A Kaleckian Theory of Income Distribution Author(s): A. Asimakopulos Source: The Canadian Journal of Economics / Revue canadienne d'Economique, Vol.

8, No. 3 (Aug., 1975), pp. 313-333 Published by: Wiley on behalf of the Canadian Economics Association Stable URL: http://www.jstor.org/stable/134236 . Accessed: 17/06/2013 14:14
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Wiley and Canadian Economics Association are collaborating with JSTOR to digitize, preserve and extend access to The Canadian Journal of Economics / Revue canadienne d'Economique.

http://www.jstor.org

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

A Kaleckiantheoryof income distribution

A Kaleckian theory of income distribution. This paper draws together the various elements of Kalecki's analysis of income distribution. It is argued that what a higher degree of monopoly makes possible and protects is the rate of return of the main firms in an industry. This degree of monopoly is reflected in the mark-ups over unit prime costs used to set prices. Employment and the level of profits are determined in this model by capitalists' expenditures. A key assumption is that a higher proportion of profits than of wages is saved. With overhead labour, changes in effective demand, as well as changed mark-ups, affect the share of profits. Une theorie kaleckienne de la distribution du revenu. L'article se veut une synthese de differents elements de l'analyse de Kalecki sur la distribution du revenu. Les accents donnes a certains aspects particuliers de sa theorie ne sont pas necessairement ceux que Kalecki lui-meme a donnes. Ses formulations des equations de prix sont deficientes. On pretend ici qu'un degre plus grand de monopole rend possible et protege non pas le rapport entre le prix et les cofuts de base, mais la rentabilite generale des entreprises d'une industrie. Les prix sont determines par la majoration des cofuts de base unitaires. Ces majorations sont affectees par le degre de pouvoir monopolistique, de meme qu'ils peuvent I'etre par le pouvoir syndical. Les composantes de cette theorie sont les relations de courte periode, une fois donnee la capacite de production et determine l'investissement par les decisions des periodes anterieures. L'emploi et le niveau des profits sont determines par les depenses de capitalistes. On peut tenir compte de l'epargne des travailleurs sans modifier l'orientation generale du modele, tant que la proportion epargnee des profits est plus grande que la proportion epargnee des salaires. Les changements de la demande effective peuvent influencer la part des profits, meme avec des majorations donnees, a cause de l'existence d'une quantite fixe de travail. Le caract6re kaleckien du modele developpe ici est teste par son utilisation dans l'examen des questions posees par Kalecki dans son article, publie de fagon posthume, sur 'La lutte des classes et la distribution du revenu.' I am grateful to J.B. Burbidge,P. Davenport, H. Flakierski, J. Henry, and S. Ingermanfor comments on an earlier draft of this paper. They are not, of course, responsiblefor any errors or for this interpretationof Kalecki's writings. The first drafts of this paper were written while I was a visiting professor at Monash University.
Canadian Journal of Economics/Revue canadienne d'Economique, VIII, no. 3 August/aofit 1975. Printed in Canada/Imprime au Canada.

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

314 / A. Asimakopulos
INTRODUCTION

Michat Kalecki's writings contain a theory of distribution that combines microeconomic and macroeconomic aspects of the economy. Income shares are influenced both by the mark-ups firms are able to establish in oligopolistic markets and by the level of effective demand. A higher level of effective demand would lead to a higher share of profits even with the mark-up unchanged, if there is overhead labour in the model. This point was made in a paper Kalecki published (in Polish) in 1933.1 Discussions of Kalecki's theory of distribution have centred on his 'degree of monopoly' and have ignored the role of overhead labour in providing a means by which changes in effective demand can influence income shares even when mark-ups are constant.2 Kalecki also tended to abstract from overhead labour in his later writings.3 The publication of Kalecki's Selected Essays on the Dynamics of the Capitalist Economy (1971), which he characterized as including 'what I consider my main contributions to the theory of dynamics of the capitalist economy' (vii), provides a convenient source from which to construct a Kaleckian theory of distribution. One of the important differences between Kaldor's 'Keynesian' theory of distribution and Kalecki's is that the former is restricted to full employment situations, while the latter is not. Kaldor (1955-6; 94) makes a distinction between 'short-run theory' and 'long-run theory' and wants to use the multiplier principle to explain variations in output and employment in the former but as a distribution theory in the latter. Kalecki denies that such a distinction between short- and long-run theory is possible. 'The long-run trend is but a slowly changing component of a chain of short-period situations: it has no independent entity' (Kalecki, 1971, 165). The building blocks for this Kaleckian theory of distribution will thus be the short-period relations. Kalecki's presentation of the degree of monopoly to explain the determination of mark-ups has been criticized as no more than a tautology, simply defining the ratio of price to prime costs as the degree of monopoly.4 There are many places where Kalecki's language seems to invite this interpretation, but 'no! problem of tautology is involved' (Kalecki, 1971, 168, em1 The paper'stitle in the English translationprintedin Kalecki ( 1971) is 'Outline of a theory of the business cycle.' The assumptionappears in a footnote: 'We assume here that aggregateproduction and profit per unit of output rise or fall together, which is actually the case. This results at least to some extent from the fact that a part of wages are overheads' (ibid., 1ln.). 2 An importantexception has been Professor Joan Robinson (Robinson, 1969), who made use of overhead, or indirect, labour in her reformulation of Kalecki's pricing equation. 3 See for example, chap. 15, 'Trends and business cycles' in Kalecki (1971), which was first published in 1968. 4 This criticism is made by, among others, Kaldor (1955-6) and Nuti (1970).

