The revolution of marginalism: how political economy became economics
Adam Smith’s definition of political economy was an ‘inquiry’ about the causes of the ‘wealth of a nation’ (Smith, 1776). In 1844, John Stuart Mill claimed that ‘political economy was a political science which studies the production and distribution of wealth’ (Mill 1844). In 1890, Alfred Marshall defined economy as the ‘study of mankind in the ordinary business of life; it examines that part of individual and social actions which is closely connected with the attainment and with the use of material requisites of well-being’ (Marshall 1890, 1).
In 1900, Pareto defined pure economics as a science for determining a state of equilibrium:
[pure economics] arrives at identifying the choices that are available to individuals, by taking into account the obstacles the latter encounter and by hypothesizing that not only can such choice be completely ordered, but they are also identical if they only differ because of the order of consumption, since the faculty to enjoy the goods is more important than the way they are enjoyed.
(Pareto  1982, 388)
In 1932, as previously already reported, Robbins defined economics as ‘the science which studies human behaviour as a relationship between given ends and scarce means which have alternative uses’ (Robbins 1932, 16).
The path from marginalism to Robbins has been the track for the rise of neoclassical economic theory as mainstream economics; and the old classical liberalism has been gradually transformed into a new liberalism.
Neoclassical economics can be dated back to the work of the founding fathers of marginalism (Menger 1871; Jevons 1871; Walras 1874), and their immediate followers (Edgeworth 1881; Marshall 1890; Pareto 1906).3
In 1871, Jevons’ Political Economy and Menger’s Grundsatze appeared without any reciprocal influence (Jevons 1871; Menger 1871). Three years later, Walras published his Elements (Walras 1874). Historians of economic thought consider these three books as the pillars of the so-called marginalistic revolution, or marginalism. Marginalism has been rightly considered the starting point of the transformation of political economy from a branch of moral philosophy (as it was regarded from Adam Smith to Marx and John Stuart Mill) to economics as a science (Milonakis and Fine 2009).
Both positivism and neopositivism deeply influenced the transformation of the discipline (Caldwell 1980; Weintraub 1985; Mirowski 2002).4 Later in the twentieth century, during the interwar period, the specific scientific feature of economics as a science became much more related to physics. Along with the development of this process, ethics was discarded from economics, as it was supposed to be alienated from any other social science (Sen 1987; Putnam 2002).
The main differences between classic political economy and marginalism can be listed as follows:
- 1 Methodological individualism: in marginalism, economic agents are individuals (either consumers or firms) and no longer social classes (rentiers, workers, capitalists) like in Smith’s, Ricardo’s and Marx’s theory.
- 2 Strict subjectivism: the marginalistic theory of value is based on individual utility, and no more on the cost of production (Smith) or on the sole labour (Ricardo-Marx).5
- 3 In marginalism, the final aim of economic theory is general economic equilibrium, and no longer an inquiry of the wealth of a nation (mainly based on the introduction of the division of labour, or alternatively, in a Marxian perspective, on the alienation of workers’ labour and on the following collapse of the rate of profit in the latest phase of capitalism).
- 4 The principle of maximization of economic agents’ expected utility function: in marginalism, general economic equilibrium is made possible when economic agents are featured as rational agents able to rank their preferences and to make an optimal choice under a condition of scarcity of available means.
- 5 In marginalism, market price is seen as the only possible mechanism able to get a final general economic equilibrium;6 in the classical school, market is still regarded as a historical institution where if exchange and trade are free, the national wealth is bound to increase.
In the 1890s a second phase of marginalism occurred. Alfred Marshall in Cambridge and Vilfredo Pareto, the successor of Walras in Lausanne, were the main representatives and most influential economists of that period.
Around 1900, the father of American institutionalism, Thorstein Veblen, coined the term neoclassical in order to define Marshall’s economy theory as ‘the best work that is being done under the guidance of the classical antecedents’ (Veblen 1919, 171.)
In fact in his Principles Marshall was able to bridge the gap between the classical school of political economy and the recent development (marginalism) within the discipline. Marshall made possible the shift of subjectivism from the demand side (the theory of consumer, upon which marginalism was grounded) to the supply side (upon which classical economic theory was rooted). In fact, Marshall introduced the concept of representative firm as an individual agent able to minimize inputs’ costs in order to maximize profits. Furthermore, Marshall introduced the notion of price as the equilibrium point between demand (as in marginalism’s theory of utility) and supply (classical theory of cost of production).
