[pdonis]

> what is the alternative?

To not commit an entire society to a single economic theory. Which means getting rid of a lot of the centralized control mechanisms that currently require just such a commitment.

> whose theories do we want to pay attention to?

Since no one has a theory that is good enough to make useful predictions, as far as public policy is concerned, the answer is "none of them".

> There's also the difficult problem of identifying when any theories have really been tested well.

This isn't a difficult problem. You test theories by comparing their predictions to what actually happens. (Note that a "theory" that makes no testable predictions fails this test.) As the article notes, economics does not do well on this test. But that's not a "difficult problem" of how to test economics; it's just that economists, and politicians who want to use economics to justify their pet policies, don't want to admit that the result of the test is that economics doesn't make good predictions.

> It's not any different from physics theory or chemistry theory in that regard.

Yes, it is, because, as you note, you can run controlled experiments in physics and chemistry, but you can't in economics.

[GhostVII]

> To not commit an entire society to a single economic theory. Which means getting rid of a lot of the centralized control mechanisms that currently require just such a commitment.

Isn't that just another economic theory?

[lliamander]

> Isn't that just another economic theory?

Yes, in the same way we could formulate a medical theory that says "we don't know enough about disease to know whether our treatments are effective, so we aren't going to create a standard diagnostic manual". Prior to the last couple centuries, that would probably have been an improvement.

No, in the sense that, compared to most economic theories, you can't really use this theory to justify a particular group being given access to government power.

[bobdole1234]

I think you're completely wrong there: You're saying that we should have left medical decisions in the hands of the local purveyors?

I think most of the gains we've gotten from the advent of evidence based medicine is building the taxonomy of medicine so we could reason about it effectively. That only works in a larger sense, small groups toiling with different nomenclatures end up wasting effort.

Much the same way economics is building the taxonomy of economies so we can compare and discuss.

Economics is going to become an entirely different animal in the future as we begin to automatically collect fine grained metrics from our automated economies.

Soon we'll be in the era of evidence based economism, and we'll see a huge increase in the effectiveness of it's utilization.

[lliamander]

> You're saying that we should have left medical decisions in the hands of the local purveyors?

Not what I was trying to say; I apologize if I was unclear. Rather, my point was that during times past, in many cases it would have been better if doctors had not applied any treatment to patients. Not because doctors were fools, but simply there wasn't enough rigor around assessing the effectiveness of treatments.

So too do I advise this for the field of economics. Not to say that we shouldn't strive for rigor, but until we have it I recommend the "doctors" (i.e. economists) refrain from applying treatments to their "patients" (i.e. the nation via policy prescriptions).

> Soon we'll be in the era of evidence based economism, and we'll see a huge increase in the effectiveness of it's utilization.

An optimistic view that I do not share. I recommend the talk "Science, Knowledge, and Freedom" by Jim Manzi[0]. In short, observations in the "soft sciences" of psychology, sociology, and economics do not generalize across different times and cultures in the way that observations about physics, chemistry, and biology do.

[0]https://www.youtube.com/watch?v=N4c89SJIC-M

[lliamander]

Indeed. It would be nice if it were a choice between government by Science vs. Superstition, because then the answer would be easy.

In reality it's more like "psuedo-scientific mismanagement" vs. "mostly civilized confusion". Neither of these scenarios is particularly appealing, but I'd rather take the confusion.