This lengthy blog post will review some of the 9th District's demographics and its 2016 and 2018 electoral performance, its current voter registration, its early voting leading up to Tuesday's general election, and then some final thoughts as to what I'll be looking for Tuesday evening.
As a reminder, the North Carolina 9th stretches from Bladen County and Cumberland County, with the city of Fayetteville, in the eastern portion of the state, westward to the city of Charlotte in Mecklenburg County.
The last time a Democrat held the seat was up to 1962, when Hugh Quincy Alexander lost his re-election bid to Republican Jim Broyhill in that year's mid-term; Republicans have held the different configurations of this district since 1963.
NC Ninth's Demographic Portrait
The Ninth District is an interesting mix of urban, suburban, and rural dynamics, along with demographic characteristics. The City of Charlotte, centered in Mecklenburg, has its southern part of the city, along with southeastern suburban areas in Mecklenburg, within the Ninth. Charlotte's suburban county of Union, right next door, is a very Republican stronghold, but as you move eastward, the district quickly becomes rural and diverse in its population. It ends in both rural Bladen County and the more urban Cumberland County and Fayetteville, which has a large military presence with Fort Bragg and Pope Air Force Base.
Demographically, the district's population is slightly less white than North Carolina as a whole, due to large segments of black/African-American and American Indian populations, located primarily in the rural counties. Based on 2017's population estimate from the U.S. Census Bureau, the NC 9th is 66 percent white, 18 percent black or African American, and 8 percent American Indian (Lumbee tribe).
U.S. Census estimates give the largest segment of civilian employment in the district to educational services/health care/social assistance, retail trade, manufacturing, and professional/scientific/management areas.
The educational diversity of the Ninth shows that over one-third of its residents are college graduates or have higher degrees:
And the annual payroll of the Ninth is dominated between manufacturing and health care, followed by finance and insurance, then retail trade:
NC Ninth's Recent Political DynamicsIn 2016, Donald Trump carried the district with 54.4 percent of the vote, winning it by 12 points over Hillary Clinton, who received 42.8 percent. This has been a typical presidential electoral performance in this version of the district, with Romney winning 55 percent and McCain winning 54 percent in 2012 and 2008, respectively.
What is also interesting is that presidential vote performance tends to have a high relationship to the Republican congressional candidate's performance, as evidence in this simple scatterplot that shows the Trump vote percentage in the district (X axis) to the GOP congressional candidate's vote performance in 2016 (Y axis):
R+8 in its Partisan Voting Index, meaning that it is 8 points more Republican than the national average.
In 2016, Republican Robert Pittenger performed slightly better than Trump, earning 58 percent of the district's vote to Democrat Christian Cano's 42 percent.
|2016 NC Ninth Congressional District General Election by County Results
|2016 NC Ninth Congressional District Republican Primary Election by County Results
Harris decided to re-challenge the first-term Pittenger in 2018's GOP primary, which led to the first Republican to not be renominated in that year's primary season. Harris won by a little over 800 votes.
|2018 NC Ninth Congressional District General Election by County Results
Ultimately, however, the 2018 election would never be certified due to electoral fraud allegations (and now indictments) in Bladen County regarding absentee by mail ballots, and the North Carolina State Board of Elections unanimously called for a new election, thus prompting this September 2019 "do-over" election.
In looking at the 2016 and 2018 general election results in the NC 9th, some patterns are evident that may be ones to watch in Tuesday's general election.
First, the county level results for the 2016 general election between Republican Pittenger and Democrat Cano:
Mecklenburg and Union counties provided 63 percent of the district's total vote, with Republicans winning 58 and 68 percent of each counties' respective totals. Rural Robeson county and urban Cumberland Counties came in third and fourth in the vote contributions, with the Democrat winning Robeson and the Republican edging out a victory in Cumberland. Overall, the Republican won the district with a 16 percentage point margin (58-42).
In 2018, however, the final margin was minuscule (905 votes), with notable shifts occurring in some key counties.
Again, both Mecklenburg and Union county provided nearly two-thirds of the district's total votes, but the final results showed a significant swing in vote performances. Mecklenburg went from 58 percent Republican to 54 percent Democratic, while Union dropped from its 68 percent Republican n 2016 to 59 percent Republican in 2018.
The vote differences and percentage differences are summarized here:
|North Carolina Ninth Congressional District County Differences in Votes from 2016 to 2018
|North Carolina Ninth Congressional District Differences in County Percentage Performances from 2016 to 2018
Democrat McCready was able to build on the 2016 vote performance in both Mecklenburg and Union counties, while Harris saw significant drops in support from 2016's performances and votes across all counties, most notably in Mecklenburg and Union counties. Now, this may be due to the 2018 Democratic "surge" in North Carolina, but the wave was not big enough to overcome the Republican partisanship due to Republican gerrymandering of the congressional districts.
In looking at the pool of eligible voters in 2018 (the district's voter registration), the party registration showed a slight advantage to registered Democrats (at 38 percent) to registered Republicans (at 31 percent). Registered unaffiliated voters were tied with registered Republicans.
However, one cannot read this as a "Democratic voter" advantage in the Ninth, due to the historic legacy of older, white, typically rural, more moderate to conservative voters having registered Democratic (because that's how their daddy and granddaddy registered and voted), but who are now much more likely to be voting Republican, despite their registration status.
In looking at who showed up in 2018's mid-term congressional year, the party registration breaks down as:
This Republican 'shift' in the electorate is normal in North Carolina, although 2018's election was rather unique, in that the state had no major state-wide contest, like a U.S. Senate or gubernatorial election. In calculating the state's and district's turnout rates for registered voters show subtle differences between the state-wide turnout rates, by race and party registration, versus who showed up in the Ninth:
White Republicans were slightly behind white Democrats state-wide, but the GOP had a slight advantage in the Ninth. Black/African-American voters in the Ninth mirrored their state-wide percentage turnout rates, while all other races in the Ninth were slightly below their state-wide turnout percentages.
2019's Early Voting in the NC Ninth
Currently, the Ninth Congressional District has a registered voter pool of 36 percent registered Democrat, 32 percent registered unaffiliated, 31 percent registered Republicans, and the remainder third parties (Libertarian, Constitution, and Green). Again, Mecklenburg and Union counties command the majority of registered voters, with 60 percent (a combined 300,000 active and inactive voters in those two counties).
Two of the eight counties have pluralities of registered Republicans, but as noted above, the rural counties (Anson, Bladen, Richmond, Robeson, and Scotland) with their registered Democratic percentages are skewed by the fact of minority populations in those counties (reliable Democratic voters, if they show up) and older white voters, who are likely more Republican voters than Democratic.
In North Carolina, especially in presidential years, voters tend to cast their ballots in the 'early voting' period by absentee ballots, most notably via the 'one-stop' method, or in-person with no excuse, or through absentee by mail (which was the cause of the fraud issues in 2018). With the impact of both Labor Day Monday and Hurricane Dorian, voting in some counties was extended to Saturday, for 18 days of in-person early voting (absentee by mail ballots will trickle in over the next several days, and if the election is close, this may be something to watch).
As of Sunday, September 8, a total of 81,547 absentee ballots (both in-person and by mail) have been accepted to be counted Tuesday evening (usually the first run of numbers we get when the polls close and the results begin to be posted).
In last year's contest, Democrat Dan McCready won the early vote (both absentee by mail and in-person), while Republican Mark Harris won on Election Day; this pattern tends to be typical in North Carolina elections:
What was not typical in 2018 was the fact that the majority of votes came in before Election Day. In North Carolina's mid-terms, voters tend to wait until Election Day, as opposed to presidential years. The following charts, starting with 2008, compare the presidential elections with the following mid-term elections and the vote methods used, along with the party registration within the vote methods for final ballots:
In 2008, 61 percent of the votes came in before Election Day, while in 2010, 63 percent were cast on Election Day.
Again, in 2012, 61 percent of the ballots cast came before Election Day, while 58 percent of 2014's ballots came on Election Day.
While the 2016 presidential election saw the typical pattern of 64 percent of the votes coming before Election Day, North Carolina witnessed, for the first time in a mid-term election, a majority (54 percent) of the ballots cast before Election Day 2018.
One other noticeable difference regarding 2018's election (which was a blue-moon election cycle with no U.S. Senate or gubernatorial contest state-wide in a mid-term) was that Democrats used an absentee by mail strategy that they had not done in previous mid-terms, or presidential years, for that matter.
It seems, with even 80,000 votes already 'in the bank' by early voting for this general election coning on Tuesday, the Ninth may revert back to a majority of votes coming in on Election Day.
