State of the 2020 Election: North Carolina
Big Picture
Donald Trump- 2,362,631 (49.83 percent, down 0.56 percent from Romney in 2012)
Hillary Clinton- 2,189,316 (46.17 percent, down 2.18 percent from Obama in 2012)
Voter Turnout- 68.44 percent (up 0.7 percent from 2012)
Similar to other states in the South, North Carolina has been considered reliably Republican for several decades. After Jimmy Carter won it in 1976, the state proceeded to back the Republican in the subsequent seven presidential elections, even staving off two competitive finishes from southern Democrat Bill Clinton. Barack Obama broke the streak in 2008 when he captured North Carolina by a mere 14,000 votes. Since then, however, it has supported Republicans.
Ultimately, North Carolina is a state Mitt Romney won in 2012 that Trump was able to retain four years later. While turnout only increased by 0.7 percent from 2012, this actually translated to a staggering 233,463 new voters. Despite performing over 2 percentage points worse from President Obama, Clinton actually received 11,164 votes more votes than him. This speaks to North Carolina’s notable population growth. Since 1880, it has reported double-digit population growth in every decennial Census. Even now, North Carolina continues to outpace the national average in population growth and it is projected to gain a new House seat following the 2020 reapportionment. Considering this, the state has become more competitive in recent decades.
Normally, I’d talk about the 2018 midterms by describing the results of the governor and/or Senate race that happened there. But as it happened, North Carolina didn’t have a race for either office in 2018. On the Senate side, this is simply due to the way its Senate terms are staggered and 2018 happened to be a year where none of its seats were up for election. And on the governor’s side, North Carolina is one of fourteen states to hold its governor’s election during presidential years. For context, in 2016, Democratic challenger Deborah K. Ross lost the Senate race to incumbent Republican Richard Burr while Democratic challenger Roy Cooper defeated incumbent governor Pat McCrory by only 10,277 votes following a several-week recount.
That leaves the results of the 2018 House races in North Carolina’s thirteen congressional districts. This is an okay indicator of electoral performance in the post-Trump “era” (I have an axe to grind with that terminology), but I’m hesitant to talk about House results within individual counties for two reasons.
One is that in the context of House races, the subsections of counties I lay out are a pretty artificial and unhelpful way to think of the results. Whereas presidential, Senatorial, and gubernatorial races are statewide and relatively easy to dis-aggregate, House races are intended to be fought at the congressional district level. And while you can break down individual districts by the counties they inhabit, these subsections are non-contiguous and stretch across multiple districts, each with very different partisan leans and incumbency advantages. This makes the discussion both overly complicated and convoluted.
And the other reason is gerrymandering. Following the 2010 redistricting, North Carolina became one of the most gerrymandered states in the country. In the 2010 midterms, the last to use the previous map, Democrats lost one House seat and the combined statewide vote with 45.3 percent, but still controlled seven of the thirteen seats. Two years later, with the new map drawn by the Republican-controlled legislature, Democrats lost three more seats despite winning the statewide vote with 50.6 percent. North Carolina received multiple legal challenges for its map on counts of gerrymandering, both on a partisan level and on a racial level. Regardless, the fact that House districts are routinely redrawn makes it difficult to assess counties because those counties’ voting behavior can be heavily influenced by incumbency advantage or factors pertaining to other counties within given districts. While statewide races can also be subject to incumbency advantage, the status of being a statewide race and the continuity of county borders simulate a more controlled environment than House races.
So for these reasons, I won’t be dwelling much on the House results for the remainder of this article. Instead, I’ll use Roy Cooper’s 2016 gubernatorial victory in place of the other results. For context, though, in 2018, Democrats lost the statewide vote with 48.4 percent; however, they were able to retain all three of their seats.
With that disclaimer out of the way, let’s discuss the possible pathways for the Biden campaign to recapture the Tar Heel state.
