Last week we took a closer look at the special elections post-Dobbs, and how they suggest good news for Democrats based on their overperformance in these districts. In this final pre-election overview, it’s time to turn back to the early voting and what that suggests for Democratic chances.
Early Voting: What we really want to know is if Democrats are meeting or exceeding their relative turnout in the latest election cycles, which would be an indicator of actual levels of enthusiasm. This means that in comparison to the 2018 midterms, we want to know if there is more early voting and if that electorate skews more Democratic than in 2018. In comparison to the 2022 presidential cycle (which always sees a greater proportion of eligible voters participating, we want to see if the electorate at least skews more Democratic.
And in fact, early voting numbers remain promising on both counts. Per this thread by Simon Rosenberg, looking largely at data from targetsmart, which also provides analysis and commentary of its own, we continue to see very positive indicators for Democrats in the early vote. Caveat: targetsmart is a Democratic-affiliated organization.
(1) There is a large gender gap, with women making up an even larger share of the early vote than they did in 2020 or 2018, and breaking even more strongly to Democrats than they did in 2018. Importantly, this is showing up in battleground states like Pennsylvania and Wisconsin. Just maybe this has something to do with the surge in female voter registration seen earlier this year, and the impact of Dobbs?
(2) Early voting totals are exceeding 2018 — as noted above, a good sign for Democrats.
(3) Democratic shares of the vote appear to be exceeding 2020 in multiple battleground states, including GA, NC, NV, OH, PA, and WI. Overall, the Democrat’s nationwide share of the early vote appears to be about 50% vs 39% for the Republicans — larger than the 48-41 advantage in 2020 or the 46-45 advantage in 2018.
This is not to say that all signs are encouraging. Targetsmart notes, for instance, that 18-39 year olds (a pro-Democratic demographic) are lagging behind their 2020 share of the early vote — and at this point there’s really no time to catch up, so hopefully they’re fired up to turn out on election day.
This week’s update: Before moving on to the final numbers and predictions, let’s take a look at some of the concerns now surrounding political polling in general. Broadly, these center around the increasing unwillingness of people — but especially those that favor Donald Trump — to speak with pollsters. This in turn feeds into pollster efforts to model “likely voters” to arrive at the polling results that actually get published. The 2020 election cycle saw the largest systematic error in decades, skewed towards Democrats. Large numbers of Trump voters either refused to talk with pollsters at all or simply lied. Does this mean there will be a similar problem when Trump is not on the ballot? Or will disproportionate numbers of Republican voters now continue the same pattern? At this point, no one knows, but pollsters are still left trying to make sure Republicans are accurately represented. This can take multiple forms, but all with the same basic approach — adjusting raw polling numbers to account for the presumed-missing GOP supporters. However, like generals fighting the last war, this leaves pollsters in the position of strategizing based on the last election. Electoral-vote provides a good brief summary of how there are multiple factors in play that complicate pollster models of the electorate. Are pollsters overcompensating in their efforts to estimate hidden GOP support? (If so, this means Democrats are better off than their numbers suggest). Is post-Dobbs anger and engagement still motivating Democratic voters, as it apparently did in the special elections? (If so, that also means Democratic numbers are underestimated). Will younger voters turn out in numbers similar to their unusually high participation in 2018 and 2020? (If yes, also good for Democrats). Will the white suburban shift towards Democrats in 2018 and 2020 hold up? (if yes..). I think you get the idea. There are a lot of hard-to predict factors in play, but I see most of them favoring the Democrats in this cycle. Unless, of course, the media narrative of “skyrocketing inflation/crime” drives up GOP turnout. We could talk about how prices have actually advanced very little since their spring/early summer surge — or about how crime rates such as homicide are actually worse in red states than blue — but that’s outside our focus.
