Monday, October 27, 2008

Early Voters in North Carolina

Although the early voting numbers in North Carolina so far have been very exciting for Democrats so far I am sorry to report that it appears they'll revert to normal.

So far 28% of early voters have been black. We weighted our poll to 21% black turnout, and that made the percentage of black early voters in our poll 31%, pretty similar to what it has actually been. That leads me to continue to believe African Americans will account for 20-22% of the electorate, as I have thought all along.

Looking at it another way, 49% of blacks in our survey said they had already voted. Only 29% of white voters said the same. If the whites planning to vote actually do follow through and do it the racial demographics of the North Carolina electorate will end up being almost identical to the state's population. Keep in mind this is still a big improvement from 2004 when blacks accounted for just 18.5% of general election voters.

The party figures for early voters in our poll also closely match what they have actually been so far. The true figures have been 55% Democrats, 28% Republicans, and 17% independents. On our poll it came out 57-25-18. The party breakdown of those who have not voted yet but plan to for our survey is 44% Democrats, 40% Republicans, and 17% independents. So the party imbalance should revert to the 13-14% registration edge Democrats in the state hold as well.

Nonetheless the numbers on who has already voted are good news for Democrats on several fronts:

-Not only is Obama winning 63-36 overall in votes already in the bank, but he's up 60-34 with independents who have turned out already. There's been a huge spike in independent voters this year, and they certainly seem to be leaning toward Obama.

-There is also very good news for Kay Hagan and Bev Perdue. There's been a lot of concern about whether Obama's supporters would just cast a vote for him and go home, or fill out the entire ballot. Among those who have already voted for Obama 92% said they also voted for Hagan and 91% said they also voted for Perdue. By a small degree that's a higher margin of party unity than the Republicans are showing- 89% of those who say they already voted for John McCain also say they voted for Elizabeth Dole and Pat McCrory.

Bottom line on the early voters: it's a lot better to go into election day with a huge lead than not!

18 comments:

  1. what gives you any indication that people who haven't shown up so far will? Really this is all speculation now over who will or will not show up. What we DO know is who has shown up and it aint no Reps. So unless I see any indication before election day that Reps are going to show up I will follow the voter registration trends in NC and those look positively ugly this year for Reps. They have lost almost 3 percent of their share of the registered electorate, while Dems share has climbed nearly as much.

    ReplyDelete
  2. Tom,
    Would it also be fair to say that it looks like turnout will also be at the same levels as 2004? Your poll says 33% of voters have already voted, which closely mirrors the 2008 EV/2004 total (34%) for NC from Michael McDonald's site:
    http://elections.gmu.edu/early_vote_2008.html

    ReplyDelete
  3. Anonymous #1,

    They don't get called unless they have a history of voting, and if they take the time to answer a 16 question survey I think they're probably going to vote.

    RS,

    I think turnout will be higher than in 2004. It is quite possible that early voters, just as they are more eager to vote, are also more eager to answer a poll. If early voters so far actually account for 30 or 31% of the electorate that would allow room for an increase in turnout.

    ReplyDelete
  4. In the Charlotte, McCrory has always gotten a good percentage of the minority votes. It appears from folks working the polls in the Charlotte, he is doing well again.

    ReplyDelete
  5. I'm of the opinion that the early voting is targeting New Voters and Infrequent voters, which is why the AA population is dramatically up. If there will be a crushing bump in performance, it may well show up on election day where there is again a massive number at the polls, of the regular every day voter, with the infrequent voter already having lodged their vote. I'm not convinced that it'll be a normal year, for this year is anything but normal. Good poll though, need to keep the fire light in NC for sure to get election day GOTV maximized.

    One upside to your poll, there is a massive number of persuadables out there in NC who will wait to vote, and who will get a taste of Obama's closing argument and hopefully the special on Wednesday night; could be the difference maker.

    ReplyDelete
  6. McCrory in Charlotte: How do they know that he's doing well with African Americans there? Are the poll workers looking at their votes after they're cast?

    ReplyDelete
  7. Poll workers, meaning folks selected by the Democrats and by the Republicans to observe the election process, not people employed by the Election Board.

    ReplyDelete
  8. Regarding RS's point:

    Are there any statistics on how many people have requested absentee ballots? It is possible that many have done so, and mailed them in already, but they haven't been received/counted yet.

    ReplyDelete
  9. Tom you may want to remember that early voting this year has already exceed that from 2004. In 2004 30% of votes were cast early, as of today 34% has been cast early in NC. Hmm if Obama is up big in early voting and early voting has already passed that from 2004, why should I not believe that early voting will be more than 40% of those cast this year? If that is so then Obama has a really good chance of winning NC. Again I have seen no indication of a "surge" in Reps at the voting booth. And without such a surge I really do not think McCain can win NC on the back of Dixicrats this year. This aint your grandpa's NC!

