Home > Uncategorized > “data-free mathematical masturbation”

“data-free mathematical masturbation”

from Imad Moosa

Econometrics is no longer about measurement in economics as it has become too abstract. The word “econometrics” is typically stretched to cover mathematical economics and the word “econometrician” refers to an economist, or otherwise, who is skilled and interested in the application of mathematics, be it mathematical statistics, game theory, topology or measure theory. Baltagi (2002) argues that research in economics and econometrics has been growing more and more abstract and highly mathematical without an application in sight or a motivation for practical use. In most cases, however, mathematization is unnecessary and a simple idea that can be represented by diagrams is made much more complex and beyond the comprehension of the average economist, let alone policy makers.

Heckman (2001) argues that econometrics is useful only if it helps economists conduct and interpret empirical research on economic data. Like Baltagi, Heckman warns that the gap between econometric theory and empirical practice has grown over the past two decades. Although he finds nothing wrong with, and much potential value in, using methods and ideas from other fields to improve empirical work in economics, he does warn of the risks involved in uncritically adopting the methods and mind set of the statisticians. Econometric methods adapted from statistics are not useful in many research activities pursued by economists. A theorem-proof format is poorly suited for analyzing economic data, which requires skills of synthesis, interpretation and empirical investigation. Command of statistical methods is only a part, and sometimes a very small part, of what is required to do useful empirical research.

The trend towards more abstract work can be seen in the contents of Econometrica. In the 1930s and 1940s, Econometrica published papers on economics, dealing with microeconomic issues like the demand for boots and macroeconomic issues like the multiplier effect of a balanced budget. In the 2012 and 2013 volumes, most of the papers are too abstract, use no data and do not provide new econometric methods that can be used in empirical work. In particular there is a high frequency of papers on game theory, which is supposed to be a branch of mathematics. Recent issues of Econometrica are dominated by what a frustrated academic economist once called “data-free mathematical masturbation”, suggesting that it was not his “source of enlightenment” (Mason et al., 1992). This is why a joke goes as follows: during the rule of Nicolai Ceausescu in Romania, the government banned all “western” economics journals – the exception was Econometrica because it had nothing to do with economics.

Click to access Moosa88.pdf

  1. Frank Salter
    July 13, 2019 at 12:55 pm

    “Econometrics is no longer about measurement in economics as it has become too abstract.” — Not so!

    Econometrics is not about the abstract representation of quantities — quite the opposite. It is only about their concrete representation. It is the invalid assumption that quantities may be manipulated as simple numbers which allows econometricians to produce their invalid interpretations of the data. What is needed is to determine appropriate groups of dimension-one and find how the groups may be combined to gain theoretical understanding. This is how physical scientists have come to their understanding of reality.

  2. July 13, 2019 at 3:32 pm

    As an interested mathematician my ‘take’ is that Keynes’ critque of econometrics still largely stands, but that the solution is not to veer away from mathematics, but to pay more attention. Keynes’ insights at the very least contributed to the development of measurement and model theories. In essence, it is dangerous to assume that everything is measureable or modellable in the sense that econometricians implictly assume, and in the case of economic data it seems ‘obvious’ that their assumptions are wrong, and it seems to matter.

    I guess I’m supporting Frank, above, but coming from a different angle. At https://djmarsay.wordpress.com/decisions/study-guides/4129-2/ I point to some insights of Peter Allen that come from yet a different viewpoint. I guess you don’t need to be a mathematician like Keynes to appreciate that there is something wrong with the mainstream status quo: but it isn’t the use of mathematics as such, rather what Keynes called ‘pseudo-mathematics’.

    • Frank Salter
      July 13, 2019 at 4:28 pm

      I do not think that there is such a thing as “pseudo-mathematics” per se. I know of one economics paper referenced more than twenty thousand times which contains an elementary mathematical error which invalidates the analysis. That all the other authors of the papers referencing the original were accepting of the mathematics and were unable to detect the very obvious mistake and its ramifications makes a serious statement about the mathematical abilities of economic researchers.

    • Frank Salter
      July 13, 2019 at 4:41 pm

      Another thought: A mathematically correct paper showing that an invalid mathematical conclusion had been made was rejected by the Economic Journal — they did not accept papers disagreeing in this way. This paper has been cited nearly twenty thousand times.

      • A.J. Sutter
        July 14, 2019 at 11:21 am

        First, a question: would you kindly identify the 20K-cited flawed paper you allude to?

        Also, a proposal for a candidate example of “pseudo-mathematics”: the use of Cobb-Douglas production functions without any specification (much less *consistent* specification) of the dimensions or units of the variables.

      • Frank Salter
        July 14, 2019 at 12:14 pm

        To A. J. Sutter: I will be writing a paper, in the near future, in which it is appropriate to identify the mathematical error in the context of my analysis. I not wish, therefore, to start rabbits running which are not under my control. If you contact me directly we can discuss the error.

        I agree that all production functions can only fit specific data. The manipulations involved with all of them are forbidden by the quantity calculus. Therefore they can be concrete representations of the specific data but not valid abstractions of reality.

  3. Helen Sakho
    July 14, 2019 at 3:08 am

    Econometrica might have been correct at the time.
    The most idiotic dictators are the ones that die surprised when shut dead. As for the separation of the micro and the macro, this is the most unnatural part of the equation. I believe that one can learn more from the woman/man begging in the streets for food than from this dismal voodoo “science”. It should be abolished altogether.

  4. Ken Zimmerman
    July 21, 2019 at 10:03 am

    In this instance, I must take the side of Moosa. Mostly mathematization is unnecessary in the social sciences, including economics. Except for numbering and geometry (which as any mathematician will tell you, is not really mathematics). Mathematics’ only significant contribution to the social sciences is to mystify them beyond the comprehension of social scientist and layperson alike. Particularly troubling for policy makers. Socio-cultural life is created by humans to serve specific and sometimes obscure purposes. Mathematics provides no assistance in revealing, showing us this creative work of humans. Nor does it help us view more clearly the results of that work. Mathematics’ role in social science is heuristic only. This also argues against using a natural-scientific approach to the study of culture and society. First, the assumption of logical reasoning in the natural-scientific approach clearly is not met in socio-cultural life. Second, empirical is not the same for the physical and socio-cultural. For me, this argues that the natural-scientific approach should not be used for the physical sciences either. But we cannot go into that here.

    One of the alternatives to the natural‐scientific model proposed for the socio-cultural and the associated mathematics is description of the institutional consistencies and conflicts that would otherwise be subject to mathematical explanations. Thereby the institutions (including their meaning) created by humans are not replaced by mathematical models. But as I said at the beginning mathematical models themselves are often useful counting, numbering, and through geometry providing a shape for socio-cultural relationships.

  5. Craig
    July 21, 2019 at 6:14 pm

    Like virtually everything else in life integrating the abstract and the the dynamic observation of the moment will yield more enlightenment of truth than merely one of those two modes of thinking/experiencing.

    Whoring only after math, statistics and abstraction is one of the reasons economists for the last century have missed the direct observation of the entirety of the economic/productive process itself and hence not cognited on the potentially incredible leveraging power to be utilized at its terminal ending point where a simple accounting and algebraic operation regarding monetary policy could simultaneously resolve the two most troublesome problems of modern economies (that all heterodox economists agree upon by the way), namely individual monetary scarcity and chronic asset inflation.

    Economies are very complex, but the actual operations of historical paradigm changes are always quite simple. Economists need to look at that fact and act….not merely go on a stupendous figure, figure, figure with math and/or abstract theoretics.

  1. No trackbacks yet.

Leave a Reply to Frank Salter Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.