This year, only two percent of the early votes have come in by mail, with the vast majority in-person (not surprising, considering the issues of absentee by mail fraud in 2018):
Absentee by mail is fairly evenly split among party registration, while registered Democrats have 'overperformed' in terms of in-person early votes, compared to their district registration. Registered Republicans overperformed slightly to their district registration, while unaffiliated voters underperformed.
In looking at the early votes by party registration within the counties:
Among the counties, Mecklenburg and Union counties are well overperforming their district percentages, with over seventy percent of the early votes coming in from just those two counties.
In comparison to 2018's early voting period, this year's early votes ended up at 56 percent of last year's total early votes.
Over the early vote period, registered Democrats tended to be above their 2018 baseline as a percentage of all early votes coming in, finishing one percentage point ahead of where they ended in 2018. Registered Republicans were one percentage point below their 2018 final percentage, while registered unaffiliated voters finished barely a point ahead of their 2018 final percentage.
The next two charts compare the actual numbers coming in with early votes (accepted absentee ballots) for 2018 and 2019:
Next, in looking at the two largest counties contributing early votes, Mecklenburg and Union, the following charts show the party registration breakdowns for the cumulative daily totals of early votes being accepted in both counties, with 2018 as a comparison.
Next is the registered voter turnout in just early voting, by county and for the district, separating out party registrations within each:
Registered Democrats again tended to have higher turnout rates in early voting, with Mecklenburg having a significant spike (over a quarter of registered Democrats have cast a ballot before Election Day), and 20 percent in Union County. The district-wide turnout rate following early voting is at 16 percent. A thought is that we should get at least that same amount of votes in on Election Day (unless something really weird happens and the majority of votes are already in), so that should put the turnout rate at 30 percent overall.
Next, I looked at the three key counties and broke down their turnout rates by both voter race and party registration within each racial category, first for Mecklenburg:
Next for Union County:
And finally, for Robeson County:
In Mecklenburg, thirty percent of white registered Democrats have already cast a ballot, while only 19 percent of white registered Republicans used early voting. In Union, 23 percent of white registered Democrats early voted, compared to 19 percent for white registered Republicans, and in Robeson, 15 percent of Black registered Democrats have already voted, compared to 13 percent white registered Democrats and 10 percent white registered Republicans.
The average age of early voters was 60 years old:
For comparison, the district's generational composition among registered voters shows that voters under the age of 38 (Millennials and Generation Z) are a plurality of the pool, with one-third of the registered voters. But in the pool of early voters, those two generations were only 11 percent of the total ballots accepted.
Finally, I looked at the 2019 early voters and how they may have voted (if they did, or not) in 2018. Of the 80,000 early voters, 93 percent of them voted in 2018: 80 percent of them voted 'early' in last November's election, while 13 percent voted on Election Day in 2018.
Here are the various vote methods, by numbers and by percentages, for each major party registration group within the district:
What I'll Be Watching For Tuesday Evening
Finally, here's a couple of thoughts as to what I'll (and maybe you) might want to watch for when returns start coming in on Tuesday evening.
1. Charlotte City Council Primaries
It's not just the Ninth Congressional District's general election happening on Tuesday, but also the primary elections for Charlotte City Council. With Charlotte having trended more and more Democratic, and the southern 'wedge' of Charlotte going from Republican safe-haven to competitive precincts, one aspect to watch may be the impact of the competitive Democratic primaries, especially for Charlotte at-large council, on the Ninth's race.
In looking at the Mecklenburg portion of the Ninth and the early vote, one can tell from the data which voters, especially unaffiliated voters, asked for a Democratic ballot for the primary and also voted in the Ninth:
Over half of the unaffiliated voters asked for the Democratic primary ballot, while only 17 percent of unaffiliated voters asked for the Republican primary ballot. That may have an impact on Mecklenburg's returns come Tuesday for the Ninth race.
2. The Mecklenburg-Union Battle...and What Other County Could Matter?
A couple of things to watch Tuesday evening when the returns come in for both Mecklenburg and Union counties.
In Mecklenburg, watch the southern Charlotte "GOP" wedge and the outlying suburban precincts in the southeastern corner of the county. This is best demonstrated by the changes happening in the area since 2008, using a modified Partisan Voting Index developed by Charlie Cook which averages a precinct's partisan voting pattern against the state's average for two elections.
