>Doug's specificly mentioned models of boss-worker relations. I do
>remember that Bowles and Gintis would have been happy to find
>quantifiable variables or their proxies, but never forced that into
>their work They instead do what most economists do--just posit some
>general (and simple) functional form so you can say stuff
>axiomatically about the relations (i.e., paying a worker above the
>rate in which they would put minimal effort will mean more effort).
>You would have to attack their choice of functional form or the
>shape of their curves (it's all about the choice of concavity vs.
>convexity and creating lemmas). The more advanced math stuff just
>posits more sophisticated functional forms onto variables and their
>relations (I agree that here it is easy to apply sophisticated
>functional forms to unsophisticated ideas or inappropriate proxies).
>I personally always try to use the simplest funtional form I can for
>any work I do. Some people have the opposite problem.
>
>So, I find it strange that anyone would find it inappropriate to use
>general math functions to say radical or neato things.
No it's not. There is absolutely no way on earth to represent accurately the kinds of things B&G are trying to represent in their models. The shape of the curves resulting from such equations is entirely a function of the assumptions you build into the models, and the kinds of numbers you plug into the variables.
No "advances" in math can compensate for the things Keynes brought up in his critique of Tinbergen's early econometrics: "What place is left for expectation and the state of confidence relating to the future? What place is allowed for non-numerical factors, such as inventions, politics, labour troubles, wars, earthquakes, financial crises?.... [H]ow far are the results mechanically and uniquely obtainable from the data, and how far do they depend on the way the cook chooses to go to work?"
Doug