Abstract

Cournot's views on the effects of international trade on social income have received little attention. Drawing upon a comprehensive analysis of Cournot's writings, this article shows that Cournot distinguished between variations in nominal income and variations in real income. He introduced the principle of compensation of demands: if the price of a commodity changes, the value of the total demand for other commodities remains unchanged. He relied on this principle to demonstrate that a unilateral removal of trade restrictions can cause real income to fall. Under specific circumstances, protective tariffs may be warranted. Cournot thus endorsed the infant-industry argument without being a die-hard protectionist. The article discusses the arguments of his critics and concludes that many commentators have misrepresented his thinking.

1. Introduction

Since the end of the Great Recession of 2008, the concept of national income and its measurement has received significant attention (Angrist, Goldberg, and Jolliffe 2021; Coyle 2016; Deaton 2013; Stiglitz, Sen, and Fitoussi 2010). Moreover, protectionist tensions between the United States and China have led economists and policymakers to reconsider the effects of a trade war (Amiti, Redding, and Weinstein 2019; Fajgelbaum et al. 2020). These recent developments offer a timely opportunity to reinvestigate the effects of international trade on economic prosperity through the writings of the nineteenth-century French economist Antoine-Augustin Cournot.

A pioneer of mathematical economics, Cournot is generally mentioned in microeconomics and public economics textbooks when students are introduced to his 1838 duopoly model consisting of a market for spring water produced by two firms (Béraud and Numa 2023). Cournot's model of strategic interaction between producers has become the cornerstone of the analysis of imperfect competition (Daughety 2008). In fact, the scope of his work transcends price theory and the study of monopoly and duopoly.

The purpose of this essay is to analyze Cournot's work on the effects of trade liberalization on “social” (i.e., national) income and discuss the debates that ensued.1 Drawing upon a comprehensive analysis of Cournot's writings, we show that Cournot relied on the principle of compensation of demands to demonstrate that a unilateral removal of trade restrictions is not necessarily a mutual source of gain for trading nations. The exporting country's real social income can go up, while the real social income of the importing country declines. Therefore, under specific circumstances, protective tariffs may be warranted. Cournot ([1863] 1981: 292) thus endorsed the infant-industry argument without being a die-hard protectionist. The significance of our research is twofold. First, Cournot's contribution constitutes one of the first attempts to study issues encompassing macroeconomics, welfare economics, and international economics with modern overtones. Moreover, it is important to note that Cournot offered an early criticism of the classical paradigm on international trade (as early as 1838) by disputing its main tenets and by challenging its prescriptive nature.

Cournot analyzes the effects of trade liberalization on social income in three publications. In 1838, in Recherches des principes mathématiques de la théorie des richesses, he defines the concept of social income; by showing that a shock in a particular industry does not affect the total demand for other commodities, he demonstrates the principle of compensation of demands; in his criticism of the classical theory of international trade, he claims that a unilateral removal (i.e., with no expected retaliation) of an import ban can cause real income to fall. In 1863, in Principes de la théorie des richesses, he relies on the results obtained in 1838 to criticize John Stuart Mill's theory of international exchange. In particular, he underlines the importance of the following assumption in Mill's analysis: the means of production that are no longer used to produce the commodity that is now imported are used in the production of the exported commodity. Finally, in Revue sommaire des doctrines économiques published in 1877, he discusses the arguments advanced by the proponents of free trade.

Following the publication of Principes in 1863 and the translations of Recherches into Italian and English in 1875 and 1897, Cournot's contributions were discussed in a questionable fashion by several authors. Karl Heinrich Hagen (1844), Roger de Fontenay (1864), Joseph Bertrand (1883), Francis Bastable (1887), Vilfredo Pareto (1892), and Francis Edgeworth (1894) incorrectly interpreted Cournot's analyses as a defense of protectionism; in fact, for Cournot studying the effects of international trade was essentially a theoretical issue. Since then, Cournot's views have been neglected in the literature on the history of international trade. For instance, he is not included in Chipman's (1965) survey. Cournot's views have not been widely discussed, with Fisher (1898), Barrault (1912), Angell (1926: 237–47), Viner (1937: 586–89), Ménard (1978), Creedy (1992), and more recently Solow (2008) standing out as rare exceptions. Note, however, that some of these authors fail to explicitly address Cournot's discussion of the principle of compensation of demands.2 The present essay aims to correct this shortcoming by presenting a more detailed analysis of the principle.

Cournot's approach is groundbreaking for his time for several reasons. He introduces the concept of social income and explains how to distinguish between changes in nominal income and changes in real income. His demonstrations are based on a basic relation between expenditure and revenue: the income of each economic agent is entirely spent. There is no money; prices are instead expressed in units of account. He obtains the following three key results, which form the basis of his criticisms of the views of Adam Smith, David Ricardo, and John Stuart Mill on international trade:

  • He demonstrates the principle of compensation of demands by showing that a shock in a particular industry does not affect the value of the total demand for other commodities (Recherches, chap. 11; Principes, bk. 3, chap. 2).

  • He shows that the establishment of free communication between two markets leads to an equalization of the price of the exchanged commodity, except for transportation costs, but does not necessarily increase the production of this commodity (Recherches, chap. 12).

  • He shows that liberalizing trade can cause adverse effects on the real social income of a country that unilaterally lifts an import ban on one of its industries (Recherches, chap. 12).

The starting point of our study is the discussion of the principle of compensation of demands.3 In chapter 11 of Recherches, Cournot analyzes the effects on social income of a shock affecting a particular commodity. He shows that the variation in social income is equal to the change in the value produced of that commodity.4 This result can be explained by the fact that the total demand for other commodities remains unaffected. We then analyze how in chapter 12 of Recherches Cournot relies on the compensation principle to study the changes in social income resulting from the communication of markets. To better understand Cournot's demonstrations, one should have a clear understanding of the authors and the ideas he targeted. Like Cournot, we focus on two tenets of the classical theory of international trade: the balance of trade and the rejection of what Cournot called the theory of barriers. Smith and his followers had maintained that the unilateral removal of trade barriers would necessarily cause national wealth to grow in the country where such a policy was implemented. They criticized the mercantilists' views of the balance of trade as follows: maintaining a trade surplus was neither desirable nor possible. Cournot subscribed to their analysis and assumed the value of imports was equal to the value of exports. However, can we conclude, like Smith, that the theory of barriers should be dismissed? Assuming that a country that decides a unilateral removal of barriers limiting imports causes exports to rise, can we conclude that the real nominal income of that country will also rise? We analyze how Cournot drew upon the principle of compensation of demands to oppose Smith's views on the theory of barriers and discuss the arguments of his critics.

2. The Principle of Compensation of Demands

Several authors maintained that a shock in a particular industry would affect the total demand for the commodities produced in other industries. For example, in his account of the 1815 crisis in England, Thomas Robert Malthus argued that the fall in the value of agricultural production had caused the recession of British manufacturing: the declining purchasing power of farmers might have reduced the demand for manufactured goods.

[This stagnation] commenced certainly with the extraordinary fall in the value of the raw produce of the land, to the amount, it has been supposed, of nearly one third. When this fall had diminished the capitals of the farmers, and still more the revenues both of landlords and farmers, and of all those who were otherwise connected with the land, their power of purchasing manufactures and foreign products was of necessity greatly diminished. The failure of home demand filled the warehouses of the manufacturers with unsold goods, which urged them to export more largely at all risks. (Malthus [1820] 1989: 493)

Similarly, Robert Torrens argued that an unusually abundant harvest decreased farmers' income and their demand for goods. He concluded that “a glut of a particular commodity may occasion a general stagnation, not merely of the commodity which first exists in excess, but of all the other commodities brought to market” (Torrens 1821: 414).

