Today CNN released a poll showing a sizable Obama bounce. We’ve blogged CNN’s biased polls before and this survey has all the hallmarks of a truly unrealistic sampling of America. I made a joke about this previously, but it looks like CNN surveyed the floor of the Democrat Convention for it latest poll. They won’t reveal the party ID of voters but look at how their representative sample views the two parties:
- Democrat Party Favorability +8 at 51 to 43
- Republican Party Favorability – 10 at 42 to 52.
There is a 0% chance America sees the two parties this way and will again deliver an overwhelming Republican majority back to the House and no worse than 50/50 split in the Senate.
Here are my other examples demonstrating the bias in sampling. The first bullet is the most damning:
- Romney has a 14-point lead among Independents (54 to 40) and is down 6 in the poll. If Romney wins Independents by 14 points he wins the election in a blowout
- Obama’s favorability at 57% versus 42% unfavorable. Favorability is different from job approval but he’s never this high
- Obama is +1 on handling the economy 50 to 49. No survey EVER claims this
- Obama is +1 among men 48 to 47. Obama is typically down 6-8 points among men. He was +1 in 2008, not today
- Registered Voter breakdown was Democrat 441 (50.4%), Republican 397 (45.5%) and Independent 37 (4.2%). Registered Independent voters make up 20-25% of the electorate.
- Even in this hugely biased poll Mitt Romney has a +1 on favorability 48 to 47.
- Romney is +2 with people over 50 on handling medicare 49 to 47
CNN and ORC International continue to show extreme bias in their polling. If the above statistics were representative of today’s electorate Obama would win re-election in a landslide, yet they can only find him with a six-point lead. Thankfully this survey is worth the same as you pay to read this blog: nothing.
Addendum: I would like CNN/ORC International to explain the following: On Page 1 they say this is a survey of 709 Likely Voters. On Page 3 they say 351 of the likely voters will vote for Obama and 340 of the likely voters will vote for Romney. That’s a split of 50% – 48% in favor of Romney, not 52 to 46 as they say on the previous page. Did they re-weight in favor of the Democrat candidate the results of a poll that was already oversampling Democrats at 50% of the registered voters? There’s a lot not right in this survey. The same discrepancies hold true for Registered Voters: 875 Registered Voters surveyed, 441 voted for Obama (50%) and 409 voted for Romney (47%) but CNN reports the registered voter result was 53 to 45 in favor of Obama. Something’s fishy.