Value judgments and the principles of economics textbook.

Heath, Will Carrington

I. Introduction

Economics principles texts almost never confuse basic theoretical issues, and yet some of the leading texts do confuse a rather fundamental point of methodology. They mistakenly identify the distinction between positive and normative with the distinction between facts and values--a misidentification that is so widespread as to be axiomatic for many economists.

The following brief essay will attempt to clear up the methodological confusion surrounding the positive/normative distinction, then offer some suggestions as to what should, and should not, constitute the textbook discussion of this important topic.

II. The Positive/Normative Distinction

Most textbook discussions of economic methodology emphasize the positive/normative distinction. Samuelson and Nordhaus |12, 9~ give the subject a fairly standard treatment:

Positive economics describes the facts and behavior in the economy. What are the causes of poverty in the United States? What will be the effect of higher cigarette taxes on smokers? How has the economic performance of socialist countries compared with that of capitalist countries? These questions can be resolved only by reference to facts . . . they are all in the realm of positive economics.

Normative economics involves ethical precepts and value judgments. Should the government give money to poor people? Should the budget deficit be reduced by higher taxes or lower spending? Should the socialist countries introduce private property and stock markets? These issues can be debated, but they can never be settled by science or by appeal to facts. There are no right or wrong answers to these questions because they involve ethics and value judgments rather than facts.

This passage confuses two very different distinctions. First is the distinction between statements of "what is" and statements of "what ought to be." This, the positive/normative distinction, correctly recognizes that "ought" statements embody one or more value judgments in addition to the positive principles of economics. The other distinction is between facts and values. We refer to this as the "fact/value distinction," but it actually denotes an epistemological perspective on value judgments; specifically, that they lie beyond the realm of objective scientific knowledge. According to this perspective the "validity" of value judgments is ultimately a matter of subjective opinion (or even faith) and as such cannot be tested in any objective sense.

Samuelson and Nordhaus are not the only authors who import the fact/value distinction into their discussion of the positive/normative distinction. McConnell and Brue |8, 6~ write that "Positive economics concerns what is, while normative economics embodies subjective feelings about what ought to be"; similarly, Gwartney and Stroup |5, 13~ tell their readers that positive economics "attempts to determine 'what is'," while normative economic statements "concern 'what ought to be,' given the philosophical views of the advocate." They add that "In contrast with positive statements, normative economic statements cannot be tested and proved false (or confirmed to be correct) . . . since their validity rests on value judgments." Parkin |9, 17-18~ states that a "difference of opinion on a positive matter can ultimately be settled by careful observation and measurement. A difference of opinion on a normative matter cannot be settled in that way." Baumol and Blinder |2, 16~ hope to give their readers "a good understanding of when the right course of action turns on disputed facts, on value judgments, and on some combination of the above." Arnold |1, 20~ states that normative economics deals with "opinions that cannot be tested," adding that it would be wrong if readers "accepted as true something that we simply state as an opinion". According to Spencer and Amos |13, 5~, "Normative economics consists of subjective and/or unsubstantiated statements that lack positive knowledge or proof." Byrns and Stone |3, 17~ assert that "Positive statements may be either true or false," while "Few normative issues are settled by looking at evidence, because value judgments involve faith and argument, not scientific proof."

These authors invoke the fact/value distinction as though it were the very essence of the positive/normative distinction. Correctly understood, the latter neither entails the former nor depends upon it. To recognize that normative statements embody value judgments is one thing, but to pass judgment on the legitimacy of those value judgments is quite another matter.

III. The Fact/Value Distinction

Let us develop a little further what the fact/value distinction means. Assertions of truth may be arranged according to the degree of stringency appertaining to them. First in this ordering would be statements of pure logic, followed by assertions of empirical fact, and finally, subjective value judgments. The question is where to establish the boundaries of objective knowledge, and the fact/value distinction draws the line between empirical fact and value judgments.(1)

Acceptance of the fact/value distinction seems quite widespread today, at least among academic social scientists, but it was not always so; only in recent decades has a fairly general consensus crystallized. Even the term "value judgment" was absent from the vocabulary of social science until roughly the second half of the nineteenth century. The Aristotelian scheme of explanations included four "causes": the material, the efficient, the formal and the final. The final cause, or the purpose for which a thing exists or a change takes place, was teleological, yet considered necessary for a full understanding of objective reality. During the Middle Ages every explanation of natural events was value-laden, and even Copernicus maintained that it was better for the stars to be at rest than for the earth to be so, because the stars are nobler and more divine.

