[OPE-L:6201] Re: econometrics

From: Gerald_A_Levy (Gerald_A_Levy@email.msn.com)
Date: Mon Nov 19 2001 - 10:35:33 EST

Re Patrick's [6196]:

> Econometrics "is after all a branch of neo-neoclassical analysis."
> Sorry, Jerry, this is just plain wrong.

In what sense am I 'just plain wrong'?

Let's consider the matter from a number of perspectives:

1) historical
       Is it not the case that econometrics was pioneered by
    marginalists? Haven't they been its greatest champions all
    of these years? Wasn't the journal *Econometrica* created
     [in 1933]  and nurtured by marginalists? (An interesting footnote,
     btw,  is that the first editor of *Econometrica*, Ragnar Frisch,
     was also the first to employ the terms 'micro' and 'macro'
     in economics).  What percentage  would you guess of all
     econometric literature has been developed using explicitly
     neo-classical terms and assumptions?

2) if not, then what?
     If econometrics is not a branch of neo-classical economics,
     then where are the anti-marginalist econometric studies?
     E.g. where are the Post-Kenyesian econometric studies?;
     where are the Marxian econometric studies? (by which I
     mean specifically Marxian studies rather than studies by
     Marxists. E.g. Desai is a Marxist, but I don't really
     consider his work *Testing Monetarism* to be a
     Marxian study. Riccardo, btw, was the first on this list
     to make that claim).

3) critique and critical appropriation
     If econometrics is not a branch of neo-classical economics,
     then where are the *critiques of econometrics* from heterodox
     economists which identify the failings of mainstream
     econometric theory and list what must be rejected and what
     must be critically appropriated and surpassed?

 4) just a technique?
      Is econometrics just a 'technique' which is theory-neutral
     and can be adapted by economists from divergent
     theoretical perspectives?  I find that claim (which you did
     not make) to be naive.  Indeed, one can think of econometrics
     as technique but it is a technique which has been both
     historically and logically tied to marginalism. In the same way,
     one could claim that cost-benefit analysis is a mathematical
     technique.  Yet, isn't cost-benefit analysis a by-product of
     marginalist welfare analysis and all that it entails? Is it also
     a 'game' that we should play?

 5) method
     Do you not think that positivist and empiricist logic underlies
     the field of econometrics?  More ominously, econometrics
     has been the technique by which marginalists have 'proven'
     their postulates.  Is there not a high ideological content to
     many of the presumptions of econometrics?

     For Allin: you claim in [6198]  that econometrics is a
     "methodology for empirical analysis".  In what sense should
      we then *share* the same empirical analytical methodology
      with marginalists?   How could a field created by marginalists
      not reflect marginalist methodological presumptions and

     What percentage of all econometric studies would you guess
      rely on *linear* models?  Yet, how many of the real-world
      subjects that econometric studies claim to analyze can be accurately
      characterized as being linear and can be grasped by linear
      models? (btw, I think that Allin was wrong in [6198] to claim
      that chaos theory is a 'theoretical  framework'  whereas econometrics
       is [just] a 'methodology for empirical analysis': don't methodologies
       for  empirical analysis *require* theoretical frameworks to give them
       meaning?;  can a methodology for empirical analysis developed using
      one theoretical framework then be adapted to other theoretical
      frameworks or does the empirical analytical methodology embody
      certain core concepts of its theoretical framework of origin?; since
      econometrics was born from the womb of marginalism how can it not
      bear its stamp and theoretical lineage?).

   6) the only game in town?
        Allin claimed in [6198] that econometrics is the only [quantitative]
       game in town. To begin with, I don't think that's a very good
       argument for playing that game. At the very least one might
       anticipate that all those who believe that econometrics represents
       a 'methodology for empirical analysis' would critically consider
       the theoretical presumptions of that method before joining the
       game. (Of course, on a more practical level this is not an
       option for most economists since they know, due to the
       hegemony of marginalism,  that econometrics is the
       'coin of the realm' and thereby serves as a 'measure of value'
        for both economists and their work.  So for reasons of
        career advancement, economists must practice econometrics
        or at least take courses in 'how to play the game' before
        being admitted into the fraternity of economists. This may be
        a practical reason for 'playing the game' but it can not be said
        to be a  legitimate defense of the game itself).

        In [6200] Andrew B mentions Tony Lawson. Lawson, of
        course, works for the celebrated Department of Applied
        Economics (DAE) at Cambridge University.  Is econometrics
        the 'only game in town' at the DAE now? Maybe Lawson and
        others have come up with better 'games' to play?

In solidarity, Jerry

This archive was generated by hypermail 2b30 : Sun Dec 02 2001 - 00:00:06 EST