Preface

This heretofore unpublished paper is now nearly 23 years old.[1] It was written at a time when what is here called “Pareto thinking” was often expressed in public debate—particularly political debate—in order to justify recommendations of policy or particular political decisions; and it was also used more widely to defend the propriety of certain commercial transactions. Vilfredo Pareto, or something called “the Pareto Principle”, might or might not have been mentioned in these connections.

I daresay that many of those who enunciated “Pareto thinking” had never even heard of Pareto himself. But what they had in common was that they thought of what is herein called “the Pareto Principle” (whether or not they thought or spoke of it under that name) as a carrying with it a kind of moral or equitable mandate that favored, or even demanded, the implementation of one or another policy decision or justified one or another commercial transaction—a mandate somehow grounded in the science of economics.

Strictly speaking, at the focus of my discussion in the paper was and is what might be more properly called the “Pareto Efficiency Principle”, which is different from Pareto’s so-called “Principle of Factor Sparsity” or “80-20 Rule”. The latter is less familiar among ordinary folk, but is also referred to as the “Pareto Principle” by political economists.

As a principle of what economists call “efficiency”,[2] Pareto’s principle may be applied either to efficiency in production, which is not discussed in the paper, or to efficiency in allocation. As indicated to my reference to policy decisions and commercial transactions—vehicles for the distributive allocation of benefits (goods and services) and burdens (costs and privations)—it is the latter that is discussed in the paper. Since it is easier to give examples based upon commercial transactions, the paper refers more often to them than general policies; but all of the examples are tied to the question of the “Pareto efficiency” of some allocative action.

Now, an astute peer-reviewer of this submission has doubted that any actual economist would view the Pareto Principle as carrying with it any normative mandate or approval for any policy or transaction—any claim that a policy or transaction that is Pareto is something that ought to be implemented or carried out. It is a principle of efficiency, this reviewer emphasizes, not a principle of equitable allocation or of normative merit, and this is something, the reviewer insists, that an economist would know. I am not sure to what extent ostensibly “Pareto” thinking is still commonly employed as a tool in economic rhetoric as it certainly was so when this paper was written, but there are indications that it has not disappeared.

Sean Ingham[3] has remarked, in an encyclopedia entry on “Pareto-optimality”[4] (often equated in economics text books with “Pareto-efficiency”), that:

Economists typically find Pareto-optimality to be extremely plausible—indeed, indisputable—as a condition that good laws, policies, and allocations must satisfy, although few would claim that it suffices to make a law, policy, allocation of commodities, and so on, good.

I find this statement fuzzy in some respects, but it is at least a claim that economists “typically” view Pareto-optimality as an “indisputably” necessary condition for the “goodness”—or let us say the propriety—of various allocative instruments, even if it is not sufficient. Ingham certainly finds that, with respect to Pareto-optimality, it is not unusual for economists to “lapse into their normative mode”, as the celebrated economist, Uwe E. Reinhardt (1937-2017), worded it when he issued this warning to his students in a Princeton University economics course in 2016:

When economists or other policy analysts lapse into their normative mode – that is, when they pretend to be using scientific methods to suggest what ought to be done and what is or is not “efficient” or “welfare enhancing” –  a red warning light should go on in your mind. There is always the chance that you are being addressed by someone knowingly and cleverly playing politics in the guise of science, or by someone commissioned by an interest group to structure information felicitously, toward a desired end – that is, by someone insufficiently respectful of the limits of economics as a science. Finally, there is always the chance you are face-to-face with an economist who has never been properly taught or, in any event, has never fully grasped the meaning of the word “efficiency” within economic analysis.[5]

Taking Reinhardt at his word, we can grant to my reviewer that by an economist who is knowledgeable, clear-minded, and honest, Pareto efficiency per se will not be thought or said to carry a justification or claim of propriety for allocative measures that support or tend toward it.[6] Given these background remarks, I now turn to the body of my paper.

I.