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

Income distribution / 315 phasis in the original) if the ratio of price to prime costs in a given situation is taken as a reflection of the degree of monopoly rather than its definition. This paper draws together the various elements of Kalecki's analysis of distribution. In the process some of his formulations, particularly the price equations, are amended. It is emphasized that what a higher degree of monopoly makes possible, and protects, is the higher profitability of firms in the industry. This is not, in general, as Kalecki's use of the term implies, the same as the ratio of price to prime costs. Over time, and in comparison with other industries, a higher degree of monopoly would enable the main firms in the industry to obtain higher rates of profits. It is thus the realized rate of return on investment which is a reflection of the degree of monopoly.5 The approach presented here. assumes that estimates of these rates of return, used for purposes of planning, are included in the mark-ups applied to unit prime costs in order to arrive at prices. Because of the amendments to Kalecki's presentation the model presented here is better described as a 'Kaleckian theory' rather than 'Kalecki's theory.' A test of its Kaleckian nature is provided by using the theory to answer the questions he posed in his posthumously published paper, 'Class struggle and the distribution of national income' (reprinted in Kalecki, 1971).
KALECKI S PRICING EQUATIONS

Kalecki distinguished between two types of short-term price changes, 'costdetermined' and 'demand-determined.' 'Generally speaking, changes in the prices of finished goods are "cost determined" while changes in the prices of raw materials inclusive of primary foodstuffs are "demand determined"' (Kalecki, 1971, 43). Supply conditions are given as the reason for this difference. The production of finished goods is elastic as a result of existing reserves of productivecapacity. When demand increases it is met mainly by an increase in the volume of productionwhile prices tend to remain stable. The price changes which do occur result mainly from changes in costs of production. ... The situation with respect to raw materials is different ... With supply inelastic in short periods, an increase in demand causes a diminutionof stocks and a consequent increase in price ... A primaryrise in demandwhich causes an increase in prices is frequently accompanied by secondary speculative demand. This makes it even more difficult in the short run for production to catch up with demand (ibid., 43-4). Kalecki's analysis of prices and his theory of distribution are concerned with finished goods. For raw materials he simply states: 'the relative share
5 Riach ( 1971) has argued that Kalecki's 'degree of monopoly' should be reformulated in this way. In the same spirit Joan Robinson's reformulation of Kalecki's pricing equation (Robinson, 1969) makes use of an expected rate of return on investment.

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

316 / A. Asimakopulos of wages in the value added depends mainly on the ratio of prices of the raw materials produced to their unit wage costs' (ibid., 64, emphasis in original). He does not go on to consider the factors tending to change unit wage costs in these industries, but it is consistent with statements made in other connections (e.g. ibid., 56, 73) to infer that they include the degree of organization of workers and the power of trade unions. Unit prime costs (composed of wages and raw material costs) in plants in manufacturing industries are assumed to be constant for a substantial range of output, and then to increase sharply when normal productive capacity is reached. They can thus be represented by reverse-L-shaped curves. Firms in these industries typically sell in oligopolistic markets and, at the prices they set, their plants generally operate at less than normal capacity. Unit prime costs are thus assumed to be constant in the range of output Kalecki considers. In setting prices firms are faced with conditions of uncertainty, with respect to both the immediate and longer-term consequences of their decisions. The options open to them are influenced by the decisions of other firms, which in turn are not independent of their own, and the prices they set now may affect their prospects in future periods. There is no simple demand curve, even for a price leader, that can form the basis for the maximization of profits by bringing marginal revenue and marginal costs to equality. Even though firms in Kalecki's model are constantly striving for profits, in view of the general atmosphere of uncertainty facing them 'it will not be assumed that the firm attempts to maximize its profits in any precise sort of manner'6 (ibid., 44). Kalecki's approach to the pricing of manufactured goods is an important element - in the Kaleckian theory of distribution developed here, but his particular formulations of the price equation are discarded. He assumed that a firm arrives at a price for one of its products by marking up its unit prime costs in order to cover overheads and achieve profits. These markups reflect, in his view, the 'semi-monopolistic influences ... resulting from imperfect competition or oligopoly' (ibid., 160). The more imperfect the market conditions faced by the firm, the stronger its position vis-a-vis other firms in the industry and potential entrants in terms of costs, product differentiation, etc., other things given, the higher the firm's mark-up. Two pricing equations appear in Kalecki (1971), both trying to relate a firm's price to its prime costs and average price in the industry. The earlier formulation (first published in its final version in his Theory of Econom,ic Dynamics, 1954) is p -mu + np, (1)

where p is the firm's price, u is unit prime cost, m and n are positive coeffi6 Kalecki's final version of his theory thus does not make use of the elasticity of the individual firm's demand curve.

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

Income distribution / 317 cients, and p is the weighted average price of all firms. The mark-up on prime costs (p - u) /u as determined by this equation is (p-u)/u = (mr) + np/u. (2)

The values of the two parameters m and n 'reflect what may be called the degree of monopoly of the firm's position' (Kalecki, 1971, 45). Kalecki combines them into a single measure mi(1 - n). He argues that different firms in the industry would have different values for these parameters and thus set different prices. Kalecki's attempt to allow for both the interdependence of the firms' pricing decisions and for the special features of the position of a particular firm is commendable, but the explanatory power of his equation is very limited. At times Kalecki writes as though firms actually arrived at prices on the basis of this equation. 'The influence of the emergence of firms representing a substantial share of the output of an industry can be readily understood in the light of the above considerations. Such a firm knows that its price p influences appreciably the average price p and that, moreover, the other firms will be pushed in the same direction because their price formation depends on the average price p. Thus, the firm can fix its price at a level higher than would otherwise be the case' (ibid., 50). This is much too simplified a representation, even at this level of abstraction, of the forces determining the prices charged by firms in these market structures. It is true that a firm is more likely to raise its price in a given situation if other firms would follow its action, but this following of a price rise is not the mechanical result of the initial change in average price. In the absence of price agreements it could be due to a similar view of the profit potential of a higher price, or to the fear of starting a price war, or to some combination of these factors. There are also technical problems with Kalecki's pricing equation. He deduces that n < 1 from the observation that 'where the price p of the firm is considered equal to the average price T we have p -mu + np, from which it follows that n must be less than one' (ibid., 45). But this conclusion does not necessarily follow. In situations where price is set by a price leader and followed by others, n would be equal to one for the price followers and thus m would be equal to zero.7 The ratio of proceeds to prime costs for a price follower would thus be equal to the ratio of the prevailing industry price to its unit prime costs. For the price leader, on the other hand, n would be equal to zero and the value of m would depend on many (interdependent) factors, such as its view of an appropriate rate of return
7 This conclusion also holds when there is product differentiation.The price equation for a price follower would still have m equal to zero, with pf - pj + d or pt = (1 + d)pl, where pt and p, representthe prices of the follower and the leader and d is the recognized price differential, expressed either in absolute terms, or as a ratio, whichever is appropriate.