Malthus and Torrens aimed to show that an increase in wheat production reduces farmers' income and their demand for other products, leading to an excess in the overall supply of goods; but they failed to take into account the fact that the fall in the price of agricultural products reduces consumer spending and allows them to increase the amount they spend on other commodities. Cournot's ([1838] 1980: 100–101) demonstrations shed brighter light on this issue by relying on the accounting equality between revenue (credits) and expenditure (debits). The price of an agricultural commodity, such as wheat, declines. The demand for wheat climbs, but, as the elasticity of the demand for wheat is low, farmers' revenues fall, which means that their demand for other commodities falls by an equal amount. However, the consumers who initially purchased wheat now pay for their merchandise at a lower price. They will use a part of this sum to purchase more wheat. The remaining expenditure will be spent on other commodities. Overall, the total demand for other commodities remains unchanged. This does not mean that the fall in the price of wheat will not be without consequence; in fact, the structure of the demand will change, not its total amount. Cournot's reasoning is summarized in table 1. In the table, p1 is the price of wheat in the initial period, p2 is its price in the second period, D1 is the demand for wheat in the first period, and D2 is the demand for wheat in the second period.

Cournot does not discuss Malthus's and Torrens's analyses explicitly, but he asserts that when a given quantity of good changes, thereby causing its price to change, the total demand for other commodities remains unchanged: “When we consider in globo the producers and consumers of the commodity in question, the same [amount of . . .] funds remains available for the demand of all other merchandise” ([1838] 1980: 101).

The demonstration rests upon the definition of social income as the sum of incomes received by all agents, the distinction between nominal changes in income and real changes in income, and the equality between expenditures and revenues. It should be noted that Cournot ([1863] 1981: 170) considers “producers of a commodity all those who use their funds, their industry, their labor, to prepare the raw materials, manufacture, transport, store, and distribute [that commodity] to consumers.” In other words, Cournot's reasoning involves vertically integrated industries. He assumes that the incomes received by the producers of commodity i are equal to the value of the demand for that commodity. The means of production are assumed to be perfectly mobile between sectors, but contrary to some classical authors like Ricardo, Cournot does not consider the total quantity of factors of production employed to be constant.

Now suppose the price of commodity i rises from p1 to p2 (table 2). Sold quantities decline from D1 to D2. Cournot assumes that the elasticity of demand is greater than 1 in absolute value. The value of the product and therefore the income of producers fall from p1D1 to p2D2. Simply put, Cournot posits that

p2>p1p2D2<p1D1.

Because the producers of commodity i face lower revenues, agents employed in that industry reduce their purchases in other industries. The fall in the demand for other commodities is equal to the decrease in their nominal income, p1D1p2D2. Considering that the price of commodity i has risen, the agents in other sectors who continue to purchase the quantity D1 of commodity i increase their spending by an amount equal to (p2p1)D2, which they will have to subtract from their demand for commodities other than i. In contrast, the agents who no longer buy the commodity will have more funds available for other purchases. The decrease in their spending is equal to p1(D1D2). As a result, the change in the demand for commodities other than i is zero, expressed as follows:

p1D1p2D2+(p2p1)D2p1(D1D2)=0.

The same amount remains available for commodities other than i. The fall in the demand of producers of i for commodities other than i is compensated for by the rise in the demand experienced by other agents.

The total demand for other products has remained the same, but its distribution may have changed. The demand for some commodities will go up, whereas the demand for other commodities will decline. These disturbances will affect the economic system until a new equilibrium is reached. Cournot concludes that it may be admitted, as an approximation, that a variation in the income of producers of i, while modifying the distribution of the demand for other goods, does not alter its total value or alters it only by a quantity that is negligible in comparison with the variation p2D2p1D1, which is experienced by the income of producers of i.

Cournot thus admits that the change in the value of social income is equal to the change in income of producers of i, which is p2D2p1D1. The consumers who demand commodity i before and after the price increase will pay p2D2 for the quantity D2, for which they used to pay p1D2 before the price change. For them, it is as if the price of commodity i remained unchanged, while their income fell by (p2p1)D2. Therefore, one can conclude that the real change in income is the change in quantity demanded valued at the base-period price:

p2D2p1D1(p2p1)D2=p1(D2D1).

This real change in income is accompanied by a change in the quantity of means of production used, specifically a change in employment.

Cournot's terminology differentiating between nominal and real changes in income may be confusing given that money is not considered in his demonstration. For him, nominal change in income means a change in income measured by any given standard. Real change in income is a change in income measured while prices are fixed; Cournot used prices of the initial period as a reference point, which tends to minimize consumers' losses. François Vatin (1998: 156–57) claims that Cournot's distinction between nominal and real changes in income is surprising, given that in chapter 2 of Recherches, he asserted that “there are only relative values; to seek for others is to fall into a contradiction with the very concept of value in exchange which necessarily implies the idea of a ratio between two terms” (Cournot [1838] 1980: 19). Vatin's interpretation is dubious, however, for Cournot does not aim to measure real social income but only the real variation in income by eliminating the effects of price changes from the nominal variation. Therefore, to carry out his demonstration, Cournot does not need to rely on a real conception of value. One can use nominal prices or a numeraire. Switching from one standard to the other only implies a scale effect that can generally be eliminated by relying on the rate of growth of real income.

3. The Smithian Analysis of International Trade

Cournot applied his analysis of the principle of compensation of demands to what he called “the communication of markets,” that is, interregional or international trade. In Recherches, he criticized Smith's and Jean-Baptiste Say's views. Surprisingly, Ricardo's contribution is not mentioned. His text focuses on two points: Smith's and Say's criticisms of the balance of trade (and the associated price-specie flow) and their rejection of the theory of barriers. His assessment of each is different. He writes, “By an admirable piece of dialectics full of rigor and flexibility, Smith decisively undermined the system known as the balance of trade which can no longer be sustained. His own error, which was pushed much further by his disciples, was to identify with this system the theory of barriers, which does not depend on it in any way” (Cournot [1838] 1980: 121). In other words, he agrees that it is not suitable for a country to maintain a trade surplus. For him, any policy seeking to maintain a permanent trade surplus is doomed to fail. He thus dismisses the arguments advanced by Thomas Mun. In line with Smith's thesis, Cournot logically assumes that the balance of trade of both countries is at the equilibrium. Smith's and Say's conclusion was that trade was mutually beneficial because no country could be forced to engage in trade if it makes that country worse off. Some authors such as Jean-Antoine Chaptal disagreed on the basis that a country could not be better off if trade involved exporting raw materials in exchange for manufactured goods, an argument that Cournot did not use. His scenario involves a country unilaterally removing restrictions on imports. His thesis is that the country's real income can fall despite rising exports.

3.1. Criticisms of the Mercantilists' Views of the Balance of Trade

Smith ([1762–63] 1978: 300) noted that, in accordance with Thomas Mun's views, many authors considered that the wealth of a country consisted of gold and silver and that a trade surplus was the only way to increase wealth.5 But Smith contended that Mun was mistaken in arguing that the government's role was to keep or increase the stock of precious metals, because free trade always supplies a country with the necessary quantity of gold and silver.

On the other hand, if the quantity of gold and silver exceeds money demand, the government cannot and should not prevent it from flowing out of the country. Thus, one fundamental argument in the passage quoted above is that, under a gold standard, the quantity of money is endogenous, and the metal standard always allows effective money holdings to match desired money holdings.

One of the key questions raised by Hume's and Smith's analyses is the stability of the price-specie flow. Smith ([1776] 1976: 433) attributed to Mun the idea that a depreciation of the domestic currency necessarily augmented the trade deficit. This argument, Smith explained, was erroneous: if it is true that the depreciation raises the cost of foreign goods, it does not necessarily increase the value of imports, because the higher cost also tends to reduce the consumption of foreign goods.

In Traité d’économie polique, Say expounded on the balance of trade, revamping his text numerous times particularly in 1814 and 1819 (Numa 2019). In Cours d’économie politique pratique, three chapters cover the topic of the balance of trade. The emphasis is on the relations between the market for money and the foreign exchange market. He explained that an excess demand for money involved a trade surplus, which, in turn, caused an inflow of precious metals that allowed traders to satisfy the excess demand for money holdings.

Cournot approves of Smith's and Say's analyses, with no further discussion. He draws upon their contributions to study how exchange rates between currencies are determined. His reasoning involves a monetary system based on a metallic standard in which each country permanently sets its monetary unit in terms of one gram of fine silver. Exchange rates therefore vary between the silver export and import points, that is, the limits set by the cost of shipping silver from one place to another. In such a system the quantity of money is endogenous: an excess demand for money automatically leads to an import of silver and an increase in the supply of money that restores equilibrium. A similar process would eliminate an excess supply of money.

