The New Palgrave Dictionary of Economics

2018 Edition
| Editors: Macmillan Publishers Ltd

Oligopoly

  • P. Sylos-Labini
Reference work entry
DOI: https://doi.org/10.1057/978-1-349-95189-5_1386

Abstract

No article entitled ‘oligopoly’ appeared in any edition of Palgrave’s Dictionary of Political Economy. It is true that the simplest case of oligopoly, that is, duopoly, was considered more than a century and a half ago, by Cournot; but such an analysis was motivated by purely theoretical interests. The fact is that only in the 20th century and especially after the Second World War did this market form become important in economic reality, as a result of two processes of economic change: the process of concentration and the process of differentiation. In those branches where the former process has asserted itself – for example, steel, basic chemical products, cement, electricity – concentrated oligopoly with relatively homogeneous products has emerged; where the latter process has prevailed, we find differentiated oligopoly; in those branches where both processes have taken place simultaneously, then mixed oligopoly has emerged. In both processes innovations have played a major role, with the proviso that in the process of concentration innovations have given rise to economies of scale, whereas in the process of differentiation the most important role has been that of technological innovations implying economies of specialization; in this case, technological innovations are combined with commercial innovations. In fact, differentiated oligopoly can be found mainly in those activities in which quality competition, commercial services and advertising have had a particularly relevant role – non-durable consumer goods, such as textiles, tyres, canned foods, soft drinks and cigarettes are often produced in conditions of differentiated oligopoly.

Keywords

Advertising Barriers to entry Competition Concentration Differentiated oligopoly Duopoly Full cost principle Great depression Imperfect competition Increasing returns Innovation Market power Mixed oligopoly Monopoly Obstacles to entry Okun’s Law Oligopoly Price determination Price leaders Price rigidity Price variation Price wars Productivity growth Returns to scale Specialization economies Sraffian economics Technical change Verdoorn’s Law 

No article entitled ‘oligopoly’ appeared in any edition of Palgrave’s Dictionary of Political Economy. It is true that the simplest case of oligopoly, that is, duopoly, was considered more than a century and a half ago, by Cournot; but such an analysis was motivated by purely theoretical interests. The fact is that only in the 20th century and especially after the Second World War did this market form become important in economic reality, as a result of two processes of economic change: the process of concentration and the process of differentiation. In those branches where the former process has asserted itself – for example, steel, basic chemical products, cement, electricity – concentrated oligopoly with relatively homogeneous products has emerged; where the latter process has prevailed, we find differentiated oligopoly; in those branches where both processes have taken place simultaneously, then mixed oligopoly has emerged. In both processes innovations have played a major role, with the proviso that in the process of concentration innovations have given rise to economies of scale, whereas in the process of differentiation the most important role has been that of technological innovations implying economies of specialization; in this case, technological innovations are combined with commercial innovations. In fact, differentiated oligopoly can be found mainly in those activities in which quality competition, commercial services and advertising have had a particularly relevant role – non-durable consumer goods, such as textiles, tyres, canned foods, soft drinks and cigarettes are often produced in conditions of differentiated oligopoly.

In the past, when the standard of living of the masses of consumers was not much above the subsistence level, there was not much scope for the factors just mentioned. With the gradual increase of per capita income, consumers’ preferences have acquired an increasing space. At the same time the possibility of advertising has been greatly enhanced by particular innovations – modern means of transportation and the so-called mass media, among which radio and television play a special role. Mixed oligopoly (concentration cum differentiation) is typical of several industries producing consumer durables such as automobiles, typewriters, refrigerators, radio and television sets, computers; mixed oligopoly can be found in several important service sectors such as banking and insurance. In addition a large number of non-durable consumers’ goods and services – including commercial services – constitute the area where differentiated oligopoly prevails; it is well to point out that as a rule there is no difference between imperfect competition and differentiated oligopoly. Analytically, the former can be seen, as a rule, as a first and the latter as a second approximation; this standpoint becomes natural if we recognize that the imperfect markets are composed by a ‘chain of oligopolistic groups’.

