GOOD, ECONOMIC WELFARE, AND THE NATIONAL DIVIDEND—PIGOU’S WELFARE TRIAD

Abstract

Arthur Pigou’s welfare treatises are exercises in practical ethics. The exercises were founded on the ethical concepts of good and economic welfare with the economist’s national dividend providing the practical instrument for solving economic problems. This paper follows this triad from its origins in Pigou’s earliest writings on ethics and economic policy, into the welfare treatises, and onto his last writings—a period of around fifty years.

I. INTRODUCTION

Wealth and Welfare (1912) was an expression of Arthur Cecil Pigou’s beliefs that policy choice is a matter of practical ethics and that doing right is choosing the greater good. “Welfare means the same thing as good,” it declares. This good, a thing of very wide range, has a component—economic welfare, or “welfare arising in connection with the earning and spending of the national dividend” (Pigou 1912, p. 3). Part I of the book continues with “important propositions” on how economic welfare is affected by the size, distribution, and variability of the dividend; later parts detail the action of policy along these dimensions.

Good, economic welfare, and dividend formed a system of related concepts—the triad—that was a constant in Pigou’s welfare thought, informing Part I of The Economics of Welfare (1920 and subsequent editions) and the reprise of “fundamental issues” of 1951. Writing on the welfare state (1954), he contended that such an entity “endeavours to promote the economic satisfaction of its citizens” (p. 1) by—in effect—acting in line with the propositions of 1912.

I trace the triad from its origins into the welfare treatises and onto Pigou’s final word to a new generation. Pigou never placed his ideas in the stream of ideas—his own or others’—but recent studies of his early writings have identified the sources of his stream. The following sections II to VII draw on them to relate how the triad was prefigured in an ur-treatise—a memorandum on Poor Law policy—that combined elements from his writings on moral philosophy and commercial policy. Sections VIII to XI follow the triad into the treatises and beyond.

Among others, Alfred Marshall and Henry Sidgwick are familiar but Pigou is seldom put into the stream associated with John Ruskin’s aphorism “There is no wealth but life” although Hla Myint (1948, pp. 201–202) once remarked that English economic thought (Pigou’s included) was “permeated” with the ideas of writers “like Carlyle, Ruskin and William Morris”—critics of political economy. On examination, Pigou’s welfare/good turns out to be very like Ruskin’s life.

Pigou (1877–1959) spent his adult life in King’s College, Cambridge. Admitted in 1896, he studied history and then moral sciences (ethics, political philosophy, and economics). ¹ His prizes show a breadth of interests: for the poem “Alfred the Great” and the essays Robert Browning as a Religious Teacher (1901a) and “The Causes and Effects of Changes in the Relative Values of Agricultural Produce in the United Kingdom during the Last Fifty Years” (see McLure 2013). By 1901 an economics career was under way, and in 1908 he succeeded to Marshall’s chair and held it until he retired in 1943. A tireless writer, Pigou had much to stimulate him: two world wars, the rise of Soviet communism, the Depression; in Britain, controversy over protection, the rise of organized labor, the emergence of the Labour Party, and development of the welfare state; in his profession, the challenge of eugenics, changes in the position of economists, and changes in the way economics was done. “Practice” always mattered for Pigou, but the character of his engagement with it changed over the decades: when very young, he wrote polemically on matters of the day and fray; when older, he avoided “partisan political debate,” warning (1935, p. 10) the beginner against the “intellectual crime” of adjusting his views so that they conform to the policy of a political party. ²

II. “PRACTICE”—ECONOMICS AND ETHICS

Practice, for Pigou (1908b, p. 13), is “the impulse to the economist’s work,” and “social enthusiasm” is the beginning of economic science. He elaborates in his inaugural Economic Science in Relation to Practice (1908b) and in The Economics of Welfare (1920).

For Pigou (1908b, pp. 8–13; 1920, p. 3), research produces “light or fruit—either knowledge for its own sake or knowledge for the sake of good things to which it leads.” Pigou was not unsusceptible to puzzles “of academic rather than of practical interest” (see section IX below) but the chief reason for studying economics is to “help social improvement” (1920, p. ix). ³ “A man” may come to economics for intellectual reasons, but, declares Pigou (1908b, p. 12), “I shall be far more glad if he comes to it because he has walked through the slums of London and is stirred to make some effort to help his fellow.” There was no conflict between being a “human being” and being an economist—being an economist was one way of being human. ⁴

Social improvement rests on a combination of values and science, as Pigou (1907a, p. 359) explained in his “Social Improvement in the Light of Modern Biology”: ⁵

It involves the questions: both what kind of society is good, either absolutely or relatively, and by what means is the desired kind of society most likely to be brought about. The former of these questions is wholly ethical. It turns exclusively upon the determination of values. … Positive science tells us what effects given causes tend to produce; it does not tell us what effects are good.

This conception of the logic of practice was in the tradition of John Stuart Mill, John Elliott Cairnes, and John Neville Keynes. Yet, in terms of Keynes’s (1891) national caricatures—English political economy as “positive, abstract and deductive” and German as “ethical, realistic and inductive” (p. 20)—Pigou belongs in the North Sea. Caricatures never tell the whole truth and Sidgwick’s Principles (1883) was the only English textbook to separate the art from the science rather than mixing the two in the treatment of particular topics.

When choosing policy it is not always necessary to choose between values. In his first substantial piece on social improvement, “Some Aspects of the Problem of Charity,” Pigou (1901b, p. 239) recognized conflicting views of the “all-important thing”—viz., “happiness” (Sidgwick) versus “character” (Marshall?)—but contended, “[I]t is unnecessary for any one engaged in the practical work of charity to decide between these two views, because his course of action would have to be very much the same whichever he adopted.” ⁶ The point reappears in Industrial Peace where Pigou (1905, pp. 3–4) notes that practical questions about arbitration and conciliation are “often answered in the same way by thinkers whose fundamental doctrines are quite irreconcilable with one another.” Later (p. 32) he mentions conflicts and the need to “weigh” considerations—without indicating how.

Robustness to difference in values never became a guiding principle for Pigou, and neither did he stand back from “fundamental controversies” about goodness—he (1908b, p. 14) saw “an urgent need for the economist that he be also a student of Ethics.” In section XI below we discuss another—Ralph Hawtrey—who agreed. ⁷

III. THE “STUDENT OF ETHICS”

Pigou was serious about ethics, contributing to both its divisions (1908a, pp. 110, 117): practical ethics, where “our ultimate goal, of course, is to promote the greatest possible amount of goodness in itself”; and theoretical ethics, where the fundamental problem is “of determining what things are good in themselves.” As an economist he engaged in practical ethics; as a moral philosopher he investigated goodness in itself and reflected on the scope and method of the science.

