First of all, we have some new unconfirmed but strong evidence that Obama may be doing what he needs to do to turn out sporadic and unlikely voters, whereas many of the Romney voters may be people that we knew would vote anyway, but who are just doing so a little bit earlier. This is of crucial importance, because if in fact Romney's higher early vote totals are coming from likely voters who would otherwise vote on election day, and if Obama is turning out unlikely and sporadic voters, then President Obama really is in a very good position to win North Carolina.
CB8421 has done some excellent work and has found vote history data which includes information on which voters voted on election day in 2008. I haven't yet had a chance to confirm this independently, but CB8421 writes:
I just compared the 2008 voters to 2012 voters by county ID and voter registration number. Below are my findings:
Out of 1.508 million early 2012 voters in NC, 545k did not vote early in 2008. Of those 545k, it looks like 460k are completely new voters. 220k of those 460k are Dems. 128k are Reps. 110k are Unaffiliated.
Of the 545k who didn't vote early in 2008, 241k are Dem, 171k are Rep, and 130k are Unafilliated. So Reps have shifted 43k from election day to early, while Dems have shifted only 20k, indicating a lot more NEW voters for Dems than Reps.
...
To get the 43k, I compared the total number of Republicans who didn't vote early in 2008 but have voted in 2012 to the total number of Republicans who voted early in 2012 but didn't vote at all in 2012. There were (as of yesterday, updating for today now), 171k Republicans who voted early in 2012 but didn't show up in the early voting list for 2008. However, only 128k of them didn't show up in the overall voting list for 2008, indicating that 43k voted on election day in 2008.
On the flip side, there are 241k Dems who voted early in 2012 but not early in 2008. Of those 241k, 220k didn't vote at all in 2008.
...
FWIW, if you remove voters who have just moved from voting on election day to voting early, the %s in NC are:
Dem 50.7%
Rep 30.0%
Other 19.3%
Other and Rep +2% from 2008, Dem -4% from 2008 as of same day before election.
Results will turn on the "Other" vote.
One other note: 0.7% of Rep were Dem in 2008. (Not much shifting the other way - only 0.1% went from Rep to Dem.)
In other words, this is going to come down to turnout. Again.
...
New numbers through this morning (remember, I only include "accepted" ballots, not all ballots:
DEM New Early Voter 275,160
LIB New Early Voter 2,536
REP New Early Voter 199,260
UNA New Early Voter 150,279
DEM New Voter 245,477
LIB New Voter 2,200
REP New Voter 145,022
UNA New Voter 125,889
Difference is the "shifting of the deck chairs". Second number is real new turnout.
Keep in mind, some of this is bound to happen anyway as old people die and people turn 18. 100,272 people who are first time voters are 18-24. 46,480 are Dem, 23,478 are Rep, 29,506 are UNA. That's about 20% of the total. 167k are African American, or 33% of the total, which is disproportionate to AA registration in the state (which I believe is closer to 25%).
Calling AA voters "unlikely" is unrealistic. They know this may be the last chance to elect one of their own in their lifetime, and they'll be turning out in the same numbers as 2008.
Again, thanks to
CB8421 for locating this data and for the excellent work!
It will take me some time to comb through this data, so what I'll probably do is just post a short update tomorrow with just the standard charts and graphs. Then later in the evening, if I have found enough good information on sporadic voters, I'll post a separate diary with that information; otherwise, I'll combine the data on sporadic voter turnout with Wednesday's update.
Monday Early Voting:
Monday early voting in North Carolina is difficult to interpret, thanks to Sandy. There was at the least a bit of rain Monday in much (all?) of NC, and rain is known to suppress turnout. In addition, the storm may have caused some counties to process their data more slowly than usual. In particular, I suspect from the distribution of the Monday votes among counties that both Mecklenburg and Durham counties are substantially under-reporting (their vote shares for the Monday votes we have in are only 8.4% and 2.5%, as opposed to the 10.7% and 4.1% they have had so far in the early voting period). Since both of those counties are very large and voted strongly for Obama in 2008, if those numbers come in tomorrow, Obama's estimated Monday vote total may go up significantly.
Thus there is some uncertainty about how reliable and complete the data we have for Monday voting actually is - we'll have to wait and see.
Click on the picture below for a full sized chart.
One thing we do know - at least 184,766 people voted on Monday, making Monday the 29th the highest turnout day so far in the early voting period except for Friday, October 26th. Even in days without a hurricane, some votes have been processed late, so that number is all but certain to go up.
However, turnout - or at least the turnout data we have so far - was significantly lower than on the same Monday in 2008 (when 210,155 people voted). If more data does not come in, this will be the first day in which turnout was less than in 2008.
Another thing to be mindful of is that in 2008, turnout was fairly similar each day from Day 12 to day 17 of early voting, averaging about 228,225 votes per day. So if 2012 is similar to 2008, we should probably expect similar levels of turnout each day through Saturday. But, of course, Sandy casts more than a little doubt on whether Monday voting is likely to be predictive of the rest of the week. Tuesday voting may also be somewhat impacted - as I write this post, it is raining across much of NC.
White Republican turnout continues to be up, and continues to diverge from 2008:
Minority and White Dem turnout is up over 2008, but does not seem to be diverging from 2008 in the same way as White Republican turnout. Of course, in absolute numbers, Minority and White Dem turnout is nearly double White Republican turnout:
Even with all the doubt about Monday voting, Obama's margin continued to increase slightly, by another estimated 2,953 votes:
The daily margin yesterday does not look promising - but is that because of Sandy and Charlotte and Durham lagging?:
And here are the estimated vote percentage charts:
Previous NC Early Voting Diaries:
Day 1 & Methodology
Day 2
Day 3
Day 4
Day 5
Day 6
Day 7
Day 9
Day 10
Day 11
9:24 AM PT: Update - 904Dem provides the eye candy:
That is actually from Broward County, Florida, but here's some from NC: