Wireless World: Cellphones and polls

CHICAGO, Oct. 29 (UPI) -- Lobbyist Gary Shapiro attended a gathering of interest groups from across the political spectrum, left to right, in Washington, D.C., this week and heard briefings by pollsters, outlining their forecasts for next Tuesday's presidential election.

"There were representatives of 100 interest groups there," said Shapiro, chief executive officer of the Consumer Electronics Association, an industry lobby. "Whenever the pollsters would speak, the people would stand up and shout, 'What about the cell phone users? Are you polling them?'"

The political leanings of denizens of the wireless world may be a decisive factor in this election, pitting incumbent President George W. Bush and Democratic challenger, Sen. John F. Kerry, D-Mass., and in future presidential contests as well, experts told UPI's Wireless World.

Research by the CEA, located in Arlington, Va., reveals around 6 percent of voters now rely exclusively on their mobile phones for telephone communications. Contrary to conventional wisdom, it is not just college kids, cocooned in their university dorms, who rely exclusively on the mobile devices.

These users do not have a landline at home. Many others, who can be reached via conventional phone lines, tend to screen calls with caller ID and answering machines, thereby insulating themselves from the pollsters.

The CEA survey indicates as many as 8 million American households may be missing this year from the political polling process.

"Number portability, advanced text messaging, voicemail features and, most important, falling prices for service plans are allowing more and more consumers to cut the cord to their traditional landline services," Shapiro told Wireless World. "This trend is, without a question, making it harder for political pollsters to get an accurate read on local, state and national races."

According to the CTIA, the international association for the wireless telecommunications industry, use of wireless minutes in the United States increased by 35 percent from June 2003 to June 2004, just as the current presidential race was beginning in earnest, and the trend is likely to continue.

"Wireless communication continues to grow at a rapid pace," said Steve Largent, president and CEO of the CTIA, in a statement.

Some leading political bloggers already are challenging the validity of standard political polling methodology, noting that it discounts certain demographic groups -- perhaps unfairly.

"This evening, I found myself becoming genuinely frustrated with all the polls," stated a posting on the HorseRaceBlog, at jaycost.blogspot.com/2004/10/scoop-on-polls.html. "I am shocked at the number of polling outfits that are generally unreliable. I have become vexed by the issue of partisan weighting."

The blogger's research indicates that polls conducted by The Washington Post weight their samples 39 percent Democratic, 35 percent Republican 26 percent independent, but the Gallup Poll, which he called the "gold standard" of polling, left its figures unweighted in terms of partisan composition. Other polls do not publish their methodology, the blogger noted.

"In an election when the difference between the candidates could be as little as 2 percent, improper partisan identification weights can result in an inaccurate prediction," the blogger wrote Oct. 27.

Add to this the inability to identify properly and locate mobile phone users, and there could be an even bigger disparity between the polls and the electoral college reality, come Nov. 2, experts said.

"There is a whole bunch of problems with the polling," said Bruce Altschuler, a professor and chairman of the political science department at State University of New York at Oswego, and author of "Private Polling and Presidential Elections."

One vexing issue for pollsters, Altschuler told Wireless World, is weighing the polls to account for the missing mobile phone users.

"Will 80 percent of mobile phone users support one candidate over another?" he asked. "Pollsters will have to make adjustments to re-weight the folks they have reached."

Another expert, Jamie McKown, a professor of political communications at the College of Charleston, S.C., and the former executive director of the Democratic Party for the state of Nebraska, said concerns over the inability to account accurately for the political leanings of mobile phone users is a "huge question" that may have a "reverberating impact" on this election.

"You are absolutely right; it is not just college kids who are relying on mobile phones," McKown told Wireless World. "My wife and I have a landline, but we just use it to connect to the Internet. For other communications, we rely on a mobile phone."

McKown said in the general populace, there appears to be an edge among Republicans in getting their voters to the polls on election day. Those who cannot be reached by phone, however, cannot be mobilized and moved to the polls, he added.

"That makes mobile phones even more of a problem," McKown said.

This could account for "a couple tenths of a percent" difference between polls and actual public sentiment, said Ron Rapoport, professor of government at the College of William and Mary, in Williamsburg, Va., and author of a number of works on survey research.

Many of these mobile phone users may never have voted before, adding yet another complication to the process of measuring public sentiment.

"The exclusion of previous non-voters is a much more serious possible omission," Rapoport told Wireless World.

Many voters do not trust the methodology of any one poll anymore, but RealClearPolitics.com, an influential political Web site, headquartered in Chicago and run by Princeton graduate John McIntyre, compiles a list of all the major national polls and publishes an average of these polls.

On Wednesday afternoon, the RealClearPolitics.com site had President Bush leading Sen. Kerry 49.0 percent to 46.4 percent, or a 2.6 percent head-to-head spread, based on five national polls, including Gallup, Newsweek and the Los Angeles Times, all of which show the president prevailing.

Research by the CEA, conducted Oct. 2 through Oct. 5, among 568 likely voters, found 37 percent of those who rely exclusively on mobile phones do not consider themselves Republicans or Democrats, but rather view themselves as independent voters.

"That's compared to 28 percent of likely voters overall," Shapiro said.

Some traditional polling organizations have tried to knock down the idea their polls are somehow flawed by the absence of mobile phone users. They even have disputed the idea that 6 percent of consumers rely solely on mobile phones.

"Pollsters can't call cell-phone users, but, fortunately, only 4 percent of the people who are cell phone users solely use them," said Lee Miringoff, director of the Marist College Institute for Public Opinion, producer of the Marist Poll. "Therefore," he told Wireless World, they get picked up via landlines. It's not a real concern."

--

A weekly series by UPI examining emerging wireless telecommunications technologies. E-mail sciencemail@upi.com