My credence that he wins the Democratic primary is 45% (pretty unstable) (he seems to be the strongest candidate so far, but there’s plenty of time for more to join the race)
My credence that he wins the general election (conditional on winning the primary) is 50% (moderately stable)
These numbers are close to 50% (and their product isn’t too far below it), so that’s good news for tractability.
I asked a Democratic political consultant (non-EA) friend about Shah and was given similar numbers—he said he thinks there’s a 20-25% chance Shah wins AZ 1 (about 50% chance of winning the Democratic nomination, and slightly lower odds in the general). He was skeptical about the benefits of having just one EA aligned house seat, though.
Also I don’t really see a strong case for sharp non-linearities here (and if anything there might a stronger case for decreasing than increasing marginal returns). So if 50 is great, naively 1 won’t be too far away from 1⁄50 as great.
There, in fact, are sharp non-linearities, and they’re quite important. Having 218 Representatives (a simple majority) is much more than 218 times better than having 1. Same for 290 (a supermajority).
Right, that’s a good point. I was imagining that EAs being a simple majority of Congress is just obviously an unrealistic pipe dream, but people who are optimistic about that should factor that in into their calculations.
My sense is that there are also sharp non-linearities even at small numbers. What I was told is: “In the house especially, you need to act in groups. The average member of the house has ~zero control over policy. They might get some oversight responsibilities via their committees… but five backbenchers may be able to extract a promise or two from leadership whereas one would simply lack the clout.”
I spent 10 minutes investigating.
My credence that he wins the Democratic primary is 45% (pretty unstable) (he seems to be the strongest candidate so far, but there’s plenty of time for more to join the race)
My credence that he wins the general election (conditional on winning the primary) is 50% (moderately stable)
These numbers are close to 50% (and their product isn’t too far below it), so that’s good news for tractability.
I asked a Democratic political consultant (non-EA) friend about Shah and was given similar numbers—he said he thinks there’s a 20-25% chance Shah wins AZ 1 (about 50% chance of winning the Democratic nomination, and slightly lower odds in the general). He was skeptical about the benefits of having just one EA aligned house seat, though.
Before you can get dozens or even several, you need to get the first.
Yes, but if getting dozens is not a realistic goal, then there are questions re: how much you should invest in getting the first.
the not realistic is a debatable assertion
Also I don’t really see a strong case for sharp non-linearities here (and if anything there might a stronger case for decreasing than increasing marginal returns). So if 50 is great, naively 1 won’t be too far away from 1⁄50 as great.
(But maybe I have a failure of imagination).
There, in fact, are sharp non-linearities, and they’re quite important. Having 218 Representatives (a simple majority) is much more than 218 times better than having 1. Same for 290 (a supermajority).
Right, that’s a good point. I was imagining that EAs being a simple majority of Congress is just obviously an unrealistic pipe dream, but people who are optimistic about that should factor that in into their calculations.
My sense is that there are also sharp non-linearities even at small numbers. What I was told is: “In the house especially, you need to act in groups. The average member of the house has ~zero control over policy. They might get some oversight responsibilities via their committees… but five backbenchers may be able to extract a promise or two from leadership whereas one would simply lack the clout.”