One more follow-up point. Rosenberg’s thread, linked above, also notes that polling averages continue to apparently be gamed by GOP-leaning pollsters flooding the zone in key states. In GA, he notes that the most recent polls from the highest-rated independent pollsters show Warnock winning, but polls put out by Republican-leaning pollsters throughout October show Walker leading — and are far more numerous than the independent polls. We may need to be even more wary of narratives about Republicans closing strong or the ‘bottom dropping out’ for Democrats.
As before, if you want to look up more detail on the methods of our polling aggregators and pundits, here they are:
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)
And now, on to the numbers and ratings. These data are current as of Sunday morning 11/6.
Senate Races
|
E-V |
RCP |
270 |
538 (polls) |
538 (deluxe) |
Sabato |
Cook |
Inside |
Az |
D +1 |
D +1.0 |
D +1.0 |
D 50.2-48.0
|
D 50.1-48.1 |
lean D |
tossup |
tilt D |
FL |
R +6 |
R +7.5 |
R +7.0 |
R 52.9-45.2 |
R 53.9-44.2 |
likely R |
likely R |
likely R |
GA |
D +1 |
R +0.4 |
D +0.2 |
D 49.4-49.1 |
R 49.5-49.0 |
tossup |
tossup |
tossup |
IA |
R +12 |
R +12 |
R +11.3 |
R 55.1-44.9 |
R 56.5-43.5 |
likely R |
likely R |
likely R |
NV |
R +6 |
R +2.4 |
R +2.2 |
R 49.3-48.1 |
R 493-48.1 |
tossup |
tossup |
tossup |
NH |
D +2 |
D +0.7 |
D +2.4 |
D 50.9-47.4 |
D 51.0-47.2 |
lean D |
lean D |
tilt D |
NC |
R +3 |
R +5.2 |
R +3.8 |
R 50.6-47.6 |
R 51.5-46.7 |
lean R |
lean R |
tilt R |
OH |
R +9 |
R +5.0 |
R +2.7 |
R 51.7-48.3 |
R 52.8-47.2 |
lean R |
lean R |
lean R |
PA |
D +1 |
R +0.1 |
R +0.6 |
D 49.4-48.6 |
D 49.2-48.8 |
lean D |
tossup |
tossup |
UT |
R +6 |
R +10 |
R +11.3 |
R 50.5-40.8 |
R 50.5-40.8 |
likely R |
likely R |
likely R |
WI |
R +3 |
R +2.8 |
R +2.8 |
R 51.6-48.4 |
R 52.1-47.9 |
lean R |
lean R |
tilt R |
Senate overview: The final pundit ratings once again have almost no change in the Senate, except for Cook shifting WI from tossup to lean R. Considering Johnson’s small but consistent lead over Barnes in recent weeks of polling, this seems reasonable. However, note that our three pundits all have just enough tossups that they will probably be able to claim they were right, no matter what happens. Maybe Fetterman wins in PA and maybe he doesn’t; Nevada could go either way, as could AZ and GA; basically any result from 48 to 51 Democratic seats will be within their range. Punditry is perhaps a bit like prophecy; be cryptic enough in your pronouncements, and you will never err.
Numbers on the poll-based sites continue to generally show small shifts favoring Republicans. However, this isn’t consistent across the board. AZ, NC, and WI show little change relative to last week. For GA, 270 and RCP show small shifts in favor of Warnock. PA is the most concerning, with 538, 270, and RCP all showing shifts in favor of Oz.
Keep in mind, though, that we still need to be way of low-quality or GOP-leaning polls impacting the averages. In the case of NV, E-V warned that a recent InsiderAdvantage poll showing a 6-point Laxalt lead should not be taken too seriously; somewhat incongruously, they posted it on their site to give Rs the big lead shown above. In the case of OH, a flood of polls over the last week showed constant 4-5 pt leads for Vance, capped off by an Emerson poll with a 9-pt lead (which is what shows up in the E-V column). It does not seem plausible that Ryan has suddenly collapsed within the last week. It turns out a pollster called Cygnal started issuing every-other-day polls with consistent 4-5 pt Vance leads (see OH-Sen polls on the 270 site). The problem here is that the results seem a little too consistent; even coming from the same pollster, multiple polls should start to show a little more variability than that.