3.2. On the Issue of Trade Barriers

Many entrepreneurs believed to have found in Mun's writings a convenient justification for trade restrictions aimed to protect them from international competition. Drawing upon their criticisms of the system of the balance of trade, Smith and his followers rejected those claims and called for the removal of trade restrictions. Cournot admitted that, in some cases, Smith's thesis was valid. For instance, he approved of Smith's ([1776] 1976: 458) famous analysis of restrictions on importing foreign wine to promote grape cultivation in Scotland. The cost of producing grapes was so high in Scotland that the demand for Scottish wine would be nil. In that case, Cournot ([1838] 1980: 122) noted, importing foreign wine would not reduce national income. The problem is to find out if the outcome is the same when a government decides to unilaterally remove an import ban on a commodity already produced in the home country.

In his Lectures on Jurisprudence, Smith intimated that international trade followed the same principles as trade between individuals and that, as a result, two trading nations would inevitably profit from it:

All commerce that is carried on betwixt any two countries must necessarily be advantageous to both. The very intention of commerce is to exchange your own commodities for others which you think will be more convenient for you. When two men trade between themselves it is undoubtedly for the advantage of both. . . . The case is exactly the same betwixt any two nations. (Smith [1766] 1976: 511)6

Therefore, in his Lectures, Smith compared the utilities of imported and exported goods. In The Wealth of Nations, his perspective was different: what mattered was the cost. The issue was the same for a family or for a large country.

It is the maxim of every prudent master of a family, never to attempt to make at home what it will cost him more to make than to buy. . . . What is prudence in the conduct of every private family, can scarce be folly in that of a great kingdom. (Smith [1776] 1976: 456–57)

But Smith's reasoning included a new argument. He relied on the ideas that a country's general industry was always proportional to the capital that employs it and that the amount of that capital was not affected by the adoption or the removal of measures restricting or prohibiting imports. Consequently, international trade could only be beneficial:

If a foreign country can supply us with a commodity cheaper than we ourselves can make it, better buy it of them with some part of the produce of our own industry, employed in a way in which we have some advantage. The general industry of the country, being always in proportion to the capital which employs it, will not thereby be diminished, no more than that of the above-mentioned artificers; but only left to find out the way in which it can be employed with the greatest advantage. It is certainly not employed to the greatest advantage, when it is thus directed towards an object which it can buy cheaper than it can make. The value of its annual produce is certainly more or less diminished, when it is thus turned away from producing commodities evidently of more value than the commodity which it is directed to produce. . . . The industry of the country, therefore, is thus turned away from a more, to a less advantageous employment, and the exchangeable value of its annual produce, instead of being increased, according to the intention of the lawgiver, must necessarily be diminished by every such regulation. (Smith [1776] 1976: 457)

Prohibiting imports thus reduced nominal and real national income. Such a disposition reduced nominal income, given that it incentivized individuals to produce commodities whose value was lower than those that would be produced in the absence of trade restrictions. In that case, the real product was greater because if the quantity available for consumption of one of the commodities was greater, the quantity available for the other remained, at least, unchanged.

Smith imagined that employment would not be affected if the government lifted a ban on imports, because labor was employed in proportion to capital. Chaptal doubted Smith's thesis, insisting that imports and exports could be produced with means of production of a very different nature. Suppose a country exported wool and, in return, received from another country some cloth of equal value that happened to be produced with that wool. Chaptal (1819, 2:243) claimed that this type of exchange would be disastrous because it requires more labor to produce cloth than to produce wool—therefore employment decreases in the country exporting wool, and generally exporting raw materials before they can be transformed also decreases employment. Therefore, he concluded, trade agreements should pay close and equal attention to the value of traded commodities and to their nature. As weaving was more labor intensive than wool production, the country producing wool would be worse off in international trade even if an agreement stipulated that the value of exported wool was equivalent to the value of imported cloth. Say did not embrace Smith's thesis. This did not mean that he espoused Chaptal's conclusions, however.

When Say evaluated the gains from imports, his approach differed from that of Smith. For him, the gain from importing foreign goods was equal to the additional cost a country would have to bear if the goods were produced domestically. Suppose zi represents the quantity of good i that is imported, pz denotes the price of one unit of the imported good, and pd is the cost that would be borne if the good was produced at home. The gain can simply be expressed as zi (pdpz). The reasoning is illustrated in figure 1. The supply and demand functions of good i are considered given. Suppose good i can be imported at price pz, which is not a function of the quantity of imports and is lower than the potential price in the absence of imports. The quantity demanded at price pz is q, and the quantity produced in the country is qz. The measurement of the gain suggested by Say is the dotted rectangle. His measurement neglects changes in the gains of entrepreneurs who produce i and overestimates consumer gains. Instead, a more accurate measurement would be triangle NEM. However, it is essential to consider what happens in industries producing the goods that will later be exported to pay for importing i.

As demonstrated by the criticisms of the system of the balance of trade, lifting a ban on imports led to higher exports. As Say ([1828–29] 2010: 624) asserted, “We benefit from imports, as they foster the production of commodities that allow us to pay for those imports.” He envisioned significant expansionary effects in exporting industries: “As the foreign product can be sold at a modest price and its consumers are numerous, it multiplies consumption, and thereby [boosts] the demand for domestic products with which [the foreign product] is bought” (609). Thus, the fall in producers' income in industries where imports replaced domestic production was compensated for by the rise in incomes of exporting industries. Neither the number of workers nor their incomes were affected; only the industry that employed them changed. In The Wealth of Nations, it was because capital remained unchanged after the removal of trade protections that employment and workers' income remained unchanged. In Say's Cours, it was the rise in the demand for exported products that kept employment at its current level.

4. Cournot and the Theory of Barriers

Cournot challenged Say's conclusion. His argumentation relies on the principle of compensation of demands. Contrary to what Say claimed, if the measures prohibiting imports are removed, the total demand for other commodities will not rise. Businesses that produce them will not be incentivized to raise output and hire more workers. Cournot developed his thinking with three targets in mind. In Recherches, his criticisms are directed at Smith and Say; in Principes, John Stuart Mill receives the bulk of the criticisms.7

4.1. Changes in Social Income Resulting from the Communication of Markets

Cournot's analysis of international trade consists of an analysis of interspatial price equilibrium: his focus is the effects of trade on markets previously isolated. More specifically, he analyzes a situation where country B unilaterally (i.e., without reciprocation from country A) decides to lift a ban on imports of commodity i from country A. His scenario therefore involves asymmetrical markets, meaning that the effects of liberalizing trade between the two countries are dissimilar. In such a situation, he demonstrates that country B's social income declines while country A's social income increases, in a context where the trade balance is in equilibrium. His framework relies on the principle of compensation of demands, which implies that the fall in the production of commodity i in country B leaves the value of the total demand for other commodities unchanged.

In the initial situation, the price of commodity i in country A, pA, is lower than the price of the same commodity in country B;8 however, B prohibits the import of i. If the ban is lifted, the commodity will be imported in B. The demand for commodity i produced in country A will go up. Production will adjust, and it can be assumed that the price will increase such that pA2>pA1,9 so that the value of its production in that country will increase such that pA2DA2>pA1DA1. Let E denote the quantity of commodity i exported and CA2 denote the consumption of i in country A in the second period; the increase in the value of i produced by country A is

pA2DA2pA1DA1=pA2(E+CA2)pA1DA1.

One can then observe that the value of the demand for commodities other than i remains unchanged:

pA2(E+CA2)pA1DA1pA2E(pA2pA1)CA2+pA1(DA1CA2)=0.

The nominal increase in country A, ΔYA, is therefore equal to the rise in the value of production of i:

ΔYA=pA2DA2pA1DA1.

