The Principles
When, in 1890, the first edition of the Principles of Economics appeared, the ground had already been prepared to ensure the book had a major impact on the economic culture of the time. Marshall was then settled in the Cambridge chair, and his pupils occupied important positions in the English academic world (the same year saw the birth of the Royal Economic Society and the Economic Journal). The influence of the classical tradition was still strong, together with that of the historical school, while the marginalist heterodoxy attracted some brilliant minds but was still a minority view. In such a situation, Marshall offered a set of elements designed to attract the convergent interest of the different streams of economic culture existing at the time: insistent reference to the classical tradition, from the Smithian theory of the division of labour to the Ricardian theory of rent; acceptance of the basic elements of the margin- alist revolution, such as the notion of marginal utility, with attribution of a central role to demand, and hence to economic agents’ preferences, within a theory of value in which prices were determined by the mechanism of equilibrium between supply and demand, already adopted by a string of post-Ricardian authors;2 insertion of this analytical structure in the context of extended discussions on the meaning of the concepts
- 5 per cent). The idea of human sciences as concerning arguments to be deduced from ‘human nature’ was replaced by that of statistical laws about what is ‘normal’. In this sense, the view of economic science as a theory of the behaviour of the rational agent (or homo oeconomicus) falls within the old view of human sciences; by substituting the term ‘natural’ with the term ‘normal’, Marshall was trying to escape such a view.
- 2 More precisely, for each commodity the normal price is determined by the point where two curves meet, graphically representing the demand and supply functions. These respectively connect the supply price (cost plus normal profit) and the demand price (the maximum price which the purchaser is ready to pay) to the quantity of the commodity under consideration.
used in the analysis and on the historical evolution of society; references to Darwinian evolutionism, which conferred an element of scientific modernity on the work and provided a flexible, open response to historical evolution as compared to the reference to physics prevailing in the theories of authors such as Jevons or Walras, which had not met with a warm reception.
From the first (1890) to the eighth (1920) edition, the Principles remained at the centre of Marshall’s theoretical work, undergoing substantive revisions; this is especially true for the fifth edition (1907), the last before his resignation from the Cambridge chair. On the evidence of the voluminous variorum editio (1961), promoted by the Royal Economic Society, we can reconstruct this path. The importance of Marshall’s revisions testifies to the difficulties he met in his work of synthesis between different approaches and in his attempt to build a theory of value that was to include simultaneously the objective (cost of production) and the subjective (utility) element and that was to be at the same time rigorous, realistic and open to historical evolution. Before discussing the difficulties Marshall came up against, it may be useful to run over the main aspects of his approach: method, the notions of equilibrium and competition, the concepts of the firm and the industry. We will then consider the problem of increasing returns and the two solutions suggested by Marshall, the representative firm and external-internal economies.
Marshall’s methodological standpoint was simple: to recognise the extreme complexity of the real world. Theory cannot but be abstract but must keep its feet on the ground. Hence his tenet, which underlay his partial equilibrium method, that short causal chains should be favoured. At each step, theory proceeds by isolating a logical nexus of cause and effect held to be the main one and thus leaves aside other effects held to be secondary, though not non-existent. This is licit, indeed necessary, for construction of each individual analytical piece. However, when we put together many logical nexus and give rise to long causal chains - as happens for instance in general economic equilibrium theory - the secondary effects left aside may have repercussions that amplify step by step, and this may cause the conclusions drawn from the theoretical analysis to be misleading. Hence Marshall relegated to a mathematical note, in an appendix to his Principles, his illustration of general economic equilibrium (one of the most rigorous of the time). In the text, on the other hand, he preferred to focus on short causal chains, and in particular on the method of partial equilibriums. The latter consisted of considering demand and supply of each good as independent of what simultaneously happens on other markets for the other goods.
The same awareness of the complexities of the real world may also be perceived in the attention Marshall paid to the construction of the system of concepts by which to represent reality. In the first books of the Principles, step by step the concepts introduced are discussed, illustrating for each the shades of meaning that rendered their contours imprecise. This is true in particular for the key notions of equilibrium and competition to which, in the intertwining of affirmations and qualifications, it is very difficult to attribute a univocal meaning. Marshall’s position oscillated between two terms of reference: on the one hand, the notions subsequently transmitted through the textbooks, the so-called Marshallian vulgata; on the other, the esoteric notions, disseminated among the circle of pupils, connected to an evolutionary view that drew more on Lamarck than on Darwin.[1] In the first case - the Marshallian vulgata - the notion of equilibrium corresponded to the static notion of equality between demand and supply and the notion of perfect competition to the presence of a large number of firms in each industry, so large as to render the size of each firm irrelevant to the dimensions of the industry as a whole and the choices of each individual firm irrelevant for the industry as a whole. In the second case - the evolutionary view - the notion of equilibrium took on dynamic features in the attempt to take account of the irreversibility characterising the actual movements of the firm and the industry along demand and supply curves;[2] the notion of competition was softened by attributing to each firm some room for manoeuvre that included, among other things, the possibility of violating the so-called law of the one price. Theoretical analysis - construction of well-structured models - inevitably led to refer to clear-cut concepts of the first kind, while in the case of the evolutionary view, as we shall see below, he remained in the realm of metaphors, which are evocative but certainly not rigorous. In other words, in the oscillation from the first to the second pole of the Marshallian construct, what was gained in terms of realism was lost in terms of analytical rigour.
