Economics has treated econometrics as a universal solvent, a technique that can be applied to any economic question, which is sufficient and, therefore, makes other applied techniques redundant.
Peter Swann in his book indicates the place of econometrics and argues against this notion and even takes this as a severe error. He advises fellow economists that they learn to respect and assimilate what he calls vernacular knowledge of the economy. His top message to economists is what the great French composer, Paul Dukas, advised his pupil: “Listen to the birds, they are great masters.” If any fellow economist asks: “don’t most economists do this already?” Then the answer by Swann is clear: “… some economists do use vernacular knowledge some of the time to underpin what they do … incidentally to make a piece of high technique more approachable … outside this limited context, economists do not tend to take the vernacular seriously."
Any argument for or against it?
To me the relation between econometrics and economic theory (or to be more appropriate the set of the discipline of economics other than econometrics) resembles the relation between tests of medical pathology and clinical medicine. A doctor anticipates the cause of the disease of a patient on the basis of her knowledge of medical science (bio chemistry, anatomy etc.) and confirms it by pathology tests. A good clinician can only do this. The doctors who are very poor in clinical diagnosis (there might be other reasons, especially you are from India) only depend on pathological tests. To these people pathology tests are something sacrosanct which it is not. Now you understand what it means. In this context a renowned economist's warning about use of econometrics may be noteworthy: if you find some result(s) via econometric analysis which is not consistent with your knowledge of theory then the result should not be0 taken seriously. (I have forgot the source).
Dear Prof Hans-Georg Petersen
According to G. M. Peter Swann, it is all about "Putting Econometrics in its Place". His main argument is that there are also other essential techniques to be considered. He is proposing a new direction and a new attitude to applied economics. And you?
Econometrics and statistics are standard economic subjects and always will be!
I think the question is very relevant. Applied Econometrics has been used extensively to address many economic issues that the economics discipline would not had been able to address otherwise. The contribution of both theoretical and applied econometrics is colossal and has raised the discipline of economics to unprecedent highs. As a result most
empirical studies focus on applied econometrics without considering other
methodologies that may be more relevant. A side effect of this trend is that often applied econometrics is not employed correctly leading to misuse of this empirical methodology.
Dear Professor Petersen, I agree with you on econometrics, but statistics are currently strongly differentiated in many areas (economic statistics, social statistics, demographic statistics, business statistics, medical statistics, mathematical statistics, and much more).
Econometrics is merely a set of statistical methods which are often used for analysing economic data, but can be used and are used in other scienticic disciplines. One of the most important method is e.g. OLS (nowadays mostly used in "refined" forms): OLS was first developed and applied by Gauss around 1800 to calculate the the orbit of Ceres and, later on, for land survey. That is 130 years before the expression "econometrics" has been invented. Of course, without computers, it was to troublesome to estimate OLS for several variables and many data. Nowadays, the exonometric packages are so comfortable, that many econometricians just do not care about economic theory and model specification, and foreget to use their own brain before starting estimation.
Econometrics is, therefore, not a universal solvent, but a statistical tool to analyse relations between economic variables (sensible in combination with theory and common sense and after carefully looking at the content of the data used).
Your question Melkamu Tadesse Wazza is decisive, with respect to choose the ‚right‘ economic method, when working on a certain problem. GMP Swann and his critical works into ‚economic precision‘ are important contributions to ‘scientific scepticism‘ , in terms of academic (official) economics and the development of the whole discipline.
My methodical point is identical with J.A. Wheeler:
Never calculate without already knowing the answer,
i.e. insight---which, unlike precise calculation, can fit inside our mind whole---is the seed of discovery and progress.
—————
The old Soviet joke applies: the military parade is headed by people in civilian suits. The president asks: what does that mean? Sir, these are economists, they can destroy every foreign economy faster than hard machinery.
—————-—
Economic statistics are like a bikini, what they reveal is important, what they conceal is vital. So: If you torture the data long enough, Nature will confess. -Ronald Coase
Dear Anton Rainer, econometrics is not a universal scientific direction. And Statistics is now a universal science. Statistical methods are applied by all sciences.
Dear Simon, you confirm my contribution. Statstistics was applied by many sciences long before the term econometrics was invented. Like mathematics (one could, of course, say that statistics is part of mathematics), it is universal in the sense, that it can be applied for rahter different scientific studies. Nevertheless, it can only be applied, if there is a theory of the respective science.
'When' and 'where' to apply' Econometrics as a method is crucial to economic study. Yet, its modern application by some teachers and their trained students creates an impression that without mathematical presentation in the theoretical language of Economics, the content of one's research study is below standards, which such notion, is completely misleading.
We need to acknowledge that unique role of 'Economic Philosophy' and that of 'Computational Economics', with an in-depth understanding of the distinctive pathway and the complimentary role they play in the Body of Economic Science.
Dear Anton, I agree with you, except for one sentence: statistics in general is not part of mathematics. Statistics as a science was developed in Germany and it was called “State Science” (Staatswissenschaft). Part of mathematics is just mathematical statistics.
I think we are getting stumbled over something not that difficult. This is the difference between theoretical and applied mathematics , econometrics, statistics and economics. Regardless who discovered the mathematical theories, our deep understanding is that theoretical mathematics is the foundation of everything , all the pure sciences and the a[plied sciences such as the many fields in engineering that are based in theoretical mathematics. For long time both theoretical econometrics, statistics and mathematics are employed in economics. Econometrics is an extension of statistics, theoretical mathematics provides the means for the discovery of additional methods than econometrics to be employed in quantitative analysis. All the software used in applied statistics and econometrics are based in mathematical theories and is the work of mathematicians and now are readily available to professors ,students and other practitioners. Thus, all mentioned disciplines constitute components that play
crucial roles in the search of understanding economic and other relations.
Professor Leamer (2010), wrote a fascinating piece in an attempt to reflect our current obsession with mathematical/econometric modelling - about the travails (struggles) of econometricians -or what I call economagicians. Their holy grail is finding the lost world of Asymptopia in which the noisy behavior of the trolls and gremlins has been silenced by the law of large numbers. In the financial markets, noise can include small price corrections in the market as well as price fluctuations–called volatility–that distorts the overall trend. However, even in Asymptopia there is the danger that statistical models fail to account for a causal factor (what economagicians call omitted variables bias). Even worse, history has stopped in Asymptopia and the structure of the econoverse is forever frozen in place, which all know to be impossible. Consequently, the economagicians find themselves trapped in the same dilemma confronting the math modelers: they must choose between truth and mathematical beauty. Most prefer beauty because it pays better.
The problem for me is that the majority of those who use econometric methodologies have a very little understanding of economic theory..They probably do not even know what the difference between price and value is...the rest is just nothing but games that academics play to keep publishing useless academic papers that make no difference to our lives whatsoever ...
Leamer (2010), “Tantalus on the road to Asymptopia,” Journal of Economic Perspectives, 24 (2), 31-46.
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude