The Distinction Between Positive and Normative Theorising, Particularly in Economics

The attempt of Enlightenment philosophers to give a naturalistic, individualistic, ‘scientific’ and universal account of our moral codes was recounted above. Intertwined was the attempt to insulate that account from any divine authority, while at the same time trying to base those codes on empirical observation of human nature. The natural-law tradition—so central to Locke’s justification of his social-contract theory—became increasingly secularised over several centuries. The effect of those attempts was merely to justify existing moral and political arrangements. Since then, much social thought has been preoccupied with finding a method that will either determine values objectively or avoid questions about values altogether.[4]

Notwithstanding the ambitions of Hobbes and Locke and their successors to found our moral judgements on science, recent social scientists have generally made a distinction between science and normative theorising. As a result, it is often claimed that value judgements lack the objective validity of science, and science must, as a methodological ideal, be kept free from them.[5] Similarly, economists have usually drawn a distinction between positive and normative economics. It has, of course, been admitted readily that the application of the ‘positive’ science of economics to real public policy problems is a normative issue. This—somewhat deceptively—usually took the form of suggesting that it was in the choice of ends that the normative issue arose, while positive economics could safely address the best way of achieving those specified ends. The fact that ends and means are usually intertwined escaped notice.

It should already be clear to the reader from the account in previous chapters that the idea of a value-free social discipline is not possible. All social knowledge and moral narratives are stories told from a particular normative perspective, employing language imbued with normative values. As Webb tells us, ‘[T]here is no human action of any importance which does not become imbued with moral and normative significance and hence develops an abstract and symbolic dimension.’[6]

The distinction between the positive and the normative is usually traced to Hume, who is taken to have held that there is a watertight distinction to be made between the realm of facts and the realm of values:

I have always remark’d, that the author proceeds for some time in the ordinary way of reasoning, and establishes the being of a God, or makes observations concerning human affairs; when of a sudden I am surpriz’d to find, that instead of the usual copulation of propositions, is, and is not, I meet with no proposition that is not connected with an ought, or an ought not. This change is imperceptible, but is, however, of the last consequence. For as this ought or ought not, expresses some new relation of affirmation, ’tis necessary that it should be observ’d and explain’d; and at the same time that a reason should be given for what seems altogether inconceivable, how this new relation can be a deduction from others, which are entirely different from it. But as authors do not commonly use this precaution, I shall presume to recommend it to the readers.[7]

In economics, the distinction was picked up by Nassau Senior and Mill,[8] and was subsequently endorsed by Weber[9] and Robbins.[10] As such, it formed part of the process by which economics shed any overt moral, historical and institutional concerns and was transformed into a mathematical discipline. This transformation reflected the growing influence of positivism with its view that true scientific knowledge was divorced from metaphysical beliefs. As already pointed out, it reflected also the growing prestige of Newtonian physics and the desire of economists to emulate what they saw as the archetypical science. Accordingly, it was claimed that economics as a positive science could treat economic processes in isolation from their social environment, narrowing the scope of the discipline significantly.[11] In this formulation, the ‘art’ of policy formulation was considered to rest in the establishment of attainable policy goals, while the ‘science’ of political economy furnished the economic frameworks by which the actions of economic actors pursuing their self-interest ensured the attainment of the desired ends. As institutionalist Wesley Mitchell (1874–1948) told us in 1918, however, this move served an important ideological purpose: ‘No one can read the Austrian writers, whose general scheme was similar to Jevons’, without feeling that they are interested in developing the concert of the maximising of utility largely because they thought it answered Marx’s socialistic critique of modern economic organisation.’[12]

Typically and more recently, Richard Lipsey and Colin Harbury made the distinction between positive and normative economics in their introductory textbook.[13]

Macintyre points out that this distinction between facts and values relies on the Enlightenment’s dismantling of the Aristotelian teleological tradition of the medieval world, so that it becomes possible to conceive of the individual as prior to and independent of social roles. In contrast, in the medieval world, the argument that a ‘ought’ could not be deduced from an ‘is’ was clearly wrong,[14] and remained clearly wrong in any world where socially defined roles continued to exist.[15] Such socially defined and enforced roles carrying normative obligations are, however, an irreducible feature of any real social system, including our own.

