[lbo-talk] "Bedlamite economics" and the EMH

Mike Beggs mikejbeggs at gmail.com
Thu Nov 12 16:23:18 PST 2009


On Fri, Nov 13, 2009 at 10:40 AM, Miles Jackson <cqmv at pdx.edu> wrote:


> The important distinction here is that the "doctrinal core" of a field like
> Chemistry or Physics has been empirically validated.  We know that the basic
> assumptions held by neo-classical economists are wrong.  One simple example:
> the notion that people are utility maximizers has been refuted in study
> after study in decades of psychological research.  It's bizarre to me that
> economists continue to cling to fundamental premises that have been
> empirically discredited.

Yeah - I agree. It's interesting to think about why that might be the case, though. If it's so transparently obvious that the assumptions are rubbish, why are they clung to? It has to do with the real structural complexity of the economic system, which requires a degree of rationalism - i.e., the building up of abstract models with internal logic - to think about. Also the uncanny half-quantifiability of economic phenomena - lots of numbers, though less commensurability than first appears.

It's funny how often economists who point the way to the relaxation of some unrealistic assumption - who look to outsiders like they are reinventing the wheel or stating the obvious - get so much acclaim from their field. Like Stiglitz and his Nobel for the economics of information asymmetry - the most cited economist in the Repec database. It is obvious that economic actors don't have perfect information, but what Stiglitz did was show how that fact could be incorporated into the models.

Nowadays behavioural economics is all the rage, relaxing the assumption of rationality, but, fatally, maintaining the methodological individualism, which I think is possibly the most damaging aspect of the core (vying with 'rational expectations'), as I briefly wrote here: http://scandalum.wordpress.com/2009/11/09/choleric-economic-man/

Cheers, Mike



More information about the lbo-talk mailing list