There was a sense of “good” that Pigou (1908a, p. 117) distinguished from “good” in itself, viz. “useful as a means to promote something thus absolutely good.” He mentioned the distinction when considering the value of knowledge or of human qualities. Thus, of those who have risen from a poor background, he (1907a, p. 365) remarked:

Among the original properties of these relatively rich presumably there are qualities which account for their rise…. How far the qualities based on these original properties are good in themselves seems to me doubtful. But they are certainly a means to good. By adding to wealth they make for happiness, and happiness is an important element in well-being.

“Means to good” in people had its day later—see section X below. Pigou would always doubt the goodness of the economic virtues.

Pigou’s ethical writings are not extensive: passages in Browning (1901a) and four short essays from 1905 to 1908 assembled in The Problem of Theism (1908a)—three published in the International Journal of Ethics. Pigou was an undergraduate when he wrote Browning, but its themes stayed with him as he made a career as an economist. His attention to poets was grounded in beliefs that “philosopher and poet confront the same problem” (1901a, p. 109) and poets have “an insight and a vision, and a hold upon concreteness, which the thinker in his study often lacks” (1901a, p. 139). ⁸

“The problem of good” treats “some points of ethical controversy” while the other essays are “critical expositions of important views”—of Jesus, Nietzsche, Robert Browning, and another poet, George Meredith. Pigou (1908a, pp. 109–110) recognized two kinds of contribution to theoretical ethics: providing “insight” into what is good and the negative task of eliminating those things that “blur and obscure insight.” Pigou saw himself as a clarifier of the original insights of others, and his contribution mainly on the negative side. Though he is the sympathetic commentator who tries to put his moralists’ views into the most coherent form, he left the reader in no doubt as to the insights, perceptions, or intuitions that he judged truest.

Henry Sidgwick (1838–1900) had been the great figure of Cambridge Moral Sciences, contributing to all its branches. In the preface to Theism Pigou (1908a, p. viii) acknowledged an indebtedness for “general philosophical standpoint,” but the debt is clearest in the metaphysical essay “The Nature of Reality” in Theism. In ethics Pigou deviated, following Sidgwick only on the first of the three issues (1908a, p. 80): the method “by which ethical inquiry ought to be pursued,” the qualities “which make up the goodness of any conscious being,” and the way “in which the goodness of one being is related to that of others.”

Pigou (1908a, pp. 80–81) identifies two methods of seeking an answer to the question What is good?—“the a priori method of deduction from the nature of things” and “the method of direct perception.” The method of deduction—associated with the Idealist philosopher Thomas H. Green and the poet Tennyson—was “mischievous and idle.” The text for “direct perception” (“insight” in his Nietzsche essay) is the Mikrokosmus of the German philosopher Hermann Lotze: “What is good and evil remains just as incapable of being reached by mere thought as what is blue or sweet.” ⁹ Pigou elaborates:

The only way to know whether anything is good is by looking at it. … We turn the eye of the soul upon it, and we perceive some things to be good and some bad, just as we perceive that some are yellow and others red. … The truer the eye of the percipient—not necessarily the intenser his mental power—the better the results will be. (1908a, p. 82)

Pigou concurs with Sidgwick and George Edward Moore on the method of intuition, though he mentions neither in this context. Incidentally for Pigou at this time, Moore (1873–1958) was a fellow laborer, not the master-guide he was for Keynes or Gerald Shove.

Considering the qualities that are good, Pigou (1908a, p. 83) holds that the only good things are “states of conscious life.” He would repeat this formula—see sections VIII and IX below—though what it signified changed over time: he began (1908a, p. 83) by protesting against Moore’s doctrine of “organic goods,” which holds that goodness applies to complexes of states of consciousness and objects, but ended (1954, p. 2) by saying that the “welfare” of an object like Stonehenge is without meaning.

What makes a conscious state good? Pigou (1908a, p. 84) lists some views:

The Utilitarians declare that the only element upon which the goodness of a conscious state depends is the quantity of pleasant feeling that it contains. Dr. Martineau, finding in the human consciousness a hierarchy of “springs of action,” declares that the goodness of a man at any time depends solely on whether or not he wills in accordance with that one of two conflicting springs which he judges to be higher…. Finally, yet another school believes that the only element upon which goodness depends is the emotion of love.

Green sits with James Martineau, while the final “school” includes Jesus and Browning.

These are not just possible views, for Pigou (1908a, p. 86) continues:

Nor are these three [pleasure, the good will, and love] the only variables upon which goodness depends. I would include also the character of a man’s ideals, his attitude towards beautiful persons and things, and, so far as it is not already embraced in love and the good will, his enthusiasm for the purpose he sets before himself.

This passage is Pigou’s fullest account of what contributes to goodness. In giving the six “variables” or dimensions independent standing, Pigou separated himself from those like utilitarians who gave primacy to one.

Of course utilitarianism was more than a view about goodness. According to Sidgwick’s Methods of Ethics (1884, p. 407), it states “that the conduct which, under any given circumstances, is objectively right is that which will produce the greatest possible happiness on the whole; that is taking into account all whose happiness is affected by the conduct.” W. David Ross (1930, p. 17) used the (less than ideal) term “ideal utilitarianism” for the position that “what produces the greatest good is right” without the requirement that the only good is pleasure—that position Ross called “hedonistic utilitarianism.” Ross had Moore in mind but Pigou was an ideal utilitarian too—sometimes in six dimensions, sometimes in three (see section VII below) but never in the one of utilitarianism. ¹⁰

Having contended that the goodness of a conscious state is “a function of several variables,” Pigou (1908a, p. 87) asks whether anything further can be known as to “the nature of this function.” Acknowledging that this is difficult territory, he spends some pages outlining possible positions before concluding (p. 89):

I think that it may in some circumstances be said that the greater the quantity present of one element A, the greater other things equal, is the addition made to the goodness of the whole by any given addition to a given quantity of another element B. I think that this relation holds between the elements happiness and virtue. But I doubt whether any other general proposition can be laid down.

Though hedged by two doses of “I think,” these are remarkably refined intuitions on the partial derivatives of the goodness function.

