That’s an interesting summary of the typical textbook’s layout. I’ll say something about it in a bit, but I’d like to play devil’s advocate first, because I think Keen would maintain that you’re understating the problem.

In Chapter 3, Keen shows how the Sonnenschein-Mantel-Debreu Theorem was systematically buried or obfuscated, beginning with its very discoveries by Gorman and Sonnenschein, and continuing in presentations of it by standard graduate and undergraduate textbooks from the 1990s (Varian, Mas-Colell, and Mankiw). And the obfuscation happened in English statements of what the theorem means for the overall theory.

I wonder if you could show what recent graduate and undergraduate textbooks say about that theorem. Could you quote where they state in English (not math) its relevance to the Law of Demand and the aggregation problem, or to market demand curves?

I think the overall problem, though, is that the crazy “models” are still taught at all in any economics textbook. Science textbooks no longer teach the geocentric model, the ether, or vitalism. They replace old models with new ones, thus advancing the state of knowledge.

I’m looking at your layout of economics textbooks and I’m wondering how that kind of treatment would be different from assuming a crazy or naïve model, such as that the Earth is flat, and “relaxing” the conditions to account for the apparent contrary evidence that the Earth is round. If the model can account for the contrary evidence only by shedding its ludicrous assumptions, why doesn’t that process demonstrate that the model is false?

If I say, “The Earth is flat, ceteris paribus,” meaning that if you assume there’s an international conspiracy to generate the mountain of evidence that shows the Earth is round, you can account for all the contrary, roundness data, how is “The Earth is flat” still a model at all? The ceteris paribus clause must deal with exceptions, not norms. If normally the Earth is known to be round, saying it’s really flat is a theological, not a scientific statement.

Likewise, if the Law of Demand can’t be aggregated unless you assume there’s only one commodity and one consumer, why isn’t that proof that the law doesn’t hold for societies? Why isn’t redefining “society” as “a group with one member” just as absurd as the assumptions needed to sustain the proposition that the Earth is flat?

If neoclassical economics has been refuted, as I think you said earlier it has been, and economics is a science, neoclassical economics shouldn’t be taught at all in economics textbooks. (It could be taught in history of economics classes, but that’s all.)

The fact that that textbook you provided presents neoclassical and Keynesian “perspectives,” then, indicates that economics isn’t exactly a science in the first place, which means the very concentration of math and the talk of “economic models” is already just dress-up and obfuscation. It’s like a geology textbook saying there’s the “Earth is round” perspective, and then there’s the “Earth is flat” one. Each perspective has its “models,” and we go from there.

That textbook of yours says many mainstream economists are both Keynesians and Neoclassicals (329). They think it’s a difference only in time scales, so that for short run questions they’re Keynesians and for long run ones they’re Neoclassicals. As Solow is quoted as saying, “At very long time scales, the interesting questions are best studied in a neoclassical framework, and attention to the Keynesian side of things would be a minor distraction.”

But if the Neoclassical framework has been refuted by economists themselves, why is that framework still applied at all? If its assumptions that sustain it are crazy, why is the framework still treated as a scientific model in the first place? Why isn’t it more like theology, QAnon conspiracy “theories,” or saying the Earth can be proven to be flat if you accept a bunch of crazy nonsense (pseudo-ceteris paribus conditions)?

This doesn’t seem like just a pedagogical problem. It’s about the nature of economics.