II. ROBBINS'S CRITIQUE OF WELFARE ECONOMICS

Robbins develops his arguments about welfare economics in three places. In chapter II he distinguishes between “Ends and means.” His central argument, that the determination of ends lies outside economics, follows inexorably from his famous definition of economics as “the science which studies human behavior as a relationship between ends and scarce means which have alternative uses” (Robbins Reference Robbins1932, 15).Footnote ⁵ In contrast to earlier writers, such as John Neville Keynes (Reference Keynes1997 [1890]) who saw economics as comprising both science and art, Robbins equates it simply with the science, and hence with the relation between means and ends, not the determination of ends themselves.Footnote ⁶

The discussion gets closer to Pigovian welfare economics in chapter III, section 6, where Robbins criticizes the approach of considering separately the production and distribution of wealth. Such an approach might follow from the materialist definitions of economics that he had considered and rejected in chapter I, but it was flawed in that it was based on aggregates that had no basis in reality: “the idea of changes in the total volume of production has no precise content” (1932, 66). Robbins rejects not only interpersonal utility comparisons (and hence measures of aggregate welfare) but also aggregates of output—crucial to the Cambridge approach. Because outputs are heterogeneous, measures of output as a whole (Marshall's or Pigou's “national dividend”) can play no role in economic theory. Instead, there should be a theory of equilibrium and variations. Production and distribution were part of the theory of equilibrium.

Finally in chapter VI the first significant argument is a seven-page critique of the way economists commonly used the law of diminishing marginal utility to draw conclusions about “social utility” (1932, 120–6). Here we find the well-known argument, on which Hicks–Allen consumer theory was based, that preferences are orderings and that as a result it is illegitimate to draw conclusions that rest upon interpersonal utility comparisons. There is no scientific basis on which to compare one person's satisfaction with that of another. Normative conclusions based on “social utility” have to be abandoned. After pointing out (in section 3) that the notion of equilibrium has no ethical implications (“that there is no penumbra of approbation round the theory of equilibrium,” Robbins Reference Robbins1932, p. 127) he moves on (in section 4) to discuss two attempts to bring ethical criteria into economics. Here, his focus is on Hobson and Hawtrey. He presents them as having “urged that the boundaries of economics should be extended to include normative studies”: that economics “should pronounce on the ultimate validity of these valuations and standards” (Hobson and Hawtrey) and that it “cannot be dissociated from ethics” (Hawtrey) (Robbins Reference Robbins1932, p. 132, citing Hobson Reference Hobson1929 and Hawtrey Reference Hawtrey1926).Footnote ⁷

Thus, although Robbins does argue with great clarity and precision against there being any scientific basis for interpersonal utility comparisons, it is part of a broader argument in which clarity about ends and means, and hence about what can be seen as scientific, is central. This explains the otherwise puzzling feature of the essay that, if it were primarily an attack on Pigovian welfare economics, he does not focus at all on Pigou. Pigou is cited but not in any of the passages discussed above. He cites Pigou's definition of economics as “the study of economic welfare” but, though he clearly rejects it along with other attempts to define economics in terms of its subject matter, he does no more than point out that extant definitions have very different implications (Robbins Reference Robbins1932, 2). He follows this up (1932, 20, footnote 2) by including Pigou in a list of economists whose definitions define economics in terms of the problems discussed by Ricardo: they confine economics to “valuations of the market” instead of “press[ing] through to the valuations of the individual.” The text focuses on German writers alone.

There are, furthermore, other places where Robbins could easily have included Pigou in the list of those whom he was criticizing but did not do so. When he lists economists who define economics as “the study of the causes of material welfare,” he cites Edwin Cannan, Alfred Marshall, J. B. Clark, and—perhaps surprisingly, given that Robbins considered his approach different in other respects—Vilfredo Pareto (Robbins Reference Robbins1932, 4). In arguing against making the distinction between production and distribution central to economics, his target was clearly Cannan. Marshall's Principles is criticized, but only by comparing “the spineless platitudes” of Book IV with the “masterly sweep of Book V,” the latter dealing with “problems that are strictly economic in our sense” (Robbins Reference Robbins1932, 65).Footnote ⁸ Even more surprising is the failure to cite Pigou in the discussion of diminishing marginal utility and interpersonal utility comparisons. He cites Francis Edgeworth and, yet again, Cannan (Robbins Reference Robbins1932, 120). Perhaps he had Pigou in mind when referring to “numberless works on Applied Economics” or “the great majority of English economists who regard these propositions [concerning diminishing marginal utility] as axiomatic” but he did not mention him (1932, 121).

