Obama Way Ahead Of McCain Among Cell-Phone Only Households

by: Chris Bowers

Sat May 17, 2008 at 17:17

In what could be a huge problem for pollsters not only this year, but in all future elections, Gallup has found a statistically significant difference in support for Obama between cell phone only households and all other households. Pollster.com has the scoop:

While the inclusion of cell phone only households makes little difference in the Clinton-McCain contest, it benefits Obama by a net four points: Without cell phone interviews, and weighted using Gallup's usual likely voter model, McCain would get 49% to Obama's 46% (clarification: this result combines six Gallup/USAToday surveys conducted so far during 2008). With the cell-phone interviews included, the result is Obama 48%, McCain 47%.

Given that cell phone only households are only 16% of the population, a 4% shift toward Obama when those voters are included indicates a wide gap between cell-phone only households and all other households in terms of Obama vs. McCain preference. Keep in mind that this was tracked across six Gallup surveys, large sample sizes were available for the subsets in this study.

Tempting as it may be, I don't want to yet conclude that this means Obama is much further ahead of McCain than current polling indicates. Four years ago, I banked on the so-called Incumbent Rule, which holds that challengers receive the vast majority of late deciders in elections against incumbents, to make a similar pro-Kerry projection, and it fell flat. Once bitten, twice shy, I guess. Still, this is hopeful news, given that Obama already leads McCain anyway.

Chris Bowers :: Obama Way Ahead Of McCain Among Cell-Phone Only Households

Tags: , , , , , , (All Tags)
Print Friendly View Send As Email

You forgot to say WHY Obama has a cell phone lead (4.00 / 2)
It's because McCain is older than cell phones:
The first fully automatic mobile phone system, called MTA (Mobile Telephone system A), was developed by Ericsson and commercially released in Sweden in 1956.

Sounds familiar (4.00 / 1)
I've heard this "pollsters are neglecting cell phone only voters!" thing in both of the last two elections-- both times to explain away bad poll results. In 2004 people were suggesting polls were undersampling Kerry because young progressive voters who would support Kerry might tend to be cell-phone-only; apparently polls weren't undersampling Kerry. In 2006 Karl Rove brushed off polls that showed bad news for the Republicans on the same grounds, that polls would be unable to sample... um... young progressive voters who would vote Republican? Not sure I ever really understood that one. Anyway, the polls that showed bad news for the Republicans were right.

Now, cell phone use is going up all the time, so I'm certain someday that stopped clock will be right and we'll have an election where cell phone usage interferes with polling; and in this case they seem to be doing something I've never seen done before, i.e., actually attempting to measure whether there is a difference in how cell-phone-only households poll. But given the history of this claim I'm tempted to be skeptical...

And incidentally I say all this as a member of a cell-phone-only household wherein both members support Obama...

Polling Getting More Hazardous (4.00 / 1)
The cellphone data is interesting, and logical, given that Obama is so strong among youth, who are more likely to be cellphone-only. Also, the recent special election in Mississippi-01 points toward an impact from higher black turnout. That one was supposed to be a lot closer than it turned out to be, and my bet is that black voters were a big part of the reason.

[ Parent ]
LOL! (0.00 / 0)
And incidentally I say all this as a member of a cell-phone-only household wherein both members support Obama...

So what does that tell ya?

[ Parent ]
probably the lesson here (4.00 / 1)
is not about cell phone users, but rather that Obama needs to turn out his supporters that are not "likely voters," especially but not limited to the "young minority" demographic.

New Jersey politics at Blue Jersey.

Land Lines? They Still Have Those? (0.00 / 0)
Wow, that's like a rotary dial, isn't it?

Not serious, I've got a land-line.  But like the article mentioned, we're one of those houses that has turned off the ringer on it because we never give it to anyone.  We get a call on it, it's a telemarketer or a wrong number.  It's only there for emergencies, and I'm considering dropping it completely now that all of our kids have cell phones.

Nearly half the people I know in the area have out of state area codes because they're keeping their old numbers rather than change.  Heck, I kept my old number for over a year on the last move, only changed because we were changing plans to get the youngest a line.  How the heck do you auto-poll someone who has an area code from the other side of the country?

Kerry Like (0.00 / 0)
Top 35 Trends that say Kerry will Take the White House in November
2) The 'Cell Phone Polling' Phenomenon: Traditional polling relies almost exclusively on landline telephone. Unfortunately, according to Charlie Cook of the Cook Political Report, as much as 18% of the electorate don't have land lines and instead rely exclusively on cell phones. The Hill gives us a little something about this demographic:

In-Stat.MDR, a wireless market-research firm based in Scottsdale, Ariz., conducted a survey of wireless users in February of this year. Of the 970 people questioned, 14.4 percent were cell-phone-only users, the majority of whom were single Americans between the ages of 18 and 24, living in mostly urban areas.
Anyone care to venture a guess as to how this demographic overwhelmingly votes?

Fat lot of good that did Kerry.