The distinction can be traced to the Cartesian mind–body dualism in which facts are said to belong to the ‘objective’ realm of the body, whereas ‘values’ are said to belong to the subjective realm of the mind.[16] More recently, the distinction is to be found in the positivist view of science, which considers that all statements other than those that are empirical, logical or mathematical are without content—are nonsense. The idea that economics—while being a scientific discipline—is also a moral discourse is inconsistent with this demand and the latter idea had to be ditched.

Under the influence of Max Weber and the logical positivists, this distinction was transformed into a dualism between facts and values. Only judgements relating to the regularities of empirical phenomena were said to be either true or false, while normative judgements could not be considered in this way—being incapable of objective truth and objective warrant—or could not be considered at all, being left to individual judgement. Carnap, for example, called all non-scientific problems a confusion of pseudo-problems, claiming that all ‘statements belonging to Metaphysics, regulative Ethics, and [metaphysical] Epistemology have this defect, are in fact unverifiable and, therefore, unscientific. In the Viennese Circle, we are accustomed to describe such statements as nonsense.’[17]

As Rorty points out, this is effectively a demand that the only language that is cognitively meaningful should resemble the language of physics. Despite the silliness of the claim, it held sway for many years and has come to seem like conventional wisdom within economics. It is also wrong—and, according to leading American philosopher Hilary Putnam, profoundly wrong, being self-refuting! Explanatory theories do not occupy a privileged epistemological position compared with normative theories.[18] Such claims rest on untenable arguments and over-inflated dichotomies.[19] The idea of an absolute dichotomy between ‘facts’ and ‘values’ depends on a strict dichotomy, applying to all judgements, between ‘analytical’ judgements—which are tautological or true by virtue of their meaning—and ‘synthetic’ judgements, which are subject to empirical falsification (terms borrowed from Kant). In reality, very few things are black or white.

Both these claims provided the foundations for logical positivism, ignoring Kant’s own claim that the principles of mathematics were synthetic and analytical. Indeed, Kant also held that moral judgements could be justified rationally—his moral philosophy being an attempt to do so. Putnam goes on to argue that the fact–value dichotomy has corrupted our ethical reasoning and our descriptions of the world. Further, he disputes whether Hume would have approved of the way in which his advice has been used in an attempt ‘to expel ethics from the domain of knowledge’, because he was an important ethical thinker himself.[20] Putnam also tells us that the original positivist view of a ‘fact’ was of something that could be certified by mere observation or a report of a sensory experience. As we have seen earlier, however, this view of fact has been discredited thoroughly. Further, the key philosophical terms used by logical positivists—‘cognitively meaningful’ and ‘nonsense’—are not observational terms, theoretical terms or logical/mathematical terms, and yet these are the only kinds of terms that they are prepared to allow in their language of science. They are, therefore, being internally inconsistent. In any event, Quine showed to the satisfaction of most philosophers that scientific statements could not ever be neatly separated into ‘conventions’ and ‘facts’, and that the idea was a hopeless muddle:

The lore of our fathers is a fabric of sentences. In our hands it develops and changes, through more or less arbitrary and deliberate revisions and additions of our own, more or less directly occasioned by the continuing stimulation of our sense organs. It is a pale grey lore, black with fact and white with convention. But I have found no substantial reasons for concluding that there are any quite black threads in it, or any white ones.[21]

Facts and values are deeply entangled throughout our vocabulary. As we have seen in Chapter 6, normative and aesthetic judgements are essential to science itself, being the ‘good reasons’ used to justify empirical belief. In any event, little in the social disciplines meets Weber’s test of universality. On the other hand, many value judgements do meet the criteria specified by Weber: a shared method and adequate data. They are, therefore, in Weber’s terms, ‘scientific’. What Weber failed to appreciate was that the terms used in the social disciplines were invariably ethically coloured—including in his own description of his ‘ideal types’. It could, however, be preferable to speak of the ‘justifiability’ of a proposition, rather than to use the honorific ‘scientific’. It is the shared standards for such truth and knowledge claims that are important, but these standards are determined socially—including, as we have already seen, among the scientific community. What is seen as true or scientifically justified is the result of an organised and contingent consensus among an intellectual or scientific community. Consequently, a normative claim is just as susceptible to justification as any empirical or theoretical claim. The consequence of this line of argument is that the conceptual distinction between positive and normative serves no convincing intellectual purpose, while serving to privilege a particular type of discourse—a political tactic in the broad sweep of discourse.

