CHAPTER 6

Alfred Marshall and the Framework of
Neo-Classical Economics

Among the Anglo-Saxon neo-classical pioneers, Alfred Marshall was a giant without rival. An examination of his analysis - despite noteworthy distinctive features - is appropriate for the purposes of setting out the central properties of neo-classicism. Though his thought was organized around a tight theoretical core, he chose to present it in a deceptively simple style. He held economics to be the study of man 'in the ordinary business of life' and that its findings should be accessible to a wide public audience. This attitude largely explains the consignment of his subtler contributions to theory to the obscurity of footnotes and appendices. Unlike most of his neo-classical contemporaries he maintained that mathematical expositions, though invaluable aids to the economist in the clarification of his own thought, were unnecessary to - and might even hinder - the communication of his findings.1

1. ALFRED MARSHALL (1842-1924)

Throughout his adult life, Marshall occupied academic positions. Apart from four years as the Principal of the University College of Bristol and a brief period as a Fellow of Balliol College, Oxford (where he taught political economy to candidates for the Indian Civil Service), he was associated with the University of Cambridge. From the chair of Political Economy (to which he was elected in 1885) he exercised a formidable influence on one of the most fertile generations of students in modern history. His inspiration and stimulus were instrumental in bringing the Cambridge economics school to a position of eminence.

The main corpus of Marshall's teaching was contained in one book - Principles of Economics. Published originally in 1890, this work went through eight editions during his lifetime. If not a prolific writer, he was an infinitely painstaking one. Marshall - to the considerable annoyance of his ablest pupil - was reluctant to commit his insights to print before they had been fully polished and before their relevance to practical problems had been established. On this trait of his master, John Maynard Keynes later commented:

Economists must leave to Adam Smith alone the glory of the Quarto, must pluck the day, fling pamphlets into the wind, write always sub specie temporis, and achieve immortality by accident, if at all.

Moreover, did not Marshall, by keeping his wisdom at home until he could produce it fully clothed, mistake, perhaps, the true nature of his own special gift? 'Economics,' he said ... 'is not a body of concrete truth, but an engine for the discovery of concrete truth.' This engine, as we employ it today, is largely Marshall's creation. He put it in the hands of his pupils long before he offered it to the world. . . . Yet he hankered greatly after the 'concrete' truth which he had disclaimed and for the discovery of which he was not specially qualified.2

Whatever the costs of delayed publication may have been, Marshall's interest in and affection for his students later paid handsome dividends. His influence extended well beyond his role in equipping them with professional skills. He urged them to be engaged with the world's problems and advised a wary attitude towards popular acclaim. 'Evil', he once wrote, 'is with [students of social science] when all men speak well of them.... It is almost impossible for a student to be a true patriot and to have the reputation of being one in his own time.'3

Though Marshall's place in history rests primarily on his contributions to economic theory, he insisted that the purpose of theoretical investigations was to illuminate practical problems. In his own career his social concerns found direct outlet in his participation (either as a member or as an expert witness) in the work of official commissions dealing with monetary questions, taxation, and the relief of the poor.

2. THE APPROACH TO THE ANALYSIS OF PRICE

For Marshall - as well as for other contributors to the formulation of neo-classical thought - analysis of the functioning of a market system began with the behaviour of consumers and producers. Throughout the discussion it was assumed that men acted rationally in pursuit of their own advantage. Consumers were held to seek maximum satisfactions; similarly, suppliers of productive services were expected to seek maximum rewards. It was not maintained, however, that economic motives were the only spurs to human action, nor that all men acted as homo economicus in the conduct of the day-to-day business of life. Most neo-classical writers - and Marshall with particular emphasis - insisted that their study was restricted to the economic aspects of human action, rather than the whole complex of man's aspirations. By the same token they did not wish to be interpreted as saying that all who participated in market transactions were rational calculators. Instead, they sought merely to establish that rationality as a behavioural postulate provided a realistic basis for the study of groups of people.

This mode of reasoning can be readily observed in Marshall's formulation of the concept of demand. As he interpreted it, 'demand' referred to the relationship between quantities demanded and prices. Normally it could be expected that buyers would be prepared to purchase more of a particular good at lower prices than at higher ones. For each good a whole range of price and quantity combinations was conceivable. These combinations could be depicted diagrammatically as a schedule (or curve) by representing price on a vertical axis and quantity on a horizontal axis.

