Through this series, I hope to keep the Dkos community informed in a single location on what the numbers and experts say about the state of the midterms. At the same time, it’s important to keep perspective. Polls are going to be off; pundits are going to make mistakes in viewing the political environment. My thoughts on polls, pundits, and predictions can be found in the first diary of this series, here. In general, my rule of thumb is that for a variety of reasons, polls today tend to overestimate Democratic support by several percentage points. As for the pundits, they may have as hard a time as any of us in understanding how major historical events — such as the Supreme Court overturning Roe v. Wade — can impact an election, making it far from routine. To give you as broad a picture as possible, and help you reach your own conclusions about the state of the midterms, I pull from a wide variety of sources. These include sites that provide polling averages; sites that use current polling and other data to predict vote outcomes; and political pundit sites. Below are brief descriptions and the links that can take you right to the sources.
FiveThirtyEight: Senate ratings and House forecast. Provides estimates of election-day vote based on both a polling average and a “deluxe” model that includes historical trends and expert ratings. (538)
Electoral-vote: Provides a simple last-week polling average for Senate races. (E-V)
RealClearPolitics: Another polling average, but broader than Electoral-Vote’s. (RCP)
270towin: Yet another polling average and algorithm. (270)
Cook Political Report: Expert ratings of each race (solid, likely, lean, or tossup). (Cook)
Sabato's Crystal Ball: Also expert ratings (safe, likely, lean, or tossup). (Sabato)
Inside Elections: Yet more expert ratings (solid, likely, lean, tilt, or tossup). (Inside)
The reason for looking at multiple polling aggregators and pundits is this: not all polling averages are the same — nor all expert ratings. The polling aggregators differ in the time frame they cover (from 1 week to a month or more, also depending on availability of polling for each race) and also differ in the pollsters they include (for example, some include partisan pollsters while others do not). Similarly, every pundit has their own “secret sauce” of factors that go into their forecast and how they weigh them. You can find details at most of the specific sites I include.
This week’s numbers were current as of Saturday 10/8. My abbreviations for the different sites should be pretty clear. For the sites providing polling averages, I transformed everything into the same format: a single number that shows the size of the Democratic (or Republican) candidates’ lead. For 538, which provides both polling and polling+ vote forecasts, I’ve kept the site’s original format of vote percentage for each candidate. Hopefully this makes it a bit easier to compare across sites.
Senate Races
|
E-V |
RCP |
270 (polls) |
538 (polls) |
538 (deluxe) |
Sabato |
Cook |
Inside |
Az |
D +4 |
D +4.1 |
D +6.4 |
D 52.1-45.9
|
D 51.7-46.3 |
lean D |
lean D |
tilt D |
FL |
R +7 |
R +4.0 |
R +5.6 |
R 51.2-46.6 |
R 52.4-45.4 |
likely R |
lean R |
likely R |
GA |
D +8 |
D +3.8 |
D +4.4 |
D 51.2-47.2 |
D 50.2-48.2 |
tossup |
tossup |
tossup |
NV |
R +2 |
R +2.1 |
R +2.4 |
D 48.2-48.8 |
D 48.5-48.5 |
tossup |
tossup |
tossup |
NH |
D +8 |
D +6.6 |
D +6.6 |
D 52.9-45.1 |
D 52.4-45.7 |
lean D |
lean D |
tilt D |
NC |
R +1 |
R +1.5 |
R +0.6 |
R 49.0-48.8 |
R 50.5-47.3 |
lean R |
lean R |
tilt R |
OH |
D +2 |
R +1.2 |
R +0.4 |
D 50.3-49.7 |
R 51.8-48.2 |
lean R |
lean R |
lean R |
PA |
D +4 |
D +4.3 |
D +4.4 |
D 51.8-45.9 |
D 50.7-47.0 |
lean D |
lean D |
tossup |
WI |
R +3 |
R +3.0 |
R +2.6 |
R 50.7-49.3 |
R 51.4-48.6 |
lean R |
tossup |
tilt R |
Senate overview: There has been no general trend, in my opinion, over the past week. Whatever tightening was happening in PA seems to have stopped (possibly due to such revelations as the fact that “Dr” Oz is an actual puppy killer). AZ may have tightened, though Kelly maintains a clear lead; also NV may have shifted more in favor of the GOP. On the other hand, Warnock’s lead in GA may be opening up again (possibly due to the endless barrage of scandals showing the dishonesty and hypocrisy of Walker). PA remains the most likely D pickup, but NC and OH remain very close and are excellent chances for Democratic pickups. Pundit forecasts tend to shift little (these do not update on a regular basis, and weeks can go by without changes). However, Inside Elections moved NC from lean R to tilt R and OH from like R to lean R, reflecting how consistently even these races are (even while moving AZ from lean D to tilt D and PA lean D to tossup, a judgment not agreed with by the other pundits).