The third and final point of ethical controversy Pigou (1908a, p. 90) considers is “whether the good of [person] A can compete with that of B or C.” Pigou’s discussion takes in the views of Green, Sidgwick, and Moore, and the main point he wants to establish (p. 92) is the arcane double negative, that “to admit that the goods of different people may compete does not involve self-contradiction.” ¹¹

Pigou’s ethical essays are extremely schematic, more like the analytical tables of contents of the treatises than the text. Of course he knew more and a hinterland was exposed when he (1901c, p. 76) criticized Westel Willoughby’s Social Justice for overlooking Sidgwick’s utilitarianism: “The Ethics and Elements of Politics are not open to any of the strictures which he passes upon Bentham and Mill.” Pigou (pp. 75–76) also noted that many of the conclusions Willoughby had reached from Green’s ethical system had been reached by Sidgwick from his brand of utilitarianism. ¹²

The ethical essays perished—Pigou never cited them and they never entered the philosophical mainstream, despite appearing in a respectable journal, being embedded in the literature, and having input from Bertrand Russell and Keynes. His ethical views were formed early and nothing suggests he altered them in any fundamental way. Good was finished—how did it tie up with the economic theory Pigou was teaching and the policy stances he was taking?

IV. LOOKING INTO MARSHALL’S PRINCIPLES

From 1901 Pigou was teaching Marshall’s Principles and he first discussed the application of its theory to ethics in 1903. ¹³ “Some Remarks on Utility” (1903) examines an obstacle: the ethical interest is in the pleasure derived but the demand theory shows how price measures the intensity of desire. After reviewing arguments from Marshall and Sidgwick, Pigou (1903, p. 68) concludes that the difference is of no practical significance for the most important cases.

Another device from the Principles was sent to the policy front in a 1904 article, “The Known and the Unknown in Mr Chamberlain’s Policy.” The title suggests political ding-dong but the device is the Bernoulli function, y = Klog(x/A) giving the satisfaction, y, a person derives from income, x, where A is the income sufficient to purchase the necessaries of life. Pigou’s interest was in the change of total satisfaction associated with the redistribution of income following the imposition of a tariff on agricultural produce (Chamberlain’s policy). Using guesstimates of the income of landowners and consumers and of the value of A, Pigou (1904, pp. 47–48) infers that “to subtract £1 from the average man and give it to the average landlord is to confer upon the latter an amount of satisfaction equal to one-fifth of the satisfaction lost to the average man.” ¹⁴

The calculation suggests a way economic welfare might have been developed, but it also brings out the visible and invisible in Pigou’s thought: the functional form versus the possibility of interpersonal comparisons of amounts of satisfaction. The latter came to light only in the 1930s when it was challenged—see section XI below.

These ideas from 1903–04 next surfaced in the welfare treatises: “Remarks” was alluded to (1912, pp. 8–9) and cited (1920, p. 24), but “Mr Chamberlain” returned only as a ghost. Another Principles device saw daylight sooner—the national dividend or “net aggregate of commodities, material and immaterial, available for consumption.” Glimpsed in Industrial Peace (1905, pp. 21, 63, 67), it soars in Pigou’s 1906 writings on protection. ¹⁵

V. THE NATIONAL DIVIDEND AND NATIONAL WELFARE

Ancestors of the propositions relating economic welfare to the national dividend appear in the article “Protection and the Working Classes” (1906c) as stages 1–2 of the “correct method of estimating the effect of Protection upon Labour” (p. 12):

[1] an inquiry into the effect of the policy upon the National Dividend, as a whole. For, prima facie, anything that enlarges that dividend is likely to be advantageous, and anything that diminishes it disadvantageous to all the agents of production in the country.

[2] an inquiry as to the effects of protection upon the distribution of the dividend among the various agents. For this may be altered in such a way that, despite the increase in the whole dividend, the share that goes to certain agents may be, not merely, proportionately, but absolutely less than it was before.

For years the prima facieness of the likelihood in [1] required no elaboration.

The final stage involved the manner in which labor receives its share:

This may be altered in such a way as to react on character and morale. If, for instance, Protection would lessen either the irregularities of employment, or the proportion of people engaged in sweated industries, the consequent improvement in the men might be well worth purchasing even at a cost of some reduction in their earnings. (p. 13)

As Pigou judged that protection would not have such effects, there was no need to locate “improvement in the men” in his six-dimensional ideal utilitarianism or to weigh such improvement against loss of earnings and of national dividend.

The article became a chapter of the book Protective and Preferential Import Duties (1906a) with a new word and title anticipating future versions of Part I—“The National Dividend and the National Welfare.” No longer limited to labor, the discussion (1906a, p. 36) moves from “advantageousness” and “disadvantageousness” to “the country as a whole” to the “welfare” of the whole. “Welfare” is not defined but appears rooted in the ordinary dictionary sense of “the state or condition of doing or being well.” The phrase “welfare of the community” had appeared in Industrial Peace, but it had no special weight either for Pigou or for other economists of the time.

The propositions flow (1906a, p. 36): “the welfare of the whole is not dependent merely upon the wealth of the whole” where “wealth of the whole” apparently refers to the size of the national dividend; among the other circumstances, “perhaps the most important is the way in which the national dividend is distributed.” Another circumstance, separate from the dividend, sits in a footnote (1906a, p. 36n1): “[The welfare of the whole] is dependent, for instance, inter alia on the desirableness of desired satisfactions, and of the desires which these satisfactions stimulate. The welfare of China might, for instance, be promoted by a subtraction from its national dividend of all the opium it now consumes.”

Desired satisfactions register in the national dividend but the “desirableness of desired satisfactions” concerns their goodness. A similar six-dimensional consideration entered when Pigou (1906b, p. 379) dissected the “fallacy” that “all men, if left free, will best advance the interests of all.” He (1906b, p. 379) maintained that the notion of “interest” is ambiguous: “It may be true that an individual is the best judge of his interest, when by interest is meant what he as a matter of fact does want, but it is not true that he is the best judge of what he ought to want.” Satisfying the desires people ought to feel had ethical value.

VI. ALL HAIL THE DIVIDEND, ALL HAIL MARSHALL

Stages [1] and [2] of the “correct method” echo elements of Sidgwick’s (1883, p. 403) art of political economy:

(1) The Art of making the proportion of output to population a maximum, taking generally as a measure the ordinary standard of exchange value … (2) the Art of rightly Distributing produce among members of the community, whether on any principle of Equity or Justice, or on the economic principle of making the whole produce as useful as possible.

Pigou made alterations—replacing “output and “produce” by the “national dividend,” dropping “equity” from (2) and making “welfare” part of the art’s overarching objective. Yet it is still surprising that Sidgwick was never noticed for the propositions. Nobody was—though Marshall got credit for the dividend and for appreciating its significance. ¹⁶

Reviewing the fifth edition of the Principles (the last to be called volume I), Pigou (1907b, p. 533) screamed: “The conception of the National Dividend is not an academic toy, but a practical instrument of great power designed for service in the concrete solution of social problems. The unavoidable but regrettable delay in the appearance of Prof. Marshall’s second volume has obscured this fact” (Pigou’s italics). After reporting Marshall’s applications of the dividend—to whether trade union policy is anti-social, to the attitude which the State should adopt towards the poor, and to innovation under collectivism—Pigou (p. 535) explodes:

The dividend constitutes the kernel of economic theory because … it is the centre of sound philanthropic endeavour. It is to an analysis of this that we are driven, when, throwing off the moral torpor of indolent optimism, we refuse, “with our modern resources and knowledge, to look contentedly at the continued destruction of all that is worth having in multitudes of human lives,” and demand from social science guidance to social reform.