Despite the fact that the Italian economist, sociologist and philosopher Vilfredo Pareto (1848-1923) is frequently named in passing as an important thinker—in particular, as a foundational thinker in microeconomics—a surprising number of contemporary economists—not to mention policy makers—are apparently unfamiliar with some of his signature ideas. Among the vulgar, Pareto’s only real legacy is the so-called “Pareto Principle”,[7] which was originally advanced as a principle of “efficiency” pertaining to—among other things—commercial transactions within a free market. Roughly speaking, the idea is that in an unrestricted market, “rational agents”, pursuing their own interests, or policy makers, supposedly pursuing the public good, would engage in transactions, or policy decisions, which yield Pareto improvements.[8] A Pareto improvement is a change which makes at least one person better off and makes no one worse off. And since rational agents in an unrestricted market prefer to make transactions in that way, a free market[9] will tend toward Pareto optimality, or so contemporary Pareto fans believe. A situation is said to be Pareto optimal if there is no way to change it that would represent a Pareto improvement.[10]

II.

In the recent climate of “free-market liberalism”, Pareto thinking has been widely applied by politicians and legislators in decision- and policy-making on a larger scale; it has, for many, become their credo, or at least their cover story. For it is vaguely implied that Pareto improvement, and convergence upon Pareto optimality, must be a good thing. It certainly sounds as if making at least one person better off while making no one worse off must be a good thing—the thing to do, whenever possible.

In their moralistic moments, certain economists and policy-makers would like us to understand that the pursuit of Pareto optimality is what really is best for everyone;[11] although when gripped by the scientific mood, these same people may be heard to assert that sound scientific and practical thinking should keep clear of moral concepts like “good”, “right” and “just”—these simply express subjective emotions—and should rely upon more tangible and measurable (and supposedly objective) notions like “preference” and “choice”. Then, Pareto allocative optimality is defended as the convergence point of what rational actors would pursue, individually and communally, in an unrestricted marketplace, free of coercion. Anyone who opposes a Pareto-driven approach is thus thought to have declared himself the enemy of individual autonomy and free choice, and the friend of coercion—although of course no “value judgement” is implied in asserting this!

III.

I believe that there are many reasons why one should not be drawn into this sort of Pareto euphoria, but most of those reasons will be left out of my present discussion. In this brief piece, I will not examine Pareto optimality, which has been the focus of much critical work, but will be primarily interested in the idea of Pareto improvement, which is in any case the more fundamental of the two notions. What I am going to ask is whether there is an imperative of morality or normative merit supporting Pareto improvements. As the term “imperative” indicates, I am asking a deontological question: a question about obligation or duty.

I will examine the strong deontological statement, which says, of a prospective action A:

(P) A ought to be done if and only if doing A would make someone better off and make no one worse off.

 

This deontological version of the Pareto Principle can thus be broken down into two claims:

(P1) If doing A would make someone better off and make no one worse off, then A ought to be done.

(P2) A ought to be done only if doing A would make someone better off and would make no one worse off.

(P1) says that an action producing a Pareto Improvement is a sufficient condition for its being normatively mandated, while (P2) says that an action leading to a Pareto Improvement is a necessary condition for its being so.

Now the question whether P could even possibly hold as a mandate of morality or normative merit is dependent upon how we understand the notion “better off”. One way to understand it is in terms of the satisfaction of wants or desires, so that it could be formulated as follows:

(Pw) A ought to be done if and only if doing A would result in at least one person’s wants being more fully satisfied and would not result in anyone else’s wants being less fully satisfied.

But this again requires further specification as to which wants are being considered. The wants in which economists have been interested are immediate or present wants at the time of a possible exchange, and this is sometimes called, technically, a person’s interest. Certain economists, at any rate, evidently subscribe to the technical definition: [12]

Y is in S’s interest =def. S wants Y

This technical sense of one’s “interest” is highly misleading; it is, in my opinion, an unfortunate terminology. That is because the more normal and natural idea of something’s being in one’s interest is not at all the same thing as its being wanted. Thus, a small child may want to play in a busy street, but that is not at all in the child’s interest. Someone may, at a given moment, want to smoke a cigarette but may not even think himself that doing so is what is in his interest.