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

318 / A. Asimakopulos on its capital over time, pressures from other firms in the industry and from potential entrants, the rate of profits tax, etc. A reformulation of Kalecki's equation for the mark-up appeared in a paper that was published posthumously in 1971.8 Instead of the linear form of equation (2), he wrote (p
-

u)/u = f(/p),

(3)

or, rearranging, p - u[1 + f(P/P)]. (4)

His attempt to allow for the factor of competition between firms by relating its price to the weighted average price for the industry as a whole, by means of the function f, does not succeed. This function has no explanatory power; it does not provide a causal relation between a firm's mark-up and average industry price. Kalecki wrote that 'f is an increasing function: the lower is p in relation to p, the higher will be fixed the mark-up'9 (ibid., 160). This observation is a non sequitur. For the firm in a given situation both u and pf are given (the effect of a higher value of p on p may be ignored for this illustration) and therefore the higher the mark-up, the higher p, and the lower p/p. Contrary to Kalecki's statement, the function f appears to be a decreasing function! This conclusion follows from the definition of these terms and points up the tautological nature of Kalecki's equation (3). Kalecki tried to distinguish between his theory and that of the 'full-cost pricing' theoiry (see Hall and Hitch, 1939), on two grounds. His 'emphasis on the influence of prices of other firms' and the belief that 'the degree of monopoly may, but need not necessarily, increase as a result of a rise in overheads in relation to prime costs' (ibid., 51, italics in the original) were held to form the basis for this distinction. The former claim can be dismissed because of his failure, noted above, to incorporate this influence in any way that, on examination, is different from the general interdependence of firms' pricing policies common to, all fully specified mark-up theories. The latter claim would only hold if his degree of monopoly was interpreted as measuring the ratio of proceeds to prime costs, rather than as indicating the imperfections in competition that permitted firms to achieve high rates of profits over time. Kalecki's writings on this point are confusing; there is a tendency to identify the degree of monopoly with the ratio of proceeds to
8 The paper, 'Class struggle and distributionof national income,' first appearedin Kyklos in 1971. It is reprinted (with some typographicalerrors) in Kalecki (1971). 9 There is a misprintin the text of Kalecki (1971): the bar has been omitted from the second T.

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

Income distribution / 319 prime costs, not only in the short-run with given productive capacity and overheads when it is synonymous, under certain conditions,0 with higher profitability, but also over long periods of time. In his examination of data on the ratio of proceeds to prime costs in manufacturing industries over historical periods he all but identifies the degree of monopoly with this ratio. He does qualify his comments, however, noting that 'the interpretation of the movement of the ratio of proceeds to prime cost in terms of changes in the degree of monopoly is really the task of the economic historian who can contribute to such a study a more thorough knowledge of changing industrial conditions' (ibid., 56). Whatever the judgment reached on Kalecki's use of the term 'degree of monopoly,' a Kaleckian theory of distribution can be developed on the assumption that it acts to protect the rate of profits of established firms in the industry.
THE DETERMINATION OF THE MARK-UP

The theory of distribution presented here is based on the assumptions that prices in manufacturing industries are generally arrived at by marking up unit prime costs and that these mark-ups are relatively stable in the face of short-term fluctuations in demand and output, at least within some substantial range of output." These mark-ups are designed to cover, over time, both overhead costs and profits. Their values would thus be dependent on the standard rates of utilization of productive capacity used to calculate standard costs'2 as well as on some expected rate of return. The rates of return built into the mark-ups are affected by what is considered to be a 'normal' return to industrial investment in the particular economy examined. In addition, those areas of industry with some degree of monopoly arising from seller concentration and barriers to entry can earn, on the average,
10 A higher price, relative to prime costs, will result in higher profits in the short period as long as it is not accompaniedby sales that are sufficientlylower to outweigh its beneficial effects on profits. More formally, we can state that a higher mark-upwill lead to higher profits in the short period if the value for the elasticity of employment of direct labour with respect to the mark-upis numerically less than one. 11 For empirical supportfor the hypothesis of mark-uppricing in manufacturing industries, see Nordhaus and Godley (1972). 12 Standardoutput is less than normal productive capacity output. Lanzillotti found that for the firms in his sample 'the standardspremised on an assumed rate of production, typically about 70 per cent to 80 per cent of capacity' (Lanzillotti, 1958, 923, n.5). Firms in Kalecki's model are assumed to plan on having some excess capacity - they provide for more than their expected rates of output. This feature is also prominent in Kaldor's models (see, for example, Kaldor, 1970, 4, and Kaldor and Mirrlees, 1962, para. 3). What firms take as standard rates of output, and thus mark-ups,would change over time with changes in the experienced rates of utilization of plant.

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

320 / A. Asimakopulos more than these normal rates of profits.'3 Provision for such returns become formalized in their mark-ups in the guise of some target rate of return.14 The 'normal' rate of return for a given time and economy is, in Kalecki's approach, 'rooted in past economic, social and technological developments."-5It very much depends on the profits that firms have been able to earn in the past and on the profit share in total income. Profits are determined in this model, as we shall see, by the investment that firms have been able to achieve in conjunction with conditions of thriftiness. Both investment and the profit share may be affected by bargaining with trade unions (the 'class struggle'). Rising money-wage rates may provoke monetary restraints which hamper investment and may lead to some erosion of markups and the share of profits. There is another feature of Kalecki's analysis that indicates a relationship between the growth possibilities for a firm and the rate-of-return portion of its mark-up. Kalecki's analysis is based on what he considered to be the characteristic features of the capitalist mode of production. The means of production are owned by capitalists who, in a closed system with government activity excluded, do all the saving. The saving of workers in a capitalist economy would be small in comparison to that of capitalists and is ignored in his writings. Kalecki distinguished between two types of capital, entrepreneurial capital and rentier capital.'6 The former is capital owned by the firm while the latter is the capital it tries to borrow. He wrote that 'the access of a firm to the capital market, or in other words the amount of rentier capital it may hope to obtain, is determined to a large extent by the amount of its entrepreneurial capital,' and 'the expansion of the firm depends on its accumulation of capital out of current profits' (ibid., 105, 106). A firm's ability to grow thus depends on the profits it can generate to finance its investment plans both directly (retained earnings) and indirectly through borrowing related to its internal funds. Given the general profitability of the economy, a firm's profit expectations are
13 Some empirical findingson the positive relationshipbetween rates of return on one hand, and seller concentrationand barriersto entry on the other, are to be found in Bain (1956) and Mann (1966). 14 For some indication of the importanceof target rate of return pricing, see Lanzillotti (1958) and Kaplan, Dirlam, and Lanzillotti (1958). 15 Kalecki, 1971, 183. These terms are used by Kalecki in referringto his approach to the rate of growth of an economy at a given time. They are also appropriatein describing a Kaleckian theory of pricing since it is an importantelement in a theory of growth. In a particularshort period various factors, such as views on what constitutes a normal rate of return, as well as available productivecapacity are taken as given. However, over a sequence of short periods, that is, over time, these views may be altered by experience. A Kaleckian model provides a framework for analysing a particulareconomy, but for the analysis to be completed the institutionalfeatures and history of that economy are required. capital and investment'in 16 See in particularthe chapter on 'Entrepreneurial Kalecki (1971, 105-9). The first version of this chapter was published in 1937.