Pareto made a step forward to the notion of economics as a science. The Italian and Lausanne-based economist introduced the notion of pure economics, centered on the description of logical actions of individuals, who are able to rank their preferences, and to follow a rational way of behaving in a contest that will be heading to general economic equilibrium. The substitution of the subjective approach with classical objectivism made it possible to analyze economics and the market through an individualistic approach, based on the idea that individuals’ behaviour is reasonably oriented to satisfy their needs.7
Marginalistic economic theory had been strongly influenced by British utilitarianism: this is evident not only in Jevons’ notion of utility, who directly quoted from Bentham,8 but also later in Edgeworth’s ideas, that Pareto would redefine as the set of efficient allocation occurring during an economic exchange.
If Marshall was able to build a strong continuity between classic political economy and marginalistic economics, the notion of economics as a science that starts from individuals’ preferences had been grounded in Pareto’s toolbox.
Charles Gide was the first scholar who became aware of the fact that along with the transformation of political economy into a scientific discipline, a new doctrine of liberalism was occurring, that was much more suitable for the emergent definition of economy. The French philosopher specifically used the term ‘new liberal school’ to address Maffeo Pantaleoni.
According to Gide, Pantaleoni was one of those ‘mathematical economists’:
adepts of ‘Neo-liberalism’ (along with Warlas and Pareto) who apply economic theory to an ‘hedonistic world’ (...) in which each contracting party will weigh in a subjective balance, infallibly exact, the final utility of the object to be disposed of and of the object to be acquired (...), a world where the law of supply and demand will bring about the maximum of utility for both individual and society, and will always send back the barometric needle, at once and without friction, to ‘set fair’ - I mean to the fair price.
(Gide 1898, 494-495)9
The role of Pareto in the building of economics as a science has been crucial. In Pareto’s Cours (1896) and Manuale (1906), ‘pure economy’ is the only scientific form of classical political economy, and it has to be described in mathematical terms even though in a different way than Walras had done. The French economist had used a deductive approach, based on the internal coherence of the logical procedure without any concern about the realism of the initial hypotheses. Following a different path, Pareto applied the method of natural sciences (experimentalism) to economic theory: by trial and error, he found that it is possible to build an economic theory free from any metaphysical residuals. The experimental approach is also be concerned about the realism of the initial hypotheses. In his Cours, economic theory is like any other natural science: the application of mechanical physics allows for a gradual entrance into the formulation of general economic equilibrium. The use of mathematics is possible, and somehow necessary, because human actions follow some regular patterns ‘qui constituent des lois naturelles’ (Pareto 1896, 397).
In the incipit of his Cours, Pareto firmly claimed that, in his book, no solution for any concrete problem will be found, and his definition of economics as a science will involve the notion of ophelimite as the measure of pure economic satisfaction.10
After the publication of his first book, Pareto’s urgency to find the objective side within economic theory led him to introduce a ‘pure experimental method’ as well as to discard the use of non-empirical categories like utility, ophelimite, value, hedonism.11 In his later Manuale (Pareto 1906), the principle of maximization of the utility function is no longer used to describe general economic equilibrium, rather that is reached starting from the application of a neutral tool: the indifference curves’ set and - as a consequence - the application of the ordinal ranking of preferences.
On March 1900, in a paper published in Giornale degli Economisti, Pareto claimed that indifference curves represent human choice when this is directly affected by experience. Hence, it is necessary to start from preferences to get general economic equilibrium. Unlike ophelimite, whose measurement presents many difficulties, indifference curves represent ordinal preferences and they allow a perfect combination of deductive and inductive methods, which is the only scientific way to formulate general laws beginning from concrete facts (Pareto 1900, 81). In that article, Pareto defined pure economy as an experimental science (Pareto 1900, 91), which uses induction to cope with concrete phenomena as well as deduction to deal with the logical relation between premises and consequences. The use of mathematics is a part of the deductive process. A third momentum is required, according to Pareto, in fact deductions have to be compared with real facts either to reveal eventual discrepancies or to confirm them.
In his Manuale, Pareto complained about the fact that a definitive and firm distinction between science and practice in economics had not happened yet. Pareto’s aim was to consider economics like a natural science: both are based on abstraction of regularities from concrete phenomena; upon those regularities, general laws can be formulated. Even though concrete phenomena (either natural or social) cannot be wholly known, general knowledge can be reached using approximation (Pareto 1906).