The final overall picture is that Senate control continues to be balanced on a knife edge, based on the polling. Projected total sites from the different sites now show 48-50 Democratic seats, depending on whose polling averages you believe. 538 now shows the Democrats slightly favored for control based on the poll-only “lite” model, but the Republicans slightly favored based on their deluxe model. However, if the Democratic overperformance from earlier in the year holds up, and other intangibles favor Democrats as I suggest above, the Senate should still remain in Democratic hands. I’m predicting 51 seats, myself.
House overview: The trend across sites was more slight movement towards the Rs in these final predictions. A majority of sites now predict outright GOP control, even accounting for seats still considered tossups. Unless there is a systematic error, the GOP is going to win control.
RCP: R 228, D 174, T 33
270: R 222, D 200, T 13
538: R 220, D 205, T 10
Sabato: R 219, D 196, T 20
Inside Elections: R 216, D 199, T 20
Cook: R 212, D 188, T 35
Gubernatorial races
|
RCP |
270 |
538 (polls) |
538 (deluxe) |
sabato |
cook |
inside |
AZ |
R +1.8 |
R +1.8 |
R 50.9-49.1 |
R 50.8-49.2 |
tossup |
tossup |
tossup |
FL |
R +11.5 |
R +11.0 |
R 54.4-43.7 |
R 55.0-43.1 |
safe R |
likely R |
likely R |
GA |
R +8.1 |
R +8.2 |
R 53.0-45.6 |
R 53.1-45.5 |
likely R |
lean R |
lean R |
KS |
D +2.5 |
D +2.5 |
D 51.6-45.5 |
D 49.8-47.3 |
tossup |
tossup |
tossup |
NV |
R +2.6 |
R +1.4 |
R 49.6-47.8 |
R 49.2-48.2 |
tossup |
tossup |
tossup |
OK |
R +2.3 |
R +2.4 |
R 50.6-46.0 |
R 53.1-43.4 |
likely R |
likely R |
likely R
|
OR |
tie |
D +0.4 |
D 43.6-42.2 |
R 43.2-42.6 |
tossup |
tossup |
tossup |
PA |
D +10.7 |
D +10.5 |
D 54.2-43.9 |
D 53.9-44.2 |
likely D |
likely D |
lean D |
TX |
R +9.2 |
R +7.7 |
R 53.4-44.5 |
R 55.1-42.8 |
likely R |
likely R |
solid R |
WI |
R +0.4 |
R +0.4 |
R 49.5-49.1 |
D 49.3-49.3 |
tossup |
tossup |
tossup |
Gubernatorial Overview: Pundit ratings only showed minor shifts, with Sabato upgrading R chances in FL and GA from likely to safe and from lean to likely, respectively. Inside Elections also shifted GA from til R to lean R.
In polling, there was no consistent pattern favoring either party. Only the WI and OK races shifted in favor of the Rs across al the polling sites. In contrast, AZ showed a slight shift towards Ds on 270 and RCP. And OR looked even more favorable, with all sites showing a pro-D shift. The numbers say this one could go either way, but the past couple of weeks have seen it looking better for Democrats.
Final thoughts: The final overall picture continues to show an extraordinarily closely divided country, which in my opinion does not bode well. Loss of the House, even with ongoing control of the Senate, basically means that nothing of value would get done legislatively for the next two years, other than judicial appointments in the Senate. If the House Republicans do what they promise, however, and spend all their time investigating Hunter Biden, trying to impeach Biden or failing that his cabinet members, and holding the debt ceiling hostage to Social Security and Medicare cuts, they are not going to endear themselves to the public. As I mentioned last week, I personally see this election cycle, and 2024, as the most dangerous for American democracy since the Civil War. But I also believe the move away from the Republicans will be real, and strong, in 2024 if the GOP House indulges its worst instincts and if Trump tries to run again.