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

Income distribution / 321 dependent on the degree of monopoly it enjoys and are indicated by its target rate of return. Where growth is an important goal for a firm, its target rate of return can serve as an indicator of its growth possibilities, that is, the rate at which it can increase the value of the capital assets under its control.'7
MACROECONOMICS AND THE LEVEL OF ABSTRACTION

The theory of distribution developed here is a macro-theory, concerned with aggregates, total investment, total consumption, and total output. It is thus very far removed from the action of individual decision-making units that decide on their own level of investment, consumption, prices, output, and so on, even though it tries to justify the aggregate relations it uses on the basis of their consistency with particular types of individual behaviour. Macroeconomic analysis thus proceeds at a very high level of abstraction and such variables as the difference in prices charged by firms in the same industry because of product differentiation, which Kalecki emphasized in setting up his price equations, are generally ignored. Product differentiation can still have a role in this model by its influence on the mark-ups that firms use to arrive at prices. Consolidation does not stop at
17 An alternative explanation for the determinationof the mark-up in oligopoly is contained in an interesting paper by Eichner (1973). He explains changes in the mark-upas responses to the demand for additional investmentfunds. There is a cost of generatingmore funds for investmentfrom the higher immediate profits that result from a higher mark-upbecause of the substitution over time by consumers of other products, the entry of other firms into the industry, and possible government intervention.Eichner expresses this implicit cost of the additional investment funds generated in this way as an 'interest' payment. This enables him to 'derive a supply curve for additional internal funds ... indicatinghow the implicit interest rate ... on these funds varies as the amount of additional funds obtained per planning period ... varies' (ibid., 1193). A supply curve for external funds is added to this function to obtain the total supply function for additional investmentfunds for the firm. The changes in the firm's investment and mark-up are then determinedby bringing in its demand curve for additional investmentfunds, which 'is simply the familiar marginal efficiency of investment curve' (ibid., 1190). The results obtained from Eichner's approach and that presented here would often be similar even though they give opposite signs for the direction of causation between the rates of return and growth. In this model a higher mark-up,due to a greater degree of monopoly, would permit a firm that is fully committing its resources to expansion to grow at a faster rate. In Eichner's model the desire for a higher growth rate, given the degree of monopoly, would lead to a higher mark-up.It would often be difficultto disentanglethese two elements from data on growth and profit rates. Another differenceis that Eichner'sapproachis not designed to explain 'the absolute price level but rather the change in the margin above costs from one pricing period to the next' (ibid., 1195). The use of a marginal efficiencyof investment schedule in a Keynes (or Kalecki)-type model has been criticized in Asimakopulos (1971). See also Shackle (1967), especially chap. 11, and Robinson (1962; 1965).

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

322 / A. Asimakopulos this point, industries themselves are combined in very broad aggregates, such as 'investment-goods' and 'consumption-goods.' Again, some of the aspects of the underlying reality, the large number of industries comprising one of these sectors, can be introduced indirectly by indicating the effects on the aggregate in question of some of the diversity they tend to obscure. For example, in explaining the possible power of strong trade unions to erode mark-ups, Kalecki argues that an industry facing much higher wage rates and attempting to maintain mark-ups would find 'its product more and more expensive and thus less competitive with products of other industries ... trade union power restrains the mark-up."8 In this paper the degree of aggregation will be carried one step further, and no distinction will be made between investment-goods and consumption-goods production sectors. The effects on our results of introducing two production sectors along these lines will be noted. The raw-materialsproducing sector will also disappear from view;19 it is not an integral part of Kalecki's theory of distribution, which is largely concerned with manufacturing industries. Firms will therefore be assumed to be fully integrated, producing all the materials required for their final output, and prime costs will thus be made up only of labour costs. Output will be assumed to consist of a multipurpose good which can be consumed or invested as capital equipment or as inventories. As soon as it is set aside for formation into capital equipment its characteristics are assumed to change. It is formed into capital equipment, with no further expenditure of labour, by accumulating this good over time.20 One final element of abstraction concerns the shares of the different plants in total output. These shares will be taken as given in any short period and will not be explained by the analysis.21 They will be represented
18 Kalecki, 1971, 161. He comments in a footnote: 'Despite the fact that for the sake of simplicity, we assumed that all wage rates are raised simultaneously in the same proportion,we consider realistically that bargainingis proceeding by industries' (ibid.). 19 It should not be necessary to emphasize that this and the exclusion of internationaltrade and governmentalactivities are importantomissions. The present model can best be viewed as a 'basic' one, to which these other features should be added. Some aspects of governmentactivity, within the context of this type of model, are examined in Asimakopulos and Burbidge (1974). 20 This device of multipurposeoutput was used in Asimakopulos (1969; 1970) and again in Asimakopulos and Burbidge (1974). 21 The differentplants are thus assumed to be operated in parallel fashion, even though they may contain equipment of different'vintages'that give rise to differences in labour productivity.This results in a perfectly elastic supply curve for output at the established price, without the need to assume that the marginal costs are constant and equal for each plant (cf. Davidson, 1960, 53). That plants of differingefficiency share in total output through imperfections in competition that offer protection for the market shares of firms owning these plants can be inferred from the data discussed by Kaldor (1970). He noted that 'the share of gross profit in value added shows a very"highvariation as ' (2). He also noted that 'the share of profits in between different"establishments"

