Production functions

The orthodox theory of the firm makes the strong assumption that it is always possible to specify a function, the so-called production function, which expresses the maximum volume of physical output that can be obtained from all technically feasible combinations of physical inputs, given the prevailing level of freely available technical knowledge about the relationship between inputs and output. It is customary to classify the inputs into more or less homogeneous classes, which ought to carry the labels “man-hours,” “machine-hours,” and “acres-per-year,” and not “labor,” “capital,” and “land,” because the inputs in question are supposed to be flow and not stock variables. On the further convenient assumption that the microproduction function so defined is smoothly differentiable and the strictly necessary assumption that the firm is profit maximizing (no value being placed on the psychic income of entrepreneurs), the theory then proceeds to derive the input demand functions as inverse forms of the marginal productivity equations. If factor and product markets are competitive, firms will hire workers, machines, and space until wage rates, machine rentals, and land rentals are equal to their respective marginal value or marginal revenue products.

If the supplies of these factor services are exogenously determined, this theory may be said to “determine” wage and rental rates. For the firm, it would be truer to say that factor prices “determine” marginal products than that marginal products “determine” factor prices.

The first edition of this book was published in 1980. Since then we have seen seven major textbooks, three books of readings, an annotated bibliography, and of course hundreds of articles, all focused on economic methodology – not bad going for a mere decade of intellectual activity in a relatively minor branch of economics.

This explosion of the literature in the methodology of economics would alone have warranted a second edition, in order to take account of new developments in the field. Moreover, my central message has sometimes been misunderstood, no doubt because it was badly expressed, tempting me to restate my argument. In addition, some of the case studies in the second half of the book were too flimsy and others needed updating. Finally, new developments in macroeconomics, general equilibrium theory, and international trade theory encouraged me to prepare a new edition.

At first, I had ambitions to double the length of the original book by new chapters on post-Keynesian economics, experimental economics, game theory, and the crisis in econometrics, resolving the clash between Bayesian and classical theories of inference. But in the final analysis, intellectual laziness and a disinclination to rush in where even angels fear to tread have produced a second edition which is only marginally longer than and different from the first.

Hume's guillotine

The distinction between positive and normative economics, between “scientific” economics and practical advice on economic policy questions, is now 150 years old, going back to the writings of Nassau Senior and John Stuart Mill. Somewhere in the latter half of the nineteenth century, this familiar distinction in economics became entangled, and almost identified with, a distinction among philosophical positivists between “is” and “ought,” between facts and values, between supposedly objective, declarative statements about the world and prescriptive evaluations of states of the world. Positive economics was now said to be about facts and normative economics about values.

Then in the 1930s, the new welfare economics came along to provide a normative economics that was allegedly free of value judgments, after which it appeared that the distinction between positive and normative economics was one between noncontroversial facts and values, on the one hand, and controversial values, on the other. The result was to enlarge traditional, positive economics to include the whole of pure welfare economics, leaving normative economics to deal with specific policy issues, where nothing much can be said about values or ends apart from what politicians tell us. What is involved here are some horrible, logical confusions that laid economists open to wholesale attack on the very idea of value-free, positive economics.

Kuhn's paradigms

We have seen that Popper's methodology is plainly normative, prescribing sound practice in science, possibly but not necessarily in the light of the best science of the past. To that extent, the Popperian methodology of falsificationism is in keeping with the received view on theories, although in many other respects it points away from the received view. In Kuhn's The Structure of Scientific Revolutions (1962), however, the break with the received view is almost total because his emphasis is not on normative prescription but rather on positive description. Moreover, the inclination to preserve theories and to render them immune to criticism, which Popper grudgingly accepts as a departure from best-practice science, becomes the central issue in Kuhn's explanation of scientific behavior. Kuhn regards normal science, that is, problem-solving activity in the context of an orthodox theoretical framework, as the rule and revolutionary science, or the overthrow of one framework by another in consequence of repeated refutations and mounting anomalies, as the exception in the history of science. It is tempting to say that for Popper science is always in a state of permanent revolution, the history of science being a history of continuous conjectures and refutations; for Kuhn, the history of science is marked by long periods during which the status quo is preserved, interrupted on occasions by discontinuous jumps from one ruling paradigm to another with no conceptual bridge for communicating between them.