To estimate the change in real income, ΔyA, Cournot subtracts from the change in nominal income the loss of purchasing power experienced by the consumers of i, (pA2pA1)CA2:

ΔyA=pA2DA2pA1DA1(pA2pA1)CA2=pA2EpA1(DA1CA2).
(1)

The change in real income is equal to the value of exports minus the fall in consumption of commodity i.10 This expression is always positive, because both the price of i and its production rose: pA2>pA1etE+CA2>DA1. Cournot ([1838] 1980: 110) admitted he does not consider “the loss experienced by that class of domestic consumers who stop purchasing the more expensive commodity, and who thus make a use of their incomes less to their liking. This loss . . . cannot be measured, and does not directly affect the national wealth.” This important remark implies that the measurement of changes in real income proposed by Cournot does not constitute a measurement of changes in social welfare.11 In contrast, Smith seemed to believe that an increase in real income was a valuable measure of changes in social welfare, as he saw the choices as voluntary under an increased opportunity set. It should be noted that the rise in real income entails a higher utilization of productive resources that Cournot does not consider as given.

Consequently, in country B, the change in the value of the demand for commodities other than i is zero:

pB2DB2pB1DB1+pB2E+(pB1pB2)DB1pB2(DB2+EDB1)=0.

Therefore, the change in nominal income of country B, ΔYB, is equal to the change in the value of its production of i:

ΔYB=pB2DB2pB1DB1.

To estimate the change in the real income of country B, ΔyB, one should deduct from the fall in the nominal income the rise in purchasing power, (pB1pB2)DB1, received by the consumers of i. We therefore have

ΔyB=pB2DB2pB1DB1+(pB1pB2)DB1=pB2(DB2DB1).

The decrease in the value of the real income of country B is equal to the decrease in the production of commodity i. Cournot ([1838] 1980: 117) underscores that if i is not produced in country B or if its production is low, social income will not significantly drop, while it will go up in country A.

Through his demonstration, Cournot aimed to reply to Say, who maintained that the imported commodity could be sold at a modest price that would attract a significant number of consumers. In return, the demand by consumers in A for commodities produced in B would rise substantially, thereby stimulating the production of those commodities. Cournot ([1838] 1980: 120) countered that, although B's imports of commodity i entail equally valued exports of other commodities produced by B, the increase in the foreign demand for B's commodities other than i is more than offset by the impoverishment of B's producers of i and by the reduction of the total sum of funds that B's nationals can apply to their total demand for commodities other than i.

Roger de Fontenay (1864), in his review of Principes, reproached Cournot for failing to consider commodity j that country B exported in exchange for importing i. In fact, Fontenay's criticism is unsubstantiated because Cournot explicitly considers these exports. Taking a closer look at Fontenay's arguments is not uninteresting, however. While Cournot deals with an asymmetrical situation where country B lifts an import ban unilaterally, Fontenay analyzes a symmetrical situation in which A reciprocates by removing a ban on imports of j from B. His conclusions were obviously different than Cournot's results. It can be shown that, even if it is agreed that the value of imports of j would be equal to the value of exports of i, the effects of such a trade agreement on social income are indeterminate.

4.2. Cournot's Criticism of the Theory of International Values

Cournot's demonstration shows that opening trade can cause a contraction of real income. It is important to determine whether such a change implies a decrease in the employment of factors of production. In Recherches, Cournot's response to this question may seem unclear. To the idea that liberalizing trade can cause real income to fall, he wrote,

It is objected that when a commodity ceases to be produced in a given territory, in consequence of importation . . . the raw materials of this commodity, the capital engaged in its production, and the hands employed in its production, find other employment; and that, reciprocally, when exportation encourages production of a commodity, the increase in production cannot occur without diverting hands, capital, and raw materials from other uses. (Cournot [1838] 1980: 122)

His response to the criticism is that, when he studies the effects of a reduction in the production of a commodity following a removal of an import ban, he never assumed that the incomes of those who contribute to its production would be lost. To make his case, he reminds the reader that his demonstration entails a vertically integrated industry: the producers of commodity i include all those who, directly or indirectly, provide the raw materials or the means of production of i (Cournot [1838] 1980: 111–12; [1863] 1981: 192–94). If the commodity is now imported, the workers who indirectly participated in its production can find equally productive employments of their inputs in other industries. Cournot's response is puzzling. Admittedly, one can assume that the factors of production are mobile, but this does not solve the issue raised by Chaptal, as mentioned above: depending on the goods, the means of production are not the same. Sheep farming requires land, but workers must be hired to process wool.

When Cournot revisits the matter in Principes and in Doctrines, his response is different. He distinguishes between two schools of economic thought.

Some conceive of manufacturing and exchange as a sort of static which comprises forces of such an intensity that men cannot control and that they can only redirect by their engines; others are dynamists or rather vitalists who consider . . . ways of stimulating forces whose internal principle is directly related to the very principle of life, in order to propel them to their highest peak of energy. (Cournot [1877] 1982: 120)

Productive forces that were employed could vary and could be used more or less intensively, whereas Mill's and Ricardo's concern in their analysis of foreign trade is the optimal utilization of a given quantity of resources. Thus, Cournot opposes the principle of transferable capital (Whewell [1850] 1856: 141), which entails that the capital that is no longer used to produce commodities that are now imported can be used to produce exported commodities.

It is undoubtedly the Scottish economist who outlined this principle, a principle representative of a problem that Cournot described as a static. Indeed, Smith ([1776] 1976: 453) wrote,

The general industry of the society never can exceed what the capital of the society can employ. As the number of workmen that can be kept in employment by any particular person must bear a certain proportion to his capital, so the number of those that can be continually employed by all the members of a great society, must bear a certain proportion to the whole capital of that society, and never can exceed that proportion. No regulation of commerce can increase the quantity of industry in any society beyond what its capital can maintain. It can only divert a part of it into a direction which it might not otherwise have gone; and it is by no means certain that this artificial direction is likely to be more advantageous.

But Smith went further. Not only was the level of economic activity incapable of going beyond what the capital of the nation could employ, but the total employment was always proportional to that capital. The quantity of labor could not be increased, but labor could be misused. Productive resources were not used efficiently when a country produced at home commodities that could be acquired from abroad at a lower cost.

Ricardo's analysis of international trade was based on principles similar to those adopted by Smith. The total demand for domestic and foreign commodities is constrained by the income and the capital of the country. Although international trade is not likely to increase the value of the national product, it can significantly increase the quantity of commodities available in the country. Considering that the value of imports is measured by the value of exports, the value of imports will remain unchanged if the country can receive two times more foreign commodities in exchange for a given quantity of domestic commodities (Ricardo [1817] 1951: 128).

Cournot does not mention Ricardo's analysis. Instead, he refers to John Stuart Mill (Cournot [1863] 1981: 219). Neither in his essay Of the Laws of Interchange between Nations (1844) nor in the first two editions of Principles of Political Economy does Mill explicitly discuss the idea that the productive resources of nations are limited and that exports cannot increase without reducing the production of commodities reserved for domestic consumption. The hypothesis is certainly contained in his analysis, but Mill expounds on it only in the 1852 edition.

In the Principles and in his 1844 essay, Mill lays out his theory of international values in terms of the reciprocal demand of one country for the commodities produced by other countries:

The produce of a country exchanges for the produce of other countries, at such values as are required in order that the whole of her exports may exactly pay for the whole of her imports. . . . All trade, either between nations or individuals, is an interchange of commodities, in which the things that they respectively have to sell, constitute also their means of purchase: the supply brought by the one constitutes his demand for what is brought by the other. So that supply and demand are but another expression for reciprocal demand: and to say that value will adjust itself so as to equalize demand with supply, is in fact to say that it will adjust itself so as to equalize the demand on one side with the demand on the other. (Mill [1848] 1965: 604)

The theory of reciprocal demand states that, with only two tradable commodities and two countries, the value of the country's demand for one commodity is equal to the value of the other country's demand for the other commodity.

In the 1852 edition of the Principles, Mill starts by reproducing the sections of the previous editions. He then notes that the theory he has just laid out is not complete.

Thus far had the theory of international values been carried in the first and second editions of this work. But intelligent criticisms and subsequent further investigation, have shown that the doctrine stated in the preceding pages, though correct as far as it goes, is not yet the complete theory of the subject matter. (Mill [1852] 1965: 607–8)

Mill's concern was the possibility that multiple (or even an infinity of) equilibria could exist. He would therefore reframe his analysis to ensure that a single equilibrium exists and that it is unique (Negishi 2014: 33–42). To solve this problem, Mill introduces three new assumptions. He first assumes that the elasticities of demand of both commodities are −1, so that the value of the quantity demanded is constant. He then supposes that the same means of production are used in the production of both commodities and in the same proportions, so that when the production of one commodity declines, the means can be used to produce the other commodity. Finally, he supposes that the costs of production are not a function of output; when the equilibrium price is bounded by autarky costs, full specialization takes place.

