Ethan C7 Profile picture
24 | Elections Nerd | Moderate Democrat | Part Time Optimist | Full Time Skeptic
Nov 4, 2024 10 tweets 3 min read
FINAL 2024 RATINGS:

PRESIDENT - LEAN HARRIS (308-230)
HOUSE - LEAN D (225-210)
SENATE: LEAN R (51-49)

Based on a D+3 Environment, as implied by an aggregate of polls, special elections, Washington primary, Selzer, etc.

Thread on the 2024 election below 🧵 Image
Image
Image
Image
If one only looked at polls, with every swing state tied, one would easily conclude the election's a tossup.

However, actual voting behavior has consistently suggested something bluer. And the NYT's explanation - very red nonvoters - has just been mostly recanted in their polls.
Aug 28, 2024 4 tweets 2 min read
Recently, some folks have asked me where donations would best help Dems flip the House. IMO, there exist clear answers: four Dem challengers in Tossup races with MUCH less $$$ than their GOP opponents:

1. Marlinga (MI-10)
2. Tran (CA-45)
3. Shah (AZ-01)
4. Bynum (OR-05) Image It makes a MUCH bigger diff to go from being outspent 2-1 to parity, than to go from parity to 1.1-1 (the case in well-funded races). And the more $$$ spent, the less each $ matters.

Rs won 3/4 of these in 2022, due in part to lack of Dem $$$. Dems can't let that happen again.
Jun 11, 2023 4 tweets 2 min read
The idea of Jan 6 being a third-rail for GOP voters is possibly the greatest info op Trump has pulled off.

It’s his biggest vulnerability imo, and despite nominal party backing, a majority of the GOP don’t really believe it. Yet he’s convinced the GOP that they can’t attack it. It’s the classic self-fulfilling prophecy of pols influencing public opinion influencing pols, except the “prophecy” isn’t even fulfilled - the pols are just gaslit into thinking it is.

RDS & establishment too cowardly and think too low of their voters to try turn against Trump.
May 27, 2023 4 tweets 3 min read
Looking back over the 2020 Senate elections atm, and I have to say, I don't think incumbency has ever been as lousy an indicator of strength as in these elections.

Dem incumbents went 7/11 for outperforming Biden.
GOP incumbents went 10/17 for outperforming Trump. ImageImage Even open & appointed incumbent seats split 2-4 between outperforming Biden and outperforming Trump.

Might need to rethink the prior that Senate incumbents outperform the top of the ticket by default. Image
May 18, 2023 11 tweets 5 min read
Catalist has released their new analysis on the 2022 electorate (HIGHLY recommend reading), and one of the most interesting findings: Dems actually did BETTER w/ Latinos in 2022 than w/ 2020 Prez, but did WAY WORSE with Asians.

catalist.us/whathappened20… Image Equally as interesting - for white voters, Dems actually held steady w/ non-college whites, and mostly suffered w/ COLLEGE-ED whites (downballot lag?).

M.w. for black voters, we have the opposite.

The Dem gain w/ Latinos & loss w/ Asians were roughly equal, college-ed or not. ImageImage
Dec 12, 2022 4 tweets 2 min read
In a general sense, both Walker and Dr. Oz were bad candidates. However, the swing maps in GA and PA look very different - Walker was garbage in the burbs, but held up in the rurals, while the opposite was true for Dr. Oz.

Feels like diff ways of being bad has diff effects. I know that it's kinda an obvious statement - of course different ways of being bad will have different effects! - but I wanna drill into that.

What kinds of being "bad" disproportionately turns off rurals? What kinds of being "bad" disproportionately turns of suburbs?
Dec 11, 2022 4 tweets 1 min read
As recently as 2018, despite going blue for Prez, the Somerset County Board of Commissioners was still 5-0 GOP. However, by 2020, every seat had flipped to Dems.

Despite GOP efforts to reclaim ground in 2022, commissioner Marano actually IMPROVED upon the Dems' 2020 performance. Mapping this cuz I'm pretty sure 2022 was the first time I ever received mailers for a countywide race. The Somerset County Dems took these races *very* seriously after nearly losing their commissioner seats in the red wave of 2021.
Nov 26, 2022 7 tweets 2 min read
Despite being a Democratic midterm, there were quite a few states House Dems actually outperformed their 2020 margins in, while the GOP notably swept through FL and NY.

After accounting for uncontesteds, this is how each state swung from 2020 House -> 2022 House. Looking at this, some regional patterns clearly emerge, with Dems performing well in the Great Plains and Upper Midwest, while the GOP dominated in the South and (most) of the West coast.

There’s good evidence that persuasion/turnout patterns by race prob caused some of this.
Nov 22, 2022 4 tweets 1 min read
Wondering whether Biden should take a trip to Anchorage during the 2024 campaign.

You know, not because he can win it, or that it’d be worth much even if he did (3 electoral votes), but for the sake of the Dems’ longterm prospects, which look surprisingly decent there. After this midterm, Wisconsin no longer looks like obvious doom for the Dems, so they still retain 25 obvious states to get Senate seats in (the 25 Biden won in 2020). But they don’t have many places to expand the map without some long term investment.
Nov 18, 2022 5 tweets 1 min read
Cumberland County, PA’s ringing some loud alarm bells this election.