To take our bearings, we must begin by defining terms.

Household production functions

The Chicago theory of the maximizing family, sometimes referred to as the new home economics, furnishes us with our last specific illustration of methodological principles. From Gary Becker's 1965 article on the allocation of time and earlier work by Jacob Mincer and Becker on fertility rates, human capital formation, and the labor force participation rates of married women, a wide-ranging research program has been developed that gives a unified interpretation to all the diverse market and nonmarket activities of families: the initial decision to marry, the decision to have children, the division of household tasks between husband and wife, the extent of participation in the labor market and even the final decision to dissolve the family by divorce.

Traditional theory views the family as a one-person household, maximizing a utility function that is defined on goods and services bought in the marketplace. The new economics of the family instead views the family as a multiperson production unit, maximizing a production function whose inputs are market goods and the time, skills, and knowledge of different members of the family. The result is not only to extend the standard tools of microeconomics to problems usually assigned to the domain of sociology, social psychology, and social anthropology but to transform the traditional explanation of consumer behavior.

The prehistory of economic methodology

A subtle but significant difference separates the methodological writings of nineteenth-century economists from those of twentieth-century ones, or rather from those of modern economists in the last forty years or so. The great British nineteenth-century economic methodologists focused attention on the premises of economic theory and continually warned their readers that the verification of economic predictions was at best a hazardous enterprise. The premises were said to be derived from introspection or the casual observation of one's neighbors and in that sense constituted a priori truths, known, so to speak, in advance of experience; a purely deductive process led from premises to implications, but implications were true a posteriori only in the absence of disturbing causes. Hence, the purpose of verifying implications was to determine the applicability of economic reasoning and not really to assess its validity. The ingenuity of these nineteenth-century writers knew no bounds when it came to giving reasons for ignoring apparent refutations of an economic prediction, but no grounds, empirical or otherwise, were ever stated in terms of which one might reject a particular economic theory. In short, the great British nineteenth-century methodologists of economics were verificationists, not falsificationists, and they preached a defensive methodology designed to make the young science secure against any and all attacks.

The Heckscher–Ohlin theorem

Ricardo found the cause of foreign trade in the relative immobility of capital across national frontiers and he explained the commodity composition of world trade by persistent differences in the productivity of labor between nations; by assuming that relative commodity prices vary proportionately with relative labor costs, he showed that free trade will cause each country to export those goods in which it possessed a comparative price advantage and that such trade will result in mutual gain as compared to a state of self-sufficiency.

Ricardian theory made no attempt to explain the underlying productivity differences that give rise to intercountry variations in comparative costs, which in turn give rise to international trade. In the modern Heckscher-Ohlin theory, these productivity differences themselves are traced to intercountry differences in initial factor endowments, which indeed are made to carry the entire burden of the explanation: the more obvious causes of the commodity composition of foreign trade, such as international differences in the quality of factors, as well as differences in production functions for given products, are deliberately excluded by assumption. The Heckscher–Ohlin theory culminates in what is now generally known as the Heckscher–Ohlin theorem (HOT) of the pattern of international trade: a country exports those goods whose production is intensive in the country's relatively abundant factor and imports other goods that use intensively the country's relatively scarce factor.

A fatal ambiguity surrounds the expression “the methodology of …” The term methodology is sometimes taken to mean the technical procedures of a discipline, being simply a more impressive-sounding synonym for methods. More frequently, however, it denotes an investigation of the concepts, theories, and basic principles of reasoning of a subject, and it is with this wider sense of the term that we are concerned in this book. To avoid misunderstanding, I have added the subtitle, How Economists Explain, suggesting that “the methodology of economics” is to be understood simply as philosophy of science applied to economics.

To ask how economists explain the phenomena with which they are concerned is in fact to ask in what sense economics is a science. In the words of one prominent modern philosopher of science: “It is the desire for explanations that are at once systematic and controlled by factual evidence that generates science; and it is the organization and classification of knowledge on the basis of explanatory principles that is the distinctive goal of the sciences” (Nagel, 1961, p. 4). There can be no doubt that economics provides plenty of examples of “explanations that are at once systematic and controlled by factual evidence,” and hence no time will be wasted defending the assertion that economics is a science.