This view, of course, contrasted sharply with the classical interpretation of 'demand'. Within a classical frame of reference this term was construed largely in a 'logistical' sense: i.e. to refer to the quantities of goods required for particular purposes. It was on this basis that classical economists could assert that population growth would increase the 'demand' for subsistence goods (or of the quantum of subsistence goods required by the economic system). The effects of consumer preferences on transactions received little attention; in the main the dominant classical assumption had been that the tastes of the bulk of the population (i.e. of the working class) were fairly rigid and the more pessimistic prognoses about the prospects for suppressing population growth rested on this presupposition. Nor had classical writers stressed the point that quantities demanded would vary in response to changes in market prices. Their sights were too closely centred on the forces influencing the 'natural price' of commodities to make this question central to their analytical programme.4

In neo-classical economics, the determination of market prices became the problem and the concept of demand as a schedule of price-quantity relationships was crucial to its solution. In Marshall's formulation, the construction of such a schedule proceeded in two stages. The first concerned individual consumers and rested on a notion labelled as 'diminishing marginal utility'. A consumer entered the market-place, it was maintained, in order to acquire satisfactions (or utilities) from his purchases. The amount of satisfaction obtainable from a unit of a commodity was closely related, however, to the number of units acquired. With the addition of each unit, it could be expected that the increment in total satisfaction (i.e. the additional or marginal utility) would decline. The rational consumer would thus be prepared to pay less for the last unit than for the preceding ones and a reduction in price would be necessary to induce him to buy more.

The full derivation of a market demand curve for a specific commodity involved a further step. The demand schedules of individual consumers had to be consolidated. The price-quantity relationships likely to prevail in the market as a whole could then be depicted. It was important to note, however, that such a construction presupposed that a number of conditions remained unchanged: in particular, the tastes of consumers, their money incomes (through which their desires could be translated into effective demand),5 and the prices of other goods. Variation in any of these conditions would shift the demand curve to a different position.

But this was not the end of the story. In a practical situation consumers have more than one good to choose from. If they were to maximize the utility obtainable from a given income they should adjust their spending patterns to ensure that no gain in satisfaction would be possible from an alternative allocation of their expenditures. The optimum result would be obtained when the last penny spent on any of the goods in question added an identical amount of satisfaction. Otherwise, a different allocation of expenditure would increase the consumer's total satisfaction. As Marshall expressed this proposition:

... good management is shown by so adjusting the margins of suspense on each line of expenditure that the marginal utility of a shilling's worth of goods on each line shall be the same. And this result each one will attain by constantly watching to see whether there is anything on which he is spending so much that he would gain by taking a little away from that line of expenditure and putting it on some other line.6

This type of argument had been latent in economic discussion since the days of Benthamite utilitarianism. The only novelty in its application to neo-classical problems lay in the explicit introduction of the concept of marginal utility. Just as the notion of diminishing returns had been hit upon simultaneously by a number of writers in the early nineteenth century, so also was the concept of marginal utility formulated independently (and at about the same time) by a number of neo-classical economists: Jevons in England, Menger in Austria, and Walras in Lausanne. Marshall, though he could legitimately claim to be among the innovators, could not support his case with published evidence. Characteristically he had chosen not to release his findings until they could be presented in a form intelligible to a lay audience.

This approach to the demand side of price formation had an important consequence: it swept under the carpet some of the organizing concepts of classicism. To most classical writers it had been axiomatic that economic value could be attributed only to tangible objects. By contrast neo-classical economists insisted that the point of an economic system was not the production of commodities, but the production of satisfactions. The measure of value was what the public would buy. Services, fully as much as material goods, could pass this test. Indeed, the whole debate about material-non-material distinctions could be dismissed. Marshall stressed this point when he wrote:

Man cannot create material things. In the mental and moral world indeed he may produce new ideas; but when he is said to produce material things, he really only produces utilities; or in other words, his efforts and sacrifices result in changing the form or arrangement of matter to adapt it better for the satisfaction of wants. ...

It is sometimes said that traders do not produce: that while the cabinet-maker produces furniture, the furniture-dealer merely sells what is already produced. But there is no scientific foundation for this distinction. They both produce utilities, and neither of them can do more. ...7

Similarly, classical notions of productive and unproductive labour were eliminated:

We may define labour as any exertion of mind or body undergone partly or wholly with a view to some good other than the pleasure derived directly from the work. And if we had to make a fresh start, it would be best to regard all labour as productive except that which failed to promote the aim towards which it was directed, and so produced no utility.8

Demand alone, however, provided only part of the explanation of price. No less important were the terms on which producers were prepared to make goods and services available for purchase. The neo-classical account of this aspect of the pricing process was developed in a manner analogous to the derivation of a demand curve. Just as consumers acquired utilities (though at a diminishing rate) through market transactions, producers suffered dis-utilities (and at an increasing rate) when making their services available. In short, production involved costs and sacrifices which, in most cases, were expected to rise as the quantity offered increased. It was recognized in passing that some persons might obtain positive satisfaction from work; nevertheless, supplies of the productive services of labour, land and capital were not likely to be forthcoming for long unless those in a position to offer them were compensated for their trouble.

This argument about the terms on which the services of land, labour and capital would be made available for production was reinforced by another consideration. In general it was presupposed that alternative uses of the various factors of production were available. Any individual buyer of these services would, therefore, be obliged to compete to acquire them. Firm X, for example, could not expect to acquire more land, labour, or capital for its purposes unless it was prepared to outbid other claimants for the same resources. The point at issue was described more formally in terms of ‘opportunity costs' - i.e. costs in the form of income the supplier of services was obliged to forego when committing himself to one activity, thus precluding other options. It was not always recognized within the neo-classical tradition, however, that this argument depended on conditions of full employment; otherwise some suppliers of productive services might have no readily available options. In such a situation the 'opportunity costs' of employment would be zero.

