Saturday, June 30, 2012

Microeconomics, Depedestalled


There has been a debate recently over whether and why macroeconomics is less scientific than its micro cousin.  I will leave macro aside for now and say, as clearly as possible, that micro is about as counter-scientific as one can get.  The problem is that utility theory, the idea that agents’ well-being can be measured by how much utility they have and that they go through life maximizing this ineffable U, is contrary to logic, evidence and the well-grounded findings of social sciences that actually study human decision-making close up.

Sen nailed the logic part decades ago.  Behavioral economics refutes the empirical presumptions.  Psychology, including its social and evolutionary branches, offers much more credible models.

Take away this utility maximization stuff and what’s left of micro?  There’s lots of excellent econometric technique, of course, and much of the aggregate theory (at the level of markets) still stands, but the agent stuff is at best a distraction and welfare economics is a zombie.  Conventional micro is propped upright only by the collective interest of its practitioners in preserving the value of their arcane skill set, and by ideological conviction.

Science it ain’t.

2 comments:

Bruce Webb said...

Micro not only fail's Popper's test for science, its main proponents never specifying what would count as "Falsification", it doesn't do well even on the older less strict Baconian measure of empiricism generally, that of verification. Instead to outsiders it's seems much closer to Ptolemaic Astronomy where failures in confirmation were simply (or not so simply in the end) 'solved' by adding a new set of epicycles to epicycles that 'corrected' the cycles all in a way that preserved the magic key: "it's all circles!"

I don't see much difference between that method and the ad hoc 'fixes' for 'utility maximization' and 'rational expectations' and 'marginal productivity'. I mean just because none of it seems to actually map one to one with the 'real world' (whatever that might be) doesn't mean we should abandon it. Because 'markets!' and 'shut up!'

Will said...

Bruce Webb:

You would do well to read up on the actual history of astronomy. "Epicycles upon epicycles" is a hoary myth, just like the notion that people thought the world was flat prior to Columbus or the story that Napoleon was short. In fact, the earth-centered astronomic model underwent virtually no changes up to the time of Copernicus.

I do agree with your point. The circularity of utility theory makes it a mathematically elegant model with no predictive power.