Hicks, Robbins and the demise of Pigovian welfare economics: rectification and amplification.

• Author: Hennipman, Pieter
• Position: Response to Nahid Aslanbeigui, Southern Economic Journal, p. 616, January 1990

1. Introduction

   It is to be welcomed that Nahid Aslanbeigui in her article "On The Demise of Pigovian Economics," which recently appeared in this journal[2], drew attention to the merits of Pigou's work in the fields of welfare economics and employment theory. It again evinces the thorough knowledge of Pigou noticeable in her enlightening comparison between his and Marshall's opinions on important issues of economic policy[1]. Her new paper testifies to the belief that Pigou's ideas tend to be underrated nowadays, while they are in fact still relevant to modem debates. Her discussion offers valuable insights.

   The part dealing with welfare economics is concerned with the emergence in Britain during the thirties of the "new welfare economics" in the Paretian style that was to supersede the prevailing Marshallian tradition, with Pigou as foremost representative. Aslanbeigui sketches four "methodological attacks" by Hicks and Robbins on Pigou, which she holds responsible for the decline of Pigovian welfare economics [2, 621-22]. She rightly centers on these authors as the principal figures in this transition, but her outline is in several aspects unsatisfactory, incomplete, too simplified and marred by some errors and misrepresentations.

   This paper aims to present a more adequate account of the topics she discusses and their significance for the transformation of welfare economics, emphasizing the distinct contributions of Hicks and Robbins to this process and comparing them in greater detail than Aslanbeigui does with the ideas held by Pigou. In addition, as a second purpose, it examines how the opinions of the main actors on the issues involved developed beyond the period covered by Aslanbeigui. The problems touched on by her are treated in this way in sections II, Ill and IV on the concept of material welfare and ordinalism, aggregate concepts, and interpersonal comparisons of utility respectively. It appears that of the four attacks described by Aslanbeigui only the one on interpersonal comparisons was relevant to the decline of Pigou's welfare economics. A subsequent section shows how later on Hicks's and Robbins's appreciation of the new welfare economics has evolved and diverged. The concluding section comments briefly on Aslanbeigui's evaluation of the replacement of Pigovian by Paretian welfare economics.(1)

2. Material Welfare and Ordinalism

   Aslanbeigui mentions as the first attack Robbins's objections to the notion that material welfare is the subject-matter of economics [36, 4 ff]. Here she repeats an error made some years ago by Cooter and Rappaport [10, 513], neglecting a subsequent correction [13, 81; 14, 143]. Robbins's criticism was mainly aimed at Cannan and not at all at Pigou, who never held this criterion for the scope of economics. He indicated as its subject-matter "economic welfare," conceived as "that part of social welfare that can be brought directly or indirectly into relation with the measuring-rod of money" [29, 11]. He distinguished it from "material welfare," defined by him, in a much broader sense than Cannan, as "a man's income or possessions," which "may be a means to welfare" [40, 288]. In his Essay Robbins clearly regarded Pigou's definition, mentioned only in passing, as different from Cannan's [36, 21 n. 1]; later he called the distinction between his scarcity definition of economics and Pigou's criticism "a very minor point indeed" [38, 344; 40, 105]. So the alleged attack never took place.

   Secondly, she contrasts Pigou's cardinal utility concept to Robbins's preference for the ordinal definition. Therefore he "approved of Hicks's revival of the Paretian theory," which Aslanbeigui illustrates with some quotations. This juxtaposition of Robbins and Hicks does not bring out as clearly as one might wish their relative merits with regard to the substitution of ordinal for cardinal utility as the dominant version. Robbins's argument was confined to a few brief statements on the general principle of ordinalism [36, 56, 75] which he learnt particularly from Wicksteed and the Austrians [40, 103-4; 28, 159-60]. His influence was slight if not negligible compared to the impact of one of the authors of what has been called the "Hicks-Allen Revolution" [4, 368], "the great technologist in this matter" [46, 83] and "the century's leading proponent of the nonmeasurability of utility" [7, 281].

   Moreover, the change was not specifically propagated against Pigou, named neither by Robbins nor by Hicks in this connection.(2) The introduction of ordinalism was certainly not intended as an attack on his welfare economics. As Hicks later made clear, ordinalism was advanced as an improvement of the theory of demand which, to his mind, "had nothing to do with welfare" [18, XIII; its application to demand theory and to welfare economics raises "quite different issues" [17, 41. That Pigou had no reason to feel endangered on this account is shown in his final reflections on welfare economics where he himself converted to ordinalism, unconcerned with the borderline drawn by Hicks [31, 289 ff.], a view he recapitulated the next year [32, 43 ff.]. Unlike Pigou, his fellow Marshallian Robertson adopted the dualist position that ordinalism is acceptable, though not necessarily preferable, in the theory of consumer behavior, but that in welfare economics cardinalism is a requirement for interpersonal comparisons which he defended tenaciously [44, 37 ff.; 45, 57-58].

   Aslanbeigui admits that the revision of utility theory does not seem "to have been a strong factor in the demise of Pigovian economics" [2, 621 n. 7] but this still overstates the case. One wonders why she lists it among the attacks on Pigou.

   The rather superficial alliance of Robbins and Hicks in the matter of ordinalism had a curious sequel. Discussing Robertson's defense of cardinalism in 1953 [44, 17 ff.], Robbins not only played down the importance of the issue, but waned in his adherence to ordinalism. If Robertson was right to argue that ordinalism precluded a comparison of differences between preferences, he said, "I will cheerfully consent to being a cardinalist" [40, 104]. He went on to appeal to "some authoritative ordinalist, such as Professor Hicks," to elucidate the difficulty. The authority called upon did not fail his erstwhile mentor[16]. Robbins did not reply to Hicks's explanation, but judging from the broad agreement, expressed some years later, with what has been called "Hicksian introspective ordinalism" [41, X; 25, XIV] it may have saved him from forsaking their original harmony.

3. Aggregates

   Next Aslanbeigui records that Robbins's methodology was "against macroeconomics and such aggregate concepts as the national dividend." Since her treatment of this typically Austrian strand in Robbins's Essay merely consists of a few quotations, it can do with some elaboration.

   Robbins indeed harbored serious reservations about the economic significance of aggregate magnitudes [36, 56 ff.; 37, 464-65](3) His basic argument was that the truly important economic phenomena, "the valuations which the price system expresses," are relationships and hence "incapable of addition. Their aggregate has...