After careful reflection, we are bound to admit that in modern industry and in services, oligopoly, in its three varieties, is the rule and competition the exception –to be found in certain industries producing sufficiently homogeneous non-durable goods and in subsidiary activities. Competition, on the other hand, is the rule in most agricultural and mineral raw materials traded in international markets.

According to the traditional (neoclassical) conception, markets in competitive conditions are formed by a great number of firms, each of which is so small as to be unable to influence prices. Each firm, then, is bound to accept the market price and pushes output up to the point at which marginal cost – which, after a point, cannot but be increasing – equals price. In fact, the increasing marginal cost, that is, diminishing returns both in the short and in the long run are a necessary feature of traditional theory. In monopoly equilibrium is reached when the decreasing marginal income equals marginal cost. Indeed, according to that theory, only two market forms are worth consideration – competition and monopoly – the former being the rule, the second the exception (The analytical tools to be used for imperfect competition are those worked out for monopoly).

The whole analysis is statical and thus presupposes given technology. To work out theoretical models consistent with dynamic analysis, we have to go back to the classical concept of competition, where freedom of entry and not the number and size of firms is crucial. If we adopt this concept, it is easy to shift from competition to non-competitive market forms, by considering obstacles of various relevance to entry. Clearly, when in a given market the obstacles to entry are serious, firms operating in that market are likely to be few; this, however, is to be seen, not as a preliminary datum, but as the likely (not necessary) result of the existence of those obstacles.

When the obstacles to entry are of little importance, then a super-normal profit will attract new firms: supply will increase and the price will fall, so that supernormal profit will tend to disappear: such a profit can persist when obstacles to entry are important.

Having chosen this approach, in a first approximation we have to distinguish, in price analysis, between agriculture and mining, on the one hand, where obstacles to entry as a rule are modest, and industry and services, on the other, where those obstacles are often considerable. Again, in the first approximation, we can state, with Ricardo, that in primary activities in the short run prices depend on demand and supply, whereas in the long run they depend on costs. If we refer to the short run and intend to work out an analysis susceptible of empirical verification, we realize that ‘demand’ can be variously interpreted; in the case of raw materials traded in the international markets, demand can best be represented by an index of world industrial production. In industry and services, instead, in the short run prices depend principally on changes in direct costs and, in the long run, on changes in total costs per unit.

The reason for this sharp difference as regards short-run variations of prices is as follows. In primary activities firms, owing to the relative freedom of entry, have no outstanding market power and cannot influence prices, which vary according to the variations of aggregate demand and aggregate supply. In the other activities, however, prices are to a non-negligible extent controlled by firms and, in particular, by those that act as price leaders. Starting from a price that is accepted by all firms – that is, from an ‘equilibrium price’ – the firms acting as leaders will modify it when the conditions of equilibrium change. There are, then, two analytical problems, conceptually different but strictly interrelated: the problem of price determination and that of price variations. In traditional terms, the former problem belongs to the area of statical analysis, the latter to that of dynamics. We can accept such a distinction provided that it implies no cleavage, that is, provided that we can pass without discontinuities from the analysis of price determination to that of price variations.

The problem of price determination implies the analysis of the equilibrium, which includes: the size of the market (that is, the position in a Cartesian diagram of the demand curve, a concept that becomes relevant when firms are no more conceived as atoms); the shape of the demand curve (that is, the elasticity of demand); technology, salaries and other administrative expenses; taxes; and the prices of durable and those of variable means of production. This is not the place to present a formal solution of the problem of price determination. Suffice it to say that the concept of entry-preventing price and elimination price are important analytical tools to be used in the construction of a theoretical model of price determination. Once the price reaches the level acceptable to all firms – the equilibrium level – each firm is in a position to calculate the markup, that is, the ratio between price and cost or, more precisely, direct cost. When the equilibrium conditions change, the price is to be changed. Normally this occurs without a price war, since such wars are costly and major firms are willing to undertake them if only the expected gains (net of risks) are higher than expected costs, an occurrence that does not appear to be frequent.

