Reflections from Jean-Charles Hourcade on the award of IAEE Marcel Boiteux Prize to the book Planetary Economics

Planetary Economics: Energy, Climate Change and the Three Domains of Sustainable Development A short essay to set the Three Domains framework in the historical context of economic theories of real-world pricing of strategic investments.

‘When Boiteux says “the prices must tell the costs like the clock tells the time’, he defines prices not as the driver of choices but as a tool that ensures their dynamic consistency, their compatibility with budgetary constraints, and indicates, at any point in time, the costs of an expanding system…

… It is a surprising paradox that climate change economics has dedicated so little attention to finance, which has played an increasing role in the dynamics of the world economy over the past three decades. This paradox must end urgently to avoid exacerbating the misunderstanding between a ‘North’ advocating for carbon prices (and border taxes to hedge the competitiveness issues raised by divergent prices) and a ‘South’ insisting on access to finance, especially in a post-Covid-19 context that sees many developing countries in debt distress.’

The award of the Marcel Boiteux prize to Planetary Economics might be a ‘sign of the times’ as it associates the name of one of the founders of modern economic calculus with the ‘three domains’ of socio-economic behaviour (and associated pillars of policy action) –  a strange concept for most of his successors.

When I started discussing the book with Michael and Karsten, it was through the lens of someone initiated to the school of ‘engineer economists’[1] fifty years ago and surprised by today’s resignation of many economists to the gap separating their recommendations from implementation in real policies. I call resignation the reflex of attributing this gap to politicians’ not understanding economics or the lack of capacity to support economic rationality against the vagaries of public debate. I would prefer a bit of introspection. In the three domains parlance, my intuition is that this gap reveals a tendency, over decades, to restrict economics to a toolkit for a world where the second domain is ‘self-sufficient’ and can be isolated from the frictions of the first and the irrational drifts of the third. This was not the view of the founders of the post-WWII economic theory as I understood it through my companionship with some of Boiteux’s collaborators in the Seventies and Eighties and with Ken Arrow during the IPCC Second Assessment Report (1992-1995).

A revealing anecdote about the genesis of economic theory as a human adventure

Boiteux is a mathematician from the highly selective École Normale Supérieure, like future Nobel Prize laureate Gérard Debreu. Maurice Allais, also future Nobel Prize winner, hired both of them. Engineer at the École Polytechnique, Allais spent WWII in the dark basements of the École des Mines. Reflecting on the causes of the French defeat, he concluded that, from an intellectual perspective, a reason was the deficit in economics knowledge. He started self-teaching economics using a few books in French (full of words with rare equations), papers on public works written by French engineers (full of equations but frugal in words) and a few books in English (not yet the lingua franca). He de facto re-invented economics and Samuelson later said that the economic theory would have gained at least one decade if Allais had written in English. In 1946 Allais found more efficient to teach economics to mathematicians than mathematics to French economists. Convinced that the brilliant ‘normaliens’ would catch up quickly if they rubbed shoulders with the most advanced economists, he obtained a grant from the Cowles Foundation for both Boiteux and Debreu. But, following a cut in the Foundation’s budget, only one position was kept, and the decision on who could stay was made with a toss. Debreu won and met Arrow, with whom he elaborated the ‘Arrow-Debreu axiomatic’ analysis of the general equilibrium theory.

These men, about 25 years old, had lived through a disaster for humanity and could not think that the real world was in a ‘general equilibrium’ or could come back to it in a short time. They were simply called to use their capacity of abstraction and mastery of topology to invent tools to support public decisions in a ‘very imperfect world.’ Most of them pursued their careers as pure academics, but Boiteux, the loser of the toss, was hired by state-owned utility Electricité de France. Just before getting absorbed in high-level responsibilities, he established the basic principles of optimal pricing at ‘peak load’ and discontinuity in cost curves. These were elaborated in three contributions written in French[2] between 1949 and 1956, one in a francophone sectoral journal[3], the others in Econometrica[4].

