Blank Slate Profile picture
Nov 7 17 tweets 3 min read
My post mortem: Exit polls tell the story. Narrow race decided at the margins. Harris hit her targets among whites in NC, PA, MI, WI, NV (enough to win EC). Lower minority electorate share in NC/PA/WI + Latino red shift killed her. In NV/MI, it's exclusively about Latino shift. AZ: missed target among whites, but also a red shift among latinos. White share dropped to 63% in AZ, so Latinos at '20 margins would've won it.
GA: Missed the white share by 2 points (28 v 30). Latino share narrowed to +15, but black vote would've been enough w/30% white vote.
Nov 6 5 tweets 1 min read
So what's the optimistic take for the Dems in this bleak time? A few thoughts:
1. Dems gained w/whites From -17 to -12, driven by a +10 margin with white college. 2. Black voters didn't move at all. Plug the '24 exit polls into '20 vote share by race demo and Harris wins by +4. Obviously the Latino hard right shift (+33 to +6) and Asians (+27 to +18) hurt Dems, but even with those shifts, just getting minorities to their '20 electorate share (34% in '20 vs 29% in '24) wins the election for Harris. Ds had a turnout problem.
Nov 6 12 tweets 2 min read
Thoughts on this election defeat. Starting point is Biden's age & unpopularity, but there are drivers to it.
1. Lack of Narrative: Biden rescued the US from the pandemic & fixed the economy. Got no credit b/c Ds didn't explain why things were bad & that progress is incremental. Obama made this case in a worse economy to win re-election. You have to stay connected to voters. Harris got this in part, but she entered late and didn't have time to develop the narrative that Biden had been ineffective in delivering.
Nov 4 15 tweets 3 min read
Prediction time!
US Presidential Election
EC: Harris 319-219 (2020 map + NC)
Popular Vote: Harris 52.1% Trump 46.6% (Harris +5.5).
Projected turnout: 159M (similar to 2020).
Demo splits: White: Harris 43 Trump 56; Nonwhite: Harris 70 Trump 28.
State predictions below: Image PA: Harris 51.4 Trump 47.9 (+3.5). Harris gets 44% white support, with improvements among college/non-college whites. Trump has pushed Dem white support down with large WWC turnout, but they'll be a smaller share, college whites are 'sick of this s**t' & minorities are engaged.
Oct 31 4 tweets 1 min read
My theory on the polling: Pollsters are locked in not undercounting Trump. Polls are engineered to make sure he hits his 2016/20 vote share. However, he's slipped among white voters and that causes pollsters to twist themselves in knots to get him to that '16/'20 vote share. The truth is Trump keeping things close in '16 & '20 had little to do with rural/WWC voters. It was about white college voters. Trump overperformed polls in both elections w/that demo but he did lose significant ground from '16 to '20, losing the demo by 3 pts. That's why he lost
May 2, 2023 4 tweets 1 min read
GOP is relying on the media to both sides the debt ceiling giving them the political capital to blame Dems & have no consequences from a default. It's a mirage. If Biden/Treasury persuade markets/creditors to dismiss the debt ceiling as legislative minutiae, GOP power evaporates. The debt ceiling 'debate' is a fraud. We borrowed the money years ago. The bill has come due. You pay the bill. It's fundamental to the economic system. If Congress authorized the borrowing it also validated the debt obligation. Therefore, it presumptively must be paid.
Nov 17, 2022 5 tweets 1 min read
This bill gives Federal protection for a marriage that is recognized as legal in a given state. That's better than the current legal regime on abortion, where someone seeking an abortion outside of their red state can be criminally punished and have their privacy invaded. This bill basically assumes that SCOTUS will overturn Obergefell and it provides fall back protection. If TX outlaws same sex marriage, the Feds will recognize a same sex marriage for a couple residing in TX that was performed in NY.
Nov 11, 2022 13 tweets 2 min read
House Update: Using MSNBC’s list of called races + a few more called by Dave Wasserman, at present Dems have 203 called seats and the GOP has 209. There are 23 tossup or uncalled seats. 14 of the 23 are in CA. Of that group, 6 of them are basically Solid to Safe D seats (CA-06, CA-09, CA-15, CA-21, CA-26, CA-34). That gets the Dems to 209.
Nov 11, 2022 9 tweets 1 min read
Short take on the House: By my count, I see Dems at 211 and the GOP at 210 called or nearly called seats (421). That leaves 14 toss-up/uncalled seats. Dems have paths to get to 218 if certain races break their way. Those 14 tossups/uncalled races are: AZ-01, AZ-06, CA-03, CA-13, CA-22, CA-27, CA-41, CA-45, CA-47, CA-49, MD-06, NY-22, OR-05, OR-06, WA-03.
Nov 9, 2022 4 tweets 1 min read
This was Biden's point when he disagreed with Obama on the Summers approach in '09/'10. I always took a skeptical eye to 'inflation' (except for gas prices) b/c econ reports told us people were employed and spending $$$. We let the media define this as a tumultuous, dispiriting time, when it should've been celebrated like the end of WWII. Biden ended the pandemic phase. He provided freedom through vaccines. He made gov't work again. We saw it in the economic boom.