The analytical steps, then, are two: the first is to understand how the equilibrium price is arrived at; the second is to understand how it varies when the equilibrium conditions change. If in the first step the concepts of entry-preventing and elimination prices are essential, in the second step it is the ‘full cost principle’ that plays the key role. Empirical enquiries have consistently shown that this principle is generally followed by managers operating in non-agricultural activities. Yet for a long period it has been considered only as a rough rule of thumb, without theoretical relevance. Probably the reasons are twofold. The first is that it contradicts the received doctrine, which is founded on marginal analysis and which, as a condition of equilibrium, assumes a rising marginal cost – the full cost principle, instead, which is based on the markup on direct costs, assumes the marginal cost to be constant and therefore equal to direct cost. The second reason is that that principle has been described as if it were a criterion to determine the price, not to modify it, but it can have a meaning only in the second case. Thus, Hall and Hitch (1939), in their pioneer empirical enquiry, report that ‘prime (or “direct”) cost per unit is taken as the base, a percentage addition is made to cover overheads … and a further conventional addition … is made for profit’. However, it is evident from this statement that the crucial theoretical problem is to explain the height of the two percentage additions – that can be unified into one percentage. Thus, given the cost elements, we have to explain the conditions that limit the discretionary powers of managers in choosing a given percentage and not another, that is, we have to explain the equilibrium conditions. Only after having explained the equilibrium price can the markup acquire a meaning. In other words, the full cost principle is theoretically meaningless as regards the problem of price determination and becomes meaningful as regards the problem of price variations: in fact, barring price wars, the markup appears to be the quickest and most rational way for firms, and particularly for price leaders, to arrive at a new equilibrium price when the equilibrium conditions vary.

The further question is to understand why direct cost and not total unit cost is taken as the term of reference to modify the price in the short run – say, year by year or even in shorter periods. The reason is that the changes in the prices of variable factors affect without much delay all firms, though not necessarily in the same proportions, whereas the changes in the other equilibrium conditions – size of the market, elasticity of demand, technology, salaries and other overhead costs – affect the firms at different degrees and in different times. These changes either affect prices in relatively long periods or do not affect them at all – substantial increases in overhead costs can be offset, not by price increases, but through productivity increases. To be sure, when these are insufficient, the increases in overhead costs can push some of the firms out of the market; this can also be the outcome of unfavourable changes in market conditions.

Changes in direct costs, then, tend to be shifted to prices in the short run. But even for this category of changes a sort of hierarchy is necessary: changes in the prices of raw materials (including the sources of energy) tend to be fully shifted on prices of finished products in both directions, since those changes tend very quickly to affect all firms. This is not so for changes in wage cost per unit, since this cost is given by the ratio between wages and productivity. Now, wage changes – if we except the areas of the so-called submerged economy – affect in a relatively short run all firms, whereas productivity increases due to organizational innovations and to technological changes determined by previous investment tend to take place at different rates in the different firms (declines in productivity are exceptional): only those changes in wage cost per unit of output have to be shifted onto prices that are common in both the upward and the downward direction. However, under contemporary conditions the shift in the downward direction will be more limited than that in the upward direction, since it is unlikely that the prices of finished industrial products in international markets will generally decrease; and it is international competition that, in industry, will limit the market power of the firms of a given country. Briefly, in the short run, the shift of changes in total direct costs will tend to be not only partial but also asymmetrical.

In the case of industrial products, then, short-run variations in prices depend on the variations of direct costs: demand does affect prices, but, as a rule, only in the long run and not in the same direction, as is the case in the short-run variations of prices under competitive conditions, but in the opposite direction, since the long-run expansion of demand makes the entry of new firms easier and opens the possibility of exploiting economies of scale. Thus, an expansion of demand tends, ceteris paribus, to reduce and not to raise the price. In the short run demand increases have no significant direct effect on prices of industrial goods; they can have an indirect effect, that is, via the prices of raw materials, when demand pressure is so strong as to affect not only finished products but also raw materials. As for finished products, demand pressure tend to affect not prices but (consistently with the Keynesian conception) the level of activity.