The phrase quoted is from the Principles (p. 721 in fifth edition; see also Pigou 1920, p. 722) but it is not linked to the dividend. Marshall favored caution in social reform: “There is evil in extreme impatience, as well as in extreme patience, with social ills” warned the marginal summary.

Pigou (1907b, p. 532) wrote as “one who has been taught by Prof. Marshall,” suggesting he knew Marshall’s plans. But there was no plan for a second volume organized around the national dividend; see John Whitaker (1990). Yet Pigou’s second volume belief served to underwrite his current work and, looking ahead, a treatise that would give the needed “guidance to social reform” and be a true second volume. ¹⁷ But first there was an ur-treatise in which ethics spoke to economics.

VII. THE UR-TREATISE AND A CALCULUS OF PRACTICAL ETHICS

In 1907 Pigou prepared A Memorandum on Some Economic Aspects and Effects of Poor Law Relief for the Royal Commission on the Poor Laws; it was published in 1910. ¹⁸ This policy piece was different: ethical considerations were promoted from the realm of footnotes and after-thoughts; economic well-being was identified and its separate study defended. The details would change but the triad was on stage.

Pigou writes not from the viewpoint of economics but of ethical science, implicitly showing the need for the ethicist to be also a student of economics. His first chapter, “The Poor Law and the Elements of National Well-being,” unpacks the proposition that any policy is “likely to have considerable effect upon the well-being of the community.” Pigou (1910, p. 981) lists the elements on which “national well-being” depends:

1. (1) the people themselves as ethical personalities;

2. (2) the direct social and other relations of people with one another, and the satisfactions that result therefrom; and

3. (3) the satisfaction obtained by the people from their economic circumstances.

Elements 1–3 make a drastic condensation of the six dimensions of “The Problem of Good”: “satisfaction”—in (3) and the second part of (2)—corresponds to the utilitarian “pleasant feeling” while (1) reflects the other five variables. The goodness of relations— the first part of (2)—seems to be a new consideration.

Pigou (1910, p. 982) sketches the effects of Poor Law policy under the three heads. The effect on “the value of the people as ethical personalities” is chiefly by affecting “the qualities commonly known as ‘economic virtues’”—presumably such qualities as self-control, thrift, and calculation. Of them, Pigou writes:

Whether [the policy] enhances or diminishes the value of the people depends on the value which various degrees of the economic virtues possess as ends in themselves, apart from their effects. Presumably, some modicum of these qualities is good, but an excess of them is likely to be bad. Up to a point an addition to economic virtue is an addition to human virtue; after a point, it is an addition to economic vice.

When it came to the “character of a man’s ideals,” Pigou was always skeptical of the qualities making for economic success.

Pigou (1910, p. 982) used old-age pensions to illustrate the second class of considerations: “some urge that they would weaken the sense of family responsibility, and would thus diminish the sum of good social relations. Others reply that the obligation to burdensome monetary support makes rather for severance than for sympathy.” Pigou found it unnecessary to offer examples of considerations under the third heading.

Pigou contemplates a calculus of practical ethics, a scheme of policy evaluation based on the three elements of well-being:

A quantitative estimate of the amount of each of them would need to be made, and the aggregate effect obtained by combining them arithmetically with due regard for signs. If this were done, the merits of all practicable lines of policy could be compared by reference to the largeness of their aggregate positive, or the smallness of their aggregate negative, effects upon well-being. (1910, p. 982)

If Pigou had known the term “social welfare function,” this is what his would have looked like. The procedure accords with the goal of promoting the “greatest possible amount of goodness in itself” in Theism.

Pigou combined enthusiasm for quantification with skepticism about what could be quantified—distancing him from such modern artifacts as the UN Development Index that incorporate some of his values. ¹⁹ He judged “there are no means of measuring the effects that fall under the first two heads” but effects under (3) were different (1910, p. 982): “The effects of different policies upon the satisfactions obtained by people from their economic circumstances … can be roughly compared.” This part of total well-being is “economic well-being.”

“Economic well-being,” like its successor “economic welfare,” is an unsatisfactory phrase, for it doesn’t mean all the well-being associated with people’s economic circumstances, only the part associated with satisfaction: “economic satisfaction” is a better fit. Such well-being is brought into relation with the national dividend in Chapter IV, “Criterion for Determining the Advantages of Various Poor Law Policies.” The criterion is the “correct method” of the protection article (see section V above) minus the third stage, which involves non-economic well-being. Pigou (1910, p. 987) indicates how to apply the criterion: “In estimating the aggregate economic influence of any Poor Law system we have therefore to examine its effects (a) upon the size of the national dividend, and (b) upon its distribution.” Economic well-being—the “satisfaction obtained by the people from their economic circumstances”—is presumably the sum of satisfactions that individuals receive from the share they receive of the goods and services annually available for consumption. Later commentators—e.g., Abram Bergson (1938, p. 324)—would say so but Pigou never did.

Pigou considered that effects under one head only could be compared and conceded that the result of such a partial comparison “need not be the one which would emerge if a complete comparison were practicable” (p. 982). However, as unknown facts are as likely as not to confirm as to conflict with known facts, he thought it reasonable to concentrate on (3):

Hence, if one policy is certainly superior to another in economic effect, it is probably superior to it on the whole; and, if one policy is probably superior to another in economic effect, it is probably, in a lower measure of probability, superior to it on the whole. For this reason, comparisons of the economic effect of different policies, though incomplete and one-sided, may still be valuable. (1910, p. 982)

This probabilistic justification of a separate study of economic welfare went into the welfare treatises.

So there was much more to the Memorandum than Poor Law policy: it applied Pigou’s six-dimensional conception of the good to social improvement and investigated a comprehensive policy science before concluding that only one element of the good, economic well-being, could be treated quantitatively—it was not supremely valuable, just uniquely measurable. These lessons were incorporated in a volume of 500 pages that added foundations and covered the whole of economic policy. A moral not drawn was that attention be confined to the dyad of economic well-being and dividend.