The very notions of industry and firm constituted a bridge between the complexity of the real world and the simplicity requirement of abstract theory. Marshall thus distanced himself from the extreme methodological individualism of the first marginalist theoreticians, favouring a classical feature, by which each commodity (‘good’, in the subjectivist terminology, which thus lays stress on their utility to the consumer) corresponds to a category that includes objects not identical amongst themselves but sufficiently similar to warrant unitary treatment, and in parallel each industry includes the firms (complex productive units) that operate in one of such commodity categories. Quite naturally, various problems arise when the categories thus defined are related to the real world: from the case of joint production to the problem of differences in the technologies adopted by different firms belonging to the same industry and on to the problem of greater or lesser similarity between the products of different firms belonging to the same industry. The latter aspect in particular renders less clear-cut, more flexible, the Marshallian notion of competition, and the notion was accordingly bent so as to allow for some degree of independence between the markets of the different firms belonging to an industry, hence some degree of autonomy in the price choices of the different firms.
Within this conceptual framework Marshall’s analytical structure was based on (short- or long-period) equilibrium between demand and supply. The demand function for each commodity is assumed to be derived from individual preferences;[3] however, Marshall tended to skate over the relationship between utility maps and demand functions; for the purpose of determining equilibrium, it is sufficient to assume as given (and decreasing, on the basis of the decreasing marginal utility postulate) the demand functions for the different goods.[4] Attention was, rather, focused on supply functions: it was in this field that Marshall tried to provide an innovative contribution in comparison to the theories proposed by the first protagonists of the marginalist revolution, and in particular by Jevons. The latter had recourse to a principle symmetrical to that of decreasing utility, the principle of the increasing sacrifice or painfulness of labour; this allowed him to obtain increasing supply curves. Such an approach, however, cannot easily be extended from study of the behaviour of individuals to analysis of industries and firms in competitive markets, and it is even harder if we stick to the method of partial analysis: each firm or industry considered in isolation can easily obtain additional hours of labour by subtracting them from other firms or industries without changing the marginal disutility of labour for the individual worker.[5]
Marshall thus proposed the road of partial equilibrium for supply side analysis and hence to construct supply curves referring to individual firms and industries. To this end he took two elements of the classical tradition and reworked them in a context different from the original one. The first was the Smithian theory of the connection between enlargement of the market and division of labour and consequently productivity increases. The second was the Ricardian theory of differential rent. Newly christened as laws of returns to scale, these two theories were simultaneously used to explain the variations of costs in response to changes in the quantity produced, respectively identified with the case of increasing returns to scale and with the case of decreasing returns. Clearly this is an artificial construct, which puts together quite different things. Furthermore, even if considered one at a time, the transposition of the Smithian and Ricardian ideas into the ambit of the theory of the firm and the industry gave rise to difficulties that Marshall saw or perceived but to which he did not attribute the importance they deserve.
Let us consider decreasing returns. Ricardo used them in the theory of rent with reference to the productivity of a means of production of a particular kind, such as land, taken to be available in a given quantity and with distinctive features for each unit of land. The Ricardian theory of differential rent did not revolve around the different productivities of the lands in the country, and so around decreasing returns for individual firms or industries, but around the problem of the distribution of national income among the social classes of workers, landlords and capitalists and in particular around the problem of determining the rent accruing to the landlords. In Marshall’s modified form, on the other hand, the theory of decreasing returns concerned the means of production utilised by specific industries. The case in which an industry is the sole subject to use a given means of production is however a very peculiar one. Once again, outside this case, the ceteris paribus clause (hence the method of partial analysis) should be abandoned.
In his earliest writings, moreover, Marshall was considering a further problem: the existence - and importance - of increasing returns to scale, which are seen as the source of economic development in the Smithian theory. The theory of the equilibrium of the firm falls apart if the assumption of decreasing returns is abandoned in favour of that, decidedly more realistic, of increasing returns to scale. A stable equilibrium is possible in this case only if the demand curve decreases more rapidly than the supply curve; but this cannot hold in the case of competition, where the price is by assumption independent of the quantity produced by the individual firm. In other words, the assumption of perfect competition is incompatible with the case of increasing returns to scale.
