Keith Rankin, 30 December 1998
"Economists [and their critics?] ... are particularly inclined to confuse the things of logic with the logic of things."
Neoclassical economics, which purports to be a science, is often criticised for placing rationalism ahead of realism. One person who constantly pursues this critique is Paul Bieleski, who publishes a newsletter under the banner of the "Anti Economist League".
The criticism of neoclassical economics is valid to the extent that the purpose of a science is to learn about the ways that the real world of experience works.
From here on, I will use the term "economics" (when unqualified) to mean "neoclassical economics".
Economics is a conceptual system of thought; an abstract human construction of an ideal form of utility, production and exchange that is rational in the sense that it is dependent on various rules of behaviour or "laws" that would be followed by "utility-maximising" beings under what western (Protestant?) cultural norms regard as "ideal" conditions. Thus economics is a science in the way mathematics is; a priori and not a posteriori. Economic analysis is therefore a method of discovering the properties of an ideal system, and not a means for prescribing increased well-being in the real world of our experience. (Many economists of course assume that we would be a more economically successful society if the conditions under which we make our choices are made to approximate their ideal conditions. That's a matter of faith; it has nothing to do with science.)
The question that then follows is "what is use of economics?". It is the same question as "What is the use of mathematics?" or "What is the use of abstract thought?" or "What is the use of classical music, abstract art or formalism in poetry?" Economics is a way of thinking about utility that may itself not have any direct "real-world" applicability. Hence it is decried by "philistines" (meaning people who give value to the concrete but not the abstract) for itself being philistine in its formal assumptions about human beings.
Economics is an example of a successful mode of rational thought, because it is logically consistent, and because it tells a number of fascinating (although unreal) stories. Systems of rational thought are true in the ideal sense, where logical consistency is the only basis for truth. But they are entirely untrue in the real world of experience. They help us to understand the real world though, because economic rationality - ie behaviour according to utility-maximising rules - does appear to play some role in human behaviour. So knowing the properties of the ideal system can give us some insight into the workings of the real system of human experience.
The problem of economics is not, therefore, its excessive abstraction; its formalism; its emphasis on rationality over substance. Rather the problem is its imperialism; even its Stalinism, as Deirdre McCloskey noted in her 1996 Auckland lecture.
This problem of intellectual imperialism exists in two forms. First, there are any number of alternative abstract systems of economics that can be constructed. Marxism is just one of such alternatives; an alternative built on equally unrealistic premises. Different premises of human decision-making lead to different ideal economic systems. Marx's premises were more sociological in character, whereas utilitarian economics is individualistic.
More realistic premises tend to lead to chaotic systems, that are perfectly rational but very hard to predict because of their complexity. Chaos models also give better insights into the world of nature than do equilibrium models like Newtonian physics. The problem is that the institutionalised dominance of equilibrium economics continues to marginalise other forms of ideal economics.
The imperialism of economics works in another more insidious way. It inhibits the development of parallel disciplines that are inductive (ie based on experience) rather than deductive (based on a set of premises); of disciplines that seek to explain and describe the real world of experience. The economic problems that we face in our daily lives, and that governments face in making policy, are problems of the world of experience. They require a form of economics that describes and explain economic activity as it occurs, and not as it might occur in Plato's or Hayek's Republic.
In natural science, realism is achieved through the process of experimentation. Rigorous experimentation is difficult in economics (i) because humans quite reasonably do not like being experimented upon, and (ii) because of the uncertainty principle that says that the observer inevitably has some impact on what is being observed. (This is true even if the observed do not know they are observed. The uncertainty principle derives from physics.)
So far, the closest thing to experimentation in economics is comparative economic history, where past societies or communities have faced similar contingencies and responded in certain similar or dissimilar ways that can be analysed and explained. (History has the advantage that the observation is purely retrospective, and is therefore not subject to the vicissitudes of the uncertainty principle; although it may be argued that some people - mainly politicians - do act today with an eye to the history books of tomorrow.) Unfortunately, part of the imperialism of economics has been to marginalise the role of history within economics.
Despite all the difficulties, a new form of experimental economics, called "behavioural economics" is emerging. (I have to thank Paul Bieleski for drawing to my attention the 19 September 1998 New Scientist article "Let's get emotional", which describes a project in California that confirms that not only do we not make choices as economics assumes, but that there is a predictable rationale behind the choices that we do make.) Through carefully devised experiments of human preference, it is becoming clear that much human behaviour, while rational in the philosophical sense, is "sub-rational" in terms of the premises of neoclassical economics.
Thus, when wages are high and jobs are abundant, people work less, not more. Likewise, when income is hard to get, people increasingly prefer paid work over other activities. After all they have expenses to meet today; expenses which cannot be met by working more tomorrow when work may be more abundant.
To give another example, when assets (eg shares, residential and commercial property) are over-valued, most of the time we demand them more, not less. But there is an imprecisely predictable point in which people's behaviour will change, and they will then try to sell rather than buy over-valued assets.
Old-fashioned economic rationalism is under threat because some of its assumptions have been proven to be irrational under real-world conditions. Hence, economics, so far, has given us very little insight into the way individuals, firms or governments should act to improve the lives of their principals. Forms of abstraction will always be an important conceptual tools for policymaking. They will be much better tools when their behavioural rules converge with the rules that we observe in the real world.
A word of warning though. We need more than a more realistic form of economic rationalism. We also need economic history, and we also need to make policy on the basis of direct experience of what happens in real worlds. Inductive and deductive approaches to economics are complements, not alternatives. The more realistic that rational economics becomes, the more chaotic will be the "solutions" to its mathematical models. Policymaking will always require a large dose of historically informed common sense.
PS See Utopia of Endless Exploitation; the essence of neoliberalism by Pierre Bourdieu (Le Monde-Diplomatique, 8 December 1998) for a Polanyiesque critique of the imperial project of economics ("an immense political project"); a critique which emphasises the role of the theory of making itself into a truth within the world of experience as well as the world of ideas.
The neoliberal project cannot however create all of the conditions required to realise the ideal. It can create the kinds of public institutions assumed by the a priori construct, but it cannot fulfil assumptions such as perfect knowledge of the technical and consumer choices available, and it cannot create an environment of zero uncertainty. And it will never create a private banking system that lends on a fully unsecured basis. Unfortunately, a utopia that doesn't work is something even worse, a dystopia. The neoliberal utopia cannot work as a real-world system; therefore successful attempts to create the required institutions and to destroy all other institutions (and the social memory of them) can only create a dystopia. [top]
Rankin File | 1998 titles