VIII. THE FIRST TREATISE: WEALTH AND WELFARE

The volume was not called The National Dividend and National Well-Being but Wealth and Welfare (1912), the euphonious coupling from 1906, lately used to effect when Pigou (1909, p. 10) urged that “the development of [the quality of patriotism and concern for friends], which the Co-partnership principle promotes, besides leading to increased wealth, is increased welfare.” ²⁰

Part I, “Welfare and the National Dividend,” organizes the old themes into a prolegomenon to an economics of welfare and the economics proper: the prolegomenon (my term) places economic welfare in the general study of welfare, and the economics goes from foundations to propositions on how the size, distribution, and variability of the dividend affect economic welfare, which in turn points to the applications in the rest of the book: the perennials of Pigouvian economics—externalities, social cost, etc.—appear in Part II on the size of the dividend. ²¹

Pigou (1912, p. 3) begins the prolegomenon by quoting Moore (1903: 6): “If I am asked ‘What is good?’ my answer is that good is good, and that is the end of the matter. Or, if I am asked ‘How is good to be defined?’ my answer is that it cannot be defined.” The identification that follows—“Welfare means the same thing as good”—is less arresting when the transitions—being good to well-being to welfare—are recalled.

Regarding (total) welfare, the Memorandum’s “well-being as a whole,” Pigou (1912, p. 3) confined himself to two “propositions” transposed from his thought on the good: “that welfare includes states of consciousness only, and not material things or conditions: secondly, that welfare can be brought under the category of greater and less.” Only the second was used in what followed—in the probability argument.

The probability argument (1912, p. 11) from the Memorandum justifies the separate study of a component of welfare:

When we have ascertained the effect of any cause on economic welfare, we may, unless, of course, we have evidence to the contrary, regard this effect as probably equivalent in direction, though not in magnitude to the effect on total welfare.… The burden of proof lies upon those who hold that, in any particular case, this presumption should be overruled.

This time Pigou found an authority, Francis Edgeworth on “unverified probabilities.” ²² The argument is repeated in The Economics of Welfare.

As this argument for a partial study was purely formal and did not depend on the nature of non-economic welfare, further discussion of non-economic welfare might seem superfluous for an economics of welfare. However, the prolegomenon was as much a celebration of non-economic welfare as a justification of the economist’s not treating it—so the triad survived.

In the Memorandum Pigou had listed the elements of well-being he now (1912, p. 5ff) catalogs, and illustrates the ways economic causes have “direct” and “indirect” effects on non-economic welfare; among the direct effects, which may be good or evil, Pigou (pp. 5–7) mentions and illustrates extensively effects on desires, on character, on occupation, and on social relations at work—effects descending from elements (1) and (2) of the Memorandum. Pigou (1912, p. 7) explains indirect effects as follows: “Causes that modify economic welfare may influence other parts of welfare, not directly, but indirectly through objective conditions of welfare other than the national dividend.”

Turning to the economics of welfare proper, the main aim is to establish the propositions relating economic welfare to the dividend (see section V) using the theoretical tools described in section IV. Pigou’s efforts were examined by later economists, and having his thought pre-digested works for and against us. Accounts like that in Bergson’s (1938) “Reformulation of Certain Aspects of Welfare Economics” clarify Pigou’s argument by making the assumptions explicit but abandon the realistic idiom of the original. Pigou wanted to tackle the “real economic world” in its “concrete actuality” and the result is a web of complications and qualifications that obscures the basic structure. ²³ Thus, when Bergson (1938, p. 324) characterized Pigou’s position symbolically, he was reading between the lines—and ignoring most of them. Pigou seemed satisfied with the result and when he restated the two propositions in 1951, it was in the new idiom; see section XI below.

Economic welfare, writes Pigou (1912, p. 3), is “welfare arising in connection with the earning and spending of the national dividend or, in other words, of those parts of the community’s net income that enter easily into relation with the measuring rod of money.” The identity suggested by “in other words” was not quite that, for the measuring instrument had limitations with Pigou (p. 3) admitting that “various good and bad qualities indirectly associated with income-getting and income-spending are excluded from it.” Economic welfare, he (p. 4) acknowledged, was only “part of a part of welfare.” Of the measuring rod itself, Pigou (p. 8) maintained that it was what made an economics of welfare possible: “The methodological principle at the basis of economic science … is the reference which it makes to a measure, namely, money.” Naturally he (pp. 8–9) remembered the old point from 1903—see section IV—that the measure was of desires and aversions instead of the satisfactions and dissatisfactions of ultimate concern.

The national dividend dominates the rest of Part I and the attention given to its conceptualization and measurement is the book’s biggest novelty. Chapter II, “Economic Welfare and the National Dividend” (1912, p. ix) opens with the claim, “For the most part economic causes act upon economic welfare, not directly, but through the national dividend,” and the statement, a cardinal point in Pigou’s welfare economics, appears in all editions of the later treatise (see 1920, p. x). It seems to be true by definition and so the need for the qualification “for the most part” is unclear.

Chapter II continues with three “important propositions” relating economic welfare to the dividend, two descending from the writings on protection. The first proposition (1912, p. 20) is: “if a cause is introduced which makes for an increase in the aggregate size of the dividend, provided that the absolute share of no group of members, in terms of the commodities which that group is accustomed chiefly to consume, decreases, the economic welfare of the community as a whole is likely to be augmented.” Apparently Pigou saw no need to demonstrate a link between dividend and satisfaction, and digresses to explain how increases in dividend, or production, come about. He (pp. 21–23) notes an additional reinforcing link, an “infant consumer” channel by which tastes are affected by consumption, and an objection (pp. 23–24) to the proposition based on the consideration that a man’s satisfaction “is, in great part derived not from the absolute, but from the comparative, magnitude of his income.” ²⁴ Pigou admitted the objection but contended, “It cannot well be maintained seriously that an increase in [a man’s absolute income] will add nothing whatever to the satisfactions which constitute his economic welfare.”

Pigou comes closest to a Bergson-like symbolic formulation when he (p. 24) justified the second important proposition, viz.,

If a cause is introduced which makes for an increase in the absolute share of relatively poor groups of persons (in terms of the commodities which these groups are accustomed chiefly to consume), provided that the magnitude of the aggregate national dividend (in terms of commodities in general) does not decrease, economic welfare is likely to be augmented.

In the justification—a “Mr Chamberlain” without Daniel Bernoulli—aggregate satisfaction is made explicit: in a community of two members with similar temperament, “it is easily shown that any transference from the richer to the poorer of the two … must increase the aggregate sum of satisfaction” (1912, p. 24). A footnote (1912, p. 25n1) uses an additive “aggregate satisfaction” function based on the assumption of similarity of temperament to show that a diminution in the inequality of distribution “probably increases satisfaction.” Pigou (p. 28ff) considered two objections: that increased wages would be spent on “worthless forms of exciting pleasure” and would stimulate population growth. He admitted that in the short run higher wages might not lead to an increased welfare “in the widest sense” and that population growth may be encouraged but suggested that both effects would only be short run.