I will not use “being in one’s interest”, therefore, in its artificial, technical and very misleading sense, but will stick to speaking of wants, when that is what I mean. Returning, then, to (Pw): suppose we think only of the wants that a person may have at a given moment. Then, I think, (Pw) loses any plausibility as a moral imperative. This is related to the fact that I am unconvinced that the satisfaction of anyone’s immediate want is per se (i.e., just because it is someone’s want) of any value.

I note here, parenthetically, that if we do not restrict the wants referred to by Pw to those that a person may have at a given moment, then Pw provides no clear instruction; even misguided instruction. Since we can want a thing at one time and not at another, and wants are fickle and often transitory, then Pw is not only subject-relative (goods are only goods-to-someone), but time-relative: to be a good to someone at t is to be desired by that person at t. It is often very difficult to discern what people might want at the moment of an allocative decision or transaction, but what they might want, or not want, later on, or “in the long run” is even more opaque, not least to the people themselves. And so it will be highly uncertain, even positively indiscernible, what the supposed normative mandate requires.[13]

At any rate, the question whether the satisfaction of anyone’s immediate want is per se of any value is a fundamental one and I think one over which there is deep disagreement. It is related to a wider disagreement about human nature and the character of human life. The tradition associated with Hume (and perhaps Hobbes) maintains (or simply assumes) that goods are created by wants or desires.[14] That is, something is a good, to someone, if and only if that person desires it. To be a good, to someone, is to be desired by that person.

IV.

There is, however, another tradition, which denies that something is a good, to someone, if and only if that person desires it. In modern philosophy, its most powerful exponent is Kant. Kant thought—among other things—that there were goods, or values, whose worth was entirely independent of anyone’s wants or desires. And he thought, moreover, that we could recognize and commit ourselves to these values, that we could be moved to pursue them, whether or not we had any desire for them. This last part is a motivational theory.

The Humean motivational theory is different. According to it, one is moved to pursue only what one wants or desires. In Hume’s language, all motives are passions.[15] This Humean motivational theory benefits from the great elasticity of the idea of a desire or a passion. Thomas Reid complained about Hume that he used these terms—desire and passion—simply to refer to any motivational state. If this is done, then of course Hume’s motivational theory will be bound to be true. But it will be true just by definition; the notions of desire and passion (or want) will lose any independent substantive content. We will not have made any interesting discovery to the effect that desires, or wants, or passions, as independently characterizable states of mind, have a causal influence on our actions—are motives—and indeed the only motives according to Hume.

The Humean point of view would be interesting if we had an independent characterization of wants, desires and passions, and could argue plausibly for the non-definitional thesis that these things, and only these things, motivate.

Now, as regards the question of self-interest and benevolent action, we can consider alternatives to the Humean motivational theory (in its substantive, non-definitional form). If I act benevolently, i.e. in the sense of promoting the interests of others, must it be because that is what I want or desire? Or because this is in any interesting sense in my interest? Perhaps so, but it is not clear. What is clear is that the motive upon which I act will always be my motive. This isn’t enough, I think, to bring my action under my interest. In any other sense (i.e. other than that I always act on my motive), it seems easy to name examples of actions that I might perform which are not in my interest, actions which it is clear to me are not in my interest. It also seems easy to name examples of actions I perform without wanting or desiring to do them. One may wish to deny or explain away these examples, and that may be possible; but one will have to be very careful not to be simply stretching the meaning of “desire” or “interest” to fit the case. What’s wanted is a carefully worked out characterization of these terms.

V.

Returning to our main theme, it seems to me that people’s interests are a more plausible basis for a Pareto-based normative imperative than are their immediate wants. I think that the mere fact that something is wanted or desired gives no basis for normative obligation, not even presumptively. The reason is that desires can be for things which are utterly trivial, or which are contrary to people’s interests (not least, the wanter’s), or which are simply malign. So, I don’t think that the fact that someone has a yen to drink a can of paint (to take an example from Donald Davidson), or to crack their knuckles, or to have sex with a certain other person creates any obligation upon anyone to bring about the fulfillment of these desires. It is where wants are attached to genuine interests or needs—or at any rate to something of a different character than mere wants—that obligations may possibly be created.

