Quinnipiac: Clinton +2 in NC, +1 in FL

[Ozymandias]

https://poll.qu.edu/2016-presidential-swing-state-polls/release-detail?ReleaseID=2401

FLORIDA

Of the 65% (575/884) of respondents who said they voted early:

Clinton - 47
Trump - 43
Johnson - 2
Stein - 0
Other - 1

Rubio - 48
Murphy - 46
Other - 2

NORTH CAROLINA

Of the 67% (583/884) of respondents who said they voted early:

Clinton - 52
Trump - 40
Johnson - 2
Other - 2

Ross - 51
Burr - 43
Other - 2

Cooper - 56
McCrory - 40
Other - 1


 

[Ozymandias]

2) The early vote lead in Florida is MoE stuff. We were already expecting Clinton to outpace Trump during early voting, the question was more whether she could get a big enough cushion before election day.

Actually, there are enough  early voters now that their MOW isn't much less than the overall MOE.

And you're right, this poll has a lower EV lead for Clinton than several other Florida polls.

But I guess the upside is that if they are under-predicting her EV lead, then she'll have an even bigger overall lead than she does now...

[Ozymandias]

And you're right, this poll has a lower EV lead for Clinton than several other Florida polls.

But I guess the upside is that if they are under-predicting her EV lead, then she'll have an even bigger overall lead than she does now...

Which means that that poll has a lower ED lead for Trump than several other Florida polls.

But I guess the upside is that if they are under-predicting his ED lead, then he'll have an even bigger overall lead than he does now...

It works both ways, you know

But surely you see the advantage of Clinton having votes banked and going into Election Day already having a lead?  Especially when she's the one with the better ground game than Trump?

[Ozymandias]

https://poll.qu.edu/2016-presidential-swing-state-polls/release-detail?ReleaseID=2401

FLORIDA

Of the 65% (575/884) of respondents who said they voted early:

Clinton - 47
Trump - 43
Johnson - 2
Stein - 0
Other - 1

Rubio - 48
Murphy - 46
Other - 2

NORTH CAROLINA

Of the 67% (583/884) of respondents who said they voted early:

Clinton - 52
Trump - 40
Johnson - 2
Other - 2

Ross - 51
Burr - 43
Other - 2

Cooper - 56
McCrory - 40
Other - 1


 

So in the presidential race only 92% of Florida respondents and 94% of NC respondent (who claimed they already voted) were willing to tell who they voted for.  That means that there is a pretty good bit of uncertainty in the result, with respect to how this block actually voted.

If you include "Other" and rounding errors it's actually 94% of FL EV and 96% of NC EV, but yes, that's right-- there's consistently a few % of EVs who don't reveal who they voted for.

As I've said before, I really don't understand why they don't just throw these non-respondents out of the poll. (Perhaps it's because they did respond to SEN or GOV candidate questions?)