Writing about protection, Pigou had contemplated the effect of irregularities of employment on “morale”—an element of non-economic welfare. Irregularity of the dividend was the business of business cycle analysis and there’s a new proposition (1912, p. 25): “if a cause is introduced which diminishes the variability, or inequality in time, of the dividend, and especially of that part of it which accrues to the poorer classes, the economic welfare of the community as a whole is likely to be augmented.” The argument for equalization across moments (pp. 401–402) parallels that for people: based on applying “the law of diminishing utility,” it goes from economic welfare in the individual as the sum of the momentary contributions to the “joint economic welfare” of different people.

Chapter III, “The Measurement of the Dividend and Its Parts” begins (p. 33) by conceding that it has been “tacitly assumed” that the concept of a change is “definite and unambiguous,” which it would be were the dividend a “large parcel of one single thing.” ²⁵ The task (p. 33) then is “to find a measure for changes in a heterogeneous dividend such that the propositions laid down in the preceding chapter in the case of homogeneous dividend remain true,” though both here and in The Economics of Welfare it was the truth of the first proposition that mattered. The discussion (pp. 32–51), which Hans Staehle (1935, p. 165) considered “the first statement of the economic theory of price index numbers,” concentrates on judging from data on prices and quantities consumed whether the satisfaction of the “representative man” has increased. There was no appeal, tacit or otherwise, to cardinal and interpersonally comparable utility as there had been in the defence of the second and third propositions.

The final chapter of Part I is “The National Dividend and the Quality of the People.” The eugenics vogue was at its height in Britain, and Pigou considered whether eugenic complications upset the propositions on economic welfare, concluding that they did not; the chapter and conclusion reappeared in The Economics of Welfare. ²⁶

IX. THE SECOND TREATISE: THE ECONOMICS OF WELFARE

The Economics of Welfare (1920) originated as “a rewritten and revised edition” of Wealth and Welfare, but it grew into an encyclopedia of economic practice. New parts on the dividend and labor and on the dividend and government finance obscured the fundamental triad. In the second edition (1924), the material on variability and public finance was dispatched to separate books but dividend and labor remained, filling more than a quarter of the contracted—though still large—work. Further editions appeared in 1929 and 1932, and there was a reprint with additional material in 1952. The tightness of the original was lost.

The significance of the new title is unclear except that it does not mean that attention is confined to economic welfare. Part I retained its function, though, apart from the chapter on the quality of the people, it was thoroughly rewritten with new elaborations, digressions, and references. The Great War left its mark in reflections on Germany—see section X below—and in a new section on defence (1920, p. 18): “Lack of security against successful hostile attack may involve ‘dissatisfactions’ of a very terrible kind. These things lie outside the economic sphere, but the risk of them may easily be affected by economic policy.”

Part I still divides into a prolegomenon and the economics proper. The prolegomenon takes up Chapter 1, still called “Welfare and Economic Welfare” and still largely about non-economic welfare and whether a separate study of economic welfare is worthwhile. The prolegomenon begins with a discourse on economic science—see section II above. The “methods of science” require measurability, and economic welfare is introduced (1920, p. 11) as the part of welfare on which those methods can operate: “The one obvious instrument of measurement available in social life is money. Hence, the range of our inquiry becomes restricted to that part of social welfare that can be brought directly or indirectly into relation with the measuring-rod of money.”

General welfare slips in without ethical fanfare and it is understood that welfare should be promoted: the object of the book (1920, p. 10) is “to make more easy practical measures to promote welfare” where welfare is “a thing of very wide range.” The two propositions from Wealth and Welfare—about “states of consciousness” and “greater or less”—are stated (p. 10) “more or less dogmatically.” Pigou (p. 14) continues to maintain that the only aspects of conscious life that can be brought into relation with a money measure are a certain limited group of satisfactions and dissatisfactions—"But conscious life … includes … other satisfactions and dissatisfactions, and, along with them, cognitions, emotions and desires. Environmental causes operating to change economic satisfactions may, therefore, either in the same act or as a consequence of it, alter some of these other elements.”

Some pages (pp. 14–18) detail how non-economic welfare is liable to be modified by “the manner in which income is earned” and by the manner in “which income is spent”—different occupations have different “ethical value” and “one [act of consumption] may exercise a debasing, and another an elevating, influence.”

The prolegomenon—and Chapter I—concludes by asking whether economic welfare is worth studying when the true objective is total welfare and whether it can be studied by economic science working alone. For the first, Pigou (1920, p. 20) restates the probability argument and, for the second, gives (pp. 20–22) a new argument deriving from Mill’s Logic to support the claim that economic science, despite its “partial and limited character,” is competent to obtain “reasonably adequate conclusions” about effects on economic welfare.

The economics of welfare proper is enlarged and subdivided. The new Chapter II expands the discussion of the link between desires and satisfactions—a problem raised in 1903. The old Chapter II, “Economic Welfare and the National Dividend,” divides into an introduction to “The National Dividend” (Chapter III) and an account of “The Relation of Economic Welfare to the National Dividend” (Chapter IV). More straightforwardly the old Chapter III, “The Measurement of the Dividend and Its Parts,” becomes Chapter V, “The Measurement of Changes in the Magnitude of the National Dividend and Its Parts.”

Chapter II (1920, pp. 23–29; fourth ed. with appendices 1952, pp. 24–30; hereinafter when an edition number is given it refers to an edition of the 1920 volume) has a new topic, a defective “telescopic faculty” whereby “generally speaking, everybody prefers present pleasures or satisfactions of given magnitude to future pleasures or satisfactions of equal magnitude, even when the latter are perfectly certain to occur.” The observation leads to a discussion of (under)investment and a governmental role in conserving the environment.

The importance of the propositions is underlined by devoting a chapter—“The Relation of Economic Welfare to the National Dividend”—to them. The tenor was unchanged but the formulations were tightened to accommodate possible difficulties; the statement of the four “main propositions” (variability now involves two) in the analytical table of contents (1920, p. x) indicates the nature of the difficulties Pigou contemplated:

[1] Any cause which, without the exercise of compulsion or pressure upon people to make them work more than their wishes and interests dictate, increases productive efficiency and, therewith, the average volume of the national dividend, provided that it neither injures the distribution nor augments the variability of the country’s consumable income, will, in general, increase economic welfare.

…

[2] Any cause which increases the proportion of the national dividend received by poor persons, provided that it does not lead to a contraction of the dividend and does not injuriously affect its variability, will, in general, increase economic welfare.

…

[3] Any cause which diminishes the variability of the national dividend, provided that it neither diminishes its volume nor injures its distribution, will, in general, increase economic welfare.

…

[4] Any cause which diminishes the variability of the part of the national dividend accruing to the poor, even though it increases in corresponding measure the variability of the part accruing to the rich, will, other things being equal, increase economic welfare.