Of even more interest is the discussion of the relation between economics and ethics in the final chapter of the Essay.

He criticizes Hawtrey and Hobson for, essentially, not accepting his definition of economics: for not accepting that there was a “logical gulf” between positive and normative studies that no amount of ingenuity could bridge (ibid.) Whilst conceding that economists needed to concern themselves with normative issues, these were, from the viewpoint of economics, “outside interests” (1932, 134). Precision in economic arguments required that discussion of ends be kept strictly separate from discussion of means.Footnote ⁹ Ends were “ultimates” on which there was either agreement or disagreement, whereas rational analysis should produce agreement on the relation between means and ends.

It is in this section that Robbins is addressing contemporary welfare economics. What is significant here is that he does this not by arguing that Pigovian welfare economics smuggles in unjustifiable value judgments but by criticizing two economists who sought deliberately to extend the scope of welfare economics to encompass ethics.

III. PIGOVIAN WELFARE ECONOMICS

Pigou was clearly the most prominent welfare economist of the period on account of his position as Marshall's successor and through the magisterial quality of his two books, Wealth and Welfare (1912) and The Economics of Welfare, which went through four editions between 1920 and 1932. The contents of these two volumes owed much to his Cambridge predecessors, Henry Sidgwick and Alfred Marshall. He drew on and developed Marshallian theoretical tools to analyze the role of the state in a manner reminiscent of Sidgwick. As O'Donnell (Reference O'Donnell1979, 588) put it, he offered “Sidgwickian philosophy couched in Marshallian methodology.” This tradition had its roots in utilitarianism—it was aggregative and hence formally required interpersonal utility comparisons in order to be able to add together different individuals’ utilities. To this extent it was intrinsically utilitarian. However, by Pigou's time, the utilitarian element had become severely attenuated (see Backhouse 2006). In his Methods of Ethics (1874), Sidgwick had tried to defend utilitarianism, but he was not successful, for he could not prove that utilitarianism was superior to egoism. Marshall, whose youthful analysis of behavior was in terms of learned routines, was more evolutionary than utilitarian (see Raffaelli Reference Raffaelli2003). For Pigou's generation, there was also the influence of G. E. Moore who moved even more decisively away from utilitarianism and any notion of theism, and his thoroughly secular Principia Ethica (1903) became the bible of the Cambridge “Apostles,” the exclusive debating society in which the views of many Cambridge economists were formed: for him the good was apprehended directly, and could not be justified by appeal to utility or any other criterion.

In his two books on welfare economics, Pigou took much from his Cambridge predecessors. Welfare comprised states of consciousness and “economic welfare” comprised that part of welfare that arose “in connection with the earning and spending of the national dividend.” Welfare increased with both the size and distribution of national income and decreasing in its variability. Using this as his framework, Pigou then proceeded to analyze the case for state intervention, finding many cases in which private enterprise would not result in a welfare-maximizing allocation of resources. Yet though his aggregative approach can be seen as utilitarian, dependent on interpersonal comparisons of utilities, it was a highly qualified utilitarianism. He talked in terms of “satisfactions” and “desiredness” rather than utilities, with their hedonistic connotations, implying that he was thinking more in terms of preferences, thereby making it clear that he made no assumption that behavior was motivated by pleasure-seeking and pain-avoidance.

Moreover, in making it clear that “economic welfare” was only a part of overall welfare—that part of welfare that could be brought into relation with the measuring rod of money—he was clearly limiting his attention to what he believed could be analyzed scientifically.

However, after recognizing that welfare depended on cognitions, emotions, and desires as well as satisfactions, and after having and discussed the importance of feeling, character, and conditions of work on welfare, he neglected them. (See Levin Reference Levin1956, 128.) National dividend was treated as a good proxy for welfare. Simply to make this assumption was to deny the practical significance of the distinction between total welfare and economic welfare that he had just drawn. Pigou also found many reasons to shy away from the radical implication that utilitarianism, given his assumptions about utility functions, implied equalizing income. Incentives mattered. In other words, having simplified his utilitarianism, by basing welfare economics on a minimal set of ethical judgments, Pigou found reasons why not all of its conclusions were to be followed. One might conclude from this that Pigou was using utilitarianism as little more than a framework within which to pursue the “scientific” and ostensibly ethically neutral economics of Marshall. Though he did not accept that the use of this framework was justified, and though he will not have accepted all the policy conclusions, Robbins will have been sympathetic with much of what Pigou was trying to do.

