Limits of Economic and Social Knowledge

S. DeCanio

• Publisher: Palgrave Macmillan
• Year: 2013
• Print ISBN: 9781137371928
• eBook ISBN: 9781137371935
• Topic: Economics
• Subtopic: Development Economics
• Index: Economics

The book aims to show that the deterministic vision embodied in conventional economic modelling is neither consistent with nor supported by the state of the art in mathematics, logic, and physical science. DeCanio recognizes that economic agents are intrinsically free and somewhat unpredictable, which is essential for economic and social theory.

1

Materialism, Determinism, and Economics: An Overview

By uncritically embracing materialism in the mistaken belief that this commitment is an intrinsic element of the scientific method, economics and the other social sciences have taken a wrong turn. The world of inanimate matter is the purview of the natural sciences, but carrying a materialist orientation over to the study of the human and social realms sets off a cascading series of oversimplifications, misconceptions, and outright fallacies that cumulatively can be fatal to genuine understanding.

There should be no confusion about the fact that modern economics does, in fact, treat humans as purely material and physically determined beings. Economic theory models the ‘agent’ as a utility function, and the maximization of utility by individuals (and of profits by firms) puts restrictions on what theory allows – thereby generating hypotheses that can be tested. The individual is treated as having fixed preferences over a set of commodities, which themselves are usually thought of as material goods, although the commodity set can be expanded to include other sources of happiness.¹ Although economics usually presents itself as studying ‘choices’ among alternatives, asserting that human behavior amounts to nothing but utility and profit maximization reduces human action to a calculus problem. Game-theoretic variants of the same idea may allow for strategic interactions among the agents (each of whom is assumed to have a stable set of preferences over possible outcomes), and can even allow for strategies that involve randomized moves or counter-moves by other agents, but the collapse of human behavior into a mathematical calculation remains the same.

Contemporary economics, along with most of the rest of academia, has (whether consciously or tacitly) adopted a set of interrelated and overlapping ideas. Various terms can be used to describe this ideological nexus: materialism – determinism – physicalism – naturalism. An issue of terminology is involved here. Physicalism (or, sometimes, naturalism) is the modern successor to the older concept of materialism. Science has established that the ‘material’ world consists of more than matter: fields, space–time, and the quantum particle zoo are all more fundamental than old-fashioned ‘matter’. Partly as a result of this, it is difficult to specify exactly what is meant by the ‘physical’ world. It is circular to define the physical world as ‘that which is the subject matter of physics’, and the Principle of Causal Closure that some modern materialists appeal to is, in the end, nothing but a metaphysical assertion. As shorthand for this mindset, I will generally use the term ‘materialism’ because of the term’s long usage and its concreteness, recognizing that the discoveries of science have taken us far beyond an imagined world made up entirely of solid objects interacting with each other through direct contact.²

The materialist ontological stance, along with the methodological reductionism that accompanies it, arose out of the great successes of physical science that predated the twentieth century revolutions of quantum theory and relativity. The materialist-determinist-physicalist-naturalist approach gives rise to a particular attitude towards the human beings who make up the subject matter of economics and social science. This attitude de-emphasizes, or denies outright, the capacity of people to make meaningful decisions when choosing among alternatives that could potentially be realized. Denial of this power is required if the study of social behavior is to be ‘scientific’ in the image of classical (pre–twentieth-century) physics.

So, right at the outset, economics comes up against the free will versus determinism debate. The scientific pose that economists have adopted requires that the choices made by the agents conform to the ‘behavioral laws’ embodied in the preference functions, constraints, and strategic options that are open to them. Scientific truth rests on the discovery (and working out the consequences) of law-like regularities. If humans are not predictable, based on application of such laws, of what would the ‘science’ of human behavior consist?

However, the materialist determinism that underlies the modeling of human behavior in neoclassical economics is no longer the default metaphysical position of modern science. Even if a physicalism that has no role for transcendent reality is the working practice of most scientists, those who are philosophically inclined realize that viewing the world exclusively in terms of massy objects interacting with one another in causally determined ways is obsolete (McMullin 2010, Clayton 2010). The content of modern science is built up from non-material fields; from space and time that are interconnected relativistically; and from a quantum phenomenology that is robustly probabilistic, non-local, and observer-dependent. Even within ‘classical’ pre-quantum physics, limits on the precision with which initial conditions can be specified create an insuperable barrier against the possibility of perfect predictability.