If we pass from partial to general analysis and adopt the framework of a Sraffian model, we are bound to distinguish between basic and non-basic (‘luxury’) products. If we decide to consider not only competitive but also non-competitive markets, we have to drop either the assumption of a unique rate of profit or the assumption of a unique wage rate (for a given type of labour). In any case, prices enter into the conditions of simple reproduction. The conditions of expanded reproduction, that is, of accumulation – to use the Marxian expression – imply, in addition, that at least a share of the surplus be employed productively, that is, invested – the velocity of accumulation being determined by that basic product that has got the lowest surplus. It is important to point out that technological progress is essential not only in the case of accumulation but also in the case of simple reproduction, since mineral products tend gradually to exhaust themselves; it is essential also in the case of a growth proportional to the increase of population, not only due to the reason just mentioned, but also due to the necessity of offsetting the tendency of diminishing returns in agriculture.

If we adopt a Sraffian model of general analysis, the study of the effects of technological changes meets with several problems, certainly serious, but, in principle, not insurmountable; in fact, some important steps in this direction have already been made by Sraffa himself. That study, instead, seems to be precluded if we adopt a Walrasian model of general equilibrium that implies a strictly static framework, in which all firms operate in conditions of diminishing returns, that is, of increasing marginal costs. Now, barring special cases, increasing returns are to be related to changes in the methods of production, even in the short run: increases in the productivity of labour can take place as a consequence of quick readjustments of the labour force and of innovating investment carried out in previous periods. A long series of empirical observations – among which may be mentioned Dunlop’s 1938 article on the movement of real wages, the ‘Verdoorn Law’ and ‘Okun’s Law’ – show that increasing, not diminishing, returns dominate modern economies and, in particular, non-agricultural activities. Thus, to admit that it is not perfect but imperfect competition and oligopoly that is the rule seems to be the only way to reconcile theoretical models and empirical enquires in both partial and general analysis.

In the short run technical progress takes mainly the form of increases in productivity of means of production and, in particular of labour; in the long run one has also to consider the production of new goods, that in the short run represent a tiny fraction of the total. The diversification of output, which in fact conditions the growth of all firms, can assume either a prevailingly commercial character, in the case where the goods are already in the market, or also a technological character, if the goods or the process through which they are produced are new. In its turn, the expansion of demand represents the condition for the introduction of two important types of technological innovations – that is, new goods and new processes implying the exploitation of economies of scale – which, after all, is nothing but another way to re-propose the Smithian proposition according to which ‘the division of labour is limited by the extent of the market’.

For the sake of simplicity, we limit ourselves to considering, as the index of technical progress both for the short and the long run, the increase in productivity of labour. The basic consequence of this increase is, at the aggregate level, a systematic divergence between the average variations of nominal incomes and the average variations of prices, with the fall of the relative prices of those goods produced in the most dynamic industries. Referring to average variations, the said divergence can take four different forms:
 

Nominal incomes

Prices

(a)

Falling

Falling more rapidly

(b)

Constant

Falling

(c)

Rising

Constant

(d)

Rising

Rising more slowly

Cases of falling prices – (a) and (b) – were frequent in the 19th century, when the process of concentration and that differentiation in industry and services had not proceeded far enough and competition was still the rule in those sectors. In the 20th century case (a) occurred during the first four years of the Great Depression; but, in sharp contrast to what was normally occurring in the 19th century, the level of activity in industry and services fell much more than prices, whereas in agriculture the prices fell violently, but the level of activity remained approximately constant. The comparison with the great depression of the 19th century – which occurred in the years 1873–9 – is illuminating.