Now I said that I thought that interests would be a more plausible basis for a Pareto imperative than wants. When I say this, it should be clear that I am not using “interest” in the technical way explained above. Something is in a person’s interest if it improves her life, is good for her, makes her genuinely “better off”, whether she happens to want it or not.

If this be granted, it may nevertheless be thought that a person’s interests come down to what he wants or desires all things considered, and in the long run: “all things considered”, because a person may have various conflicting wants, so fulfilling some at the expense of others may not be in his interest (e.g. if the ones fulfilled are valued less than the ones that aren’t); “in the long run”, because what one wants now may lead to unforeseen, and undesired consequences: consequences not desired by that person himself later on, when they come to light. So the idea is that interest is determined by desire, but not—or not merely—by immediate, individual desire.

Many (among them Thomas Reid and John Stuart Mill) have found this persuasive, but I do not. For if a person has a distorted or limited perspective, he may conceivably fail to build a genuinely fulfilling life for himself even though in his life, his actual desires are fulfilled all things considered and in the long run. A person may have a set of desires which is not the best set that he could have had, or even a very good set. And it makes perfect sense to say that his interest would have been better served—he would have been better off—had he had a different set of desires. So a couch potato, or a Gamma in Huxley’s Brave New World, would have been better off with a different set of ideas than those they actually have. And even if many of those different desires had been unfulfilled, a person’s life might have been better than in his actual life where all of his desires are fulfilled.

J.S. Mill warned that we should not confound the two very different ideas of happiness (as he called it) and contentment. Contentment is the state where one has no, or few, unfulfilled desires. It is a kind of desiderative Pareto point. Happiness—by which Mill meant not a state of mind, but well-being—living a good life—is a different thing.

My argument so far may be questioned, or objected to, at many points. But if we accept at least temporarily, for the sake of further discussion, that if (P) holds as a moral imperative, we must understand “better off” to mean genuinely better off (all things considered and in the long run, whatever that comes to). What it does not come to is merely having one’s desires fulfilled.

That said, I want to stick in a word in favor of wants and desires—or perhaps, better, preferences in the sense that economists like to use this term—as metrics for being better off. A person may not have a good idea about what would make her better off; and her preferences may not be in accord with her real interests. This metric of preferences is therefore imperfect and incomplete. But in general (although rebuttably), individual people are, or should be taken to be, the best judges of their best interest or what would make them better off. Preferences are the most important general metric we have for this; and in practice, in a society with an enlightened citizenry, it is the metric upon which it is best to rely as a general rule. But it is not the only metric and it does not carry absolute privilege. And when the limitations and imperfections of this metric (not least the way in which it is bound to function in a morally dysfunctional society) are overlooked or denied the standard of preferences may become malicious.

VI.

Now, how does (P) come out as an imperative of normative merit in allocation-making, given what I will call the strong reading of “better off”? Let us consider (P1) and (P2) separately. Together, we recall, they add up to (P).

(P2) is, at least, unconvincing as an imperative of that kind. For (P2) says that an action A ought to be done from the point of view of normative merit only if doing A would make somebody better off and would make no one worse off. But it seems quite clear that we often have obligations which, if carried out, will make at least some people worse off, for instance, the moral obligation to hire the best qualified candidate for an academic job, which will (or might) make less qualified candidates worse off.[16] More dramatic cases involve the morally necessary sacrifice of some people in cases of war or shipwreck, but it is the relatively banal examples which are the most telling. There is practically no choice that we make which does not make someone worse off to some extent: certainly no large-scale or political policy decision will be of that kind. But many such choices and decisions are morally defensible and indeed even obligatory.