IV. WELFARE ECONOMICS AS A CRITIQUE OF ECONOMIC SCIENCE

A more radical approach to welfare economics came out of Oxford, where T. H. Green put forward an ethical creed based instead on idealism, using the language of broad-church Christianity (though some of his critics questioned whether he had in fact abandoned this) and where John Ruskin was developing a clear ethical critique of economic values.Footnote ¹⁰ Oxford was the Christian Socialism of men like Bishop Charles Gore with its roots in the Oxford Movement. Between them, these thinkers inspired a group of economists who brought much more specific ethical judgments to bear on questions of welfare than did members of the Cambridge tradition. These included J. A. Hobson, who left Oxford to become involved in various ethical societies in London,Footnote ¹¹ and R. H. Tawney, a lifelong Anglican and Christian Socialist.

The most comprehensive and systematic welfare economics arising out of this tradition came from Hobson, whose Work and Wealth (1914) was presented as completing the task identified by John Ruskin: “to determine what are in reality useful or life-giving things, and by what degrees and kind of labour they are attainable and distributable” (Munera Pulveris, quoted in Hobson Reference Hobson1914, 10).Footnote ¹² He wrote of the need to develop a human standard of value and to analyze “organic welfare,” taking account of the organic structure of society, the value of whose parts depended on the whole. Hobson agreed with Pigou's remarks (quoted a couple of pages above) about the difference between welfare and economic welfare, but saw this difference as highly significant. The main reason would appear to be that Hobson was willing to use a broader range of value judgments than Pigou, many of those taken directly from Ruskin, even though these could not be quantified. He attacked the idea of economic science, with its links to quantification and the search for exactness.Footnote ¹³ The values involved came from shared human experience: “the nature and circumstances of mankind have so much in common, and the processes of civilisation are so powerfully assimilating them, as to furnish a continually increasing community of experience and feeling. It is, of course, this fund of “common sense” that constitutes the true criterion [of welfare]” (Hobson Reference Hobson1914: 321).

Hobson, a self-confessed “heretic,” is sometimes treated as an isolated individual. Robbins's selection of him as a target becomes more significant if it is noted that he represented but one of a group of English welfare economists, all expounding views to which Robbins's arguments applied.Footnote ¹⁴ Perhaps the most widely known representative of this group of “Oxford” inspired welfare economics was R. H. Tawney's analysis of the acquisitive society, defined as a society in which priority was given to protecting economic rights, in particular property rights, whilst leaving economic functions to fulfill themselves, except under exceptional circumstances (Tawney Reference Tawney1920: 17). Tawney argued that goods and activities should be judged according to the contribution made to the public purpose. People were not isolated individuals but parts of societies that had common goals and purposes, or moral principles. After saying that increased production was important, he claimed that “plenty depends upon co-operative effort, and co-operation upon moral principles” (Tawney Reference Tawney1920: 5). However, individualism had destroyed these moral principles and the purpose, without which society could not exist.

An echo of Green's idealism is found in Tawney's rejection of the utilitarian criterion as too individualistic—denying the existence of any common end for society (Tawney Reference Tawney1920: 17). Society was not an economic mechanism but a community of wills (Tawney Reference Tawney1921: 227). Given this view that societies had, or should have, common purposes, he was able to argue that some goods and activities were better than others, and that part of wealth was waste, which should never have been produced when there was still useful work to be done (Tawney Reference Tawney1920: 21).

A similar line was taken by Henry Clay, a colleague of Tawney's in the Workers Education Association in his very widely used Economics: An Introduction for the General Reader (1916; 1918). Clay did not go so far as Tawney in his view of society as involving a shared purpose, or in arguing for a functional view of society, in which rewards are clearly linked to the functions performed, where these are valued in relation to society's goals. However, he shared Tawney's view that social welfare has to be judged against shared moral values and the dominance of commercial activities can undermine those moral values. For both of them, there is simply no basis on which positive and normative economics could be separated. Normative judgments had to be made on the basis of moral values that are themselves a function of the economic organization of society.

Clay distinguished between “economic wealth” and welfare. Economic wealth—the output of the economic system—comprises both good and bad wants and is, in principle, measurable (1918, 415–7). In contrast, welfare depends on ethical views, counting wants differently according to whether they were good or bad. It was subjective and might not be measurable. He found many reasons why there might be little relationship between welfare and economic wealth, notably that preferring less would raise welfare.