Putting aside services, which offer a picture similar to that of industry, the percentage changes in prices and production of agriculture and industry during the two great depressions (I and II) were as follows:
 

United Kingdom

United States

Prices

Production

Prices

Production

Agriculture

I

II

−18

−44

+3

0

+31

−54

+4

+2

Industry

I

II

−29

−21

−5

−16

−33

−23

−5

−48

If we except the period of the great depression of the 20th century, which was in all senses an exceptional event, with productivity varying in a very irregular fashion, in the 20th century cases (c) and (d) – rising nominal incomes with constant prices or prices rising more closely – were the rule. Now, it is not indifferent that the fruits of technical progress have one type of consequence or the other on prices and incomes.

When prices of all goods fall, the means of production (Sraffa’s basic products) become cheaper and this stimulates the expansion of all firms, including those that do not introduce innovations. On the other hand, when prices fall demand increases automatically in real terms.

Let us now consider what happens when productivity rises but prices do not fall. If, in such circumstances, nominal incomes do not rise, the whole increase in productivity tends to translate itself into a decreasing level of employment; to have at least a stable level of employment, nominal incomes should rise in proportion to the increase in productivity; and this is not an automatic process. It is unlikely that wages and salaries rise if there is not a systematic action of trade unions, unless the process of differentiation and the consequent increasing fragmentation in the labour market have become so widespread as to favour wage increases even without a generalized pressure of trade unions. On their side, non-labour incomes will increase only if investment or government expenditure increases, or both. Investment can increase only if new investment opportunities arise, due to technical innovations, whereas government expenditure can increase as a political decision. On the other hand, with stable prices, the firms that do not introduce innovations cannot receive the stimulus arising from the means of production becoming cheaper. As a result, the process of growth tends to become more and more unbalanced, unless a general expansion of demand – originated by innovations and/or by government – takes the place of the stimulus afforded by an overall fall in prices.

In short, owing to the obstacles to entry, in most non-agriculture activities the ‘competitive mechanism’ for the distribution of the fruits of technical progress (falling prices, stable nominal incomes) has been more and more substituted by the ‘oligopolistic mechanism’ (stable prices and increasing nominal incomes). In the new conditions, the process of growth requires increasing intervention of public powers, but not necessarily in the form of increasing public expenditure. That intervention can consist of taxation (to afford incentives or to put brakes), or can support the prices and the incomes of those activities, like agriculture, least affected by those two processes, or can promote the source of technological innovation, that is, scientific research, or – to give another important example – can create conditions favourable to development of small firms, not only with fiscal and credit incentives, but also by supplying real services – especially commercial and technical assistance. All these measures of public powers can push up the growth of the volume of investment to the velocity required to avoid an increase of unemployment or gradually to reduce it to the frictional level.

If the countervailing influences of public interventions process are not strong enough, in the new conditions the process of growth tends to become more unbalanced not only from the standpoint of the different industries (since those that do not carry out innovations directly have no more the stimulus determined by the declining prices of the means of production they use), but also from the point of view of income distribution. In fact, the downward price rigidity tends to create special margins in certain industries or in certain firms. These rising margins do not necessarily become above-normal profits; they can become, too, above-normal wages or salaries, depending on the relative strength of the opposing parties. Instead of the above-normal incomes, the advantages for workers can also take the form of a greater stability of employment; similarly, the advantages of capitalists can take the form, rather than of above-normal profits, of more stable profits.

It seems that in recent times the process of differentiation has become more important than the process of concentration and the economies of specialization seem to have become more important than the economies of scale. This new development in industry has been promoted by at least three changes: (1) the growth of electronics and allied industries; (2) the reaction of increasing masses of workers in advanced countries against the monotony of assembly lines and other methods of mass production; (3) the growing differentiation in consumer preferences originated by the increasing per capita income. In services, differentiation has always been important and in recent times has become even more important; at the same time, services become the most important section of the economy in the so-called post-industrial societies. Considering the declining relative weight of agriculture and mining in advanced societies, we have to conclude that the area of flexible prices tends to shrink and that of rigid prices to expand – I mean flexibility or rigidity in the downward direction. In particular, the area of rigid prices tending to expand refers more and more to services and less and less to industry; this phenomenon, that has important consequences also on the overall behaviour of prices, up to now has received very little attention.