Cell Phone Databases - Betting Pool, Anyone? (4.00 / 1)
Marketing and polling firms have a pressing need for data on and segment of the population that has consumer preferences or political preferences significantly different from other segments of the population.  If the different preferences can be determined from some other communications (e.g., internet surveys), the other sources will be used for surveying this segment of the market.  But if it ever becomes clear that cell-phone only households have a distinctive pattern of consumer preferences or political preferences that can only be surveyed through sampling of a database of cell-phone numbers, then there will be big bucks in compiling and selling databases of cell phone numbers.

I'd wager that marketing firms will be compiling cell-phone only databases by 2012, or at least by 2016 -- unless some other more comprehensive databases not even conceived yet by those of us in our 60s are being used instead.  I will not be surprised if someone posts that marketing firms are already negotiating with cellular and digital phone providers to compile comprehensive databases.  Anyone want to establish a baseline for an "Open Left" pool on the year when a polling company first offers its customers polls based on databases that include comprehensive lists of cell phone numbers?

Given the increased turnout in this election among young voters who have no residential land lines, this is a significant finding.  Since Kerry failed to "turn on" the youth vote enough to increase "turn out," there was no noticeable effect in 2004, or in the McCain-Clinton polling match-ups. To follow EW's pithy observation, Bush, Kerry, and Clinton, like McCain, are all older than cell phones!

This factual information simply reprises the situation in 1936, when the Literary Digest (I think the sponsor) conducted a poll of households with telephones and found that Alf Landon was way ahead of FDR.  There was a huge difference in 1936 in the voting preferences of households with telephones as compared to households that did not have telephones in 1936. Therefore the population of voters with telephones reported a majority preference for Landon, although the population of all voters overwhelmingly supported FDR, and the population of voters without telephones supported FDR by super-landslide majorities.

The study that Chris cites shows an enormous majority for Obama among cell-phone only households.  If one sixth of the population accounts for a four per cent (4%) increase in proportional support for Obama, I think this means that the cell-phone only households support Obama by a margin 24% greater than the other 5/6 of the households that have a residential land line.

For example, if 44% of the sample without cell phones support Obama, but 48% of the total sample support Obama, then the one sixth of the population that has only cell phones support Obama by 44% + 24% = 68%.

I agree with the comments that caution us not to put much weight on the unique political preferences of cell-phone only households in interpreting polls this election year.  But this factor could easily play a big role, particularly in some states, as the campaign unfolds this fall.

Update (0.00 / 0)

Just discovered that the four-point gain was net points, not absolute points.  The absolute point gain was 2%.  The margin without the cell-phone only responders was McCain 49%, Obama 46%.

The margin including the cell-phone only responders was Obama 48% vs McCain 47%.

This means, if my math is correct, that Obama's margin was 12% greater in the cell-phone only households.  Adjusted, the margin among cell-phone only is 46 + 12 = 58%.  For McCain, the number is 49 - 12 = 37% among cell-phone only responders.

That's a 21-point lead for Obama in this sample of cell-phone only voters.  That does make more sense.  I thought 68% was implausibly high.

[ Parent ]
it's be nice if the blogosphere could have "where do polls come from" day (0.00 / 0)
and at the same time focus on collection methodology, statistics, the history of honest/dishonest polls in politics and how they are used in general, pushpolling, that sort of thing. most people who use polls to support some claim seem, well, let's say not as up to snuff as they could be, when it comes to when a poll is, and is not, useful support for an argument.

my own experience with numbers and polling includes working for a firm when i was young, well, actually hanging out with a gf and some friends of mine in the building where they worked and made calls for one summer. short version: they'd all get high between shifts, during shifts, during lunchbreaks, etc., and i thought (as a dumb teen) that it would be Kewl to hang with such interesting stoners. i bring this up because i got to see just how often this bunch of underpaid, undermotivated, not exactly sober folks "gathered information" that would later become the basis for this or that new theory of trend in some market. they did political calling too; anyone could commission a poll or sample or study from this company and they'd start making the calls. a short time later i myself worked for a telemarketing firm, and gosh the morals and professional standards there were even worse.

by the time i got to grad school and did a long, painful year of grad level statistics, i came to understand that sometimes it makes sense to look at poll numbers and try to extrapolate from them, but a lot of the time, not so much.  after the last few years of seeing just how corrupt the media and wall st and madison ave have been allowed to become, in this age of republican cronyism, ponzi schemes, anti-intellectualism, and flat out for-profit propaganda production...let's just say that as far as politics go, the only polls i care about any more are the elections and primaries. i caution people all the time: are you really sure you understand, in a math/science kind of way, "where that poll came from?" how it was produced, by whom, for what money, and how it will be (was intended to be?) used? chris is Man enough to admit when he's wrong and his predictions haven't panned out. i'll say that he's too hard on himself; a lot of the time i bet he's been fed faulty data from which he's based his conclusions; i trust his smarts over a lot of people, including for-pay pollsters with republican/SCLM motivations and connections.

IHMO: the cell phone skew is a big deal, and few polls are taking it into account. of course they don't: older folks don't use them like the young, and the older generation more closely mirrors/responds to the SCLM of 'what america thinks' so of course the pollsters don't want to upset that.


Open Left Campaigns



Advanced Search

Powered by: SoapBlox