On microfoundations

Some recent post on the necessity of microfoundations that I found interesting.

1. Issues about microfoundations.
2. Why you don’t need rigorous microfoundations.
3. Update: Zombie Marx – illustrating the similarities between Marxism and Neo-classical economics.  ht Guan.

I agree with these in part, especially when it comes to macroeconomics (whether this is appropriate depends on your view on the nature of how we should treat macroeconomic aggregates).  But as with all things it is a matter of balance.

To quote myself (as I am a touch short of time to find new words):

Even given this, in truth macroeconomics still needs to be trying to understand the behaviour and structure behind any relationships they observe (or at least the behaviour and structure they are assuming when coming up with policy advice) – and this corresponds to the central research program of macroeconomics as it stands.  Many of us – me included in my weak moments  – keep trying to find relationships that seem either theory or history invariant, where we can just either use the data or the concept and say “this is what will happen”.  While this is a useful heuristic, this isn’t what these macroeconomic scientists are (or should be) relying on!

This is especially important for the social sciences, as compared to the physical sciences, given that our “target” is unobservable and its relation to measurable targets can only be worked out by (implicit or explicit) reliance on some model.  Transparency about our micro-macro relations which define not just the way we can predict a measurable target, but also its relation to our true unobservable target, captures an important element of why some form of “microfoundations” are very important.

A team of physicists can get us to the moon, but can’t tell us whether we should go.  Economics, and social sciences, are not just constrained because they can’t prevent certain social occurrences (recessions) – but because even if they could, its relation to what we value and whether its something society should do is very unclear.  This is why economists get obsessed with very specific questions, as the nature of value forces us to look at things in very specific, and conditional, ways!

In terms of monetary policy (which is different from most other economics), market monetarists would likely go as far as saying we can avoid (demand side) recessions, but the concern of economists about the related policy actions, and the inaction of analysts and politicians, is what ensures that we don’t.

Update:  On thing that has occurred to me following the above Marxist piece, and this conversations between economists:

@TVHE I hope so. Agent based models seem to be pretty good at approximating the stylised facts. Is the era of analytical solutions over?

— JamesZ (@jzuccollo) October 23, 2013

Is that the question of “appropriate microfoundations” makes close to zero sense unless we can figure out what question we are asking, and trying to answer, in the first place.  I suspect that much of the disagreement stems from the suspected scope and purpose of given questions in the social sciences!  We touched on this idea above by stating that economists are trying to answer very specific, conditional, statements – as Arrow points out here.