Here we should pause for a moment to notice the general inapplicability of the Pareto approach to large-scale decision-making, not only from the deontological, but even from the more strictly and narrowly economic and political points of view. The approach makes a fair amount of sense in its primary application, which is to the isolated commercial transaction: Fred and Roger decide in a free and uncoerced manner to exchange x for y; since they both agree freely to the exchange, both of them must think that they will be better off for making it, and neither will think himself worse off. But here our attention is restricted only to those who make the transaction, Fred and Roger, leaving out of account any others who may be affected by it (for instance, Shalimar the slave-girl might be the y which Roger exchanges for Fred’s sum of money x; and she might be made much worse off by the exchange). Economists commonly refer to such considerations as “externalities”. When a theory is such that, in application, “externalities” are bound to matter much more than the parameters posited by the theory, then it is the validity of the theory that is thrown into doubt, rather than the factors categorized as externalities.

As soon as we move from the narrow consideration of the preferences of the transaction-makers taken in isolation, to include the preferences (not to mention the genuine interests) of third parties affected by a transaction, we quickly get an idea of just how few transactions there are which actually make no one worse off. Taking all affected parties into account, the number of real Pareto improvements is vanishingly small, even for individual transactions; for large-scale policy-making decisions, I cannot think of a single real case. The realization of what even a small transaction may mean for third parties also raises serious questions about the Pareto approach as a panacea for those who value individual autonomy and free choice.

VII.

At any rate, having disposed of (P2) as a plausible moral imperative, we are left with (P1). This says that if an action A would make someone better off and no one worse off, then A ought to be done. This, too, has evident problems.

For instance, two alternative actions, A and A’, might be actions of the kind described, but of benefit to different persons. Say that by doing A, some person S would be made a bit better off and no one worse off; but by doing A’, S’ would be made much better off and no one worse off. Then, evidently, there would be no moral imperative supporting A. And if a perceptible alternative A’ would make everyone much better off and no one worse off, then, certainly, if there is anything that ought to be done, it would evidently be not be A. Suppose now that A’ would make most people much better off, but would make a few people a bit worse off. Then if there is any normative mandate, it would seem, once again, not to attach to doing A—insofar, that is, as the foundation of this imperative is supposed to be that of making people better off.

However, in favor of (P1), it might be maintained that it is, nevertheless, what is sometimes called a prima facie moral imperative. In other words, if an action represents a Pareto improvement on the strong reading of “better off”, it is something that arguably ought to be done, although one which may be overweighed by other moral obligations: for example, the obligation to make the existing allocative landscape tolerably just.[17]

With the caveats included, I have no particular objection to offer to this, and I cannot see that even Kant would need to have one, given a suitable understanding of “better off”. But it will then be one principle among many, and one of meager significance since (1) so few of the possible actions that we might take lead to Pareto improvements of the kind imagined by the devotees of this principle and since (2) its normative input will be so heavily qualified and so easily overweighed by conflicting considerations of propriety.

VIII.

Before closing, I would like to add some remarks about the Pareto Efficiency Principle as a normative principle. Part of the cover story of contemporary economics, in its several “mainstreams”, is that any such principle is non-normative. My peer-reviewer sought to disagree with me by saying that no proper economist would think of Pareto’s principle as carrying a normative mandate applicable to allocative decisions. I suppose that Pareto himself would resist thinking of it as such. “It is a principle of efficiency,” my reviewer insisted, “not a normative principle or moral imperative.”[18] Actually, my reviewer, in saying this, was more in agreement with me than in disagreement. (But the disagreement is important.)

To begin with the agreement, what I was critiquing was interpreting the Pareto Efficiency Principle” as carrying a mandate for implementing certain allocative decisions or policies, or at least as carrying a recommendation favoring their implementation. Now my reviewer maintained, in agreement with me, that this should not be done, and opined that no proper economist would do so. I maintained, however, and still maintain, that this was done by a variety of policy makers and economists to support the pretense that these recommendations were issued by science, and therefore respectful of Weberian Wertfreiheit. “Efficiency” is widely claimed by those economists who reflect upon their modus operendi, to be a non-normative matter; but their commitment to this claim is belied by their at the same time using Pareto-efficiency as a reason justifying certain allocative decisions as the “right” ones—and even as the normatively mandated ones—to make, without any arguments that back up the drawing of such inferences from the “fact” of efficiency.

