Hmm, okay. This does update me a bit. But also I suspect that the actual versions of the concepts you were using were more like the sociopolitical versions of these concepts than the econ-brained versions.
For example, you could invest in Anthropic just the same whether or not you approved of their plans, meaning that investing isn’t a credible signal (in the economic/game-theoretic sense) of approval. In order to make it a credible signal you’d need some mechanism by which you are harmed if you invest without approving—e.g. if you’re part of a community which punishes those who knowingly cause harm; or if you’re part of a community with norms of not investing in things without approving of them. But these kinds of explanations are squarely in sociopolitical territory.
(You could argue that it’s a credible signal insofar as investing in Anthropic increases your chance of dying. But a) now you also need to credibly signal that you believe this, and b) it seems hard for the numbers to work out in a way which makes this credible unless you’re using FDT/UDT, which is very outside standard econ.)
More generally, the whole idea of communicating things is very hard to pin down in economic terms, since any individual sentence is not a credible signal of the meaning of that sentence without some sociopolitical context that incentivizes honesty (unlike, say, an antelope jumping, which is a credible signal of its health).
I expect I could generate similar arguments for your other bullet points but hopefully this conveys my broad perspective.
More generally, the whole idea of communicating things is very hard to pin down in economic terms, since any individual sentence is not a credible signal of the meaning of that sentence without some sociopolitical context that incentivizes honesty (unlike, say, an antelope jumping, which is a credible signal of its health).
There’s a big 2000 onwards econ game theory literature on communication and information design.
I think the formal analysis of these concepts in economics is far from adequate (I haven’t deliberately kept up with the literature but expect that if there has been significant advances I would have heard about them), but that’s probably reflective of the actual difficulty of fully grounding these ideas in micro foundations (like game theory), not lack of attention or bad incentives among economists. (In other words I think you’ll probably have a hard time trying to do this yourself, unless you’ve got some promising novel approach that they’ve missed.)
In practice I mostly depend on my intuitions on these topics, which is partly informed by the formal/econ analysis I’ve seen, but perhaps you’re right that they’re “more like the sociopolitical versions of these concepts”.
but that’s probably reflective of the actual difficulty of fully grounding these ideas in micro foundations (like game theory), not lack of attention or bad incentives among economists
I think it’s a combination of both. Agent foundations research (which I consider this to be) is genuinely difficult, but also game theory has been around for a long time now so I expect there would have been much more progress if not for bad incentives and lack of attention.
Some discussion related to this from the 2009 sociology/game theory book The Bounds of Reason (which I haven’t yet finished):
“The assumption that humans are rational is an excellent first approximation. Contemporary sociology…reject[ing] the rational actor model is the reason sociological theory has atrophied since the death of Talcott Parsons in 1979.”
“The most fundamental failure of game theory is its lack of a theory of when and how rational agents share mental constructs. The Bayesian rational actors favored by contemporary game theory live in a universe of subjectivity and instead of constructing a truly social epistemology, game theorists have developed a variety of subterfuges that make it appear that rational agents may enjoy a commonality of belief (common priors, common knowledge), but all are failures.”
Regardless of the exact source of the difficulty, though, the problem I’m most concerned with is how economic concepts are given more deference because they have formalisms, even though the formalisms are often importantly incomplete or misleading for describing real-world phenomena.
Hmm, okay. This does update me a bit. But also I suspect that the actual versions of the concepts you were using were more like the sociopolitical versions of these concepts than the econ-brained versions.
For example, you could invest in Anthropic just the same whether or not you approved of their plans, meaning that investing isn’t a credible signal (in the economic/game-theoretic sense) of approval. In order to make it a credible signal you’d need some mechanism by which you are harmed if you invest without approving—e.g. if you’re part of a community which punishes those who knowingly cause harm; or if you’re part of a community with norms of not investing in things without approving of them. But these kinds of explanations are squarely in sociopolitical territory.
(You could argue that it’s a credible signal insofar as investing in Anthropic increases your chance of dying. But a) now you also need to credibly signal that you believe this, and b) it seems hard for the numbers to work out in a way which makes this credible unless you’re using FDT/UDT, which is very outside standard econ.)
More generally, the whole idea of communicating things is very hard to pin down in economic terms, since any individual sentence is not a credible signal of the meaning of that sentence without some sociopolitical context that incentivizes honesty (unlike, say, an antelope jumping, which is a credible signal of its health).
I expect I could generate similar arguments for your other bullet points but hopefully this conveys my broad perspective.
There’s a big 2000 onwards econ game theory literature on communication and information design.
Maybe the seminal modern paper here is https://web.stanford.edu/~gentzkow/research/BayesianPersuasion.pdf
Don’t have opinions on how insightful it is, but its definitely popular—maybe you will get something out of it.
I think the formal analysis of these concepts in economics is far from adequate (I haven’t deliberately kept up with the literature but expect that if there has been significant advances I would have heard about them), but that’s probably reflective of the actual difficulty of fully grounding these ideas in micro foundations (like game theory), not lack of attention or bad incentives among economists. (In other words I think you’ll probably have a hard time trying to do this yourself, unless you’ve got some promising novel approach that they’ve missed.)
In practice I mostly depend on my intuitions on these topics, which is partly informed by the formal/econ analysis I’ve seen, but perhaps you’re right that they’re “more like the sociopolitical versions of these concepts”.
I think it’s a combination of both. Agent foundations research (which I consider this to be) is genuinely difficult, but also game theory has been around for a long time now so I expect there would have been much more progress if not for bad incentives and lack of attention.
Some discussion related to this from the 2009 sociology/game theory book The Bounds of Reason (which I haven’t yet finished):
Regardless of the exact source of the difficulty, though, the problem I’m most concerned with is how economic concepts are given more deference because they have formalisms, even though the formalisms are often importantly incomplete or misleading for describing real-world phenomena.