Cournot’s Trade Theory and its Neoclassical Appropriation: Lessons to be Learnt about the Use and Abuse of Models

Download full paper

This paper is closed for comments.


The ubiquitous partial equilibrium model of trade owes its origins to the work of Augustin Cournot. While Cournot’s name is a staple in economic textbooks, usually in relation to oligopoly theory, his work on trade has vanished from view. This is all the more surprising, given that the partial equilibrium model that he developed, to show the effects of trade, is still in use today, but without attribution. This paper aims to address why this is the case. Key to understanding the neglect of Cournot’s trade theory is to realise that he arrived at very different conclusions on the effect of trade than did his neoclassical successors, notwithstanding their deployment of a similar set of tools.

Posted for comments on 14 Jul 2016, 9:33 am.

Comments (2)

  • Achim Truger says:

    This is a highly relevant, very well-argued and inspiring paper that in my view should be published. However, I do have some doubts about parts of the argument and there is some room for improvement. The author might consider the following remarks that go from the more fundamental ones to some minor typos.

    1. Remarks about why the argument may be too narrow

    In particular section 6 of the paper is extremely inspiring. It is convincingly argued that Cournot’s approach to trade theory and his conclusions as to the effects of free trade in the two-country-case really make sense against the background of Cournot’s methodological position. Indeed, the arguments given suggest relevance for the paper that goes far beyond the particular application of trade theory and that reaches from the correct interpretation of partial/general equilibrium theory to general questions of measuring welfare and welfare economics as well as to neo-Keynesian disequilibrium approaches.

    However, what is clearly a strength in terms of the general relevance of the paper may turn into a weakness concerning the particular argument as to why mainstream economists and textbooks today do not attribute the basic partial equilibrium model of trade to Cournot. The quotes given in the paper, in particular in sections 5 and 6, naturally refer mostly to economists that specifically criticize Cournot’s trade theory as wrong or protectionist. This is plausible, given the rather narrow application to trade theory. However, given the extremely high general relevance of Cournot’s methodological views, a different explanation for neglecting his contribution to trade theory and the analytical apparatus would simply be that his general views – absence of utility/welfare foundations, disequilibrium phenomena and unemployed resources – were dismissed in the further development of general equilibrium economics. The most important question to be answered first would then be why his general views were dismissed and the particular application to trade theory would then only be of secondary priority.

    With regard to standard mainstream textbooks, of course these do not even usually stress the history of economic thought and when using the concept of consumer/producer surplus do not even always mention Alfred Marshall. So the neglect of Cournot may be due more to the overall development and approach of standard mainstream economics than to a particular aversion to his trade theory or their economic policy conclusions.

    Maybe, therefore, the author could distinguish better between the general relevance and the particular application to trade theory. But this point does not undermine the relevance of the paper or the argument.

    2. Doubts about the attempt at reconciling Cournot and neoclassical theory

    The paper in section 8 tries to reconcile Cournot’s view with neoclassical theory by seeing them as limiting cases and suggesting that empirical evidence may be used to decide the case.

    It is not clear, however, how convincing empirical evidence should be looked for, given the complexity of the issue. Maybe the author could suggest some method. In the current stage the paper sometimes reads as if the neoclassical general equilibrium framework and interpretation were still the (tacit) benchmark, e.g. when framing the question as one of using compensated or uncompensated demand curves. Most probably Cournot’s opposition against utility and welfare judgements were much more fundamental and “translating” them in such a way does not do justice to him. Perhaps taking Cournot’s methodological views more seriously would lead to more fundamental opposition to standard economics with trade theory as only one particular application.

    The last sentence of section 8 states that “Cournot’s theory, with its emphasis on production and productive forces, reveals strongly mercantilist influences.” This maybe true but it runs against the crucial argument of the paper that Cournot’s results do not stem from error or mercantilist prejudice, but from a consistent application of a methodologically sound framework. I would suggest dropping the sentence. By the way, the emphasis on production need not necessarily be mercantilist it could also be (Post-)Keynesian.

    3. Clarifying specific arguments

    Perhaps some references to standard textbooks and their treatment of free trade theory in partial equilibrium analysis could explicitly be given to prove the case.

    Footnote 3 may be difficult to understand, because it presupposes that readers are familiar with the fact that Cournot did ignore utility when using demand and supply analysis. Perhaps this fact should be clearly stated.

    In section 4 perhaps some more quotes from Cournot’s original text may be helpful.

    On page 8, the claim that Cournot may even be seen as an imperfect precursor to Keynes is fascinating, but might be substantiated a bit more.

    The claim about the compensated demand curve should be better explained.