These considerations provided the raw materials from which a market supply curve could be constructed. Inasmuch as it could normally be assumed that firms could obtain greater quantities of the necessary productive inputs (labour, land, and capital) only at increased cost, it followed that they could be expected to expand their offerings of outputs only when higher prices made this course of action worthwhile. In short, it was postulated that firms normally operated under conditions in which the addition to total costs associated with producing additional units of output (i.e. marginal costs) were rising. This conclusion was further buttressed by the view that the addition to the total product obtainable from adding a unit of one productive input (while the quantum of others was unchanged) was likely to decline. The structure of marginal costs, in turn, determined the shape of the supply curve. But, just as the demand curve was expected to shift should the conditions underlying it alter, the supply curve would also shift if costs of production changed. And, just as the market demand for a particular product was derived by aggregating the demands of individual consumers, a market supply curve could similarly be arrived at by consolidating the supply curves of firms producing identical outputs.

The treatment of costs developed in this analysis could not stand in sharper contrast with the notions employed by the classical and Marxian traditions. Exercises in reducing costs to labour inputs now vanished from the scene. Their place was taken by the general account of sacrifices incurred in the supply of any of the productive services. In the neo-classical scheme the former primacy of labour in the explanation of costs was completely eliminated.

With his twin concepts of supply and demand, Marshall had the tools he needed for his explanation of price. It was at the point of intersection between these two curves that the equilibrium price (i.e. the price towards which the market would naturally tend to gravitate) was established. A price above the equilibrium would produce, a situation in which sellers would be prepared to offer more than buyers would take; the resulting disappointments of sellers would, under competitive conditions, lead to reductions in the offer price to a level at which the market could be cleared. Conversely, a sub-equilibrium price would produce frustrations for some potential buyers; the normal path of adjustment would be one in which competitive bidding would push the market price towards equilibrium. Marshall likened these two curves to the blades of a pair of scissors and observed that 'we might as reasonably dispute whether it is the upper or the under blade of a pair of scissors that cuts a piece of paper, as whether value is governed by utility or cost of production'.9

To modern readers acquainted with the microeconomics section of a standard textbook the Marshallian account of price formation may now appear to be too familiar - perhaps even too self-evident - to require extensive defence or justification. At the time of its formulation, however, it was a considerable innovation. Not only was it a major departure from the classical and Marxian labour-based account of value, but it was also constructed to counter over-zealous reactions against classical approaches on the part of some early neo-classicists. Jevons, for example, had asserted that utility and demand considerations alone were sufficient for an adequate explanation of price. Marshall rejected both the classical and cruder neo-classical positions. Demand (based on utility) and supply (based on costs of production) were both indispensable to the explanation of market prices.

One further analytical consequence of Marshallian procedure deserves mention. From this perspective the distinction upon which so much classical discussion had turned - that between value (the natural price) and the market price - evaporated. The search for an invariant measure of value over prolonged periods was abandoned. What mattered were prices as they were determined in a competitive market process.

3. THE THEORY OF DISTRIBUTION

For Marshall and his neo-classical contemporaries the analysis of distribution was essentially a problem in the pricing of productive services. Its solution was sought along lines analogous to those followed in explaining the pricing of products. In the case of both inputs and outputs, the interaction of supply and demand established equilibrium prices.

This approach was built around a three-fold classification of the basic productive factors - land, labour and capital - to each of which was assigned a unique distributive share. (Some writers added a fourth productive factor; Marshall suggested that organizational skill might be so regarded). In this scheme of things wages were defined as the reward for human effort. This definition, unlike the classical one, did not restrict wage payments to a working class. Salary incomes and an imputed 'wage to management' in owner-operated establishments also fell within the neo-classicist's wage classification. Interest accrued to the owners of capital as a reward for 'waiting' - i.e. for the sacrifice involved in foregoing present consumption in favour of prospective future gains. While rents were associated with the productive services supplied by land, the classical pre-occupation with agricultural land was shaded towards the background. In the neoclassical era the site values of urban land came into prominence.

In this re-definition of distributive shares the concept of profits with which the classical and Marxian traditions had worked largely disappeared. Much of the income earlier traditions assigned to profits was now absorbed as a wage to management and as interest. Though neoclassical economists did not share a common concept of profits, most (including Marshall) held that pure profits (i.e. rewards to business in excess of the normal wage of management, interest on invested capital, etc.) should be regarded as symptomatic of a temporary disequilibrium or of the existence of monopoly.

