Home > Uncategorized > Economic history — a victim of economics imperialism

Economic history — a victim of economics imperialism

from Lars Syll

Economic history was one of the first fields to succumb at least partially to so-called economics imperialism, the phenomenon through which the methods and models of economic theory have taken over other social scientific subject fields.  The economic historian’s craft, which probably always tended to err in any case towards retrospective reassurance that things were never as bad as they appeared at the time, was thus at least to some extent overwritten by the economic theorist’s often unflinching faith in the universal applicability of their explanatory models …

ecimpIn its orthodox form at least, economic theory reduces all behaviour to that which is consistent with the solution to a partial differential equation.  Such equations will no longer typically be visible in the resulting model, because to say that they are hardly at the cutting-edge of mathematical sophistication is obviously a rather big understatement.  But the presence of economic agents who act slavishly in this way is likely to be a necessary feature of those models.  They are always either maximising or minimising something, akin to the position of turning points of mathematical functions taking the form y = f(x).  No challenge is permitted to this conception of agency, because it is simply assumed that nature has ordained the economic agent to act in this way.  If they are not either maximising or minimising something, then they cannot be considered to have acted economically at all.

Economic history written through the perspective of this style of orthodox economic theory consequently has all of its behavioural uncertainty removed at the point of origin.  Historical uncertainty may still be admissible, but what of the scope for fleshing out the feelings of alienation that so often follow from being caught in a moment of intense historical uncertainty?  Presumably this has to be severely curtailed if every economic agent at every moment of time is blessed with pure certainty of behaviour.

Matthew Watson

  1. Frank Salter
    December 18, 2017 at 10:29 am

    I believe the Matthew Watson quotation to be a fair description of the understanding which most economists attribute to the mathematics found in economic theorising. However, is this the real truth of the situation. I think not. So can we unpick some of the misunderstandings prevalent in economic discourse?

    When tested, the statement “In its orthodox form at least, economic theory reduces all behaviour to that which is consistent with the solution to a partial differential equation” has no real validity. This is part of the sleight of hand practised by the majority of theorists by providing “mere[ly] corroborative detail, intended to give artistic verisimilitude to an otherwise bald and unconvincing narrative”. They introduce the terminology of Newton and Leibniz in their equation forms but the implied differential equations are not solved. Curves are fitted to the data and then causation is claimed — incorrectly! Usually only the coefficients of the fitted equations are tabulated. Graphs, showing the data and the fit, are relatively rare, unless the intention is to show just how much noise the data contains.

    Physicists and engineers will finish an undergraduate degree course knowing how to solve differential equation analytically and numerically. Then, in reality, they discover that the majority of the differential equations encountered have no analytical solution at all. For example the three body problem has no general closed-form solution for every case. Numerical methods are required. For partial differential equations, numerical solutions are the norm. This may be why economic theorists do not actually solve relevant differential equations. They would be forced into complex descriptions with many graphs — if they ever made such an attempt. I speak from experience. My first published paper presented numerical solutions to transient heating problems requiring thinking in five dimensions — you really do not want to try writing papers like that.

    So what is all this mathematics doing. Most of it is to prove a solution exists to “models” being hypothesised. The ubiquitous “assume a convex twice differentiable function” serves only to provide that a maximum value exists. Diagrams are drawn claiming to sketch some hypothesised solution to differential equations which are never solved. Some of the interpretations of these sketches are based on diagrams which are simply not true!

    So what needs to be done? It is: apply the scientific method and falsify incorrect hypotheses; test what remains against the empirical evidence; what finally remains will be new paradigms. These paradigms will be the ones which endure.

