https://orcid.org/0000-0001-7628-4888
![Sample our Economics, Finance,Business & Industry journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5931/TOC-Subject-Sample-2021-Economics-Finance-Business-Industry.jpg"})
Abstract
After the Second World War, optimal pricing in the public sector became an important topic internationally. The welfare enhancing properties of marginal pricing were a key concern, yet, the technical computation of marginal costs also proved difficult. It was unclear how to compute marginal costs, mainly in view of the discontinuities of the cost function. In the context of post-war reconstruction and of practically implementing a marginal pricing policy, this technical debate was closely linked in France to the “Calais traveller paradox” and the emergence of a new generation of engineer-economists contributing at the same time to the theoretical debate and to the practical implementation of marginal cost pricing. Maurice Allais and Marcel Boiteux, as well as Gabriel Dessus and Roger Hutter contributed to developing national solutions that also spread theoretical thinking internationally. This debate connects with the history of economic calculus and the rise of public economics, as well as the possibility of computing optimal welfare enhancing prices in the face of market failures.
Acknowledgements
The authors would like to thank Christine Allais, who gave us access to some valuable documents from the private papers of Maurice Allais, and the archivists at Électricité de France, who helped us with the archival work on Marcel Boiteux. We are also grateful to the editors of this special issue for their advice and support, especially Maxime Desmarais-Tremblay for his many helpful remarks on both the form and content of the article. The paper has also benefited from the comments of the participants in the workshop “From Public Finance to Public Economics” held in Graz in September 2022. Finally, we are indebted to Manuela Mosca and two anonymous reviewers who helped us to strengthen the arguments of the article.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 “Services publics industriels” being the standard French expression at the time.
2 Two instruments in the service of national reconstruction; SNCF had already been operating for several years (founded in 1938), while EDF was formed in 1946 as part of the wave of nationalisations.
3 This article draws on documents from EDF archive (in particular the “Dossier de M. Marcel Boiteux”) and from Maurice Allais’s private papers in Saint-Cloud, material to which we have had access by courtesy of his daughter Christine Allais. In Maurice Allais’s archive, we used in particular the minutes of the GRECS meeting of 31 October 1946 (there is still no comprehensive indexing of Allais’s archive). Among the numerous items in the EDF archive, we consulted materials on Marcel Boiteux, Gabriel Dessus, Roger Hutter and Maurice Allais (including private correspondence between Boiteux and Allais in particular), the main figures beyond the GRECS meeting and later the UNIPEDE meetings.
4 Engineers’ reasoning was closely anchored to the development of the railway system, which they made one of their favourite fields of analysis, as shown by Dupuit’s work (see Ekelund and Hébert Citation2012a; Numa Citation2012; Poinsot Citation2016).
5 Jacques Drèze (Citation1995) located the shift from public finance to public economics around 1950, notably pinpointing the work of Allais and French engineers at EDF (together with contributions by Vickrey and Samuelson). If by the beginning of the 1960s the label “public economics” was quite unusual for US economists (Hauchecorne Citation2020, 185), the expression was more or less in use in France as a domain encompassing both taxation and public expenditure (public finance), as well as management of the public sector (building and pricing).
6 Allais’s singularity begins with his curriculum. He had been trained as a theoretical physicist at Polytechnique and then joined the École des Mines (not Ponts et Chaussées). Hence, his school formation in economics was scarce (he just followed Divisia’s 1932 class) and was largely self-taught (on this topic, see Allais Citation1989). Notably, Allais did not read directly Dupuit’s contribution until later in the 1950s. On how Allais’s work relates to and differs from Dupuit’s, see Baumstark and Bonnafous (Citation2000) and Béraud (Citation2010).
7 In the English speaking world, the Dupuit method was associated with the Marshallian notion of consumer surplus. See, for instance, Hotelling (Citation1938a).
8 The period of the Vichy regime was not a simple parenthesis in the development of economics in France, as is too often assumed, but rather the establishment of new intellectual patterns, tools and methods that weighed on the post-war period (see Brisset and Fèvre Citation2020, Citation2021).
9 Boiteux was a mathematician trained at the Ecole Normale Supérieure, another elite state school, but not related to engineering. However, because of the influence that Allais had on his way of thinking, Boiteux is generally considered the quintessential French engineer-economist for the recent period (the other example being Edmond Malinvaud).
10 This result, for which Allais was awarded the Nobel Prize in 1988, was proved in a more general way by means of different formal tools (topological analysis in particular) by Arrow and Debreu in 1954 (see Düppe and Weintraub Citation2014). On Debreu’s meandering intellectual debt to Allais in this respect, see Fèvre and Mueller (Citation2023).