Because welfare would rise if people wanted less, wants could not be taken as given.

It was not just Oxford that produced approaches to welfare that were more radical than those of the Cambridge School. Hawtrey came from Cambridge, a member of the Apostles who shared Moore's belief that good and bad were elementary properties that were perceived directly. In an argument criticized by Robbins, he proclaimed that “economics cannot be dissociated from ethics” (Hawtrey Reference Hawtrey1926: 184), on the grounds that to say that anything, whether wealth or utility, was the end of economic activity was to commit oneself to an ethical proposition. To presume that pleasure or happiness was the right end to pursue was to participate in the “cult of individualism” (Hawtrey Reference Hawtrey1926: 182). Instead, welfare judgments, therefore, had to be based on “the common ethical judgments of mankind”—on those judgments that are common to all ethical systems (Hawtrey Reference Hawtrey1926: 188).

What these economists have in common is a willingness to base welfare conclusions on a wider set of ethical judgments, for which the main justification was the claim that they were widely accepted. This approach was more radical and, especially when presented as an attack on economic science, potentially much more dangerous than that of Pigou. Pigovian welfare economics might be flawed in that there was no basis for arguing in terms of aggregates (whether of utility or output), but it represented an attempt to argue scientifically about welfare. In contrast, the economists discussed in this section offered a much more radical attack on economic science. In attacking Hobson and Hawtrey, Robbins was implicitly attacking the views of the whole group.

V. PERCEPTIONS OF WELFARE ECONOMICS BEFORE THE ESSAY

There is no doubt that what, for want of a better shorthand, can be called the “Oxford” view of welfare economics was not taken seriously in Cambridge. In correspondence, Marshall dismissed Hobson as being in too much of a hurry—one might say a journalist rather than an academic economist.Footnote ¹⁵ Neither did the younger generation, such as those at Cambridge and the London School of Economics who later contributed to the Review of Economic Studies, take notice of his work. Tawney's reputation outside socialist circles was based on his economic history. Their writing might offer important interventions in political discussions but they did not need to be taken seriously as economic theory. As for Clay and Hawtrey, their work was buried in an elementary textbook and a collection of essays: both were influential as applied economists whose work on practical problems bore no obvious relation to their writings on welfare. This seems to confirm the conventional, retrospective view, according to which welfare economics was, prior to the revival of the Paretian approach in the 1930s, essentially Pigovian.

However, whilst this may have been the view at Cambridge and amongst the young economic theorists, it ignores the fact that “Oxford” welfare economics was taken seriously elsewhere, in particular in the United States and among institutionalists (then a broad-based, influential grouping within the profession, not a heterodox minority).Footnote ¹⁶ Wesley Mitchell (Reference Mitchell1969), in a lecture probably first delivered around 1918, picked out as the representative of welfare economics not Pigou but J. A. Hobson, exploring his work in a lengthy chapter. Walton Hamilton not only wrote a review article on Clay (1919a) but elsewhere (1919b: 318) referred to the English “welfare school” comprising Webb,Footnote ¹⁷ Hobson, Tawney, Cannan, and Clay. Pigou was not even mentioned. Paul Homan (Reference Homan1927, pp. 776, 790; Reference Homan1928), surveying the state of economic theory, focused on Hobson as the author of one of “the most influential attempts to modify the content and purpose of economic theory … the only economist who has developed any comprehensive body of dissident thought.” Pigou, in contrast, was described merely as perpetuating Marshall's scheme of thought (Homan Reference Homan1927:789). Even someone more critical (Wolfe Reference Wolfe1931) considered that Wealth and Life merited a 15-page review article (as did the editor of the American Economic Review). Alternatives to Cambridge welfare economics were also taken seriously in China (Liu Reference Liu1934) and Japan (Nishizawa Reference Backhouse, Nishizawa, Backhouse and Nishizawa2010). As for the others, Clay's elementary textbook was no match for the work of Marshall or Pigou but was taken very seriously by contemporaries. The British (1916) and the American (1918) editions were widely reviewed, including a ten-page review article in the Journal of Political Economy (Hamilton Reference Hamilton1919a). Though his reputation lay elsewhere, Hawtrey was clearly anything but a marginal figure, which may explain why The Economic Problem was the subject of a review article by Robbins, much of which was later incorporated in his Essay.