The interaction between Leontief and Sraffa: No meeting, no citation, no attention?

Abstract

Samuelson often regretted that Leontief and Sraffa never cited each other (true), and seemed to pay no attention to the other's work (false). In the Foley interview Leontief suggested he never met Sraffa (false). Archival evidence shows that in the 1940s Sraffa studied Leontief's classic The Structure of American Economy; he also owned the rare mimeographed supplement, and did some calculations on Leontief's first input–output table. Leontief and Sraffa met in Cambridge (UK) in 1950 and later. In the 1980s Leontief wrote an ambitious empirical paper on technological change, rejected by the AER, and not widely read. It studied some Sraffian topics without Sraffian terminology. I construct a hypothetical reswitching example using Leontief's statistics.

Keywords:

• Sraffa
• Leontief
• input–output analysis
• reswitching

JEL Classification:

• B24
• B31
• C67

Acknowledgements

An earlier version of this paper was presented at the 19th Annual ESHET Conference at the University of Roma Tre, 14–16 May 2015, where I received several stimulating comments, especially from my discussant Nerio Naldi. I also thank the literary executors of the Kahn Papers (David Papineau), the Kaldor Papers (Tony Thirlwall), and the Sraffa Papers (John Eatwell). I like to express special gratitude to Jonathan Smith, Archivist at Trinity College, Cambridge, and to his colleagues; to Patricia McGuire, Archivist at the Modern Archive Centre, King's College, Cambridge; and to the staff of the Harvard University Archives, for very competent help in my archival searches. I always appreciated the insightful mails by Olav Bjerkholt. I thank Faye Duchin, Luigi Pasinetti and Robert Solow for information on their contacts with Leontief or Sraffa. Two anonymous referees provided many detailed and expert remarks, which improved my previous versions. All remaining errors in the present text are my own responsibility.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 At the start of my ESHET conference presentation in Rome, I introduced the following quiz. Guess the name of X, an important twentieth century economist. X grew up in an intellectual and wealthy family. X was an only child. His father was a professor; his mother came from a Jewish family. In his early career (in the 1920s) X translated a book on monetary economics into his native language. In the same period he had problems with the authorities of his non-democratic home country, and he emigrated. This was a good decision; he spent many decades at an Anglo-Saxon top university. In the 1920s X was less enthusiastic about partial equilibrium models than most of his colleagues, and wanted to create a general system that expressed the interdependence between all sectors of the economy, with special emphasis on the circular nature of production. To handle this circularity, he constructed a system of simultaneous equations, which he published only much later after unexpected delays. X was an early member of the Econometric Society, but later he wrote a letter to express his disappointment. X had read widely about the history of economics, but he presented his own system with few bibliographical references (but with a conspicuous reference to Quesnay). X had some unpleasant experiences with some security services. The reports of his security investigators contained some bizarre blunders, for example, exaggerating his list of published books. At the ESHET conference a Russian colleague then guessed X was Leontief; I suppose all the Italians thought of Sraffa. In point of fact, both solutions were correct, and remained so even after I added the very specific information that the birthday of X was 5 August. Sraffa (1898–1983) was exactly seven years older than Leontief (1905–1999).

2 A detailed comparison of the early works of Leontief and Sraffa is provided by Kurz and Salvadori (2006).

3 Elsewhere it is called “venial mistakes”: see the memorial paper “A portrait of the master as a young man” (Samuelson 2004a, p. 7). An earlier conference version (Samuelson 2000b) contained a nearly identical text, but used a different title: “Our Wassily: W.W. Leontief (1905–1999)”.

4 One of the referees correctly mentioned that most of the “oddity” seems to stem from Sraffa. This is one of the reasons why my paper has to pay special attention to Sraffa's side of the interaction.

5 The Leontief Papers at Harvard include reading lists of Leontief's courses. These lists contain many books and articles on input–output theory and empiry, but again nothing by Sraffa.

6 Both the academic and the family conditions were relevant. The Kaldor Papers contain several letters (from February, March and April 1950) by Leontief's wife, Estelle Marks, asking help from Kaldor's wife, to find a suitable school for her only child Svetlana Leontief. This teenage daughter wanted no boarding school, but a day school; not a rigid old fashioned institution, but a modern and agreeable one (Nicholas Kaldor Papers, Modern Archive Centre, King's College, Cambridge, NK/3/30/136/3-8).

