One week ahead of the closely-watched re-do election in south-central North Carolina’s 9th congressional district, a new poll of 500 likely voters from RRH Elections shows the race very tight. State Sen. Dan Bishop (R) edges businessman and 2018 general election nominee Dan McCready (D) by a margin of 46%-45%, well within the poll’s 4% margin of error. Libertarian Jeff Scott and Green Party candidate Allen Smith trail with 2% and 1% respectively, and 5% remain undecided. The survey also shows that Trump’s approval rating is just barely positive in the conservative-leaning district. In hypothetical 2020 matchups, Trump holds a modest lead over Joe Biden in the presidential race, while LG Dan Forest (R) edges Gov. Roy Cooper (D) in the gubernatorial race in the district. The poll of 500 likely voters was conducted by RRH Elections through landline and cell phone calls, and has a margin of error of 4%. RRH Elections is solely responsible for data analysis for this survey, including demographic weighting. The survey results were weighted by gender and geographic location only. RRH Elections is a group of volunteer hobbyist bloggers that has no connection to any candidate or organization active in this race. This poll was entirely crowd-funded by donations from the readers of our blog. Please visit us at http://rrhelections.com.
Flip over for full results and crosstabs!
Topline Results:
| 2019 Congressional Rerun Election | |
| Dan Bishop (R) | 46% |
| Dan McCready (D) | 45% |
| Jeff Scott (L) | 2% |
| Allen Smith (G) | 1% |
| Undecided | 5% |
| Trump Approval | |
| Approve | 47% |
| Disapprove | 45% |
| No Opinion | 8% |
| 2020 Presidential | |
| Donald Trump (R) | 50% |
| Joe Biden (D) | 44% |
| Undecided | 6% |
| 2020 Gubernatorial | |
| Dan Forest (R) | 45% |
| Roy Cooper (D) | 44% |
| Undecided | 10% |
Demographics & Crosstabs:
| Voter Status | Total | Bishop (R) | McCready (D) | Others | Undecided |
| Already Voted | 7% | 48% | 41% | 11% | N/A |
| Certain to Vote | 71% | 47% | 44% | 4% | 5% |
| Likely to Vote | 22% | 44% | 47% | 3% | 6% |
| Trump Approval: | Bishop (R) | McCready (D) | Others | Undecided |
| Approve (47%) | 90% | 7% | 1% | 2% |
| Disapprove (45%) | 5% | 89% | 2% | 4% |
| No Opinion (8%) | 25% | 14% | 25% | 36% |
| Gender: | Total | Bishop (R) | McCready (D) | Others | Undecided |
| Male | 48% | 47% | 42% | 4% | 8% |
| Female | 52% | 46% | 47% | 3% | 3% |
| Race: | Total | Bishop (R) | McCready (D) | Others | Undecided |
| White | 73% | 47% | 44% | 3% | 6% |
| Black | 22% | 42% | 46% | 6% | 6% |
| Hispanic/Other | 5% | 55% | 45% | 0% | 0% |
| Age: | Total | Bishop (R) | McCready (D) | Others | Undecided |
| 18-29 | 8% | 45% | 49% | 3% | 3% |
| 30-49 | 30% | 50% | 46% | 2% | 6% |
| 50-69 | 41% | 45% | 46% | 4% | 5% |
| 70+ | 21% | 45% | 45% | 5% | 6% |
| Region/County: | Total | Bishop (R) | McCready (D) | Others | Undecided |
| Cumberland | 8% | 53% | 39% | 0% | 7% |
| Mecklenburg | 36% | 42% | 48% | 3% | 7% |
| Robeson | 9% | 45% | 42% | 13% | 0% |
| Union | 29% | 51% | 42% | 2% | 5% |
| Others (Anson, Bladen, Richmond, Scotland) | 17% | 45% | 46% | 2% | 1% |
Methodology:
All survey design and data analysis was conducted by RRH Elections. GAJ Solutions placed live calls to cell and landline phones of likely voters in North Carolina’s 9th Congressional District on August 26-28, 2019, yielding a survey of 500 complete responses. Those who attested that they were unlikely to vote in the special rerun election were not included in the survey. The poll was weighted by gender and geography (county) only; random deletion was not used as a method of weighting.