    4. Some typos and formalities

    • p.2, para 2, line 3: “S2” and “D2”, using superscript “2” confused me to look for footnotes, so maybe subscripts could be used both in the text and in figure 1.
    • p.3, para 4, line 2 from below: “used deployed”, delete either “used” or “deployed”?
    • p.5, para 2, line 3: “thesis of his 1838 book”, insert “his”.
    • p.7, para 3, line 4: “In page 166” shouldn’t it be “On page 166”?
    • p. 8, para 3, line 14: “This would be a reasonable”, insert “be”.
    • p. 9, para 1, line 3: “country’s trade could be balanced”, insert “be”.
    • P. 10, para 2, line 4: “presupposes that that”, delete one “that”

    Achim Truger, Berlin School of Economics and Law and Macroeconomic Policy Institute at Hans-Böckler-Foundation, Düsseldorf

  • Eithne Murphy says:

    My thanks to Achim Truger for: his careful reading of my paper, his positive remarks, his many suggestions and big challenge. I am going to respond in reverse order from later, minor remarks to the earlier major observations.

    Starting with the more minor issues, I note the typos. Furthermore, most of the clarifications that Achim requested in the third part of his Comment can (and should be) addressed, with two exceptions. He mentions quoting more from Cournot’s text in section 4. This is difficult to do, since Cournot’s terminology is different to what I use. Instead, I can refer the reader to Cournot’s text, pages 150-157, where he works through the algebra and derives his conclusion. Achim also asked that I develop the claim that Cournot was an imperfect precursor to Keynes. I believe that I have done this as much as I can (in the context of this paper), especially in the previous two paragraphs.

    In the second part of his Comment, Achim suggests dropping the last sentence in Section 8, a throw-away remark about mercantilist influences. In hindsight, I agree. Regarding the rest of Section 8, I would defend my assertion that the different normative conclusions derived from the application of a common set of tools, make sense for both neoclassicals and Cournot, given their respective views on how consumer welfare is evaluated and how markets are presumed to work. I don’t see that I am necessarily taking the neoclassical general equilibrium view as a tacit benchmark, but even if it were the case, I think the contrast with Cournot exposes the extreme assumptions (especially regarding how markets function) underpinning the neoclassical perspective. I accept that letting empirical evidence act as a final arbiter between the extreme conclusions of the neoclassical view and that of Cournot is challenging. Possibly (and at a risk of sounding too neoclassical), one could argue that if the income elasticity of demand for a particular good is high, then it is more likely that the neoclassical measurement of consumer gains is overstated, but this would be less of a problem if a good has a low income effect. On the production side, it depends on the nature of markets in general and the time frame. Will those displaced from an activity find an alternative one that remunerates them as well? All I would add in this regard is that applied trade analysis (that purports to measure economic reality) implicitly adopts the neoclassical perspective when calculating allocative efficiency effects (see section 2 in the paper) and this is not justified.

    The first part of Achim’s Comment throws down the gauntlet by asking why Cournot’s general views were dismissed, thus implying that the neglect of his views on trade was of secondary importance. My first response to this is to say that Cournot’s work has not being neglected in many areas of economics, especially his partial equilibrium work on profit maximisation in different types of markets. (See Cournot oligopoly, but also his work on monopoly and unlimited competition). His analytical apparatus in this sphere is still influential today. I accept that mainstream textbooks tend to ignore history of thought, and that even Marshall’s partial equilibrium work on trade is not acknowledged. But to my mind, this is less of an issue, as Marhsall’s normative conclusions and how he derived them are still reproduced today. So his world view lives on in trade theory. This is not the case with Cournot and trade, and all the more surprising given his impact on other areas of microeconomics.

    Where I think Achim’s remarks have some bite is when he mentions the neglect of Cournot’s methodological views and his conception of general equilibrium. I accept that his chapter on trade (ch. 12) needs to be understood in the light of his methodology and general equilibrium views. However I think that I do show this in the paper. And obviously, those economists who rejected his trade theory as wrong, did so because they rejected his model of markets, his views on welfare, and what he called social income. That such views made him question the benefits of free trade would not have endeared him to his neoclassical antagonists either. My objective was to show that his analysis was internally consistent and that therefore the rejection of his trade theory was not justified.

    Cournot’s methodology and his general equilibrium perspective are addressed in Chs. 1 and 11 in what is a twelve chapter book. And, as Achim pointed out, neither has been influential in the development of general equilibrium analysis. I believe both would repay further analysis, especially chapter 11, where Cournot alludes to the limitations of his tools, but nevertheless seeks to see how far he can deploy them in answering useful questions. (And the questions he raises in this chapter are very pertinent). However, I think this is another paper.

    Again, thanks to Achim for his insightful comments. He forced me to re-read some of Cournot again, and has inspired me to consider more directly on the neglect of Cournot’s methodological views.