Ryan Cummings retweeted
Personal News:
my debut book 'WHAT CAN WE AFFORD: A TOUR ACROSS AMERICAN INEQUALITY
& A SEARCH FOR THE ROAD AHEAD'
COMING MARCH 9TH is
AVAILABLE ON PREORDER
all original reporting. literal sweat blood & tears went into finalizing it
no time for elite policy pretense.
lets go
my debut book 'WHAT CAN WE AFFORD: A TOUR ACROSS AMERICAN INEQUALITY
& A SEARCH FOR THE ROAD AHEAD'
COMING MARCH 9TH is
AVAILABLE ON PREORDER
all original reporting. literal sweat blood & tears went into finalizing it
no time for elite policy pretense.
lets go
23
51
226
41,647
Ryan Cummings retweeted
Not to get too political, but do any red states have BBQ that competes with blue states?
621
2,000
43,345
2,252,714
Even at a valuation of $1.1T, Anthropic would still need to generate annual revenues ~$300B by the end of the next 5 years, which would be >$30b what Meta is currently making.
Any valuation >$1T demands revenues keep going up and to the right; plateauing would be catastrophic.
Any valuation >$1T demands revenues keep going up and to the right; plateauing would be catastrophic.
Anthropic IPO buyers, per Bloomberg, now see the valuation at as low as $1.1 trillion
7
7
74
16,465
Ryan Cummings retweeted
What happens to the US economy if the AI bubble bursts? Lots of people think the pain would be akin to the dot-com bust: shallow & short-lived. SIEPR's @econJaredB + @weakinstrument run the numbers in the @nytimes and explain how this time could be so, so much worse. More in 🧵👇
1
3
5
534
Ryan Cummings retweeted
Who is going to be the first reporter to profile Jake Johnson and his campaign?
A high school math teacher, who grew up in a working-class household with 10 siblings, decides to run for Congress with a campaign almost entirely staffed by his former students in a district that went for Trump by 12 points. And then wins because of the precinct that he taught in for nearly two decades! That’s politics at its best. @jmart where are you on this??
A high school math teacher, who grew up in a working-class household with 10 siblings, decides to run for Congress with a campaign almost entirely staffed by his former students in a district that went for Trump by 12 points. And then wins because of the precinct that he taught in for nearly two decades! That’s politics at its best. @jmart where are you on this??
The final ballots that put us over the top came in from the precincts of the high school where Jake taught @DomSokotoff, myself, and thousands of students. It was such a great moment at 1:30am last night when we saw that those were most of the remaining ballots.
7
60
317
10,633
1/n A little more than a year after our initial bubble call, @econJaredB and I are back in the NYT examining the natural next question after "are we in a bubble?" what happens if it pops?
1
4
20
1,975
10/n: Third, and finally, is that the ability of the current Administration to effectively respond to a crisis doesn't look good. The midterm results this week may have changed that, although a competent executive branch goes a long way when the market is collapsing.
1
0
2
142
Ryan Cummings retweeted
I say this as a quant myself: reject the false aura of scientific certainty.
Once again, from me and Bonica in @newrepublic: newrepublic.com/article/216304…
Once again, from me and Bonica in @newrepublic: newrepublic.com/article/216304…
2
8
63
2,559
Ryan Cummings retweeted
1) Confounding--moderates and progressives run in different districts and are different. Many confounders are obvious:
AES is Muslim
Troy Jackson ran against strong incumbent
TX is a high-variance swing place
Ceiling effects in solid blue areas
Discuss qualitatively
AES is Muslim
Troy Jackson ran against strong incumbent
TX is a high-variance swing place
Ceiling effects in solid blue areas
Discuss qualitatively
2
5
62
2,275
Ryan Cummings retweeted
Breaking my twitter fast to say why you should ignore all (yes all) Quantitative Hot Takes on the "progressives vs. moderates" question
1) Confounding
2) SUTVA/spillovers
3) Measurement of moderation
Use qualitative thinking here and accept uncertainty
1) Confounding
2) SUTVA/spillovers
3) Measurement of moderation
Use qualitative thinking here and accept uncertainty
I would like to see someone take a look at what happens to “progressive” vs “moderate” electoral performance when controlling for same- or opposite-party control of the state government (maybe someone has and I haven’t seen it)
MN in particular makes me wonder about this: Flanagan has been held up as a major progressive underperformer, but Klobuchar also underperformed the national shift by ~4pp (and only did about ~2pp better than Flanagan—not nothing, but not gigantic).
MI has a similar story, FWIW; Dems up and down the ticket mostly didn’t put up the kind of colossal shifts that we saw across the South—including the moderates.
There’s an obvious plausible explanation—Walz was a drag for MN Dems, and in a general anti-incumbent environment, being the state party in power attenuated some of the anti-federal-incumbency energy against the GOP.
And just broadly, the biggest shifts we saw were in places like Texas, Mississippi, South Carolina, Kansas—places where the GOP has been deeply entrenched for a long time.
All of which is to say, because progressives are more likely to run in and win primaries in already-blue states (and moderates disproportionately likely to run in red districts) it seems to be that there could be a confounding factor in progressive “underperformance” in the form of thermostatic reaction to whatever party is in power in a given state. Has anyone done an analysis looking at this specifically? @lxeagle17 @JakeMGrumbach
13
75
312
23,873
Also, if this is the beginning of a larger drawdown (impossible to say if so--market could rebound tomorrow, especially since a lot of selling today was likely mechanical), then we are still very much in the early innings. Dot-com drawdown was particularly long and painful.
Pretty brutal day; worse than 99.5% of other days since 1929.
A cold comfort, but it's still better than Liberation Day and the worst days of 2008/the dot com bubble.
0
0
2
474