Discussion:
The rerun election in North Carolina’s 9th congressional district, stretching from Charlotte through its southeastern suburbs and to the rural sandhills in the south-central part of the state, has been closely-watched nationally, and our new RRH Elections poll shows that the focus on the race is justified. Republican State Senator Dan Bishop edges out businessman Dan McCready, who was the Democratic nominee in the invalidated 2018 general election for this district, by a margin of just 46-45, well within the poll’s 4% margin of error. Two minor candidates are in the low single digits, while 6% remain undecided. The key divide here is on partisanship and support of Trump, as Bishop gets an overwhelming share of those who approve of the President while McCready takes a similarly overwhelming share of those who disapprove. Aside from that cleavage, the race is highly notable for how similarly Bishop and McCready are performing in this survey across a wide variety of demographic groups. Bishop seems to take an implausibly high share of the votes of Black voters, a result that is likely due to sampling error in the small sample size for that demographic subgroup. Bishop and McCready keep the race close across the entire district, with no particularly strong divide between the urban and suburban western part of the seat and the more rural east. They also keep the race close among both men and women. Looking ahead to the 2020 races, President Trump leads Joe Biden by a modest 6-point margin in the district, less than the margin he won by three years ago in this seat. In the gubernatorial race, Republican Lieutenant Governor Dan Forest and incumbent Democratic Governor Roy Cooper are essentially tied in this district, with Forest edging ahead by a single point. As this seat is to the right of the state, these results could be a hopeful sign for Democrats looking to carry the statewide races in North Carolina in 2020. For next week’s special rerun election race, however, much will depend on which side is more effective at getting its voters to turn out, and this race remains likely to be decided by a very small margin.
Poll Script:
There is a special election on September 10th for the US
Congress. Are you likely to vote in it?
Already Voted Early/Absentee
Certain to vote
Likely to vote
Which of the following candidates for Congress in the
September 10th special election are you most likely to vote for?
Republican Dan Bishop
Democrat Dan McCready
Libertarian Jeff Scott
Green Party candidate Allen Smith
I am totally undecided
Do you generally approve or disapprove of President Trump’s
performance in office?
Approve
Disapprove
No opinion
Thinking ahead to the 2020 election, if the presidential
election were between Donald Trump and Joe Biden, who would you be more likely
to vote for?
Republican Donald Trump
Democrat Joe Biden
I would be totally undecided
Thinking ahead to the 2020 election, if the gubernatorial
election were between Democratic Governor Roy Cooper and Republican Lieutenant
Governor Dan Forest, who would you be more likely to vote for?
Republican Dan Forest
Democrat Roy Cooper
I would be totally undecided
What is your gender?
Male
Female
What is your race?
White
Black/African-American
Other race
What is your age bracket?
18-29
30-49
50-69
70 and older
About RRH Elections:
RRH Elections is a Republican-leaning Elections Blog that discusses electoral politics, not policy, with particular attention to congressional and state races. We cover electoral news daily and provide insight into the electoral landscape. Using our in-house expertise, we have produced ten polls, eight of which have predicted the margin in their races to within four points. RRH Elections is run by a team of 9 unpaid hobbyists, all with no connection to any candidate or group active in this race. Funding for this poll comes entirely from the donations of our readers. If you enjoyed this poll, please visit our site and donate so that we may bring you more surveys like it in the future. Our site can be found at http://rrhelections.com. To contact us for comments or questions about this poll, the best way is by email at rrhelections@gmail.com.
37 Comments
Damn. This is gonna be interesting.
One thing that I think deserves notice: Bishop is leading among those who have already voted and are certain to vote. That does not bode well for McCready, who only leads among those “likely” to vote, which is a category that would include many of the lower-propensity Democratic-leaning voters. However, McCready still has a chance, given that this is within the margin of error and that the Democratic base seems to be enthusiastic. Trump’s approvals are also poor, and that may boost McCready as well.
22-Independent from Colorado. Some agreement with both parties.Everything is within the margin of error in that respect though, and remember the MOE on those that have already voted in particular is huge (like 20%)
R, MD-7, Jorgensen 2020 I guess.Of course, but if I had to call this race right now, I would say that Bishop wins by a narrow margin, perhaps 2-3% or something similar. McCready could get around 48-49%, and Bishop, probably 50-53%. This of course, depends upon how the “undecideds” break, and given prior electoral history in the South, I believe that the majority of them will break for Bishop.
22-Independent from Colorado. Some agreement with both parties.I wouldn’t pay too much attention to the crosstabs. For example, this poll has Bishop only losing Black voters by 4, 18-29 year olds by 4, winning Cumberland and Robeson, but also has McCready pulling even with White voters, and only losing by 9 in Union County. These kinds of things tend to cancel each other out in the end.
20, Democrat. Represented by Nancy Pelosi. It's the economy stupid!I’m not giving “all” to this poll, as it provides us with only a rough indication of where this race may be placed. But one cannot completely discount some of those results. Turnout will be critical for this race, and given that Trump will be holding a rally for Bishop (I don’t think it has happened yet), that might give Bishop enough of a push to get over.
