The Economics of Imperfect Competition and Employment

In Chapter 1 of the companion volume, I have traced Joan Robinson’s contributions to economic theory and philosophy. There, on numerous occasions, I referred to the paramount influence of Keynes and Kalecki on her. As time went by, she became more and more Kaleckian in her interpretation of Keynes and in her perceptions of the great intellectual upheaval associated with his name. As early as 1952, referring to Kalecki’s independent ‘discovery of the General Theory’, Joan Robinson (1952, p.159) wrote that ‘by now it is impossible to distinguish what one has learned from which’. In fact, writing in 1948, she (1951, 1978 p.134) admitted that when she refers to the General Theory

The purpose of this chapter is to suggest that the methodology of non-cooperative game theory, which is so prominent in partial equilibrium analyses of imperfectly competitive markets, can fruitfully be adopted to study traditional macroeconomic issues that have usually been approached via non-strategic general equilibrium models or treated solely as disequilibrium phenomena.

In his Mathematical Psychics (1881), Edgeworth demonstrated in so-called Edgeworth equivalence theorem that the outcome of an exchange economy where traders act cooperatively is identical to the Walrasian equilibrium of perfect competition where traders act non-cooperatively as price-takers, when the number of traders of the same type is infinitely large. Certainly it clarifies an implication of Walrasian economics and is a pioneering work of the recent theory of a large economy composed of infinitely many small traders. ¹When the number of traders is limited, Edgeworth argued, traders themselves cannot determine the outcome of an exchange economy cooperatively and the utilitarian principle should be introduced to arbitrate disagreement between traders. Since the infinity of identical traders is a sufficient condition in the proof of Edgeworth’s equivalence theorem, however, a natural question is whether it is also a necessary one. If it is not necesary, what is essential for Walrasian economies is not so much the infinity of the number of small traders as perfect information and no friction, also assumed in Edgeworth’s theorem.

During the past few years, imperfect competition seems to be coming into its own. Models of product differentiation, in particular, are now a standard vehicle in which to discuss many issues in industrial organization and in international trade. Yet this resurgence of interest in the area has failed to dispel a long-standing concern, shared by many economists, that imperfect competition is ‘empirically empty’-the particularly pungent dismissal of the theory in George Stigler’s Nobel address of 1982 is a recent echo of older debates.

Is there any diagram more familiar to economists than the following?

Analyses of industrial competition have attained a new vigor with the application of game-theoretic methods. The process of competition is represented in models that reflect genuine struggles for entry, market power, and continuing survival. Dynamics and informational effects are captured explicitly, although so far only in simplified formulations. Recognition of the importance and intricate complexities of competitive processes began a half-century ago in the work of Joan Robinson (1933), and the first game-theoretic formulations were developed by Shubik (1959) a quarter-century later, but the flowering of this approach began in the 1980s with the recognition that informational asymmetries are crucial ingredients to bring the flavor of struggle to the ensuing dynamic processes. Models that admit both dynamics and private information formulate competition as essentially a bargaining process in which credible communication is limited to costly actions. Each firm’s claim to survival is signalled by its willingness to offer lower prices longer than others. Firms’ struggles for the advantages of monopoly power bring benefits to consumers by dissipating a substantial part of the subsequent profits in the battle to obtain them. This is not the entire story of competition, of course, since also important are, for example, races for cost advantages, product development and differentiation, as well as imposition of search and switching costs on customers to sustain monopoly pricing; nevertheless, it brings theories that describe more realistically the Darwinian aspect of competition.

The aim of this chapter is to examine the effects of entry of new firms in a symmetric differentiated products industry, in particular the effect on the profits of firms already in the industry. Purposely, our models are kept close to those of a Chamberlinian industry, but with this exception: because of the implicit indivisibilities, it would be only by accident that profits are zero at free entry equilibrium (as they also are in Joan Robinson’s models). Typically, the entry of an additional firm would (under symmetry assumptions) result in negative profits for everyone, but if an additional firm does not enter profits at equilibrium are positive.