Solid Democratic Counties
Anson (southern part of state, on South Carolina border)
Bertie (eastern part of state)
Cumberland (Fayetteville city proper and surrounding communities)
Durham (Durham and Chapel Hill city proper, Research Triangle)
Edgecombe (Rocky Mount city proper, shared with Nash County)
Guilford (Greensboro city proper, Research Triangle)
Halifax (northern part of state)
Hertford (northeastern part of state, on Virginia border)
Mecklenberg (Charlotte city proper and surrounding communities)
Northampton (northern part of state, on Virginia border)
Orange (Research Triangle)
Vance (Raleigh metropolitan area)
Wake (Raleigh city proper, Research Triangle)
Warren (northern part of state, on Virginia border)
Washington (eastern part of state)
First, there’s the solidly Democratic counties. For the purposes of this list, I decided to expand the category by lowering the Democratic vote share threshold from 60 percent to 55 percent. I did this because I think there are some prominent counties in North Carolina with a lot of voters that fell in the 55–60 percent range, too solid to be considered “swing” but also not solid enough to fall under the old categorization of “solid”. With that in mind, there are fifteen counties in this subsection that accounted for 37 percent of all votes cast statewide in 2016 and almost half of all the votes Clinton received from North Carolina.
A good chunk of this subsection is urban. Seven of the state’s ten largest cities are in this subsection. Since the 2000s, Charlotte has been one of the fastest-growing major cities in the United States; in the last decade, its population grew by more than 50 percent. A lot of this has to do with a surge in new jobs, especially in banking and tech, but it also boasts a reputation for being more affordable than other major cities with comparable job prospects such as New York and San Francisco. Today, Charlotte is the second-largest city in the South (if we exclude Texas, of course) and a heavy hitter for Democratic candidates.
Another major area in this subsection is Raleigh and the Research Triangle. Spread out across several counties in the central part of the state, the Research Triangle is defined by its high concentration of prestigious universities (such as Duke, NC State, UNC at Chapel Hill, and Wake Forest), major cities (including Raleigh, Durham, Chapel Hill, Greensboro, and Winston-Salem), and employers. The Triangle has a thriving economy with thousands of jobs from hospitals, tech companies, research and development firms, universities, and the state government. And given its strong mixture of urban, suburban, and highly-educated voters, it also fields a healthy competitive edge for Democratic candidates.
Outside the major metropolitan areas, this subsection includes predominantly black suburbs in the more rural sections of the state, particularly along the Virginia border in the north and the South Carolina border in the south. These include Anson, Bertie, Halifax, and several others. A lot of these counties used to be driven by their tobacco and cotton plantations, which (unfortunately) relied on slavery. Nowadays, the warm climate does attract retirees; however, these counties remain mostly black.
In 2016, Clinton won 61.6 percent of the vote in this subsection with a turnout rate of about 69 percent (on par with the state average). This is a 1-point improvement from President Obama’s performance in 2012 and of the 75,490 new voters that didn’t vote in 2012, Clinton won 62,678 of them, which is pretty impressive; however, it won’t be enough to win the election on its own.
Demographically, it’s not difficult to see where potential Democratic supporters come from. In 2016, 40.5 percent of the subsection’s population was nonwhite (higher than the statewide average of 29 percent) and 10.4 percent was Hispanic (slightly higher than the statewide average of 9.2 percent). This in itself is a great display of diversity, but what’s interesting is that while most of the nonwhite voters in the subsection come from the major urban areas, the rural counties actually have a higher percentage of nonwhite residents than these areas. Nonwhites only make up 41.9 percent of Mecklenberg’s population and 31.4 percent of Wake’s, but there are nine counties in this subsection where nonwhites make up more than 50 percent of their population.
Regarding educational attainment, this subsection is a tale of two extremes. On one end, the top four counties in the state for adult residents with at least a bachelor’s degree are represented (Orange, Wake, Durham, and Mecklenberg). Three of them are in the Research Triangle while the fourth covers the city of Charlotte. In all four of these counties, bachelor’s degree attainment is over 40 percent, placing them above the statewide average of 30.5 percent. On the other hand, nine counties are in the bottom third in the state (and four of them are in the bottom ten). In each of these counties, less than 15 percent of adult residents hold a bachelor’s degree. It should be mentioned that a lot of these counties that have few college graduates are those that more rural, historically agricultural, and predominantly black.
This educational divide also translates to economic indicators. The top four counties for educational attainment were also in the top ten for per capita income in 2016, with Orange being first in the state at $60,044. They also enjoy fairly low unemployment (with the highest being Mecklenberg, ranked 22nd lowest in the state at 4.7 percent) and low poverty (with the highest being Durham, ranked 39th lowest in the state at 16.1 percent). Conversely, many of the other counties in this subsection struggle with these indicators. In 2016, nine counties each had a per capita income below $35,000, an unemployment rate above 6 percent, and a poverty rate above 20 percent. Warren County, for example, has the second lowest per capita in North Carolina ($28,793 in 2016), the tenth highest unemployment rate (7.3 percent), and the fifth highest poverty rate (26.4 percent).