By the time of Newton, interpretations of nature's processes were given in terms of "efficient" causation mostly; the "final" cause had lost its place in science as such. This "modern" conception of science was gradually assimilated into the social sciences, due largely to the work and person of Max Weber. Science meant to Weber the exploration of cause and effect relationships meticulously kept separate from all value judgments.

The attempt to separate value judgments from pure cause and effect relationships is entirely appropriate, and not especially controversial, but Weber actually went much further. When he assumed the editorship of Archiv fur Sozialwissenschaft und Socialpolitik, Weber declared |15, 55-57~ that "to judge the validity of . . . values is a matter of faith" and that "it certainly does not fall within the province of an empirical science in the sense in which it is to be practiced here" (emphasis in the original). The search for generally valid value judgments was not to be the task of the Archiv. Nor could it be "in general the task of any empirical science," for "|s~uch a thing would not only be impracticable; it would be entirely meaningless as well."

Though it seems that many, perhaps most, social scientists today would agree with Weber that the search for scientifically valid value judgments is "entirely meaningless," it should be recognized that some respected twentieth-century scholars have held forth against this view. The eminent political scientist Eric Voegelin, for instance, maintained |14, 11~ that the fact/value distinction "made sense only if the positivistic dogma was accepted on principle; and it could be accepted only by thinkers who did not master the classic and Christian science of man," adding that "Only when ontology as a science was lost, and when consequently ethics and politics could no longer be understood as sciences of the order in which human nature reaches its maximal actualization, was it possible for this realm of knowledge to become suspect as a field of subjective, uncritical opinion."

Among economists, Wilhelm Ropke was perhaps the most eloquent critic of the fact/value distinction. He believed that some value judgments are more subjective than others--that is, more subject to individual tastes, preferences and fancies |10, 11~:

The popular indignation of scientists at value judgments seems to suggest that it were more or less like a matter of the taste in neck-ties. In the case of neck-ties our judgments . . . will be highly various, arbitrary and subject to individual fancies . . . But taking all possible value judgments together we observe that the degree of subjectivity may range from anything like zero to 100 per cent. In a great number of cases, i.e., in that of the more vital and comprehensive judgments of value, the degree of subjectivity becomes so negligible as to give them an objective character. Save again for the "idiot fringe", we all agree on them because they are part of the normal make-up of man.

Clearly Ropke wished to establish the boundaries of objective knowledge somewhere within the sphere of value judgments. Who but a lunatic would prefer disease to health, Ropke would ask, or choose the actions of a Hitler over those of a Mother Teresa? Why should value judgments about such matters not be accorded the status of objective fact? But important questions remain unanswered, perhaps unanswerable: Precisely how does one know when a value judgment is "less" or "more" subjective, or how to argue the difference with someone else who fails to see it? And how would one objectively determine "the normal make-up of man"? There is little reason to expect a consensus on these and similar kinds of questions.

The ambiguity and inevitable "messiness" of serious moral discourse tends to repel many economists, particularly those most enamored of the simplicity and precision of formal economic analysis. Yet in recent years it seems that moral philosophers and economists are again learning from one another, having discontinued virtually all interdisciplinary communication during the time coinciding (but not coincidentally!) with the heyday of positivist methodology in both disciplines. Hausman and McPherson |6~ have surveyed the recent contributions to normative economics by the diverse works of such moral theorists as Richard Brandt, John Broome, Ronald Dworkin, James Griffin, Russel Hardin, Richard Hare, John Harsanyi, Shelly Kagan, Tibor Machan, John McDowell, Thomas Nagel, Robert Nozick, Derek Parfit, Peter Railton, John Rawls, John Roemer, Amartya Sen, and Peter Singer, to name just a few.(2) Hausman and McPherson |6,678~ urge all economists to acquaint themselves with at least a smattering of modern moral philosophy because it will "help exorcise unjustifiable skeptical doubts (to which economists have been prone) about the possibility of rational argument concerning values." But their cautionary words |6, 712~ also bear repeating here: "Very little in ethics is completely uncontroversial . . . We've tried to provide a rough map of the terrain, but we make no guarantees that those who venture in will not find some surprises. It's tough and tangled territory."