7 I am indebted to Olav Bjerkholt for additional information on Schultz and Sraffa in 1933, and I refer to Gehrke and Kurz (2002) for a well-documented study on Sraffa's difficulties with Hollander.

8 Items in Sraffa's Library (now conserved in Trinity College) are indicated by the name Sraffa and a number.

9 I have not found any Leontief diaries at Harvard.

10 Sraffa's diaries nearly always used the spelling “Leontieff” instead of “Leontief”; Dobb's diary of 1950 used three different spellings in one month. Leontief wrote “Graaf” instead of the more official “Graaff”, etc. I always try to quote literally.

11 This refers to Austin Robinson's party in Sidney Sussex College, in honour of two American visitors: Wassily Leontief from Harvard, and Bernard F. Haley from Stanford.

12 Michael Postan was a professor of economic history in Cambridge, member of Peterhouse College, an anti-communist emigrant from the USSR, but even the marxist students were attracted by his unusually fascinating lectures (Hobsbawm 2002, Chapter 17).

13 Leontief interview by Silk (1976, p. 155).

14 Leontief interview by Foley (1998, p. 118).

15 For many details on Sraffa and von Neumann, see also Kurz and Salvadori (2001a, 2001b).

16 Ironically, Hicks already had an opportunity to learn about von Neumann's economic system, when he met von Neumann in 1933: “I know, from personal recollection, that he had these things in mind in September 1933, when I met him with Kaldor in Budapest. Of course I did not understand what he was saying!” (Hicks 1960, p. 676). Many years after the Hicks critique by Morgenstern (1941), the importance of the von Neumann system was explained by Hicks himself in his survey of linear theory (Hicks 1960). Moreover, after he received a copy of Sraffa's (1960) book, Hicks sent a long letter to Sraffa with several remarks on similarities between von Neumann and Sraffa. This long letter of 3 September 1960 is reproduced in full by Kurz and Salvadori (2001b, pp. 177–8).

17 Cameron's letters are quoted in the very informative paper by Millmow (2014, pp. 10–1). Amartya Sen's many personal contacts with Sraffa for all his undergraduate years were much more productive: “I ended up learning a great deal from him on a large variety of economic, political and philosophical issues”, and Sen also called Sraffa “an exceptionally stimulating educator who left a lasting mark on a great many students in Cambridge” (Sen 2004, p. 26). For more comprehensive studies on Cambridge as a place in economics, see Marcuzzo et al. (2008) and Johnson and Johnson (1977).

18 Sraffa sent out a large number of complimentary copies; see the long lists in the Sraffa Papers, D3/12/113. In this file some of the handwritten drafts appear in reverse chronological order, and reveal that Sraffa first had put a question mark after the names “Leontieff” (p. 69) and “Goodwin” (p. 67), but ultimately he included both in the more polished final list (p. 65).

19 The business seminars of 1970 and 1971 in London are mentioned in the Leontief Papers, HUG 4517.6, Box 11, Folder “London Seminar May 18—19” and HUG 4517.6, Box 13, Folder “England”. The latter folder contains the letter to Liesner.

20 The confidential appraisal of Leontief's doctoral dissertation (Leontief 1928) contained a similar remark about the young Leontief: “… the candidate received no guidance whatsoever from his academic teachers. He arrived at his present position quite independently, one might say, despite them. It is very likely that he will maintain this scientific point of view also in the future.” This appraisal was written by Ladislaus von Bortkiewicz and much later published by Leontief (1991).

21 See Parys (2013) for more details. The improvised closing speech of 24 September 1931 by Ragnar Frisch (quoted by Bjerkholt 2015, p. 1164) implicitly referred to Sraffa's talk on Lloyd. In his speech Frisch ironically remarked: “I know that there is a theory of price none of us shall forget: the price theory of Monsieur Lloyd!”. Most probably Sraffa already had left Lausanne earlier. (I thank Nerio Naldi for details on an Italian police report which shows that Sraffa's car crossed the Italian border on 24 September 1931).

22 Leontief Papers, Accession 12255, Box 10, Folder “Ancient Correspondence 1930–32”.

23 See the complete program and the abstracts of the papers in the Econometrica issue of July 1971.

24 The Sraffa Papers (A/1/8:18) contain a United Nations admission card with a stamp, dated 5 July 1947, of the “2nd Session of Economic Commission for Europe”, and Sraffa's diary on the same day mentions “Apertura conferenza”. However, this “conference” can hardly be qualified as a purely scholarly one; it had to take care of many practical political problems, as the UN-ECE included both Western and Eastern European countries.

