|
post-autistic economics |
Gilles
Raveaud (co-founder of Autisme-Économie)
In this article, we present
the reasons for our discontent as students of economics in France (1.) and find
that they have a broad cause : the (desperate) quest of economists for
scientificity (2.). Our proposal is to turn the teaching of economics in a
completely different direction, where teaching of controversies would be
given priority (3.). We then explain why teaching through controversies is
preferable to current methods (4.), and, well, how to do it (5.). Finally, we offer a very tentative
insight into the theory of knowledge, by pointing out affinities between the
classical works of Max Weber and those of Tony Lawson. We then suggest that
this more general framework of comparisons (in a well defined sense) may
prove a useful guide, not only for teaching, but for research as well. 1. Reasons for our discontent I will briefly review here the reasons for our
discontent, as expressed in our petition (full text on our website: www.autisme-economie.org) : - First, we criticized the construction of
“imaginary worlds” by economists. That is, worlds which do not have any link
with any plausible mechanism in reality. Such worlds (the famous “models”)
are just developed for their own sake, because of their tractability. We no
longer want to be taught such fairy tales, the aim of which is not to explain
« reality », but just to show the ability of the writer to
construct a « nice model ». It may be fun for the authors, but we
do not want to be part of the game. - Second, we highlighted the excessive use of maths
in the curriculum. Here, the situation is well known and does not require
further explanation. Let me just emphasize the fact that in France, the
fascination with maths is particularly developed, and the level required of
students extremely high in most cases. - Third, we stressed the fact that most
« explanations » given in lectures are derived from neo-classical
economics. Marxism and Keynesianism, just to name a few alternate views,
which were living schools of thought in France not so long ago seem to have
completely disappeared[1].
Nowadays, virtually no student has the opportunity to study Marx as part of
the common curriculum. Furthermore, the explanations given are not thought of
and presented as possible
explanations of complex phenomena (say, unemployment, or international
trade), competing with different approaches. They are supposed to be the only
possible economic analysis of such phenomena, in the following sense: even if
people do not actually behave according to these views, it is because other
elements (social, traditions, imperfections, etc.) blur the picture. But the
proper economic part of their activity is as defined. As one can see, the situation is disastrous[2].
As so many writers have already given possible reasons for such a situation
in economics, we will simply focus on one aspect of the problem: the quest
for scientificity for economists. There are two reasons for selecting this
problem: first, these are the terms which have been used by teachers reacting
to the first teachers’ petition, which supported our views. Second, we
consider it as the central source of our problems, even if this view is only
a conjecture and is subject to revision. 2. The underlying causes: seeking
for scientificity A few weeks after launching our petition in May
2000, we were supported by a petition from teachers, signed by approximately
140 teachers, with some very famous people among them (Aglietta, Orléan,
Boyer, etc.). This text from the teachers stressed the lack of pluralism in
economics and rejected the overall domination of neo-classical economics on
courses. Later on, another text, signed by only a dozen
economists, but nearly all of them important in the Establishment, reacted
against both the students’ and teachers’ petitions. Their purpose was, as
they put it, to « preserve the scientificity of economics ». This
had to be done in the most traditional of ways, following the known
recipe. The following principles
were insisted upon in this tract: - First, the “precise identification and definition
of concepts and behaviours which characterize economic activity (consumption,
production, investment, etc.) and the formulation of basic hypotheses
concerning such behaviours.” - Second, “the formulation of theories expressed as
formalised relationships between the concepts previously defined ” - Third, “the verification of such theories through
experience. This can only take place through confrontation with history, as
quantified by statistics and econometrics. They then denied that there was any illegitimate
domination by neo-classical economics : any views are valid, as long as
they respect this canvas. One can note that their reaction did not engage with
our criticism: where we talked about teaching,
they talked about science. They did
not answer any of our points : imaginary worlds, the excess of maths and
lack of pluralism. Instead, they gave us a mythical description of science,
where everyone is simply doing their best with their limited means (isn’t
that rational ?). In short, leaving the classrooms, the debate was now
entering the field of methodology. So much the better, as this suddenly gave coherence
to our protest. We had focused on specific points which, through related, did
not lead to any necessary coherence. Now, with the help of the
« counter-appeal », we had more than coherence: we had a common cause.
