I’ve linked to this article before, but it is very good so here it is again.
What many more quantitative-oriented people miss (sometimes in the most comical fashion) is that just because you’ve used calculus to solve some self-posed problem or tossed some statistics at something, if you have poor data, ridiculous assumptions, have misunderstood the problem or otherwise engaged in some other fundamental cock-up, you have learned less than nothing.
Because you have wasted time and increased the entropy of the universe, you are now actually behind where you were when you started.
In the case of economics, its history since the 1930s has been to focus on making the real world conform to increasingly-complex models rather than assaying that world as it is.
This happens in many fields, but by far eonomics has been the most detrimental to the world — more destructive than many wars.
The illusion of veracity can be achieved by mathing something up. Making a field opaque with unnecessary intellectual hurdles is a great way for practitioners to exclude others who might question their conclusions (“You can’t understand transfinite extension of the mu-calculus, therefore your opinion on why the rich shouldn’t be allowed to own you is invalid!”).
Stating an equation doesn’t magically transubstantiate questionable assumptions to verity; running regression analysis on a data corpus doesn’t reveal anything when your input is garbage.