Science is a learning process, a social process, which develops in some subcultures, and is characterised by the acceptance of an ethic—a strong value system.[22] Knowledge of the social system is an essential part of the social system itself. Consequently, objectivity—in the sense of investigating a world that is unchanged by the investigation of it—is also not achievable. The social disciplines do not merely investigate the world; they simultaneously help create the world that they are investigating. At this point, it is appropriate to recall the point made in Chapter 1 and again in Chapter 6 that the stories we tell and the vocabulary we use create our understanding of who we are and how we should act. What scientific discourse creates becomes a problem of ethical choice. Even the epistemological content of science has an ethical component. Under these circumstances, the concept of a value-free science is untenable.

Leading Swedish economist Gunnar Myrdal (1917–92)—consistent with the constructionist view outlined in Chapter 2—endorsed this point of view:

Throughout the book [The Political Element in the Development of Political Theory] there lurks the idea that when all metaphysical elements are radically cut away, a healthy body of positive economic theory will remain, which is altogether independent of valuations…This implicit belief in the existence of a body of scientific knowledge acquired independently of all valuations is, as I now see it, naïve empiricism. Facts do not organise themselves into concepts and theories just by being looked at; indeed, except within the framework of concepts and theories, there are no scientific facts but only chaos. There is an inescapable a priori element in all scientific work. Questions must be asked before answers can be given. The questions are all expressions of our interest in the world; they are at bottom valuations. Valuations are thus necessarily involved already at the stage when we observe facts and carry on theoretical analysis and not only at the stage when we draw political inferences from facts and valuations.[23]

Valuations are critical to the determination of facts and to all stages of inquiry. Consequently, value commitments are inevitable in the social disciplines. In particular, every social discipline carries an implicit definition of what it is to be human, to provide a focus for its research and to distinguish its field from those of the logician, physicist or biologist.[24] There are, however, no grounds for deciding what is an acceptable definition. It follows that the argument that neoclassical economics is a formal system—which merely explores the implications of its assumptions, the idealisations on which it is based—is not convincing. No one develops such a system for pure pleasure, but to provide a guide to policy decisions or as a justification of their ideological beliefs. Additionally, the assumptions themselves incorporate normative valuations. Furthermore, the social disciplines are moral disciplines by the very nature of the problems they deal with. Scarcity, conflict, inequality, domination, exploitation and war necessarily create problems for a stable and legitimate social order.

In any event, ‘social science’ is part of the Enlightenment tradition that instrumentalises nature and is now tending to instrumentalise human society itself. The claim for the value neutrality of science—particularly social science—is simply another highly questionable aspect of modernity. For Rorty also, the distinction between facts and values can be sustained only if there is a value-free vocabulary that renders sets of ‘factual’ statements commensurable.[25] There is, however, no such vocabulary. He argues that in choosing Galileo’s vocabulary as a model, science and philosophy have confused its apparent lack of metaphysical comfort and moral significance with the fact that it worked within a particular narrow range. Consequently, the positivists sought to eliminate subjective elements by avoiding terms that could not be linked definitionally to the terms denoting primary qualities in Galileo’s and Newton’s vocabularies. This is the seventeenth-century myth of nature’s own vocabulary—the idea that only a certain vocabulary is suitable for describing human beings or human societies and that it is the only vocabulary in which they can be understood. For Rorty, therefore, the issue between those who seek an objective, value-free, truly scientific social science and those who think it should be acknowledged as something more hermeneutical is not a disagreement about ‘method’ but a disagreement about the sort of terminology to be used in moral and political reflection.[26] To say that something is better understood in one vocabulary than another is simply a claim that a description in the preferred vocabulary is more useful for a particular purpose.

