The cell phone effect

As the pace of polling picks up in the Presidential race - about 40 (YES FORTY) state and national polls were released yesterday - I thought it might be useful to consider what these polls may be missing.

According to Pew Research, back in 2004, according to exit polls, 7.1% of voters were cell-phone only voters. That number is expected to double or more for this election.

So, what effect would adding the cell-phone voters have on the polls? On the overall numbers of the poll, according to Pew, the effect may be small. In a recent survey they conducted, when cell phone-only individuals were blended with the land-line individuals that typically respond to the major surveys, the mean difference among a range of questions was only 0.7%.

The problem is that, among certain demographic groups, that number changes dramatically.

As one might suspect, most of the under-represented groups in land-line only surveys lean or lean heavily toward Barack Obama. According to the most recent government estimate, more than 25% of those under age 30 use only a cell phone. Other demographic highlights from a 2007 SF Chronicle article:

-- Latinos at 15.3 percent were the most likely of ethnic groups to sever their landline and rely on cell phones exclusively,

-- Fifty-four percent of unrelated adults with no children went without a landline compared with 10.5 percent of adults with children.

-- Renters (26.4 percent) were more likely than homeowners (5.8 percent) to use wireless.

Let's see. Under 30 population generally expected to support Obama? (check) Latinos generally expected to support Obama? (check) Renters expected to support Obama? (check)

According to Salon,

Heretofore [the polling] industry has dismissed the cellphone-only population with a troika of "yes, buts." Yes, they're undercounted, but 1) they don't vote anyway; 2) their numbers are still small; and 3) we can find acceptable substitutes in the land-line population.

And to be honest, there is a fourth, still more powerful rationale that remains unstated: "Yes, they're undercounted, but it's too damn difficult and expensive to reach them."

As Salon points out, there is, of course, the fear that these groups - especially young voters - won't vote anyway so leaving them off of polls isn't that big of a deal. But if the primaries showed us anything about Barack Obama, he inspired the young people to 'prove 'em wrong' on election day. Turnout among high school and college voters and caucus goers was essential to Obama's success this spring.

The Salon article I quote above estimates that the inclusion of the cell-phone only crowd could boost Obama by 2-3 percentage points. Even if it's not that great - even if it's at 0.7%, as Pew conservatively estimates (afterall - some conservatives are 'wireless' as well) - those voters could well prove the pollsters wrong.