    ReplyDelete
  10. Early votes probably will be more than 40% of the total- 33% is the percentage of people who intend to vote this year who have ALREADY voted. The other 67% may vote in the last week of early voting or may vote on election day. I imagine in our final preelection poll the percentage of people saying they voted already will be well over 40%

    ReplyDelete
  11. It looks like there will be more than 2 million early voters this year. If that represents only 40% of total votes, that means we'd expect 5 million votes to be cast, which would be 80% of all registered voters. That sounds awfully high to me. It seems more likely that early voters will account for closer to half of all voters, which would then put overall turnout in a more believable 65%-70% of registered voters. It also means for McCain to win, he'll have to do as well or slightly better with election day voters than Obama does with early voters. It will be interesting to see what the party breakdowns are after Saturday. McCain will want to see the gap close considerably. If Democrats retain a significant edge it may be hard for McCain (and Dole) to make it up.

    ReplyDelete
  12. This is sort of on topic here--I posted this question at 538 but think maybe Tom can answer for me since he has been looking at the early vote in NC.

    One of the things that a lot of republican commentators or analysts say is that Obama's operation is all smoke and mirrors because in the end the new voters or the Obama voters, will not show up. But given how effective the GOTV for the early vote has been so far from the Obama camp can we not now see that this is a machine that actually works (in the general election context, since the primary proved they had a good operation but it was much smaller)? Look at the numbers in FL, NC, VA, GA etc Are these days of early voting not excellent multiple dry runs for the one day they really have to get out the vote with all they have got?

    Thanks.

    ReplyDelete
  13. From your polling it appears the one demographic that is holding out to vote are Republicans. Your post surmises that "assuming" these Republicans turn up to vote on election day as they express they will in the poll then you get the topline numbers revealed. I have a sneaking suspicion that a lot of these Republicans will not turn up on election day. Call it anticipated defeat syndrome. Much like votes cast on the West coast tend to wane in the evening as the election returns start coming in from the East coast, I suspect that some of these Republicans will not bother to show up figuring that the election is probably already lost (heck look at the infighting among Mccain and Palin's staff) so why bother take the extra time to cast what is perceived as nothing more than a protest vote. It is just a theory, but my gut tells me that there will be some of that this election.

    ReplyDelete
  14. The polling here - including the huge disparity in Early Vote turnout - is very reminiscent of the polling leading up to the NC primary, with Obama running up a huge Early lead over Clinton. Part of the reason that PPP was the most accurate pollster in that cycle was the fact that they polled a larger sample of people who had actually voted already - rather than just people who said they were going to vote, where the margin was much closer.

    In NC's primary, Obama did "bank" enough Early Votes to win before May 6. Clinton cut down his margin that day, although obviously not nearly enough to matter. She could actually have "won" on election day and still lost the election - as long as she didn't get above (roughly) 55%, Obama already had too large a margin in the Early voting for her to overcome.

    I don't think that Obama will be able to bank quite the same lead in the Early voting here, just because the race is much tighter. But I do think that if this week's voting follows the turnout of the last two, and if he pulls down the same margins that he's polling here, then as long as election day turnout isn't above 90% (no way), I think Obama could still win if he keeps McCain below (roughly) 53% and gets at least 45% on 11/4. That's roughly what this poll is showing in the "plans to vote," and I actually expect a much more even split. Turning out every vote is still key, but this polling looks very good - not quite as good as heading into the primary, but nearly!

    ReplyDelete
  15. What qualifies as a history of voting? Does this mean you are not polling newly registered voters? How about newly registered voters who voted in the primary?

    ReplyDelete
  16. Interestingly with another 200 thousand votes yesterday the mix remains 55%D, 28% R, and 17% U, with 28% Black/African American and 56% female. I agree that these percentages should close in, but Republicans better hope the D vote is down close to 50% after Saturday. If the Ds are still at 54%-55% when early voting is finished, it is going to be very hard to catch up.

    ReplyDelete
  17. tom, what do you expect black turnout to be in North Carolina this year? What was it in 2004? Do you have a guess for nationwide? I know that in '04, it was 60%. What do you think black turnout will be this year nationwide?

    ReplyDelete
  18. Well, Tuesday was another 200 thousand day. Total votes have passed 1.6 million. Finally we're seeing a little tightening, 54%D vs. 29% R, although Black/African American remains at 28% and Females at 56%. With four days left it now appears there's a good chance that more than half the NC votes will be cast early.

    ReplyDelete