25 See Speich Chassé (2014, pp. 203–4) for a list of all the participants.

26 I thank Nerio Naldi for providing a copy of Sraffa's paper.

27 The Sraffa Library contains no copy of Leontief's important doctoral dissertation, which was published in the Archiv für Sozialwissenschaft und Sozialpolitik (Leontief 1928).

28 Both Kahn and Leontief were born in August 1905. Kahn's long letter to Sraffa was sent from New York, 17 February 1933 (Sraffa Papers, C 150).

29 Leontief Papers, Accession 12255, Box 4, Folder “Cambridge (sic) University Press”.

30 A similar announcement appeared at the start of Appendix II (Leontief 1941, p. 141).

31 The two letters from Leontief to Keynes are conserved in the Harvard University Archives, Leontief Papers, HUG 4517.5, Box 1, Folder “General Correspondence: 1932–1941”.

32 Leontief Papers, HUG 4517.30, Box 4.

33 John Maynard Keynes Papers, King's College, Cambridge, JMK/EJ/1/3/313.

34 See Naldi (1998, pp. 513–4) for letters by Keynes (to Lydia Lopokova) and by Pigou (to Sraffa) on Sraffa's first equations. See also Naldi (2005, pp. 387, 398).

35 I am indebted to Jonathan Smith for raising this question.

36 The story about Leontief's notations is even more sophisticated, because in Leontief (1936) we did not find two subscripts yet, but a notation with letters like in Sraffa (1960, p. 4). And a fourth variant of notation can be found in the German dissertation by Leontief (1928).

37 Every reader of Leontief (1941) can find the huge folded tables for 1919 and 1929 in the pocket at the back of the book. The 1919 table was also published in Leontief (1936).

38 Part of the gross output of American agriculture (the amount 22 147) was delivered to American human consumers (the amount 2209), a much larger amount was absorbed by agriculture itself (feed for the cows, for example), or delivered to the sector Slaughtering and Meat Packing, or to Exports, etc.

39 By “From wages” Sraffa means the amount paid by the sector to wages and salaries (as computed by Leontief).

40 Sraffa leaves the entry for $d$ blank for all sectors, but in later computations for $e=c+d$ he uses the amount of Leontief's ‘Capital and entrepreneurial services’ (equal to 9818 in agriculture)

41 Sraffa implicitly takes the sum $c$ + $d$=$e$, and then he explicitly defines $f$ and $g$ as above.

42 Sraffa considers also alternative summations over 42 terms, the last one using the figures of “undistributed” items like taxes etc. It is obvious that Sraffa spent a lot of time on computing all these ratios. All rows and columns were handwritten by Sraffa. His arithmetical ‘technology’ must have been extremely primitive, compared to Leontief's use of the most advanced machinery at Harvard; see Leontief (1948).

43 In my one commodity example the maximum rate of profits R = e/(a–e). Hence, the reciprocal of R equals (a–e)/e = (a/e) – 1 = (3/1) – 1 = 2, and thus R = 1/2.

44 Leontief Papers, Accession 13712, Box 2, Folder ‘Cambridge Journal of Economics’

45 Robert Solow (letter to me, 22 March 2016) informed me that in all the many years he knew Leontief well, “Sraffa and his ideas never came up”. With respect to my current paragraph 9.2 on Leontief (1986a), Solow mentioned: “I think that Leontief would have said that those are not ‘Sraffian topics in disguise’, they are perfectly straightforward topics in conventional economics”.

46 Leontief Papers, Accession 13712, Box 11, Folder “American Economic Review”.

47 A footnote by Roncaglia (1990, p. 477) refers to a 1984 Leontief mimeo with the same title. Roncaglia makes a short remark on the analytical resemblance between the Leontief mimeo and Sraffa's wage-profit lines. Kurz (2011) concentrates on a very short paper (Leontief 1985) in the Scientific American, which also contains a few Sraffian elements, but no statistical tables. In the same period two other writings on technological change were published by Leontief (1986b) and by Leontief and Duchin (1986), but these are less relevant for the Leontief–Sraffa story.