Poll Showing Highest Obama Approval in Two Years Has Flaws

National   Steven Ertelt   May 11, 2011   |   11:07AM    Washington, DC

A new poll showing President Barack Obama with his highest approval rating in two years is generating considerable buzz in the nation’s capital today, in part because it has been found to be severely flawed.

The new Associated Press-GfK poll, taken on the heels of the killing of terrorist mastermind Osama bin Ladin, shows what AP calls “worrisome signs for Republicans” because “the president’s standing improved not just on foreign policy but also on the economy, and independent Americans — a key voting bloc in the November 2012 presidential election — caused the overall uptick in support by sliding back to Obama after fleeing for much of the past two years. ”

The poll found 60 percent of Americans give Obama a positive approval rating and more than half of Americans say he deserves re-election. A majority call Obama a strong leader who will keep America safe, 73 percent say he can handle any future terrorist threats, and 52 percent say Obama is doing well on the economy, despite rising gas prices and a recovery that can only be described as sluggish at best.

While 53 percent say he deserves re-election, 43 percent say he should not get a second term.

The Associated Press-GfK Poll was conducted May 5-9 with 1,001 adults nationwide and it has a 4.2 percentage point margin of error. While polls of adults typically yield more favorable results for Democrats like Obama than polls of voters or likely voters, several conservative pundits note the survey has significant shortcomings.

Jim Geraghty of National Review notes the polling sample AP used is problematic.

“It is of adults, which isn’t surprising … you don’t have to be a registered or likely voter to have an opinion on the president,” he writes. “But then you get to the party ID: 46 percent identify as Democrat or leaning Democrat, 29 percent identify as Republican or leaning Republican, 4 percent identify as purely independent leaning towards neither party, and 20 percent answered, “I don’t know.”

“For contrast, the AP’s immediate preceding poll was 45 percent Democrat, 33 percent Republican; the likely voter pool in October 2010 was 43 percent Democrat, 48 percent Republican. The poll’s total sample in October 2010 split 43 percent Democrat, 40 percent Republican,” Geraghty continued. “With a poll sample that has a 17 percentage point margin in favor of the Democrats, is anyone surprised that these results look like a David Axelrod dream?”

Ed Morrissey of HotAir, also weighed in on the skewed polling sample.

“Move over, CBS.  Hang up the kid-leather gloves, WaPo/ABC.  There’s a new sample-skewing sheriff in town, and it’s the Associated Press.,” he said, calling the poll an outlier. “Oddly — or perhaps not — the AP report doesn’t include a link back to the survey’s raw data.  In order to find it, one has to go to GfK’s site for its AP polls.  The partisan breakdown in the sample is found about halfway through the PDF, and it explains a great deal about how Obama managed to get such a high boost in this poll while others showed shallow bumps that had already started to subside.”

Morrissey says the 17-point gap favoring Democrats in the poll is “more than twice what was seen in the 2008 actual popular vote that elected Obama.”

“It’s pretty easy to get Obama to 60% when Republicans are undersampled by almost half,” he says. “Frankly, this sample is so bad that no real insights can be gleaned from it.”

UPDATE: The Associated Press responded to Geraghty at National Review, telling him they use “consistent methodology.”

AP claims: Some conservatives criticized the AP-GfK poll as heavy with responses from Democrats that skewed the results. AP-GfK polls use a consistent methodology that draws a random sample of the population independent of party identification. Such identification is not static and tends to fluctuate over time along with other political opinions. However, the change in party identification in the current AP-GfK current poll is not a statistically significant shift from the previous poll in March and could not by itself explain the poll findings.

Geraghty responds I appreciate the AP noting the criticism. They emphasize that their sample hasn’t changed significantly from the immediately preceding poll, where the sample split 45 percent Democrat, 33 percent Republican. But the question isn’t really whether the sample changed too much from their poll in April; the question is whether the sample accurately reflects the American public at large, and whether we indeed have 1.63 Democrats in this country for every 1 Republican. If their sample had an unrealistic proportion of Democrats one month, it’s entirely possible they can get a similar unrealistic proportion the following month.