North Carolina Democrats put up a valiant fight, but Mitt Romney managed to edge out President Obama 50.46% to 48.29% in the Tar Heel state in 2012. Except for provisional ballots, which have yet to be tallied, Romney won NC by 96,489 votes, (2,252,830 to 2,156,341).
Without a doubt, even while they celebrate the national outcome, many North Carolina Democrats are ruminating gloomily upon local defeats. But on the positive side, North Carolina Democrats demonstrated that NC truly is a purple swing state, and that 2008 was not a fluke. As we will see, if we look closely at the 2012 results, we can find signs that over the longer term, Democrats in North Carolina have a bright future full of many hotly contested statewide races ahead of them.
North Carolina remained competitive to the bitter end in 2012, unlike several other states that Obama contested or considered contesting in either 2008 or 2012, such as Indiana, Missouri, Arizona, and Montana. The final result in North Carolina was closer than in numerous other swing states, such as Iowa, Wisconsin, Pennsylvania, Colorado, Nevada, New Hampshire, and possibly Virginia (depending on the last few votes dribbling in).
If President Obama had been able to maintain the lead he built up after the Democratic convention, and if he had performed better in the first debate, he may well have won North Carolina's 15 electoral votes.
When all is said and done, NC will now have only a modestly R+ PVI - probably around R+3. The demographic trends that put NC in play in 2008 have not stopped, and in 2016, it is a pretty safe bet that the Democratic presidential nominee will once again contest North Carolina. At any rate, if North Carolina is not contested, it will be because Democrats are repeating John Kerry's 2004 mistake of playing on too narrow of a presidential map.
Unfortunately, it is NC Republicans' move. And their next move will likely be to attack voting rights using potential voter suppression measures like voter ID and restricting early voting. Take it from a Texan - NC progressives will be well advised to lawyer up and prepare to fight off any and all attacks against voting rights in the arena of public opinion and in the courts.
How did turnout change from 2008 to 2012?
To figure out how Romney was able to narrowly win NC, we will first of all look at how turnout changed. Overall, turnout in NC increased from to 4,310,789 in 2008 to 4,464,995 in 2012. That's an increase of 154,206 votes (a 3.6% gain).
First of all, for reference, this map shows the total population by county in NC:
And this map shows the percent growth rate of the population from 2000 to 2010:
The map below shows the percentage increase in total turnout by county. Red counties have a decrease in turnout and green counties have an increase.
All across rural Appalachian western NC, we can clearly see that turnout slightly decreased in almost every county - except for liberal Buncombe county (Asheville).
In eastern rural DC, most counties had slight turnout increases, but there were also some counties with slight turnout declines.
Unsurprisingly, the largest percentage increases were in the Charlotte area and the Triangle area. There were large percentage increases in both urban counties like Wake and Mecklenburg counties, and also in suburban/exurban counties like Johnston and Union counties.
Now, how does the change in turnout look in terms of raw numbers of actual people voting, rather than as a percent increase or decrease?:
This map changes the emphasis, clearly bringing Mecklenburg County (35,276 more votes in 2012 than in 2008) and Wake County (39,845 more votes in 2012 than in 2008) into sharp focus.
Population growth, and hence also turnout increases, were heavily concentrated in those two counties, which happen to be two of the most strongly Democratic counties in North Carolina.
Other urban Obama-voting counties such as Durham (+7,310), Guilford (+10,341), and Forsyth (+6,606).
There were turnout increases in suburban/exurban counties as well (Union +7,754, Gaston +5478, Johnston +5206), which at first seems like a very good thing for Republicans. However, a fair amount of the population growth in these counties is from Democratic-leaning voters.
How did NC voters swing from 2008 to 2012?
The map below shows the swing in Obama's percentage margin in each county:
The percentage swing to Romney was clearly by far the strongest in Appalachian western North Carolina, where swings of 5% to 10% were the norm. But remember, we just saw that turnout decreased in Western NC. There was also a swing to Romney in eastern NC, but it was much more muted. Finally, there were modest swings to Romney in Charlotte, the Triangle, and the Triad - but they were significantly smaller than the national swing to Romney.
And the map below shows the change in Obama's raw vote margin by county:
This map looks a bit different. Even though Wake County swung much less in percentage terms than Appalachia, Romney's gains in Wake County were more important to his ultimate victory than the gains in any one small rural Appalachian county, because Wake county has many more people. Still, we should be a bit careful in over-interpreting this: although Romney got 22,796 more votes in Wake County than McCain, Obama also got 13,945 more votes in Wake county than he got in 2008.
In addition, Romney had relatively large margin gains in many relatively populous Piedmont counties, in Johnston county, and in the Charlotte exurbs. Finally, Romney's margin gains in rural western and eastern NC were small in any one individual county, but they added up.
North Carolina's Continued Democratic Trend
But now let's look at the results in NC in a slightly different way. Instead of looking at the swing to Romney in isolation of context, let's look at the swing in comparison to the national average swing. This is the "trend" relative to the national average.
At the time I am writing this, President Obama is ahead by 61,297,013 to 58,402,867 in the national popular vote (50.45% to 48.06%). This will change slightly as more results come in, but should be close enough to the final result for our purposes.
In 2008, President Obama defeated John McCain by 52.87% to, 45.60%, with a margin of 7.27%. So subtracting Obama's 2012 margin of roughly 2.38%, we find that the overall national swing was a 4.89% margin shift to Romney.