The modern theory of price discrimination began with the work of Pigou (1920). Joan Robinson devoted two chapters of her book The Economics of Imperfect Competition (1969) to the problem of (‘third degree’) price discrimination. Her account examines the conditions that make price discrimination possible, presents a graphical analysis of the discriminating monopolist’s pricing decision which has become the standard textbook treatment, and ends with an inquiry into the consequences of price discrimination for both allocative efficiency and distributional equity. Although Pigou’s and Robinson’s contributions have proved of lasting value, the theory of price discrimination has by no means remained unaltered.¹

Business practices often take advantage of the heterogeneity in the consumer population: product lines and price schedules are especially designed to discriminate among consumers on the basis of their taste and/or income diversity.

That there is at least some analogy between models of spatial and of monopolistic competition, or even that they are, in some fundamental sense, ‘the same’, has been commonly thought for many years. It was first suggested (to our knowledge) by Hotelling (1929); Chamberlin (see particularly (1957), pp.7, 23–4, 47–51, 124–9) appears to have regarded them as essentially the same; and many subsequent writers have presumed that a result obtained in one model would have a natural, if unspecified, twin in the other model. The purpose of the present paper is to investigate this presumption.

The nature of product differentiation in Joan Robinson’s Economics of Imperfect Competition remains unclear fifty-odd years after its publication. Her thinking was shaped by the cost disputes in the British literature of the 1920s that reacted to the recently formalized analysis of the firm in pure competition. Because, it was asserted, decreasing costs were experienced in industry more frequently than not, the limits to the size of the firm must be dictated by a declining demand curve, not rising marginal cost and horizontal demand functions. As Sraffa (1926) argued explicitly in the most influential article in the debate, the analysis of the firm’s decision-making must be reformulated in a universal theory of monopoly. Implicitly, the decreasing cost advocates urged that if the inspired fuzziness of Marshal-lian economics had to be formalized, the monopoly model, not that of pure competition, was the more relevant tool. Robinson’s debt to Sraffa is amply recorded in her book.¹

This essay is related to two topics on which Joan Robinson¹ made important contributions. The first is the theory of imperfect competition. At the time it was written, The Economics of Imperfect Competition was heralded (along with Chamberlin’s contemporaneous book The Theory of Monopolistic Competition ²) as doing for microeconomics what Keynes had done for macroeconomics. But the revolution to which it was supposed to give rise never occurred, and it was almost forty years before the themes she developed there again became the focus of theoretical research.³

The empirical observation made by most students of credit markets is that rationing is being practiced in those markets. The early developers of the theory of imperfect competition, including Joan Robinson, viewed this as a reflection of monopolistic forces in the credit markets. Research over the past two decades has been suggesting that the credit market possesses some unique features which differentiate it from other markets and this perspective has altered our view of it. Two lines of research have been developed. One observed that lending institutions necessarily need to take the risk of default by their borrowers, and for this reason the interest rate charged by them depends upon the characteristics of the borrower and the amount borrowed. Thus, an upward-sloping interest rate schedule is nothing but a reflection of the ever-present risk of default rather than the imperfection of the credit markets (see Freimer and Gordon, 1965; Stigler, 1967; Stiglitz, 1970; Jaffee, 1971; Baltensperger, 1976; Kurz, 1976; Keeton, 1979; and others). A second line of research focused on the problems generated by the presence of asymmetric information in this market, resulting in either adverse selection of borrowers or in the existence of an incentive for borrowers to take greater risk of default than desired by the lenders (see Jaffee and Russell, 1976; Stiglitz and Weiss, 1981; Hammond, 1986).

Is there a paradox in the economics of Joan Robinson in the post-Keynesian period?¹ Apart from Richard Kahn, she worked with Keynes more closely than anyone else during the years of preparation of the General Theory of Employment, Interest and Money, a book which explains unemployment in terms of money and interest and which Keynes initially entitled ‘The Monetary Theory of Production’ (Keynes, 1979, p.49). In the early post-General Theory years, Mrs Robinson published several books and numerous journal articles expounding, with Keynes’s approval, the doctrines of the General Theory. In her published writings she was always a loyal follower and defender of Keynes’s theory. Yet in the post-Keynesian phase (after 1946) of her career money occupies no significant place in her extension of the General Theory to capital accumulation, economic growth, and the distribution of income. There appears to be a paradox in the sense that, after deep immersion in the monetary economics of Keynes, the absence of money as a strategic factor in her theory runs counter to expectations. The present chapter examines this apparent paradox. Section 1 deals briefly with the period prior to the preparation of the General Theory; Section 2 with the preparation and early aftermath of the General Theory; Section 3 with the post-Keynesian period following Keynes’s death in 1946; and Section 4 contains some general observations and conclusions.