The marginal cost pricing theory undermined by discontinuities in the costs curves and increasing returns?

Let us see how Boiteux tells the story[5]. He explains that there was “unanimous agreement on the need to rebuild rational (electricity) tariffs ab ovo”. At the theoretical level, there was a favourable opinion on the principle of marginal cost pricing. But how to move from principle to practice? Boiteux then takes the case of the hydroelectric plant, which was “even more appalling [than that of the thermal plant]: either the plant discharged [below its full potential], and the marginal cost of the Kwh was zero, or it ‘turbined’ the entire flow of the river and the marginal cost was no longer defined since there was no question of supplying one more Kwh.”

The only way out of this uncomfortable situation was to imagine that the installations remain exactly adapted to the demand and that the supply of an additional Kwh leads to readjust them in a precise – albeit theoretical – manner. The marginal cost would then incorporate the cost of this readjustment. Boiteux then shows with a few graphs the need to differentiate the price according to the place of the demanded kwh on the load curve and establishes the basic rule of equating price, the short-term marginal costs and the marginal development cost of the system.[6]

But this solution works only with decreasing returns as total production increases. This was roughly the case when the construction of hydroelectric dams was a pillar of the reconstruction of the French electricity system, because it was rational to schedule their building in decreasing merit order, from the most to the least cost-efficient. What would happen later, with the economies of scale expected from the deployment of thermoelectric power? The theory seemed on shaky ground: “The optimality of an economy (in the sense of Pareto) surely does not coincide with the general equilibrium of this economy (in the sense of Walras) and, vice versa, coincides only under the assumption of convexity of the consumer preference orders on the one hand and convexity of the production sets on the other… in the case of increasing returns, the sale at marginal cost does not cover the average cost”. This lead Boiteux to “add to the usual constraints of the Pareto optimum an additional constraint known as the ‘second rank’, the budgetary balance, and study the formal properties of the solution in order to draw lessons on the corrections to be made to marginal costs… and to address the principled response to those who considered that selling at marginal cost was irrelevant as long as it covered average cost“. His finding was the now well-known Ramsey-Boiteux rule that tells us that a public monopoly in charge of maximising the collective interest under the constraint of budgetary equilibrium must set price differentials in inverse proportion of the inelasticity of demand.

The marginal cost pricing and the Ramsey-Boiteux rule as tools to manage the tensions between the second and third domains?

What are the links between this seemingly technical discussion and the three domains framework? The reasoning starts from a consensus on marginal cost pricing and the impossibility of calculating meaningful marginal costs. Boiteux then imagines a theoretical readjustment of the installations and defines the marginal development costs as the ‘envelope’ of the short-term marginal costs of an optimised system. Thus, the reconnection between short-term and long-term marginal cost is possible, despite the discontinuities in the cost curve, only ‘imagining’ the system’s ‘expansion plan’ over the long run. The plan ultimately depends upon strategic choices about what the system should be over the long term. This is typically a ‘third domain’ problem.

Another connection between second and third domains is addressed in the Ramsey-Boiteux rule, present in all good public economics handbooks. This rule states that the rate of deviation from the first best pricing depends on the importance attached to the budgetary constraints imposed on a natural monopoly and the accepted gap between the average costs and the accounting costs. This gap ultimately reflects a high-level political choice on the level of public debt (and the corresponding trade-off between the consumer and the taxpayer) that is deemed acceptable. Boiteux concludes: ‘That said, this whole structure is only meaningful if, in addition, a judicious investment policy ensures that the choice of equipment and its capacity are optimal”. So, no meaningful calculation or marginal cost pricing without strategic visions. Further, because of the path-dependency of the systems, history matters. This is nicely expressed by Gabriel Dessus who hired Boiteux at EDF:  “The historicity of decisions comes into play… the marginal cost of the equipment will only be defined in the context of a specific (global) project. It is a fallacious ex post reconstruction to say that the project was determined by the marginal cost as an abstract pre-existing parameter.”