When put altogether, this subsection reflects a high degree of diversity across race, ethnicity, education, and socioeconomic status. Despite this, there isn’t a lot of middle ground relative to the rest of the state, especially regarding education and socioeconomic status; the counties included are either near the top of the state or near the bottom. Luckily for the Biden campaign, the mixture of demographics is highly favorable to Democrats. During the 2016 governor’s race, Roy Cooper outperformed Clinton in this subsection by winning all fifteen counties with 63.2 percent of the vote. While this doesn’t sound much, marginal gains matter a lot in a state as competitive as North Carolina.
And as of 2020, this subsection has 37.2 percent of the state’s registered voters, an enormous share of the electorate. When broken down by party, registered Democrats in this subsection outnumber registered Republicans with 43.8 percent versus 21 percent. Granted, this margin is down from previous elections (in 2016, for example, registered Democrats were up 46.1 percent versus 22.8 percent); however, this is still an impressive advantage and one that will be critical for the Biden campaign this November.
Overall, the Biden campaign will continue to need to perform well in these counties to win statewide. Ideally, it should push to win these counties by a larger margin, particularly Wake and Guilford counties, where Clinton received less than 60 percent of the vote. There are clear opportunities for Biden to get these votes, particularly from suburban voters, college-educated voters, and nonwhite voters.
Swing Counties
Buncombe (Asheville city proper and surrounding communities)
Chatham (Research Triangle)
Forsyth (Winston-Salem city proper, Research Triangle)
Granville (northern part of state, on Virginia border)
Hoke (Fayetteville metropolitan area)
Lenoir (eastern part of state)
Martin (eastern part of state)
Nash (Rocky Mount city proper, shared with Edgecombe County)
New Hanover (Wilmington city proper and surrounding communities)
Pasquotank (northeastern part of state)
Pitt (Greenville city proper and surrounding communities)
Robeson (southern part of state, on South Carolina border)
Scotland (southern part of state, on South Carolina border)
Watauga (northwestern part of state, on Tennessee border)
Wilson (eastern part of state)
In addition to there being fifteen solidly Democratic counties, there are fifteen swing counties in North Carolina this time around. In 2012, President Obama won twelve of these counties with 52.6 percent of the vote and turnout of about 67.4 percent. In fact, applying the classifications to the 2012 results, five of these counties would not have been considered “swing” counties in 2012. But in 2016, Clinton only won nine of these counties with 50.7 percent of the vote and comparable turnout. This translates to a loss of 1,673 votes in a subsection where 26,172 more ballots were cast than four years ago. While it should be noted that Clinton didn’t lose too many raw votes in these counties, this is overshadowed by the fact that there were far more voters that could have been won, making this a missed opportunity for the campaign.
This subsection is an interesting mixture between rural counties and medium-sized cities. These cities include Asheville (a popular tourist destination in the mountainous western part of the state known for its live music), Fayetteville (located near Fort Bragg, one of the world’s largest military installations), Greenville (a central hub in the state’s Coastal Plains region), Rocky Mount (a smaller city with a thriving manufacturing base), Wilmington (a coastal city with a prominent port), and Winston-Salem (home to Wake Forest University and near the Research Triangle).
On the other hand, there are plenty of rural counties that are considered swing. Most of these are concentrated in the eastern half of the state. For the most part, these counties are pretty diverse racially, with a strong mixture of white, black, and Native American residents (especially in Robeson County). Considering this geographical diversity, winning this subsection requires being able to appeal to both urban and rural sensibilities.
Demographically, this subsection doesn’t have the same mixture of nonwhite or Hispanic voters as the solidly Democratic counties. In 2016, 32.7 percent of the population was nonwhite and 8.2 percent was Hispanic. This is still slightly above the state average, but it doesn’t match the same level as the other subsection. Outside the cities, the rural counties have a more even split between whites and nonwhites than the rural counties in the solidly Democratic subsection. As it result, this requires a more bifurcated strategy to appeal to these distinct demographics.
Similar to the solidly Democratic counties, this subsection spans the entire spectrum of educational attainment. For instance, four counties are in the top ten statewide and two of them are in the bottom twenty (not in the bottom ten, though). But whereas the solid counties are concentrated solely on the extremes with little in the middle, many counties in this subsection are in the middle of the spectrum, meaning there’s less extremes. At least 10 percent of the adult population in every county has a bachelor’s degree or higher and six counties are above the state average.