The growth of scientific and quasi-scientific knowledge has not been as beneficial socially as the Enlightenment imagined it would. The ethos of scientific rationality has consistently undermined and eroded the particular, the local, the implicit and the traditional in the name of individual human emancipation.[27] As scientific knowledge and technical expertise have grown ever more specialised, scientific experts are often able to wield power and authority through their monopolisation of esoteric knowledge and the prestige that this knowledge brings. This is the very criticism I have made of economists throughout this book. Additionally, the uncritical pursuit of social scientific knowledge works to reinforce the existing powers in society that fund that research.[28]

The common thread in this critique is the realisation that the social disciplines have an intrinsic connection with the moral and political life of society.[29] While the social scientist has an obligation to view reality as dispassionately as possible, our perceptions of reality and our assumptions about it are radically moral. There is no neutral platform of pure science utterly free from value commitments. Rather, social science is a product of the development of a particular kind of society and its lexicon. The development of Enlightenment economics clearly took place in parallel with the development of the market system and served to justify that system morally and scientifically. Nowhere is that connection more closely observable than in the period of the ideological conflict between capitalism and communism, when economics was deployed as a ‘scientific’ justification for the capitalist system.

Consequently, there is a growing body of opinion that again sees social disciplines and economics in particular as moral inquiries. Furthermore, the particular idealisations on which neoclassical economics is based are themselves based on particular ideological commitments, particularly individualism. The distinction becomes even less convincing when placed alongside the real public policy questions on which economists provide advice. Inevitably, they involve leading normative questions. The nature of property rights developed by a society clearly involves normative choices. Pareto-optimality—the major criteria for policy choice used by economists—is dependent on, and biased in favour of, the existing distribution of power, wealth and property rights, and consequently is not normatively neutral. Likewise, the choice of the goal of maximising the value of output is a normative one. Importantly, the price structure itself is not neutral. It is a function of the distribution of income, wealth and power.[30] The question of the regulation of unfair conduct is also a normative one, and the arguments used in that discourse are normative. Indeed, the advocacy of economic efficiency as the general goal of public policy is plainly a moral choice.

Importantly, the policy world is also one in which the distinction between ends and means quickly breaks down. While the distinction could have served to draw attention to the normative content of policy advice, in practice it has been used to camouflage the moral judgements being made by economists and the normative presuppositions of the market system—behind the cloak of alleged scientific objectivity. Consequently, while the idea and ideal of value neutrality persists, the confidence put in it is misplaced.

Among prominent contemporary economists, Daniel Hausman and Michael McPherson agree that the simple picture of the economist providing value-free technical information does not fit the economist who is asked for advice.[31] They summarise that economists should care about moral questions for at least the following four reasons.

  1. Behaviour, and hence economic outcomes, is influenced by the moral values of economic agents. Economists rarely describe moral commitments without evaluating them, and they affect that morality by how they describe it. They should, therefore, think about the morality that should be accepted, as well as the morality that is, in fact, accepted in society.

  2. Standard welfare economics rests on strong and contestable moral presuppositions. The standard definition of a social optimum compares social alternatives exclusively in terms of their outcomes—rather than the rightness of their procedures—and identifies the goodness of outcomes with satisfaction of individual preferences. These commitments are neither neutral nor uncontroversial. Consequently, they question the moral basis of the concern with efficiency, and whether it is any less controversial than the moral commitments that lie behind equity.

  3. Politicians and non-economists talking about welfare employ concepts that do not translate easily into the language of standard economic theory. Ideas of fairness, opportunity, freedom and rights are more important in policy making than individual preference rankings. Equating welfare with the satisfaction of preferences—which could be short-sighted or ill-informed—begs questions of justified paternalism. They question the quality of a world in which our humanity is always under the control of rational calculation.

  4. In practice, positive and normative concerns are often intermingled in policy advice.

Hausman and McPherson point out that economics embodies a commitment to a certain mode of modelling and to a normative theory of prudence. The theory of rationality is already a part of the theory of morality. The view of rationality that economists endorse—utility theory—might not even be compatible with moral behaviour, and does not provide a rich enough picture of individual choice to permit one to discuss the character, causes and consequences of moral behaviour. As Saul says, ‘If you confuse self-interest with ethics, you stumble into a false rationality— instrumentalism—in which ethics is meant to be profitable.’[32]