Mill ([1848] 1965: 604) demonstrated that, in a two-good, two-country scenario, when supply and demand are equal for one of the goods, the value of the demand of country A for one good is equal to the value of the demand of country B for the other good. In Mill's example, England buys linen from Germany, which, in return, purchases cloth from England. Suppose pi is the relative price of cloth, DiG(pi) denotes Germany's demand for cloth, and DjE(pi) is England's demand for linen. The principle of reciprocal demand entails that

piDiG(pi)=DjE(pi).

Considering that Mill assumes the elasticity of demand of both goods is −1, one can write that the value of Germany's demand for cloth based on the international price is equal to the value of the demand for the good at autarky price piG in Germany:

piDiG(pi)=piGDiG(piG).

The value of England's demand for linen is equal to the value of England's supply of cloth. This supply is equal to the quantity of cloth that can be produced by capital and labor previously employed in linen's production:

DjE(pi)=pi[DjE(piE)piE].

One can conclude that

piDiG(pi)=DjE(pi)          pi[DjE(piE)piE]=piGDiG(piG),

which is in line with Mill's conclusion ([1852] 1965: 611): “The whole of the cloth which England can make with the capital previously devoted to linen, will exchange for the whole of the linen which Germany can make with the capital previously devoted to cloth.”

In Principes, Cournot ([1863] 1981: 216–44) criticizes the theory of international exchange developed by Mill. He will later reiterate his arguments more briefly in Doctrines ([1877] 1982: 114–18). His criticisms are focused on the theory of international values developed by Mill in his 1844 essay and in the first sections of chapter 28 of book 3 of the Principles, and Cournot also criticized the sections added by Mill in 1852.

Cournot disapproves of the fact that, to develop a theory of international trade, Mill and his predecessors started their analysis by considering the simplest scenario, the one “in which the exchange only involves two items between two markets. . . . In this case, they were mistaken: because exchange, even when affecting only two items, cannot take place without influencing, through its reaction, the whole of the economic system in both markets” (Cournot [1863] 1981: 224). Cournot believes that it is not possible to appreciate the effects of these reactions without relying on the principle of compensation of demands, that is, the idea that exporting or importing a commodity leaves the total demand for other commodities unchanged.

Starting off with bilateral exchange, Cournot contended, gives the false impression that trading nations are in a symmetrical situation. This hypothesis led to the conclusion that Smith had long suggested: exchange is possible only if it improves the situation of both parties. Ironically, however, the theory of international exchange does not allow one to rule out the arguments of the proponents of government intervention, because the defenders of the theory acknowledge that the gains from international trade are very unequally distributed across nations: it is even possible that only one country ends up securing almost all gains. In his Essays, Mill insisted on the fact that, by means of taxes on imports or exports, a government could change the distribution of the gains from trade between nations.

By taxing exports . . . we may, under certain circumstances, produce a division of the advantage of the trade more favourable to ourselves. In some cases, we may draw into our coffers, at the expense of foreigners, not only the whole tax, but more than the tax: in other cases, we should gain exactly the tax,—in others, less than the tax. (Mill 1844: 21)

By expanding his text in 1852, Mill sought to reply to those who pointed out that the demand functions could allow for a multiplicity or even an infinity of equilibria without being able to know what point the equilibrium would reach. Cournot recognizes that possibility; for him, this was not specific to the theory of international exchange. But he also argues that the existence of multiple equilibria should not lead to the dismissal of his theoretical approach: what matters is that the equilibrium is unique locally:

The variations of the economic system . . . are usually contained in very strict limits so that there is no reason to admit the possibility . . . of multiple equilibria in the interval of these limits. (Cournot [1863] 1981: 219)

Mill's first error is that he took the criticisms of his theory too seriously. The second one is his inappropriate response by adding the proposition that “the whole of the commodities which the two countries can respectively make for exportation, with the labour and capital thrown out of employment by importation, will exchange against one another” (Mill [1852] 1965: 611).12

All in all, Cournot rejected Mill's theory of international trade for three reasons:

  • Favorable international exchange makes available resources that were initially used to produce the good that is now imported, but these resources are not necessarily employed.

  • If one assumes that the average cost of production is an increasing function of the quantity produced, then the equality between the value of imports and the value of exports does not imply that the value of the means of production thrown out of employment by imports is equal to the value of the means of production employed in the production of exported goods.

  • Depending on the goods, the means of production are different and are not used in the same proportions. It is thus very likely that some means of production that were used to produce goods that are now imported will remain unemployed. Of course, such a situation will normally cause their price to fall, but whether this adjustment will enable them to find alternative employments in other sectors is unproven.

4.3. Cournot on Protectionism

In his 1838 Recherches, Cournot appears skeptical about generalized free trade. He attacks the classical theory of international trade by primarily focusing on the theoretical implications of trade liberalization. By the time Principes appears in 1863, his reservations about free trade have grown to the point that he now favors some form of trade protections. His support for protections is restricted to specific circumstances, motivated by the belief that emerging productive forces should be protected from international competition at least for some time. He thus comes to endorse the infant-industry argument.

Two major developments likely explain this shift in Cournot's thinking between 1838 and 1863. The first is the publication of Friedrich List's magnum opus Das national System der politischen Oekonomie in 1841 (translated into French in 1851).13 List is known to have popularized the infant-industry argument. The second development is the 1860 free-trade agreement between France and England (also known as the Cobden-Chevalier treaty). The treaty removed French import prohibitions and lowered tariffs between the two countries. This example of trade liberalization probably forced Cournot to take a stance on policy matters, something he had been reluctant to do up to this point, as shown in the closing section of Recherches (Cournot [1838] 1980: 125).

Cournot's views on trade protection blend industrial policy imperatives and economic nationalism. His rationale for government intervention in international trade is twofold. In some cases, Cournot asserts, private interests and public interests do not coincide; therefore laissez-faire principles do not apply. More specifically, what Cournot disputes is the idea that the pursuit of private interests will always and necessarily coincide with public interests under free trade. Government can intervene so that private interests are aligned with public interests (Cournot [1863] 1981: 293). Further, international competition between nations cannot be compared with (and reduced to) competition between individuals. Many vital interests are at stake in the former situation, particularly considering the rivalry between nations. If a nation loses ground, its citizens would not be able to develop and defend their interests (320–21).

Consequently, government can impose trade restrictions to serve the interests of the nation and those of its citizens. His rationale for protecting emerging industries is as follows:

The measures adopted by a government to protect its citizens against foreign competition can be accepted . . . when they stimulate the productive forces of the protected nation. . . . As the national industry develops thanks to a protective system, the resulting development of its population, its activity, its needs, will be able to open up to foreign production . . . which will create greater outlets than what was available under the protective regime. It is true that this whole theory of the development of productive forces by means of protection, education, and encouragement entails that protection is only temporary, and that the educative [period] ends [at some point]. (Cournot [1863] 1981: 321–22)

Similarly, he wrote,

Nobody doubts any more, in principle, the utility of a temporary encouragement.14 . . . When a new industry has just been born (even that which is destined to have the most vigorous growth), must it not at first be sheltered, and get acclimatized, and all those employed by it, must they not make their apprenticeship? . . . Time is needed to give a new direction to consumers' preferences, needs, and habits to expand the market for the new product, to the extent possible given the most economical conditions of production. Therefore, with no encouragement or special protection from the government, the infant industry might be strangled in its early stages. (292)

However, Cournot was perfectly aware of the dangers of protectionism. The practice shows that, once granted to a few industries, protections tend to be extended to a growing number of recipients under the pressure of special interests: “The more the protective system becomes completed, the more difficult it becomes to appreciate its consequences and measure its effect. . . . The maxim laissez-faire, even if it does not have (as some would like it) the value of an axiom or a theorem, must definitely be preferred in many cases as an adage of practical wisdom” (Cournot [1863] 1981: 295).

