North Carolina-Civatas/SurveyUSA Trump +4

[pbrower2a]

https://twitter.com/Nate_Cohn/status/760922001338933248

Civitas/SurveyUSA has Trump+4 in North Carolina, but black voters are woefully underrepresented (18%, shld be >=22)

https://twitter.com/Nate_Cohn/status/760922404923244544

Poll is also too old, too male,. This is a sample off the voter file, which in NC has race/age/gender, so there's no excuse

Plus blacks support Trump 32%

Everyone go look at the NC voter file...

Black vote is at 22%, Civitas thinks it's 18%.
Female vote is 53%, Civitas thinks it's 50%.
Can't find any age statistics but 60% of the voting population being >50 seems too high.

I'm not gonna pull a StatesPoll but this poll is off in a lot of regards.

There needs to be a complete and total shutdown of Civitas polls until they figure out what is going on.

SurveyUSA interviewed 400 likely November voters from the entire state of North Carolina 07/31/16 through 08/02/16 using blended sample, mixed mode.

RV ≠ LV

This election is unique and the turnout might change a lot among different groups.

But still it is likely to be an outlier.

I will not use this poll. You can count on PPP getting a poll of North Carolina fairly soon, anyway. That will be worth the wait.

[pbrower2a]

https://twitter.com/Nate_Cohn/status/760922001338933248

Civitas/SurveyUSA has Trump+4 in North Carolina, but black voters are woefully underrepresented (18%, shld be >=22)

https://twitter.com/Nate_Cohn/status/760922404923244544

Poll is also too old, too male,. This is a sample off the voter file, which in NC has race/age/gender, so there's no excuse

Plus blacks support Trump 32%

Everyone go look at the NC voter file...

Black vote is at 22%, Civitas thinks it's 18%.
Female vote is 53%, Civitas thinks it's 50%.
Can't find any age statistics but 60% of the voting population being >50 seems too high.

I'm not gonna pull a StatesPoll but this poll is off in a lot of regards.

There needs to be a complete and total shutdown of Civitas polls until they figure out what is going on.

SurveyUSA interviewed 400 likely November voters from the entire state of North Carolina 07/31/16 through 08/02/16 using blended sample, mixed mode.

RV ≠ LV

This election is unique and the turnout might change a lot among different groups.

But still it is likely to be an outlier.

I will not use this poll. You can count on PPP getting a poll of North Carolina fairly soon, anyway. That will be worth the wait.

So you're only using polls you like now

As soon as I saw evidence that the Utah poll was a fake I rejected it.

....PPP polls North Carolina more than any other state, and nobody polls North Carolina as often as does PPP. Civitas has been all over the board. Screwy crosstabs get screwy results.

[pbrower2a]

Does anyone really believe that Trump will get 30% of the black vote?

Getting a 4% edge in North Carolina? Maybe. An unusually-small black vote that goes 32% for Trump? That isn't going to happen.