As already stated, Marshall had already recognised the existence of this dilemma in his essays published in 1879; much of his analytical effort in the Principles and in subsequent revisions of the book was devoted to solving the dilemma, but the suggested solutions went in different directions.
First, we have the solution later developed by Pigou and Viner, and then adopted in most textbooks, based on the assumption of U-shaped average and marginal cost curves: initially, when production increases, cost curves are decreasing because increasing returns prevail; from a certain point onward, decreasing returns take the lead, and costs start increasing. Under competition and in the long run, the equilibrium point for the firm corresponds to the minimum of the average cost curve. The development of the industry then depends on the economies of scale internal to the industry itself (since the demand curve is decreasing for the industry, so that an equilibrium is possible even if the supply curve is also decreasing, provided that the speed of decrease is lower than that of demand) but external to the individual firms that compose the industry (so as to retain for them the possibility of a competitive equilibrium that needs cost to increase with the quantity produced, from a certain point onwards). Such a construct may thus be criticised both for its lack of realism and for its connection to a static notion of equilibrium.
Marshall, while hinting at the line of reasoning outlined previously, suggested, rather, a second path, consisting in the theory of the representative firm and recourse to biological metaphors. The core of the argument is this: the industry is made up of many firms which, like trees in a forest, are at different points of their life cycle: some, the young ones, enjoy increasing returns and develop although in a competitive environment; others, the mature ones, have already reached dimensions at which the elements of growth and decay balance out; yet others are decaying. In a world composed of individual firms distributed among the different stages of development, the representative firm, of average dimensions, turns out to be midway in its development process and can thus be identified with a firm experiencing increasing returns, even if overall the population of firms is stationary.
The weakness in this construct is not simply the difficulty of translating it into a well-structured analytical model; rather, it lies in the difficulty of accepting the assumption of the life cycle of firms. Marshall referred to the sequence of three generations in control of the firm: the founder, endowed with above-normal organisational and innovative ability; his immediate heirs, brought up in the hard school of the founder and used at least to rigorous management of the family business; and then the third generation, brought up in prosperous conditions and less ready to make the sacrifices that are often necessary in a competitive environment characterised by continuous technological change and hence by the need to save and invest.[6] Such an assumption refers to a world of small firms managed by their proprietor but does not apply to public companies. In the last editions of the Principles, Marshall appeared aware of these difficulties. Various among his followers nonetheless remained faithful to the construct of the representative firm, including Robertson, who re-proposed it in 1930, provoking Sraffa’s (1930) sarcastic reaction.
On this count at least, the Principles constituted a failure. However, various other elements of Marshall’s edifice are fully entitled to remain part of modern economic theory; we may recall, for instance, the notion of elasticity. And it should be added that Marshall’s greatness as an economist lies also (and perhaps mainly) in his awareness of the limits of his analytical constructs, which have however been accepted without critical scrutiny by many of his followers.
- [1] Jean-Baptiste de Lamarck’s (1744-1829) thesis, rejected by Darwin, was the heredity ofthe characteristics acquired in life by an organism as response-adaptation to the environment in which it lives. Darwin’s thesis was, of course, that the characteristics best adaptedto existence (and above all to reproduction) in the end prevail because of a process ofnatural selection. Lamark’s theses had been re-proposed, confounded with Darwin’sevolutionism, as a tool for the analysis of society by the sociologist Herbert Spencer(1820-1903), very influential at the time.
- [2] Marshall derived the ‘evolutionary’ notion of equilibrium from the theory of population,which can tend to a stationary age structure ofthe population through constancy over timeof birth and death rates, even in the presence of a growing or declining population size.
- [3] In the context of partial analysis, the marginal utility of money was assumed to beconstant.
- [4] Recalling John Stuart Mill, in this respect Marshall stressed the need to develop a newscience, ethology or the study of human habits and customs and their gradual changes inthe course of time.
- [5] On the other hand - we may now add on the basis of Sraffa’s 1925 remarks - if we werealso to take account of infinitesimal changes in the painfulness of labour, stemming fromchanges in production levels of a single firm or industry, such changes would equally affectall industries and firms in the economy. As a consequence, it would not be possible to usethe ceteris paribus clause which is the basis of partial analysis; in particular, when faced withgeneralised changes in prices it would not be possible to assume as given the demandcurve for the individual industry.
- [6] In that period (eleven years after the first edition of Marshall’s Principles but six yearsbefore the fifth edition, in which the idea of the representative firm reached fulldevelopment) the idea of the life cycle of firms found literary expression in the famousnovel by Thomas Mann, Buddenbrooks (1901).
