The Cell Phone Effect
Jeff Alworth
When pollsters conduct telephone surveys and call only households with a landline, do they fail to capture an important part of the electorate who use only cell phones? For the past half-decade or so, pollsters believed the answer to this question was no. However, Pew just released an extremely interesting report (okay, interesting to poll geeks) suggesting that there is a marginal bias in landline surveys. These surveys underestimate support to Obama--and quite likely, to Jeff Merkley as well.
Pew looked at surveys they conducted with landline-only samples and combined landline/cell-phone samples. They found that:
In each case, including cell phone interviews resulted in slightly more support for Obama and slightly less for McCain, a consistent difference of two-to-three points in the margin.
There has been such a strong correlation between age and the cell-phone-only population that pollsters have been able to weight their findings for age and eliminate bias. But Pew found that the young cell-phone-only voters are more Democratic than their landline-using counterparts.
Among landline respondents under age 30, there is an 18-point gap in party identification - 54% identify or lean Democratic while 36% are Republican. Among the cell-only respondents under age 30, there is a 34-point gap - 62% are Democrats, 28% Republican. The difference among registered voters on the horserace is similar: 39% of registered voters under 30 reached by landline favor McCain, compared with just 27% of cell-only respondents. Obama is backed by 52% of landline respondents under 30, compared with 62% of the cell-only.
It's not a huge deal, but it could be decisive in a tight campaign. Take for instance, oh I don't know, the US Senate race between Gordon Smith and Jeff Merkley. In that new SurveyUSA poll Kari reported on earlier, we can see how similar the age demographics are for Obama and Merkley. Among the 18-34 cohort, Merkley has an 18-point lead, 54% - 36%. That's similar to the advantage Obama enjoys among young voters, where he's +24%. (SurveyUSA conducts their polls through a random-dialing system, catching unlisted numbers and cell phones, so the 44% - 42% lead they measure for Merkley would be unaffected by the cell-phone bias.)
It is definitely worth reading the small print when you see other polls, though--as this report suggests, methodology matters. Last week's Portland Tribune/Fox 12 poll, for example, appears to be be an all-landline sample, and it had Smith up by three. By all accounts, the race is extremely close--Pollster's rolling average has it at a dead heat, 42.2% to 42.3%. But perhaps there's a bigger gap between the two than we know. It's a trend worth watching.
 |
More Recent Posts | |
Albert Kaufman |
|
Guest Column |
|
Kari Chisholm |
|
Kari Chisholm |
Final pre-census estimate: Oregon's getting a sixth congressional seat |
Albert Kaufman |
Polluted by Money - How corporate cash corrupted one of the greenest states in America |
Guest Column |
|
Albert Kaufman |
Our Democrat Representatives in Action - What's on your wish list? |
Kari Chisholm |
|
Guest Column |
|
Kari Chisholm |
|
connect with blueoregon
Sep 25, '08
I'm reminded of the infamous headline, "Dewey Wins"!
Any poll that does not include cell phones marginalizes the under 30's. Most of them do not have landlines and are therefore excluded. It would be intuitive that these folks are probably more scewed to the progressive/democrat side of the equation than the Hannity side.
Sep 25, '08
Not just the under-30s, although that generation is more likely to have ditched landlines completely. I know at least a dozen people over 40 that have cell phones only. I've been cell-only for nearly 16 years now and have never regretted it---partly because I don't get the polling or phonebank calls.
Good reminder of something that may well be skewing the data we all read every day...thanks!
1:46 p.m.
Sep 25, '08
It's not just that it's younger voters--that would be, and has been, easy to weight for. It's that younger cell-phone users are more liberal than young landline users. Or anyway a greater number ID as Democratic.
What would be especially interesting is to see if they're actually more libertarian. A far greater number of techies are liberatarian than other groups, so it's possible that this effect will not uniformly skew away from Dems. If Ron Paul were in the race, it might be that all bets are off.
Sep 25, '08
Any poll that does not include cell phones marginalizes the under 30's.
Of course, they marginalize themselves by not voting their numbers. My guess is that any polling bias that exists -- and it probably does -- hasn't had much impact because the under-30 crowd votes less often than they say they will. And I would guess (without any evidence whatsoever) that the more liberal under-30 voters vote even less than their more conservative under-30 counterparts. So although the polling bias likely exists, it may be self-correcting if those same voters don't show up on election day.
