The Pew Research Center made waves Friday with a new poll showing President Obama leading by a double-digit percentage – but the polling outfit is coming under criticism for allegedly over-sampling Democrats and skewing the results toward the president.

The poll itself showed Obama with 51 percent support over Mitt Romney's 41 percent. The president has consistently led Romney in Pew polling over the past six months, once by as much as 12 percentage points.

But beneath the top-line number in the latest poll were some curious figures. Pew reported polling 459 people who identify as Republicans, but 813 who identify as Democrats. The rest were independents.

Polling this year typically includes more self-identified Democrats than Republicans. Pew reported the difference in this survey, though, to be 13 percentage points – a wider margin than is typical.

“It’s just ridiculous,” Democratic pollster Doug Schoen said, when asked about the Pew figures.

Reliable figures for the percentage of registered Democrats to registered Republicans are not available, since not every state registers voters by party. But Schoen estimated that Democrats probably only have a 3-or-4 point advantage in that regard. He said the advantage was 7-8 points in the 2008 race, and certainly wouldn’t have risen since then.

“There’s systemic bias in the Pew poll,” Schoen concluded.

As to why that might be, he posited that sometimes Republicans are less inclined to answer polls.

But Pew adamantly defended its method in a lengthy and detailed response on its website Friday afternoon.

The polling outfit acknowledged the 13-point difference among Democrats polled was higher than usual – typically, Pew reported, it’s about 7 points.

But Pew said that although surveys adjust for demographic factors, it would be wrong to smooth out the party identification numbers. Pew said those numbers “ebb and flow with candidate fortunes,” noting the measure is not the same as party registration.

“Party identification is another thing entirely. Most fundamentally, it is an attitude, not a demographic. To put it simply, party identification is one of the aspects of public opinion that our surveys are trying to measure,” the statement said.

Pew acknowledged the hubbub surrounding its latest poll but deflected it.

“In every campaign cycle, pollwatchers pay close attention to the details of every election survey. And well they should. But focusing on the partisan balance of surveys is, in almost every circumstance, the wrong place to look,” the Pew statement said.

Pew also defended itself against complaints that it uses registered voters and not likely voters, noting that most registered voters end up going to the polls.

The 10-point lead for Obama was certainly out of step with other recent surveys. The Real Clear Politics average of national presidential polls puts the Obama edge at 2.7 percentage points.

READ FULL SOURCE ARTICLE

Editor's Note: The only polls that matter are the two that are executed by the campaigns (which we will never see) and the one that takes place on election day.


The BasicsProject.org informational and educational pamphlet series is now available for Kindle and iPad. Click here to find out more...

The New Media Journal and BasicsProject.org are not funded by outside sources. We exist exclusively on tax deductible donations from our readers and contributors. Please make a tax deductible donation today.