Ultraempiricism?

The year 1938 saw the publication of The Significance and Basic Postulates of Economic Theory by Terence Hutchison, and with it the explicit introduction of Popper's methodological criterion of falsifiability into economic debates. That Hutchison should have recognized the significance of Popper's demarcation criterion as early as 1938 is itself remarkable: Popper's Logik der Forschung (1934) was then almost completely unknown, and even so famous a popularization of the philosophical ideas of the Vienna circle as Ayer's Language, Truth, and Logic (1936) completely missed the significance of Popper's critique of the verifiability principle of meaning. To some extent, even Hutchison failed to realize the novelty of Popper's thinking: although he cited Popper frequently, he laid down the fundamental criterion that economic propositions that aspire to the status of “science” must be capable, at least conceivably, of being put to an interpersonal empirical test without any acknowledgment to Popper (Hutchison, 1938, pp. 10, 19, 26–7, 48, 49, 126, 156). Hutchison's principal target of attack was apriorism in all its varieties, but in assailing the postulates of orthodox economics that were said by Mises and Robbins to be intuitively obvious, he overstated his case and so spoiled what might have been a decisive effort to reorient the methodology of interwar economics.

Hard core versus protective belt

We turn next to a theory that requires the sort of full-scale treatment that it has rarely received. The birth of human capital theory was announced in 1960 by Theodore Schultz. The birth itself may be said to have taken place two years later when the Journal of Political Economy published its October 1962 supplement volume on “Investment in Human Beings.” This volume included, among several other path-breaking papers, the preliminary chapters of Gary Becker's 1964 monograph, Human Capital, which has ever since served as the locus classicus of the subject. Thus, the theory of human capital has been with us for more than twenty-five years, during which time the flood of literature in the field has never abated, at least not until the 1980s. The first textbook exclusively devoted to the subject appeared in 1963 (Schultz, 1963). After a lull in the mid-sixties, the textbook industry started in earnest: between 1970 and 1973 as many as eight authors tried their hand at the task, accompanied by the publication in rapid succession of seven anthologies of classic articles on human-capital-and-all-that; more recent years have seen three more textbooks (Psacharopoulos and Woodhall, 1985; Psacharopoulos, 1985; and Cohn and Geske, 1990). It may be time, therefore, to ask what all this adds up to.

The received view

Anyone consulting some current textbooks in the philosophy of science will soon discover that the philosophy of science is a very strange subject: it is not, as might be expected, a study of the psychological and sociological factors that promote and encourage the discovery of scientific hypotheses; it is not an examination of the philosophical views of the world that are implicit in leading scientific theories; it is not even a reflection on the principles, methods, and results of the physical and social sciences, describing at the highest level of generality the pinnacles of scientific achievement. Instead, it appears to consist largely of a purely logical analysis of the formal structure of scientific theories, which seem to be more concerned with prescribing good scientific practice than with describing what it is that has actually passed as science; and when it mentions the history of science at all, it is written as if classical physics were the prototype science to which all other disciplines must sooner or later conform if they are to justify the title of “science.”

This characterization of the philosophy of science is now somewhat out of date, reflecting as it does the heyday of logical positivism in the interwar years. Between the 1920s and 1950s, philosophers of science did more or less agree with what Frederick Suppe (1974) has called “The Received View on Theories.”

Introduction

We are now ready to put our knowledge of methodology to practical use in the appraisal of economic theories. In so doing, we must always begin by stating what Popper calls the “problem-situation” to which the theory is supposed to be a solution. This obvious point is all too frequently neglected. Next, we must decide what it is that the theory actually predicts. This too is an obvious point and yet, as we shall see, it may be a very difficult question to answer. But since we have come so far, we must attempt to assess the evidence bearing upon the theory's predictions without, however, neglecting the nature of the “explanation” that underlies these predictions. Does the theory provide a causal mechanism that takes us systematically from the actions of economic agents and the operations of economic institutions to the outcomes predicted by the theory?

None of these questions can be fruitfully discussed if all that is available to us is a single theory. Scientific theories can only be meaningfully assessed in terms of competing hypotheses for the simple reason that methodology provides no absolute standards to which all theories must conform: what it does provide are criteria in terms of which theories may be ranked as more or less promising.