The last proposition is included more for “analytical completeness” than for “practical importance.”

The prolegomenon changed little in later editions of The Economics of Welfare (1920)—unlike the rest of Part I. In the second edition the variability propositions departed to the new Industrial Fluctuations (1927, p. 217), covered there in a sentence. The pivotal chapter on the relation of economic welfare to the national dividend became three: one on reactions through population and one for each of the remaining propositions. Surprisingly the propositions became less visible: the first survived (unbilled) in the analytical table of contents (second ed. 1924, p. x) but the discussion in the text (second ed., 1924, pp. 72–75; fourth ed. with appendices 1952, pp. 82–86) was much reduced. The second main proposition also received less emphasis: in the contents (second ed., 1924, p. xi) it appeared as, “Except in very special circumstances such transferences [in favour of the poor] must increase economic welfare.” However, the preface to the third edition (1929, p. v) restored some of their glory when it reinstated the statement from the first edition summarizing the argument of Part I, viz., “that the economic welfare of a community of given size is likely to be greater (1) the larger is the volume of the national dividend, and (2) the larger is the absolute share of that dividend that accrues to the poor.”

Changes to the treatment of the national dividend had more direction. Comparing versions of Part I across treatises and editions of the second treatise, the dynamic was in the increasingly close attention paid to the dividend. In the preface to the second edition, Pigou (1924, p. v) mentioned puzzles associated with the concept: “[They] are, no doubt, of academic rather than of practical interest. But it is none the less important to resolve them if we can, and the difficulty of doing so is great.” These words would make a fitting epigraph for what was, from 1929, a self-contained monograph of four chapters and fifty pages—what Paul Samuelson (1950, p. 21) called a “classic discussion” making “substantial contributions to the modern theory of economic index numbers.”

X. MAKING GOOD MEN

In the 1920s Pigou revisited some early themes about the welfare that wasn’t economic welfare. He had always distinguished good as a means from good in itself and had applied the distinction to people—see section III. The Economics of Welfare (1920, pp. 12–14) noted that people have qualities that are good in themselves and instrumental qualities, contrasting the modern Germany of engineers and businessmen with the old Germany of poets and philosophers, the materialist West with the spiritual East. The moral of the contrasts was that “efforts devoted to the production of people who are good instruments may involve a failure to produce people who are good men.” The larger point was that “an economic cause may affect non-economic welfare in ways that cancel its effect on economic welfare” (p. 12).

In a lecture on eugenics Pigou applied the same distinction to the “production” of people by biological rather than cultural means. He starts by asking what is meant by “improvement,” by a “good society.” He (1923, p. 305) interpreted the question as: What do we mean by good in society? “A society is a group of persons. If then we are to settle what we mean by good in society, we must decide first what we mean by good in a single man.” Pigou had faced the issue of goodness in a single man long before—abstractly in “The Problem of Good” and concretely in the Memorandum.

Of society he now wrote:

What we aim at is a society that is in the highest possible degree good in itself; containing persons whose qualities are good in themselves; who are happy—for happiness is clearly a good; whose mutual relations are intimate and friendly—for sympathy is clearly among the greater goods: the sort of society perhaps that Morris has conceived in his dream of John Ball. (1923, p. 305)

I don’t think Pigou ever quoted John Ruskin but the famous passage from Unto this Last (1907, p. 185) translates easily into Pigouese when the first sentence is read as “The most valuable thing is well-being.”

There is no wealth but life. Life, including all its powers of love, of joy, and of admiration. That country is the richest which nourishes the greatest number of noble and happy human beings; that man is richest who, having perfected the functions of his own life to the utmost, has also the widest helpful influence, both personal, and by means of his possessions, over the lives of others.

Pigou (1923, p. 306) reminded his eugenist audience that “the artists and poets of Athens could not have adorned the world if there were not available somewhere the qualities that are necessary to provide the means of subsistence and defence from danger.” From such considerations came a formula (p. 306): “To secure the greatest sum of ultimate good, we need a balance: alongside of the qualities that directly contribute to that sum, we need also those that indirectly as means contribute to it and make it possible.” Applying this formula to eugenics and, writing as an economist, he asked:

Is there reason to expect that children born in the lower economic strata of society will, when account is taken both of goodness in itself and of capacity to fill an essential place in the economic organism, possess inherent qualities (1) less good in themselves and (2) less efficient as means to the good of the whole, than children born in the higher strata? (1923, p. 307)

Pigou attended only to the efficiency aspect of the question, concluding (p. 308) with some diffidence that “the true welfare of society is likely to suffer … if the proportion of children born among the lower social strata exceeds substantially the proportion born among the higher.”

Pigou returned to the cultural production of people when he reviewed Beatrice Webb’s and Sidney Webb’s Soviet Communism: A New Civilisation? Responding to their notion of the “remaking of men” Pigou (1936, pp. 93–94) said this in praise of the Soviet Union: “to have enshrined in the policy of a great country the doctrine that it is life, not machinery that matters, which matters in the end, that the supreme commodity is man itself, and that the approaches to civilisation should be free to all and not the privilege of a few, is to have made a unique contribution to history.” In the conclusion to his own Socialism versus Capitalism, Pigou (1937, p. 138) wrote that he “would take a leaf from the book of Soviet Russia and remember that the most important investment of all is investment in the health, intelligence and character of the people.”

The reference to William Morris’s Dream of John Ball in the eugenics lecture is the only explicit reference I have found in Pigou’s writings to any author in Myint’s “broader tradition” (see section I). ²⁷ Yet his distaste for commercial values and admiration for the ideals of co-partnership and guild socialism seem to put him in that tradition. His sense of professional duty led him to add concern with the means of realizing its ideals: cf. his (1920, p. 17) remark on guild socialism: “The fact that schemes of industrial reorganisation on these lines are exposed to serious practical difficulties, which their authors do not as yet seem fully to have faced, does not render any less admirable the spirit of this ideal.” Marshall was more likely to refer to authors in the tradition, but Donald Winch (2009, p. 256) judges his compliments “double-edged” and designed to undermine the position of those authors. Pigou was more straightforward and had only one (softer) edge.

