[This essay first appeared in the American Economist 17 (Spring 1973): 35–39. See also in The Logic of Action One.]
Economics, as a science, attempts and claims to be purely value-free; that is, separate from the personal, valuational, or political proclivities of the economist. And yet economics and economists are continually making political pronouncements; economics per se is shot through with value-loaded assumptions, usually implicit, which then emerge as political conclusions and recommendations. It is my contention that this procedure is illegitimate and unscientific, and that it is incumbent on economic theory to purge itself of all vestiges of the unsupported value judgment. As a science, economics can and should stand apart from such value judgments. But since all political policy recommendations necessarily involve value judgments, does this mean that the economist must never make any policy recommendations or, indeed, never use any terminology that is value-loaded? Not necessarily.
There are only two possible kinds of philosophical status for value judgments. Either they are all necessarily purely subjective and personal whims on the part of the valuer, in which case for the economist to remain a scientist he must indeed refrain from all policy recommendations whatever. Or these judgments may well be part of a general ethical system which is rationally and objectively demonstrable; in that case, it is perfectly legitimate for the economist, when he applies his scientific theory to public policy, to use this ethical system to arrive at economic policy recommendations. Let us take an example from medicine. A “purely” scientific, value-free medical procedure enables a physician to say that Treatment X will cure disease Y. As an applied scientist, the physician can then take this knowledge and combine it with the ethical judgment that “cure of disease is good” and indeed is the goal of his treatment, and then conclude with the “policy” conclusion that he should apply Treatment X. In this case both the patient and the physician are proceeding, implicitly or explicitly, on the basis of a deeply shared ethical system; their value judgments are neither personal nor arbitrary, but stem from a shared ethical system which pronounces health and life as great goods for man, and death and disease as corresponding evils.1
The point is that in medicine all parties proceed from the basis of a deeply shared ethical system. In the case of economics, this is scarcely true; here there are many competing and clashing values and value-systems held in society. Hence, the applied economist is in a more difficult situation. If an economist does not have an ethical system, but only subjective and arbitrary values, then it is incumbent upon him as a scientist to ruthlessly keep them out of his work. In short, the economist who lacks an ethical system must refrain from any and all value-loaded or political conclusions. (This statement, of course, is itself a value judgment stemming from an ethical system which holds that science must confine itself strictly to the search for, and the exposition of, truth.) But suppose on the other hand that an economist also holds an ethical system. What then?
It must be emphasized that if ethics is a rational and demonstrable discipline, it is self-subsistent, i.e., its principles are arrived at apart from economics or any other particular science except itself. As in the case of medicine, the applied economist would then have to take this ethical system and add it to his economic knowledge to arrive at policy conclusions and recommendations. But in that case it is incumbent upon the applied economist to state his ethical system fully and with supporting argument; whatever he does, he must not slip value judgments, ad hoc, unanalyzed, and unsupported, into the body of his economic theory or into his policy conclusions. And yet this is precisely what the bulk of economists have been doing. They, and economic theory along with them, habitually make a host of value judgments which are smuggled into their analysis, and which then permit them to make policy recommendations, implicit or explicit, without presenting or defending a coherent ethical system. Because they cannot, like physicians, work from a universally shared ethical system, it is incumbent upon economists to present a coherent and supported ethical system or forever hold their valuational and political peace.
There is no room here to cover more than a few of the outstanding examples of the smuggling of unsupported value judgments into economic analysis. In the first place, there is the familiar case of the “Pareto optimum.” If A and B trade two goods or services, they each do so because they will be, or rather expect to be, better off as a result of the trade. Surely it is legitimate then to say that A and B are both better off, and “therefore” that “society is better off,” since no one demonstrably loses by the exchange? It is implicit, and even explicit from the use of the value-loaded term “optimal,” that this exchange is therefore a “good thing.” I am sympathetic to the view that this exchange is a good thing, but I do not believe that this can be concluded merely from the fact of exchange, as the Pareto Optimum does. In the first place, there might well be one or more people in existence who dislike and envy A or B, and who therefore experience pain and psychic loss because the object of their envy has now improved his lot. We cannot therefore conclude from the mere fact of an exchange that “everyone” is better off, and we can therefore not simply leap to the valuational idea of social utility. In order to pronounce this voluntary exchange as “good” we need another term to our syllogism: we must make the ethical pronouncement that envy is evil, and should not be allowed to cloud our approval of the exchange. But in that case we are back to the need for a coherent ethical system. I believe, as an “ethicist,” that envy is evil, but I see no willingness among economists to admit the need for, much less to set forth, any sort of coherent ethical position.