22-Independent from Colorado. Some agreement with both parties.I believe the Trump rally is scheduled for next Monday.
Just remember that the crosstabs have huge MOEs, particularly the ones that are <10% of the full poll. The MOEs on the Cumberland and Robeson numbers for example are north of 15%.
Polling is more art than science in a lot of respects and making sure these sorts of high-MOE crosstab sniff tests smell equally bad in both directions if part of how we determine we're getting the best model.
R, MD-7, Jorgensen 2020 I guess.Gonna be a close one.
Maybe someone morw familiar with NC voting patterns than I could answer this. Am I mistaken that Democrats have tended to historically done better with early voters than ED voters in the state? I seem to remember McCready having a lead early in the night with early votes and then then race got to a dead heat once ED votes were counted, or is my memory just bad?
I am an Amy Klobuchar Democrat.If McCready is only leading among black voters by four points he’s totally fucked.
D. MD-08. Né: Davybaby.And if he’s nearly pulling even with White voters he’s winning by double digits. He’s not only going to win Black voters by four, that’s just silly.
20, Democrat. Represented by Nancy Pelosi. It's the economy stupid!Doesn’t the “already voted” raise some serious red flags given the high numbers of D and black voters who have cast ballots already?
Which is what I was suggesting above. But again, it is clear that this race is close to 50-50, with there being a slight advantage for Bishop.
22-Independent from Colorado. Some agreement with both parties.About what I expected in this poll. Special elections always scare me.
I see this going within 5000 votes and flip a coin. Gut says McCready. I don’t bet against dems in specials.
MI-08 - Industrial Midwest style conservative. Anybody but Biden.It’s impossible for McCready to only be leading with blacks by 4%. This poll really has some wacky cross tabs!
The geographic crosstabs are what I’m paying attention to. Union is 63% Trump so I doubt that it only goes for Bishop by 10. I think he’ll win it by at least 15. However, McCready cannot win unless he either wins Meck by more or swamps the rurals, neither of which he is doing. I think it’ll be a 3 point Bishop win
New Account for inactive Old Poster.Trump Republican. Big Tent good, loyalty a must.
Very interesting poll indeed. After the underwhelming GOP performance in the poll this morning, I was worried that we were going to have a big Dem lead here. Thank goodness that’s not the case. And for those worried about internal cross-tabs, just bear in mind that many polls have odd cross-tabs that make little sense. I have learned to not totally discredit a poll with seemingly weird cross-tabs.
Anyway, President Trump is very smart to make a campaign appearance or two.
SC-3, Conservative RepublicanWhat were the dates the calls were made?
I am an Amy Klobuchar Democrat.GAJ Solutions placed live calls to cell and landline phones of likely voters in North Carolina’s 9th Congressional District on August 26-28, 2019, yielding a survey of 500 complete responses
Follow me on Twitter: @Izengabe_Poll has been added to wikipedia. Moving up in the world!
I am an Amy Klobuchar Democrat.Well some interesting observations on the crosstabs. The Black vote is given at 22%, but by registration Black voters are 19.4% and to date have only made up 18.3% of the early vote. So this may cancel out to an extent Bishop doing as well as 42% of the Black vote, which seems questionable.
Also in the crosstabs Bishop leads by 7 points in the early vote a bigger margin than with those yet to vote. Yet in the actual early vote numbers to date, registered Dems outnumber Repubs by 9.7 points, better than the registration figures where they lead Repubs by 5.4 points. In 2018 McCready won the early vote but Harris won the election day vote by about 7 points, so the poll is showing a contrary trend to what happened in 2018.
But as Shamlet points out there are big MOEs with the crosstabs which when all taken together can cancel each other out. But I thought it would be interesting to throw in what I am seeing in the early vote to date.
Overall the poll is showing a very close race which is what I expected and which is what happened in 2018. Election day turnout will be the key here – good GOTV operations are critical.
Huh. The combination of this poll (which is OK) combined with the Outer Banks one, which has an absolutely dire result for the GOP, reminds me of the same-day SC/GA specials from 2017, where the result in rural SC, which was not considered to be competitive, was totally dire for the GOP, while the suburban battleground held for another day. That’s somewhat hard to square with the crosstabs of the most enthusiastic voters being pro-Bishop and least enthusiastic being pro-McCready (it implies much greater enthusiasm from Democrats, as in 2017-18), though this might be a poll where you just have to look the other way from the crosstabs and take the toplines for what they’re worth
Thanks for the polls, shamlet!
McCain Republican. OH-16. Male, Russoanglohispanophone.Interesting thing when comparing Trump approvals vs. Biden’s numbers. NC-3 has Approve+4 and Trump+11, and NC-9 has Approve+2 and Trump +6.