The well-known Pareto’s distinction between logical and not-logical actions represents the starting point of the introduction of general economic equilibrium: individuals’ logical actions, massively and regularly repeated, to get material satisfactions can be expressed in mathematical terms. They allow the building of pure economics, which has no practical aim.12
While writing his Manuale, Pareto was involved in a debate on the nature of economics along with the Italian philosopher Benedetto Croce (Croce 1900, 1901; Pareto 1901, 1902). Their well-known querelle, which was hosted in Giornale degli Economisti between 1900 and 1902, allows a description of the transformation of economics into a formal science in a broader cultural sense.
As Kirzner pointed out:
Croce was aware of a new general element in his experience of human affairs. This element was not moral, it was not technical, nor did it coincide with any other already-named abstractions. In his formulations, Croce made a vigorous attempt to present this abstraction to the attention of the world by ascribing to it the word ‘economic’ (....) If in the effort to provide an adequate definition of economics, an attempt is made to analyse the concept of economy, for example, one necessarily becomes involved in a problem of economic science itself [emphasis added].
(Kirzner 1976, 11)
Croce had started the debate. The Italian philosopher, influenced by historical Hegelism, and supporter of a conception of economics as a ‘practical form of the Spirit’ that deals with ‘useful things’, was well aware of the revolutionary feature of the advent of marginalism within economics. He considered Pareto as one of the most important economists of this new doctrine, able to show that ‘economics has nothing to do with history and with practical issues: it is a science following its own principle, the economic principle’ (Croce 1900, 16). In his first paper, Croce rhetorically asked Pareto what was the nature of this economic principle: Croce’s main objection was that the mechanical definition (provided by Pareto) of the economic principle (able to make measurability possible) is problematic and obscure. Croce insisted on the fact that economic choice is a practical activity and it depends on human will. According to the philosopher, economics cannot follow mechanical laws. Furthermore, Pareto did not clearly explained what logical actions really are. Following Hegel, Croce claimed that logics belongs to a theoretical realm, while economic phenomena are practical activities, and they may be related to ethics (Croce 1900).
Pareto’s reply to Croce was centred on the idea that there is a sort of uniformity within economic phenomena. This uniformity allows a definition of economic laws, which can be expressed in mathematical terms. The same process happens within natural sciences: the use of mathematics allows an objective description of relations amongst quantities, and it has nothing to do with relations amongst mental concepts. Pareto admitted that something else can be present in economic phenomena (Pareto 1901), but this does not change the scientific nature of economics (and other sciences). In fact, any science deals with functional relations, not causal relations, which occur amongst phenomena.
In his following reply, Croce pointed out that Pareto was still confusing the nature of a science with the way to explain scientific results: the formal representation of economic phenomena is not science, because it does not explain the nature of economic relations and phenomena (Croce 1901). Croce’s critique went even beyond this point. Almost in a paradoxical way, the philosopher accused the economist of being metaphysical in his reduction of human facts (as economic phenomena are) to physical facts, and he added that, even in physics, facts can only be described, but never wholly understood (Croce 1901).
In Pareto’s following and last reply, which concluded their debate, the economist rejected Croce’s accusation that he was metaphysical in his definition of economics as a science. ‘I am the most nominalist amongst nominalists’ Pareto claimed (Pareto 1902, 131); and he added that there were no differences between the regularities amongst physical facts and human phenomena: scientists’ aim was to find them out and to describe how they work. Mathematical description was the most suitable and simple way to do it.13
Pareto purged economics from any metaphysical residual as well as he purged economic agents from any psychological influence, by coping with preferences in an axiomatic way. Later during the interwar period, mathematical economists developed this approach (see below). Many other non-mathematical economists expounded upon it in a non-formalistic way: Robbins shared Pareto’s opposition to a psychological subjective theory of value, as well as Pareto’s attitude of considering economics as the science of rational behaviour under certain circumstances (Bruni and Sudgen 2007).
Lionel Robbins’ contribution between the late 1920s and early 1930s is important in order to describe the transformation of the discipline which today we recognize as neoclassical economics.
Robbins’ main concern was that infinite disputes amongst economists have deprived economics of a proper definition about its nature and its significance. In 1929, Robbins wrote in his diary that a proper definition of economics was still unsettled, and he pointed out the wide variety of available definitions of economics. According to him, there was a ‘too vague’ sociological definition, based on exchange and money measurement (Pigou, Mises, and Schumpeter); also, there was a ‘too narrow’ definition of economics, as the study of material welfare (Cannan); finally, there was a ‘misleading and confusing’ definition of economics, focused on the notion of scarcity and the operation of economizing (Cassel and Schumpeter). Robbins also added that the use of some words, such as ‘welfare’, is dangerous because of its ethical implications (Howson 2011, 144).