Philosophers have struggled for millennia over the question of human free will, and it would be presumptuous of me to claim to add anything new. Instead, I wish to draw the attention of those who are concerned with economic theory and policy to the pressing relevance of the philosophical issues involved. Most economists are blithely indifferent to these matters. Some economists with wider-ranging intellectual interests, such as Hayek and Keynes, have not been content with the superficial analysis of ‘choice’ within a deterministic framework. But curiosity about the first principles of human behavior is almost entirely absent from contemporary economic theory or, for that matter, from economics education at either the undergraduate or graduate level. The reason for such intellectual complacency is not necessarily bad faith on the part of economists; rather, the problem stems from the specialization and compartmentalization of knowledge, as well as the professional incentives (even imperatives) that preclude examination or discussion of fundamentals. Nevertheless, the philosophical issues find, willy-nilly, a resolution, and if there is no effort by those with some economic sophistication to address these issues, the resolution will be carried out by politicians, journalists, technocrats, or others who are even less suited to the task.

The free will controversy is not the only philosophical problem pertinent to economics. All current economic modeling is mathematical, so issues having to do with the limits of mathematical proof, with the kinds of results that can be reached by computation, and with the properties of the underlying variables that are both the inputs and outputs of models, ought to be of concern to working economists. The existence proofs of economic general equilibrium theory are largely concerned with establishing whether certain equation systems have solutions, but questions are hardly ever raised about whether there are limits to what such equation systems are capable of describing, or about whether those bounds have any consequences for real-world actions.³

Deterministic policy analysis

It is unfortunate that economists are oblivious to the contradiction that lies at the heart of so much of their practice. Economics claims to study choice in the face of scarcity, but the ‘choices’ are really nothing but a series of calculations whose outcomes can be worked out in advance. Yet a glaring absurdity of the deterministic outlook arises in considering what economists might mean when they conduct ‘policy analysis’.

Under straightforward determinism, the very idea of ‘policy choice’ is empty. Everyone in the system behaves as an automaton, including those who are in power or who are structurally located in one of the decision-making slots in the system. Presumably it might be possible to study the operation of the system as a whole, including the formation of economic ‘policy’ subject to inexorable deterministic laws, but what would be the point? It would have no human significance – just as, under strict determinism, human action in general lacks meaning.

Alternatively, it might be possible to think of economists, analysts, or policymakers as standing above the human scene, armed with econometric estimates of the utility functions (and their consequent individual demands for particular goods and services), and able to manipulate policy levers in such a way that the welfare of the agents operating at the mundane level is maximized. This is indeed the conceit of economic modelers who seek to set optimal policies to reduce greenhouse gas emissions, levels of spending on various kinds of public goods, or regulatory policies to correct market failures. In such exercises it does not matter that the human agents are reduced to utility functions whose well-being can be calculated and perhaps compared; the economist, from a position of aloof superiority, pronounces that his models show the way to Pareto improvement.

In this fantasy world there is no room for genuine choice. Or, rather, the only real choices are those made by the economist/policymaker in selecting which outcome is to be preferred. The economist can claim to have the welfare of ordinary citizens at heart and, in many cases, this may be true, but there is no escaping the conclusion that an economist operating in this mode is acting on behalf of social automata who have no real minds of their own.

This approach is internally incoherent because it ignores the factors that are determining the policymakers’ preferences and actions. Consistency would require that ‘policymaking’ is nothing more than the actions of decision-making agents attempting to implement their own particular preferences. A great deal of what passes for politics nowadays does have this quality, of course; the leading political actors do not genuinely engage at the level of ideas, but rather seek to frame their own agendas in ways that will appeal to the mass electorate. The mechanistic models of economics encourage this behavior by taking preferences as exogenously given and denying that there is any non-material ground for valuing individual or collective action. Everything is a matter of strategy and calculation, with goals measured exclusively in terms of material outcomes. ‘Policy analysis’ is nothing but an elaborately embellished framing or spinning of issues to further the acquisition and holding of power.

