[drewcrawford]

This is a common misconception. It is "traditional" macroeconomics (e.g. Keynes) did not engage with microeconomics at any point [1].

It was not until the 1970s and the Lucasian rationalist critique that microfoundations became an important inquiry. And that shift didn't occur because people were sentimentally attached to micro (if anything, they were attached to prevailing macro theories), it happened because Lucasian rationalism explained an empirical phenomenon (a shift in the Phillips curve) that was not explained by the other approaches. Since that time support for microfoundations has waxed and waned a few times, in response to several other empirical shifts.

In practice, economics is much like the other sciences, with a slightly wider diversity of thought. Like the other sciences, most people only understand a caricature of it that is given through the lens of politics. Unlike the other sciences, more people seem inclined to accept this caricature and criticize it on that basis, which is unfortunate.

[1] Everyone, but see https://www.jstor.org/stable/2231707?seq=1#page_scan_tab_con... for a plausibly-readable overview

[neilwilson]

The problem is that the notion that 'microfoundations' can be linked to the macro economy by searching for 'deep parameters' suffers from precisely the same critique.

It's even proved wrong within the body of theory that embraces it. The Sonnenschein-Mantel-Debreu theorem.

Economists then do a lot of hand waving and point to 'data', without acknowledging that the 'data' is collected from a system that has been under the influence of their ideas for over 40 years and therefore will, inevitably, be the right shape.

If you collect data from a man in a straitjacket, unsurprisingly it will look like a man in straitjacket. It can tell you nothing about ideas that involve removing the straitjacket.

Non-falsifiability of theorems is the hallmark of macroeconomics. Because it is an ideology dressed up as a science.