But in North Carolina, the swing to Romney was substantially less than that - only 2.49%. Subtracting 2.49% from 4.89%, we find that the trend relative to the national average in North Carolina was 2.4% towards Obama. That means that if there had been no overall national swing to Romney, Obama would have likely improved his margin in North Carolina by about 2.4% (which would have meant a win in NC by 2.7%, up from his .3% win in 2008). In other words, North Carolina showed in 2012 that it is continuing to trend Democratic.
Now, how does that 2.4% Democratic trend break down by county?:
(This map actually looks a bit redder than it should, because if a county has a +.4% Democratic trend, that is rounded down to 0 and shown as light red.)
What this map essentially shows is that Republicans gained ground relative to the national average in (some, but not all) stagnant rural areas of NC - and that's it.
With the exception of several counties, Western rural NC swung Republican more than the national average. In addition, some rural counties in the central NC Piedmont region swung to Romney more than the national average swing. Finally, a grand total of 7 coastal counties swung to Romney. And that's it.
Obama gained substantial ground, relative to the national average swing, in rural eastern NC counties with large minority populations.
But in addition - and more importantly for the future - every single major growing urban area in North Carolina trended Democratic.
Mecklenburg County (Charlotte) trended 3 points Democratic, and most surrounding suburban/exurban counties also trended Democratic.
Orange County (Chapel Hill), Durham County (Durham), and Wake County (Raleigh) all trended Democratic as well - by 2%, 6%, and 2% respectively. Even overwhelmingly Republican Johnston county trended 1% Democratic.
What's more, every other relatively large and growing urban area in North Carolina trended Democratic. Buncombe County (Asheville) trended 3 points to Obama, Forsyth County (Winston-Salem) trended 2 points to Obama, Guilford County (Greensboro) trended 3 points to Obama, Cumberland County (Fayetteville) trended 7 points to Obama, and New Hanover County (Wilmington) trended 2% Democratic.
In other words, anywhere where there is any population growth at all trended Democratic.
Meanwhile, the only parts of North Carolina which trended Republican are the same parts of North Carolina which are casting an ever-decreasing share of the statewide vote. In 2012, Republicans chained themselves more strongly than ever before to the declining demographic of white, rural, conservative voters - especially in Appalachia.
Democrats in North Carolina are a growing demographic, while Republicans in North Carolina are a declining demographic. And for precisely this reason, Democrats in North Carolina have excellent reason to be optimistic about their long term future.
Initial Evaluation of my Prediction
On the day before election day, I predicted:
Barack Obama will carry North Carolina's 15 electoral votes by a vote of 2,263,022 to 2,258,945, a margin of 4,077 votes. In percentage terms, that means Obama will win by 49.58% to 49.50%. 27,232 votes will be cast for Gary Johnson and 14,761 votes will be cast for write-in candidates. There will be 45,388 undervotes.
...
Obviously Obama's predicted margin of victory is FAR, FAR within the penumbra of uncertainty. As I see it, North Carolina is as much of a tossup as a tossup can possibly be. The primary measure by which I will judge the accuracy of this prediction by is the predicted vote margin of 4,077 votes for Barack Obama.
...
What should we look for on election night?:
1) If Obama wins early votes by about 150,000-200,000, he is in position to win NC, but it will be very close. If Obama wins early votes by more than 200,000 votes, though it will still probably be close. If Obama wins early votes by less than 150,000, then it will be tough (though not necessarily impossible) for him to win North Carolina.
2) If Obama is winning about 33% of white voters, he is in the position he needs to be to win NC, but it will be close. If he is much above that, it's looking good for an Obama victory in NC. If he is much below that, Romney will likely win NC.
Provisional ballots still need to be counted, but for the most part all the votes are in and we can evaluate my predictions. Here are the actual results, shown in the same kind of table:
And here is the difference between my prediction and the actual result - in order words, the error:
First of all, I predicted that 4,563,960 people would vote in the presidential race. That would have been a turnout increase of 5.9%, which is the same rate at which North Carolina's voting eligible population grew since 2008. However, turnout increased more slowly than that, and only 4,464,995 people voted - 98,965 fewer than I predicted. The fact that I had turnout too high helped to throw off many of my other numbers.
Secondly, I took all the undervotes away from election day rather than partly from election day and partly from early voting. That was a real facepalm.
Thirdly, I was worried that I was underestimating Obama's support among early voters, because polls seemed to be showing that Obama was doing better with early voters than my prediction system indicated. For this reason, I arbitrarily shifted 25,000 votes for Obama from election day to early vote, and shifted 25,000 votes for Romney from early vote to election day. As it turned out, my original early vote projection was pretty close to the actual results - but was actually (slightly) over-optimistic.
Fourthly, I overestimated election day turnout. We still don't know exactly who voted on election day, but it is probably a good bet that the people who I thought would show up to vote but who stayed home were Democratic-leaning sporadic, unlikely and new registrant voters.
Fifthly, I underestimated the swing of white voters to Romney. A two point swing among white voters was built into my prediction, through the mechanism of declining Democratic party registration and increasing Unaffiliated party registration. Exit polls initially showed that White voters were giving 33% of their support to Obama, which would have been high enough for a very close race. However, as results came in and the exit polls were revised, Obama's White support in the exit polls dropped to 31% - and that was low enough to push Romney over the top.