Voters Agree: Media Intentionally Skewing Poll Data to Help Obama

Voters Agree: Media Intentionally Skewing Poll Data to Help Obama

When Public Policy Polling asked North Carolinians if pollsters were intentionally skewing poll data to help President Obama, I bet they weren’t expecting this: 

 Yes: 46%

No: 38%

Broken down, 77% of Republicans, 55% of Independents, and even 19% of Democrats agreed that the media were intentionally skewing polling data and failing to provide an accurate state of the race.

Ironically, the question was asked in a polling sample that used a whopping 48% Democrat sample and 14-point Democrat advantage over Republicans overall (in 2010, Republicans had a 1-point turnout advantage). The absurd pro-Democratic oversample was the only thing stopping the “yes” vote from blowing past that 50% mark. But still, a D+14 sample wasn’t big enough to give Obama a lead in the state where President Clinton tried to drag him across the finish line at the nominating convention last month.

Other interesting findings:

  • Obama’s support with black voters is down 3 points from 2008, while Romney is currently doing 2 points better than McCain did. Obama’s margin of victory was so small in 2008 that if everything else stayed the same from 2008 the 5-point swing among blacks would be enough alone to give Romney the win in North Carolina.
  • Not only was the sample D+14, but apparently North Carolina has a sever shortage of male voters: Women made up 56% of the total sample. Women support Obama by 12 points while men back Romney by 17. But I’m guessing the media would get huffy if we questioned their demographics skews once again.
  • Romney’s strength among Independents continues to play out. He holds a 54-36% advantage over Obama among the key voting group. But just like in Ohio and Florida, Independent voters suddenly are not all the rage with the media, especially since they no longer favor Obama.

Follow Dustin Hawkins on Twitter.

COMMENTS

Please let us know if you're having issues with commenting.