My big fear for November is 1) that there are a lot of white Democrats who are saying they'll support Obama but won't actually do so when they step into the ballot box because they can't bring themselves to vote for a black man, and 2) the under-30 crowd where Obama receives such strong support won't show up on election day. There are so many elections where we've talked about the "energized youth vote" but been jilted at the altar when it comes time to actually commit.
Anything less than a 10-point lead for Obama going into November, and I'm going to be extremely nervous.
2:42 p.m.
Sep 25, '08
"There are so many elections where we've talked about the "energized youth vote" but been jilted at the altar when it comes time to actually commit."
Didn't we have record youth turnout in 2006?
Side note, Scott Keeter hired me onto his staff when he was the Director at VCU's survey house. Smart cookie, great guy. If he says there's now a bias, there is.
Sep 25, '08
I have a hard time believing that 90% of people under 30 identify strongly with either major party. Was that 62/28 split among only respondents who were identified with some political party? It's hard to tell from the summary at Pew. But I'd bet this excludes Independent voters, which of course tilts the cellphone group even further in Obama's favor.
Jeff -- If Paul's recent endorsement of the explicitly theocratic candidate doesn't ruin his support among this libertarian group, nothing will.
Sep 25, '08
You should probably also count people with caller id among those who are not polled. At our house we screen the multiple calls received every night from toll free numbers, the Gallup Poll and others. We are an over 50 household.
Sep 25, '08
More oldsters of the boomer generation are now using cell phones only.
Here's a poll just out, CBS/NY Times, Obama up 47-42. http://www.cbsnews.com/stories/2008/09/25/opinion/polls/main4478890.shtml Reportedly CBS does sample cell users, but at what percentage of the sample? According to Nate Silver at (fivethirtyeight.com) it makes a difference of 2.8 per cent. I suspect it's larger.
Sep 25, '08
My wife and I are both 55. We have not had a land line in many years. We each have a cell phone for work and a personal cell phone. Only a moron would pay Qwest another 45 dollars a month for nothing.
Obama wins!!!!!!!!!!!!!!!!!!!!
10:27 p.m.
Sep 25, '08
Yes we had record turnout of younger people in 2006 and in this year's primary.
2004 numbers for youth turnout by state:
youth turnout
Average 47.7%. Oregon was 15th at 49%. Minnesota was#1 at 69%.
From the Student PIRG's
The Student PIRGs’ New Voters Project is the nation’s largest youth voter mobilization program. Since 2004, we have registered more than 600,000 young people and made more than 650,000 peer to peer voter turnout contacts to get young people to the polls on Election Day. Due in large part to our efforts, the youth vote increased by 4.3 million votes, or 9% in 2004 and an analysis of our work in 2006 found that in the student dense precincts in which we worked with our allies, youth voter turnout increased on average by 157%.
They are not at all the only organization registering and engaging younger voters. The Oregon Student Association is in the process of registering 30,000 more young people and have an unmatched record in getting them to turn out.
It is a misperception that young people are unengaged and do not vote. They ebb and flow just like the rest of the electorate and it is not uncommon for them to register as NAV, which makes them hard for parties to track.
Young people are very aware of what is riding on the decisions that will be getting made in November and they have no intention of sitting it out.
12:26 a.m.
Sep 26, '08
I don't doubt the cell phone effect, but it does seem to me there are countervailing effects with Obama that at least so far have overcome it.
Obama is a chronic underperformer vs. polls prior to election day. It happened time and again during the primary. It seemed like his vote total was regularly 5% below what the polls would have suggested,
That in and of itself doesn't mean the cell phone effect is an illusion, but it does suggest that at the very least it is outweighed by other effects (such as youth non-voting.)
Many a political campaign has died at the altar waiting for the youth vote to come down the aisle. Will young voters in 2008 leave Obama in the lurch?
I dunno, but it wouldn't be anywhere close to a shock if McCain wins this thing.
4:28 a.m.
Sep 26, '08
Obama is a chronic underperformer vs. polls prior to election day. It happened time and again during the primary. It seemed like his vote total was regularly 5% below what the polls would have suggested,
That statement is actually wrong in an interesting way. Obama underperformed the polls in some states and overperformed in others. Nate Silver discusses these phenomena at fivethirtyeight.com.
<hr/>