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

Income distribution / 323 by the relative amounts of direct labour employed by each plant. These direct-labour employment ratios will be the same for all comparisons made with respect to any particular short period.
THE MODEL

The following notations will be used: w a L1 L1 L,, L p p Y W

C A X S ! I

money-wage rate, average output per unit of direct labour, level of employment of direct labour, maximum possible employment of direct labour with existing plant, level of employment of overhead labour, level of total employment, price of the output, mark-up over unit prime costs, value of gross output, the total wage bill, value of total gross profits, value of total consumption, constant element in capitalists' consumption in real terms, small fraction expressing dependence of a portion of capitalists' consumption on current profits, value of gross saving, value of gross investment, and gross investment in real terms.

The unit prime costs of the ith plant, in the range of employment where they are constant (L1l < Lf1), are equal to w/ai. Average output per unit of direct labour in the economy a is equal to 1q = 1 aiL1ilL1, where n is the total number of plants utilized. Our assumption about output shares means that Lv/L, is assumed to be unchanged in all the short-period comparisons. Average unit prime or direct costs can thus be represented by w/a. It is assumed that the price is arrived at by a price leader setting a mark-up over unit prime costs and that other firms follow this price. The ratio of this price to average unit prime costs is stable in the face of short-term fluctuations in demand and output. The mark-up over this average unit prime cost u is equal to p/[(w/a) - 1], so that we can write p= (1 + Ft)(w/a). (5)

relation to the value added shows the same kind of scatter among "firms"as it ... It is perfectly possible, therefore, that the does among "establishments" "marginalequipment"of the least efficient firm is one that is considerablyolder than the "marginal" equipment of a high profit firm' (3).

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

324 / A. Asimakopulos There are thus three elements in the determination of the price level in this model: the mark-up, the money-wage rate, and the average output per unit of direct labour. The latter is given by present-day operating knowledge and the technical features of the plant and equipment inherited from investment in earlier time periods and utilized in the present. Money-wage rates are set as a result of bargaining between employers and trade unions. This wage level is not explained by the model, it is taken as given, and the cffects of different values for it will be considered. There is a banking system in this model, even though it does not appear explicitly. It provides the money supply, holds deposits of households, and lends working capital to firms. Banks' interest income represents a share of total profits and is assumed to be remitted to their shareholders. When different levels of output and prices are considered the supply of money by the banks is assumed to be 'elastic at the ruling rate of interest.'22Firms can generally obtain finance for their planned investment. Credit restraints, and the inability to achieve planned rates of investment, only arise in inflationary situations. The remaining features of this model can be quickly sketched in. The value of output produced in the particular short period (e.g. three months or a year) examined is Y = p a LI, L, < L1. (6)

It will be assumed that Keynes's short-term expectations, which determine the degree of utilization of existing plant, are always borne out by events because they can be quickly adjusted to conform to actual conditions. Thus the value for the amount produced in the particular short period turns out to be equal to the value of the amounts demanded for consumption and investment, at the set price. The value of total gross output produced is divided on income account into two categories, wages and gross profits.

Y = W+
W = wL.

(7)
(8)

Total employment L is made up of two categories, total direct labour L1 which appears in equation (6), and total overhead or indirect labour Lo. In order to operate a plant at any non-zero degree of utilization, a certain number of workers are required (for the ith plant this number would be represented by Lo, and for all plants, Lo = I,.'-1 L0I). This number, unlike that of direct labour employed, is independent of the degree of utilization of the plant (as long as it is utilized). Therefore, our assumption of constant output per unit of direct labour in the output range considered
22 This assumptionis implicit in many of Kalecki's papers (for example, Kalecki, 1971, 13 and 159n.).

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

Income distribution / 325 means that labour productivity, output per unit of total labour employed, is an increasing function of total output in this range. LrLo+Li, (9)

where L < total labour force. The one time-lag that appears in all of Kalecki's writings on effective demand and distribution is that between investment decisions and investment (see Kalecki, 1971, 2, 166), and it appears in this model. On the other hand he occasionally abstracts from the lag in consumer expenditure,23 and it will be ignored here. The effect of the latter assumption is to allow the multiplier to take its full value in the short period in which investment is increased. Kalecki abstracts from 'workers' savings, which are definitely unimportant' (ibid., 166), while capitalists' consumption contains an important element that is 'a slowly changing magnitude dependent on past economic and social developments' (ibid., 167) as well as another element that is proportionate to profits. The equation for the value of total consumption expenditures will be written here as C= W+ XAI +pA. Gross saving is thus equal to S (- P1) (1-A X) -pA. (10)

Investment in real terms in the short period is exogenous, determined by investment decisions made in an earlier period subject to two possible constraints. One is that planned investment plus the associated consumption do not exceed the normal productive capacity of the available plants or require more labour than can be provided by the labour force, and the other is that the 'inflation barrier' is not reached. This constraint would operate if pressures in the labour market led to rapidly rising wages and prices, to which the monetary authorities respond by imposing severe credit restrictions that would prevent firms from achieving their planned investment. 7= pI. (12)

Finally, we have the necessary equality between saving and investment: I - S. (13) Gross saving, as given by equation ( 11), is desired saving, while gross investment in equation (12) is intended investment, and thus equation
23 In making this assumptionhe notes that 'this ... is realistic with regard to workers' consumption, but not so with regard to that of capitalists. However, as long as the time-lag between investment decisions and investment is emphasized, disregardingthat between profits and capitalists'consumption does not distort the analysis' (Kalecki, 1971, 166).