This approach to distribution represented a clear rejection of the class-oriented scheme around which classical and Marxian models had been organized. Neoclassical theory rested on a functional interpretation of distributive shares which linked income payments to the productive contribution of the various factors. These definitions provided ammunition for a further attack on Marxian analysis. Marshall drove the point home forcefully:

It is not true that the spinning of yarn in a factory, after allowance has been made for the wear-and-tear of the machinery, is the product of the labour of the operatives. It is the product of their labour, together with that of the employer and subordinate managers, and of the capital employed; and that capital itself is the product of labour and waiting; and therefore the spinning is the product of labour of many kinds, and of waiting. If we admit that it is the product of labour alone, and not of labour and waiting, we can no doubt be compelled by inexorable logic to admit that there is no justification of interest, the reward of waiting; for the conclusion is implied in the premiss.10

Marshall might, of course, have added that his own conclusions followed from the premisses he had chosen - premisses that imparted to property shares of income a legitimacy that Marx had not been prepared to allow.

Once these distributional categories had been defined, supply and demand forces in the market could be appealed to as the basis on which rewards to the suppliers of productive services were established. It was, of course, recognized that each of the markets in which productive factors were priced had special properties. The labour force, for example, was highly differentiated by varying skills and abilities; the market, however, could usually be relied upon to recognize differences in the productive contribution of various types of labour and to establish appropriate wage differentials. In any event, classical exercises in reducing labour to a standard unit were superfluous. The treatment of capital presented a complication of another sort. As Marshall recognized, a distinction between the accumulated stock of capital and the flow of new investments was required because the economic implications of payments to the owners of old and newly-created capital were quite different. As he saw the matter:

That which is rightly regarded as interest on ‘free' or 'floating' capital, or on new investments of capital, is more properly treated as a sort of rent - a quasi-rent - on old investments of capital. ... And thus even the rent of land is seen, not as a thing by itself, but as the leading species of a large genus. . . .11

All this was a far cry from the classical concern over the long-period behaviour of distributive shares. In Marshall's hands, distribution theory was primarily a special case of the pricing of productive inputs in the marketplace.

4. THE THEORY OF PRODUCTION

Neo-classical production theory addressed itself to two principal issues. The first concerned the manner in which any producer set about combining the productive factors. The second dealt with the adjustments a producer might be expected to make when market conditions altered.

The first of these points could be handled quite straightforwardly with the aid of analytical tools already considered. Individual business men were regarded as rational calculators seeking to maximize their earnings. So long as competitive conditions prevailed they were powerless to influence the prices of their products. The objective of profit maximization thus amounted to an attempt to minimize costs. Technically, a number of possible combinations of various productive factors could produce whatever volume of output might be desired. The rational manager would naturally select the least-cost combination.

These rules were simple enough. The analysis of the producer's response to a change in market circumstances was more intricate. In particular, it presented the problem of time which Marshall described as 'a chief cause of those difficulties in economic investigation which make it necessary for man with his limited power to go step by step; breaking up a complex question, studying one bit at a time, and at last combining his partial solutions into a more or less complete solution of the whole riddle'.12 This task of disentanglement involved an examination of the consequences of minute changes on the assumption described in Marshallian shorthand as ceteris paribus: i.e. that all the underlying factors remained unchanged.

For Marshall's purposes three time periods were distinguished from one another. The first he described as `the market period', a period too short for the producer to make any adjustments in his output in response to a change in prices. The second - labelled 'the short run' - permitted output to be adjusted by changing the intensity with which a given plant was utilized. More workers might be hired (or the present labour force induced to work longer hours) and additional raw materials acquired. All of these measures would enable output to be enlarged in response to an increase in demand. These adjustments, however, would probably be associated with rising marginal costs. If an increase in demand was expected to be sustained, it might well be worth the firm's while to expand capacity in order to reduce its costs. The time period required to effect this adjustment was described as 'the long run'.

The nature of these divisions of economic time deserves a moment's reflection. At first sight it might appear that these categories bore a resemblance to the notions of time with which classical economists had worked. Any such appearances would be quite misleading. Classical writers had been interested in historical change. The time distinctions introduced by Marshall were divorced from calendar time, resting instead on logical distinctions.13 If asked to specify the length of the 'long run', Marshall would reply that it was the time span sufficient to accomplish adjustments in the scale of plant necessary to produce a new market equilibrium after an earlier one had been disturbed. In a practical case, the length of this period would depend on the circumstances of individual firms and industries. The 'long run' for a steel fabricator and for the corner hairdressing establishment would not coincide.

These logical distinctions between moments of economic time opened the door to a new and interesting set of theoretical possibilities. After all, it was quite conceivable that in the long-run - when the scale of plant could be altered and the utilization of all productive factors varied - several outcomes with respect to levels of costs might follow. Changes in scale, for example, might be associated with rising, declining, or constant unit costs. The most interesting case was the one in which average costs declined with the enlargement in the scale of plant; this situation was described as 'increasing returns to scale'. By and large the classical economists had anticipated that 'constant returns to scale' would normally prevail; in other words, that the size of the individual production unit had no effect on average costs. They had, of course, given much attention to the gains in productivity arising from growth in the size of the economy (and the progressive sub-division of labour associated with it), but this scale effect was quite different from the neoclassical concern with individual enterprises. Mill and Marx, to be sure, had caught glimpses of the cost reducing effects of large industrial concentrations though they had not fully worked out their implications.

