but then why are we told that supply & demand “sets prices” when there is no evidence of, for instance, a shock to the system upsetting prices which then follow a standard supply & demand curve to restabilise?

I feel like I've seen that in person. When the pandemic hit here, petrol prices hit the floor, like less than half of what they usually are. After a while they went back up as the production end was slowed down to match. So there's evidence, albeit not evidence you can really fit a whole curve to.

Well, because such a situation would not be “convex”, so economics has written an escape clause for this phenomenon whereby it can avoid ever making any real predictions about what the market will do in a situation that isn’t already “stabilised”.

It could be convex. That's not even an economics term, it's math. In terms of gradient decent - like a market adjusting - it just means that the curves are "round" as opposed to having weird little pockets that it can get stuck in. A counterexample would be something where decreasing the price actually decreases sales, with Veblen goods being an obvious example - nobody would buy Gucci if it was just slightly more expensive than a normal bag. Graphed out, that looks like a hump in a curve (supplier profit, I think), until Gucci bags come out the other side where they're cheap enough to compete as normal bags, as opposed to status items.

It creates a "local minimum" where the market can stabilise (it usually is -ise in British spelling) in a way that's counterproductive. You can talk about it in terms of second derivatives, too, but unless you're actually crunching numbers the marginal change of marginal price is unintuitive and unhelpful.

As you add more dimensions things get more complicated. Giffen goods happen when each good individually follows convex logic, but not when you rotate the resulting (hyper-)surfaces into a coordinate basis where cross elasticities become important. The math stays the same, though.

And also, if it’s just short line segments, why is it fashionable to use a curve that implies more data than exists? Again, I would say that it is to ape the aesthetics of science without actually doing it.

I mean, nobody's ever observed a part of the Sombrero potential other than the very bottom, but Higgs theory doesn't make sense without the rest. That on it's own isn't a point against economics, or QFT. In fact, the physics example is arguably more of a stretch, because it's supposed to be precise, as opposed to just a model.

You could critcise economics for only offering models, but that's all of social sciences, and will be so for the foreseeable future.

And it’s very strange that you turn to your personal anecdotal experience to say that you’ve “seen supply & demand play out”. If that were true, then you should be able to gather the data and present it. The fact that data doesn’t exist should tell you that you are probably using confirmation bias to evaluate your personal experiences.

I expect someone has been recording it, actually. I suppose I'll actually go looking out of respect for you.

For instance, the housing market is something where ideally there would be very inflexible demand, because people don’t need more than one bed, typically.

I know many people with a guest bedroom.

Selling a property you can’t rent out is a pretty bad financial decision, you’re better off waiting until the housing market flips and it becomes a seller’s market again.

Hah, no. Nearly as often the market will keep going the way you don't want, and the whole time you're not earning on your principle. Eventually, it has always gone up again, but there's no time limit on that - look at Detroit - and if the population stops growing in your country real estate becomes deflationary for good. Meanwhile, if you had just put it in an index fund you're going to earn 10% interest per year, on average. If you go bonds it might only be 4%, but it's almost guaranteed.

If you have literally any money saved, "don't try to time the market" is the first thing that every advisor will tell you. It's a dirty open secret that the big investors on wallstreet don't actually beat the market, either.

The reality of all markets is that they are subject to boom & bust cycles which supply & demand cannot account for.

At the small scale, yes. The economy is noisy. There are also cases where it does bad stuff at the big scale, but they tend to correspond to nonconvexities.

I'll look into Australia when I have the time. I have to go pretty quick, though.

And this is where the authors of that article talk about how orthodox economics institutions will launder heterodox theories. The point of doing this would be to maintain their prestigious position as the arbiter of all economics knowledge, as a result of which any damage such information might do to existing insitutions can be dampened by orthodox economists putting their own spin on the information. If they never acknowledge where the theory comes from, they get to rewrite it however they want.

That sounds right.

That’s distinctly unscientific.

Cynical and unfair? Definitely. Unscientific? No, academics is pretty much always like that; hard sciences too. You think you get tenure just by being the most honest?

Science itself is fine if worthy ideas make it to the top. Making the process fair is still a pipe dream. Gregori Perelman straight up refused the Abel Medal over this kind of shit.