Let’s take this case seriously[3]. In the construction of a « science » so
naively understood, one has few options. The phenomena studied, even analysed
as « stylised facts », are so complex that you have to choose among
them if you want to produce a « result ». From claiming to be
global, the analysis in fact turns partial. Then, if your purpose is to
construct logical relationships between « objects » (to be
defined), you can hardly do without maths. But then again, as you progress,
things will get just so complicated - leading either to many solutions or
none at all - that you’ll have to spend your time simplifying your
hypotheses. Besides, for reasons which cannot be explained here,
the dominant view in neoclassical economics is that everything starts with
the individual, you then have to define your « individuals » as
narrowly as possible. Actually, you will not deal with
« individuals », nor even economic « agents » any
more : you simply declare them to be their (unobservable) preference
curve, and that’s it. You then have a solid basis on which to ground your
calculations. Does this lead to any analytical success? We know that it
hardly does. Even the logical
basis of such a simple phenomenon as the « law of supply and
demand » cannot be established[4]. The logical consequence is that the whole approach
should be re-thought. But as the « scientificity » of economics is
to preserved, this is not an option. You have to follow this road, produce
results, refine your hypotheses, in order to « make science
progress ». You are not going to give up the possibility of constructing
a science, are you? Following the magi’s’ scientific recipe has three
consequences: - First, models are not supposed to be realistic:
their aim is not to give insights into real phenomena, but to participate in
the global and fascinating construction of a science. One can no longer
criticise models on the grounds of irrelevance. From the point of view of this major achievement, which is
the construction of a general economic science, universally valid, these
models are, on the contrary, perfectly relevant. But, yes, in a sense, they
are « imaginary ». - Second, science understood this way has to be
mathematical. There is no precise reason for that, apart from the fact that
any science is mathematical, isn’t it ? You would not call sociology a
science, would you ? As economists, theoreticians deal with quantitative
data (inflation rates, balance sheets, etc.), so there is no reason for them
not to use them. More fundamentally, their purpose is to construct a general
and logical theory, on the model of classical physics[5]
– so they need maths. And, as they need maths for research, maths needs to be
taught[6].
So there cannot be an excess in the use of maths. The only possible
difficulty is not enough maths, or inadequate maths. But not too much. - Third, as their aim is, I hope you’ve understood
it by now, the construction of a « science », the idea of competing
theories is nonsense. Well, of course, you can debate specific problems, propose
different tools, but this game is required to take place on the same field.
You cannot play against someone situated on a different pitch, can you? We
need common rules, a common language. So you have to learn this language
first of all: after that, you will be absolutely free to make any use of it,
honest. The points we highlighted are not here by random.
Even more, as we hope to have shown it, they
are not a problem for those who subscribe to the line of « scientificity ». They are, on the
contrary, perfectly coherent. That is, our critics can no longer concern
themselves only with teaching: they have to encompass theory as well, and
methodology. 3. Looking for controversies Let’s turn back to the image of science insisted
upon by those economists : aren’t there any conflicts of power, any
competing views on the good in economics? Aren’t there completely opposed
conceptions of what a society and an economy is among economists? Does
everybody agree on the way to produce scientific analyses of economic
problems? To put it frankly, the conception of science those economists have
is really naïve: many works, written by sociologists studying natural sciences, have shown how
knowledge is constructed through
the conflict of theories[7].
Moreover, they’ve shown how the permanence of such debates is intrinsic to
science, and is, one could say, its engine. No science is a calm and
eternally defined pitch. On the
contrary, its lines are constantly evolving, and sometimes scientists do not
even agree on the shape of the ball ! Would economics be an exception? Fortunately not.