In his Essay, Robbins’ aim was twofold: ‘to arrive at a precise notion concerning the subject-matter of Economic Science and the nature of the generalizations of which Economic Science consists. Second, it attempts to explain the limitations and the significance of these generalizations’ (Robbins 1932, vii). As it is well known, Robbins’ formula in defining economics was exposed in the first chapter of his book: ‘economics is the science which studies human behaviour as a relationship between ends and scarce means which have alternative uses’ (Robbins, 1932, 16).
In this definition, Robbins embedded Menger’s and Mises’ individualism along with Pareto’s technicalities. In his general definition of economics, Robbins was able to merge the formal aspect of the discipline with its implicit subjectivism. Furthermore, he defined economics unequivocally as a science able to describe and to forecast a special behaviour, the behaviour imposed by a situation of scarcity of means to achieve given ends.
The scientific character of economics has been especially underlined by Robbins’ definition of economics’ neutrality about ends. In his words: ‘economics is entirely neutral between ends’ (Robbins 1932, 24).14
The use of purely formal conceptions within economics was fundamental when Robbins defined the nature of an economic good: ‘Whether a particular thing or a particular service is an economic good depends entirely on its relation to valuations. Thus wealth is not wealth because of its substantial qualities. Wealth is wealth because it is scarce’ (Robbins 1932, 47).
Given the formal nature of the relationship between ends and means, Robbins described the economic laws upon which economic science rests: price theory as well as the law of diminishing marginal utility directly derive from the definition of economics as the disposal of scarce goods with alternative uses.
The pure theory of equilibrium is the final achievement of economics:
[it] enables us to understand how, given the valuations of the various economic subjects and the facts of the legal and technical environment, a system of relationships may be regarded as tending. It enables us to describe that distribution of resources which, given the valuations of the individual concerned, satisfies demand most fully. But it does not by itself provide any ethical sanctions (...) There is no penumbra of approbation round the Theory of Equilibrium, equilibrium is just equilibrium [emphasis added].
(Robbins, 1932, 127)
At the end of his book, Robbins went back on the necessity to free economics from any ethical issue. Likely pushed by what has recently been called emotivism amongst economists (De Martino and McCloskey 2016), Robbins almost obsessively stressed the urgency to fix the boundaries of economics within the assumption of human rationality, and to let ethics remain strictly outside them. Robbins wrote:
There is nothing in Economics which relieves us of the obligation to choose. There is nothing in any kind of science, which can decide the ultimate problem of preference. But to be rational, we must know what it is we prefer. We must be aware of the objective implications of the alternatives of choice. For rationality in choice is nothing more and nothing less than choice with complete awareness of the alternatives rejected.
(Robbins 1932, 136)
Deeply influenced by Mises, Robbins was able to give a precise definition of positive economics, which was based on the subjectivism of individuals’ choice in a context of objective scarcity, without any ethical implication. Robbins’ postulates were the existence of an order of preferences, the existence of more than one factor of production, and the uncertainty about future conditions of scarcity. Even though in his postulates Robbins did not feature ‘the fundamental assumption of economics, namely, the postulate of rational behaviour’ (Giocoli 2003, 86), and his book is much more complex than a simple manifesto of mainstream economics (Masini 2009), Robbins built up the most persuasive compromise for a scientific definition of economics.
Robbins’ book was one of the most influential methodological works in twentieth century economics, and its influence needs to be understood within the context of the 1930s development of marginalism (Wade Hands 2009). This was especially true after the enthusiastic acceptance and endorsement Robbins’ work got from Stigler (Stigler 1942) and Samuelson (Samuelson 1948). In addition, Koopmans (Koopmans 1957) explicitly linked the axiomatic method with the Robbins definition of economics (Backhouse and Medema 2009b). Furthermore, in 1962, Friedman implicitly invoked Robbins when he claimed that: ‘an economic problem exists whenever scarce means are used to satisfy alternative ends’ (Friedman 1962a, 6). Robbins got a similar endorsement from Becker, whose definition of economics was ‘the study of the allocation of scarce means to satisfy competing ends’ (Becker 1971, 1).
From Becker onward, the use of Robbins’ definition of economics (Backhouse and Medema 2009b, 2009c) had a huge influence even outside the discipline: it narrowed economics to the theory of constrained maximization or rational choice, and justified the imperialism of economics among the other social sciences (Boulding 1969; Tullock 1972; Cartwright 1999; Lazear 2000; Backhouse and Medema 2009a).