This brings me to the position of the bulk of free-market economists, such as the Chicago School, who favor the free market but claim to do so not on ethical grounds, but purely on the grounds of “efficiency.” I maintain that it is impermissible to advocate the free market without bolstering one’s economic analysis with an ethical framework. Indeed, in some cases it is even impossible to set forth a coherent free-market approach without taking a frankly ethical position, and a position which goes beyond the almost universally-held utilitarian viewpoint of economists. Let us ponder our above-mentioned voluntary exchange between A and B. The free market economist advocates a world where such exchanges are legitimate and not interfered with. But any exchange implies an exchange of titles to private property. If I buy a newspaper for 15 cents, what has happened is that I have ceded my ownership of the 15 cents to the newsdealer, who in turn has granted his ownership of the newspaper to me. But this means that to advocate our right to make this exchange, means also to advocate the propriety, and hence the justice, of the existing property titles in the first place. To pronounce it “good” for myself and the newsdealer to have the right to make the exchange, means also to pronounce it “good” and just for each of us to own the 15 cents and the newspaper to begin with. Yet economists are not willing to make this extension, for to do so would mean adopting a systematic concept of justice in property titles, which would involve the adoption of a system of political ethics. Economists have generally regarded such ethical systems as beyond their province; but if so, it is illegitimate for them to advocate a free market at all.
Let us illustrate: suppose that in our presumed exchange between A and B, A has sold to B a watch which he has stolen from a third party, C. Here it becomes clear that it is illegitimate to cheer this voluntary exchange from the sidelines. For since A had stolen the watch, it was not his legitimate property, and therefore he had no right to keep it or sell it; the watch was not his legitimate title to do with as he wished. But if this is true in the case of the watch, then it would also be true in other less directly flagrant cases of unjust property titles.
Furthermore, not only is it illegitimate for the economist to advocate a free market without also adumbrating a theory of justice in property titles; he cannot even define a free market without doing so. For even to define and expound upon the free market model, the economist is describing a system in which property titles are being exchanged, and therefore he must also define and expound upon how these titles are arrived at in the first place; he must have a theory of original property and of how property comes into being.
This problem of justice in property titles also exposes a fatal flaw in the concept of the “Unanimity Principle” as a supposedly value-free guide for the applied economist. Thus, Professor James Buchanan and others have declared that it is legitimate and presumably value-free for the economist to advocate a public policy, provided that everyone can agree on such a policy. Once again, and even more than in the case of the Pareto Optimum, this position is scarcely self-evident when subjected to analysis. For the implicit assumption of the Unanimity Principle is that all existing property titles are just. The Unanimity Principle would mean, for example, that it would be illegitimate to confiscate A’s watch even though he had stolen it from C. But if we regard A’s property title as illegitimate, then we must say that A’s watch should be confiscated, and returned to C. Once again, our ethical systems intrude ineluctably into the discussion.
The well-known Compensation Principle, adopted by most economists as a supposedly value-free route for making political recommendations, is in even worse straits than the pure Unanimity Principle. (A fortiori, the “weak” version of the Compensation Principle—that compensation does not actually have to be made hut only be conceptually possible—seems to me to have no rational foundation whatever.) For the Compensation Principle assumes also that it is conceptually possible to measure losses and thereby to compensate the losers. But “utility” is a purely subjective and unmeasurable concept, and being purely psychic it cannot be measured, either conceptually or in practice. If I buy the newspaper, all that can be known is that my utility from the newspaper is greater than from the 15 cents, and vice versa for the newsdealer. There is no way of measuring these utility gains, for utility is not a quantity but a rank order of subjective valuation.
Let us take, for example, the hypothetical proposition that the imposition of a tariff on zinc is “good” or socially useful because the gainers can (and even do) take their gains from the tariff, recompense the losers, and still have monetary gains left over. But suppose that I, as a convinced adherent of free trade and opponent of tariffs, declare that my psychic loss from the imposition of a zinc tariff is so great that no feasible monetary compensation could compensate me for that disutility. No one can say me nay, and therefore the Compensation Principle falls to the ground. Conversely, the same could be true for the idea that repeal of the tariff on zinc could be advocated in some sort of value-free manner on compensation grounds. Once again, I might be such a dedicated protectionist that I could not feasibly be compensated for my psychic loss stemming from repeal of the tariff. The Compensation Principle falls in either case.