One knows how economics has been the scene of terrific fights over the years
since its very origin, whenever one locates that beginning. Moreover, the
point of such battles was often over the very foundations of the discipline
(the forms of the lines of the pitch), not only its methodology (the rules of
the game) and objects (shape of the ball, if any). Such controversies do
exist in economics, they are numerous and persistent. What does « controversy » mean? A strict
definition of it could be the following : A formulates a theory, that B
criticises, A replies, and so on. Examples of such controversies are for
example, the « Vining-Koopmans » controversy on the place and role
of maths and theory in economics. The controversy started with Koopmans
qualifying Burns and Mitchell Business
Cycles of « measurement without theory », because it was not
based on neo-classical grounds. In his reply, Vining pointed out that
Koopmans’ judgement relied on a stringent definition of economics, according
to which only his approach (neo-classical theory) was valid, although he had
no firm results which showed the productivity of such an approach. The
controversy concerned the definition of acceptable theory and the reasons why
neo-classical economics should be preferred[8]. As one can see, this controversy has never been
settled : there is no « scientific answer » to the problems
raised above, even if the majority of economists support one view rather than
the other at a certain time. This is what economics is really about : conflicting views about problems and the way
to deal with them ; competing claims on which policy to support. This is
just so obvious that one is surprised to have to recall this. Newspapers are
filled with economists contradicting one another on the policies to be
followed, because their theories differ. Everyone knows that there is no such
thing as an economic science, at least not in the sense understood above, but
the students are required to believe there is. Only economics students are
supposed to believe that there is a knowledge about economic facts which is
coherent, precise, and which leads to clear policy recommendations. 4. Teaching through controversies Controversies, even understood in the restricted
sense given above, are numerous in economics (see the classical debates on
value ; the question of the nature and measurement of capital ; Keynes and
Hicks on the IS-LM model ; Romer and Solow on growth theory ; etc.)
But one can understand controversies in a broader sense : that is, the
competing views on recurring problems in economics. Theories of consumption
are an excellent example of this: starting with the Keynesian
« consumption function », one can trace back the attempts by Brown,
Modigliani and Duesenberry to « save » the Keynesian function. One
can then turn to Friedman’s radical criticism, with his « permanent
income hypothesis », and its revision through the life-cycle model. All these models are not to be presented as trials
and errors on the road to an ever better theory : on the contrary, they
are to be presented as what they are,
that is, conflicting views on an economic problem of central importance. In
particular, such theories, and the differences between them, simply cannot be
understood if one is not aware of their consequences in terms of economic
policy. Teaching these models should be quite different than at present; each
theory should be presented on an equal footing, given its chances in the competition
of science. In particular, one should focus not only on the
« model » itself (if any), but also on three crucial aspects of the
problems surrounding the use of the model: the context of the debate ;
the role and place of empirical data ; the consequences for economic
policy. As one can see, teaching is much more demanding in
this context, as it is supposed to encompass all the dimensions of the
problem under consideration. In fact, what we are suggesting here may already
exist, in some shape or other, but, even if it is the case, it is done
through various courses, which are not linked to one another. Statistics are
taught for their own sake, with no or little reference to economic
theory ; the history of thought very often overlooks the mathematical
contents of theories ; and economic theory is presented as being
autonomous from everything else. The complete absence of economics textbooks
constructed in the way we propose is a clear indication of this disastrous
« division of labour » among researchers and teachers. As a result of the present state of teaching, the
student does not acquire any knowledge of the economy he lives in, nor of the
competing theories that set out to explain it. With what we propose here, it
could be hoped that both lacks would be filled : controversies make
theories lively and real ; and their empirical content gives us
knowledge of « facts », as we show below. 5. Consequences Teaching should focus on controversies, past and
present. The idea of neutral tools would quite simply disappear.
