As y'all know this is something I rant about often, so its nice to see someone else doing the same over at Gene Expression. They also get into the criticisms of math in econ, in the post and the comment sections. On that subject, I commented:

I think there are two different views of math in economics here. One can criticize the obsession of modern economics with complicated math without denying that math is a wonderful tool for solving problems, and if used correctly, for understanding the world.

I agree with godless that denying the role of *some* math in econ is just ridiculous. Even if you believe in pure empiricism and no model-building, we still need Statistics to understand our empirical results. Math is useful!

But this does not at all contradict the point that a lot of economics is very divorced from the real world, and sometimes focuses on math puzzles instead of understanding humans. As someone with a math degree and a great love for econ, I find the idea that you can "prove theorems" about a real economy to be laughable. You have to make assumptions which are approximations, you have only proved something about your model, not the real world. This may be useful, but don't fool yourself that you've definitively "proved" something.