The relation between the Compensation Principle (as well as the related Unanimity Principle) and theories of justice can be starkly demonstrated from the example of slavery. During the debates in the British Parliament in the early nineteenth century on abolition of slavery, the early adherents of the Compensation Principle were maintaining that the masters must be compensated for the loss of their investment in slaves. At that point, Benjamin Pearson, a member of the Manchester School, declared that “he had thought it was the slaves who should have been compensated.”2 Here is a stark example of the need, in advocating public policy, of an ethical system, of a concept of justice. Those of us who hold that slavery is unjust would always oppose the idea of compensating the masters, and indeed would think rather in terms of reparations: of the masters compensating the slaves for their years of oppression. But what is there here for the wertfrei economist to say?
A similar argument applies to the Coase-Demsetz analysis of property rights and external cost. Coase-Demsetz declare that “it doesn’t matter” from the point of view of allocation of resources whether, for example, a railroad is given the property right to pour smoke onto the land of neighboring farmers, or the farmers are given the property right to require compensation for the invasion of their land by the railroad. The implication is that the effect is “only” a matter of distribution of wealth. In the first place, of course, the decision “matters” a great deal to the railroad and the farmers. I contend that it is totally invalid to dismiss such “distribution effects” as somehow unworthy of consideration by the economist, even though it is clear that ethical considerations are directly relevant to any treatment of such distribution. But apart from this, the Coase-Demsetz analysis is not even correct for short-run allocational problems (setting aside its validity or invalidity for long-run allocation) if we realize that social costs are psychic to the individual and therefore cannot be measured in monetary terms. One or more of the farmers, for example, may love his land so deeply that no feasible monetary compensation for the smoke loss could be made by the railroad. As soon as we admit these psychic costs into the picture, the Coase-Demsetz analysis becomes invalid even for the short-run allocation of resources. This is apart from another consideration: that in law, an invasion of property can be stopped completely by court injunction and not merely be compensated after the fact.
This brings us to the entire analysis of neighborhood effects in the economic literature. It is simply assumed without adequate support, for example, that external economies should be internalized. But why? What is the ethical groundwork for this position? Let us take an example of an external economy which no economist has suggested we internalize—not out of logical consistency but simply from empirical convenience. Women, let us say, purchase and use cosmetics; this use has a great deal of external spillover effects in conferring psychic benefits among a large part of the male population; and yet these males are “free riders”; they are not paying for the cosmetics. The neighborhood effect theorist, to be consistent, must claim that “too little” cosmetics are being used; that men are free riders on the female use of cosmetics and therefore should be taxed to subsidize females in their use. There are, of course, many problems with this doctrine, apart from those that we have already stated. The “internalizing” theorist must assume illegitimately that he can measure, even conceptually, how much men are being benefited, and gauge the precise amount of tax and subsidy. But apart from the conceptual impossibility of doing this, there are other grave problems involved in all attempts to apply such a principle for governmental action. One is that some men may dislike cosmetics intensely, and that they are therefore being penalized still further by the subsidy program. And furthermore, the very use of government implies a whole host of questionable political value judgements: for example, that government action per se involves neither psychic costs nor ethical injustice.
But there is a flaw even more directly germane to the concept of internalizing external economies. For by what ethical standard is the production and use of cosmetics “too low”? Too low for whom, and by what ethical standards? The very concept of “too low” is a value judgment which is by no means self-evident, and arrives here unsupported by any sort of ethical system.
Professor Demsetz goes on to advocate an allocation of property rights in accordance with whichever allocation involves lower total social transaction costs, such as costs of enforcing the given property right.3
But once again, there are two grave flaws in this position. One, that since social costs embody psychic costs or disutilities for each individual, it is impossible to measure and hence to add them up interpersonally. But apart from this, such a gauge for the allocation of property rights brusquely sets aside any consideration of the justice of property titles. But this itself is an ethical position unsupported by the economist. In the case of slavery, for example, it might well be found that the monetary cost of enforcing slave titles is lower than the monetary cost of each freed slave defending himself from re-enslavement. For those of us who claim that slavery is unjust, such considerations would be piddling as compared to the dictates of justice. But for an economist to try to decide such questions as the allocation of property rights by discarding considerations of justice must be totally unscientific and illegitimate.There is only space here to touch very briefly on a few other examples of the illegitimate use of implicit value assumptions in economics. One example is the long-standing aim of the Chicago School—at least until Milton Friedman’s recent essay on the “Optimum Quantity of Money”—to achieve a constant price level, either in the short or the long run. But little has been written to justify this goal. The value of the goal is scarcely self-evident, particularly when we consider the fact that a growing, unhampered economy will lead to secularly falling prices and costs, with the resulting higher living standards spread throughout the ranks of the consumers. And if falling prices should be a consequence of an increased demand for money, then again it is surely not self-evident that it is the business of government to deliberately thwart the desire of the public for a higher level of real cash balances—anymore than it is the business of government to thwart the desires of consumers for any other goods or services.
