Tom Walker wrote: >>>The critique has been that econometrics CAN'T show that the economic data are or aren't "consistent with one's theory".<<< saith I: >>No, the critique of econometrics (or at least the informed critique) says we can't prove anything using econometrics.<< Tom quotes Lakatos (on Duhem-Quine thesis): >'no experimental result can kill a theory: any theory can be saved from counter-instances either by some auxiliary hypothesis or by a suitable reinterpretation of its terms'.< and adds: >And, because econometrics is non-experimental, the scope for such theory-rescue is widened considerably.< The fact that "no experimental result can kill a theory" is due to the way "scientific research programs" (Lakato's term) are set up, not due to econometrics. But what if one doesn't participate in the dominant scientific research program? I know I don't participate in that program, since I think that that program reflects (in an alienated way) the class nature of capitalist society. I think that the critique of econometrics (much of which is valid) suffers from the same disease as the adulation -- the mystique, the fetishism -- of econometric which inflicts the official economics profession. The official economists think that econometrics is _wonderful_, a magic bullet for showing how _scientific_ "we" are. The antiofficials then trash econometrics as horrible, and assert that it should never be used, etc. But that's confusing a symptom with the disease. The basic sickness is with the economics profession (which reflects the basic sicknesses of capitalism), not with econometrics _per se_. To some extent, to trash econometrics in a unvarnished way is to distract us from the basic problem. I think that econometrics has been misused over and over again (and will be misused in the future). It's also not a "neutral tool," since it treat the empirical world as simply data. But with a little sophistication, it's better than tables of numbers or charts. It seems to me if one accepts Tom's root-and-branch dismissal of econometrics, one should reject all empirical research involving numbers. Jim Devine [EMAIL PROTECTED] & http://clawww.lmu.edu/Faculty/JDevine/jdevine.html Bombing DESTROYS human rights. US/NATO out of Serbia!