Nov 9, 2022 10 tweets 2 min read
The goal for Dems in 2024 is to appeal to the people who like them: younger voters under 50, minorities and college educated whites. Don't sacrifice hitting their political sweet spots to pacify 50+ voters or Trump voters in the hopes of managing margins. It won't work. Cheri Beasley lost because the electorate was just too white. It needed to be younger and more minority. She would've won b/c her overall white share was surprisingly high.
Nov 8, 2022 10 tweets 2 min read
2022 Forecasts, Scenarios and Predictions Vol 3: US House. Just about every analyst has made the GOP a lock to win the House. I think it’s a tossup (there could be tipping points either way). The available data, imho, don't support an overwhelming advantage for the GOP. What gives the GOP an advantage is that at the current 222-213 margin, the GOP needs only 5 net gains to get to 218 seats and claim a majority. That’s why folks are in a rush to award the GOP the House. Imho, they have to earn it by beating front line Dem incumbents.
Nov 7, 2022 26 tweets 5 min read
2022 Forecasts, Scenarios and Predictions Vol 2: US Senate. Polling has been all over the map with partisan GOP polls making an emphatic statement of where they think the races are, and non partisan polls basically showing the opposite. Here are my takes: PA: Fetterman +3 (50-47). Fett has been ahead by 2 to 4 points since Oct 1. He's been averaging around 46% white support, and has strong white support even in GOP polls like IA. He’s also winning minorities by typical Dem margins and indies by around +6 to +7. .
Nov 7, 2022 8 tweets 2 min read
2022 Forecasts, Scenarios and Predictions Vol 1: Everyone knows that this is a weird election cycle. Some folks are going to be right. Others wrong. A lot of the ET takes feel like nothing more than vibes to fit a predetermined narrative It is what it is. So here are my takes. Congressional PV: Dems 50.3, GOP 48.2. Polls that provide data show Dems getting into the low 40s among white voters, which is more or less the magic threshold to win nationally. Polls during the final week have all seen the Dems in this range, with the GOP around 54%-56%.
Nov 6, 2022 5 tweets 1 min read
ABC/WAPO has Ds down 48-50 among LVs and 48-49 among RVs. This is a +3 improvement for the Dems among LVs (no change among RVs). In addition, there are reasons to think this poll understates the Dems' strengths heading into e-day. First, this sample has 12% of GOPers vs 14% of Dems/Dem leaners having already voted. However, per Prof McDonald's tracker, in 23 tracked states which report ballots by party ID, Dems have cast 43% of all ballots and Rs have cast 34% (nearly a +2M advantage).
Nov 6, 2022 7 tweets 2 min read
The Grassley numbers seem to align to the 2020 exit polls (which were R+10 by party ID share for IA) in a year where the overall national electorate was D+1 (by party ID share) per exit polls. We are looking at a R+1 to D+3 electorate. IA-02 and IA-03 look very competitive and winnable for the Dems. Axne (IA-03) has outperformed the state's congressional vote in both 2018 and 2020. She may do so again (in a district that got slightly more D vs '20).
Nov 2, 2022 5 tweets 1 min read
NC EV: 1.44M early votes cast, including 1.33M in person votes. Dems' cast ballot lead is at +7.4, which is higher than '20 (+5.6). Avg cast ballots/day is around 100k. At that rate we're looking at under 1.9M total EV (lower than '18). The total vote in 2018 was 3.76 million (EV was 2.03M or 54% of the total vote). Assuming the EV/E-day split is 55-45 or 60-40, the trajectory of the total NC vote may fall short of '18.
Nov 1, 2022 10 tweets 2 min read
About 79 polls on the 538 tracker having Oct dates. Raw average: Ds 45.1, Rs 44.7 (D+0.4%). projects.fivethirtyeight.com/polls/generic-… If one excludes partisan GOP polls, the D margin goes up to +1.2%. LV polls only are GOP +0.6%. Excluding LV polls (RV +A only) is D+1.6%. That's essentially a range of GOP +1 to D+2 if you round to the nearest whole number.
Oct 30, 2022 6 tweets 1 min read
NC EV: Looks like Dems had a big Saturday turnout as their overall EV margin went up from +7.8% to (39.2%-31.1%) over the GOP (+8.1%). Total vote is 1,126,702. At current rates, the total EV will be in the 1.9M to 2M range, close to but maybe short of '18 (2.03 M). This race is really tough to figure out, because the EV is pointing to a lower turnout election relative to the last 2. Both sides have a plausible scenario where they bring supersized turnout of their strong demos to win.
Oct 29, 2022 5 tweets 1 min read
The GOP has completely broken the polling averages this month w/junk polls so as to render them effectively meaningless. It’s as bad as 2012. However, that’s probably good news for Dems and bad for the GOP. GOP efforts to corrupt the polling averages are so over the top, so ‘schlock and awe’ that more Dems are simply ignoring the polls, showing less worry and are just voting. The political narrative they want to drive isn’t sticking.
Oct 29, 2022 7 tweets 2 min read
NC EV: 1,063,378 total votes cast, including 975,131 in person votes. Avg in person/day is 108k. At this rate, we're looking at 1.9M-2M early votes, which is pretty close to '18: 2.03M. Dems lead the EV 39.1% - 31.3% (+7.8%), which is better than '20 (+5.6%) but worse than '18 (+12.9%). That said, there were no major statewide races on the '18 ballot, so '20 is probably the better guide, and that suggests incremental Dem improvement so far.