The two columns for 538 illustrate the different conclusions that can be reached from polling data alone versus data plus expert knowledge. In almost all instances, the polls alone favor the D candidate more strongly than the ‘deluxe’ model that incorporates expert opinion, prior voting results, etc. This is also part of the reason that I believe the polls in general favor Democrats a bit too much. However, I also think it’s a mistake to react too strongly to one bad polling result for a Democrat — there is too much uncertainty in methods and responses to take any single poll as a very accurate snapshot of a race.
Overall, forecasts continue to suggest a Senate balance of power almost exactly what it is now. E-V again shows a 51-49 split for the Democrats; 538 still forecasts 50 or 51 Democratic seats; others show Democrats even or a seat ahead; and only RCP, with its conservative lean, forecasts a GOP takeover (with a 52-48 split). If you’re wondering where you can send some money, defending the Democratic seat in NV looks the most critical right now; PA remains the best place to go on offense.
House overview: The slow shifted towards the Democrats over the past couple of months may have halted, at least for now. 538 predicts a 70% chance of GOP control, compared to 68% last week. However, merging all categories other than “tossup,” most sites still show the GOP falling short of control:
RCP: R 220, D 180, T 35
270: R 219, D 206, T 10
Sabato: R 215, D 196, T 24
538: R 215, D 208, T 12
Cook: R 211, D 194, T 30
Inside Elections: R 211, D 205, T 19 (now with all 435 seats)
Obviously some sites are much more generous than others in ratings tossups, although some may have a bias of rating districts others see as D-leaning in this category (ahem *cough* RCP). At the same time, Democrats would have to pretty much sweep the tossup seats to maintain control. Unless we see renewed momentum in the remaining few weeks, keeping control will be very difficult.
Gubernatorial races
|
RCP |
270 |
538 (polls) |
538 (deluxe) |
|
|
|
AZ |
R +1.1 |
D +0.8 |
D 50.4-49.6 |
D 50.0-50.0 |
|
|
|
FL |
R +7.0 |
R +8.6 |
R 52.4-45.4 |
R 53.7-44.1 |
|
|
|
GA |
R +6.4 |
R +5.4 |
R 51.8-46.7 |
R 52.3=46.2 |
|
|
|
KS |
D +2.0 |
NA |
D 50.9-45.7 |
D 49.5-47.1 |
|
|
|
NV |
R +1.6 |
R +1.6 |
R 48.6-48.3 |
D 48.8-48.1 |
|
|
|
OR |
R +2.4 |
R +2.0 |
R 39.3-38.4 |
D 39.1-38.6 |
|
|
|
PA |
D +10.4 |
D +11.6 |
D 54.2-43.7 |
D 53.9-44.1 |
|
|
|
TX |
R +8.6 |
R +7.6 |
R 52.1-45.4 |
R 54.2-43.4 |
|
|
|
WI |
R +0.5 |
R +0.6 |
D 49.8-48.5 |
D 49.9-48.4 |
|
|
|
Gubernatorial Overview: Based on feedback from last week’s poll, this week I’ve added Oregon and Texas into our overview. As before, the governor’s races are definitely shakier for Democrats than the Senate. D-held governorships in OR, NV, and WI are in need of defense. OR is more difficult to call, though, as independent candidate Johnson is a former Democrat. If incumbent governor Kate Brown can convince Johnson supporters, mostly Democrats, that they need to rally to her to avoid a GOP win, then this seat becomes much more likely for the Ds. AZ is surprisingly close, as GOP candidate Kari Lake is a full-bore MAGA election denier and anti-choice lunatic. However, this still seems the best chance for a D pickup. If you’re looking for a Democratic gubernatorial candidate to support, I believe AZ (offense) and NV and WI (defense) are the most urgent needs. Abrams and O’Rourke remain substantially behind in their races, and the OR legislature should remain D-controlled, while Shapiro seems well on his way to victory in PA.
Discuss! What Senate an gubernatorial races seem most important to you, and why?