromise," Blumenthal said. "You interview a sample of people because it’s too expensive to interview everyone."
For pollster’s, switching can be a headache. Cell users may live far away from where they signed up for service, skewing valuable geographic data. Federal law forbids pollsters from calling mobiles with automatic dialers - so pollsters must pay workers to dial numbers and do live interviews. Pew found that more than 40% of those who answered the phone were too young to vote. And - though it hasn’t happened - some pollsters wonder about liability claims: what if a distracted respondent on the go were injured while answering a survey?
It’s too early to tell if cell phones will have any impact on this year’s presidential contest, said Scott Rasmussen, president of Rasmussen Reports, which doesn’t do cell phone polls.
"Polling is easy in theory but difficult in practice," Rasmussen said. "Cell phones are a factor, but there are other factors as well."
For example, he considers it hard to predict whether young people will turn out in higher numbers this year, since models predicting turnout for older adults don’t work the same way for younger voters.
"It’s too close to call this year," he said.
If pollsters who skip cell phones are wrong, though, Breslin’s critique just might come back to haunt them.
This November, young voters could have the biggest impact on an election in decades. Cell-only voters now comprise at least 15 percent of the population, and disproportionately support Democrat Barack Obama, according to Pew Research Center survey research director Scott Keeter. That skews national polls by as much as 2 percent - within the standard margin of error, but "worth reporting," he noted.
"For the first time, we’re actually seeing a difference between cell-only voters and land line voters when you take into account age, - Keeter said. Three Pew polls in the summer of 2008 showed a much stronger Obama lean among cell-only electors under 30 (59 percent, vs. 32 percent for Republican candidate John McCain) than in the land line group (53 percent Obama, 38 percent McCain).
It’s a sensitive issue for pollsters, who stake their reputations on getting the numbers right.
"On November 4, we have our Super Bowl," Nielsen said. "One or two points is a big issue. You get that wrong, and you lose all credibility."
Pew and Gallup have started folding cell phone data into their results. Gallup is using cell phone polling to track the habits of 1,000 Americans, as part of an ambitious 100-year project to provide a comprehensive look into the lives of Americans. "If you’re going to do a mammoth undertaking, you’re going to do it right," Nielsen said.
At least four other polling efforts, including Time/SRDI and ABC polls, are calling cell phones in presidential polls.
But most pollsters say a 2 percent bias isn’t enough to justify the extra cost of cell phone polling.
"Every survey is a compromise," Blumenthal said. "You interview a sample of people because it’s too expensive to interview everyone."
For pollster’s, switching can be a headache. Cell users may live far away from where they signed up for service, skewing valuable geographic data. Federal law forbids pollsters from calling mobiles with automatic dialers - so pollsters must pay workers to dial numbers and do live interviews. Pew found that more than 40% of those who answered the phone were too young to vote. And - though it hasn’t happened - some pollsters wonder about liability claims: what if a distracted respondent on the go were injured while answering a survey?
It’s too early to tell if cell phones will have any impact on this year’s presidential contest, said Scott Rasmussen, president of Rasmussen Reports, which doesn’t do cell phone polls.
"Polling is easy in theory but difficult in practice," Rasmussen said. "Cell phones are a factor, but there are other factors as well."
For example, he considers it hard to predict whether young people will turn out in higher numbers this year, since models predicting turnout for older adults don’t work the same way for younger voters.
"It’s too close to call this year," he said.
If pollsters who skip cell phones are wrong, though, Breslin’s critique just might come back to haunt them.
Contributed by NYU Livewire.
|
 |
|