For Marshall, increasing returns to scale associated with the application of high technologies presented an awkward problem. Economies of scale implied that a small number of large producers could operate with lower unit costs than could a large number of small firms producing the same quantum of output. Hence, one of the premisses of a competitive market - namely, that the number of firms producing similar products was sufficiently large to deny market power to any individual seller - was challenged. Bigness might indeed erode the basis of the competitive order and threaten its preservation. Marshall saw the issue when he wrote:

In fact when the production of a commodity conforms to the law of increasing return in such a way as to give a very great advantage to large producers, it is apt to fall almost entirely into the hands of a few large firms; and then the normal marginal supply price cannot be isolated on the plan just referred to, because that plan assumes the existence of a great many competitors with businesses of all sizes, some of them being young and some old, some in the ascending and some in the descending phase. The production of such a commodity really partakes in a great measure of the nature of a monopoly; and its price is likely to be so much influenced by the incidents of the campaign between rival producers, each struggling for an extension of territory, as scarcely to have a true normal level.14

The availability of scale economies had consequences both for the industrial structure of the economy and for the structure of neo-classical reasoning. At the analytical level it precluded a clear and unambiguous operational definition of a supply schedule. Marshall perceived the implications of this complication (and criticized others for their failure to do so) in the following language:

Some ... have before them what is in effect the supply schedule of an individual firm; representing that an increase in its output gives it command over so great internal economies as much to diminish its expenses of production; and they follow their mathematics boldly, but apparently without noticing that their premises lead inevitably to the conclusion that whatever firm gets a good start will obtain a monopoly of the whole business of its trade in its district. While others avoiding this horn of the dilemma, maintain that there is no equilibrium at all for commodities which obey the law of increasing return; and some again have called in question the validity of any supply schedule which represents prices diminishing as the amount produced increases.15

For this part, Marshall attempted to build an analysis in which the essentials of the competitive equilibrium model could be preserved despite this challenge to its realism.

5. THE PROSPECTS FOR THE COMPETITIVE ORDER

Two Marshalls - one the abstract theorist, the other the practical observer of everyday economic life - were blended in all of his writing. This duality was most conspicuous in his treatment of market structures and the competitive process.

As a formal theorist Marshall saw the potential danger to the competitive order latent in the growth of large productive units with considerable market power. But as an observer of events he argued that a number of factors tended to moderate the social and economic consequences of such concentrations. Characteristically, Marshall maintained that when analytical tidiness and descriptive realism appeared to be at odds with one another ordinary observation should claim precedence. Theory might be indispensable but it also had inherent limitations. No theoretical construction could embrace `all the conditions of real life' for then `the problem is too heavy to be handled'; but he feared that if only a few aspects were selected for study, then 'long-drawn-out and subtle reasonings with regard to them become scientific toys rather than engines for practical work'.16

At the descriptive level Marshall distinguished between two types of market structure. One he described as the `special' market, a sphere in which individual firms could operate largely in isolation from immediate competitors. These circumstances might arise, for example, through geographical isolation or as a by-product of the existence of a special clientele served by a particular seller. But the ‘special' market was also surrounded by a larger and more embracing ‘general' market. Marshall invoked these distinctions in an attempt to reconcile the world of business behaviour with a model in which effective competition was an analytical requirement.

Marshall's strategy for salvaging his competitive plan from the threats implied by the technology of increasing returns also rested on presuppositions about the nature of business enterprises - and, most importantly, his view that firms could be likened to biological organisms. Both had a life cycle which included phases of expansion (and perhaps even of supremacy), and phases of decline, decay and - ultimately - death. Ownership and control of business enterprises might be handed down over the generations, but in the process the vigour of those who led it during dynamic periods was unlikely to be perpetuated. Marshall depicted the situation as follows:

Nature still presses on the private business by limiting the length of the life of its original founders, and by limiting even more narrowly that part of their lives in which their faculties retain full vigour. And so, after a while, the guidance of the business falls into the hands of people with less energy and less creative genius, if not with less active interest in its prosperity. If it is turned into a joint-stock company, it may retain the advantages of division of labour, of specialized skill and machinery: it may even increase them by a further increase of its capital; and under favourable conditions it may secure a permanent and prominent place in the work of production. But it is likely to have lost so much of its elasticity and progressive force, that the advantages are no longer exclusively on its side in its competition with younger and smaller rivals.17

These `natural' factors were not the only ones checking the growth of firms and limiting the exercise of market power. Other constraints were inherent in the make-up of the ‘special' markets within which firms could enjoy unique privileges. Marshall insisted that these advantages could not long be retained by an expanding firm. In this connexion he wrote:

.... many commodities with regard to which the tendency to increasing return acts strongly are, more or less, specialities: some of them aim at creating a new want, or at meeting an old want in a new way. Some of them are adapted to special tastes, and can never have a very large market; and some have merits that are not easily tested, and must win their way to general favour slowly. In all such cases the sales of each business are limited, more or less according to circumstances, to the particular market which it has slowly and expensively acquired; and though the production itself might be economically increased very fast the sale could not.18

Or, as Marshall again emphasized the point: ‘There are many trades in which an individual producer could secure much increased "internal" economies by a great increase of his output; and there are many in which he could market that output easily; yet there are few in which he could do both. And this is not an accidental, but almost a necessary result.'19 Expansion of a firm beyond the confines of its special market would also expose it to the competition of rivals. The market protection it had formerly enjoyed would be sacrificed as producers from the `general' market checked its economic power.