2020 Prez: Trump+10.6
2020 AG: Heidelbaugh+11.3
2022 Senate: Oz+5.5

Even Scott Perry is only gonna outperform Trump by 1.4% here, vs ~5% everywhere else. Despite overperforming most in Cumberland in 2020. Image For reference, this county is where Biden got one of his largest swings from 2016 to 2020 in, from Trump+17.8 to Trump+10.5. It’s redder downballot, so given 2022, it seems pretty likely Cumberland’s in for another big swing to the left in 2024.
Nov 14, 2022 4 tweets 1 min read
Game over for Lake, and game over for Hodge.

Dems win AZ Gov. GOP wins AZ-01. With the latest batch in CA-41 coming in big for Calvert, it’s looking very unlikely Rollins can come back.

Can’t officially call it, but very likely the GOP holds CA-41
Nov 11, 2022 4 tweets 1 min read
Doing the math in WA-03, currently Kent is down 5117 votes with ~54051 votes left -> must win them by 9.5%.

If these votes went the same way as their county in 2020 (they won't), they'd go Biden+1.3%.

Thus, he needs every new batch to be 10.8% redder than the county in 2020. Fwiw, the batches that have been coming in have not been anywhere near red enough for Kent to win.

He'd better hope the later batches break that pattern. Otherwise, MGP is set to pull off the upset of the year.
Nov 11, 2022 7 tweets 3 min read
The State of Play:

With the likely reelection of both Mark Kelly and CCM, Dems have likely kept the Senate.

In the House, the field has changed. To win a majority, Dems now need to win all but one of: AZ-01, AZ-06, CA-13, CA-22, CA-27, CA-41, WA-03. Difficult, but doable. There exist outside shots Dems can still win OR-05 and CA-45. However, it seems pretty unlikely in both.

Arguably, CA-27 is also a longshot, but others seem to disagree, and tbh I'm not quite sure myself.
Nov 11, 2022 8 tweets 3 min read
In the two completed counties in Biden+0.3 AZ-06, Ciscomani (R) is winning Trump+55 Graham (partial) by 54%, and Trump+34 Greenlee by 27%.

These are tiny, and the big Pima portion is redder downballot, so it still favors GOP, but wondering if we wrote this one off too early...? More votes just got counted in Pima, which is now Engel+7.4%.

This portion is 70% in, and went for Biden by 8.6%. Engel needs to be winning here by about 8.3% to win - I need to see what the remaining ballots are in Pima, but she improved her margin with the last batch.
Nov 9, 2022 4 tweets 1 min read
Trump+26 Polk County, NC is 95% in.

Budd+22 Trump+28 Pamlico is > 95% in.

Budd+30
Nov 8, 2022 6 tweets 2 min read
FINAL Predictions: Based on an R+1 Environment
(= R+2.5 House pop vote b/c uncontested seats)

House: 223R-212D
Senate: 50R-49D-GA Runoff
Gov: 28R-22D I recognize that this is more bullish on Dems than other forecasts, and imo the reason is that I don't think the polls are underestimating the GOP.

The data roughly points to an R+1 enviro (5% swing right from 2020), and thus I rated the House accordingly.
Nov 6, 2022 4 tweets 1 min read
How many 2020 Trump counties does Josh Shapiro carry in PA-Gov?

Some possibilities (Prez, AG):
Berks (Trump+8, Heidelbaugh+5)
Luzerne (Trump+14, Shapiro+0.2)
Beaver (Trump+17.7, Heidelbaugh+4)
Cumberland (Trump+10.5, Heidelbaugh+11.4) For reference:

Berks is in SEPA and contains Reading

Luzerne is in NEPA and contains ancestrally Dem Wilkes-Barre

Cumberland is in SCPA and contains the Harrisburg burbs

Beaver is in ancestrally Dem WPA, just NW of Pittsburgh
Aug 24, 2022 4 tweets 1 min read
Re: how a general elex enviro may differ from a special elex enviro, lemme pull up the special elex from EDay 2021.

While Rs outperformed by ~12% in NJ & VA, they only outran by 1.4% on avg in the concurrent specials, tho there was a wide variation. The avg of post-Dobbs specials have been ~5% to the left of 2020. Thus, if we apply the same diff from 2021, you'd get about an EVEN environment in November, which lines up w/ expectations.

However, this is a example of why the special elex enviro likely won't transfer to EDay.
Aug 12, 2022 13 tweets 3 min read
The hopium train chugs along Image Well, the hopium's coming and it don't stop coming Image
Jun 12, 2021 4 tweets 2 min read
I think a good case study for this can be found in Kentucky - what happens when you run someone with a similar personality to Trump but the same old conservative economics on issues like education and unions?

Answer: you become Matt Bevin and lose reelection in a Trump+26 state. OTOH, what happens when you run on Trumpian issues but without his personality?

Im…not sure? I can’t think of any recent parallels.

Hawley fits this, but he reinvented himself this way *after* he won in 2018.
Mar 3, 2021 4 tweets 2 min read
Looking through my old pre-election county rating maps.

Honestly, aside from the disaster that was the Wisconsin map, these really weren't too bad imo, given the polling at the time. Tho in hindsight, I really have no idea why I put Berks County, PA in Lean R, or Cumberland, Beaver, and Luzerne into Likely R. Erie probably should've been Lean D as well.

Definitely should've known better.