The objective status of value judgments is an enormously, beguilingly complicated issue, and an adequate examination of it would take us very far afield; it is not, at any rate, the purpose of this note. The point to be made here is that the axiomatic assertion of the fact/value distinction, as one finds in many textbook discussions, fails to do justice to the subject, and to those scholars who have thought so incisively and written so eloquently about it.

IV. Value Judgments and the Principles of Economics Textbook: A Suggestion

Authors should drop the fact/value distinction from principles-level textbook discussions of the positive/normative distinction. Normative analysis embodies value judgments in addition to the positive principles of economic science, and of course it is appropriate to emphasize this point in a textbook discussion of economic methodology. It is appropriate also to emphasize that these value judgments are not among the principles of economic science. But little else need be said about value judgments or the question of their epistemological status.

A good illustration of the approach recommended here is offered by Henderson and Poole |7, 11-12~. They stress that "normative analysis depends centrally upon value judgments", but nowhere do they assert that value judgments as such cannot be evaluated. And their claim that "|e~conomists do not bring to normative analysis any special claim to expertise regarding preferences or goals" nicely delineates economics from moral philosophy, but wisely leaves the matter at that. Ruffin and Gregory |11~ also refrain from pronouncing on the objective status of values. Their discussion |11, 13~ of normative economics emphasizes instead the economist's instrumental role--sorting out the likely effects of this policy or that--with special attention to the Hayekian "principle of unintended consequences". They see this principle not as "a council for despair". Rather, "the principle tells us that the challenge |of normative analysis~ is to develop good economic policies, not superficial solutions." Likewise Colander |4, 23~ stresses the instrumental rationality of normative analysis when he writes that "in normative economics, economists try to figure out objectively what will make society better."

One final point. An admittedly subjective normative proposition underlies this whole discussion: that economic instructors generally, and not only textbook authors, ought to be more clear, and perhaps less glib, about the assumptions that underlie economic analyses. Certainly those who insist upon making the fact/value distinction one of those assumptions ought to present it explicitly for what it is: a philosophical perspective on the status of value judgments. Then students can judge for themselves whether it seems legitimate or false, helpful or not.

Intellectual honesty would seem to demand that the fact/value distinction be presented as a debatable proposition; at the very least, clarity of thought requires that it not be confused with the positive/normative distinction. Students, and their teachers, deserve both in their principles textbooks.

1. This discussion borrows from Ropke |10, 3~.

2. For further discussion and particular references, see Hausman and McPherson |6~.

References

1. Arnold, Roger. Economics. New York: West Publishing, 1992.

2. Baumol, William and Alan Blinder. Economics: Principles and Policy. New York: Harcourt Brace Jovanovich, 1988.

3. Byrnes, Ralph T. and Gerald W. Stone. Economics. New York: Harper Collins College Publishers, 1993.

4. Colander, David C. Economics. Homewood, Ill.: Richard D. Irwin, Inc., 1993.

5. Gwartney, James and Richard Stroup. Economics: Private and Public Choice. New York: Dryden/Harcourt Brace Jovanovich, 1992.

6. Hausman, Daniel M. and Michael S. McPherson, "Taking Ethics Seriously: Economics and Contemporary Moral Philosophy." Journal of Economic Literature, June 1993, 671-731.

7. Henderson, Vernon and William Poole. Principles of Economics. Lexington: D.C. Heath, 1991.

8. McConnell, Campbell and Stanley Brue. Economics: Principles, Problems and Policy. New York: McGraw-Hill, Inc., 1993.

9. Parkin, Michael. Economics. New York: Addison-Wesley, 1993.

10. Ropke, Wilhelm, "A Value Judgment on Value Judgments." Revue de la Faculte des Sciences Economiques d Istanbul Nos. 1/2, 1942, 1-19.

11. Ruffin, Roy J. and Paul Gregory. Principles of Economics. New York: Harper College Publishers, 1993.

12. Samuelson, Paul and William Nordhaus. Economics. New York: McGraw Hill, 1992.

13. Spencer, Milton H. and Orley M. Amos, Jr. Contemporary Economics. New York: Worth Publishers, 1993.

14. Voegelin, Erik. The New Science of Politics. Chicago: University of Chicago Press, 1952.

15. Weber, Max. The Methodology of the Social Sciences. New York: The Free Press, 1949.