These considerations led Marshall to the optimistic conclusion that economies of scale were unlikely to present a severe challenge to the maintenance of a competitive order. The same factors which enabled firms to enjoy a limited degree of market power (the existence of `special' markets) also restrained the trend toward bigness. From purely theoretical considerations, a quite different conclusion might be drawn. Characteristically, Marshall cautioned against judgements based solely on a priori reasoning and recommended `treating each important concrete case very much as an independent problem, under the guidance of staple general reasonings. Attempts so to enlarge the direct applications of general propositions as to enable them to supply adequate solutions of all difficulties, would make them so cumbrous as to be of little service for their main work. The "principles" of economics must aim at affording guidance to an entry on problems of life, without making claim to be a substitute for independent study and thought.' 20

While Marshall was not prepared to buy analytical rigour at the price of contact with reality, his institutional portrait of business behaviour was not without limitations. The picture he offered of the restraints to expansion of firms might have been reasonably accurate in late nineteenth and early twentieth century England. His notion of the life cycle of firms, however, is much less plausible when applied to the modern corporation. Its institutional structure, in which management and ownership are largely divorced, creates a survival power that may approach immortality. Nor is Marshall's argument well-suited to production for mass markets. The 'special' markets he had in mind were built on Edwardian 'custom' tastes. In an era of mass consumption, in which the tastes of the public for a wide range of consumer outputs are not highly differentiated by social status, Marshall's views of the exclusiveness and selectivity of elite special markets must be considerably qualified.

Marshall's approach to the theory of the firm has left a dual legacy. Parts of his analysis have been elaborated into formal models of the equilibrium conditions generated by a regime of perfectly competitive firms. Other portions have provided a springboard for the more institutionally-oriented doctrines of workable competition in which it is held that the important results of a perfectly competitive system can be approximated even in a market structure which is not dominated by a large number of small firms.

6. THE AGGREGATIVE STRAND OF MARSHALL'S THOUGHT

Though Marshall's attention was primarily directed to microeconomic problems, aggregative themes still occupied a place in his thought. In his view, the major microeconomic question was the determination of the general price level. Short-term fluctuations in output and employment were peripheral matters; when they occurred they were expected to be temporary and slight.

His analysis of the general price level was developed around a version of the 'quantity theory' of money. Much of the earlier discussion of this point of doctrine had proceeded from the tautological statement that the quantity of money multiplied by the number of times it was spent in a given time period (the velocity of circulation) would necessarily be equal to the average price level multiplied by the total number of transactions; this expression, after all, amounted to no more than two ways of viewing total expenditure. Marshall modified this procedure by shifting the focus from the rate at which the money supply turned over to an examination of the money balances held by the community. In the hands of one of Marshall's pupils this way of viewing money was later to open up fresh analytical horizons. Marshall's own results from the use of this 'cash balance' approach however, were essentially no different from those that had been reached via the 'velocity of circulation' route. He maintained that the amount of money held was regulated by the institutional arrangements of the economy and, on ceteris paribus assumptions, could be treated as a constant. In his words:

... whatever the state of society, there is a certain volume of their resources which people of different classes taken one with another, care to keep in the form of currency; and, if everything else remains the same, then there is this direct relation between the volume of currency and the level of prices, that, if one is increased by ten per cent, the other also will be increased by ten per cent.21

The effect of this procedure was to reinforce the essential requirement of Say's Law: that all income would be spent. The possibility of leakages into idle balances could, for practical purposes, be ignored. Money was interesting primarily in relation to spending and to the general price level, rather than for any connexion it might bear to the level of interest rates. This conclusion, of course, gained additional strength from Marshall's insistence - which was common to the neo-classical tradition as a whole - that interest rates would be established through the interaction of the supply of loanable funds (fed by saving) and the demand for loanable funds (stimulated by the productivity of capital). Moreover, the rate of interest could be relied upon to produce an equilibrium between decisions to save and to invest. Should the demand for loanable funds increase, the rate of interest would rise, making it more attractive for people to reduce consumption spending and to save. Conversely, should the public choose to save more, the rate of interest would fall. Investors would then be induced to increase both their borrowings and their expenditures on plant and equipment. Further, this way of looking at the matter implied that the intersection of the curves of supply and demand for loanable funds determined the equilibrium rate of interest. The position of these curves, in turn, was established by the thriftiness of the community (on the supply side) and by the productivity of capital (on the demand side).

