If Kaiser can get a Little Jiffy, you can too.

(I agree with you! I'm saying that I think many people avoid going into the actor-level, which makes it easier to accept certain ideas).

I think the problem is that many define "power" at such a macro-level that it ends up being ambiguous. "an actor i does this _because_ another actor j influences them" is much less straightforward than saying "these structures generate power imbalances."

“Credibility revolution” is really about observational designs. The “experimental revolution” I think is related but distinct (see 2019 vs 2021 Econ nobels)

the problem with sociological explanations is that if you really want to make the argument that such and such imbalances are at play, you need a plausibly specified motivational model that makes such imbalances an explanatory factor in the decision calculus of an individual actor.

that I agree with (though not believe in).

I think it's perfectly fine to take a position that says anything other than an experiment is already doomed. I just think that applies to everything, not just one or the other specific method.

perhaps. but one way or another none of these methods is more than X, no? IV is just a regression. DID is just a panel specification. the essence of the causal revolution is (or should be) less on the specific method and more on how you conceptualize your design.

well, if you look at it like that, everything is gone other than experiments.

well, more likely, sociology is neither that nor the other, not both:

sure, it's from John Levi Martin's Thinking Through Statistics.

let's square it with "the social reproduction of status."