This range in education also translates to socioeconomic outcomes, where this subsection stretches the entire range of outcomes. Highly-educated counties, such as Chatham, perform exceptionally well, ranking second highest in per capita income ($58,600 in 2016), third lowest in unemployment rate (4.2 percent), and tenth lowest in poverty rate (11.7 percent). Other counties either have an average or mixed combination of outcomes. Watauga is one example of this, where despite having the fifth highest concentration of college graduates (41.6 percent), it ranks 62nd in per capita income ($34,244 in 2016), tenth lowest in unemployment rate (4.5 percent), and 13th highest in poverty rate (24.3 percent). And some counties, such as Robeson, struggle a lot socioeconomically. It’s ranked 87th in adults with a bachelor’s degree or higher (at only 13.2 percent), which translates to the seventh highest unemployment rate in the state (7.9 percent in 2016), the lowest per capita income ($27,604 in 2016), and the highest poverty rate (a staggering 27.8 percent).
Overall, this subsection reflects a lot of the geographic, economic, and demographic diversity demonstrated in the solid Democratic counties; however, this subsection has a less stark divide within that diversity by including plenty of counties in the middle of the state’s spectrum. And despite Clinton’s slight underperformance in this subsection, this region still remains in reach for Democratic candidates. The same day voters went to choose between Clinton and Trump, Cooper won twelve of these counties with 53.8 percent of the vote. And in three of those counties, Cooper received 55 percent or more of the vote share, moving them from the swing category to the solidly Democratic category. Similar to that of the solidly Democratic counties, small improvements matter a lot in a competitive state and the Biden campaign should recognize as many opportunities to pick up votes as possible.
As of 2020, registered Democrats continue to outnumber registered Republicans 40.3 percent to 25.8 percent; however, this margin is down from 2016 (46.1 percent versus 22.8 percent) and 2012 (47.7 percent versus 27 percent). On some level, this reflects a decrease in partisan registration altogether as more voters choose to register as independent or “no party”; however, it also demonstrates how despite the numerical advantage, Democrats do not have a solid lock on these counties. Therefore, it’ll be important for the Biden campaign to target these counties and win them over through his policies and his criticism of the Trump administration.
Flipped Counties
Bladen (Obama-Trump, southeastern part of state)
Gates (Obama-Trump, Virginia Beach/Norfolk metropolitan area, on Virginia border)
Granville (Obama-Trump, northern part of state, on Virginia border)
Martin (Obama-Trump, eastern part of state)
Nash (Obama-Trump, Rocky Mount city proper, shared with Edgecombe County)
Richmond (Obama-Trump, southern part of state, on South Carolina border)
Robeson (Obama-Trump, southern part of state, on South Carolina border)
Watauga (Romney-Clinton, northwestern part of state, on Tennessee border)
Finally, there are eight states that flipped to the other party between 2012 and 2016. Seven of them are Obama-Trump counties, a grouping that’s been discussed in previous posts for other states. Remarkably, one of them is a Romney-Clinton county, making it part of a trend that received relatively little attention following the election, but provide some key insights into the race.
Five of these counties overlap with the swing counties. Among these overlaps include Robeson, which was discussed earlier as one of the poorest counties in the state, and Nash, which shares the city of Rocky Mount with Edgecombe, a solidly Democratic county. This demonstrates that at least some of these areas are pretty competitive, but for the most part, these counties are fairly rural and tend to congregate near the state borders. One example is Gates County which is close to Virginia Beach, which has gone Republican in every presidential election since 1968 despite being the largest city in its respective state.
In 2016, Clinton only received 46.9 percent of the vote in the seven Obama-Trump counties, down 5.9 points from Obama’s performance in 2012. This translates to about 11,321 fewer votes in a subsection where turnout (both in percentage and raw numbers) closely matched that of four years prior. As for Watauga, the state’ only Romney-Clinton county, Clinton only improved President Obama’s vote share by 0.17 percentage points. This translates to 1,136 more votes.
Demographically, there’s a distinct character between Watauga and the seven Obama-Trump counties. Despite overlapping with the swing counties, Watauga is the least racially diverse county within that subsection with only 5 percent of its population being nonwhite and 3.4 percent being Hispanic. On the other hand, almost half of their combined population of the Obama-Trump counties is nonwhite and 7.2 percent is Hispanic; each of them has a nonwhite population greater than 35 percent. Considering these factors, one wouldn’t expect these counties to swing so far to the right; however, one possible explanation is that the white voters in these counties went overwhelmingly for Trump in 2016. According to exit polls, Trump won North Carolina white voters by 31 points, more than his national margin of 21 points.