Economic analysis as ‘negotiation language’ to align the 1st,2nd and 3rd domains of policy action

This is why when Boiteux says “the prices must tell the costs like the clock tells the time”[7],he defines prices not as the driver of choices but as a tool that ensure their dynamic consistency, their compatibility with budgetary constraints, and indicates, at any point in time, the costs of an expanding system. Boiteux does not confuse this tool with the strategic choices necessary to set up the drivers, including the political credibility, of the system’s expansion. A significant historical example was abandoning the graphite-gas option to provide nuclear energy, “with US Westinghouse technology”. He explains clearly[8] that, together with (relatively weak) economic arguments, the main reason for selecting the US option (with strong guarantees of recuperating the technological know-how) was the political risk of letting the core of the French electricity system dependent on a technology controlled by a powerful institution, the CEA (Commissariat pour l’Energie Atomique), not totally disconnected from the French military system.

The question then is about the status of economic calculus, if it depends on necessarily controversial ‘visions’ of the future and if, as M. Allais said, “the economist as such, nor the technical expert, cannot say what the end goals should be. It is up to the political system to specify the goals pursued. The economist can only say one thing: whether the sought goals are actually compatible with each other and whether the means employed are actually the most appropriate for pursuing them.’[9]

Are the above reflections typical of the old days of the planning illusion now overtaken by the full development of market mechanisms? This would come to misinterpret the real message of a generation of engineer economists (to which we must add at least Tjalling Koopmans, Oscar Lange, and Leonid Kantorovich), which is a conceptual duality between the benevolent planner metaphor and the perfect market metaphor. If a ‘benevolent planner’ does not address the disconnection between short-term marginal costs and long-term development costs, the question remains on the apt market structure to reconnect them. It is not sure that the market reforms undertaken in the European Union will easily reconcile the conflict between an increasing amount of options with zero operating costs and the necessity for  wholesale  markets to fund consistent capacity and storage, a difficulty pointed out by Joskow in the US context[10].

Ultimately, Claude Henry, an heir of this tradition, concludes that economic analysis is a negotiation language amongst actors[11]. The first domain considerations represent an intrinsic dimension of this negotiation. This was the case when Boiteux succeeded to build a large alliance around a nuclear programme based on the Westinghouse technology bringing together the industry, the French Treasury and almost all trade unions. Boiteux summarises this with a little coquetry: “Had I won the toss, I do not know if I would have obtained the Nobel Prize[12], but I am sure that Debreu would never have the patience to discuss with the CGT” (the majority trade union at EDF, ideologically close to the Communist Party, which received between 20% and 25% of the votes and was capable of deciding very blocking strikes). Obviously, the nuclear choice can be contested and, reading the Boiteux own memories ‘Haute Tension’, it appears clear that, when deciding the scale of the programme and its downsized revision later[13], the ‘regulator’ was dominated by the ‘regulated’ i.e., the political body had lost part of the control. This presents the core issue of the political organisation of public choices and, in this specific case the organisation of public deliberations that allow for revising initial decisions instead of leading inevitably to a lock-in. This short text aims simply at reminding the role economists had in connecting the three domains of policy action.

Lessons for today in the context of climate change

But to resume this role, economists must accept they have to get rid of the false comfort of a second domain isolated from the first and the third. I am perhaps wrong, but I venture to give my interpretation of a trend that explains why Laurence Tubiana, who coordinated the adoption of the Paris Agreement on the French side, posed the provocative question Michael reports in the preface of the book: “Has economics actually helped, or hindered, our response to the global challenge of energy and the environment?