11 The “only road” according to Allais (Citation1947a, Citation1947b) was competitive planning. He occasionally spoke of “competitive socialism” (Allais Citation1947d), defined as a planning of the structures of the economy (money, credit and currencies, investment, social security, etc.) aimed at maximising social return. On Allais’s competitive planning, see Diemer and Nedelciu (Citation2020).
12 On the French side, the concept of natural monopoly can be traced back to the work of Cournot, Dupuit and Walras (see Mosca Citation2008).
13 In fact, until 1947, Meade as a civil servant tried to promote marginal cost pricing in British public industries. On the reasons for the success of Allais and his followers at EDF compared to Meade’s setback in England, see Chick (Citation2002).
14 On January 1938, the Front Populaire government decided to merge the five major French railway companies to form the SNCF. At the time, the French network had 515,000 railway workers and 42,700 km of track.
15 Allais was familiar with both issues: from 1937 to 1943, he was in charge of the mining sector in Nantes, and had a hand in the railway system that served it. On Allais’s long-term contributions to transport economics, see Bonnafous (Citation2020).
16 This seminar, which was not purely academic, was widely open to businessmen, senior civil servants and engineers, all of whom were linked through the Polytechnic network (Yon Citation2014a, 98). On the driving role of the seminar in economic research, see Cherrier and Saïdi (Citation2021).
17 Regarding the transportation of commodities, the price depended directly on the value of the commodity (ad valorem): the more expensive the good, the more expensive its transport (in practice the SNCF applied a charge per kilometre according to six categories).
18 According to Allais, the SNCF’s rail transport activity (passenger and freight) belonged to the non-differentiated sector where competition was impossible (unlike the road freight transport sector, for example, where competition is possible). Allais insisted that the difference between these two types of sector was a physical question and not an economic one per se.
19 Setting prices proportional to marginal costs was also suggested by Frisch and Hotelling in their exchange (Hotelling Citation1939a, Citation1939b; Frisch Citation1939a, Citation1939b). However, Lerner and Samuelson later argued against this, showing that “equality really is necessary for consistency in the system as a whole, taking account of the prices of the factors of production. It is apparent that if the prices of consumers’ goods were not equal to marginal cost, but factor payments were maintained equal to the marginal products of the factors, the relationship between work and leisure would be altered and the marginal conditions throughout the system would not be met” (see Ruggles Citation1949, 108).
20 Polytechnique is known in France as X. Both Allais (X-1931) and Hutter (X-1930) joined in the public Corps des Mines, the finest assignment available to X graduates. Starting in 1943, Hutter became interested in Allais’s work at the instigation of Robert Le Besnerais, General Director of the SNCF, with the aim of providing theoretical tools for more rational transport pricing (Glachant Citation1989, 110).
21 It is quite hard to know exactly who imagined the story in this exact wording. It is usually credited either to Gabriel Dessus or Maurice Allais. It is likely that the story of the Calais-traveller was developed by the entire group during Allais’s seminars as a joint attempt to increase their mutual understanding regarding the definition of marginal costs. Part of the story, especially concerning the importance of saturation, may even be traced back to Dupuis himself (see also Mosca Citation1991).
22 Dessus had a clear penchant for storytelling. Twice he opened his contributions on marginal costs with quotes from Lewis Carrol’s The hunting of the snark (marking by hand his own papers “SNARK 1,” “SNARK 2”) and often filled his articles with pedagogical stories (Archives EDF 891 066; see also Dessus Citation1950). Besides the Calais traveller, Dessus penned another apologue concerning a mountain village which, for its heating, exploits alternatively the surrounding forest and a coal mine; this story illustrates the difference between pricing at the average vs. marginal costs (Archives EDF 891 066, “Les principes généraux de la tarification dans les services publics” 1949: 17–18).
23 Hotelling addressed this issue in the section entitled “Marginal cost depends on extent of unused capacity,” saying that “when a train is completely filled, and has all the coaches it can haul, the marginal cost of carrying an extra passenger is the cost of running another train. On the other hand, in the more normal situation in which the equipment does not carry more than a small part of its capacity load, the marginal cost is virtually nothing” (Hotelling 1938b, 264).
24 For an English version of Boiteux’s Citation1949 article, see Boiteux (Citation1960).
25 Unfortunately, this sentence did not appeared in the English translation (Nelson Citation1964, 59–90).
26 The harshness of Allais’s words towards his former assistant is not very surprising. In the same years, Allais also engaged in a heated argument (mixing elements of theoretical interpretation and recognition of the primacy of contributions) with his other CNRS assistant, Gerard Debreu, as the latter embarked on a career in the United States (see Fèvre and Mueller Citation2023).
27 The question of pricing at EDF was not limited to the price of services but similar discussions also took place for wages (see Bénistand Citation2021).