Reihardt, in the lecture text cited earlier (see footnote 5) cites Kenneth Arrow’s observation that

A definition is just a definition, but when the definiendum is a word [like “efficient”] already in common use with a highly favorable connotation, it is clear that we [economists] are really trying to be persuasive; we are implicitly recommending the achievements of optimal states [such as Pareto-optimality][19]

In other words, despite their dutiful protestations, economists who appeal to “efficiency” as a justifying reason for implementing certain allocative decisions without any further supporting argument, have turned “efficient” into a normative term and left Wertfreiheit in the dust. That way, they avoid revealing their supporting reasons, which might appear odious if exposed to the cold light of day. Reihardt himself says, in that same lecture text, that:

Honorable economists will always be quick to apprise their audiences and clients of the distributional impact of proposed policy changes. Alas, the real world is full of “seasoned economists” who do not, either because they wish to push their own ideology, or because they are paid to abstract conveniently from the distributional consequences of proposed policies  –  e.g., proposed health reforms or trade policies. “This policy is efficient,” they may loudly proclaim, adding inaudibly under their breath “abstracting from distributional effects, of course.” Sadly, some policy makers buy this stuff, either because they innocently believe the economist, or because they find it profitable or ideologically soothing to do so.

I suspect that Pareto himself was rather in the forthright camp, which is why he is anathema to a large number of political economists who admire his scientific work but dislike the way in which he wanted to see it applied.

Whether or not we turn out to agree with it, we should take seriously Reinhardt’s view that (quoted earlier in footnote 5), that:

Modern applied welfare economics may look like objective science, because it typically is cloaked in mathematical symbols or graphs. At its core, however, economic welfare analysis is but one particular distributive ethic. It is just one of many different moral doctrines.[20]

And Reinhardt says, rather open-mindedly to his Princeton University students:

You may or may not buy that particular distribute ethic. Whether or not you do may possibly depend on—you guessed it—your wealth and ability to offer high bid prices for things.

IX.

What will by this time be apparent to moral theorists or, indeed, to anyone who has thought carefully and systematically about such matters, is that the Pareto Principle, as a Consequentialist moral norm for decision-making, invites comparison with more familiar Utilitarian principles. A good Utilitarian is, I think, bound to conclude, that the Pareto Principle, which is in an obvious way Utilitarian in spirit, has no independent authority but can be given weight as a moral norm to the extent, and only to the extent, that it is sanctioned by more fundamental (and overarching) Utilitarian principles. The arguments given above against the weight of (P1) and (P2) are, at bottom, Utilitarian arguments. It is, however, questioned by more deontological thinkers, whether any kind of Utilitarian theory is capable of giving a satisfactory account of normative obligation or, in words I have used here (echoing Kant), of “normative imperatives”. No position is taken on that issue in this paper. Rather, an argument concerning the limitations of the Pareto Principle as a moral norm is framed within the general theoretical context of Utilitarian, or at least Consequentialist, reasoning to which the Pareto Principle itself belongs. In other words, the Principle is critiqued on its home ground, so to speak. The exploration of the deeper questions concerning allocative imperatives would obviously require a great deal more time and space than I have at my disposal here. And for extraneous reasons, I wanted to keep Vilfredo Pareto in focus throughout. In short, my objective in this paper is relatively modest, but not, I think, trivial. Deeper discussion of the issues raised here must await another occasion.

 

Endnotes & references

[1] Originally composed as a lecture text under the title, “Equity and the Pareto Principle—Does the Pareto Principle Have Moral Force?”, presented to students and teachers at the SOCRATES Intensive Program, “Equity versus Efficiency in Political and Economic Decision-making”, University of Rennes I, 16-26 July 2000 and, within the same period presented as an invited paper to the Department of Philosophy of the University of Rennes 1, 24 July 2000. Revised for presentation under the title, “Does the Pareto Principle Have Moral Force?”, to the Department of Philosophy of the University of Genoa, 10 April 2001 and much later lightly revised for presentation to the Seminar of the Faculty of Economics, University of Iceland, 25 February 2011. Subsequently submitted to Nordicum-Mediterraneum in response to a call for papers on Vilfredo Pareto to be published in commemoration of the centenary of Pareto’s death on 19 August 1923 and revised in response to critiques made in blind peer review.