Another example is the problem of rational pricing for governmental services. Thus, in recent years, much valuable work has been done advocating market-clearing prices for such services as streets, roads, and subways; for example, that pricing be graduated in accordance with peak hours and the degree of congestion on the roads. All this makes a great deal of sense, but one vital assumption is missing: that there is nothing wrong with the fact that an increased amount of revenue will thereby accrue to the coffers of government. The implicit value assumption here is that there is nothing wrong economically or ethically with an increased amount of social resources being siphoned off to government. For those of us who do not take such a sanguine ethical view of government, this consideration must be an important factor in our policy conclusions.
In the area of government, indeed, there has been much discussion of the difficulties of national product accounting, but little has been said of the implicit—and scarcely self-evident—value assumption at the heart of the treatment of government. The blithe assumption that government expenditure on its own salaries can in any way measure government’s contribution to the national product encapsulates what some of us would consider a highly naïve view of the functions and operations of government—indeed a view that places one’s ethical imprimatur on every one of government’s activities. In these days of military overkill, and of pyramid-building on a grand scale, there are not very many people who would still automatically accept Lord Keynes’ famous dictum that building pyramids is just as productive an expenditure as anything else. In fact, anyone who believes that government expenditure contains at least 51% waste—surely not a very unreasonable assumption by anyone’s reckoning—would construct national product accounts by subtracting government expenditures as a burden upon production and upon society, rather than adding it as a productive contribution.
Finally, there is the generally held view that an economist can provide technical advice to his client while remaining purely value-free. I submit, on the contrary, that servicing a client’s ends thereby commits the economist to the ethical value of the end itself. Often it is held that by simply furnishing advice on the pursuit of goals or values held by the majority of the public, the economist remains uncommitted to values. But surely value-freedom means free of values, period; and the fact that the majority of the public might have such values does not make commitment to them any less value-laden. To take a deliberately dramatic example, let us suppose that an economist is hired by the Nazis to advise the government on the most efficient way of setting up concentration camps. I submit that by doing so, the economist has, willy nilly, adopted the pursuit of “better,” i.e., more efficient, concentration camps as a goal. And he would be doing so even if this goal were heartily endorsed by the great majority of the German public. To underscore this point, it should be clear that an economist whose value system led him to oppose concentration camps might well then give such advice to his clients as to make the concentration camps as inefficient as possible, i.e., to sabotage their operations. In short, whatever advice he gives to his clients, the economist’s value-commitment, for or against the clients’ project, is inescapable. But if this is true for concentration camps, it is true also for the myriad of other and usually less significant projects that his clients may have in mind.
I would like to cite a passage on this question from the last essay of the great Italian economist Luigi Einaudi. Einaudi wrote that the economic advisers to government “indispensable, extremely learned, extremely informed, the experts, the only people who know the jargon, have become…one of the seven plagues of Egypt, a disgrace to humanity.” A “plague,” Einaudi wrote, because of the typical economist’s view that “I have performed my duty fully when I have decided whether the proposed means or other alternatives are consistent with the end prosecuted by the politician.” Einaudi then commented: “No. The economist has failed in that case to perform the essential part of his task…The economist…has not the right to be neutral or to hide under an unreal distinction between means and ends. He must declare himself for that end to which he is closest; and must prove what he assumes.”4
It is important to stress what this paper is not saying: I am not taking the position, now fashionable in many quarters, that there is no such thing as a value-free economics, that all economic analysis is inextricably shot through with value assumptions. On the contrary, I believe that the main body of economic analysis is scientific and value-free; what I am saying is that any time that economists impinge on political or policy conclusions, value-judgments are then implicit in their analysis, and I have referred to some of the many instances where such unsupported value-judgments have entered into their discussion. My conclusion, then, is that economists must either make their value judgments explicit and defend them with a coherent ethical system, or strictly refrain from entering, directly, or indirectly, into the public policy realm.
- 1In some cases, of course, Treatment X may lead to other effects that both patient and physician may consider “harmful”; again both share a judgment stemming from a shared ethic about the evils of injury to the human organism. Both parties will then have to judge the treatment by weighing these contrasting effects.
- 2William D. Grampp, The Manchester School of Economics (Stanford: Stanford University Press, 1960), p. 59.
- 3Thus, see Harold Demsetz, “When Does the Rule of Liability Matter?” Journal of Legal Studies, Vol. I, No. 1 (1971), pp. 25–28; and Demsetz, “Some Aspects of Property Rights,” Journal of Law and Economics (October, 1966), p. 66.
- 4Luigi Einaudi, “Politicians and Economists,” Il Politico (Pavia) (June, 1962), pp. 258, 262–63.
