delong at econ.Berkeley.EDU
Wed Oct 11 14:44:40 PDT 2000
>At 12:53 PM 10/11/00 -0400, Doug quoted:
>>academy said by way of example. "Prior to McFadden's contributions,
>>empirical studies of such choices lacked a foundation in economic
>>theory. Evolving from a new theory of discrete choice, the
>>statistical methods developed by McFadden have transformed empirical
>Prior to McFadden's contribution facts were inconsistent with the
>infallible economic theory. McFadden invented a gimmick (or a
>"problemshift" in Imre Lakatos' lingo) to massage the facts so they do.
>This is an important contribution, because economic theory is a form of
>managerial ideology that creates an illusion that captains of industry and
>people who call themselves government know what they are doing and are
>fully in charge of the situation. The facts that are inconsistent with the
>theory undermine that illusion and thus must be assimilated.
>Indeed, as John Kenneth Galbraith aptly observed, economics is the business
>of providing needed conclusions to those in the position to pay for them.
>That is why they have a "Nobel" prize in economics and not, say, in social
>sociences or, heaven-forbid, sociology whose main business is debunking
>such needed conclusions (or "translating" them as C Wright Mills used to say).
Say, rather, that McFadden built a framework that you can use to
estimate what demand will be when people must make discrete
choices--whether to commute by car or by subway, for example. And
Heckman built a framework that you can use when you don't observe a
substantial chunk of your sample for reasons that depend on the
choices they do make.
More information about the lbo-talk