i have a math degree, and i didn't take any formal economics courses in an actual school until i went back to work on a second undergraduate degree in computer science. and, i started off in physics, so i'm fundamentally empirical in how i interpret reality. i remember having this exact reaction when i first looked at these post-keynesian models in a first-year classroom - even as the simplest, most basic model, the idea that this was even remotely useful in reality was instantly realized as laughable - and i actually was the person that asked the prof if there was any empirical evidence for this, and was at least lucky enough to get a derisive, scornful response that no, of course not, and even teaching it nowadays is just silly. that's really the most important thing i got from taking a few formal courses - you can call modern economics speculative philosophy if you'd like, but it often doesn't even pretend that it's applying scientific methods. it's a bad joke.

 

that said, there's a similar debate in physics about the utility of teaching newtonian mechanics and the overwhelming consensus is that it has too much historical value to drop it altogether. i don't even disagree, really. but, the classical models should be presented later in the course sequence as a historical breadth for experts, and the 100 courses should really be more contemporary, so as to be useful for both general audiences and math geeks taking breadth requirements.

but, steve keen apparently doesn't know what croutons are.

somebody buy the man a salad.