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

326 / A. Asimakopulos (13) indicates not only the necessary ex post equality between saving and investment but a position of short-period equilibrium as well. There are nine unknowns (p, Y, LI, W, H, L, C, S, I) in this model. We take as given the values for w, a, j., Lo, X, A and I. Solving for the variables of interest, we find that profits for the period can be derived by combining the last three equations and rewriting: T (1 + 1)W((I + A) a(l -A) (14)

Equation (14) rearranges an equation (II = pl + XIT + pA) that expresses Kalecki's conclusion with respect to the determination of profits. Profits are determined by the value of capitalists' expenditure - their gross investment and consumption.24This conclusion reappears in all his writings on distribution.25Profits in real terms are not a function of the mark-up. This can be readily deduced from (14) by dividing both sides by p. To obtain the total wage bill and to derive income shares it is necessary to solve for total output, that is, the total amount of direct labour employed. By combining equations (5) to (9), and making use of (14), we obtain
L=

Lo + (1 +

,r)

a(l-X)

(15)

A more revealing form of this expression for the amount of direct labour employed is obtained by substituting, from equation (5 ), p/w for ( 1 + ,t) /a. It becomes L = 1FLo + p( + A)] w(1 - x) /1 (16)

24 This sole dependenceof profits on the expenditureof capitalists disappears if workers'saving is significant.In that case total profits would be equal to the sum of investment, capitalists'consumption expenditure,and workers' consumption expenditureout of their share of profits less workers'saving out of their wage incomes. Capitalists would no longer 'earn what they spend' because workers do not 'spend what they earn.' The introductionof workers' saving would not disturbthe main features of a Kaleckian theory of distribution. This point is illustratedin the appendix. 25 In his paper, first published in 1933, he wrote: 'Thus capitalists, as a whole, determine their own profits by the extent of their investment and personal consumption.In a way they are "mastersof their fate"; but how they "master"it is determinedby objective factors, so that fluctuationsof profits appear after all to be unavoidable. Capitalists' consumption is a function of the gross accumulation.The gross accumulationwhich is equal to the production of investment goods is determinedby investmentorders which in turn were undertakenin a past period on the basis of the profitabilityin that period, i.e. on the basis of the gross accumulationand the volume of capital equipment in that period' (Kalecki, 1971, 13). The factors underlyinginvestment decisions, alluded to above, will not be examined here since their considerationwould turn this paper into a full-fledged investigationof the theory of economic growth.

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

Income distribution / 327 The amount of direct labour employed is thus seen to be equal to the product of the reciprocal of the mark-up and the sum of indirect labour employed in the plants utilized, and investment and capitalists' consumption expenditures expressed in wage units. The gross profit share in value of gross output can thus be written as ,u(I + A) (17) )LO +(1 + I)(I +A)' The profit share in total output is a function of the mark-up even though the level of profits in real terms is not. This level is determined solely by capitalists' expenditure in real terms. The degree of utilization of capacity, as reflected in the amount of direct labour employed, is also a function of the mark-up. A higher mark-up, other things given, would lower the degree of plant utilization and increase the profit share. Higher capitalists' expenditure, whether on investment or consumption, because of the presence of overhead labour not only increases profits but also increases the profit share in total output.26This share is not determined solely by the mark-up; it varies with short-term fluctuations in the level of output. Its increase is consistent with a constant real-wage rate. This Kaleckian theory of distribution combines two of Kalecki's important contributions: his recognition of the role of the degree of monopoly in the setting of mark-ups, and his demonstration of the role of capitalists' expenditures in determining profits and the level of employment. Concentration on the degree of monopoly aspect of Kalecki's theory of distribution (e.g. as in Kaldor, 1955-6) does not do full justice to his writings in this area. The reintroduction of overhead labour into Kalecki's model means that labour productivity, output per unit of total labour employed, would vary directly with demand. Consequently an increased demand for labour due to higher capitalist expenditure leads to a higher profit share as well as to higher output, even though the mark-up is constant.27 The a(1 26 This is the one conclusion from this model that might have to be modified if separate investment-and consumption-goodsproducing sectors are introduced. Higher values for investment or capitalists'consumption expendituresmight alter the ratio of investmentto consumptionexpendituresin the model, thus changing the weights to be used in arrivingat the over-all ratio of profit to output. If the mark-upsdiffer in the two sectors, it is possible that the over-all ratio of profit to output will fall, even if this ratio increases in each sector. 27 This Kaleckian theory of distributionbelongs, of course, to the genus, 'Cambridge'theory. As Kaldor (1970, 5) has phrased it: 'Accordingto the "Cambridge" theory, the labour market does not behave in accordancewith the postulates of neo-classical theory: with a rise in the demand for labour, there is a rise in the share of profit, and a fall in the share of wages, and not the other way round. (This of course is not inconsistentwith a rise in absolute wages - in wages per man - if output-per-headalso rises with rising employment.) This is because a high d'emandfor labour is associated with a high rate of investment and a high rate of profit on capital. Empirically,the evidence here

II Y

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

328 / A. Asimakopulos model predicts that real-wage rates are more stable than output per worker in the course of short-term fluctuations in output.28 Unlike Kaldor's version of the 'Cambridge' theory of distribution this one is not tied to the assumption of full employment of labour; higher investment brings about a higher profit share even though the mark-up is constant and employment is also increased.29 Kaldor's results can be readily obtained from this model by assuming that labour is fully employed and then deducing the effects on the mark-up and on income shares of a higher level of investment. With both labour and productive capacity available to increase output the employment multiplier as derived from equa- X) - the corresponding income tion (15) is equal to (1 + jt)/jta(l multiplier is equal to (1 + pt) /4(1 - X). There is no conflict in the two 'uses' of investment in this model. It determines, given the propensities to save, both the level of employment and the distribution of income.30
CLASS OF STRUGGLE AND INCOME THE DISTRIBUTION