And when broken down further by educational attainment, it isn’t super difficult to see how a candidate like Trump can find serious inroads in these counties. Five of the seven Obama-Trump counties are in the bottom half of the state by percentage of adults with at least a bachelor’s degree. While none of these counties are in the bottom ten, none of them have this figure above 23 percent. Among North Carolina’s white voters without a college degree, Trump won by 44 points, which outpaces his already impressive national margin of 39 points. Considering this, Trump was able to overcome a deficit from a large black and non-white population with a dominant victory among white voters. Conversely, Watauga is in the top five statewide for educational attainment, where over 40 percent of its adults have a college degree. And although Clinton didn’t receive a plurality vote, she still managed to win this mostly white county.
And then there’s the socioeconomic situation. I already mentioned Watauga County as having a mixed economic situation: slightly below average per capita income, relatively high poverty, and relatively low unemployment. As for the Obama-Trump counties, they do cover a fairly broad stretch of the spectrum, but on the whole, they skew towards the bottom half in the state for the three indicators. Robeson, another county that overlaps with the swing counties, has the lowest per capita income, highest poverty rate, and one of the highest unemployment rates in the state. While the other Obama-Trump counties aren’t approaching that extreme, they also aren’t performing well on any of the indicators.
Perhaps the craziest fact is that despite not having a consistent history of supporting Democratic candidates in presidential elections, the seven Obama-Trump counties actually have more registered Democrats than registered Republicans. And it isn’t that close either. In 2016, registered Democrats outnumbered registered Republicans by almost 40 percentage points (57.8 percent versus 20 percent). And even now, while registered Democrats have gone down, they still have a massive advantage in 2020 (49.6 percent versus 22.5 percent). Once again, this is another strong advantage for Democrats heading into any general election; however, lack of enthusiasm for Clinton and Trump’s ability to win among other demographics in this subsection neutralized that advantage. Even so, this breakdown is definitely something to keep an eye on during this election.
During the 2016 gubernatorial race, Roy Cooper only won two of the Obama counties with 48.5 percent (a two-point improvement over Clinton) and Watauga County with 52.6 percent of the vote (a 5.5-point improvement). While one can interpret that to mean that it’s possible to win statewide without these Obama-Trump counties, it also makes it a lot more difficult. Cooper had to perform several percentage points better than Clinton in this subsection and everywhere else in the state just to win by the narrowest of margins. For the Biden campaign, it should strive to do more so that it doesn’t come down to the wire.
Ideally, the Biden campaign should work to win back a lot of the Obama-Trump voters that Clinton couldn’t capture in 2016. While there’s some evidence that Trump is making marginal gains among black voters, the salience of race issues in this year’s election could be a strong motivating factor for voters in this subsection. The Biden campaign should work to maximize black turnout in these counties, which will go a long way towards winning back these counties. White voters in these areas may be more difficult to win over; however, considering the relatively high level of poverty in these counties, Biden can promise these voters a larger economic response to the pandemic through additional stimulus packages. It may not work for most of them; however, even if he can pick up some independents and disaffected Trump supporters, this will help Biden at the margins.
Conclusion
At the end of the day, it’s important to remember that North Carolina still has fairly strong conservative impulses. For example, Trump won the state’s suburban voters by 24 points in 2016, more than the 4-point margin he enjoyed nationally. In another example, Clinton only won the state’s college graduates by 1 point, down from her 10-point victory nationally. While North Carolina may be a state in demographic transition, that doesn’t erase the fact that it has only backed one Democratic presidential candidate since 1976 or the fact that Republicans continue to control both of its Senate seats, nine of its thirteen House seats, and both chambers of its state legislature.
Joe Biden is definitely in a position to improve upon Clinton’s 2016 performance and there’s a good likelihood that the combination of demographics and political shifts will be enough to carry him the state. But it’s important to remember that for a Democrat to win a state that typically votes Republican, his campaign will really need to work these voters in the last few weeks to ensure that Trump doesn’t allow himself to be seen as a “lesser of two evils”.
So that’s all I have for North Carolina. In my penultimate post, I plan to move further south to discuss Georgia.
