I have screamed for the last several years that the problem with economics isn’t that Macro is problematic; DSGE is a detour but—as Noah Smith shows here—it can be used as an access road.
The problem is assuming that “micromotives” must be universal. For all of the emphasis about the glories of using natural logs because they reflect “what happens in nature,” economists who insist that micro principles must be foundational to macro behavior
are idiots fail to study nature.
Now, via Michael Swanwick (whose wife Marianne taught me more about China in ten minutes than all of the analysts and economists who wrote about it did in the previous ten years), we see that even robots understand that individual behavior is not sustainable in a society.
If only economists could (re-*)learn that, they might get back on the road to being a science, instead of being lost in the woods of a perpetually self-indulgent wankfest.
The flavor of this discussion is indescribable. In its total estrangement from our political and social life today [and] its wilfull disregard of all known facts about human nature.
You cannot have a “social science” if your foundations are anti-social. No matter how many differential equations you use.
*I’m fairly certain it is a matter of re-learning. The Kuznets Curve, for instance, indicates an understanding that collective action does not reflect just “the sum of the parts.”