NATIONAL

The 'Kaleckian' nature of the model developed here3' can be illustrated by deriving the results Kalecki obtained in his posthumously published article, 'Class struggle and distribution of national income.' Kalecki argued that, in a closed system, if 'wage rates in all industries
supports Cambridge,and not the implications of neo-classical theory: the share of profit and the level of employment are positively correlated,not negatively' (Kaldor, 1970; 5, emphasis in original). This consequence of our model is in agreementwith empirical observations of the behaviour of real-wage rates and labour productivityover the cycle. See Kuh (1960) and Neild (1963). This Kaleckian model is thus, in Rothschild'sphrase, 'fully in,the Keynesian spirit' (Rothschild, 1971, 25). Kaldor's (1955-6) results can be readily obtained from this model by assumingthat labour is fully employed and deducing the effects on income shares of a higher level of investment. The mark-up would in this case be positively related to the level of investment in terms of wage units, as shown in note 35 below. Lydall (1971), in criticizing Kaldor's theory of distribution,has argued that 'if the Keynesian theory of employment is to retain its validity we must assume y [the share of profits in national income] constant (or nearly so)' (92). This argumentis refuted by our Kaleckian-Keynesianmodel. It provides an explanation for the level of employment without requiringthe profit share to be constant. Keynesian theory does not require the over-all propensityto save in the economy to be constant. It may vary with changes in the distributionof income even though the propensitiesto save of the differentclasses in the economy are constant. A similar model, which included governmentexpenditure and taxation, -was used to investigate the short-periodincidence of taxation in Asimakopulos and Burbidge (1974). They duplicate and extend the results obtained by Kalecki in his 1937 article, 'A theory of commodity, income and capital taxation' (Kalecki, 1971, 35-42).

28 29

30

31

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

Income distribution / 329 increase in the same proportion, 1 + a times ... all prices will also increase 1 + a times provided that functions f in industries to which they are relevant are unchanged ... It follows that if these conditions were fulfilled we should arrive at the ... conclusion ... that a general increase in money wages in a closed economy does not change the distribution of national income' (Kalecki, 1971, 161, emphasis in the original). (The 'functions f' are the mark-ups.) Investment and rentiers' expenditures are assumed to be unaffected in real terms as a result of the higher money-wage rates. His conclusions can be easily deduced from equation (17) above. The profit share in total income is constant, no matter what the value of w, as long as jt, X, and (I + A) have unchanged values.32 Kalecki's analysis does not end with this mechanical result. He believed that in certain circumstances mark-ups depend on trade-union activity. This occurs because wage rates are not raised simultaneously in all industries and bargaining tends to proceed industry by industry. 'High mark-ups in existence will encourage strong trade unions to bargain for higher wages since they know that firms can "afford" to pay them' (161). Kalecki believed that competitive pressures in their markets from those in other industries will lead them to lower mark-ups and thus raise prices by less than the increases in unit prime costs.33 Kalecki distinguished between two types of wage increases. 'Normal' wage increases that usually leave the mark-ups unchanged in practice prevent the increases that would otherwise occur 'because of the rise in productivity of labour' (162). Spectacular wage rises that are due to increases in bargaining capacity are assumed to depress the mark-ups somewhat, so that, as can be seen from equation (17), redistribution of national income from profits to wages occurs. In the process, with the volume of investment and capitalists' consumption expenditures unchanged in real terms. total employment and output will be higher, as can be easily deduced from equation ( 15). Although Kalecki's model in his 'Class struggle' paper distinguishes between three Departments of production - Department i producing investment goods, ii producing consumption goods for capitalists, and iII wage goods - all his results can be obtained with our
32 If a higher money-wage rate leads to a fall in the value of capitalists' expenditurein real terms, e.g. because a portion of their incomes is fixed in money-terms and they cut their consumption expenditure in real terms in the face of higher prices, then the higher wage rates would depress the profit share. This increase in their share of total gross income is not all gain for workers, however, since the level of employment, as indicated by equation (15), would also be lower in such a case. 33 Profits in real terms would also be adversely affected during periods of rising wages and prices if firms based mark-upsand prices on historical costs. Nordhaus (1974), who uses mark-uppricing in his investigationof profit shares, argues that 'most businessmen ... base their actual calculations of prices, sales, and profits on historical cost' (187).

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

330 / A. Asimakopulos one-sector model. We can conclude, as does Kalecki, 'It follows from the above that a wage rise showing an increase in the trade union power leads - contrary to the precepts of classical economics - to an increase in employment. And conversely, a fall in wages showing a weakening in their bargaining power leads to a decline in employment. The weakness of trade unions in a depression manifested in permitting wage cuts contributes to deepening of unemployment rather than to relieving it' (163). It is important to note, with Kalecki, that the 'redistribution of income from profits to wages is feasible only if excess capacity is in existence. Otherwise, it is impossible to increase wages in relation to prices of wage goods because prices are determined by demand' (164). When it is no longer possible to increase output in the short period because of the absence of unemployed labour or unused productive capacity, the class struggle would be reflected, not in the mark-ups, since they would largely be determined by demand in the face of inelastic supplies,34 but in the shares of total output accounted for by capitalists' expenditure. A stronger bargaining position of workers would be reflected in a diminution of capitalists' expenditure in real terms and thus in a higher share for workers in total income. The key elements in the struggle over income shares in a Kaleckian theory of distribution, given the propensities to save, are thus the mark-ups and capitalists' expenditures in real terms. Trade unions will succeed in increasing workers' share in total output if, in successfully bargaining for higher money-wage rates, they can lower mark-ups and/or capitalists' expenditures. This may present trade unions with a dilemma in so far as investment expenditures would be cut if money-wage rates were sharply increased, say, as a result of credit restrictions imposed to try to curb the resulting inflationary pressures, even if employment is maintained. The improvement in the workers' share and level of income in the present may be at the expense of future income that will be affected by the lower rate of current investment.
34 If situations in which maximum possible output is being produced are compared, equation (15) should be reversed to show the mark-up u as a function of effective demand. Let LlL denote the maximum amount of employment of direct labour possible, either because of full employment of labour or because of available productivecapacity, we have + (I + A) a(l -X)L . and a(l A)Ljm - (I + A)ad (I + A)
a(l
-

II_ Y

A)LIm

The mark-up is no longer exogenous and does not appear in the equation for the profit share. This share varies directly with the value of capitalists' expenditurein real terms and inversely with the maximum employment of direct labour.

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

Income distribution / 331

A positive propensity to save out of wages Some of the assumptions made in setting up the model in the text were introduced for the sake of simplicity in presentation. They can be relaxed without affecting the general nature of the conclusions. With respect to saving the critical assumption is not the absence of workers' saving but the difference in the propensities to save' out of the two categories of income, profits and wages. This can be illustrated by considering a situation in which the propensities to save out of wages and out of distributed profits (rentiers' incomes) are the same. The propensity to save out of profits would still be greater than the propensity to save out of wages, because of the retention by firms of part of profits. Let s be the common propensity to save of individuals and /8 the proportion of profits distributed. The equation for total (desired) gross saving is now S = 1I(1
-

3) + sPfT +

SW,

(la)

and this replaces equation (11). All the other equations describing the model are unchanged. By substituting the equations S = I and I = pI, in the above equation for S, and rearranging,we have
I =

p? + (1-

s) 'an1-SW.

