How the fk do we even prove any of the economic concepts.

>>16722524 (OP)
Categorize the situation first then do statistical analysis.

Economic regions with similar policies and environment but different outcomes can be compared.

>>16722524 (OP)
The World Bank is constantly running experiments on shithole countries. Economics is more of a real science than astronomy.

Randomized control trials are the gold standard. Half of a randomly selected group gets a policy treatment and half of a randomly selected group doesn't get, then you do statistical analysis on the outcomes. Usually stuff like difference-in-difference regression, discontinuity design, etc.

>>16722543
the existence of the "world bank" and how malleable "economics" actually is as a result is proof its not a science at all.

>>16722524 (OP)
Just accept they are human social constructs. Can a cow be equal to 100 chickens? .Its nonsense, but you have to do trades.

Curve fit the existing data, assume the future will conform to this.

>>16723223
the fuck does epidemiological approach have to do with economics lmfao

>>16722524 (OP)
Through statistical inference like >>16722541 said, which is what science does with anything that cannot be directly observed but is postulated to have observable effects.

FE all possible EQ attributal relationships from test data and sample people in isolated simulations then use gans to form the perfect model with constraints that are set by the features and use generative ai

>>16722524 (OP)
well you create parallel universes and start changing parameters in them and measure the results

Many-World economics

>>16724584
maybe physics really is a real science, like economics

>>16724584
We should apply this to every field, actually. Law ought to assume that every possible legal outcome exists in a multiverse of verdicts.

>>16722524 (OP)
You don't. As an EE who does systems/signal processing, I'm always amazed when I see economists borrowing our concepts for their time series analysis. Constant velocity models and interacting multiple models make sense when you're talking about local motion of a distant target like an airplane or a submarine or something. They make a lot less sense when you're using a Kalman filter or a Bayesian Gaussian mixture filter to do economic predictions.

I think they just do fairly basic maximum-likelihood and Bayesian estimation strategies and then hope that whatever they are interested into roughly maps onto the concepts from statistics and estimation theory they are borrowing.

>>16722524 (OP)
Austrian economics is self-evident and does not require any proof beyond understanding human nature. All men behave rationally at all times. Everything else is scammers.

>>16724790
what is human nature?

>>16724790
>All men behave rationally at all times
Not sure if troll or just a very self unaware autist. Austrian is the common sense but the whole point of economics exand social sciences in general is the fact that most people are fucking idiots and easily manipulated. How many covid shots did you get?

>>16722524 (OP)
You just make whatever assumptions you need to make it all sufficiently simple.