This line of reasoning, while supporting a Say's Law interpretation of aggregative economic activity, did not preclude the possibility of economic instability. Though no disturbances of the scale experienced in the 1930s clouded the horizons of Marshall and his neo-classical contemporaries, they did observe modest cycles of boom and bust. How were these fluctuations to be explained? In Marshall's view the main answer was to be found in the psychology of the business community. Waves of optimism and pessimism seemed endemic to it. When business men were bullish the demand for loans increased. This phase might generate capital spending on many high-risk undertakings, some of which were doomed to failure. And when they did fail the bubble was pricked. Pessimism replaced optimism as the dominant mood; investment and economic activity generally would be curtailed. As Marshall described the process:

The recent history of fluctuations of general credit shows much variety of detail, but a close uniformity of general outline. In the ascending phase, credit has been given somewhat boldly, and even to men whose business capacity has not been proved. For, at such times a man may gain a profit on nearly every transaction, even though he has brought no special knowledge or ability to bear on it; and his success may probably tempt others of like capacity with himself, to buy speculatively. If he is quick to get out of his ventures, he probably makes a profit. But his sales hasten a fall of prices, which must have come in the course of time. Though the fall is likely to be slight at first; yet each downward movement impairs the confidence which had caused the rise of prices, and is still giving them some support. The fall of a lighted match on some thing that smoulders has often started a disastrous panic in a crowded theatre.22

Credit cycles, however, were still incapable of converting ‘partial' over-production into ‘general' over-production. Given time, the economic system would adjust itself to its normal full employment level of operations. No special action on the part of government was required to accomplish this result and, indeed, direct government intervention might make matters worse. The tendency toward instability could, however, be moderated by anticipatory action on the part of the monetary authorities whose proper role was to minimize discrepancies between prevailing interest rates and the rate that would be established through the normal interplay of the supply of and demand for loanable funds.

The main thrust of Marshall's aggregative analysis thus buttressed faith in the capability of the economic system, if left alone, to eradicate involuntary idleness. In the final edition (1920) of his Principles, however, Marshall added one dark hint that the analytical basis for this conclusion might ultimately need to be revised. Following an orthodox neo-classical discussion of the relationship between productivity, thrift, and the rate of interest, he inserted a note of qualification:

... every one understands generally the causes which have kept the supply of accumulated wealth so small relatively to the demand for its use, that that use is on balance a source of gain, and can therefore require payment when loaned. Everyone is aware that the accumulation of wealth is held in check, and the rate of interest so far sustained, by the preference which the great mass of humanity have for present over deferred gratifications, or, in other words, by their unwillingness to wait. And indeed the true work of economic analysis in this respect is, not to emphasize this familiar truth, but to point out how much more numerous are the exceptions to this general preference than would appear at first sight.23

He elaborated this point in a footnote with the words:

It is a good corrective of this error to note how small a modification of the conditions of our own world would be required to bring us to another in which the mass of the people would be so anxious to provide for old age and for their families after them, and in which the new openings for the advantageous use of accumulated wealth in any form were so small, that the amount of wealth for the safe custody of which people were willing to pay would exceed that which others desired to borrow; and where in consequence even those who saw their way to make a gain out of the use of capital, would be able to exact a payment for taking charge of it; and interest would be negative all along the line.24

These heterodox afterthoughts did not mar the tranquillity of Marshall's grand design. They did foreshadow - far more than Marshall himself could possibly have suspected - the assault on neo-classical aggregative premisses launched by Keynes in the 1930s.

7. MARSHALL ON LONG-PERIOD ECONOMIC CHANGE

Within the framework of neo-classical theory, long-period economic change had little place. Marshall addressed himself only briefly to the subject with a discussion of `the secular period' of the economy. In all essential respects this time dimension was identical with the one with which classical writers had been preoccupied.

From his vantage point in time Marshall could observe that the gloomier classical prognoses on the fate of the economy had not, in fact, been borne out. The stationary state had not emerged; despite increases in population, real incomes of workers had improved; capital accumulation had proceeded, but it had not been accompanied by a widespread displacement of labour. Nor had the growth in demand for foodstuffs given landlords a stranglehold over the economy. The expansion of international trade (and particularly the opening up of low cost sources of food supply) had been partially responsible for the outcome.

Despite all this, Marshall shared the general classical conclusion that rents would tend to rise during the course of sustained economic expansion. In his interpretation, however, this phenomenon was associated less with the natural limits to the fertility of the soil than with growth in demand for business and residential sites. Indeed, rising rents in urban situations had the more serious implications for the cost structure. The application of high technologies to agriculture held out prospects for productivity improvements that would forestall a re-distribution of income in favour of agricultural landowners.

Marshall's treatment of wages also departed substantially from the main classical line. He would have no part of the Malthusian `iron laws'. On this point he followed the path charted by Mill by rejecting the view that population growth would necessarily frustrate sustained improvement in real wages. It was Marshall's expectation that workers would grow in skill, energy, and self-respect and that their productivity and their incomes would be correspondingly enhanced.25 Similarly, Marshall dismissed Ricardian and Marxian anxieties about the effects of capital accumulation on employment. Much of any short-term competition between capital and labour, he contended, would be offset by the growth in demand for workers in the capital goods industries. Moreover, the cost-reducing effects of mechanization were clearly a blessing: competition could be relied upon to ensure that price reductions - the gains from which would be shared by all segments of the community - would follow. Marshall, of course, was on stronger ground when arguing this point than were the early classicists. The latter had assumed that real wages would always be so close to the subsistence level that there was little room in the worker's budget for goods produced by higher techniques (and thereby subject to price reduction). In Marshall's world, it could be more plausibly maintained that the benefits of lower price would be much more widely diffused.