« Micro » and « macro » lessons would be replaced by a
single and large course on « economic theories », presenting them
in the way described above, in historical (and not logical) order. There
would then be the time to present Classical and Marxian approaches seriously,
spending time on reading texts, and not simply solving silly ready-made
exercises. In practice, this implies that neo-classical theory should not be
taught in the first year, and perhaps even not in the second year. Besides,
we propose a broad course on political
philosophy, explaining the ethical and political backgrounds and the
consequences of the theories proposed. A canvas would therefore be laid down for a
stimulating debate inside economic departments : theories would be
presented as being in permanent confrontation with one another, the course on
economic theories following their historical development and reasons for
success and decline. As we stressed above, the purpose of such a course would
be to present both the historical
context of the theories and their
precise analytical content[9].
This would be completed by the political philosophy course, which would show
the permanence of questions such as the role of the State in economic
matters; the fairness of the market; the definition of money; etc. and the
varying answers given to such questions through history. Statistics and econometrics would then be taught in
close relationship with such lessons. Even if these subjects are necessarily
autonomous in the sense that they need to present their own techniques,
tutorials (at least) should be entirely devoted to the examination of
problems linked with what is presented in the two courses defined above (how
does one measure the « natural rate of unemployment »? How did it
evolve in the UK or the US or Gambia? What is « purchasing power
parity » and how to measure it? How can one measure the value of the
Keynesian multiplier nowadays in a particular country? What is the value of
the rate of profit ; is it declining? etc.). In particular, it is
important that students should work on real data, with its imperfections, and
not, once again, on artificial sources[10]. A common and simple consequence of all this is that
we would get rid of all ready-made
exercises. Well, one or two may be tolerated, as they are useful aids to
understanding a theory
precisely. But the aim of teaching should no longer be the mastering of such
tools. They should remain what they should never have ceased to be, that is,
analytical tools with limited relevance. The object and purpose of economics
and its teaching is the world outside: what do we know about it? How do we
know it? Do we all agree on this knowledge, and if not, why? What do our
theories have to say about it? And above all, to be honest: how to make it a
better place to live in?
|
|
[1] The change was quite brutal : a few years ago, the author of these
lines had the chance to read texts from authors like Kaldor, Robinson, Marx,
Marshall, etc. in first year, but he no
longer has the opportunity to teach them to his students.
[2] And to be honest, our protest is
fairly incomplete: one could mention for instance the way statistics and
econometrics, useful tools as they may be, are taught in France nowadays. That
is, students spend hours demonstrating the convergence proprieties of
estimators, but we never run real data in tutorials. Or the political
consequences of the models proposed, etc.
[3] What follows may be read as an oversimplified
and caricatured synthesis of some aspects of Lawson’s (1997) Economics and Reality, Routledge. We
present some of his arguments more seriously in the final section.
[4] See various books by B. Guerrien
(in French only, unfortunately), and on related topics, Alan Kirman, (1989),
"The Intrinsic Limits of Modern Economic Theory: The Emperor has No
Clothes", The Economic Journal, 99,
395, 126-139.
[5] See Mirowski, (1989), More Heat than Light, Cambridge
University Press.
[6] And if all the students are not
pleased, never mind. As a French teacher put it, « it does not take 15000
people to build a science »
[7] See the programme of science studies, and in particular works
by Bloor and Latour (even if their views are themselves quite different).
[8] For a full story, see the original
texts in The Review of Economics and
Statistics, vol. 29, published in 1947. Mirowski, in (1990?), “The
Measurement without theory controversy : defeating rival research programs by
accusing them of naive empiricism”, Economies
et Sociétés n°11, pp. 65-87. puts the story in the broader perspective of
the conflict between the NBER and the Cowles Commission.
[9] One could imagine a course done by
two teachers, one a specialist of the history of thought, and the other a
micro- or macroeconomist. It could even vary with the subject, some lessons (on
the Phillips curve, say) being done by a statistician and a macroeconomist,
others (e.g. : theories of value) by an historian of thought and a
microeconomist, etc. In France, when we suggest this, everybody laughs :
« What ? You want to make different people work together ? What a joke ! ». Unfortunately, this
does not make us laugh. Let’s hope that the situation is somewhat better in the
UK...
[10] In fact, it is hard to understand why
it is not already the case, as this is just common sense.