First, PVI map for 2004 and 2008 shows the south Charlotte Republican 'wedge' as a baseline (the yellow lines are the city council districts, with the two bottom ones being the traditional GOP stronghold). Also notice the southeastern corner of the county, outside the yellow lines.
Next, the PVI map for 2008-2012, the Obama elections:
Notice the lack of deep reds, but the more 'toss-up' greens starting to be present in that southern portion of the city and county.
Finally, the PVI map for the 2012 and 2016 elections:
Beyond north Mecklenburg's significant transition (it appears) and the deepening blues across Charlotte, the real hold-outs of deep reds are in a pocket of southeast Mecklenburg, outside of Charlotte. The red 'wedge' of south Charlotte has pretty much collapsed by this point, with some precincts turning light blue and a significant number of them turning green, or competitive/toss-up.
This collapse has particular relevance to the next thing I'll be watching Tuesday evening: Union County. As a strong Republican county that is typical of surrounding suburban counties in North Carolina (Trump won 63-32 in 2016), Union County has become an increasingly Charlotte-based suburban area. If the suburbs are the potential battlegrounds in 2020, it will be worth watching if Dan Bishop can keep Mark Harris performance of 60 percent from 2018, or is there Republican slippage in that country, especially in the precincts on the Union-Mecklenburg county lines.
If it's a battle between Mecklenburg going Democratic and Union going Republican (but perhaps more slightly), then one of the rural counties could hold the final tally. One that I'll be watching is Robeson County, and especially black/African-American and American Indian turnout there.
Speaking of Dan Bishop's performance, there may be a more crucial area to watch come Tuesday evening.
3. Republican Dan Bishop's State Senate District in Mecklenburg
The Republican candidate in Tuesday's election is Dan Bishop, a current state senator from District 39 and a former Mecklenburg County commissioner and state representative. Early speculation was that with a Republican candidate with known roots and name recognition in the race, it might flip Mecklenburg back to the GOP column.
But Bishop was the primary author of House Bill 2, regarding local cities and the issue of public facilities and transgender citizens. The issue was headline news for several months, and Bishop is still seen as the primary architect of the policy.
His state senate district, number 39, is one of four in Mecklenburg:
Notice how District 39 captures so many precincts in south Charlotte, and then juts northeast into the suburbs along Union County. Based on the 2012-16 PVI map above, only the top corner of senate district 39 is still deep red.
This GOP collapse can also be looked at through election returns with Dan Bishop. In 2016, he garnered 56 percent the vote in a different district, one primarily based in the once Republican wedge of south Charlotte.
|North Carolina State Senate District 39 with Republican Dan Bishop in 2016 General Election
But in 2018, following a mandated redistricting by the U.S. Supreme Court, Bishop's district continued to include south Charlotte, but then made the north-eastern inclusion of suburban Mecklenburg County along the Union County line.
|North Carolina State Senate District 39 with Republican Dan Bishop in 2018 General Election
Again, district-wide, Bishop received 53 percent of the vote, but was the only Republican state legislative candidate to survive a Mecklenburg Democratic wave last November. And notice how many of the precincts flipped to blue, compared to 2016's general election.
In looking at the precincts that were in both the 2016 and 2018 general elections, Bishop experienced an average of a six percentage point drop in support from 2016 to 2018.
The greatest drop was nine in precinct 144 (going from 58.6 percent (out of 2,314 votes cast) to 49.6 percent (out of 2,112 votes cast)), with precinct 71 on the other end going down two (from 64 percent (out of 1,215 votes cast) to 62.3 percent (out of 1,115 votes cast)).
Bishop has closely aligned himself with President Trump, with the president coming to Fayetteville (in the eastern portion of the district) for a Monday rally. If the national narrative of white women, particularly those who are college-educated and suburban, are turning against the president, then it is likely that Bishop's senate district, one that is 44 percent white female, could send some signals of how the president's rhetoric and performance over the past few months is fairing, and how Republicans are either rewarded or punished for fully supporting the president.
If there is another Greenville NC rally type of atmosphere in Fayetteville, with chants of "send them back," then Tuesday's election may be a direct 'canary in the coal mine' experiment for 2020, not just in the NC Ninth, but also in the NC Third Congressional District, located on the coast of the state.