(2a)

The level of profits no longer depends solely on capitalists' expenditure because workers no longer spend what they earn. It is equal to capitalists' expenditure plus workers' consumption out of their share of distributed profits minus workers' saving out of wage incomes, as can be seen from equation (2a). In order to express the level of profits as a function of the exogenous variables it is necessary to obtain from the full model an expression for L, in terms of the exogenous variables. Solving for Lt from equations (5) to (9) and (2a), we obtain - (1 + L Ll

=~~

+ s-s) )I + aLO(1+ sp) + s] a[t(l -

(3a) (a

When this expression is substituted for L in equation (2a) we can obtain (since W = wLo ? wL1)

+ j)w(pJ([(1

saL+)

(4a)

Finally, in order to obtain the profit share in total income, expression (4a) is divided by the equation for Y(= (1 + IA)wL1) with equation (3a) substituted for L1. It is equal to H Y I-saL0 (1 + /i)I + aLO(1- 3+sf -sY(5) (5a)

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

332 / A. Asimakopulos It can be seen from equation (4a) that, as in the main model, the level of profits in the short period is positively related to the level of investment. By differentiating this equation with respect to s and /3 it can be readily deduced that the level of profits is inversely related to the propensity to save of individuals and directly related to the proportion of profits distributed. With workers' saving, however, the level of profits as well as the profit share is affected by the value for the mark-up. A higher value for investment in terms of wage units increases the profit share, even with the mark-up constant as in the basic model, because of the presence of overhead labour. Saving not in the desired relation to income All the results derived in this paper have been based on a necessary assumption for short-period equilibrium, that saving is in the desired relation to income. When this condition is not satisfied, the equation for actual saving, for the saving propensities used in this appendix, would then be written as S= (1-la) + sIH + sW + S*, (lb)

0. where S* #& The equation for profits then becomes H [1-p8 + s,f] = pS*
-

sW.

(2b)

All the other expressions can be readily derived. Disequilibrium saving St is seen to modify the effects of the planned level of investment in these expressions. For example, if S* is positive the employment in the short period examined resulting from any given values for I and the propensities to save is lower, and both profits and profits share are lower. By specifying the factors determining S*, for example the time lag between changes in income and changes in consumption, S* can be 'explained.' Introduction of S* into the analysis does not change the nature of the conclusions.
REFERENCES

Asimakopulos, A. (1969) 'A Robinsonian growth model in one-sector notation.' Australian Economic Papers 8, 41-58 Asimakopulos, A. (1970) 'A Robinsonian growth model in one-sector notation - an amendment.' Australian Economic Papers 9, 171-6 Asimakopulos, A. (1971) 'The determination of investment in Keynes's Model.'

This JOURNAL 4, 382-8

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions

Income distribution / 3 3 3 Davidson, Paul (1960) Theories of AggregateIncome Distribution (New Brunswick,New Jersey: RutgersUniversityPress) Eichner, Alfred S. (1973) 'A theory of the determination of the mark-up under oligopoly.'Economic Journal83, 1184-1200 Hall, R.L. and C.J. Hitch (1939) 'Price theory and business behaviour.' OxfordEconomic Papers 2, 12-45 Kaldor,Nicholas (1955-6) 'Alternativetheoriesof distribution.' Review of Economic Studies 23, 83-100 Kaldor, Nicholas and J.A. Mirrlees(1962) 'A new model of economic growth.' Review of Economic Studies 29, 174-90 Kaldor, Nicholas (1970) 'Some fallacies in the interpretation of Kaldor.' Review of Economic Studies 37, 1-7 Kalecki, Michat (1971) Selected Essays on the Dynamics of the Capitalist Economy (Cambridge:CambridgeUniversityPress) Kaplan,A.D.H., J.B. Dirlam, and R.F. Lanzillotti(1958) Pricing in Big Business- A Case Approach (Washington:BrookingsInstitution) Kuh, E. (1960) Profits,Mark-Upand Productivity(Washington:Study Paper No. 15, Joint Economic CommitteeStudy of Employment,Growth and Price Levels) Lanzillotti,R.F. (1958) 'Pricingobjectivesin large companies.'American Economic Review 48, 921-40 Lydall, H. (1971) 'A theory of distributionand growthwith economies of scale.' Economic Journal 81, 91-112 Mann, H.M. (1966) 'Sellerconcentration,barriersto entry and rates of return in thirty industries,1956-60.' Review of Economicsand Statistics 48, 296-307 Neild, R. (1963) Pricing and Employmentin the TradeCycle (London: CambridgeUniversity Press) Nordhaus,William D. and Wynne Godley (1972) 'Pricingin the trade cycle.' Economic Journal82, 853-82 Nordhaus,William D. ( 1974) 'The falling share of profits.'BrookingsPapers on Economic Activity, 1974 169-208 Nuti, D.M. (1970) "'Vulgar economy"in the theory of income distribution.' Reprintedin E.K. Hunt and Jesse G. Schwartz,eds, A Critiqueof Economic Theory (Hammondsworth,Middlesex: Penguin, 1972) Riach, P.A. (1971) 'Kalecki's"degreeof monopoly"reconsidered.' AustralianEconomic Papers 10, 50-60 Robinson, Joan (1962) Essays in the Theory of Economic Growth (London. Macmillan) Robinson,Joan (1965) 'Kaleckiand Keynes.'Collected Economic Papers 3 (Oxford: Blackwell) Robinson,Joan (1969) 'A furthernote.' Review of Economic Studies 36, 260-2 Rothschild,Kurt W. ( 1971) 'Differentapproachesin distribution theory.' Kyklos 24, 10-29 Shackle, G.L.S. (1967) Years of High Theory (Cambridge:Cambridge University Press)

This content downloaded from 147.96.1.236 on Mon, 17 Jun 2013 14:14:07 PM All use subject to JSTOR Terms and Conditions