5. Cournot and His Critics

Among the late nineteenth-century criticisms of Cournot's views on the variations of social income resulting from international trade, the most interesting are Pareto's (1892) and Edgeworth's (1894). In this section, we focus on the important contribution of Pareto, who is often mentioned in the literature but rarely appraised.15 In their articles, the authors examine to what extent national income is an accurate measure of social welfare and whether such a concept is a pertinent tool to assess the effects of a specific economic policy.16 Their criticisms are concerned with three topics in Cournot's writings: the measurement of real income, the relationship between changes in social income and changes in social welfare, and Cournot's rejection of the principle of transferable capital.

5.1. Measuring Real Income

What Pareto and Edgeworth criticize is Cournot's definition of a change in real income. In the example given by Cournot ([1838] 1980: 102–3), the price of commodity i climbs from p1 to p2, causing the demand to fall from D1 to D2 and reducing the incomes of producers. The decrease in nominal income is p2D2p1D1. To convert this change in nominal income into a change in real income, Cournot notes that, for the consumers who continue to purchase the commodity despite the price hike, the situation is as if their income had dropped by the quantity (p2p1)D2. The change in real income is p1(D2D1). Irrespective of the loss experienced by individuals who cease to purchase a commodity perceived as being too expensive, Cournot underestimates the decline in social income. Pareto claims that by neglecting such losses, Cournot has reached “strange” conclusions.

Albeit less critical of Cournot's analysis than Pareto, Edgeworth was somewhat perplexed. He wrote,

I [cannot] explain why, upon the interpretation of real revenue here suggested, the loss due to a rise of price should be formulated as (D1D2), multiplied by p1 rather than p2; except so far as in the method in question there must be always something arbitrary in the selection of the price to be operated with. (Edgeworth 1894: 628)

Although Cournot is not very explicit on this issue, Pareto and Edgeworth insinuate that, to calculate variations in real income, Cournot always values quantities at base-period prices. This is incorrect, however. When Cournot ([1838] 1980: 105) analyzes the effects of a price drop resulting from a decrease in production costs, he values the rise in quantities produced at second-period prices. He writes that the increase in real income is p2(D2D1). As for a price increase, Cournot ([1838] 1980: 102) writes that the fall in real income is p1(D1D2).17 Therefore, in both cases, Cournot minimizes the changes in real social income. One can conclude that the variation in the real social income—which is not an actual estimation—is between p1(D2D1)and p2(D2D1).

Concluding that Cournot values quantities at base-period prices, Pareto highlights what he views as “contradictory” results. The first is that the change in real income in a given period is not equal to the sum of its changes in the subperiods. Suppose y is the real income; then

y3y1(y3y2)+(y2y1),

given that p1(D3D1)p2(D3D2)+p1(D2D1). This problem will later be addressed when statisticians and economists develop methods of index calculation. One solution is to estimate chain indexes whose weighting varies over time. In the case mentioned by Pareto, one should estimate the change in real income between period 0 and period 2 as p1(D2D1)+p2(D3D2) or as p2(D2D1)+p3(D3D2).

According to Pareto, Cournot's formula implies that, if the initial price is zero, the change in real income is zero: “From this we deduce that if the air, which is now free, should become a scarce commodity which had to be paid for, the real decline in social income would be zero” (Pareto 1892: 5). Another case would lead to a similar result. If the production of a commodity remains unchanged, the change in real income is zero when the price changes. “Hence, if bread becomes dearer and consumption remains at the same level as before, the real decline in income is zero” (5). Yet, in both cases the state of the society is profoundly changed. However, what the first scenario suggests is not that the choice of price is arbitrary but that it should be appropriate to the problem under consideration. When new commodities are produced in an economy, they cannot be considered if they are valued at base-period prices, that is, using a reference period where they did not yet exist. In the second scenario, it is not real income itself but the distribution of real income that changes: bread producers have gained what consumers have lost. Given how the concept of real income is constructed, it does not take into account how the distribution of income evolves over time. As Pareto pointed out, the concept of change in real income can be problematic; but this does not imply that it must be ruled out. In this controversy, Cournot's thesis eventually prevailed.

5.2. Social Income and Welfare

The gist of Pareto's criticism rests on the idea that changes in social income do not accurately account for variations in social welfare. He uses the following apologue to make his case. Having conquered nation A in war, nation B imposes on it an annual tribute of a certain quantity D¯ of wheat, which is then distributed free of charge to people living in B. Let p1 andp2 denote the price of wheat in nation B in the first and in the second period, respectively. And let q1 andq2 denote the quantities of wheat produced by B in these two periods. To simplify, Pareto assumes that the consumption of wheat remains constant. One can write q1=q2+D¯.

Cournot's compensation principle entails that, from one period to the next, the total demand for commodities other than wheat remains constant. Nominal income declines by an amount equal to the fall in the value of wheat produced in B: p2q2p1q1.18 Individuals who continue to purchase wheat produced in B spend less (wheat is cheaper). Their gain in purchasing power is (p1p2)q2. Individuals who receive free wheat paid for by A experience an increase in real income by an amount equal to p1(q1q2). The purchasing power of consumers rises globally by the amount (p1p2)q2+p1(q1q2)=p1q1p2q2, which is exactly equal to the decline in nominal income. Real income is unchanged.

Pareto's conclusions are twofold. For him, any reasoning that leads to such absurd conclusions—it is a matter of indifference for B whether it receives the tribute imposed on A—is obviously erroneous. Pareto (1892: 11) claims that “the error consists mainly in taking, as the index of a good or bad economic state of the country, this quantity defined as the real variation in national income.” He adds that “Cournot condemns free trade because it leads to a real decline in national income, and thus shows that, according to him, such a decline is harmful” (10).

Despite his scathing attacks, Pareto's conclusions are fragile. His argument is based on the fact that the payment of the tribute is not taken into account in the calculation of nominal income. Although the payment of the tribute does not raise real income according to Pareto's measurement, one cannot infer that the citizens of B are indifferent about receiving the tribute imposed on nation A. Some of those who receive wheat for free will be very satisfied. For their part, the producers of wheat will be dissatisfied, and so will be those forced to go out of business and those compelled to lower their prices. Furthermore, Cournot refrains from interpreting the fall in real income as harmful, nor does he consider a rise in real income a benefit. In the closing section of Recherches, he is adamant that, to understand the effects of measures that prohibit or restrict freedom of trade,

it is not enough to accurately analyze the influence of such measures on the national income; their tendency as to the distribution of national wealth should also be investigated. We have no intention of tackling this delicate question. . . . If we have tried to fight the doctrine of Smith's school as to barriers, it was only from theoretical considerations, and not in the least to make ourselves the advocates of prohibitory and restrictive laws. (Cournot [1838] 1980: 125)

5.3. The Principle of Transferable Capital

Cournot received a lot of criticism for embracing the idea that the quantity of factors of production employed is variable. In particular, his readers were puzzled by the fact that, in his analysis of the removal of trade protections, the means of production previously used to produce the good that is now imported can remain idle. Karl Hagen (1844) is the first author who raised the issue. He argues that, if an item is no longer produced or if the item is produced in smaller quantities, capital and labor previously used to produce the good will find alternative employments in other industries. Conversely, if the importation of a good previously authorized is prohibited, domestic production will not be able to develop without diverting capital and labor from other employments. The argument is taken up by Edgeworth (1894: 629) and Pareto (1892: 12), among others; it will later be criticized by Henry Émile Barrault (1912) and James Angell (1926: 238–47).

In his response to Hagen, Cournot ([1863] 1981: 211–12; [1877] 1982: 114) underscores the fact that his reasoning entails vertically integrated industries, which de facto includes the possibility of a transfer of resources to other employments. If commodity M can be used to produce commodities A, B, C . . . and if the demand for commodity A decreases, greater quantities of M will be used to produce B, C . . . Thus, Cournot replies that he preemptively addressed Hagen's criticism. However, his response left Edgeworth (1894: 629) and Jacob Viner (1937: 587) unswayed. Cournot maintains that he has admitted that a factor of production used to produce a given commodity can easily be employed in the production of another. He therefore admits that the mobility of the means of production can be perfect. Nonetheless, in his example involving the removal of trade barriers, the demand for other commodities remains unchanged, thereby leaving the resources used to produce the commodity that is now imported idle. Cournot ([1863] 1981: 220) notes that, in such a scenario, an adjustment process will take place and the prices of the means of production that remain unemployed will go down; his analysis does not go any further. Cournot repeatedly puts the emphasis on his disapproval of the premise adopted by Smith, Ricardo, and Mill, which considers the quantities of inputs employed as given. The implication is that capital and labor that are no longer used in the production of commodities now imported are necessarily employed in the production of exported commodities. By discarding this hypothesis, Cournot aimed to study the potential effects of a unilateral removal of trade barriers on social income in a more pertinent fashion.