Responding simply ‘yes’ or ‘no’ would not do justice to the depths revealed by the question, or indeed, the answers in our book, which demonstrates the richness of materials accumulated by economists over the last decades to inform climate policy thinking. I will illustrate, however, the usefulness of a bit of self-criticism through a detail of importance for shaping the ‘world vision’, the ‘weltanschauung’, the ‘vision du monde’ behind debates about climate policy architecture. This detail is Boiteux’s forgotten embarrassment about discontinuities in the cost curves. All our models rank technical options in function of their leveraged costs and suggest aligning carbon prices to the cost of the last option to be mobilised in a scenario of carbon constraint. But the discounted annualized costs, in addition to minimising the amount of upfront investment costs, assume that the short-term marginal costs equate to the long-term development costs at any point in time and cover the required investments. The discontinuity problem thus vanishes, there is no risk for options with high up-front costs and no need for finance. This approach assumes that carbon price signals will suffice to redirect investments and the price to pay for this is to neglect the microeconomic paradox (lack of investment in activities with a reasonable 4% to 6% return rate) behind the structural infrastructure investment gap highlighted by the International Monetary Fund[14]. The causes of such situation are primarily the risk averse behaviours that penalise high capital costs and long maturity options, and the obstacle to accessing capital given the current functioning of the financial and monetary systems[15].

It is a surprising paradox that climate change economics has dedicated so little attention to finance, which has played an increasing role in the dynamics of the world economy over the past three decades. This paradox must end urgently to avoid exacerbating the misunderstanding between a ‘North’ advocating for carbon prices (and border taxes to hedge the competitiveness issues raised by heterogeneous pricing) and a ‘South’ insisting on access to finance, especially in a post-Covid-19 context that sees many developing countries in debt distress. Not only is neglecting the discontinuities in cost-curves an obstacle to a ‘well-behaved’ functioning of the markets, the lack of attention to the causes of this misunderstanding also results from an over-interpretation of the second theorem of welfare about the separability between equity and efficiency. Most IAM models picture the world’s carbon markets generating a unique global carbon price assuming implicitly that financial transfers compensate for the adverse distributive effects of such a price and do not have negative macroeconomic consequences. This neglects the A.K. Sen caveat about the consequences of the progressive separation between ethics and economics. Sen warns that separating these two traditions would be dangerous for the former.

No doubt that something has to change in our professional practices. I will not attempt to specify what this ‘something’ is, beyond the attraction to elegant mathematical formalisations as signs of legitimacy in economic journals and the discomfort we experience when questioning a result, a theorem so far considered robust becomes less useful than we thought, and the certainties of conventional wisdom to which we had become accustomed (even to criticise them!) disappear. This discomfort is reinforced by the difficulty of conceiving new solutions that are both theoretically robust and operational, and the price to pay to convince our peers of their legitimacy. It might be that, over the past three decades, accepting this discomfort seemed unjustified as the theory of rational expectations provided a theoretical basis for sticking to well-established tools. Under this hypothesis, the third domain is not a problem since the long-term solution will reveal itself progressively and the most troublesome edges from the first domain will disappear over time too.

Twelve years after the subprime crisis, the Covid-19 pandemic creates a context in which the rational expectation has lost a good part of its credibility. It is therefore timely for climate change economics to reframe its agenda in view of putting some rationale in the connection between the three domains of policy action. After all, this connection comes back to the old Seneca sentence ‘ignoranti quem portus petat, nullus ventus suus est’ (no wind is favourable to one who does not know to which harbour is sailing). We have seen that founders of modern economics like Boiteux were conscious of that. Not to re-invent the wheel, it would be appropriate to remember the efforts of this generation, to be like nani gigantum humeris insidentes (dwarfs standing on the shoulders of giants to see further[16]). Its main message is not to confuse the difference between first and secondbest option, which is a mathematical construct, with a hierarchy of the most economically rational options, which can be implemented given the technical, political and social asperities. The mathematical construct gives a very important benchmark for public discussion, but very early (1956) Lipsey and Lancaster[17] explained that the best option in a first world (which resembles a world where the second domain would have no interference with the first and the third) might not be the best in a world which is far from a tabula rasa. If we forget this message, we will not help the homo-economicus public to find the best economic conditions, to avoid the ‘homo instinctus’, trapped where the past history placed it, blocking the transformation of our development modes, and to facilitate the emergence of shared strategic visions capable of mobilizing the ‘homo evolvens’ for a just and sustainable low carbon world.