Other ways of describing human behavior in materialist terms that do not entail self-consciously maximizing agents are possible. For example, the societies of the social insects – ants, bees, and termites – display remarkable structural and other regularities, including division of labor, hierarchy, and self-sacrifice by individuals (see Moffett 2011 and the references therein). Models of these insect collectives have been developed based on purely local interactions of individuals with their neighbors, on the roles of chemical trails and signals, and so forth. It is plausible to think that some of the structural features of human societies might also be explicable in similar terms – nearest-neighbor interactions, myopic rules of thumb, institutionally grounded conventions, biological necessities, and the like. The relatively new field of agent-based modeling proceeds along these lines. Artificial societies can be constructed out of computer simulations in which simple behavioral rules produce structures and dynamics with familiar and realistic features (see Chapter 3).

Granting that such an approach may generate insights, what would policymaking mean if the real world conforms to the typology of agent-based modeling? One possibility is the same as has been discussed before: ‘policymakers’ standing over and above the actual society and adjusting the parameters or input variables of the agents’ rule sets so as to bring about desired outcomes. This would make the policy world separate from the world of the agents and would essentially posit a distinct class of rulers who possess the free will that is denied to the rule-following agents. Alternatively, the government agents might be subject to rules that are the same (or different in specific ways) as the other agents, but still lacking any freedom of choice. This is essentially the same as other forms of hard determinism. Thus, while the agent-based approach is interesting and offers some new possibilities for modeling (primarily because the explosion of computational power enables running simulations with multiple agents, relatively complex rules, and evolutionary development), in the end it is just another variety of deterministic reductionism.

Confronting the free-will/determinism dilemma

Of course, it is conceivable that human beings really are only complicated mechanisms whose actions will become entirely predictable as the sciences of psychology and economics advance.⁴ Thus:

If explanation in the social sciences is conceived naturalistically – as no different in kind from explanation in the natural sciences – it would seem that the notion of a metaphysically free will can find small purchase, and must be rejected as a holdover from religious views of the world, condemned to obsolescence by the rise of science. (Quinn 2006)⁵

More will be said later on about the ‘obsolescence’ of religion brought about by the rise of science as the dominant world view. For the moment, it is enough to make the common-sense point that materialist determinism is inconsistent with the overwhelming subjective sense of freedom we all share. No philosophical contortions can produce a Houdini-like escape from this subjective certainty. A great deal of philosophical literature on the mind–body problem, efforts in current neuroscience to associate brain states and neural activity with subjective experiences, and theories of consciousness like epiphenomenalism are all efforts to evade or deny the subjective certainty that we are free to control, at least to some degree, our own actions. As Hans Jonas puts it succinctly:

They [the most complex and subtle life-forms] contain something else, as we know first hand, for example, through our present inquiry into the nature of things. For there exists the dimension of the subjective – inwardness – which no material evidence by itself allows us to surmise, of whose actual presence no physical model offers the slightest hint. The physical cannot represent or clarify the subjective dimension with its concepts – indeed, it does not even seem to grant any room for the participation of inwardness, as undeniable as it is, in external occurrences. (1996, p. 169)

The debate over freedom versus necessity has been going on for as long as we have any historical recollection. Leucippus said that ‘nothing happens in vain (matên) but everything from logos and by necessity’ (Berryman 2010, citing Diels and Kranz, 1951). One of the objectives of the chapters that follow is to show that in light of modern developments, determinism cannot be established according to the standards of science. Non-trivial instances abound in which the behavior of free-willed agents and purely dete...

Table of contents

• Cover
• Title
• 1 Materialism, Determinism, and Economics: An Overview
• 2 Consequences of Computational Limits
• 3 Simulating Simple Societies
• 4 Economics and Physical Science
• 5 Economics, Behaviorism, and Utilitarianism
• 6 A Case Study and Cautionary Tale: Climate Policy
• 7 Politics and Governance
• 8 The Gift of Freedom
• Notes
• References
• Index