6. Conclusion

Cournot was aware of the limitations of his partial equilibrium approach, citing the need to approach international trade issues in terms of general equilibrium. However, he noted that studying a few variables in isolation was a necessary detour. Indeed, “any demonstration ought to proceed from the simple to the complex” (Cournot [1838] 1980: 42). He used specific assumptions, and his analysis is concerned with a specific situation: a unilateral removal of trade protections.

His purpose was not to develop a new theory of international trade but to analyze the pertinence of prevalent theories with a new method of economic investigation (Cournot [1863] 1981: 308). He cautioned against any generalization of the classical dogma praising the alleged merits of free trade. Cournot was essentially a pragmatist: while praising freedom and competition in general, he seemed to disdain dogmas, which led him to espouse the infant-industry argument.

Summarizing his criticisms of Cournot's analysis, Pareto (1892: 12) concludes as follows: “His whole theory of social income is vitiated by this error [the rejection of the principle of transferable capital], rests on falsehood, and is useless to study.” Unfortunately, many economists accepted Pareto's arguments. Just as Joseph Bertrand (1883) and Edgeworth (1897) wrongly concluded that Cournot's duopoly theory was erroneous, and just as Léon Walras ([1863] 1987) wrongly asserted that Cournot merely borrowed, without changing anything, from Ricardo's political economy, Pareto unsuccessfully searched for an error that did not exist. Cournot certainly relied on specific premises to carry out his analysis of the effects of the communication of markets on social income. His views can be debated using different hypotheses; in and of itself, however, this does not mean that Cournot's demonstration was erroneous.

Cournot's peculiar way of defining real income can be questioned, but it is not so different from the approach currently used by national accounting specialists. Admittedly, the price system used by Cournot to value physical quantities is problematic, but the method used to deal with problems arising from his analysis is acceptable. Moreover, although real income is not an appropriate measure of social welfare, criticizing Cournot for overlooking changes in consumer surplus in his calculation of the variations in social income is not convincing. Just as unconvincing is the argument that he ruled out the principle of transferable capital. Smith, Ricardo, Mill, and Whewell hypothesized that the means of production that were no longer used to produce the commodity that was now imported were employed in the production of the exported commodity. It makes little sense to criticize Cournot because he studied the effects on social income when this hypothesis was relaxed.

Notes

1.

Cournot ([1838] 1980: 113) wrote, “The term ‘social income’ may . . . be substituted for ‘national income’ which . . . includes . . . the total sum of individual incomes, rents, profits, and wages of every kind, in the whole extent of the national territory.” Translations of Cournot’s writings are ours unless otherwise indicated.

2.

Notable exceptions are Theocharis (1983: 182–236) and, to a lesser extent, Gomes (1990: 10–22).

3.

The term principle of compensation of demands does not appear in Recherches, although it is clearly discussed. The term is introduced in Principes (Cournot [1863] 1981: 179).

4.

With an error not exceeding quantities of the second degree” (Cournot [1838] 1980: 151; emphasis in original).

5.

In The Wealth of Nations, Smith’s arguments were similar to the thesis put forward by David Hume in Of the Balance of Trade in 1752 (Hume [1752] 1955). However, given that Cournot referred to Smith, it is more logical to examine Smith’s arguments instead of Hume’s.

6.

In the lecture given in 1762–63, he used similar arguments. See Smith (1762–63) 1978: 390.

7.

In Recherches, Smith is mentioned in eight instances and Say is mentioned in four instances; in Principes, Smith is cited five times, Say is cited twice, and John Stuart Mill is cited seven times.

8.

Keep in mind that all prices are expressed in units of account. There is no money in Cournot’s demonstration.

9.

Cournot’s reasoning involves unlimited competition between producers of i in A. He assumes that the marginal cost is an increasing function of output (Cournot [1838] 1980: 70).

10.

Equation (1) involves prices of different periods, which is puzzling: exports are valued at second-period prices, while the fall in consumption is valued at base-period prices. This result can be explained by the fact that Cournot considers only the loss experienced by individuals who continue to purchase commodity i.

11.

Note that Cournot does not consider the concept of surplus that Dupuit (1844) will develop a few years later.

12.

Cournot quotes the French translation of Mill’s Principles by Hippolyte Dussard and Jean-Gustave Courcelle-Seneuil (Mill 1854, 2:193). As Gérard Jorland points out in his edition of Cournot’s Principes, the translation is inaccurate. In fact, Mill’s original text mentions “labour and capital.” However, Jorland notes that even in its original formulation Mill’s statement is incomplete because there are other means of production besides capital and labor.

13.

It should be noted that List is mentioned by Cournot in the introduction to Principes and in the preface to Doctrines. Although the seeds of Cournot’s skepticism toward generalized free trade were planted in Recherches, Cournot’s reading of List most likely influenced his views in favor of protectionism (Barrault 1912: 120n4; Gomes 1990: 17).

14.

This statement shows that Cournot was probably aware that some of his main targets—such as Say and Mill—endorsed the infant-industry argument. On Say’s support for the infant-industry argument, see Numa 2019. As for Mill, see Irwin 1991.

15.

As one referee correctly pointed out, Pareto’s criticisms of Cournot can be explained by the fact that protectionist views were popular in Italy at that time.

16.

Irving Fisher subscribes to Pareto’s opinion in his article “Cournot and Mathematical Economics,” which, for a long time, constituted the main reference for an assessment of Cournot’s contribution to economics. Fisher (1898: 132) concludes, “In the final chapter Cournot applies his ideas of income to international trade, and attempts to show in particular that a protective tariff may, under special circumstances, increase the national income. Inasmuch as the idea of income is so arbitrary and faulty, little or no importance attaches to such speculations.”

17.

Solow (2008: 109, 111) notes that the fall in real income is a Laspeyres-type expression, whereas the increase in real income is a Paasche-type expression.

18.

It should be noted that Pareto omits the payment of the tribute that increases nominal income.

