This continues an ongoing series about the 2020 Senate elections (my most recent update on the Presidential race is here). This week we add forecasts of total seats from both 538 and from Daily Kos itself. We’ll review the total seats each of the sites we’re tracking forecasts, then the polling averages provided by 270towin (270) and RealClearPolitics (RCP). Numbers are from 9/25:
538: D 51, R 49. Toss-ups (less than a 60% chance for either candidate) are IA and ME.
DailyKos: D 48, R 49, toss-ups 3 (as of 9/21). This is based off of the DKos competitiveness map for Senate races, not the polling averages. Toss-ups are IA, ME, and NC.
Princeton Election Consortium: D 52, R 48 (updated daily, no change). MT, KS, and AK are still the Senate “Moneyball” states. This site’s forecast is remarkably stable. On the other hand, the overall picture has changed very little for months.
CNN (no update): 60% chance of Ds winning at least 4 seats (AZ, CO, ME, and NC are the likely pickups) while losing just 1 (AL). IA rated a 50-50 toss-up and a 40% chance for the Ds in MT and the Perdue seat in GA. This all puts the Ds at a minimum of 50 seats.
Electoral-vote.com: D 51, R 47, 2 ties (updated daily; +1 Tie). D pickups are AZ, CO, IA, ME, and NC; R pickup is AL; the tie is AK.
Rachel Bitecofer: D 50, R 46, toss-ups 4 (as of 9/15; no change). Ds pick up AZ, CO, ME, and NC; Rs pick up AL. Toss-ups are AK, GA-Special, GA-Perdue, and IA.
Inside Elections: D 48, R 48, toss-ups 4 (updated 9/18; no change). Ds pick up AZ and CO, Rs pick up AL. Toss-ups are IA, ME, MT, NC.
Sabato's Crystal Ball: D 49, R 49, toss-ups 2 (updated 9/21; D +1). Ds pick up AZ, CO and ME; Rs pick up AL. Toss-ups are IA and NC.
Cook Political Report: D 48, R 47, toss-ups 5 (updated 9/23; D +1). D pick up AZ and CO, Rs pick up AL; toss-ups are GA-Perdue, IA, ME, MT, and NC.
RealClearPolitics: D 46, R 47, toss-ups 7 (updated daily; R +1). Ds pick up CO; Rs pick up AL. Toss-ups are AZ, GA-Perdue, IA, ME, MI, MT, and NC.
The stability over time is as consistent for Senate races as Biden’s polling lead over Trump (which has hovered around 7% for several months). Half of the 10 models we are now tracking put the Ds at 50 seats or more; none of them do so for the Rs. Almost everyone thinks AZ and CO will go Democratic; everyone thinks AL will go GOP. This by itself would result in a D 48 — R 52 Senate, but almost everyone also agrees on several key toss-ups: typically IA, ME,and NC, with GA-Perdue, MT and now AK making up a second tier. As we will see, Graham’s SC seat now appears to be a toss-up as well. All of these seats are GOP-held, which means the Rs are playing defense everywhere. Now, the individual races:
Now for the individual races:
MN (Smith/D vs Lewis/R): D +10.4 (270), D +8.7 (RCP). I’m going to drop this one, as it shows no signs of becoming competitive.
AZ (Kelly/D vs McSally/R): D +8.7 (270), D+5.5 (RCP). A single 1-point poll causes Kelly’s lead to shrink; all others in the last few weeks have him ahead 8-10 points.
CO (Hickenlooper/D vs Gardner/R): D +7.4 (270). Hickenlooper maintains a comfortable lead. The refusal of RCP to post any polls of this race is baffling.
NC (Cunningham/D vs Tillis/R): D +6.0 (270), D +5.2 (RCP). Cunningham’s lead opens up again.
ME (Gideon/D vs Collins/R): D +5.4 (270), D +6.5 (RCP). As I suspected, the Fabrizio Ward/Hart Research poll was a fluke; Gideon’s average opens back up.
MI (Peters/D vs James/R): D +5.2 (270), D +4.5 (RCP). Peter’s lead is smaller than I think any of us would like, but over 5 points is still solid.
IA (Greenfield/D vs Ernst/R): Tie (270), D +2.6 (RCP). A slew of new polls are more favorable to Greenfield; as I again suspected, the older Fabrizio Ward/Hart Research poll proves to be out of step.
AK (Gross/I vs Sullivan/R): Tie (270), R +5.0 (RCP). Both still based on single polls, but 270’s is more recent. New polls are needed here before we can be confident of what’s happening.
KS (Bollier/D vs Marshall/R): R +1.3 (270), R +2.0 (RCP). No change — again, new polls are needed.
GA (Ossoff/D vs Perdue/R): R +2.0 (270), R +3.3 (RCP). Again the Fabrizio Ward/Hart Research poll from last time is an outlier; the average swings back to slightly favor Perdue.
MT (Bullock/D vs Daines/R): R +2.0 (270), R +1.6 (RCP). Daines maintains a slim lead.
Next, the two most awful GOP senators of today:
SC (Harrison/D vs. Graham/R): R +0.5 (270), R +2.3 (RCP). Based on the 3 most recent polls for RCP, the most recent 2 of which are ties. Graham is definitely in serious trouble, and we reasonably hope we’ll soon see the last of him.
KY (McGrath/D vs. McConnell/R): R +13.5 (270); R +8.5 (RCP). We’re not so lucky with McConnell; more polls show that his lead really is double digits after all.
With just over 5 weeks to the election, all the candidates are running out of time to dramatically change their races. If these averages hold, we will definitely have 50+ Democratic senators.