Pigou was exercised by the question: Can economics with its narrowness do good? He (1920, p. vii) asked on behalf of himself and the student contemplating ways of improving the world, telling the latter that the search for the knowledge by which social evils may be restrained “is the task, to find it perhaps the prize, which the ‘dismal science of Political Economy’ offers to those who face its discipline.” Thomas Carlyle, it is implied, had scorned a powerful but demanding method for doing good: combining two tags recently resurrected—as “the first serious optimist” his aim was to create a “sanguine science.” ²⁸

XI. EXPLAINING AND DEFENDING THE TRIAD

For decades Pigou had little explaining and defending to do, for Part I wasn’t much discussed. Presumably it was seen to be beside the point. The sympathetic Allyn Young (1913, p. 673) noted points of disagreement with the prolegomenon but did not develop them because Pigou “hedges his doctrines about with so many safeguards that the really vulnerable points are, at most, few; and, secondly, because these debatable points play but a small role in the general argument and conclusions of the book.” The unsympathetic Edwin Cannan (1921, p. 207) asked of the propositions, “Must we read 108 pages to make sure that we are right in believing” that “a big, well-distributed and steady income is better than a small, ill-distributed and violently fluctuating income, especially if the fluctuations fall chiefly on the poor.”

Pigou answered two critics in later editions of The Economics of Welfare. His contemporary Ralph Hawtrey (1879–1975) also saw the need for the economist to be a student of ethics. ²⁹ Hawtrey (1926, p. 189) proposed enlarging the scope of economics to include ethics—with a conception of welfare comprising “all those experiences which possess ethical value in themselves.” As this was Pigou’s own conception, he (third ed., 1929, p. 17n3) was unmoved:

Mr. Hawtrey has criticised my analysis upon the ground that it implicitly makes equal satisfactions embody equal amounts of welfare, whereas, in fact, satisfactions are of various degrees of goodness and badness. … There is, however, no difference in substance between Mr. Hawtrey and myself. We both take account of those variations of quality. Whether it is better to say, of two equal satisfactions, that one may in itself contain more good than the other, or to say that in themselves, qua satisfactions, they are equally good, but that their reactions upon the quality of the people enjoying them may differ in goodness, is chiefly a matter of words.

Both would have condemned opium eating—see section V above—in different words.

In the next edition Pigou (fourth ed., 1932, p. 84n) answered a criticism of the first welfare/dividend proposition from Georges-Henri Bousquet (1929): “Bousquet argues that economic welfare depends on the relation between incomes and needs, and that an increase in income involves, after time for adjustment has been allowed, such an increase of needs that the original relation between income and needs is re-established.”

Pigou saw something in the point but, given current levels of poverty, had to disagree: “The goal of economic betterment is not a mere illusion” he declared.

The 1930s saw the development of a “welfare economics” that had no interest in welfare, only in the dyad of economic welfare and the dividend, and, having different presuppositions, questioned Pigou’s handling of that. ³⁰ After another war and much unrelated work, Pigou returned to the subject in 1951 with related material in Alfred Marshall and Current Thought (1953) and “Some Aspects of the Welfare State” (1954).

Pigou wrote “Some Aspects of Welfare Economics” (1951a) for the American Economic Review. He (1951a, p. 287) identified two drivers of the new Welfare Economics, “as it likes to be named”—“some semi-philosophical questions about utility” and “some significant logical problems which arise out of the fact that real income is made up of a number of different things, the quantities of which vary in different proportions.” The logical problems were treated elsewhere, and his American Economic Review response—to unnamed critics—was a general statement on “fundamental issues” that included an abridged prolegomenon and a treatment of the two propositions in the new lean style.

While welfare in general was outside the scope of the new welfare economics, Pigou (1951a, p. 288) found it necessary to insist:

As it seems to me, welfare must be taken to refer either to the goodness of a man’s state of mind or to the satisfactions embodied in it…. [I]t is generally felt, in a vague way, that some sorts of satisfaction are in their nature better than others, and that quite irrespective of whether or not they entail dissatisfactions later on. If this is right, a situation containing more satisfaction is not necessarily “better” than one containing less.

Pigou did not restate his argument that a separate study of economic welfare could coexist with a larger concern with welfare—perhaps because the new welfare economists weren’t interested.

Regarding economic welfare, Pigou (p. 288) insisted we have “to decide whether or not it is the sort of thing to which the notions of greater or less and increase or decrease can properly be applied. For, if they cannot, Welfare Economics, every part and aspect of it, vanishes and leaves not a wrack behind.” This view was first expressed in the Memorandum discussed in section VII above.

One semi-philosophical question Pigou had not seen coming was the impossibility of interpersonal comparisons raised by Lionel Robbins (1938), and Pigou (1953, p. 45) half-suspected it of being frivolous: “it has been maintained, with what degree of seriousness I cannot say, that utilities enjoyed by different persons are not comparable.” His defence was:

On the basis of analogy, observation and intercourse, interpersonal comparisons can, as I think, properly be made; and, moreover, unless we have a special reason to believe the contrary, a given amount of stuff may be presumed to yield a similar amount of satisfaction, not indeed as between any one man and any other, but as between representative members of groups of individuals, such as the citizens of Birmingham and the citizens of Leeds. (1951, p. 292)

Pigou took the logical problems associated with the measurement of real income seriously, treating them in “Real Income and Economic Welfare” (1951b), a reply to Samuelson’s (1950) “Evaluation of Real National Income.” Pigou (1951b, p. 16) acknowledged that “serious defects” had been found in his analysis and he wanted to say “in my own language … how these things seem to me to stand now.” They did not stand well, for his analysis of the relationship between price/quantity data and economic welfare led him to conclude: “Thus it is only … where quantity of resources and technical conditions have changed and tastes and the purchasing power are alike for all purchasers and have not changed, that inferences about economic welfare are possible” (1951b, p. 20). The “practical instrument of great power” had crumbled.

For the profession Samuelson (1950, p. 28) offered a consolation: “Lucky it is that the remaining fifty-odd chapters of the Economics of Welfare do not depend in an essential way upon the results of the early chapters of Part I dealing with the national dividend.” It can’t have seemed so lucky to Pigou, for whom the dependence made the study of economic welfare what it was.

XII. IN SUM …

Pigou started from the beliefs that everyone should do good and that it was his personal responsibility to do good as an economist. Consequently it was necessary to understand what good is and how it could be promoted using the instruments at the economist’s disposal. And hence the triad of Wealth and Welfare. It didn’t please—either as a whole or element by element. Individual readers may have reacted positively—Norikazu Takami (2014, p. 381) quotes Hugh Dalton, “[Wealth and Welfare] was a book that helped me, more than any other, to formulate my own approach from ethics, through politics, to economics”—but few took on its project of founding economic policy on ethics or Pigou’s personal way of combining Ruskin with Marshall and Sidgwick. John Hobson (1914), who avowedly started from Ruskin, missed the welfare in Wealth and Welfare and saw the dividend masquerading as a measure of welfare. Later came others who were comfortable with the dyad of economic welfare and the national dividend but found all sorts of difficulties with Pigou’s account of it. At the end Pigou was dissatisfied with his treatment of the dividend but not, I suspect, with the rest of his approach.