It remains true, however, that the increasing rigidity of prices of goods and services determines the need for an increase in demand large enough to avoid a decline in employment, if population grows. Now, with the diffusion of high education, with the space for a rapidly increasing number of goods opened up by the increasing per capita income, in recent times the potentialities of development have increased. But such potentialities can remain unexploited if they are left to spontaneous market forces; given the rate of interest, all depends on investment stimulated by technological innovations that promise to be profitable and that can be devised and carried out by private firms without the support or the stimulus afforded by public powers. If those investments are not enough to promote an increase of demand capable of generating an increase in income at least equal to that in productivity, unemployment gradually grows. It is well to emphasize that the main obstacles to a policy of economic growth arise not by diminishing returns, but either from the side of the public deficit, if the increasing supply of bonds pushes up the rate of interest; or from the side of the foreign deficit, which pushes up the value of foreign currencies, giving rise to a special kind of inflationary pressure. Such problems are aggravated by the fact that the two deficits, to some extent, reinforce each other: for instance, large firms tend to borrow abroad, owing to the high internal rate of interest. But these are matters that go beyond the limits of our theme.

See Also

Bibliography

  1. Andrews, P.W.S. 1949. Manufacturing business. London: Macmillan.Google Scholar
  2. Bain, J.S. 1956. Barriers to new competition. Cambridge, MA: Harvard University Press.CrossRefGoogle Scholar
  3. Baumol, W.J. 1967. Business behavior, value and growth. Rev. ed. New York: Harcourt Brace and World.Google Scholar
  4. Bhagwati, J.N. 1970. Oligopoly theory, entry-prevention, and growth. Oxford Economic Papers 22: 297–301.CrossRefGoogle Scholar
  5. Chamberlin, E.H. 1933. The theory of monopolistic competition: A reorientation of the theory of value. Cambridge, MA: Harvard University Press.Google Scholar
  6. Dunlop, J. 1938. The movement of real and money wage rates. Economic Journal 48: 413–434.CrossRefGoogle Scholar
  7. Eichner, A.S. 1976. The megacorp and oligopoly: Microfoundations of macrodynamics. Cambridge: Cambridge University Press.CrossRefGoogle Scholar
  8. Fellner, W. 1949. Competition among the few: Oligopoly and similar market structures. New York: Knopf.Google Scholar
  9. Galbraith, J. 1957. Market structure and stabilization policy. Review of Economic and Statistics 39: 124–133.CrossRefGoogle Scholar
  10. Hall, R.L., and C.J. Hitch. 1939. Price theory and business behaviour. Oxford Economic Papers 2: 12–45.CrossRefGoogle Scholar
  11. Kaldor, N. 1935. Market imperfection and excess capacity. Economica NS 2: 33–50.Google Scholar
  12. Kalecki, M. 1943. Costs and prices. Repr. in M. Kalecki, Selected essays on the dynamics of the capitalist economy 1933–1970. Cambridge: Cambridge University Press, 1971.Google Scholar
  13. Modigliani, F. 1958. New developments on the oligopoly front. Journal of Political Economy 66: 215–232.CrossRefGoogle Scholar
  14. Robinson, J. 1933. The economics of imperfect competition. London: Macmillan; New ed, 1969.Google Scholar
  15. Sraffa, P. 1960. Production of commodities by means of commodities. Cambridge: Cambridge University Press.Google Scholar
  16. Sylos-Labini, P. 1956. Oligopoly and technical progress. Trans., Cambridge, MA: Harvard University Press, 1969.Google Scholar
  17. Sylos-Labini, P. 1984. The forces of economic growth and decline. Cambridge, MA: MIT Press.Google Scholar

Copyright information

© Macmillan Publishers Ltd. 2018

Authors and Affiliations

  • P. Sylos-Labini
    • 1
  1. 1.