[2] A principle rather than the principle, since economic “efficiency” may be characterized in other ways than Pareto’s, although few alternatives are discussed.

[3] Sean Ingham is (or was when he contributed the entry) an Assistant Professor of Political Science at the University of Georgia and evidently a regular contributor to various respectable encyclopediae.

[4] The Encyclopedia Britannica on line; see https://www.britannica.com/topic/Pareto-optimality

[5] The source for this is a lecture text, or handout, entitled “The Concept of ‘Efficiency’ in Economics” that has very recently come into my hands. The main lesson is 18 pages, followed by three appendices, two pages on “The Efficiency of Production”, six pages on “How Economists Bastardized Benthamite Utilitarianism and Became Shills for the Well-to-Do”, and one page on “The First and Second Theorem of Optimality in Welfare Economics”, where he references Kenneth J. Arrow, “Uncertainty and the Welfare Economics of Medical Care”, American Economic Review, December, 1963, pp. 942-943. The conclusion of the second appendix reads as follows: “Modern applied welfare economics may look like objective science, because it typically is cloaked in mathematical symbols or graphs. At its core, however, economic welfare analysis is but one particular distributive ethic. It is just one of many different moral doctrines.” This is not worded as a claim about contemporary economics generally, nor is it a claim about the Pareto Efficiency Principle as such. But it reflects an orientation that is very close to my own, and Reinhardt’s discussion of the uses of Pareto’s Efficiency Principle make a number of the critical points made in this paper and more. Unfortunately, Reinhardt seems never to have published his lecture text, leaving room for Nordicum-Mediterraneum to publish mine. It does make me wish that I had been fortunate enough to be a student in his classes.

[6] And note that this does not rule out all arguments that the correctly demonstrated Pareto efficiency of an allocative measure could support normative claims of that sort, even if I am skeptical concerning that possibility; but that is a separate matter.

[7] As noted in the Preface, I have in mind what might be more properly called the “Pareto Efficiency Principle”.

[8] It should be noted that Pareto did not believe that those making allocative decisions acted as rational agents; indeed, he was profoundly skeptical concerning the rationality of economic actors. Thus, Pareto‘s Efficiency Principle was not understood as a descriptive principle, giving a description of the behavior of economic actors, but as a rational standard, and thus a normative principle although not thereby a principle of equity or moral propriety.

[9] A “free market” is nowadays widely understood to mean one in which exchanges and transactions are unrestricted or unregulated. As the economist and economic historian,  Michael Hudson, likes to point out, this is not what Adam Smith meant by the term, although Smith is often cited as its originator. By a “free market”, Smith meant a market free of control by rentiers. Without worrying about Smith, even brief reflection suggests that no market can really be unregulated. What a market is is precisely a venue for the sale and exchange of goods and services, activities that proceed according to certain rules. It is a theatre of contractual exchange, and contracting is defined, and bound by, rules. Moreover, the ideologues of what they call the “free market” insist that free market exchanges must be voluntary, by which they mean uncoerced, but they appear to be blind to the fact that that in itself is a regulatory requirement.

[10] Paul Samuelson (1915-2009), “likely the most influential economist of the latter half of the 20th century“ according to The Economist, described allocative efficiency, in a glossary, as “A situation in which no reorganization or trade could raise the utility or satisfaction of one individual without lowering the utility or satisfaction of another individual. Under certain limited conditions, perfect competition leads to allocative efficiency. Also called Pareto efficiency.”

[11]Even if it comes out much better for some than for others.

[12]The attraction of this definition is partly the antecedent, (supposedly) “Humean”, persuasion of many people that all goods and values are the creatures of wants or desires—as discussed below—and partly the attempt by economists to be “scientific”. Although at first blush a “want” or “desire” might seem to be the kind of internal, subjective, emotional state of mind that many economists think of as unworthy of the attention of proper science, the use of “wants” as a metric may be thought acceptable to the extent that “wanting” can be defined operationally, by reference to what people select in the marketplace: to put it crudely, if you bought it, you wanted it.