    • December 18, 2017 at 5:03 pm

      Even us applied mathematicians are products of our tutelage at the knee of our masters, who invariably share their notions of the history of “our” chosen field. I was taught that applied mathematicians modern (1970) view was that linear problems, no matter how complex, were boring because they were reducible to matrix inversion or to eigenvalue polynomials suitable for numerical methods. Moreover, uniqueness of solutions ensured that once a solution was found, you were done.
      Nonlinearity, on the other hand, was where the action is. Model equations like the Navier Stokes PDE describing fluid motion is a perfect example. The model is deterministic, it is exact, and it is not wrong at all. The equations describing fluid motion are nonlinear so the solutions are mind blowingly complex. There are multiple solutions as well, so each solution must be tested for stability to determine which solution we see in the real world. The number of solutions can vary with a parameter value. For example at low dimensionless velocity (Reynolds number) flow in a pipe is steady, parallel and stable; velocity perturbations return to steady parallel flow. At higher speeds the motion is metastable. Large enough perturbations cause transition to turbulent flow, a motion so chaotic and complex, it’s description must be reduced to giving statistical properties even though the governing equations are deterministic.
      Today, it is through a glass darkly that we still glimpse the potential power of nonlinear methods to describe our world. Progress is still hard. Stability, error estimates, chaotic dependence on boundary data, and numerical methods are still major topics. But few appreciate the predictive power of what’s known about dealing with nonlinearity now. Perturbation method is used to “linearize” weakly nonlinear systems. The quantum wave equation is one such, and the solution by this technique allows the design and fabrication of semi conductor devices.
      You young Turks reading this will not waste your time looking into the history of nonlinear analysis and listening to those like Steve Keen who did the same.

      • Frank Salter
        December 19, 2017 at 8:02 am

        As you have outlined, the physical world and economic reality are describable in terms of partial differential equations. This truth appears to be overlooked or possibly unconsciously avoided because of the sheer difficulty of dealing with PDE’s by most economic theorists. I have made a start in “Transient development” in RWER-81. As yet it has elicited little response, but the comments page for RWER-81 is still broken thus preventing comments from being made!

        I keep sounding off (and will continue so to do) about applying the scientific method to economic analysis. Possibly, if economists had better appreciated the power of the method and insisted on its application, neoclassical analysis would have been dead and buried by now. Logical argument and persuasion alone appear insufficient and papers such as Zambelli (2017) provide difficult reading. Possibly the many counterexamples available and sheer repetition will finally gain traction and allow economics to get out of its rut, and move forward with effective solutions to the many problems experienced by most of the world’s population.

        Reference
        Zambelli, Stefano (2017). “The aggregate production function is NOT neoclassical”. In: Cambridge Journal of Economics, doi:10.1093/cje/bex011. eprint: /oup/backfile/content_public/journal/cje/pap/10.1093_cje_bex011/3/bex011.pdf.

  2. robert locke
    December 18, 2017 at 10:57 am

    If econometricians produce evidence that the French steel industry was thriving in the decades before 1914, I am using a discipline mostly developed by people who did not live in France at the time. But if I read a report from German engineers in 1916 who studied the plants and equipment of the French steel industry, the bulk of which was in German hands during WWI, and these experts report that the French steel industry was backward, and those parts of it that were not replied on imported German technology — who am I to believe? The experts on the ground or the results of econometrics? I encountered this contradictory evidence all the time. And if I tell economics about it, they don’t even read my studies.

  3. Dan Andrews
    December 25, 2017 at 1:22 am

    Unemployment is measured and reported as number of people unemployed divided by the labor force. The labor force includes both employed and unemployed people. Unemployment rate equals unemployed/labor force. Alternative measures of unemployment include discouraged workers, marginally attached workers. People who are working part time for economic reason increases unemployment by 9 million since 2015. There are three different categories of unemployment, which are structural, frictional and cyclical. Structural unemployment exists due to mismatch of skill set of an individual and the skills required for a specific job. Frictional unemployment results from temporary transitions in the job markets with workers switching between jobs. This can also occur from workers not pursuing a job in hope of better wages or better job situations in near future. Cyclical unemployment can result from economic contraction in the economy.

  1. No trackbacks yet.

Leave a 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 )

w

Connecting to %s

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