Joan Robinson never turned her attention to the writings of the school of new classical economists. It would not be difficult, however, to imagine the nature and tone of a hypothetical essay on such a topic. Her view that the inflation process is largely driven by the evolution of money wages which, in turn, is predominantly influenced by social and political factors, led to a strong rejection of the natural rate hypothesis. The self-conscious application of Walrasian theory which characterizes the new classical economists (not to mention their appellation) would have been particularly galling to her with her well-known views on the sterility of this class of analysis.

A well-established feature of the business cycle, which has been known for a long time, is the high correlation between output and the variation of inventories (for some recent evidence, see, e.g., Chanut and Laroque, 1979; Blinder, 1985). It is interesting to contrast the interpretations that the two main strands of macroeconomic theory give of this stylized fact.

To even the most casual reader of the daily business pages, it must be fairly obvious that financial markets react (or overreact) to much news which clearly (or, at least, arguably) has little or no direct bearing on ‘fundamentals’. Some prominent, current items: the threats issuing from various Middle Eastern factions; the Fed’s weekly announcement of previous changes in the money supply; the outcome of this fall’s Congressional elections; the great deficit debate. Contrary to one’s instinctive reaction as a trained economist, several recent studies have convincingly demonstrated that such market behavior is not necessarily ‘non-economic’ or ‘irrational’. Prices and returns-and hence market allocation itself-may very well consistently depend on purely expectational phenomena. Moreover, self-fulfilling beliefs about market forces may have significant consequences for individual welfare. In short, ‘sunspots’ may matter, and they may matter very much.¹

Contemporary macroeconomists when asked about points of theory or details of policy, rarely answer with a single voice: the extraordinary diversity of their views is as dismaying to professionals as it is to the public. On only one matter do they all join in the same chorus: lower interest rates would speed up economic growth, reduce unemployment, ease the burden that debtors must now bear and probably reduce the country’s trade deficit. Recently, of course, all interest rates from those on overnight loans to those on terms of twenty-five years or more have fallen but they remain very high. Imagine how Keynes, who had anticipated with pleasure further declines in rates from the very low levels of 1935 and a hastening of the ‘euthanasia of the rentier’, would feel if he knew that in February 1986 long-term corporate bonds in the United States yielded 9.67 per cent (and only 3.4 per cent in 1935) and ninety-day treasury bills, 7.06 per cent (as against 2.25 per cent in 1935).

Joan Robinson was appropriately contemptuous of the notion that unemployment could be explained by the stubborn refusal of workers to accept the market real wage. In one memorable passage, she wrote,

[T]he orthodox conception of wages tending to equal the marginal disutility of labour, which has its origin in the picture of a peasant farmer leaning on his hoe in the evening and deciding whether the extra product of another hour’s work will repay the extra backache, is projected into the modern labour market, where the individual worker has no opportunity to decide anything except whether it is better to work or starve. (Joan Robinson, 1966, p.2)

What are the Questions? Where Do We Go From Here? What Now? Such queries decorate Joan Robinson’s later policy essays. Yet the point is critical, not constructive, and one looks in vain there for a program of her own design. For the reader seeking to be led, it is slightly frustrating. Evidently, there is work to do.

Joan Robinson was not only interested in economic theory for its own sake, but was deeply concerned with human welfare in all its aspects. She sought to improve the human condition through a better understanding of the interactive forces that move human society. In addition to problems of economic development and fair distribution of income and wealth, she was also particularly concerned with the squandering of resources in the arms race and with ways to reduce international tensions and avoid war.