David F. Hendry, 1980; Econometrics: Alchemy or Science?
本文为计量经济学理论大师Hendry的经典论文，从幽默而自由的行文来看应该是Hendry当年在 the London School of Economics的一篇讲座记录，目的是介绍计量经济学。文题主要源于著名的凯尔斯对Tinbergen的批判，"That there is anyone I would trust with it at the present stage, or that this brand of statistical alchemy is ripe to become a branch of science, I am not yet persuaded". 所以文章从介绍炼金术与科学的定义开始，到介绍计量经济学，它的缺陷所招致的各种批判以及各种回应。当然，最后他说：I commend its study to you.
1. Alchemy and Science
By general definition, alchemy is an art of transmuting base metals into noble ones. It has played an important role in the European history, for the development of European religion, chemistry, and also philosophy (alchemy can also be understood as the process of turning base souls into noble ones). Science, however, is the inductive and deductive process between data or phenomena and theories/concepts. By Stanislas de Rola (1973), "true" alchemy is actually a secret art striving for the "absolute".
2. Econometrics
Its main objectives is "the advancement of economic theory in its relation to statistics and mathematics". The necessity of econometrics mainly originates from the absence of experiments in economic study, which promotes, according to Frisch (1933), the division between data collectors, data users, econometric theorists and mathematical economists.
3. Econometrics as Alchemy
Hendry uses a typical case, where econometrics step-by-step refines a model to fit the relationship between some exogenous variable C and P (prices) , but it finally turns out that C is simply cumulative rainfall. The powerful but dubious results shade much doubt on econometrics as being scientific: "caveat emptor"（一经售出，后果自负）. Nevertheless, this can't denigrate the decency of the whole subject, because the false models can be explained inside the discipline.
4. Econometrics' Problems
Lots of problems of the linear regression model are listed: using an incomplete set of determining factors; building models with unobservable variables, estimated from badly measured data based on index numbers; obtaining "spurious" correlations from the use of "proxy" variables and simultaneity as well as the "mine sprang under the contraptions of optimistic statisticians"; being unable to separate the distinct effects of multicollinear variables; assuming linear functional forms not knowing the appropriate dimensions of the regressors; mis-specifying the dynamic reactions and lag lengths; incorrectly pre-filtering the data; invalidly inferring "causes" from correlations; predicting inaccurately; confusing statistical with economic "significance"of results and failing to relate economic theory to econometrics.
Hendry also indicates that considerable progress has been made (till 1980). By quoting Leontief (1971), "the scientists have their machines while the economists are still waiting for their data", Hendry stresses the inability to experiment in economics again, and blames the limited government budget invested in economic studies.
With further explanations of the structures of econometrics, Hendry says that:
"Econometricians may well tend to look too much where the light is and too little where the key might be found. Nevertheless, they are a positive help in trying to dispel the poor public image of economics (quantitative or otherwise) as a subject in which empty boxes are opened by assuming the existence of can-openers to reveal contents which any 10 economists will interpret in 11 ways."
扫码加好友，拉您进群
全部版块
我的主页
收藏