[1] The term is less used now than some decades ago. For A.K. Sen, economic science comes from both the tradition of ‘engineer economists’ whose aim was to help governments to conduct public works in an efficient manner (canals, road networks, railways, electricity) and the ‘moral science tripos’ in the times of Adam Smith discussing the content of the Aristotelian ‘good society’. Amartya Sen, On Ethics and Economics, Basil Blackwell Oxford and Cambridge, 1987.

[2] I do not give this detail out of nostalgia for a time when French was competing with English as a lingua franca but because it reveals that the drawback of being written in French was perceived as fully compensated by the novelty of the contribution.

[3] M. Boiteux: la tarification des demandes en pointe, application de la théorie de la vente au coût marginal (1949), Revue générale d’électricité.

[4] M. Boiteux: le revenu distribuable et les pertes économiques : Econometrica Apr 1951, Volume 19, Issue 2. https://doi.org/0012-9682(195104)19:2<112:L"DELP>2.0.

[5] Le calcul économique dans l’entreprise électrique : Revue de l’énergie, n° 390, février – mars 1987

[6] This purely abstract construct supported later the ‘tarif bleu’ and ‘tarif vert’ of EDF, the selling prices of electricity for respectively households and manufacturing enterprises according to the time slot

[7] Marcel Boiteux et Louis Puiseux (1969) Neutralité tarifaire et entreprises publiques, Extrait du Bulletin N°12 de Institut International d’Administration Publique, Paris

[8] Marcel Boiteux: Haute Tension, Editions Odile Jacob 1993

[9] Maurice Allais: foreword of ‘La politique de l’énergie’, séminaire du Professeur Maurice Allais, extraits des Annales des Mines d’avril 1960 à septembre 1961, Centre National de la Recherche Scientifique 1962

[10] PL Joskow, Challenges for wholesale electricity markets with intermittent renewable generation at scale: the US experience, Oxford Review of Economic Policy 35 (2), 291-331

[11] Henry C. [1984], « La micro-économie comme langage et enjeu de négociations », Revue économique, 35 (1), p. 177-197. He develops the idea of a Janus role of economic rationality (between facts, beliefs and values) in Investment projects and natural resources: Economic rationality in Janus’ role Ecological Economics Volume 1, Issue 2, May 1989, Pages 117-135

[12] A bit of coquetry because Boiteux was elected President of the Econometric Society in 1958 while he had written only three academic papers, in French, of about 20 pages each. About the last one, he writes: “This is what my scientific work boils down to, to use the big words… my relative notoriety in the world (of economists) is only due to these twenty pages. The article did indeed have a certain impact: it was the first foray into what would later become the theory of the second best.”

[13] Alain Peyrefitte, former collaborator of De Gaulle, deputy and ministry in the Pompidou’s government explains in ‘Le Mal Français’ (Plon 1976; Fayard 2006) how the nuclear program was voted ‘under hypnosis’ by the parliament in 1973.

[14] Abiad, A., Furceri, D., & Topalova, P. (2014). IMF survey: The time is right for an infrastructure push. International Monetary Fund. https://www.imf.org/en/News/Articles/2015/09/28/04/53 /sores093014a

[15] For more information on this point see chapter 2 of Jean-Charles Hourcade, Yannick Glémarec, Heleen de Conninck, Fiona Bayat-Renoux, Kilaparti & Aromar Revi, Scaling-up Climate Finance in the context of post-Covid 

[16] This expression from Bernard de Chartres (XII° century) was used by Isaac Newton and is the name given to Apollo 17

[17] Lipsey, R. G.; Lancaster, Kelvin (1956). “The General Theory of Second Best”. Review of Economic Studies24 (1): 11–32. doi:10.2307/2296233JSTOR 2296233.