The 116th Congress is one-third over, but we’re still tying up the last loose end from the 2018 midterms. That’s tonight’s special election in North Carolina’s 9th Congressional District, where a do-over was ordered after a scheme to steal the election by operatives working for last year's Republican nominee was unmasked.
As Daily Kos Elections always does on big election nights, we’re offering county-level “benchmarks” that help you interpret what kind of numbers Democrat Dan McCready needs to hit to stay on track to win once returns start coming in after polls close at 7:30 PM ET.
We do this by using 2016 presidential election results, both in each county and districtwide, as our baseline. We then adjust county-level results so that they add up to the narrowest possible Democratic win districtwide. This helps viewers gauge how the election is going, even in the early phases of reporting, when only a few counties (possibly with very different leanings than the district as a whole) might make up a disproportionate share of who has reported so far.
The first column below gives the name of each county (“pt.” means only part of the county is in the district); the second column shows what percentage of the vote total each county provided three years ago, so that you can assess which are the important ones to track. The third column shows the 2016 presidential results in each county, and the fourth column shows the adjusted results McCready needs to get over the line. (There are Libertarian and Green candidates on the ballot, who may eke out 1% or so apiece, so we’re modeling toward a 49-48 win for McCready.)
COUNTY |
% OF VOTE |
2016 RESULTS (D/R) |
WHAT MCCREADY NEEDS TO WIN |
DISTRICTWIDE |
100 |
43/54 |
49/48 |
MECKLENBURG (PT.) |
31.8 |
47/50 |
53/44 |
UNION |
30.4 |
33/64 |
39/58 |
ROBESON |
11.8 |
47/51 |
53/45 |
CUMBERLAND (PT.) |
10.0 |
48/49 |
54/43 |
RICHMOND |
5.6 |
44/54 |
50/48 |
SCOTLAND |
4.0 |
53/45 |
59/39 |
BLADEN (PT.) |
3.4 |
39/60 |
45/54 |
ANSON |
3.1 |
56/43 |
62/37 |
On paper, this looks like a pretty difficult district for McCready, and it is: Hillary Clinton lost here by a 54-43 margin in 2016. While the 9th isn’t one of those rural districts that swung precipitously in the GOP direction in 2016, it also isn’t one of those once-much-redder suburban districts that took a turn toward the Democrats in 2016; Barack Obama, in 2012, turned in an almost identical performance, losing to Mitt Romney 55-44 under the current boundaries.
A more positive sign, though, is that McCready significantly outperformed Clinton’s numbers last year. There are several reasons for that: One is that there was no Republican incumbent here in 2018, because Harris defeated the more establishment-flavored congressman, Robert Pittenger, in the GOP primary. Another is that McCready has earned a reputation as a moderate, which may give him some additional appeal to swing voters who might not have considered Clinton.
The main reason, though, is the same factor that applied nationwide throughout the 2018 elections, and has in fact applied to nearly every first presidential midterm: the harsh nature of reflexive fluctuations in public opinion, where supporters of the party in the White House get complacent or disillusioned, partisans in the minority get enraged, and swing voters swiftly turn on the same president that they just elected, seeking to balance him with a Congress run by the opposite party.
The fact that last year’s voided results showed McCready trailing by only a very narrow margin, in fact, gives us the opportunity to use a second set of numbers — the actual county-by-county numbers from the 2018 election, which, conveniently, were divided by just a 49.3-48.9 spread. These figures can serve as a sanity check on the adjusted benchmarks from above. Since many of the actual 2018 numbers, at least in the largest counties, are nearly identical to the 2016-based benchmarks, that would seem to confirm the model.
COUNTY |
% OF VOTE |
2018 RESULTS (D/R) |
DISTRICTWIDE |
100 |
49/49 |
MECKLENBURG (PT.) |
34.0 |
54/44 |
UNION |
30.9 |
39/59 |
ROBESON |
10.9 |
56/41 |
CUMBERLAND (PT.) |
9.3 |
51/47 |
RICHMOND |
5.0
|
50/48 |
SCOTLAND |
3.7 |
56/42 |
BLADEN (Pt.) |
3.3 |
41/57 |
ANSON |
2.8 |
58/41 |
Where the numbers deviate a bit, you can see where McCready either outperformed Clinton (most notably in Robeson County, in the district’s east), or underperformed them relative to the benchmarks in the first table (most notably in the sparsely populated rural counties in the district’s middle, such as Richmond).
In fact, those patterns tell us a lot about the demographics of these parts of the district. Robeson County is prime “ancestral Democrat” territory, where Democrats still have a registration advantage and where they’re still successful at electing lower-level officials such as state legislators (who can differentiate themselves from the national party better than candidates in federal races can).
On the federal level, though, they’re facing diminishing returns, with Clinton losing here despite Democrats carrying the county in every presidential race from 1972 to 2012 (even Walter Mondale!). Robeson County is fairly unusual in that it has a Native American plurality of 39%; 26% are non-Hispanic white and 24% are black. However, the local Native American population—mostly members of the Lumbee tribe—is largely evangelical Christian and has few college graduates. As a result of growing social conservatism, this part of the state has begun behaving more like the rest of the rural South.
Meanwhile, some of the smaller counties in the middle of the district have larger African American populations. Anson County—the closest this district has to a solidly blue county—is a case in point, at 45% white and 49% black. These counties, though, have low turnout, especially in midterms, as seen in how they made up a notably smaller percentage of the overall vote total in 2018 than they did in 2016. McCready’s challenge in the small counties will therefore simply be getting voters to show up, even at 2018 levels.
However, while a map might make the 9th District look like it’s mostly rural, in fact, two-thirds of the district’s population lives in the Charlotte metropolitan area, in Mecklenburg and Union Counties in the district’s western corner.
Union County is primarily exurban and is still pretty reflexively dark red. Mecklenburg, however, is where Charlotte is located and is the state’s most populous county. Mecklenburg as a whole is solidly Democratic (countywide, Clinton won 62-33), but the southeastern portion of Mecklenburg that’s gerrymandered into the 9th includes the whitest and most affluent portion of Charlotte and its close-in suburbs such as Mint Hill. Mecklenburg as a whole is 48% white and 44% college graduates, but the portion of Mecklenburg that’s in the 9th is 80% white and 62% college graduates.
Interestingly, that’s the portion of the district that Bishop is from. As that part of the Charlotte area starts to vote the way other affluent and well-educated, white-majority parts of the Sun Belt’s suburbs do (like, say, Georgia’s 6th District in Atlanta’s suburbs or Texas’s 7th in Houston’s suburbs), it’s an open question as to whether someone like Bishop could even get elected there in his same legislative district going forward. Bishop could well still win districtwide tonight, though, thanks more to Republican strength further out in Union County.
As with most House-level races, polling has been spotty, mostly taking the form of leaked internals; the one recent nonpartisan poll, from late August, found McCready leading Bishop 46-42, though even Democratic polling has reportedly found a closer race than that. All prognosticators who’ve weighed in consider this race to be a tossup. In other words, it’s one of those rarest of beasts: a House special election where there’s no clear favorite.