[13] The matter of “long run” interests is touched upon in a later section of this paper.

[14] As far as Hume is concerned, I have maintained that this is not actually Hume’s view, however often it may be attributed to him. See my “Reason, Passion and the Influencing Motives of the Will” in Saul Traiger, ed. The Blackwell Guide to Hume’s Treatise (Oxford: Blackwell Publishing, 2006), 235-55. In an earlier version of this paper, I claimed that Adam Smith belonged to this same tradition, but, upon reflection stimulated by a peer reviewer, I find that it is unclear whether or not Smith followed this supposedly-Humean doctrine.

[15] Unlike the view of goods commonly attributed to him (see discussion above), this view is pretty certainly Hume’s actual view.

[16] Here, my peer-reviewer objects that I do not use the notion of “worse off” in the way that an economist—or perhaps Pareto himself—would understand it. In the context of Pareto-guided allocation, a person’s being made “better off” or “worse off” by an allocative decision must be understood in terms of his ex ante “endowment”—that is, what he has or lacks prior to the allocative decision’s being made. Prior to the hiring decision, none of the candidates for the job had the job; thus, possession of the job was not a part of any of their “endowments”. The successful candidate is made “better off” by the hiring decision, since having the job was added to her endowment; but, since nothing was subtracted from the unsuccessful candidate’s endowment, he is not made “worse off” by the hiring decision, the reviewer explains. In response to this point, I could withdraw the example, for there are plenty of others available to illustrate my point; indeed, some are given in the sentence that follows. But I think that it is important to say something about the reviewer’s critique of the example, because it is important to the strategy of defending Pareto efficiency as the touchstone to “good” or “proper” allocative decision-making. From Pareto’s perspective—and I believe that this can actually be attributed to him as well as to the community of economists—an allocative decision is an “improvement” in the landscape of allocations if it adds to the endowment of at least one person while not subtracting anything from anyone else’s endowment. The endowments in question are those existing at the time that the allocative decision is implemented. This entirely—and deliberately—ignores the question of how the existing landscape of allocations, that is to say, of “endowments” was reached or whether it is defensible per se in terms of, say, the general welfare. An overall situation that is Pareto-optimal (maximally Pareto-efficient) can be outrageous from the standpoint of social or economic justice and inefficient in terms of the functioning of society (with “efficiency” understood in an ordinary, common-sensical, way. This is masked by the predication of allocative “improvement” or “optimality” upon “endowments” whose credentials are left deliberately unscrutinized. A further point to make with respect to my reviewer’s critique of my mundane (and indeed, familiar) example, is that there is no principled way of describing what is, or is not, in a person’s “endowment”. For instance, prior to the hiring decision, might have considered his endowment to include an opportunity for gaining employment, which was then lost when someone else got the job. Would we not all understand what was meant were he to complain that, “She got the job at my expense”? My reviewer defines the endowments in the example in a way that is at odds with the “satisfied preferences” view of being made “better or worse off”: the successful job candidate had one of her preferences fulfilled and none thwarted by the decision. The unsuccessful candidate had none of his preferences fulfilled, but one thwarted. This points to a certain indeterminacy, and even perhaps to a certain disingenuousness, in the use of the “existing endowment platform” for the assessment of allocative decisions.

[17]I imagine, in fact, that we would have what Kant described as imperfect duties, which are even weaker than indicated here; but I aim here to keep the discussion as simple as possible. Whether there are actually such “Pareto duties”, whether they would be what some theorists call pro tanto, as opposed to prima facie duties, perfect or imperfect, and so on, are issues which it would take a much larger space to address.

[18] This is a paraphrase, but I believe that I do not distort the meaning.

[19] Citation to Kenneth Arrow, “Uncertainty and the welfare economics of medical care,” American Economic Review, 1963; p. 942. Bracketed material added by Reinhardt and by myself.

[20] This ethic is clearly not restricted to applied welfare economics; but it is there that any such ethic is most directly relevant to the allocation of benefits—goods, services, and wealth—i in its real social, political and technocratic context.