My very
first post on this blog discussed the mathematical imperialism that has
recently infiltrated the discipline of economics. In that post, I distinguished
between pluralism in subject matter (e.g. paying more attention to issues like
inequality) and pluralism in methodology (e.g. accepting historical approaches
as legitimate), and argued that, at present, ‘New Economic Thinking’ relates
primarily to the former. I promised, moreover, to explain in a subsequent post
why I think that methodological diversity is also crucial to the renaissance of
economics. Over ten posts later, that is precisely what I intend to do now,
albeit by way of analogy, and in a rather tongue-in-cheek fashion.
Following Adam Smith, who is widely considered to be the ‘father
of modern economics’, I accept that some degree of specialisation – in this
case, with regard to methodology – is acceptable, and even desirable.
Mathematics and econometrics (hereafter ‘MAE’) certainly have their place, and
it stands to reason that the field of economics benefits when certain
economists choose to specialise in using these tools. Smith (1776, Book I,
Chapter 1), after all, argued that specialisation would increase productivity
by improving workers’ dexterity and by facilitating innovations at each stage
of production [1]. Specialisation in MAE thus produces better
mathematicians/econometricians and better maths/econometrics. For a number of reasons, however, Smith’s logic also implies
that this specialisation should not proceed indefinitely.
First, there are clearly limits – or at least diminishing returns – to the productivity gains that can be accrued by further specialisation in a given task or methodology. You can only get so good at casting or filing a pin (clearly, I don’t remember the actual stages of production involved in pin-making described by Smith), or, for that matter, doing MAE. Although innovations in casting or filing can surpass these limits, they are not costless – you might get away with paying a worker the same wage to do more work (or more productive work) in a given amount of time, but machines have to be built, purchased, operated, and serviced. The result is that a whole production process, including its own division of labour, emerges to focus on improving the particular task of casting or filing. A parallel in economics, for example, is that teachers require an increasingly higher level of training in MAE. Meanwhile, textbooks must be continually rewritten to reflect mathematical and econometric ‘advances’, giving rise to a vast industry in its own right, just as peer-reviewed journals must expand, change focus, or multiply to keep apace. At some point, the costs of this sub-division of labour must outweigh any benefits to economics of marginally superior (and increasingly less-superior) MAE.
First, there are clearly limits – or at least diminishing returns – to the productivity gains that can be accrued by further specialisation in a given task or methodology. You can only get so good at casting or filing a pin (clearly, I don’t remember the actual stages of production involved in pin-making described by Smith), or, for that matter, doing MAE. Although innovations in casting or filing can surpass these limits, they are not costless – you might get away with paying a worker the same wage to do more work (or more productive work) in a given amount of time, but machines have to be built, purchased, operated, and serviced. The result is that a whole production process, including its own division of labour, emerges to focus on improving the particular task of casting or filing. A parallel in economics, for example, is that teachers require an increasingly higher level of training in MAE. Meanwhile, textbooks must be continually rewritten to reflect mathematical and econometric ‘advances’, giving rise to a vast industry in its own right, just as peer-reviewed journals must expand, change focus, or multiply to keep apace. At some point, the costs of this sub-division of labour must outweigh any benefits to economics of marginally superior (and increasingly less-superior) MAE.
A more glaring problem with unlimited specialisation in
maths and econometrics, however, is that the whole point of specialisation, according
to Smithm is to allow for more production to occur for a given amount of
manpower. What we see in economics, by contrast, is the exponential proliferation
of economists specialising in MAE. This proliferation, moreover, does not
simply reflect the proliferation of the discipline itself, which would not necessarily
defy Smith’s logic; rather, it is occurring to detriment of other
methodologies. This relates to a third problem, which is that specialisation in
a given task is only beneficial within a broader
division of labour. After all, if specialisation in MAE is beneficial, then
specialisation in other methodologies must be also [2].
In my initial post, I alluded to the proverb of the man
with a hammer (i.e. skills in MAE) who thinks that everything is a nail. What
we observe in economics is a vast population of workers, each equipped with a
state-of-the-art hammer with which she has been trained to the level of expert.
All of these workers are huddled around a single worktable, each trying
desperately to push their way to the front so that they can impress the boss by
hitting something (and don’t come at me with the benefits of competition, which
Smith also emphasised – the whole point of a firm, at least according to
mainstream economics, is to internalise transactions that would normally occur
through the market). Meanwhile, all (at least most) of the other work stations are
undermanned, laden with untapped potential for improvements in productivity.
Thus, in a sort of reverse Lewis-model scenario, workers are flooding into the
surplus-labour ‘MAE sector’, adding little in the way of productivity, despite
other sectors remaining underpopulated. Furthermore, those work stations that are enjoying a much-needed increase in manpower, such as behavioural economics, remain disconnected from the other stages of production, with the lamentable result that the mainstream continues to rely on demonstrably false assumptions in order to facilitate hammering (i.e. modelling). The gross inefficiencies of this factory of
economics are reflected in the quality of its final product, with which its
consumers – that is, students, employers, and the like – are
increasingly dissatisfied.
Finally, it should be noted that, despite singing the
praises of specialisation within a division of labour, Smith also warned
against its adverse effects on individual workers. Indeed, he (1776, Book V,
Chapter 1, para. 178) went so far as to state that repetitively performing a
narrowly-defined task would render a worker not only “stupid and ignorant”, but
also “incapable of…conceiving any generous, noble, or tender sentiment”, even
to the point of adversely affecting her “private life”. At the risk of (severe)
generalisation, this is indeed what we observe in economics – flocks of idiot-savants who succeed in degrading the subject to a mere branch of mathematics, so devoid of any
real-world foundation or application that academics and students alike have
called for a ‘post-autistic economics’.
Even specialists in MAE must have some knowledge
of actual economies, as well as economic history, economic thought, and other
methodologies. Indeed, Smith (ibid.) was so concerned about the adverse effects of
specialisation on workers’ mental health that he believed government
intervention would be required to curb it. Perhaps something similar is
required in economics!
Notes:
[1] Another channel by which specialisation would increase productivity,
according to Smith, was by eliminating the time wasted by transitioning between
stages of production. This channel does not apply to the current discussion
except insofar as time is required to learn new methodologies, which could be
spent using the methodologies in which each economist is already proficient.
[2] Obviously, I am assuming here that other methodologies are useful, which is precisely what I am trying to argue. As stated at the outset, this is (admittedly circular) argument by analogy, not deduction. Incidentally, the exclusive use of deductive logic to the neglect of inductive approaches has been identified to be one of the major pitfalls of contemporary economics.
Reference:
Smith, A. (1904 [1776]). An Inquiry into the Nature and Causes of the Wealth of Nations (5th ed.). London: Methuen & Co. Retrieved from http://www.econlib.org/library/Smith/smWNCover.html
[2] Obviously, I am assuming here that other methodologies are useful, which is precisely what I am trying to argue. As stated at the outset, this is (admittedly circular) argument by analogy, not deduction. Incidentally, the exclusive use of deductive logic to the neglect of inductive approaches has been identified to be one of the major pitfalls of contemporary economics.
Reference:
Smith, A. (1904 [1776]). An Inquiry into the Nature and Causes of the Wealth of Nations (5th ed.). London: Methuen & Co. Retrieved from http://www.econlib.org/library/Smith/smWNCover.html