Another prominent contrast between Marshall's conclusions and those of the classical orthodoxy concerned the doctrine of the falling rate of profit. This proposition, which had occupied a central position in classical thought supplied the underpinnings to fears about the ultimate emergence of the stationary state. Marshall's treatment c this question was, of course, adjusted to a different set c distributional categories. Profits could no longer be regarded in the classical sense (i.e. as the income of the capitalist class); instead the rate of interest was held to b a more appropriate measure of the return to the supplier of capital. Marshall acknowledged that the rate of interest would tend to fall as accumulation proceeded, but only to the extent that increments to the capital stock were subject to diminishing returns. Such tendencies, however might well be offset by technological progress. Marshal maintained that there was every reason to believe that technical improvement would proceed at faster rates that the classical economists had anticipated. In any event, the average rate of return on capital throughout the economy was not particularly pertinent to the investment decision. The carriers of economic progress were men who sough out avenues for reaping above-average returns on capital In short, it was the result of calculations at the margin that mattered.

8. MARSHALL AND ECONOMIC POLICY

By his own account, Marshall was originally attracted to the serious study of economics by a desire to understand the causes of poverty and the means by which it could be alleviated. He emerged from his investigations convinced that:

the social and economic forces already at work are changing the distribution of wealth for the better: that they are persistent and increasing in strength; and that their influence is for the greater part cumulative; that the socio-economic organism is more delicate and complex than at first sight appears; and that large ill-considered changes might result in grave disaster.26

His sympathy for the sufferings of the mass of humanity had by no means diminished. But these impulses were now substantially tempered by the belief that radical measures to alter the existing economic order would be unwise. In particular, he opposed a socialistic programme on the grounds that:

the collective ownership of the means of production would deaden the energies of mankind, and arrest economic progress; unless before its introduction the whole people had acquired a power of unselfish devotion to the public good which is now relatively rare. And ... it might probably destroy much that is most beautiful and joyful in the private and domestic relations of life. These are the main reasons which cause patient students of economics generally to anticipate little good and much evil from schemes for sudden and violent reorganization of the economic, social and political conditions of life.27

While the market system as portrayed by Marshall was largely benevolent, his analysis also demonstrated that in certain situations unregulated markets could not be relied upon to yield socially desirable results. Prominent among the exceptions were the cases in which - for technical reasons - competition would prove to be wasteful and inefficient, if not a practical impossibility. The `natural monopolies' (a term which Marshall associated primarily with such public services as water supply, power generation, etc.) could not usefully be organized in accordance with the competitive plan and the case for government regulation (if not public ownership) in these instances was clear. He was reluctant, however, to recommend government intervention in those sectors of the economy in which increasing returns to scale threatened to produce industrial concentrations, even though these circumstances implied that individual firms could enjoy considerable market power and that prices would not be competitively determined. This problem, he contended, deserved continuing study. His general position on the life cycle of firms led him to the conclusion that the potential market power of large business units was unlikely to be abused for long.

Though he was disposed to view the market as a sensitive instrument through which an economy's resources could be efficiently allocated, he also recognized that its performance could be improved. For this purpose, improvements in public education were particularly important. Consumers and producers could then conduct their affairs more intelligently by enhancing the rationality of their choices. Moreover, improved economic education would do much to eradicate one of the blights of an unregulated market system - the bouts of speculation which gave rise to harmful fluctuations.

Marshall was also prepared to entertain the possibility that governments could play a useful role in improving the allocative efficiency of markets. Would not the sum of social satisfactions be increased, he asked, if the productive resources of society were shifted in favour of lines of production subject to increasing returns and away from those in which decreasing returns prevailed? Greater outputs could then be obtained from the existing stock of resources. Governments could encourage a re-allocation of resources along these lines through appropriate taxes and subsidies. He advanced this suggestion, however, with the utmost caution, pointing out that such policies could be justified only when it could be shown that the gains in satisfactions arising from expanded outputs in the subsidized sectors more than offset the losses in utility associated with higher tax levies on others. He recognized that this criterion would be difficult to apply to practical cases.

Conceivably the introduction of maximization of aggregate utility as a goal of public policy could also be used to support recommendations for a redistribution of income. If it could be assumed that the marginal utility of money was likely to be greater for a poor man than for a rich one, it followed that society's aggregate satisfactions would be enlarged through a redistribution from rich to poor. Marshall did not draw this conclusion. He did recommend a less systematic scheme of income redistribution for further study when he wrote of the possibilities of ‘economic chivalry'. Such a regime would tax the rich to ameliorate the distress of those still trapped in poverty.