Considering that the country is Approve -10, seems like that implies anywhere between Biden +3 and Biden +6. Which sounds right to me.
I am starting to get the feeling that Biden’s electoral advantage is dimming a little.
Henry Wallace Republican | Open-Minded for 2024What’s with the “totally undecided” wording? It makes him sound like a stoned out surfer.
Also, no choice to pick if actually not likely to vote, such as reaching a cell number of somebody who used to live in the district but has now moved away but still has the same phone number?
48 M, MO-02That’s an attempt to encourage leaners to pick a candidate. “Undecided” often ropes in leaners who aren’t 100% committed. “Totally undecided” implies either true indecision or not having given it much thought.
There was a ‘not likely to vote” option, but people who chose it were dropped from the survey.
R, MD-7, Jorgensen 2020 I guess.Based on today’s court ruling, I think the chances that the winner of this special election is in Congress 16 months from now are pretty low.
25, D, NC-04That appeared to be state house and state senate district ruling this morning; not congressional.
A brand new case would have to be filed if they want to do so for Congressional maps and the primary date is March 3rd, 2020.
And that’s if defendants don’t get it tossed early by state district for dealing with the exact same subject already ruled on by the federal courts.
What you are likely to see is in 2021 or 2022 some state district court tossing the post 2020 NC Congressional maps; even if said map concedes an additional seat to the Democrats.
And new challenges to state house / state senate as well.
48 M, MO-02That’s true, but one is already in process, and I don’t think there’s any reason the N.C. courts would hold back given Roberts’ SCOTUS opinion. Courts have already delayed primaries for Congress earlier this decade.
No South Charlotte-to-Fayetteville district would exist in a non-partisan or Dem-leaning map, Fayetteville/Robeson area would be in its own majority minority district and NC-09 would probably be contained in the Union/Cabarrus/Mecklenburg area, which Richard Hudson would win in a primary.
25, D, NC-04I just did a quick, whole counties-only, draw of this. The best I can come up with (using a 13-district map) is a D+3 seat that was D+1 in 2016. I’ll try it with 14 seats an attempt to adjust for growth patterns over the last ten years, but it’s probably still going to end up being a right-trending swing seat.
27, R, TN-09Vote for the asshole, it's important.
The best I could get with 14 seats is D+4.2 with traditional PVI and D+2.5 in 2016. A Democrat would likely win and hold that, but it probably wouldn’t hold over the decade.
27, R, TN-09Vote for the asshole, it's important.
A further refinement of the 14-set map that splits two counties instead of one (I’m going for PA judicialmander style maps here) comes out at a traditionalPVI of D+5.2 and D+3.5 in 2016. Given what kind of areas are in the seat, I think that Republican would capture it at least one before 2030 if current trends hold.
27, R, TN-09Vote for the asshole, it's important.
Political numbers cruncher John Couvillon is projecting NC-9 total turnout will be about 25% of the 506,109 registered voters, or about 133,000. Latest early vote/absentee figure to date is 46,351. Based on what happened in 2018, he thinks about 56% of the total vote will be early on site or absentee. Early on site voting ends Sept. 6.
https://twitter.com/WinWithJMC
Curious how his numbers will turn out as there was a sizable turnout today across the counties, part of it could be due to the hurricane coming and everyone being encouraged to try and vote early. Granted, more D rural counties will be affected in NC-09, with Union and Mecklenburg portions not getting much action.
Part of that is any already mailed in ballots that didn’t get delivered Saturday were delivered today due to the holiday.
I’m also note sure if there were polls open on Memorial Day for in person.
Let us hope that if any of the early voting is electronic, that the software & hardware used won’t lose ballots if the power goes out and that they are all protected by a strong enough surge protector. (Along with hoping that if there’s any early voting locations on the barrier islands that ballots turned in there don’t get ruined by flood damage.)
48 M, MO-02Why would you not weight your poll by partisanship and race? Those are the two leading factors in determining electoral choice.
We didn’t weight by race because the racial makeup is consistent with reasonable projections in this district. We would have if the numbers were significantly off of what we’d expect the electorate to be, but they weren’t so we decided to employ our philosophical predilection to stick as closely with the data we have wherever possible.
We don’t weight by partisanship EVER, because self-identified partisanship (or worse, party registration) is not a particularly reliable metric. Registered Democrats/Republicans, especially in a seat like this with a lot of coalition change, will differ from self-ID’d Democrats/Republicans, which will differ further from actually reliable D/R voters, and so it’s near impossible to determine what exactly the proportion should be. In this case we feel that Trump approval is actually a better metric of the electorate’s partisanship because it’s a simple reliable metric that actually correlates with D/R voting.
R, MD-7, Jorgensen 2020 I guess.