Thursday, December 5, 2013

Brad DeLong -- sensible thoughts on "microfoundations"



The one thing about modern macroeconomics that I find really hard to comprehend is that theorists seem to jump through hoops to get models with a certain kind of mathematical consistency in them, even though this guarantees that the models make very little contact with reality (and real data). In fact, it guarantees that these theories are based on sweeping assumptions that we know not to be true in the real world. I've written about this mystery before. A theory having "microfoundations" -- which macro theories are supposed to have to be considered respectable -- is thereby a theory that we know is not consistent with what we know about real human behavior and economic reality. Economists, it seems, only allow themselves to believe in things they know NOT to be true!! How wonderful and counter-intuitive! (I can see why the field must be appealing.)

Imagine if the engineers building and managing modern GPS and other global navigation systems were, for bizarre historical reasons, constrained to go on using a map of the globe essentially like that pictured above: The Square and Stationary Earth of Professor Orlando Ferguson.

Brad DeLong has an interesting comment on this topic today.

6 comments:

  1. Isn't the point of microfoundations simply that provided we build a coherent macroeconomic edifice upon the foundation of our current understanding of microeconomics, then any new information that allows us to improve our model of the economic behaviour of individuals (eg by incorporating increasingly sophisticated social psycological predictions as to economic behaviour), then the macro model will improve in step, automatically, providing us with a much larger scope for predictive ability - and thus the ability to spot flaws in new theories in a more efficient and consistent manner.

    ReplyDelete
  2. I think this ought to be the point of microfoundations, yes. And this would be the case if microfoundations as currently considered were based on an authentic empirical picture of how people behave, in some crude approximation. Instead, it is generally based on an axiomatic approach based on a vision of rational behavior and intertemporal optimization, which doesn't have much empirical basis at all... does it?

    ReplyDelete
  3. Well, no. Trying to do microeconomics without a better grasp of social psychology than we currently have is like trying to do statistical mechanics without including quantum theory. You might get some of the simple stuff right, but try anything more difficult and the theory just falls to pieces.

    None of this contradicts the purely methodological justification for attempting to keep micro and macro as closely interlinked as possible though. It may not give us the right answers (currently) but it IS the right method through which to search for better answers, from an academic point of view.

    ReplyDelete
  4. It seems like modern economic theory is a bit like the geocentric theory of cosmology. On the surface it seems to make sense of most of the data we observe ("sun and stars seem to be orbiting around the earth"), but whenever new variables pops up (like the parallax effect of stars), whether from behavioral economics or any other discipline (sociology, evolutionary biology, etc.), economists simply cannot come up with a sensible answer that would explain these phenomena.

    What's necessary therefore is a revolution in the way we think about the economy (or humanity) as a whole and our place in it. A paradigm shift that, in scope, would parallel the move from a geocentric theory of cosmology to a heliocentric one - a radical difference in how we perceive the universe and our place in it, yet a view that is much more consistent with reality.

    Consider the Flow Economy (www.geopolitics.us/toward-a-flow-economy/) as such a economic paradigm. It effectively addresses most if not all the inherent problems of the current economic paradigm and puts us on a path to sustainable economic growth and prosperity.

    ReplyDelete
  5. I find really hard to comprehend is that theorists seem to jump through hoops to get models with a certain kind of mathematical consistency in them, even though this guarantees that the models make very little contact with reality (and real data). borrowing options

    ReplyDelete
  6. I suppose I've selected an unbelievable and interesting blog.



    Financial Adviser Bournemouth

    ReplyDelete