References

Amiti, Mary, Stephen J. Redding, and David E. Weinstein.
2019
. “
The Impact of the 2018 Trade War on U.S. Prices and Welfare
.”
Journal of Economic Perspectives
33
, no.
4
:
187
210
.
Angell, James W.
1926
.
The Theory of International Trade: History, Criticism, and Restatement
. Harvard Economic Studies 28.
Cambridge, Mass.
:
Harvard University Press
.
Angrist, Noam, Pinelopi K. Goldberg, and Dean Jolliffe.
2021
. “
Why Is Growth in Developing Countries So Hard to Measure?
Journal of Economic Perspectives
35
, no.
3
:
215
42
.
Barrault, Henry Émile.
1912
. “
Les doctrines de Cournot sur le commerce international
.”
Revue d'histoire des doctrines économiques et sociales
5
:
110
24
.
Bastable, Francis.
1887
.
The Theory of International Trade with Some of Its Applications to Economic Policy.
Dublin
:
Hodges, Figgis
.
Béraud, Alain, and Guy Numa.
2023
. “
Antoine-Augustin Cournot and the Emergence of Mathematical Economics
.” In
A History of Economic Thought in France
, edited by G. Faccarello and C. Silvant.
London
:
Routledge, forthcoming
.
Bertrand, Joseph.
1883
. Review of
Théorie de la richesse sociale
, by Léon Walras, and Recherches sur les principes mathématiques de la théorie des richesses, by Augustin Cournot. Journal des savants,
September
,
499
508
.
Chaptal, Jean-Antoine.
1819
.
De l'industrie française
.
2
vols.
Paris
:
Renouard
.
Chipman, John S.
1965
. “
A Survey of the Theory of International Trade.” Pt. 1, “The Classical Theory
.”
Econometrica
33
, no.
3
:
477
519
.
Coyle, Diane.
2016
.
GDP: A Brief but Affectionate History
. Rev. and expanded ed.
Princeton, N.J.
:
Princeton University Press
.
Cournot, Antoine-Augustin. (
1838
) 1980.
Recherches sur les principes mathématiques de la théorie des richesses
. Vol.
8
of Œuvres complètes, edited by G. Jorland.
Paris
:
Vrin
.
Cournot, Antoine-Augustin. (
1863
) 1981.
Principes de la théorie des richesses
. Vol.
9
of Œuvres complètes, edited by G. Jorland.
Paris
:
Vrin
.
Cournot, Antoine-Augustin. (
1877
) 1982.
Revue sommaire des doctrines économiques
. Vol.
10
of Œuvres complètes, edited by G. Jorland.
Paris
:
Vrin
.
Creedy, John.
1992
. “
Cournot on Trade between Regions and the Transition from Partial to General Equilibrium
.”
History of Economics Review
18
, no.
1
:
10
18
.
Daughety, Andrew F.
2008
. “
Cournot Competition
.” In
The New Palgrave Dictionary of Economics
, edited by Steven N. Durlauf and Lawrence E. Blume,
2408
16
. 2nd ed.
London
:
Palgrave Macmillan
.
Deaton, Angus.
2013
.
The Great Escape: Health, Wealth, and the Origins of Inequality
.
Princeton, N.J.
:
Princeton University Press
.
Dupuit, Jules.
1844
. “
De la mesure de l'utilité des travaux publics
.”
Annales des ponts et chaussées
8
, no.
116
:
332
75
.
Edgeworth, Francis Y.
1894
. “
Theory of International Values.” Pt. 3
.
Economic Journal
4
, no.
16
:
606
38
.
Edgeworth, Francis Y.
1897
. “
Teoria pura del monopolo
.”
Giornale degli economisti
15
:
13
21
, 307–20, 405–14.
Fajgelbaum, Pablo D., Pinelopi K. Goldberg, Patrick J. Kennedy, and Amit K. Khandelwal.
2020
. “
The Return to Protectionism
.”
Quarterly Journal of Economics
135
, no.
1
:
1
55
.
Fisher, Irving.
1898
. “
Cournot and Mathematical Economics
.”
Quarterly Journal of Economics
12
, no.
2
:
119
38
.
Fontenay, Roger de.
1864
. “
Principes de la théorie des richesses par M. Cournot
.”
Journal des Économistes
, 2nd ser.,
43
:
231
51
.
Gomes, Leonard.
1990
.
Neoclassical International Economics: An Historical Survey
.
New York
:
St. Martin's Press
.
Hagen, Karl Heinrich.
1844
.
Die Nothwendigkeit der Handelsfreiheit für das Nationaleinkommen: Mathematisch nachgewiesen
.
Königsberg
:
Gräfe und Unzer
.
Hume, David. (
1752
) 1955.
Of the Balance of Trade
. In Writings on Economics, edited by Eugene Rotwein,
60
77
.
London
:
Nelson
.
Irwin, Douglas A.
1991
. “
Retrospectives: Challenges to Free Trade
.”
Journal of Economic Perspectives
5
, no.
2
:
201
8
.
Malthus, Thomas Robert. (
1820
) 1989.
Principles of Political Economy
. Variorum ed. Edited by John Pullen.
Cambridge
:
Cambridge University Press
.
Marshall, Alfred.
1890
.
Principles of Economics
.
London
:
Macmillan
.
Ménard, Claude.
1978
.
La formation d'une rationalité économique: Cournot
.
Paris
:
Flammarion
.
Mill, John Stuart.
1844
. “
Of the Laws of Interchange between Nations and the Distribution of the Gains of Commerce among the Countries of the Commercial World
.” In
Essays on Some Unsettled Questions of Political Economy
.
London
:
John W. Parker
.
Mill, John Stuart. (
1848
) 1965.
Principles of Political Economy with Some of Their Applications to Social Philosophy
. Vols.
2–3
of Collected Works of John Stuart Mill, edited by J. M. Robson.
Toronto
:
University of Toronto Press
.
Mill, John Stuart. (
1852
) 1965.
Principles of Political Economy with Some of Their Applications to Social Philosophy
. Vols.
2–3
of Collected Works of John Stuart Mill, edited by J. M. Robson.
Toronto
:
University of Toronto Press
.
Mill, John Stuart.
1854
.
Principes d’économie politique avec quelques-unes de leurs applications à l’économie sociale
. Translated by Hippolyte Dussard and Jean-Gustave Courcelle Seneuil.
Paris
:
Guillaumin
.
Negishi, Takashi.
2014
.
Developments of International Trade Theory
. 2nd enhanced ed.
Tokyo
:
Springer
.
Numa, Guy.
2019
. “
Jean-Baptiste Say on Free Trade
.”
History of Political Economy
51
, no.
5
:
901
34
.
Pareto, Vilfredo.
1892
. “
Di un errore del Cournot nel trattare colla matematica
.”
Giornale degli economisti
, 2nd ser., 3rd year,
4
:
1
14
.
Ricardo, David. (
1817
) 1951.
Principles of Political Economy
. Vol.
1
of The Works and Correspondence of David Ricardo, edited by P. Sraffa.
Cambridge
:
Cambridge University Press
.
Say, Jean-Baptiste. (
1828
–29) 2010.
Cours complet d’économie politique pratique
. Variorum ed. Edited by E. Blanc, P.-H. Goutte, and J.-P. Potier. Vol.
2
of Œuvres complètes de Jean-Baptiste Say.
Paris
:
Economica
.
Solow, Robert.
2008
. “
Cournot and the Social Income
.” In
Augustin Cournot: Modelling Economics
, edited by Jean-Philippe Touffut,
106
15
.
Cheltenham
:
Edward Elgar
.
Smith, Adam. (
1762
–63). 1978.
Lectures on Jurisprudence: Report of 1762–3
. Edited by R. L. Meek, D. D. Raphael, and P. G. Stein. In vol.
5
of The Glasgow Edition of the Works and Correspondence of Adam Smith.
Oxford
:
Clarendon Press
.
Smith, Adam. (
1766
) 1976.
Lectures on Jurisprudence: Report Dated 1766
. Edited by R. L. Meek, D. D. Raphael, and P. G. Stein. In vol.
5
of The Glasgow Edition of the Works and Correspondence of Adam Smith.
Oxford
:
Clarendon Press
.
Smith, Adam. (
1776
) 1976.
An Inquiry into the Nature and Causes of the Wealth of Nations
. Edited by R. H. Campbell and A. S. Skinner; textual editor, W. B. Todd. 2 vols. Vol.
2
of The Glasgow Edition of the Works and Correspondence of Adam Smith.
Oxford
:
Clarendon Press
.
Stiglitz, Joseph, Amartya Sen, and Jean Paul Fitoussi.
2010
.
Mismeasuring Our Lives: Why GDP Doesn't Add Up
.
New York
:
New Press
.
Theocharis, Reghinos D.
1983
.
Early Developments in Mathematical Economics
. 2nd ed.
Philadelphia
:
Porcupine Press
.
Torrens, Robert.
1821
.
An Essay on the Production of Wealth
.
London
:
Longman, Hurst, Rees, Orme and Brown
.
Vatin, François.
1998
.
Économie politique et économie naturelle chez Antoine-Augustin Cournot
.
Paris
:
Presses Universitaires de France
.
Viner, Jacob.
1937
.
Studies in the Theory of International Trade.
London
:
Allen & Unwin
.
Walras Léon. (
1863
) 1987. Review of
Principes de la théorie des richesses
, by Augustin Cournot. In Mélanges d’économie politique et sociale,
155
59
. Vol.
7
of Les œuvres économiques completes [of Auguste Walras and Léon Walras], edited by Jan van Daal and Donald Walker.
Paris
:
Economica
.
Whewell, William. (
1850
) 1856. “
Mathematical Exposition of Some Doctrines of Political Economy
.” Second memoir.
Transactions of the Cambridge Philosophical Society
9
:
128
49
.